Steel Tips Committee of California Parte 3

STRUCTURAL STEELEDUCATIONALCOUNCIL
TECHNICAL INFORMATION& PRODUCTSERVICE

October 1993
l',leavy Structural Shapes
in Tension Applications
INTRODUCTION
On January 1, 1989 the American Insti-
tute of Steel Constructi on issued a
supplement to both the 8th edition ASD
Manual and the 1st edition LRFD Manual.
The supplements covered procedures for
welding Groups 4 & 5 structural shapes and
sections built up from plates over 2" in
thickness, subject to primary tension
stresses.
These supplements have been part of
the 9th edition 1989 ASD Specifications and
apply equally to structures designed by the
LRFD procedure. Recently, in many cases,
the AISC provisions have been applied to
sections other than those that are in Group
4 & 5, and to members other than those that
are subject to primary tensile stresses, and
spliced using full penetration welds.
The application of the AISC provisions
should be carefully assessed, since the
requirements can lead to increased costs of
material, fabrication, and inspection. Misap-
plying or altering the provisions can increase
the cost of construction without necessarily
increasing the reliability of the structure.
This TIPS is based on a Bethlehem
Steel Corporation Technical Bulletin dated
June, 1992 of the same title and the AISC
Specifications. It is hoped that this will help
clear up any misunderstanding that occur
with design engineers regarding the use of
heavy structural shapes in tension and
spliced with full penetration welds.
BACKGROUND
In response to problems encountered in
some situations with the use of heavy W-
shapes and thick plates in non-column
applications, the Al SC Committee on Speci-
fications has developed new provisions
covering material properties, splicing de-
tails, thermal cutting and welding of such
shapes (References 1,2 & 3).
The new provisions are applicable to
ASTM A6/A6M Groups 4 and 5 rolled shapes
and to welded shapes built up from plates
with thickness exceeding 2 in., if such rolled
or welded shapes are subject to primary
tensile stresses due to tension or flexure,
and if they are spliced using full-penetration
welds. The provisions include material tough-
ness requirements for beam copes and weld
access holes to facilitate welding operation
and to minimize restraint, and fabrication
requirements which include preheat for ther-
mal cutting and for making groove-weld
splices.
Solidified, but still hot, weld metal con-
tracts significantly as it cools to ambient
temperature. Shrinkage of large welds be-
tween elements, which are not free to move
to accommodate the shrinkage, causes
strains in the material adjacent to the weld
that can exceed the yield point strain. In
thick material, the weld shrinkage is re-
strained in the thickness direction, as well
as in the width and length directions, caus-
ing triaxial stresses to develop that may
inhibit the ability of ductile steel to deform in
a ductile manner. Under these conditions,
the possibility of brittle fracture increases.
When splicing ASTM Groups 4 and 5
rolled sections or heavy welded built-up
members, the potentially harmful weld
shrinkage strains can be avoided by use of
bolted splices, fillet welded lap splices or a
splice using a combination welded and bolted
detail. However, details and techniques that
perform well for materials of modest thick-
ness usual l y must be changed or
supplemented by more demanding require-
ments when welding thick material. Aisc,
the provisions of Structural Welding Code
AWS D1.1 are minimum requirements that
apply to most structural welding situations.
When designing and fabricating welded
splices of Groups 4 and 5 shapes and simi-
lar bui l t -up cross-sect i ons, speci al
consideration must be given to all aspects of
the welded splice detail. These are as fol-
lows:
Notch-toughness requirements
Steel Tips - Page 2
shall be specified for tension mem-
bers. AISC Specifications require
that tension members of Group 4 &
5 structural shapes to be supplied
with Charpy V-Notch testing and
that the impact test shall meet a
minimum average value of 20 ft.
pounds absorbed energy at +70F.
AISC Specification J2.7 (ASD) re-
quires pre-heating of Group 4 & 5
shapes and built-up members made
of plates more than 2" thick to 350F
or greater when making groove
weld splices.
Generously sized weld access
holes are required to provide relief
from concentrated weld shrinkage
strains by avoiding the close junc-
t ure of welds in ort hogonal
directions and to provide adequate
clearance for the exercise of high
quality workmanship in hole pen-
et rat i on, wel di ng & ease of
inspection. AISC Specifications
state that weld access holes shall
have a length from the toe of the
weld preparation not less than
1 1/2 times the thickness of the
material in which the hole is made.
The height of the weld access hole
shall be adequate for the deposi-
tion of sound weld metal and provide
clearance for weld tabs.
Thermally cut surfaces of beam
copes and weld access holes in
Groups 4 & 5 shapes and built up
shapes over 2" thick shall be ground
to bright metal and inspected by
either magnetic particle or dye pen-
etrant methods.
Pre-heating for thermal cutting is
required to minimize the formation
of a hard surface layer. AISC Speci-
fications require a pre-heat of 150F
for thermally cut edges of Group 4
& 5 shapes and built up shapes
with material thicker than 2".
In addition to tension splices of truss
chord members and tension flanges of flex-
ural members, other joints fabricated of
heavy sections subject to tension should be
given special consideration during design
and fabrication.
It is important to note that the provisions
of notch toughness, grinding of thermally
cut edges, etc. contained herein apply only
to Group 4 & 5 structural shapes and built up
sections with materialthickness 2"or greater.
Further, these provisions apply only to those
heavy members subject to primary tensile
stresses due to direct tension or flexure and
spliced with full penetration welds. Examples
are tension chords of trusses and beams in
a ductile moment resisting space frame.
Columns in which the dead Icad compres-
sion stresses exceed the tensile stresses
due to lateral Icad flexure are normally not
full penetration welded, do not go into net
tension and are not governed by these pro-
visions (P/A _>Mc/I).
The September 1993 AISC Specifica-
tion for LRFD construction adds some
clarification to the welding requirements for
Group 4 & 5 shapes with this sentence from
Section A3.1c: The requirements need not
apply to ASTM A6 Group 4 & 5 shapes and
built-up members with thickness exceeding
two inches to which members other than
ASTM A6 Group 4 & 5 shapes are con-
nected by complete-joint penetration welded
joints through the thickness of the thinner
material to the face of the heavy material.
Additional requirements for joints in heavy
rolled and built-up members are given in
Sections J1.5, J1.6, J2.6, J2.7 and M2.2.
This applies equally to structures designed
by the ASD specifications.
AISC PROVISIONS:
STRUCTURALSHAPES
For ASTM A6 Group 4 & 5 rolled
shapes to be used as members
subject to primary tensile stresses
due to tension or flexure, tough-
ness need not be specified ffsplices
are made by bolting.
If Group 4 & 5 rolled shapes are
spliced using full penetration welds,
the steel shall be specified in the
contract documents to be supplied
with Charpy V-Notch testing in ac-
cordance wi th ASTM A6,
Supplementary Requirement S5.
The impact test shall meet a mini-
mum average value of 20 ft. lbs.
absorbed energy at + 70F and
shall be conducted in accordance
with ASTM A673 with the following
exceptions:
a. The center longitudinal axis of the
b.
specimens shall be located as near
as practical to midway between the
inner flange surface and the center
of the flange thickness at the inter-
section with the web mid-thickness
(see sketch below).
Tests shall be conducted by the

producer on material selected from
a location representing the top of
each ingot or part of an ingot used
to produce the product represented
by these tests.
Steel Tips - Page 3
PLATES
For plates exceeding 2" thick used for
built-up members with bolted splices
and subject to primary tensile stresses
due to tension or flexure, material
toughness need not be specified.
If such members are spliced using full-
penetration welds, the steel shall be
specified in the contract documents to
be supplied with Charpy V-Notch test-
ing in accordance with ASTM A6,
Supplementary Requirement S5. The
impact test shall be conducted by the
producer in accordance with ASTM
A673, Frequency P, and shall meet a
minimum average value of 20ft.-lb.
absorbed energy at +70F.
GENERAL
The above supplementary tough-
ness requirements shall also be
considered for welded full-penetra-
tion joints otherthen splices in heavy
rolled and built-up member subject
to primary tensile stresses.
The above supplementary tough-
ness requirements need to be used
in conjunction with good design
and fabrication procedures. Addi-
tional AISC requirements include:
a. Splices in heavy sections
b. Beam copes and weld access holes
c. Mixed weld metal
d. Preheat for heavy shapes
e. Thermal cutting
The reader is referred to the following
sections of the Al SC Specification contained
in the 9th Manual ASD Sects. A3.1c, J1.7,
J1.8, J2.6, J2.7 & M2.2 for further informa-
tion. Aisc referenced is the September, 1993
Load and Resistance Factor Design Speci-
fication sections, A3.1c, J1.5, J1.6, J2.6,
J2.7, & M2.2. The specifications for welding
Group 4 & 5 shapes and built-up sections
over 2" thickness apply equally to member
designed by the ASD or LRFD methods.
GROUP 4 & 5 SHAPES
ASTM A-6 Section 3.10 defines hot rolled
shapes into size groups from 1 to 5 based on
web thickness. These groupings are shown
in Table 2 of the AISC Manual of Steel
Construction, both ASD & LRFD. It is noted
again that the special welding procedure
requirements contained in this TIPS apply
only to Group 4 & 5 shapes and welded
shapes built up from plates over 2" thick-
ness, subject to primary tension stresses
and spliced with full penetration welds. A
listing of Group 4 & 5 shapes is shown
below.
Steel Tips - Page 4
GROUP 4 GROUP 5
W40x362 to 655 incl
W36x328 to 798 incl
W33x318 to 619 incl
W30x292 to 581 incl
W27x281 to 539 incl
W24x250 to 492 incl
W21x248 to 402 incl
W18x211 to 311 incl
W14x233 to 550 incl
W12x210 to 336 incl
W36x848
W14x605 to 730 incl
SUMMARY
The AISC Specifications, both for ASD
and LRFD speak to state-of-the-art knowl-
edge of the practice of welding Group 4 & 5
structural shapes and built-up sections hav-
ing material thickness 2" or greater subject
to direct tension or tension due to flexure
and spliced with full penetration welds.
Notch-toughness requirements, size of
weld access holes and specification for pre-
heating and treatment of thermally cut holes
are covered in the specifications.
The AISC Specifications provide only
part of what is needed to minimize the
potential for failure. They must be used in
conjunction with good design and fabrica-
tion procedures.
1) Bethlehem Steel Corp. Technical Bulletin.
"Use of Heavy Structural Shapes in tension
Applications," dated June 1992.
2) AISC "Specification for Structural Steel
Buildings," June 1989.
3) AISC "LRFD Specification for Structural
Steel Buildings," September 10, 1993.
DISCLAIMER
The reader is cautioned that professional
judgment must be exercised when data or
recommendation presented herein are ap-
plied. The publication of the material
contained herein is not intended as a repre-
sentation or warranty on the part of the
Structural Steel Educational Council or any
other person named herein, that this infor-
mation is suitable for general or particular
use, or freedom from infringement of any
patent or patents. Anyone making use of
this information assumes all liability arising
from such use. The design and detailing of
steel structures is within the expertise of
professional individuals who are competent
by virtue of education, training, and experi-
ence for the application of engineering
principles and the provisions of this specifi-
cation to the design and/or detailing of a
particular structure.
Steel Tips - Page 5
STEEL-EDUCATIONAL COUNCIL
TECHNICN. INFORMATiON & PRODUCTSERVICE
SEPTEMBER, 1991
Design Practice to
Prevent Floor Vibrations
by
Farzad Naeim
Design Practice to
Prevent Floor Vibrations
About the Author:
Farzad Naeim, Ph.D., S.E., is Director of Research and Development for the Structural
Engineering firm ofJohn A. Martin and Associates, Inc., Los Angeles, California. He has been
in charge of design review and analysis of numerous complex projects across the United States
for Vibration as well as other serviceability concerns. Dr. Naeim has regularly lectured on
various aspects of structural design and earthquake engineering at University of Southern
California and California State University, Northridge. He is an active member of several
professional organizations and has more than 30 publications covering a wide spectrum of
structural and earthquake engineering applications.
INTRODUCTION
The current trend towards longer spans and lighter floor systems, combined with
reduced damping and new activities, such as aerobics exercises, have resulted in a
significant increase in the number of floor vibration complaints by building owners and
occupants. This has increased the degree of attention paid during the design process, to
preventing, or reducing floor vibration problems.
The purpose of this publication is to provide design engineers with a practical yet
comprehensive review of the criteria and methods available to prevent floor vibration
problems.
Because of the complexities involved in human response to vibration and the different
objectives persued by various investigators, the predictions of the methods presented
here are not always consistent. Unfortunately, a general consensus on the relative
accuracy and reliability of these methods does not yet exist. However, it is hoped that
collective review, application, and comparison of these methods will help to form this
seriously needed consensus in the near future.
Annoying floor vibrations may be caused by occupant activities. Walking, dancing,
jumping, aerobics, and audience participation at music concerts and sporting events are
some prime examples of occupant activities which create floor vibrations.
Operation of mechanical equipment is another cause for concern. Heating, ventilation,
and air-conditioning systems (HVAC) as well as washing and drying machines, if not
properly isolated, can cause serious vibration problems.
Most of the sources contributing to reported human discomfort rest on the floor system
itself. However, human activities or machinery off a floor can cause significant floor
vibrations. On more than one occasion, aerobics on one floor of a high-rise building has
been reported to cause vibration discomfort at another level in the building. The
vibrations caused by automobiles on parking levels below have been reported to disrupt
sensitive laboratory work on upper floors. Other equipment and activities off the floor
that can contribute to a floor vibration problem are ground or air traffic, drilling, impact
of falling objects, and other construction related events.
FACTORS INFLUENCING VIBRATION PERCEPTIBILITY
Several factors influence the level of perception and the degree of sensitivity of people to
vibrations. Among them are:
(a) .Position of the human body. Consider the human body coordinate system defined
in Figure 1. Here, the x-axis defines the back-to-chest direction, the y-axis defines
the right side to left side direction, and the z-axis defines the foot -(or-buttocks-
)to-head direction. According to ISO9,o, the frequency range of maximum
sensitivity to acceleration for humans is between 4 to 8 Hz for vibration along the
z-axis and 0 to 2 Hz for vibration along the x- or y- axes. While z-axis vibration is
most important in the design of offices and other workplaces, all three axes
become important in the design of residences and hotels where sleeping comfort
should be considered.
Suppong
surface
Z
X
Y
x
surface
X, Qvo Q=
x-axiS
y-axis
z-axis
X
Z
Sup0orting
surface Y
-- acceleration in the directions of the x-, .v-, z-exes
=
= right side to left side
= foot-(or buttocks-ito-head
Figure 1. Directions of basicentric coordinate systems for vibrations influencing
humans. TM
2
(bi
(c)
(d)
(e)
(f)
Excitation source characteristics such as amplitude, frequency content and
duration.
Exoosure time. As shown in Figures 2 and 3, human tolerance of vibration
decreases in a characteristic way with increasing exposure time9.
Floor system characteristi such as natural frequency (stiffness; mass), and
damping.
Level of exoectancv. The more one expects vibration and knows about its source
the less startling the vibration becomes. Because people expect more vibration in
workshops than in hotel lobbies, they can put up with more in the former than in
the latter. Anxiety and discomfort can be reduced if occupants are made aware of
the nature of vibrations and are assured that they are not a threat to their safety
and well being.
Tvoe of act/v/tv engaged in. The level of perception varies with the nature of
activity that one is engaged in such as office work, dinning, walking, or dancing.
CATEGORIES OF HUMAN RESPONSE
ISO9 classifies human response to vibrations into three categories:
(a) limit beyond which the comfort is reduced ("reduced comfort boundary")
(b) limit beyond which the working efficiency is impaired (" fatigue-decreased
proficiency boundary")
(c) limit beyond which the health or safety is endangered ("exposure limit")
These categories were derived from various studies conducted for transportation
industries and generally reflect a much higher level of tolerance than what would be
acceptable in a building environment. According to ISO 2631-2o:
"Experience has shown in many countries that complaints regarding building
vibrations in residential situations are likely to arise from occupants of buildings
when the vibration magnitudes are only slightly in excess of perception levels. In
general, the satisfactory magnitudes are related to the minimum adverse comment
level by the occupants and are not determined by any other factors, such as short-
term health hazard and working efficiency. Indeed, in practically all cases the
magnitudes are such that there is no possibility of fatigue or other vibration-
induced symptoms."
Murray's 3 categorization of human response is more design oriented and hence more
useful. He defines four response categories, among which the first t wo are acceptable as
far as design is concerned:
(a) Vibration, though present, is not perceived by the occupants.
(b) Vibration is perceived but does not annoy.
(c) Vibration annoys and disturbs.
(d) Vibration is so severe that makes occupants ill.
3
20
16
12.5
10
8.0
6.3
0
kO
!
4[
Fi gure 2.
3.15
2.5
2.o
1.6
1.25
1.0
0.8
0.63
0.50
0.40
0.315
0.25
0.20
0.16
0.
0.1O
Fi gure 3.
f /
/ /
1.6 /,' / , .
/ - / /
-g- """"- - - ' 7 : '-5 '-- I -- -
0.6J . / f / / / ' .,
oJ, / " / / j ' j / .
. . ." / / / / / / . , -j
,- / / / . /
oo3 16 " / / / -/ i
O,Oi , '-
/ -
8h / / ,0o,..,
0,0 5 I i ' / ....................
f lilCLeeteltaO(I values by 2 16 dB htjhl
0,016 J - - ....-, .......... u,
I % I I I J I I
0.016 0./, 0.5 0.63 0,6 1.0A.25 1.6 2.0 2.5 3.15 /,.0 5.0 6.3 B.0 10 12.5 16 20 25 31.5 (.0 50 63 80
f fe(x o. Cenlte beQuy o! oe Ihledoctave t)4rKI HI
Longi t udi nal (az} accel er at i on l i mi t s as a f unct i on of frequency and exposure
time ( f at i gue- decr eased pr of i ci ency boundar y) .
. . . . . _ . _. . . . I ' I r
""- ! I i : I
1 xgn ' '. ,, ,,k
. . . . . i
' !
' - , ,
accMermen vM by 2 (6 dB h,gh).
- ' .,,,cea co,on bOuKm" e.,ee ' . -
I
o
'0.01
s lo b i s-, b 00,? i '
25
16
10
6.3
~/,.0
E
; 2.S
E
-:-1.6
)
i
l.0
0.63
0,1.
O.315
0.25
0,16
0,1
0.063
0.0
0.25 0.5 1 1.5 2 2.5 6 8 10 16 2,
ama h
Longi t udi nal (az) accel er at i on l i mi ts as a f unct i on of exposure time and
frequency ( f at i gue- decr eased pr of i ci ency boundar y) .
4
ISO INTERNATIONAL STANDARD 2631 PROVISIONS
ISO 2631-2 TM provides a number of human perceptibility base curves for floor velocity
and acceleration. According to ISO, at vibration magnitudes below the base curves,
adverse comments, sensations, or complaints are very rare. They note however, that
this does not mean that the values above the base curves will give rise to adverse
comments or dissatisfaction. Since the magnitude which is considered to be satisfactory
depends on the circumstances, ISO suggests specifying satisfactory vibration levels in
terms of multiples of these base curves. Base curves for foot-to-head, back-to-chest,
and side-to-side accelerations are shown in Figures 4 and 5.
In terms of human response, ISO divides vibrations into t wo classes: (a)transient (also
called impulsive) and (b)continuous or intermittent. Transient vibration is defined as a
rapid build-up to a peak followed by a damped decay, such as vibration caused by the
impact of a single heavy object on a floor system. It can also consist of several cycles of
vibration at approximately the same amplitude, providing that the duration is short ( less
than about 2 seconds).
Continuous vibration on the other hand is vibration which remains uninterrupted over the
time period under consideration. Intermittent vibration is defined as a string of vibration
incidents, each of short duration, separated by intervals of much lower vibration
magnitude (for example vibration caused by a group of people walking or elevators
operating).
In an appendix to ISO 2631-2, a set of state of the art multiplication factors frequently
used with the ISO base curves are presented. These factors which lead to magnitudes
of vibration below which the probability of reaction is Iow are summarized in Table 1.
In many situations the same building space, residences and hotel guest rooms, for
example, may be used in both standing and lying positions. For these cases, ISO 2631-2
suggests using a combined standard that represents the worst case combination of z-
axis and x/y axes conditions. The combined standard curves for acceleration response
are presented in Figure 6. Notice that the multiplication factors in Table I have already
been applied to these curves.
COMPUTING FLOOR SYSTEM CHARACTERISTICS
Unless otherwise noted the following assumptions are used in this publication for
calculating floor system vibration characteristics:
(1) Full composite action is assumed to exist between the concrete slab and steel
beam regardless of the number of shear studs present 2.
(2) The beam is modeled as a single degree of freedom (SDOF) system.
(3) The transformed moment of inertia (It) is calculated using Murray's
assumptions 2, 3.14.
As pointed out by Allen3.4.s, it is better to calculate the first natural frequency, f,
on deflection:
[1] f _ I /stiffness I .x/--
27t' m---sss - 27t
based
0.1
0.063
0,0/.,
0.025
0,016
%
0.01
E
g 0,006 3
o.oo .
<
0,002 5
[ i I
I
/
': Z . . . . .
t
I
I
,, f i
i
0,0016
0,001
Figure 4.
1.6 2.5 I, 6.3 10 16 25 .0 63 100
C e n t r e f r e q u e n c y o f o n e - t h i r do c t a v e b a n d s . H z
Building v i b r a t i o n z-axis base c u r v e for acceleration TM
%
.6
g
' t -
Il
. <
Figure 5.
0.16 -
0.1
0,063
0, 0/ .
0,025
0,016
0.01
0.0063
0.00
0,002 5;
O,O01 6 I
0.001
1
i
: I / '
1.6 2,5 Z; 6.3 10 16 25
C e n t r e f r e q u e n c y o f o n e - t h i r d o c t a v e b a n d s , H z
i
: i
i
[
i ,. -
1
63 100
Building vibration x- and y - a x i s base c u r v e for acceleration TM
TABLE 1 --- Ranges of multiplying factors used in several countries to specify satisfactory
magnitudes of building vibrations with respect to human response{1)
(from ISO 2631-2: 1989)
Continuous or Inter- Transient Vibration
PLACE TIME mittent Vibration (Excitation with several
occurrences per day)
Critical Working Areas
(for example some
hospital operating- Day
theatres, some 1 1 (2,3)
precision laboratories) Night
Day 2 to 4 30 to 90 (4,5,6,7!
Residential
Night 1.4 (4! 1.4 to 20
Day
Office 4(a) 60 to 128 (s)
Night
Day
Workshop(9) Night 8(8.1o) 90 to 128 (8.1)
1) Table leads to magmtudas of vibration below whGch the probability of reaction s Iow. (Any acoustic noese
caused by vibrating walls is not considered.)
2) Also includes quasi-stationary vibrations caused by repetitive shocks. Shock is definecl in ISO 2041: 1975.
clause 3, and is sometimes referred to as transient (impulsive) vibration.
3) Magnitudes of transient vibration in hospital operating-theatres and critical working places pertamn to periods
of t,me when ooeratqona are in progress or critical work is being performed. At other times, magnitudes as hgh
IS thoGe for residence are satisfactory provided that there is due agreement and warning.
4) Within residential areas there are wide variations in vibration tolerance. Specific values are dependent upon
social and cultural factors, psychological attitudes and expected interference wi th privacy.
5) The "trade-off" between number of events per day and magnitudes is not well established. The following
prowsional relationship shall be used for cases of more than three events a day pending further research into
human vibration tolerance. This involves further multiplying by a number factor Fn = 1,7 N-0 5 where ./is the
number of events per day. This "trade-off" equation does not apply when values are lower than those given by
the factors for continuous vibration. When the range of event megni tud# is small (wnthin I haft amplitude of the
blrgest), the arithmetic mean can be used. OthenAtiee only the largest need be considered.
6) For discrete events wi th durations exceeding 1 s, the factors can be adjusted by further multiplying by a
duration factor, Fd:
Fd = T - 1.22for concrete fl oo and T is between 1 end 20
Fo ,, T-0. for wooden floors and Ti s between I end 60
where Ti s the duration of the event, in seconds, and can be estimated from the 10 percentage ( -20 dB) points
of the motion time histories.
7) In hard rock excavation, where underground disturbances cause higher frequency vibration, a factor of up
to 128 has been found to be satisfactory for residential properties in some countries.
8) The magnitudes for transient vibration in offices end workshop areas should not be increased without con-
sidering the Ixsibility of significant disruption of working activity.
9) Vibration acting on operators of certamn processes, such as drop forges or crushers whi ch vibrate working
places, may be in a separate category from the workshop areas comddered here. Vibration magnitudes, for the
operators of the exciting processes, which are specified in ISO 2631-1, will then apply.
10) Doubling the suggested vibration magnitudes for continuous or intermittent vibration and repeated tran-
siam vibration (fourth column) may result in adveme n t and t t may increase gn',antfy if the levels are
quadrupled (where available, dose/rekoonse curves can be consulted).
7
where A is the mid-span deflection of an equivalent SDOF system due to its own weight
and g is the gravitational acceleration (386.4 in./sec2). For a floor system, & may be
approximated by
(A8 + AG)
[2] = 1.3 + &S
where AB is deflection of floor beam due to flexure and shear,AG is the deflection of the
girder at the beam support due to flexure and shear, and AS is the shortening of the
column or wall support. The constant 1.3 in the above equation applies to both simply
Supported and fixed-end beams. For fixed-cantilevers a value of 1.5 should be used. In
the calculation of A, continuous beams on pin supports should be treated as simply
supported, since vibration nodes exist at the supports.
If shearing deformations are negligible, then the transformed moment of inertia of the
floor beam, It, may be used to estimate its natural frequency:
[3]
.x /gEIt
f = K W L 3
where K = for simply supported beams. Values of K for various end conditions are
readily available from tables such as those contained in Reference [7]. It is the
transformed moment of inertia of the composite beam section, E is the modulus of
elasticity of steel (29000 ksi) and W is total weight supported by the beam. Usually a
sustained portion of the live load (about 10% to 25% of the total design live load) is
included in this weight estimate. Finally, L is the span length of the beam. For
computation of It, the effective slab depth (de) is assumed equal to the depth of a
rectangular slab having the same weight as the actual slab, including the concrete in
valleys of the decking and the weight of the metal deck (see Figure 7).
The effect of girder and column support flexibilities on the first natural frequency of the
system, may also be approximated by:
1 I 1 1
[4] - + +
f 2 (fb)2 (fg)2 (fs)2
where fb, fg, and fs are the natural frequencies of the beam, girder, and column supports
each computed individually.
The reader should note that floor systems are complex and have multiple natural
frequencies. The above simplified procedures usually provide a good estimate of the first
natural frequency. However, depending on the activity of concern, this might or might
not be the natural frequency of greatest concern. For example, for most non-rhythmic
activities (i.e. walking) it is very unlikely that the column supports will have a significant
participation in the response. For these cases, the natural frequencies of great interest
are those of the floor beam alone, the girder alone, and the combined beam and girder
system. On the other hand, all three natural frequencies (i.e. beam; beam +girder;
beam + girder-i-support) should be considered in design for rhythmic activities.
0.63
0,4 I
0,25 .
0,16
L
0,1 ;
%. 0,063 t
0,04
' - 0 , 0 2 5
0,016
001
0,0063
0.004
0,002 5
0,0016
0.001
1
. See Table !
c! -
- - 60-'
32
H16
2
1,
i t
1,6 2.5 /+ 6.3 10 16 25 40 63
Frequency ' centre frequency of one-third octave band, Hz
100
Figure 6. Combi ned di rect i on criteria curves f or vi brat i on in bui l di ngs lo
1, Spacing S b
_3L._ 1 L Beam Spacing S li,
. . . . . I Z] :,
e e e e i* , r e
I T
ACTUAL MODEL
Figure 7. Tee-Beam model f or comput i ng transformed moment of inertia TM.
9
EXAMPLE 16: Estimate the natural frequency of the following floor beam. The girder and
column support motions are small and can be ignored.
GIVEN:
BEAM: W21x44
SPAN -- 41'
LIVE LOADS:
, ,
SLAB: 2 in. metal deck + 3
slab weight = 41 psf fi
f ' c = 3000 psi
Concrete weight = 115 pc
-0" SPACING = 10' -0"
Office ......... 50 psf
Partitions ..... 20 psf
Misc. -......... 10 psf
light weight concrete
- - - > Total Live Load = 80 psf
SOLUTION:
Support motions are negligible and the beam is not deep. Hence, the shearing
deformations may be ignored as well and we can use the It formula to calculate f.
de
S
14 i S / n i
525",
20.66"
vi
.
.j.--W21 x44
As =13.0in
Is =843
d = 20.66in.
BEAM MODEL
EC = ( W c ) l ' 5 c = (115 pcf)l .5-3ksi = 2136ksi
Es 29000 ksi
n - Ec 2136 ksi - 13.6 de --
actual slab weight
concrete weight
41 psf in
= 115 pcf (12)
Distance from c.g. to slab top, Yt is calculated as:
I 10' x12 ,,,, .20.66"
(2)(' 1-.6 )(4.3 )" + (13.0 in2) t + 5.25")
Yt = (10'x12) = 5.6"
13.6 (4.3") + 13.0i n 2
The transformed moment of inertia is:
__1_1 4.3",2
10'x12)(4.3")3 + (10'x12)(4.3)(5 6- + 843 in4 +
It = ( )( 13.6 13.6 ' 2 I
,20.66"
+ (13.0in2)( + 5.25"-5.6") 2 = 2648i n 4
= 4.3"
10
Assuming that10% of the design live load acts as a sustained load during vibration, the
participating weight is calculated as:
WDL = WSlab + WBeam = (0.041 ksf)(41')(10') + (0.044 k/ft)(41') = 18.6 k
WLL = (0.10)(0.080 ksf)(41')(10') = 3.3 k
W = WDL + WLL = 18.6 + 3.3 = 21.9 k
Hence, the natural frequency is:
_ ! gEIt
_ . , = 5.3.z
EXAMPLE 2: For the typical interior beam shown below, estimate the first natural
frequency by:
(al using [3] and [4]; (b) using [1] assuming column shortening is inconsequential; (c)
using [1] assuming column shortening of AS= 0.50 inches should be considered.
Assume the beam self-weight and 10% of the design live load are included in the 80 psf
estimate of floor weight.
(
, 4 0 ' - 0 " ,
_ j : W21x50 (It =3533 in ) z
Iom under
DO ....
' -
( 3' ML' rALZX
DCSIGH LIllE
W21x50 (It=3533 in4)
SOLUTION:
Since both beams and girder are shallow, shearing deformations may be ignored.
(a) For the beam:
JgEIt
For the girder:
fg = K
WL3
W = (80 psf)(10')(40') = 32,000 lb = 32 kips
(2) j1386.4)(29000)(3533)
"V =
W = 2(32 ki ps)+ (0.055)(30')= 65.65 kips
(2) /i380 4)(29000)(4485)
' (65.62" = 6.36 Hz
* Calculated based on a total column height of 130 ft. and an average sustained axial stress of 12 ksi.
A LG (130 ft)(12)(12 ksi)
= --"- = -- 0.64 in.
E 29000
The factor 1.30 is applicable to A for frequency calculations since uniform mass distribution along the
A 0.64
column height is assumed: As - 1.--3 - 1.3 - 0.50 in.
11
(b)
lc)
1 1 1
From [4]' - +
f2 (5.25)2 (6.36)2
- - - > f = 4.05 Hz
The beam deflection at midspan is:
5wL4
AB = 384Eit
5(32)(40x12) 3
= = 0.45 in.
384(29000) (3533)
The girder deflection at the beam support (1/3rd point) is:
5PL3 5(32)(30xl 2)3
AG - 162EIt 162(29000)(4485) 0.35 in.
The natural frequency of the system is then determined:
(AB + AG) (0.45 + 0.35)
A = = = 0.62 in.
1.3 1.3
I 1 /386.4
f - = - 3.97 Hz
Adding column shortening to the natural frequency calculation:
A = 0.62" + 0.50" = 1.12"
I 1 38J--.4
f - 2to ' = 2-- ' 12
- 2.96 Hz
FLOOR VIBRATION DUE TO WALKING
To model the impulse caused by a person walking, a standard heel drop impact hasbeen
defined2,. This is the impulse initiated by a person weighing 170 pounds who supports
his weight on his toes with the heels raised about 2.5 inches, and then suddenly drops
his weight through his heels to the floor. A plot of the resulting heel drop impact and a
typical floor response to such impact are shown in Figures 8 and 9, respectively.
Several investigators have suggested methods to evaluate and design for floor vibrations
caused by heel drop impacts2.8, 1.13,14.17018. Among them, Murray's acceptability
criterion TM enjoys the most wide-spread use by structural designers in United States.
In this section, six such methods are introduced and applied to a sample floor vibration
design example.
Murray's Acceptability CritErion
MurrayTM provides a step-by-step procedure for evaluating potential floor vibration
problems in residential and office environments. Design tables have been published
which simplify application of this technique6. The method is based on field measurements
and human response studies performed on approximately 100 floor systems. For
commercial environments, the use of the criteria suggested by an ASCE Ad Hoc
committee chaired by Ellingwood [1986] and covered later in this publication is
recommended.
12
Figure 8.
d01
ii
d
0
FIO
I J I
600
SUIIO
Sl i t
I I I I %1
I
O
lO !'0 310 40 SO ,lO
yI&I{. MI
Average plot of force versus time for heel impact4
! :
I I
INITIAL r i a l A C C l t l l A l l O t t
i - i - - / ' -
.... OA&IPlNGI A I l O
t * A , -
/ r -
t!^t ^
I
I !
I -
Z
o
f f
6
I .i t 3 .O
ti&Il e. S
Figure 9. Typical floor response to heel impact (High frequencies filtered out). 4
TABLE 2 --- Suggested ranges for available floor system damping
(after Murray 12.13.14)
Source Damping Comments
Bare Floor 1% - 3% Lower limit for thin slab of lightweight
concrete; upper limit for thick slab of
regular weight concrete
'Ceiling 1% - 3% Lower limit for hung ceiling; upper limit for
sheetrock on furring attached to beams
Mechanical 1% - 10% Depends on amount and attachment
....Systems
Partitions 10% - If attached to the floor at three points or
20% more and not spaced more than every five
floor beams.
13
The procedure for applying Murray' s acceptability criterion is as follows6:
(1) Estimate the total amount of damping that will be available, Davai I. Murray' s
estimates of available damping whi ch are based on observation only are shown in
Table 2. If the total available damping is greater than 8 to 10%, the beam is
satisfactory and further investigation is not necessary.
(2) Compute composite section properties and the first natural frequency of the beam,
f. If f is greater than 10 Hz, the beam is satisfactory regardless of the damping
provided.
(3) Compute the initial maximum amplitude of the beam, Aot, due to a standard heel-
drop impact as:
L3
[5] Aot = (DLF)max x (8--t )
where all units are in kips and inches and (DLF)max is the' dynami c load factor.
Values of DLF for various natural frequencies are listed in Table 3.
(4) Account for the stiffness contribution of adjacent beams by estimating the total
effective number of beams, Nef f, where:
[6] N e f f = 2.97-0.0578Idle] + 2.56x10-8[L--tl
where S is beam spacing and de is the effective slab thickness, both in inches (see
Figure 7).
(5) Divide Aot by Nef f to obtain a modified initial maximum amplitude, Ao, which
accounts for the stiffness of adjacent beams:
Aot
[7] Ao - Neff
(6) Estimate the required level of damping, Dreqd as:
[8] Dreqd = 35Aof + 2.5
(7) Compare values of DavaiI and Dreqd:
[9] If Dreqd---< Davail - - -> The beam is satisfactory
If Dreqd > Davail - - -> Redesign is recommended
If the available damping cannot be estimated, Murray suggests the comparison
summarized in Table 4.
14
TABLE 3 --- Dynamic Idad factors for heel-drop Impact. 14
f, Hz DLF F, Hz DLF F, Hz DLF
1.00 0.1541 5.50 0.7819 10.00 1.1770
1.10 0.1695 5.60 0.7937 10.10 1.1831
1.20 0.1847 5.70 0.8053 10.20 1.1891
1.30 0.2000 5.80 0.8168 10.30 1.1949
1.40 0.2152 5.90 0.8282 10.40 1.2007
1,50 0.2304 6.00 0.8394 10.50 1.2065
1.60 0.2456 6.10 0.8505 10.60 1.2121
1.70 0.2607 6.20 0.8615 10.70 1.2177
1.80 0.2758 6.30 0.8723 10.80 1.2231
1.90 0.2908 6.40 0.8830 10.90 1.2285
2.00 0.3058 6.50 0,8936 11.00 1.2339
2.10 0.3207 6.60 0.9040 11.10 1.2391
2.20 0.3356 6.70 0,9143 11.20 1.2443
2.30 0.3504 6.80 0.9244 11.30 1.2494
2.40 0.3651 6.90 0.9344 11.40 1.2545
2.50 0.3798 7.00 0,9443 11.50 1.2594
2.60 0.3945 7.10 0,9540 11.60 1.2643
2.70 0.4091 7.20 0,9635 11.70 1.2692
2.80 0.4236 7.30 0,9729 11.80 1.2740
2.90 0.4380 7.40 0,9821 11.90 1.2787
3.00 0.4524 7.50 0.9912 12.00 1.2834
3.10 0.4667 7.60 1,0002 12.10 1.2879
3.20 0.4809 7.70 1.0090 12.20 1.2925
3.30 0.4950 7.80 1.0176 12.30 1.2970
3.40 0.5091 7.90 1.0261 12.40 1.3014
3.50 0.5231 8.00 1.0345 12.50 1.3058
3.60 0.5369 8.10 1.0428 12.60 1.3101
3.70 0.5507 8.20 1.0509 12.70 1.3143
3.80 0.5645 8.30 1.0588 12.80 1.3185
3.90 0.5781 8.40 1.0667 12.90 1.3227
4.00 0.5916 8.50 1.0744 13.00 1.3268
4.10 0.6050 8.60 1.0820 13.10 1.3308
4.20 0.6184 8.70 1.0895 13.20 1.3348
4.30 0.6316 8.80 1.0969 13.30 1.3388
4.40 0.6448 8.90 1.1041 13.40 1.3427
4.50 0.6578 9.00 1.1113 13.50 1.3466
4.60 0.6707 9.10 1.1183 13.60 1.3504
4.70 0.6635 9.20 1.1252 13.70 1.3541
4.80 0.6962 9.30 1.1321 13.80 1.3579
4.90 0.7088 9.40 1.1388 13.90 1.3615
5.00 0.7213 9.50 1.1434 14.00 1.3652
5.01 0.7337 9.60 1.1519 14.10 1.3688
5.20 0.7459 9.70 1.1583 14.20 1.3723
5.30 0.7580 9.80 1.1647 14.30 1.3758
5.40 0.7700 9.90 1.1709 14.40 1.3793
15
TABLE 4 --- Required damping comparison chart
(after Murray 12.13,4)
Computed Required Damping Range
Dreqd < 3.5%
3.5% < Dreqd < 4.2%
Dreqd > 4.2%
Comments
System will be satisfactory even if
supported areas are completely free of
fixed partitions.
Designer must carefully consider the office
environment and the intended use.
Designer must be able to identify an exact
source of damping or artificially provide
additional damping to be sure the floor
system will be satisfactory. If this can not
be accomplished, redesign is necessary.
EXAMPLE 36: Use Murray's acceptability criterion to investigate the adequacy of the
floor beam of Example1 for walking induced vibration. This is a floor beam in an office
building where the girder and column support motions are small and can be ignored.
SOLUTION:
(1)
(2)
(3)
(4)
(5)
Estimate available damping:
(Floor at 1%) + (Ceiling at 1%) + (Mechanical at 3%)
> DavaiI = 5% < 8% > Continue the analysis
Calculate natural frequency, f:
from EXAMPLE 1: f = 5.3 Hz <: 10. Hz
Since f is less than 10 Hz, the analysis procedure is continued.
Compute the initial maximum amplitude of the beam, Aot:
For f = 5.3 Hz from Table 3 (DLF)max = 0.7580
L3
Aot -- (DLF)max x (8---0-t ) = (0.7580) x I-
Calculate Neff:
(41 'xl 2)3
80(29000)(2648)-] = 0.015 in.
Neff = 2.97- 0.0578 + 2.56x10-8 =
; o o [lOxl l
4.3" J + 2'56x10-8[ 2' J
Calculate the modified initial maximum amplitude, Ao:
Aot 0.015 in
ao - - - 0.0078"
Neff 1.92
= 1.92
16
(6)
(7)
Estimate Dreqd:
Oreqd = 35Ao f + 2.5 = 35(0.0078)(5.3) + 2.5 - 3.9%
Compare the values of DavaiI and Dreqd:
Dreqd = 3.9% <: DavaiI = 5.0 - - - > The beam is satisfactory
Ellingwood et. al Recommendations for C0mmerial Environments As a part of a report
issued by an Ad Hoc ASCE committee on serviceability research, a vibration criterion for
commercial floor systems, for example in shopping centers, was recommended l.s. The
criterion is considered satisfied if the maximum deflection for a 450 lb force applied
anywhere on the floor does not exceed 0.02 inches. Both the Canadian Standards
Associations and Murray4 recommend that the natural frequency of commercial floor
systems be kept greater than 8 Hz in order to minimize the possibility of resonance due
to walking.
EXAMPLE 4: Determine if the floor system of Example 2(b) satisfies Ellingwood et. al.
recommendations as a part of a shopping center floor system. Assume the number of
effective tee-beams, Neff = 1.96.
SOLUTION:
Examine the maximum deflection due to a 450 lb load on the beam:
* . 4'girder 4'support
max = "beam t 2 + 4
0.450L3 1 , 0.450(40' x12) 3)( 1 ,
Abeam = (. 4---t )(N--) = t48x29000x3533 1--1 = 0.0052 in.
0.450L3 ( 0.450(30' xl 2)3 ,
4.girder = (' 4-t ) = 48x29000x4485J = 0.0034 in.
0.0034 0.00
4.max = 0.0052 + 2 + 4 - 0.0069 in. < 0.020 in."
However, since the floor system natural frequency of 3.21 Hz is significantly less than
Murray's suggested value of 8.0 Hz, redesign is recommended.
> O.K.
wi88-Parmelee Rating Factor Criterion
Wiss and Parmelees conducted a laboratory study to investigate human perception of
transient floor vibrations. 40 volunteers were subjected to platform motions designed to
simulate floor vibrations due to heel-drop impact. An empirical formula was developed
which related human response to the floor system's maximum displacement amplitude
A0, the first natural frequency, f, and available damping, Davai I, such that:
fao 0.265
[10] R = 5.08 [ 0.217]
(Davail)
where R is the mean response rating, interpreted as follows:
17
1] imperceptible
[11] R 3 / barely perceptible
= = /distinctly perceptible
strongly perceptible
severe
The Wiss-Parmelee rating factor was adopted by the United States Department of
Housing and Urban Development as a criteria for acceptability of floor systems where a
limit of R<__ 2,5 was established, The Wiss-Parmelee rating method, which is also
referred to as the GSA/PBS acceptability criterion has been criticized for not being
sufficiently sensitive tofloor system damping13,7.
EXAMPLE 5: Determine if the floor beam of Example 3 is acceptable according to the
GSA/PBS criterion.
SOLUTION: With f, A0, and DavaiI already known from Example 3, we can directly
proceed with calculation of the Wiss-Parmelee rating factor:
fao 0.265 [(5.3 Hz)(0.0078")10.265
R = 5.08 [ 0.217 ] = 5.08 0.217 "
(Davai I) (0.05)
= 2.59 > 2.50
> Beam not acceptable according to GSA/PBS provisions.
Modified Reiher-Meister Scale
As early as 1931, Reiher and Meister reported results of their investigation of human
perceptibility to steady state vibration s. Their studies covered a forcing frequency range
of 3 to 100 Hz and a displacement amplitude range of 0.0004 inches to 0.40 inches. In
the early 1960's, Lenzen suggested that if the Reiher-Meister amplitude scale was
multiplied by a factor of 10, the resulting scale would be applicable to lightly damped
floor systems (damping less than 5% of critical). The resulting scale which correlates
human perceptibility with natural frequency and displacement amplitude, is called the
Modified Reiher-Meister Scale and is shown in Figure 10. As a result of studies
conducted on numerous beams, Murray in a 1975 paper 12 suggested that
"steel beam-concrete s/ab floors w/th 4% to 10% critica/ damping which plot
above the upper one-ha/f of the distinctly perceptible range w/I/result in
complaints from the occupants; and systems in the strong/)/perceptible range will
be unacceptable to both occupants and owners."
The Modified Reiher-Meister scale is frequently used by designers along with an
additional method (for example Murray's acceptability criterion) to pass judgement on
border-line situations. The main criticism of this scale is its lack of explicit consideration
of damping, which is considered to be the most important factor involved.
18
EXAMPLE 6: Use the Modified Reiher-Meister scale to determine a vibration perceptibility
level for the floor beam of Example 3.
SOLUTION:
With f - 5.3 Hz and A0 = 0.0078 inches enter the Modified Reiher-Meister chart
of Figure 10. The beam plots below the distinctly perceptible range and hence is
acceptable.
Canadian Standards Association Scale (CSA)
Based on the extensive research work by Allen and Rainer= an annoyance criteria for
floor vibrations in residential, office and school room environments was adopted by the
Canadian Standards Association (CSA) and was included as Appendix G to CSA
Standard S16.1-1974 (Steel Structures for Buildings -- Limit States Design). This criteria
sets limits on peak acceleration experienced by the floor system in terms of its natural
frequency and available damping (see Figure 11).
For design purposes, the peak acceleration, 3' may be estimated from the now familiar
maximum displacement amplitude, A0, assuming a harmonic floor response at the floor's
first natural frequency:
7 = (2tt f)2 (A0)
The chart in Figure 11 consists of a base curve for continuous vibration, and three limit
curves for walking vibration, for 3%, 6%, and 12% available damping. A floor system
plotting below the corresponding limit curve is considered satisfactory.
EXAMPLE 7: Use the CSA scale as devised by Allen and Reiner to determine
acceptability of the floor beam in Example 3.
SOLUTION: For f = 5.3 Hz and A0 = 0.0078 inches, estimate the peak acceleration:
y = (2 f)2(A0) = (2 x 5.3)2(0.0078) = 8.64 in/sec2 = 2.2 %g
Enter the chart of Figure 11 with these values. The required damping suggested by
the chart is less than the 5% provided. Hence, the beam is satisfactory.
Tolaymat's Criterion
Tolaymat 7 reviewed esults of 96 composite floor systems studied by Murray as a basis
for his acceptability criterion3, and suggested a new rating system that is claimed to
provide a better correlation between test results and reported human perceptibility levels.
In contrast to most other methods covered in this section, which are based on study of a
single heel drop impact, Tolaymat used a series of impacts to simulate excitation caused
by walking humans.
According to this approach, a floor system is rated acceptable if it satisfies one of the
following two conditions:
19
I1!
i I i Ill
'x! III
_ L I l l N
Ill
..% x X i x xlr'"'illii'
I S IQ ZO $0 100
F r e q u e n c y , CPS
Figure 10.
10g
SO
20
31.
0
tt
S
Z
0
2
1.0
0.$
Figure 1 1.
Modified Reiher*Meister perceptibility chart.
I I I I I
/
/
/
/
- /
DAMPING IATIO 12%J
- /
/
/
0AMPING lAllC
/
/
/
?
/
/
0.1
IIAIlG 3%
------ C!ITEIIA FOI
WALKING VIIIATION$
AS GIVEN IY #ILEI,,
IMPACT TEST
" CIIITIIIIA POI
CONTINUOUS VIIRATION
(10 TG 30 CYCLIS)
AVE'AGE2/t,X IC--
VVV
I I I .J !
2 4 ir, to 20
FREQUENCY, N4
CSA annoyance criteria chart for floor vibrations 2.
20
A2
[13] (1) A0 __< 1.15 with Amax <__.0.015 in.
[14] (2) (Amax)X(f) < 0.050
where A0 and f are as defined previously, A2 is the second heel drop maximum
amplitude and Amax is the absolute maximum heel drop amplitude, both in inches.
While on the surface the application of this approach seems simple, the reader should be
reminded that determination of A2 and Amax ,in general, requires calculation of the
dynamic response of a SDOF system (i.e. floor beam) to a general excitation (i.e. a series
of heel drop impacts). A procedure not suitable for hand calculations. A rather simple
computer program, however, can do the job and a diskette containing one such program
accompanies Reference 15.
FLOOR VIBRATION FROM RHYTHMIC ACTIVITIES
Coordinated rhythmic activities such as dancing, audience participation in arenas and
concert halls, and most importantly aerobics can result in undesirable levels of vibration.
For rhythmic activities, it is resonant or near resonant behavior that results in significant
dynamic amplification and hence human discomfort. The most rational design strategy is
to provide enough of a gap between the natural frequency of the floor system, and the
dominant frequencies excited by planned human activities to reasonably assure that
resonance will not occur. Multi-purpose facilities, such as floor systems in aerobics gyms
and office space on the same floor, pose the most difficult vibration design task.
Allen3.4.s has reported the most comprehensive design guidelines on this subject. His
recommendations have been reflected in the recent serviceability criteria supplement to
the National Building Code of Canada. Not surprisingly, the material presented in this
section is mainly based on information contained in References 3,4, and 5.
While for most rhythmic activities, consideration of the first harmonic (main frequency)
of the activity is sufficient, for aerobics and other coordinated jumping exercises, the
second and third harmonics can make significant contributions and should be considered
in the analysis. Figure 12 shows such a third harmonic resonance which was caused by
aerobics activity at 2.25 Hz on a 6.7 Hz floor system4.
TIME RECORD
5.19 - - - I I I 2.87
3.32 2.39
1.46 1.91
, . .
-2.28 0.96
-4.14 V ' [ 0.48
0 0.4 0,8 I 1.6 2.0
TIME, s
Figure 1 2.
B
FOURIER TRANSFORM
I I I I
11'IIRO
HARMONIC
RESONANCE
2 4 6 8 10
FREQUENCY. Hz
Vibration of a 6.7 Hz floor due to aerobics at 2.25 Hz4.
21
According to AlienS:
"Resonance is the most important factor affecting aerobics vibration, hence natural
frequency is the most important structural design parameter. The problem is to
get the natural frequency away from the three harmonics."
Design steps to prevent floor vibration from rhythmic activities may be summarized as
follows:
(1) For each type of activity, determine the dominant range of forcing frequency, ff
(see Table 5). Notice that for aerobics and jumping exercises, the first three
harmonics should be considered.
(2) Select a maximum acceptable limit for floor acceleration, a0. Use the values
recommended in Table 6, or ISO charts as discussed previously
(3) Select a dynamic load factor, (x. See Table 5 for guidance. Estimate the
distributed weight of the participants, Wp. When only a portion of span is used for
the activity the load Wp can be estimated by taking the total load on the partially
loaded span and distributing it uniformly over the entire span. Table 5 may be used
to arrive at a reasonable estimate for Wp.
(4) Compute the total floor load, wt by adding the normally sustained, non-active
load and Wp.
(5) Compute the natural frequency of the floor system, f, using an appropriate method
such as one of the methods discussed in this publication.
(6) Check the following criterion for the minimum natural frequency of the floor
system:
, 1.3 (ZWp
[15] f _> ff I + ao/g wt
where ao/g is the acceleration limit discussed in step 2 above, expressed in
percent of gravitational acceleration. The factor 1.3 in [15] is subject to the same
discussion provided for [2].
For aerobics and jumping exercises, the first three harmonics of the forcing
frequency should be considered. However, since these harmonics add together,
the factor 1.3 in [15] should be increased to 2.0. Hence, the governing criterion
for aerobics becomes:
, 2.0 (x Wp
[16] f (i)(ff) I + a0/g wt
where i= 1,2,3 is the harmonic number. Condition [16] should be satisfied for each
of the three harmonics.
Furthermore, Allen3 recommends that floor systems in assembly occupancies that do not
meet the minimum natural frequencies of Table 7 should be evaluated more carefully.
22
TABLE 5 --- Suggested design parameters f or rhythmi c events3.4,s.
Activity
Dancing
Lively concert
or sport event
Forcing
frequency
ff, Hz
1.5 - 3. 0
1.5 - 3. 0
Weight of participants*
Wp, psf
12.5 (27 ft2/couple)
31.3 (5 ft2/person)
Dynamic load
factore.
(z
I
0.5
0.25
Dynamic load
OrWp, psf
6.25
7.83
Aerobics
1st Harmonic 2 - 2.75 4.2 (42 ft2/person) *** 1.5 6.30
2nd Harmonic 4 - 5.50 0.6 2.52
3rd Harmonic 6 - 8.25 4.2 (42 ft2/person) *** 0.1 0.42
4.2 (42 ft2/person) ***
* Density of participants is for commonly encountered conditions. For special events the density of
participants can be greater.
** Values of (x are based on commonly encountered events involving a minimum of about 20 participants.
Values of = should be increased for well-coordinated events (e.g. jump dances) or for fewer than 20
participants.
*** Suggested revision to the 1985 supplement of CSA codes.
TABLE 6 --- Recommended acceleration limits for vibration due t o rhythmi c
Occupancies affected
by the vi brati on
Office and residential
Dining, Dancing, Wei ght-l i fti ng
Aerobics, rhythmi c activities onl y
Mi xed use occupancies housing
aerobics
Acceleration limit,
percent lravity
0.4 t o 0.7
1.5 t o 2. 5
4to 7
2
m
activities 4
TABLE 7 --- Mi ni mum recommended natural assembly fl oor frequencies, Hz3.
Dance f l oors*, stadi a,
Type of fl oor constructi on gymnasi a** arenas**
Composite (st eel - concrete) 9 6
Solid Concrete 7 5
Wood 12 8
* Limiting peak acceleration 0. 02 g,
** Limiting peak acceleration 0. 05 g.
EXAMPLE 8: Determine the mi ni mum natural frequency needed for a composi te fl oor
system in a gymnasi um t o be used exclusively for aerobics and other similar exercises.
23
EXAMPLE 8: Determine the minimum natural frequency needed for a composite floor
system in a gymnasium to be used exclusively for aerobics and other similar exercises.
The total normally sustained load on the floor including the dead weight and th'e weight
of non-participating audience is estimated at 80 pounds per square foot.
SOLUTION:
(1)
(2)
(3)
(4)
(5)
(6)
Following the forementioned step-by-step procedure:
From Table 5, select a reasonable value for forcing frequency, say 2.5 Hz.
Since the floor is to be used for aerobics and rhythmic activities only, from Table 6
an acceleration limit of 4% to 7%g is reasonable. For this example we select an
acceleration limit of a0 = 0.05g.
We use the suggested values from Table 5 for weight of the participants, dynamic
load factors, and dynamic loads. Hence, dynamic loads for the three harmonics
are 6.30, 2.52, and 0.42 psf, respectively.
wt = WD.L. + Wp = 80 + 4.2 -- 84.2 psf
This step does not apply to this problem.
Check [16] for each of the three harmonics
1st harmonic:
,/ 2.0 6.30
f > (1)(2.50) I + 0.05 84.2 -
2nd harmonic:
. 02_..5 2.52
f (2)(2.50) 1 + .
5.00 Hz
7.41 Hz
3rd harmonic:
. 2.0 0.42
f >. (3)(2.50) I + 0.05 84.2 - 8.21 Hz Controls
The floor system should be designed to have a first natural frequency larger than
8.21 Hz. Notice that Table 7 suggests a minimum natural frequency of 9 Hz for this
case.
ACKNOWLEDGMENTS
The author's work on this publication was sponsored in part by John A. Martin and
Associates, Inc. Critical review of the manuscript by Dr. Roger M. DiJulio and Mr.
James Marsh is gratefully appreciated.
24
REFERENCES
[1] Ad Hoc Committee on Serviceability Research, "Structural Serviceability: A Critical Appraisal and
Research Needs," Journal of Structural Engineering, ASCE, Vol. 112, No. 12, Dec., 1986.
[2] Allen, D.E. and Rainer, J.H., "Vibration Criteria for Long Span Floors," Canadian Journal of C/vii
Engineering, The National Research Council of Canada, Vol. 3, No. 2, Jun., 1976.
[3] Allen, D.E. , Rainer, J.H.and Pernica, G., "Vibration Criteria for Assembly Occupancies," Canadian
Journal of Civ# Engineering, The National Research Council of Canada, Vol 12. No. 3, 617-623,
1985.
[4] Allen, D.E. , "Building Vibrations form Human Activities," Concrete international: Design and
Construct/on, American Concrete Institute, Vol 12., No. 6, 66-73, 1990.
[5] Allen, D.E.," Floor Vibrations from Aerobics," Canadian Journal of Civil Engineering, The National
Research Council of Canada, Vol 12. No. 3, 617-623, 1985.
[6] AISC Marketing, Inc., "Floor Vibrations in Steel Framed Buildings," Pittsburgh, PA, September,
1989.
[7] Belvins, mm, "Formulas for Natural Frequency and Mode Shapes," Van Nostrand Reinhold, 1979.
[8] Ellingwood, B. and Tallin, M., "Structural Serviceability: Floor Vibrations," Journal of Structural
Engineering, ASCE, Vol. 110, No. 2, Feb., 1984.'
[9] International Organization for Standards, "Evaluation o; human exposure to whole-body vibration --
Part 1: General requirements," International Standard ISO-2631/1-1985(E), Geneva, Switzerland.
[10] International Organization for Standards, "Evaluation of human exposure to whole-body vibration --
Part 2: Continuous and shock-induced vibration in buildings (1 to 80 Hz)," International Standard
ISO-2631/2-1989(E), Geneva, Switzerland.
[11] Lenzen, K.H., "Vibration of steel joist-Concrete slab Floors," Engineering Journal, AISC, Vol. 3., No.
3., 133-136, 1966.
[12] Murray, T.M., "Design to Prevent Floor Vibrations'," Engineering Journal, AISC, Vol. 12, No. 3.,
1975.
[13] Murray, T.M., "Acceptability Criterion for Occupant-Induced Floor Vibrations," Engineering Journal,
AISC, Vol. 18, No. 2, 62-70, 1981.
[14] Murray, T.M., "Building Floor Vibrations," T.R. Higgins lectureship paper presented at the AISC
National Steel Construction Conference, Washington, D.C., June, 1991.
[15] Paz, M, "Microcomputer Aided Engineering -- Structural Dynamics," Van Nostrand Reinhold, 1986.
[16] Reiher, H. and Meister, F.J., "The Effect of Vibration on People," Published in German in 1931,
English Translation in Report No. F-TS-616-R.E.H.Q. Air Material Command, Wright Field, Ohio,
1949.
[17] Tolaymat, R. A., "A New Approach to Floor Vibration Analysis," Engineering Journal, AISC, Vol.
25, No. 4., 137-143, 1988.
[18] Wiss, J.F. and Parmelee, R.H., "Human Perception of Transient Vibrations," Journal of the
Structural Division, ASCE, Vol. 100, No. ST4, April, 1974.
25
STRUCTURAL STEEL EDUCATIONAL COUNCIL
TECHNICALINFORMATION& PRODUCTSERVICE
' t _ /
i j I . 7 i : : : : '
! . ,:- '3
-)
':-- " .C..
c.....
' ]
APRIL 1992
,:'% . = j
/
{ ?r'-'j ::'" :' i"
.... :'", i'" ] : " : : : : i ! 9 : : :
By W.A. Thornton, Ph.D., P.E.
specially in today's climate of reduced
construction activity, it is important to
do everything possible to reduce costs.
Through the careful design of structural
connections, fabrication and erection costs can
be reduced.
Bracing connections constitute an area in
which there has been much disagreement con-
cerning a proper method for design. Research
conducted during the past decade is just now
being distilled into a consistent method of de-
signing connections based on equilibrium mod-
els for the gusset, beam, and column that re-
quire that yield not be exceeded globally on any
gusset edge or section, and also on any section
in the column or beam.
Careful selection of. connections
can substantially reduce the fabri-
cation and erection costs on many
steel construction projects
While there are many possible equilibrium
models, three are presented here and then
applied to the design of a connection to deter-
mine their cost-effectiveness.
Mo dsl I . This is the most common and
simplest of all equilibrium models. The force
distribution on the gusset, beam, and col-
umn are shown in Figure 1. As with all
equilibrium models, this model guarantees
that the gusset, beam, and column are in
equilibrium under the brace load P. If the
work point coincides with the gravity axes of
V p
H
W.P.
/
, , / IR
Mc .
V p
',,._.,. Mil-Hee
EQUILIBRIUM MODEL
I
Fig u rc Z: Model 1, the simplest equilibrium model
V V
e P p
J EQUILIBRIUM
ec ,, 2 _
Figure 2: .Model 2, one of several z4.[SC.Models
r ' , I I II mi I lira
w.P
R-'N
e
; , l
R
H,- H-H V,-O
Vc-V Hc= 'cV
e
c =
Figure 3: Force Distributions for Model 2
i c - E Q U I L I B R I U M M O D E L
-, 5
Fb;u;'c 4: Model 3, an equilibrium model with no couples
V p
I [ J
w.P
the members, equilibrium is achieved with
'no connection induced couples in the beam,
column, or brace. Model 1 has been referred
to sarcastically as the "KISS" method (Keep
It Simple, Stupid).
,, [,odel 2 (AISC Model). This model is oneof
several adopted by AISC based on the re-
search during the past decade. The force
distributions for the gusset, beam, and col-
umn are shown in Figures 2 and 3, and as
with Model 1, these force distributions guar-
antee that the gusset, beam, and column are
in equilibrium under the brace load P. If the
work point coincides with the member grav-
ity axes, equilibrium is achieved with no
connection induced couples in the beam,
column, or brace. Model 2 is a little more
complex for calculations than Model 1, but it
yields less expensive designs.
Model 3. This model is the result of the
author's search for an equilibrium model for
W.P
R'\
I R v'-h
--F---; vZ' J i
Hs--P Ye-
r-/(,+ q:.,8.e,F.
Figure 5: Force distributions for Model 3
WlS,,106 -"
352__...
.;',,'"' J; w,4,6o5
Figure 6 Datafor illustrative example connection
bracing connections that achieves equilib-
rium for all components of the connection
with linear forces only, i.e., no couples. It is
the most efficientit yields the least expen-
sive designs---of the three models presented
here but is also the most complex in terms of
calculations required. Note, however, that
this is not a serious problem because a
computer program makes the calculation
aspect of all three models of little impor-
tance. The force distributions for the gusset,
beam, and column of Model 3 are given in
Figures 4 and 5.
The beam shear R in Figures 4 and 5 is
shown applied to the beam-to-column connec-
tion, If the shear is large, it may be desirable to
distribute it to the gusset-to-column connec-
tion as well. In this case the gusset serves as
a haunch and the gusset-to-beam forces must
be adjusted to effect the desired distribution of
R.
2 Steel Tips Apri l 1992
2'/. 4 -t z
WiSx 106 3 s,ots *WI 4x605
2- / . 6x 6 I I -
Fi.,;:tn' 7: ;olution to example connection using Model I
2- / Bx 6' 1 '--, 8
.
BOLTS 9 _,,,
2-t 4x4xSed
Figure 9.' Solution to example connection using Model 3
2- / . 6x6 x I ---,
I Wl 8xl O 6 3 sots
BOLTS,- *';,'f% "' /
HOLES:STD 3 s , o . s - ' s,.__,/
I d " 2-L6x4x,,
i
k
F
I 4 ' 6 0 5
I I I
Fig:ire 8: Solution to example connection using Model 2
Figure 6 provides an example. The column
is a W14x605, the beam a W18x106, and the
brace a W12x87 with 450 kips.
Figures 7, 8, and 9 give the completed
designs for Models 1,2, and 3, respectively. A
cost comparison shows that Model 3 gives the
most economical design, while Model 1, the
"KISS" method, gives a design that costs ap-
proximately 28-30% more and Model 2 gives a
design that costs approximately 13% more.
Using a lighter column section, a W18x119, to
assess the effect of drilling the heavy flange of
the W14x605 reveals similar results.
To see the effect on a project of using Model
1 rather than Model 3, consider a 40-story build-
ing with eight bracing connections per story. If all
these connections were similar to those shown
in Figure 6, the cost of using Model 1 rather than
Model 3 would be (840-658) x 8 x 40 = $58,240.
i i i i i i i i
Fisurt' 11): Same cost stiffened and unstiffened column
No transverse beams
iCJ
i i ,
., 0
WI4IO9(A36) W14x90 ( A36)
One pair stiffeners = 200 lbs. steel
(Fillet welds)
This is assuming one bay of bracing on each of
the four faces. If two bays per face were used,
the extra cost of Model 1 would be about
$116,0O0.
Columns, when part of an unbraced frame,
are designed for bending moment as well as
axial force. The designer uses a rigid frame
analysis computer program, which also possi-
bly does a code check using the beam column
interaction equations or he performs the latter
operation manually. What the designer gener-
ally does not consider in his column design is
the "panel zone" between the column and the
transverse framing beams and this can be a
costly oversight.
Steel Tips Apri l 1992 3
Figure 10 shows a W14x90 column 34'-long
with fillet welded stiffeners and a same cost
W14x109 with no stiffeners. However, if a
Wl 4x99 column will work, a less expensive job
will result. The W14x109 also may be less
expensive if extra erection costs associated
with beams framing to the weak axis of the
W14x90 due to the stiffeners are considered.
Figure 11 shows the same W14'x90 column as
Figure 10, but here the designer has specified
full penetration groove welds of the stiffeners to
the column. This doubles the cost of the stiffen-
ers and means that an unstiffened W14x132
will cost about the same as the stiffened W14x90.
Now, looking at the sections between W14x90
and W14x132, we see that we have available a
W14x99, a W14x109, and a W14x120, all of
which will yield a less expensive design if they
satisfy the beam column design equations.
Figl.,'c i : Same cost stiffened and unstiffened columns
No transverse beams
P L zx7xl'-oSe
WI490 (A36) WI4xt32(A36)
(Full penetration welds)
Fief, rc I2: Rules of thumb: same cost columns
no transverse beams
- Pti:,x7xI:OSe
PL'4xll xa'-IO
S' CTION A-A
SECTION B-E}
Wl4x90 (A36)
I ll.
W14x145(A36)
4 doubler plates + 4 pairs of stiffeners = 1900 lbs steel
(Fillet welds)
Figure 12 shows the "fabricators nightmare"
of stiffeners and doublers. A clean W14x145
costs no more than the stiffened and doubled
W14x90, and all of the W14 Sections in be-
tween will give less expensive designs if they
satisfy the beam-column equations.
For the convenience of designers, Figure 13
gives the cost in lbs. of steel, as well as the cost
of column splices. Column weights can be
increased by approximately the amounts shown
here without increasing costs because, as pre-
viously mentioned, the stiffeners and the dou-
blers will tend to increase erection costs. (Note
that erection costs are not included in Figures
10 through 13.)
Figure 14 takes a different view. here the
connection with the stiffeners and doublers is
given per tributary length of column. As an
example, Figure 15 presents a W24x55 framing
to a column flange. The design moment is M =
212k-ft, which is just slightly less than the full
strength moment of the W24x55(A36), which is
226k-ft. The W14x90 column, which is deter-
mined to be adequate for M = 212k-ft and the
design axial load, requires stiffeners and dou-
blers. The W14x120, which is also adequate for
the design moment and axial force, requires no
stiffeners or doublers. Since 120- 90 = 30 lbs,
which is less than the 79 lbs from figure 14, the
W14x120 is the more economical choice. As
Figure 15 shows, $180 is saved per connection.
If there were 1,000 similar connections on the
job, savings would be approximately $180,000.
The stiffeners and doublers of the column
cost studies previously discussed are the result
One Pai r Stiffeners = 200 lbs steel
( f i l l e t welded)
(full penetration welded)
One Pai r
Doubler Pl at es = 550 lbs steel
One Doubler Pl at e = 280 lbs steel
One Col umn Spl i ce = 500 lbs steel
4 Steel Tips April 1992
Length of Increasein weight per foot with
column no increase in cost of "clean"
tributary to column
d connection StiffenersFillet Stiffeners
i with stiffeners Wel ded GrooveWelded
and doublers
._._.._ 10 95 135
12 79 113
4 Si' ]I:FENERS
14 68 96
2 DOUa.ES
16 59 84
NOWEAKAXiS
B Co.Ec.o 18 53 75
20 48 68
Column Selection Design Aid
of requirements for beam-to-column moment
connections, especially when full-strength mo-
ment connections are specified, as in Figure 16
for doublers. Since stiffeners and doublers can
add significant costs to a job, design engineers
should not specify full-strength moment con-
nections unless they are required by loads or
codes, e.g., ductile moment resisting frames for
high seismic loads.
For wind loads and for conventional mo-
ment frames where beams and columns are
sized for stiffness (drift control) as much as for
strength, full strength moment connections are
not required. Even so, many design engineers
will specify full strength moment connections,
adding to the cost of a structure.
Designing for actual loads has the potential,
without any increase in column weight, to dras-
tically reduce the stiffener and doubler require-
ments. On one recent 30-story building, a
change from full moment connections to a de-
sign for actual loads combined with using Fig-
ure 17 for doublers reduced the number of
locations where stiffeners and doublers were
required to several dozen from 4,500 locations
with an estimated cost savings of approximately
$50O,OO0.
The uniform design load (UDL) is a great
crutch of the engineer because it allows him to
issue design drawings without putting the beam
reactions on the drawings. Instead, often the
fabricator is told to design the beam end con-
: Example of use of column selection design
STORY HEIGHT 12'0
CokJrnn
wi cok.
Since 120 - 90 = 30 lbs. < 79 lbs.,
Saved 79 - 30 = 49 lbs./fi, x $.30/Ib. = $15/ft.
Therefore, per 12' of column, $15 x 12 = $180 saved
Building with 1,000 locations = $180,000 saved
:: . . . . . : Design for the full strength of the beams
db
M,
L / I M I
dc _
t =
WIEQ'D ' c
M2 FULL STRENGTH MOMENTS
Doubler Plates
Commonly Seen Requirements
nections for one-half UDL, or some other per-
centage to account for composite design, un-
less greater reactions are shown. Unless con-
centrated loads are located very near the beam
ends, UDL reactions are generally very conser-
vative. Because the reactions are too large,
extremely strong connections, such as double
framing angles, will often be required.
Single angles, because the bolts are in
single shear, will have about half the strength of
double clips for the same number of rows of
bolts. But if actual reactions are given, it will
almost always be found that a single angle
connection will work, perhaps with a couple of
extra rows of bolts.
Figure 18 is part of an industrial building with
dead Icad of 140 psf and live Icad of 250 psf.
Beam 1 of Figure 18 is shown in Figure 19. The
total Icad on Beam 1 is 82 kips and the actual
reactions are thus 41 kips. The one-half UDL
reaction is 45 kips, which is pretty close. Now
look at the connections. The minimum double
clip connection on this coped beam has four
rows and is good for 81 kips, almost twice the
actual reaction. Many designers routinely re-
quire "full depth" connections, i.e. six rows. The
six row double clip connection is good for 116
kips, almost three times the actual reaction.
However, a five row single angle is good for 52
kips, which is okay for the actual and the one-
half UDL reactions.
As this example illustrates, single angles
will work even in heavy industrial applications,
and they are much less expensive than double
clips, especially for erection. In Figure 20, the
connections for this W24x55 beam have the
same strength and have a differential cost of
$10 for fabrication. But, including erection, the
single angle beam costs approximately $25
less than the double clip beam. For a 30-story
building 200' x 200' with 25' bays and 200
beams per floor with tabs, there is a savings of
200 x 30 x 25 = $150,000.
Returning to Figure 18, suppose Beam 1 is
subjected to the same Icad of 82 kips total, but
32 of the 82 is a concentrated Icad located at
mid-span, such as from a vessel. Figure 21
shows the actual reaction of the beam, now a
W24x76, is still 41 kips, while the one-half UDL
reaction is 56 kips--which is 37% greater than
the actual reaction. This means while a five row
Figure 17: Design for the actual loads
Doubler Plates
' -Vt = Couu x,,4o S.eAR
t = 7 32 .95cl b
Will often eliminate doubler requirement Figure 18: Partial plan o[ industrial building fioor
Figure 19: Comparisons W21 x68 / - -3.25k/ft
for Beam I
k 25'-0 )
Beam Section Loads Reactions Connections
Uniform Conc. Total Actual 1/2 UDL Double Clips Single Clip
min.# CAP max# CAP # of CAP
kips/ft, kips kips kips kips of rows (kips) of rows (ki ps) rows (kips)
1 W21x68 3.25 0 82 41 45 4 81 6 116 5 52
Bolts 7/8 A325N, Clips 4 x 31/2 x /8, Welds 1/4" fillet
J".C,.=..;;:.-Z;...Tz:;';,='TT-"L=;--=-;'-= LJ--/ ~: . . . . .' 7 = - . ; = i i - : : :7. V: :%2: !7.,>.T:/i'J..'. :!??JUF::`.J;`-*va`7;` . . . . . .
. , :, : . Cost of same strength single and double clips
SINGLE CLIPS
W24,55
L 0' -0
I
V
DOUBLE CLIPS
24,55
k
I
Fabrication - $10 per beam less for single clips; Erection - $15 per beam less for single clips
Total Cost Reduct i on - $25 per beam usi ng single clips
Fi,urc 21: Comparisons
for Beam 1 (prime)
32K
t F
min.# CAP max.# CAP # of CAP
kips/fi, kips kips kips kips of rows (kips) of rows (ki ps) rows (kips)
,1 (prime) W24X76 2 32 82 41 56 4 83 7 137 5 52
Bolts 7/8 A325N, Clips 4 x 31/2 x %'8, Welds t / 4 " fillet
Figm'e 22: Comparisons
for Beam 2
82 K 82K
I - I
Uniform Conc. Total Actual 1/2UDL Double Clips Single Clip
min# CAP max,# CAP #of CAP
2 W33x118 0 82 164 82 114 6 150 9 210 8 92
Bolts 7/80 A325N, Clips 4 x 31/2 x 3/8, Welds ,5'16" fillet
single angle connection is okay for the actual
reaction, a six row connection with a capacity of
66 kips would be required for the one-half UDL
reaction.
Figure 22 shows the disparity between ac-
tual and one-half UDL reactions for Beam 2.
Again, single angles are sufficient.
This Tips was printed from an article that ap-
peared in the AISC magazine "Modem Steel
Construction. "A complimentary subscription to
"Modem Steel Construction" may be obtained
by contacting AISC, Chicago.
Steel Tips April 1992 7
S'7 TM ' '-'-I " :
auL, 1 J R A L S T E E L E D U ' " "'-:'"'x' "'
470 Fernwood Drive
Moraga, CA 94556
(510) 631-9570
SPONSORS
Adams & Smith
Allied Steel Co., Inc.
Bannister Steel, Inc.
Bethlehem Steel Corporation
C.A. Buchen Corporation
Butler Manufacturing Co.
G.M. Iron Works Co.
The Herrick Corporation
Hoertig Iron Works
Hogan Mfg., Inc.
Junior Steel Co.
Lee & Daniel
McLean Steel, Inc.
Martin Iron Works, Inc.
Nelson Stud Welding Co.
Oregon Steel Mills
Palm Iron & Bridge Works
pDM Strocal, Inc.
Reno Iron Works
H.H. Robertson Co.
Schrader Iron Works, Inc.
Southland Iron Works
Stockton Steel
Stott, Inc.
U.S. Steel Corporation
Verco Manufacturing, Inc.
Vulcraft Sales Corp.
The local structural steel industry (abovesponsors) stands ready to assist you in
determining the most economical solution for your products. Our assistance can
range from budgetprices andestimated tonnageto cost comparisons, fabrication
details and delivery schedules.
Funding for this publication provided by the California Field Iron Workers Administrative Trust.
TECHNICAL INFORMATION & PRODUCTSERVICE
JUNE 1992
What Design Engineers Can Do
To Reduce Fabrication Costs
This Tips wasprinted from an article that appearedin
the AISC magazine "Modern Steel Construction." A
compl i mentary subscription to "Modern Steel
Construction" may be obtained by contacting AISC,
Chicago.
Bill Dyker, manager of engineering, and John
D. Smith, vice presi dent of sales, with AISC-
member Garbe Iron Works, Inc., Aurora, IL:
When welding base plates to columns in the
shop or field, designs should not show "all-around"
,illet welds unless the weld is required to resist a
moment or a large uplift. Usually, welds as shown on
Page 4-130 of the AISC Manual of Steel Construc-
tion, ASD--Ni nth Edition are adequate. Similar con-
sideration should be given to attaching cap plates to
columns.
Desi gns, espe-
cially CAD generated
i ssues--shoul d be
reviewed to ascertain
that all pieces can be
erected without undue
restraints Also, the re-
viewer should be alert
to avoiding framing
conditions that cause
deep copes at the
ends of beams. This
A compilation ofcomments from
experienced fabricators across
the country
Avoid framing conditions that cause
deep copes at the ends of beams.
The cost for a few extra pounds of
steel per foot for a deeper beam
often will be easily offset by the
savings from not having to design
and fabricate special connections
and cope reinforcements.
solution occurs especially when relatively shallow
beams support deep beams and both beams may or
may not be at the same elevation. The cost for a few
extra pounds of steel per foot for a deeper beam
often will be easily offset by the savings from not
having to design and fabricate special connections
and cope reinforcements.
If beam-column moment connections require
doubler plates in the column web, consider increas-
ing the column size to eliminate the need for such
plates. The costs to design, detail, approve and
fabricate doubler plates could be offset by the cost
of a heavier column.
Designs should not arbitrarily call for stiffeners
in webs of columns at moment connections. With
the variety of computer programs available, (e.g.
AISC's CONXPRT program), the designer should
be able to describe
on the drawings which
columns require stiff-
eners and the stiffener
sizes. Again, perhaps
a slight increase in the
unit weight of a col-
umn would offset all
the costs--detailing
through fabrication--
associated with web
sti ffeners. These
costs really begin to
escalate when doubler plates are needed along with
stiffeners. Further complicating the situation and
adding to costs is when connections must be pro-
vided for beams attaching to column webs in the
vicinity of stiffeners.
Designers should not arbitrarily call for beam
connections that virtually fill the "T" distance of
beam webs when lesser connections will accom-
modate the loads.
Partial Penetration Groove Weld Full Penetration Groove Weld
i
Total Cost Per Weld
$145
Total Cost Per Weld
$350
Amount of weld in partial-penetration weld calculated using ,it/6 + 1/8.
Costs include: preparation; root pass; backup bar or back gouge;
cleaning of welds between passes; cut-off of run-out tab;
weld material; and inspection.
Average wage rates calculated at $60/hr.
Source: Drawings courtesy of Garbe Iron Works; calculations courtesy of American Bridge
When desi gners run into an uncommon framing
condition, they shoul d consult a fabricator for ideas
on how to solve the problem.
Designers shoul d not call for A325-SC bolts
unless they are meeting the criteria for the use of
such bolts as listed in the ASTM A325 Specification.
In its publications, AISC is placing more emphasis
on using bearing type connections. This will be
reflected in the forthcoming revised Engineering
For Steel Construction.
Galvanizing should be limited to members whose
exposure to the elements could result in structure
failure from excessive rusting and/or where painting
is impractical. Whether or not it is the intent of
Masterspec, its wordi ng in regards to galvanizing
calls for pieces to be galvanized which do not fit the
aforementioned conditions. Galvanized steel re-
quires special fabrication and has an inherently
higher cost.
Desi gners shoul d not arbi trari l y i ncl ude
Masterspec or other standard specification sections
in the design documents, as many times such
sections conflict with the drawings. Similarly, when
designers prepare their standard details and j ob
notes, they should present information pertinent to
the project. Standard j ob notes should be updated
periodically to reflect current industry practice.
Keep base plates and column sizes as uniform
as possible without grossly oversizing.
Do not specify high strength steel "here and
there"--keep it in groups, e.g. "columns" or "main
girders."
Keep connections, angles, and plates as stan-
dard as possible with the fabricator allowed to
choose bolted or welded.
Show steel plainly on structural drawings with-
out blending it into the architectural design.
2 Steel Tips June 1992
I
Thomas Schlafly, Director of Fabricating
Operations & Standards, AISC:
A few sample connections should be sized at
the same time that beams are sized to avoid prob-
lems such as requiring more bolts than the connec-
tion can handle.
Consider and note construction tolerances on
design drawings and provide adjustments in appro-
priate details. For example, make embedded plates
6" to 8" larger than the beam and connect with long
slots to accommo-
date concrete toler-
ances. This will al-
low the steel to re-
main in plumb even
if the concrete is
slightly off and will
simplify curtainwall
erection. Another ex-
ample would be to put curb plates on in the field or
to provide other means of adjustment.
Shade the pieces coming out of the page in
sketches to catch interferences and difficulties with
members in the "third" dimension.
A good rule of thumb to remember is that the
more pieces in a detail, the more expensive it is to
fabricate. Al so, a
poundof weldment
is worth about $30
to $60. Likewise, ex-
pansion joints are
very expensive and
should be avoided if
possible.
Rafters shoul d
be run up hip roofs
rather than horizontally, if possible, because the
geometry is much simpler.
Larry L. Mednick, president, with AISC-
member Globe Iron Construction Co., Inc.,
Norfolk, VA:
Structural and architectural drawings should be
coordinated so they agree (e.g. on details and
dimensions).
Develop details that work to minimize the re-
quired coordination between trades, such as those
found in AISC's Load and Resistance Factor De-
sign of Simple Shear Connections and Allowable
Stress Design of Simple Shear Connections (for
more information, call 312/670-2400 ext. 433).
Try not to use a lot of different size
sections to save a small amount of
steel. It may cost more to buy and
track the different section sizes than
is saved by the reduced weight.
, i I
Don't hesitate to ask a fabricator for alternate
connection types and details that work best and
minimize costs and still meet AISC-accepted proce-
dures.
Bob Petroski, P.E., vice president/general
manager/chief engineer with AISC-member
Hercules Steel Co., Inc., Fayetteville, NC:
We sometimes find big discrepancies between
where the architect and the engineer show steel. It
is very important that drawings be coordinated
dimensionally be-
tween the architect
and engineer.
Too many ad-
denda in the j ob
stage create head-
aches for the fabri-
cator. We like to see
a current set of draw-
ings when the job is issued; not a bid set of drawings
with a series of addenda.
Engineers need to be realistic when it comes to
welding. Don't call for full penetration welds if they're
not needed. Don't call for continuous welds if you
can use intermediate welds. Remember, when
there's a lot of welding on light pieces, it can result
Incomplete design drawings should
not be released simply to meet a
schedule. In the long run, it will
cause more delays as the fabricator
is forced to check the design.
in distortion.
Engineers should
show non-typical con-
nections on the draw-
ings so all fabricators
are bidding on the
same design.
I don't believe in
using a lot of different
size sections to save
a minimal amount of steel. It sometimes costs more
to buy and track the different section sizes than is
saved by the reduced weight.
Eugene Miller, retired structural engineer,
formerly with AISC-member American Bridge
Co. and AiSC-member Trinity Industries,
Houston:
Don't call for full penetration welds when partial
or fillet welds will do the job.
Incomplete design drawings should not be re-
leased simply to meet a schedule. In the long run, it
will cause more delays as the fabricator is forced to
check the designs and the detailer has to stop and
ask questions.
Steel Tips June 1992 3
I
Designers should consult with fabricators when
using a special type of weldment.
When design drawings are revised, the changes
should be properly highlighted on a design docu-
ment, rather than presented in sketch form.
Computers shouldn't have the last word in the
selection of member sizes. Designs should be prag-
matically reviewed to avoid ill-working solutions,
such as where a W31 frames into a W12.
Designers should write specifications for indi-
vidual projects rather than simply using hurriedly
adopted off-the-shelf specs.
Barry L. Barger, Vice President-Production,
AI SC-member Southern Iron Works, Inc.,
Springfield, VA: (these comments are excerpted
from his paper, "Practical Engineering In Shop
Fabrication And Erecti on--How It Can Benefit The
Owner," presented at the 1991 National Steel
Construction Conference)
Noted below are items typically shown or noted
on contract drawings that severely limit or prohibit a
full range of connections or may force the exclusive
use of framing angles even if the fabricator is
allowed to choose the
types of simple shear
connect i ons for a
proj ect. In all in-
stances, the problem
can simply be over-
come by showing the
required reaction for
the members on the
contract drawings:
Requi ri ng 7/8"-
Tubes and pipes make economical
column members. They are an
excellent choice when stiffness about
both axes is required. And they have
less surface area than equivalent
wide flange members.
diameter bolts when 3/4"-diameter are sufficient.
Requiring uniformly loaded beams to carry 125%
of the end reaction.
Using friction bolts (slip-critical) when bearing
bolts are adequate.
Specifying the minimum number of double rows
of bolts for each section size.
Listing unrealistically high reaction multiplication
factors on composite beams.
While it must be remembered that the engineer
of record is at liberty to be as conservative as he
wishes, the above requirements will always signifi-
cantly add to the project's cost and may not be in the
best interest of the owner.
I would also like to give a few tips that may either
make a job go easier and avoid problems. Most of
them are just simple common sense, but are often
forgotten.
For erection stability, if using leveling nuts, do not
use less than four anchor bolts.
Place column splices 4' above the top of steel so
that perimeter safety cables may be attached
before the next floor is erected.
If skewed hole patterns are required, try to skew
them in the connection material rather than the
main member.
When a job is designed in A572 Grade 50, list the
small beams (W8xl 0, etc.) that can use A36 so
the fabricator can take them from stock.
If making last minute changes, design in A36 so
that material may be located quickly from ware-
houses.
Stop stiffeners shortwhere you can so they do not
have to be fitted.
David T, Flicker, retired structural engineer,
formerly with AISC-member The Berlin Steel
Construction Co,, Inc,, Berlin, CT:
Tubes and pipes make economical column
members. They are an excellent choice when stiff-
ness about both axes
is required.
They can be used
as hollow members or
filled with concrete
However, there is no
great advantage to fill-
ing small tubes or pipe
wi t h concret e. A
TS3x3x1/4 x 10' has
a capacity of 37 kips
when filled with concrete and 33 kips when unfilled.
For larger columns, there is an advantage, though.
For example, a pipe 8" standard column 12' long has
a capacity when filled with 3,000 psi concrete of 217
kips, whereas the unfilled capacity is only 155 kips.
Hollow structural sections have a number of
advantages over wide flange shapes:
Tubes and pipes have less surface area than
equivalent wide flange members. For example,
listed here for comparison are the surface areas
per linear foot of three common sizes:
W8x31 = 3.89 sq. ft
TS8x8x1/4 = 2.65 sq. ft
pipe 8 std = 2.26 sq. ft
This can be a significant cost factor if the mem-
bers require an exotic surface coating or fire proofing.
4 Steel Tips June 1992
mill i I I I I I r l ,,t
AK;GLES
i ....l
END
L H
TF_E
L . - - - ] . .z.
THI U - Z PLATES 2 ANGLES
Tube Connections
Tubes offer excellent resistance to torsional forces
and can be used to support eccentric loads such
as relieving angles for brick veneer, stone, or
precast concrete.
Tubes also make efficient bracing members. They
also can be combined with other structural shapes
to produce some startling aesthetic effects.
In recent years, tube and pipe prices have be-
come more competitive. Availability, however, is
sometimes a questionmark--a fabricator or sup-
plier should be consulted.
Flagging Changes. Changes on all plans and
shop drawings issued subsequent to the date of the
contract should be flagged so that revisions can be
easily located. Designers, fabricators and erectors
alike should observe this suggestion.
Tolerances. It is essential that erected toleraces
are compatible with systems and materials being
supported by the steel frame so that adjustment for
the trades that follow is possible. The published
fabrication and erection tolerances will usually ac-
complish this. However, if special tolerances are
required, they must be clearly indicated on the
plans.
Multiple details. Shop bolting and welding on
the same beam increases material handling, an
important element in fabrication cost.
Fillet welds. If possible, fillet welds should be
5/16" maximum, or a size that can be made in one
pass. Oversized welds add unnecessarily to the
cost of fabrication and erection and also may cause
distortion.
Bol t uniformity. Minimizing the number of di-
ameters and types of bolts on a given job lessens
the chance for a mixup in the shop or field and allows
more efficiency in drilling or punching operations.
Anchor bolts. The possibility of foundation
errors will be reduced when the fewest anchor bolt
and base plates sizes are used and when anchor
bolt spacing is kept uniform throughout the job.
Partial depth stiffeners. Consider using par-
tial-depth beam and column stiffeners where they
are adequate rather than full-depth fitted stiffeners.
Composite beams. To make composite beams
economical, at least 6 to 7 lbs. of total beam weight
per stud should be saved.
Relieving angles. The thickness of relieving
angles is normally 5/16" or 3/8". If calculations
indicate a greater thickness than this, the basic
design assumptions should be reviewed and per-
haps a different approach attempted.
Odd sections. Before specifying odd sections,
the designer should contact a local fabricator to
determine their availability.
Steel Tips June 1992 5
STRUCTURAL STEELEDUCATIONALCOUNCIL
TECHNICAL INFORMATION & PRODUCTSERVICE
December 1993
Common Steel Erection
Problems
and Suggested Solutions
by
James J. Putkey
Acknowledgements
The author wishes to thank the following persons for their input, review,
and comments on the content of this Steel TIPS publication:
Members of the Structural Steel Educational Council
Dave McEuen, California Erectors, Bay Area, Inc.
William C. Honeck, Structural Engineer with Forell/Elsesser Engineers, Inc.
The information presented in this publication has been prepared in accordance with recognized
engineering principles and construction practices and is for general information only. While it is
believed to be accurate, this information should not be used or relied upon for any specific
application without competent professional examination and verification of its accuracy, suitability,
and applicability by a licensed professional engineer or architect. The publication of the material
contained herein is not intended as a representation or warranty on the part of the Structural Steel
Educational Council, or of any other person named herein, that this information is suitable for any
general or particular use or of freedom infringement of any patent or patents. Anyone making use
of this information assumes all liability arising from such use.
COMMON STEEL ERECTION PROBLEMS AND
SUGGESTED SOLUTIONS
List of Problems
No. Pa_=eNo.
Erection
5. One-Bolt Connections 11
6. Columns or Bents Tied Together With Non-Bolted Steel Joists 15
7. Steel Joists Without Bolted Bridging 16
8. Columns or Bents Tied in With Timber 17
9. Steel Columns or Partial Bents Not Tied In 18
10. Non-Self-Supporting Steel Frames 19
11. Column Splices Too Low or Too High Above Floor 20
12. Columns Interrupted by Beams 21
13. Columns Offset From Beam Framing 22
14. Revisions and Alternates Not Flagged on Drawings 23
15. Doubl e- Fr amedBeam Connections to Girder 24
16. Doubl e- Fr amedBeam Connections to Column Web 25
Bolting
17. Mixed Bolts 27
18. Mixed Bolt Diameters 28
19. Reuse of High-Strength Bolts 28
Welding
20. Prequalified and Non-Prequalified Weld Joints 29
21. Extending Continuity Plate for Back-up Bar 30
22. Welded Connections to Inside of Column 31
23. Restrained Welded Joints 33
24. Fi el d- Wel dedCurb Angles 35
Decking
25. Steel Floor Deck Spanning Uneven Surfaces 36
General
26. Project Specifications 37
Anchor Bolts
1. Low Anchor Bolts 5
2. Misplaced Anchor Bolts 7
3. Rotated Anchor Bolt Pattern 9
4. Inadequate Anchor Bolts for Column Erection 10
COMMON STEEL ERECTION PROBLEMSAND
SUGGESTED SOLUTIONS
Introduction
Preface
About two years ago a structural engineer asked me
the following question, "Why don't you write a book-
let on steel erection? We keep seeing the same
erection problems occur over and over again, and it
would be nice to have a reference for erectors,
fabricators, and structural designers to either avoid
a problem or to present a solution to a problem."
The question was posed to two steel erectors, and
they both thought such a publication would be an
excellent idea. The end result is this Steel TIPS.
Many publications exist that inform the structural
designer on how to select types of steel, design
economically, reduce fabrication costs, and how to
design various types of structures or portions of
structures. But what source of information is avail-
able to the designer when the steel erector makes
an inquiry regarding the steel design or experiences
problems that require the designer's input? These
inquiries or problems may result from:
Erection or fabrication errors.
Erection procedures or sequences.
Faulty work of other trade contractors.
Design that can lead to safety problems.
Erection equipment loads into the structure.
Changes or alternates requested by the owner.
Now, looking ahead in the construction timetable,
one might logically ask the following questions,
"What source of information is available to the
structural designer to produce a design that can
avoid these erection problems? What are the
details to avoid? What are the desired details? Why
doesn't the steel industry provide structural design-
ers, and others, with solutions to common design-
related problems experienced bythe steel erector"
Purpose
The purpose of this Steel TIPS is to provide struc-
tural designers and steel erectors with a basic and
convenient source of solutions to common steel
erection problems that involve the structural de-
signer.
Organization and Content
To provide structural designers with solutions to
common steel erection problems, 26 common prob-
lems with suggested solutions are provided. The
problems are divided into six categories: anchor
bolts, erection, bolting, welding, decking, and gen-
eral. In each category a specific problem is shown
by its title. The problem is then described and the
suggested solution is given.
The content of this Steel TIPS does not address the
various methods of erecting steel. If the designer
needs to design a structure with unusual features,
or with a required erection procedure or sequence,
then a sponsor firm of the Structural Steel Educa-
tional Council might be consulted to make certain
the unusual features can be economically erected.
The erection problems presented are not only "com-
mon'' problems, but may also be considered basic,
reoccurring problems. So the content is chosen to
be especially useful to the new structural designer
(and maybe experienced designers).
Some of the problems or portions of problems
addressed in this Steel TIPS are mentioned or
addressed in previous Steel TIPS, or in the AISC
publications Modern Steel Construction, and Steel
Design Guide Series. These problems and their
solutions are now conveniently gathered into this
publication.
4
1. Low Anchor Bolts
Problem
Anchor bolts are sometimes set with their tops
lower than the detailed elevation. Two situations
can exist: 1) the bolts are placed so Iow that the top
of the bolt is below the top of the base plate and the
anchor bolt nut cannot be engaged, or 2) the bolt
top extends above the base plate, but not high
enough to allow full thread engagement of the nut.
Setti ng Tol erances. Section 7.5 of the AISC
Code of Standard Practice requires the owner to
set anchor bolts in accordance with approved
anchor bolt plans. [1] The Code provides for a
+1/2-inch tolerance for the elevation of the top of
anchor bolts. The contractor setting the anchor
bolts should be able to meet this tolerance, but
errors can occur. Section 7.5 in the Commentary
on the Code of Standard Practice discusses the
installation of anchor bolts. [2]
Bol t Detailing. Anchor bolt detailing is discussed
in Chapter 7 of AISC Detailing forSteel Construc-
tion. [3] To match the minus 1/2-inch tolerance
noted in Section 7.5 of the Code, the steel detailer
should allow for at least a 1/2-inch projection of the
bolt above the top of the nut. If the anchor bolt is
set 1/2-inch Iow, the nut will still obtain full thread
engagement. However, when the minus 1/2-inch
tolerance is exceeded, the problem of a Iow
anchor bolt exists.
S o l u t i o n
Extending the Bol ts. Anchor bolts that are set
Iow are commonly called "short anchor bolts."
Short bolts need to be corrected by making them
longer. Two methods of making the bolts longer
are threaded couplers and welded extensions.
The "Steel Interchange" feature in Modem Steel
Construction, January 1993, and "Some Practical
Aspects of Column Base Selection," SteelDesign
Guide Series 1: Column Base Plates, discuss
these two methods. [4, 5] For either correction
method, the erector must work with the structural
designer (and general contractor). If the anchor
bolts are designed to resist uplift, in addition to
providing column stability during erection, then the
structural designer may require special proce-
dures. See AISC Manual of Steel Construction,
Specification J10, page 5-172, for loads on an-
chor bolts. [6]
Preventative Sol uti on. A"preventative" solution
that anticipates Iow anchor bolts is to design and
detail anchor bolts with additional bolt projection.
Examples include:
The structural designer shows a 1-inch bolt pro-
jection above the top of the nut in the base plate
details on the structural drawings. This 1-inch
bolt projection allows bolts to be set an additional
1/2-inch lower than the minus 1/2-inchsetting toler-
ance provided by the AISC Code of Standard
Practice, and still obtain full thread engagement.
The steel fabricator details anchor bolts with the
top of the bolt one bolt diameter above the top of
the nut. So for bolts larger than 1-inch diameter,
even more bolt projection is furnished than the
above example. For example, the detail of a 2-
inch diameter bolt will show the top of the bolt
detailed 2 inches above the top of the nut.
Full Thread Engagement. Short anchor bolts
that prevent full thread engagement can be a
frustrating problem. First, the question arises,
What is full thread engagement?. Section IIl.F in
Chapter 2 of AISC Quality Criteria and Inspection
Standards discusses full thread engagement for
high-strength bolts. [7] Section III.F refers to
Section 2(b) of the "Specification for Structural
Joints Using ASTM A325 or A490 Bolts," on page
5-265 of AISC Manual of Steel Construction. [6]
Section 2(b) states, "The length of bolts shall be
such that the end of the bolt will be flush with or
outside the face of the nut when properly in-
stalled." The same criteria could apply to nuts on
anchor bolts.
Second, what action is necessary if the top of the
bolt is just below the top of the nut? Instead of
lengthening the bolt, the nut might be welded to the
bolt by filling in the space between the top of the
bolt and the top of the nut with weld metal. How-
ever, welding the nut to the bolt is not always
allowed, particularly if high-strength, heat-treated
bolts and nuts are used, and the bolts are subject
to tensile loads. See "Steel Interchange" in the
December 1992, May 1993, and July 1993 issues
of Modern Steel Construction. [8, 9, 10] If the
erector can prove the '"fill-in" weld is adequate, the
structural designer may approve this welding pro-
cedure. But to provide proper column support, the
weld may need to be made before the lifting line is
released from the column.
5
Instead of welding the nut to the bolt, the erector
might consider limited air carbon arc gouging of
the base plate surface under the nut to provide full
thread engagement. This procedure must be
approved by the structural designer.
Practical Procedure. A practical procedure is for
the erector to review the as-built survey of the bolt
elevations before starting erection, determine bolts
that are set Iow, work with the contractor to resolve
how to correct them, decide who is to make the
corrections, and make corrections before erection
crews arrive at the jobsite.
2. Misplaced Anchor Bolts
Probl em
The erector discovers anchor bolts are:
Incorrectly spaced.
Located off the established column lines.
Tilted (out of plumb).
Bent over flat, damaged, or even broken off.
Installed with the bolt pattern rotated 90 degrees.
See the problem, "Rotated Anchor Bolt Pattern."
Installed to include anycombination of the above.
Installation Conditions. The installation of an-
chor bolts is not an easy task under the best of
conditions. If the foundation contractor has a firm,
level, dry, and uncongested job site, then the steel
erector will probably find properly installed anchor
bolts. But we all know most foundation sites are
not in the above listed condition. So misplaced
anchor bolts may be expected.
S o l u t i o n
Survey of Bolts. The first line of defense for the
steel erector against misplaced anchor bolts is to
review the as-built anchor bolt survey before steel
erection starts. Then the steel erector will know if
any corrective work is required, have the correc-
tive work performed before steel erection starts,
and not be faced with the frustration and delay
expense of correcting the bolts while erecting the
columns.
Setti ng Tol erances. Section 7.5 of the AISC
Code of Standard Practice specifies tolerances for
setting anchor bolts. [1] These tolerances ac-
knowledge that bolts will not be set exactly as
shown on the anchor bolt plan. To allow for
misplaced bolts, holes in the base plates or holes
in the framing angles from the columns to the base
plates are allowed to be made oversized. See
Table 6-1 on page 6-12, ManualofSteelConstruc-
tion, Vol. II, Connections. [11] For example, 23/4-
inch diameter holes are allowed for 13/4-inch diam-
eter bolts. Thus, the oversized holes will allow the
erector to overcome some misplacement of the
anchor bolts.
If the bolts are misplaced too much for the over-
sized holes to overcome, then corrective work
must be performed. The type of corrective work
depends on the function of the anchor bolts. All
anchor bolts serve to locate the columns and
prevent overturning of the columns during steel
erection. Some anchor bolts tie the column to the
foundation to resist uplift, overturning, and shear
from building design loads. The latter functions
may require more extensive corrective work for
misplaced bolts. In any event, inform the struc-
tural designer of the corrective work.
If bolts are misplaced up to 1/2inch, the oversized
base plate holes normally allow the base plate and
column to be placed near or on the column line.
For example, the 23/4-inch diameter base plate
hole for a 13/4-inch diameter anchor bolt allows for
a 1/2-inch adjustment of the base plate. If the bolts
are misplaced by more than 1/2 inch, then correc-
tive work is required.
Anchor Bolts Designed to Prevent Overturn-
ing of Column During Steel Erection. For
anchor bolts designed to prevent overturning of
the column during steel erection, corrective work
may include:
Slotting the base plate or column angle holes.
Fabricating a base plate to match the misplaced
bolts.
Fabricating an oversized base plate with stub
bolts welded to the base plate in the correct
location, and then welding the base plate to the
rotated bolts.
Making an "s" bend in the bolts. (But not too
sharp of a bend.)
Chipping away the concrete to make a larger "s"
bend.
Burning off the bolt and placing new expansion
bolts.
Burning off the bolt and welding a new bolt to the
side of the projecting stub.
An extensive discussion on misplaced anchor
bolts is given in "Some Practical Aspects of Col-
umn Selection," by David T. Ricker. [5] A discus-
sion on the design and use of column bases and
base plates is contained in Chapter 6 of the AISC
Manual of Steel Construction, Volume II, Connec-
tions. [11]
Another solution that anticipates anchor bolt mis-
placement is for the structural designer to detail
oversized holes in the base plates that are even
larger than the oversized holes allowed by Table
6-1 on page 6-12, Manual of Steel Construction,
VoL II, Connections. [11] Plate washers with bolt
holes 1/16 inch larger than the bolt diameter are
then welded to the base plate. This solution allows
additional tolerances in setting the anchor bolts.
The plate washer is placed between the top nut
and the top of the base plate, and is welded to the
base plate after the column is erected and aligned.
A bottom plate washer is required above the
bottom leveling nut. This bottom plate washer is
not really added material because it will also be
needed with the standard oversized holes. See
the following detail for anchor bolt, nut, and plate
details.
Anchor Bolts That Resist Uplift, Overturning,
and Shear. For anchor bolts designed to resist
uplift, overturning, and shear from building design
loads, corrective work may be limited to:
Slotting the base plate or column angle holes.
Fabricating a base plate to match the misplaced
bolts.
Chipping out the concrete, removing the mis-
placed bolts, and concreting in new, correctly
placed bolts (in the extreme case).
COLUMN
EXTRA OVERSIZE HOLE
IN BASE PLATE
WASHER
- C%OUT%_.-E-
CONCRETE
///
PLATE
B
GROUT FORM
x_ PLATE WASHER . NOT REQUIRED IF
/ / /
LEVELING NUT SHIMS ARE USED
/ / / , /
,ANCHOR
/ /
BASE PLATE DETAIL
Exercise caution before using this detail. If the
anchor bolts are designed to resist column shear
forces (see below), the anchor bolts must be
designed to resist bending because shear forces
to the bolts are applied at the plate washer--which
may be a few inches above the surface of the
concrete.
3 . R o t a t e d A n c h o rB o l t P a t t e r n
P r o b l e m
The erector discovers anchor bolts placed with the
anchor bolt pattern rotated 90 degrees from the
detailed orientation.
I I l
O 1 - - O
I I I
Detailed
Orientation
I I I
- - 0 - - + O--
- - + .. +--
- - O - - + - 0 - -
I I I
As-Built
Orientation
Sol ut i on
Uniform Spacing. One means to prevent rotated
anchor bolt patterns is to use uniform bolt spacing.
As stated by David T. Ricker in Steel TIPS, 'q'he
possibility of foundati on errors will be
reduced...when anchor bolt spacing is kept uni-
form throughout the job." [12] If a square anchor
bolt pattern is used, a rotated pattern cannot
occur. So the ultimate uniform spacing is to design
a square anchor bolt pattern--if possible.
Anchor bolt patterns that are rotated 90 degrees
may be corrected using the procedures listed for
misplaced anchor bolts.
S u r v e y of Bolts. When the anchor bolts are
surveyed before fabrication, the base plates may
possibly be fabricated to match the bolt spacing, or
the base plates may possibly be rotated on the
columns. Correction methods are discussed in
"Some Practical Aspects of Column Base Selec-
tion.'' [5]
Case H i s t o r y . On a 20-story building in San
Francisco, California, the steel erector surveyed
the as-built location of anchor bolts. The
contractor's superintendent, John, was an "old
timer" and took much pride in his work. He
carefully explained to the surveyor, with his fore-
men present, that he personally supervised the
anchor bolt installation. All the bolts were at the
correct elevation, were exactly spaced, and were
"right on" the column lines. After the survey was
complete, the surveyor reported the results to
John, with his foremen present. The surveyor
stated all the bolts were at the exact elevation,
correctly spaced, and "right on" the columns lines.
John smiled. But when the surveyor told him the
bolts on column lines B2 and B3 were rotated 90
degrees, his smile disappeared. And no matter
how he measured the bolts, they were still rotated
90 degrees.
9
4. Inadequate Anchor Bolts for Column Erection
Problem
After reviewing the anchor bolt and column base
design, the erector discovers the anchor bolt and
base plate design do not provide for adequate
resistance to overturning of columns during erec-
tion. This problem can occur when:
Only two anchor bolts are provided, leveling
plates are not used, and shims or wedges cannot
be placed under the base plate.
The structural designer or detailer has not made
provisions for the anchor bolts to resist lateral
forces on the free-standing columns.
S o l u t i o n
Overturni ng of Col umn. After the column is set
on a leveling plate, or on anchor bolt leveling nuts,
or on shims, or on a base plate, the anchor bolt and
base plate design must be capable of resisting
overturning caused by lateral forces on the col-
umn. The lateral forces may consist of wind, other
steel members striking the column, erection equip-
ment striking the Column, or even ironworker con-
nectors at the column top. Chapter 6, page 6-12,
in the AISC Manual of Steel Construction, Volume
II, Connections, mentions overturning due to acci-
dental collisions during erection. [11] Overturning
is also discussed in "Some Practical Aspects of
Column Selection." [5] Overturning is usually not
a problem when the anchor bolts and column base
are designed to resist overturning and uplift from
building design loads.
Prevent the Problem. The best method to pre-
vent the above problem is to perform proper plan-
ning for anchor bolt and base plate design. Proper
planning means:
The erector lets the steel detailer know what
lateral loads the column base design must resist.
A specified lateral load from any direction at the
column top is provided to handle wind or objects
striking the column.
The erector coordinates foundation construction
with the general contractor to make certain shims
may be placed under column bases with onlytwo
anchor bolts when leveling plates are not used.
The steel erector requests four-bolt anchor bolt
patterns when leveling plates are not used. The
column is then landed on four supporting leveling
nuts. Shims under the base plate may also be
added to help resist overturning.
If proper planning is not performed, the steel
erector may face a safety problem while erecting
the columns. The column may need to be guyed-
off before the lifting line is released. But guys also
present another safety hazard because guys are
not easy to see and something may run into or
strike the guy. Steel struts similar to tilt-up wall
struts may be used. Struts present a less hazard-
ous situation because they are easily seen and
take up less space.
Tall, unsupported columns may require an erec-
tion engineer to analyze the column base (anchor
bolts). For example, an airplane hangar had 90-
foot high columns with trusses at the top. The
column bases had multiple anchor bolts that tied
the column base to the foundation. The columns
were 30 inches deep and 12 inches wide. The bolt
design provided adequate support in the strong
direction of the column, but inadequate support in
the weak direction. The steel erector solved the
problem by erecting a column "bent" consisting of
two columns and the fill-in beams. This "bent"
gave adequate resistance in the weak direction.
The ironworkers still did not trust support in the
strong direction, so they added wire rope guys.
After all, the ironworkers had to be at the column
top to connect the trusses.
Case Hi story. Even with the proper column base
design, the steel erector must still be cautious. On
one industrial building, the owner scheduled a
small ceremony for the first column erected. The
column was set on four anchor bolt leveling nuts,
the top nuts were tightened, and the column was
then released from the lifting line. The column
promptly fell over because the column had only
been tack-welded to the base plate. What a way
to start? Needless to say, the steel erector made
a big impression at the ceremony.
5. One-Bolt Connections
Problem
While reviewing the design drawings, the erector
discovers the structural designer has provided a
connection with no bolts, or with only one bolt.
Code Requirements. The Construction Safety
Orders, Section 1710(c)(1 ), states:
During the final placing of solid web struc-
tural members, the load shall not be re-
leased from the hoisting line until the mem-
bers are secured with not less than two
bolts, or the equivalent at each connection
to keep members from rolling and to sustain
anticipated loads. Bolts shall be drawn up
wrench tight. [13]
The term "solid web structural member" is in-
tended to mean a beam, channel, girder, or even
a column standing vertically connected at one
end. The two bolts are required to keep the beam
from rolling and to sustain erection loads. Almost
all bolted members are designed with at least two
bolts just to take the design load. However, some
welded members may show no bolts.
Work Practices. Apart from the requirements of
Section 1710(c)(1), two bolts are also required to
allow the ironworker connector to "connect" the
beam in a safe, quick, and economical manner.
The ironworker will place the tapered shaft of a
spud wrench in one bolt hole, place a bolt in the
second bolt hole, and then be able to remove the
spud wrench shaft to place the second bolt, if the
second bolt is required. If only one bolt hole is
provided, the connector obviously cannot use that
bolt hole for both the connecting spud wrench and
a bolt. Certain steel members can be erected
without any bolts, or with only one bolt. However,
erection costs are increased because the member
must be held with the load line until the single bolt
can be placed, or, in the case of no bolts, a
temporary weld is made.
Tubes. Steel tubes, commonly used as bracing
members in braced frames, may be shown on the
design drawings without any erection bolts, or with
only one erection bolt. See the following two
details for examples of this situation.
t II
l i t
No Bolt
II
T IL
One Bolt
Solution
Provide for Two Bolts. The erector should make
provisions in its estimate for at least two bolt
connections on all members. During steel detail-
ing, the erector should coordinate with the fabrica-
tor, detailer, and structural designer to make pro-
visions for the required two bolts.
Tube Bracing. Steel tube bracing members
present special problems to the erector. A typical
tube bracing design provides for a slotted end to fit
over a gusset plate. The tube is then fillet-welded
to the plate. As mentioned under "Problem," the
design drawings may show no bolts, or only one
bolt to allow for erection of the tube.
The steel tube should not be subject to the provi-
sions of Section 1710(c)(1 ) because it is not a solid
web member. Further, the slotted ends will keep
the tube from rolling when the load line is released.
However, the ironworker connector still needs at
least two bolts at each end of the tube to safely
make the connection.
Erection Angles on Tubes. The following Detail
A shows how one erector solves the two-bolt
problem by using erection angles atthe ends of the
tube. The two bolts in the erection angle at the top
end of the tube allow the ironworker connector to
safely connect that end of the tube first. The two
slotted bolt holes in the connection angle at the
bottom of the tube allow that end of the tube to be
connected with a spud wrench. This method
presents the following problems:
Long slots in the tube are difficult to make in the
shop and difficult to fit-up and weld in the field.
The tube is required to be erected by first posi-
tioning the tube in the same vertical plane as the
gusset plates and then swinging it in into posi-
tion--a task not readily accomplished, if at all.
Panel geometry may not allow the tube to be
erected unless the tube angles, gusset plates,
and tube slots are specially shaped and the
bottom gusset plate is shipped loose.
Pl ates on Tubes. The following Detail B shows
how structural engineer William C. Honeck solves
both the two-bolt problem and difficult erection
problem by using plates shop-welded to the ends
of the tube. This method has the following advan-
tages:
The fabricator makes a block and short slot at
each end of the tube instead of the difficult long
slot.
The difficult positioning of the long slot to the
gusset plate is eliminated.
The tube is easier to erect and can always be
erected because it is simply brought in sideways.
Long Bolts. An alternate solution is to place two
erection bolts through the tube and gusset plate.
This solution has problems because when the
long bolts are tightened to fit up the slot to the
gusset plate, the tube sides may bend in.
Other Tube Connections. For other tube end
connections, see the article by Lawrence A. Kloiber
titled, "Designing Architecturally Exposed Steel
Tubes," in the March 1993 issue of Modern Steel
Construction. [14] However, the one-boltconnec-
tions illusl]'ated in that article are not recommended.
2
NOTES: 1. REQUEST APPROVAL FROM ENGINEER TO
LET ANGLES REMAIN IN PLACE.
2. THIS DETAIL IS MEANT TO ILLUSTRATE THE
USE OF ANGLES ON THE ENDS OF THE TUBE.
SEE COMMENTS IN THE "SOLUTION" FOR
PROBLEMS WITH THIS DETAIL,
k
Z ERECTION BOLTS
/ ,
-/
V l/
Z 3x3x 3, 8 WITH TWO
LONG SLOTTED HOLES IN ANGLE
DETAIL A
ANGLES ON ENDS OF TUBE - SLOT IN TUBE
13
THIS DETAIL PRODUCES A VERY SMALL ECCENTRICITY
THAT CAUSES BENDING IN THE BRACING MEMBER,
THIS BENDING SHOULD BE CONSIDERED IN THE
DESIGN OF THE BRACE.
REGULAR HOLES FOR
TWO ERECTION BOLTS
TUBE PL
OPTIONAL
FIELD OR
SHOP WELD
GUSSET PL
/
TUBE PLATE ON GL OF TUBE
AND WORK GL
LONG SLOI-FED HOLES IN TUBE PLATE
FOR TWO ERECTION BOLTS
DETAIL B
PLATES ON ENDS OF TUBE
14
6. Columns or Bents Tied Together With Non-Bolted Steel Joists
Problem S o l u t i o n
The design drawings show columns or bents (par-
tial steel frames) tied together with steel joists that
have welded end anchorages (no bolts). This
condition is unacceptable to the erector because:
The Construction Safety Orders, Secti on
1710(c)(3) states:
In steel framing, where bar joists are utilized,
and columns are not framed in a least two
directions with structural steel members, a
bar joist shall be field-bolted at columns to
provide lateral stability during construction.
[13]
The welded connection provides no fit-up for
spacing adjacent columns or frames.
The Steel Joist Institute (SJI) requires bolted end
anchorages for joists at column lines to provide
lateral stability during construction.
If the erector discovers column line joists with
welded end anchorages, the erector should:
Condition its bid for bolted end anchorages.
Work with the detailer, fabricator, joist supplier,
and structural designer to provide bolted end
anchorages.
If for some reason the column line joists are
delivered to the jobsite without bolted end anchor-
ages, the erector must provide the required bolt
holes in the field.
The SJI Standard Specifications Load Tables and
Weight Tables for Steel Joists and Joist Girders,
and Technical Digest, No. 9, Handling and Erec-
tion of Steel Joists and Joist Girders are must
references for joist design, fabrication, and erec-
tion. [15, 16]
15
7. Steel Joists Without Bolted Bridging
Problem
Open web steel joists are furnished without bolted
bridging required for proper and safe erection.
Code Requirements. The Construction Safety
Orders, Section 1710(c)(4) states:
Where Iongspan joists or trusses, 40 feet or
longer, are used rows of bridging shall be
installed to provide lateral stability during
construction prior to slacking of hoisting
line. [13]
Industry Procedures. The Steel Joist Institute's
(SJI) Standard Specifications Load Tables and
Weight Tables for Steel Joists and Joist Girders
gives various requirements for erecting joists. [15]
For example, Section 6, "Handling and Erection,"
for K-Series steel joists requires bolted diagonal
bridging to be installed on certain joists before the
hoisting cables are released. The SJI Technical
Digest, No. 9, Handling and Erection of Steel
Joists andJoist Girders, also discusses stability of
joists and required bolted bridging. [16]
S o l u t i o n
Joi st Design. The structural designer must be
cautious when designing steel joists or using pre-
engineered joists. If the designer shows bridging
details, then care must be taken to follow the
handling and erection requirements of the Steel
Joist Institute. The Institute's requirements meet
the requirements of the Construction Safety Or-
ders.
The erector should review the design drawings
and work with the fabricator and joist supplier to
make certain that the required bolted bridging is
furnished.
Assembl e Joi sts. The erector can assemble
groups of joists on the ground, complete with
bridging, and erect the assembled group to stand
alone as a laterally stable unit. This method of
erecting joists also solves the problem of erectors
working on highly unstable joists. Section 6,
"Handling and Erection," in Reference 15, also
states:
When it is necessary for the erector to climb
on the joists to install the bridging, extreme
caution must be exercised since unbridged
joists may exhibit some degree of instability
under the erector's weight.
Case Hi story. On one project, a metal deck
foreman happened to walk on the top chord of a
newly erected joist that had no bridging installed.
The joist moved laterally and the foreman fell off.
The joist erector was following proper erection
procedures, and had reviewed those procedures
with the metal deck contractor. The foreman had
a momentary lapse of safety procedures. This
example illustrates that the required joist erection
procedures are not to be taken lightly by the
structural designer or erector.
16
8. Columns or Bents Tied in With Timber
Problem
The structural designer produces a building de-
sign that uses a combination of timber beams and
steel bents (partial steel frames) in order to reduce
costs. The timber beams tie the steel bents
together. The combined frame is usually laterally
stabilized by horizontal and vertical plywood dia-
phragms in the timber direction.
Erecti on Supports. The steel erector has the
problem of determining how to temporarily sup-
port the steel bents. The steel framing is obviously
a non-self-supporting steel frame as specified in
Section 7.9, 'q'emporary Support of Structural
Steel Frames," in the AISC Code of Standard
Practice. [1] The erector must furnish adequate
temporary supports as required by the Code. The
erector is also governed by Section 1710(a), "Brac-
ing,'' of the Construction Safety Orders. [13]
Solution
Designate in Contract. First of all, the structural
designer must realize the problems inherent in a
combination design of steel frames and timber tie-
in beams. Section 7.9.3 in the Code of Standard
Practice states in part, "Such frames shall be
clearly designated as 'non-self-supporting.'" [1] If
the structural designer does not make that state-
ment in the contract documents, then the steel
erector may make a claim against the owner.
All-Steel Frame. One solution to the temporary
support problem is for the steel erector to ap-
proach the fabricator, contractor, and structural
designer to replace the timber beams on the
column lines with steel beams. Then, at least the
erector will have an all-steel frame that will be
easier and safer to temporarily support. Ofcourse,
the best solution from the steel industry's view-
point is to ask the structural designer to replace all
the timber beams with steel beams.
Support Methods. If the structural designer
cannot change or modify the design, then the steel
erector is faced with the problem of determining
how to erect the steel and furnish temporary
supports that provide the required lateral stability
with the least hazardous working conditions. Any
method the erector chooses to erect the steel and
timber will present greater safety hazards than the
hazards in erecting an all-steel frame.
Some methods the erector can follow are to:
1. Erect the steel bents supported in all directions
and then leave the jobsite. This solution presents
a hazardous condition because other trades might
run into or remove the supports---especially if wire
rope guys are used. Temporary horizontal steel
struts between the steel bents will allow the use of
less hazardous wire rope "X" bracing in lieu of the
undesirable wire rope guys.
2. Work with the carpenters and erect the steel
concurrently with the timber beams. This method
presents the hazards of two trades working to-
gether, and one relying on the other--not the best
of conditions. Temporary supports will still be
required, and the ironworkers and other trades will
probably not end the project on the best of terms.
3. Use a combination of methods 1 and 2.
Case Hi story. On a recent project, a combination
steel bent and timber beam structure with four
levels of steel was used. The erector chose
method I above--erect the steel, guy it off, and
leave the jobsite. The bents were supported with
wire rope "X" bracing in the steel frame direction
and wire rope guys in the timber beam direction. In
the timberbeam direction, the columns were guyed-
off at three floor levels to anchors in the concrete
basement floor. Guys at the third level were so
steep, their ability to prevent lateral displacement
was questionable. Fortunately, the frame did not
collapse. However, the carpenters had to con-
stantly make adjustments to the plywood dia-
phragms in order to keep the building plumb. The
question might be asked, "Would an all-steel frame
have been more efficient and economical?"
9. Steel Columns or Partial Bents Not Tied In
Probl em
sign that uses a combination of steel and other
building materials. Steel columns may be com-
pletely tied in by timber or concrete, or partial steel
bents may be tied in by concrete. A variety of
designs may exist, but all of them require tempo-
rary supports by the steel erector.
Si mi l ar Problem. This problem is similar to the
problem, "Columns or Bents Tied in With Timber."
But this problem presents a more hazardous con-
struction condition because the steel is, for the
most part, unsupported free-standing columns
with an irregular steel beam pattern.
Erection Supports. The steel erector has the
problem of determining how to temporarily sup-
port the steel members. The steel members are
obviously a non-self-supporting steel frame as
specified in Section 7.9, "Temporary Support of
Structural Steel Frames," in the AISC Code of
Standard Practice. [1] The erector must furnish
adequate temporary supports as required by the
Code. The erector is also governed by Section
1710(a), "Bracing," of the Construction Safety
Orders. [13]
Sol uti on
Designate in Contract. The structural designer
must realize the problems inherent in a combina-
tion design of steel and other materials. Section
7.9.3 in the Code of Standard Practice states in
part, "Such frames shall be clearly designated as
_non-self-supporting._" [1] Although the steel
members are obviously non-self-supporting, the
structural designer must make a statement in the
contract documents that the frames are non-self-
supporting, or the owner may be subject to a claim.
Hazardous Methods. The temporary supports
determined by the steel erector will present a
varying degree of safety hazards depending on
the type of supports. One method of temporary
support is to guy-off the columns with wire rope
guys that are anchored to the concrete floor or
concrete footings. This solution presents an ex-
tremely hazardous condition because the wire
rope guys will interfere with construction opera-
tions of the steel erector and the othertrades. If the
wire rope guy is struck by construction equipment
or materials being hoisted, or if the wire rope guy
is accidentally slacked-off by a worker who thinks,
"It is in the way," a disastrous accident can occur.
Such an accident did occur on a high-rise building
in Toronto, Canada, when a wire rope guywas cut
by another trade because it was in the way. Wire
rope guys are also subject to a multitude of prob-
lems that must be constantly monitored. For
example, the wire rope clamps must be properly
placed and checked to make certain they have not
been loosened. Turnbuckles must also be con-
stantly observed to make certain they have not
been slacked-off or tampered with. Wire rope
guys may be the most economical and easiest
type of temporary support to install, but they present
the most hazardous safety condition.
Another safer, temporary support is to provide
rigid struts from the steel members to the concrete
floor or footings. Struts are more visible than wire
rope guys and can take more physical abuse.
Case History. The steel erector should take
advantage of adjacent existing structures to stabi-
lize the steel being erected. For example, on one
project 200-foot-long trusses were erected around
three sides of an existing hangar. On two sides of
the hangar the new columns were temporarily
braced to the existing columns with angle frames.
These frames:
Stabilized the long free-standing columns.
Located the columns for vertical alignment.
Stabilized the truss bents until bottom chord
members could be connected.
No wire rope guys were required, which made the
steel erector and the contractor very happy.
18
10. Non-Self-Supporting Steel Frames
Problem
sign where the completed steel frame is not stable.
Section 7.9.3 in the AISC Code of Standard Prac-
tice defines this type of steel frame as a non-self-
supporting steel frame. [1] The AISC definition is:
A non-self-supporting steel frame is one
that, when fully assembled and connected,
requires interaction with other elements not
classified as Structural Steel to provide sta-
bility and strength to resist loads for which
the frame is designed.
Desi gnate in Contract. Such frames are re-
quired to be clearly designated as "non-self-sup-
porting" in the contract documents. The Code of
Standard Practice defines contract documents to
mean the contract, plans, and specifications. The
structural designer must convey the "non-self-
supporting" designation, preferably on the struc-
tural drawings (plans). If the structural designer
does not make such a designation on the draw-
ings, then the owner may receive claims for extra
work from the steel erector and contractor. If the
drawings are not so designated and the steel
erector does not realize the non-self-supporting
condition, and if a construction failure occurs, then
the structural designer may wish the steel frame
had been designed as a self-supporting frame.
New Code. The AISC recently issued a new
version of the Code of Standard Practice, effective
June 10, 1992. This version replaces the Septem-
ber 1, 1986 version contained in the Manual of
Steel Construction. [6] Significant changes are
made to Section 7.9.3, "Non-Self-Supporting Steel
Frames." Hopefully, these changes will alleviate
controversies that resulted from varied interpreta-
tions of language in the September 1, 1986 ver-
sion.
Erector Furnishes Supports. The steel erector
is required to furnish and install temporary sup-
ports for the erection operation for both self-sup-
porting and non-self-supporting steel frames. See
Section 1710 of the Construction Safety Orders,
and Section 7.9 of the AISC Code of Standard
Practice. [13, 1] For many erectors, furnishing
temporary supports for self-supporting frames is a
difficult task. Furnishing supports for a non-self-
supporting frame may tax the resources of the
erector. Then if the non-self-supporting frame is
not designated as such in the contract documents,
and the erector does not realize this condition until
work is started, the erector may have extreme
difficulty in erecting the frame.
S o l u t i o n
Designate on Drawi ngs. The most obvious
solution, and the course of action required by steel
industry practice, is for the structural designer to
designate non-self-supporting frames in the con-
tract documents. See page 26 of "Structural Steel
Construction in the '90s," in Steel TIPS. [17] If the
"non-self-supporting" designation is made on the
drawings, erectors will be able to determine during
the bidding or negotiating period if they can cope
with the problems presented by such frames.
Analyze Frames. As a second line of defense, the
erector might be wise to use the services of an
erection engineer to analyze any suspicious-look-
ing frames. Even the most experienced erectors
may miss the fact that a frame is non-self-support-
ing when that designation is not made in the
contract documents. Section 1710(b) in the Con-
struction Safety Orders, requires a civil engineer
currently registered in California to prepare an
erection plan for trusses and beams over 25 feet
long. [13] Hopefully, the engineer would discover
that the frame is non-self-supporting.
Examples. Some examples of non-self-support-
ing frames are:
Concrete shear walls that attach to a non-mo-
ment steel frame--after the steel is erected.
Column line beams that need metal deck for
lateral support to carry axial or vertical loads--
and the deck is not in place.
Floor framing that needs metal deck to transfer
horizontal l oads--and the metal deck is not in
place.
Roof trusses that help provide lateral stability by
frame acti on--but the bottom chords cannot be
connected until alloof loads are applied.
Tilt-up walls attached to the non-self-supporting
steel framo and the walls have no lateral sup-
port. See Section 7.9.3 in the Commentary on
the Code of Standard Practice. [2]
19
11. Column Splices Too Low or Too High Above Floor
Problem
On one tier building, the column splices are de-
signed at 6'-0" above the top of steel. On another
tier building, the splices are designed at 3'-6"
above the top of steel. The 6'-0" splices are too
high to allow the connectors, bolters, and welders
to work without scaffolding or floats. The 3'-6"
splices are not high enough to allow safety wire
rope attachments for exposed floor edges at the
periphery of the building or at interior floor open-
ings.
S o l u t i o n
Splice Design. Desi gnthecol umnspl i ces atl east
4'-0" above the top of steel. This height will:
Allow attachments for the top safety wire rope to
be placed on the column. The attachment for the
top wire rope needs to be located to provide the
correct height for both the erector and contractor,
if the contractor wants to use the wire rope
installed by the erector without moving the wire
rope.
Column attachments for the safety wire rope need
to be placed so the wire rope is located between 42
and 45 inches above design finish floor height as
required by Section 1710(e)(3) of the Construction
Safety Orders. [13] The 4'-0" splice meets this
height requirement for most cases. The determin-
ing factor is the floor thickness. If the floor is too
thick, the height of the col umn splice should be
increased. The 4' -0' height is recommended in
Chapter 6, page 6-19, of the Manual of Steel
Construction, Volume II, Connections, and by
Barry L. Barger in "What Design Engineers Can
Do to Reduce Fabrication Costs." [11, 12]
Allow the erectors, bolters, and welders to work *
without scaffolding or floats. The article, "Value
Engineering and Steel Economy," by David T.
Ricker, in Steel TIPS, discusses splices that are
too high. [18]
Provide for uniformity in shipping, unloading,
sorting, and erecting columns. If column splices
are designed at different heights above the floor
elevation on the same floor, or are designed with
the same tiers spliced at different floors, then
erection costs will increase.
Erector Requests. The structural designer should
consider requests from the erector to increase or
decrease the designed column splice heights.
20
12. Columns Interrupted by Beams
Problem
On a two-floor shopping center building, the col-
umns, rather than being one continuous piece
from the base plate to the roof, are interrupted by
the beam framing. The structural designer has
used the interrupted-column-framing system to
utilize continuous, supported beams.
Suspended Beams. In addition to thecontinuous
beams, the design utilizes cantilevers with sus-
pended beams between the two cantilevers. See
elevation sketch below. This type of design in-
creases erection and plumbing costs even more
than just continuous beams because the bent
units must be plumbed individually to allow the
suspended beams to be erected.
More Di ffi cul t Erection. The interrupted columns will
make steel erection more difficult and more costly
because:
I SUSPENDED
ii ii/ 1
II II
U H
BEAMS
I I
Elevation
Vibration. Continuous beam framing, especially
with cantilevers, may produce a design with ex-
cessive vibration. This vibration is not really an
erection problem, but the ironworkers will
notice and comment on the vibration. And
I surely, if the ironworkers feel the vibration,
the tenants will also feel the vibration.
Solution
In this particular case, the steel erector
must work with the design presented.
However, on future projects the structural
designer should realize the interrupted-
column design will increase the erector's
cost. Any savings visualized by using
continuous beams may be negated by the
More pieces are required to be erected.
The columns are more difficult to plumb and keep
plumb.
The complete frame is more difficult to plumb.
The sequence of and direction of erection may
be limited.
increased erection cost. The structural designer
may want to perform a brief value engineering
exercise on using full-length columns versus inter-
rupted columns.
21
13. Columns Offset From Beam Framing
Problem
On a tier building, some of the columns are offset
from the beam framing grid line. See sketch
below. This offset will present erecting and deck-
ing problems to the steel erector.
OFFSET
mB
T
J
T
F F
FLOOR PLAN
Solution
If possible, the structural designer should arrange
the framing so the columns and beams are tied
together on the main column lines without offsets.
Keeping the framing on common column lines
allows for more efficient loading, erecting, and
decking procedures. Also, lateral loads from erec-
tion equipment are more easilytransmitted through
the floor framing system.
22
14. Revisions and Alternates Not Flagged on Drawings
Problem Solution
Design drawings are issued without revisions high-
lighted, marked, or flagged to clearly indicate the
revisions. The fabricator and erector do not notice
the revisions. During construction, the structural
designer, contractor, or owner asks, "Why is that
door framing there?," or in an extreme case, "Isn't
the weld on those box columns too small?"
or
During bidding, the bid form and specifications
request and describe alternates, but the drawings
do not clearly indicate the alternates. As a result,
the fabricator and erector miss the scope of an
alternate. During construction, the structural de-
signer, contractor, or owner asks, "Where is eleva-
tor No. 6 going to fit?."
Indicate on Drawi ngs. The structural designer
must clearly indicate revisions and alternates on
the design drawings by:
Using the standard symbol for a revision.
Placing a "cloud" around the revision or alter-
nate, and identifying the cloud with the revision
symbol or alternate number.
Using some other highlighting or flagging method
to show the revision or alternate.
By Fabricator and Erector. The fabricator and
erector must follow the above practice whenever
they make revisions to their shop, erection, and
erection scheme drawings.
Flagging. "Flagging" revisions is discussed by
Bob Petroski, Eugene Miller, and David T. Ricker
in the article, "What Design Engineers Can Do to
Reduce Fabrication Costs," in Steel TIPS. [12]
23
15. Double-Framed Beam Connections to Girder
P r o b l e m
If two opposing beams, each wi th double frami ng
angl es, connect to the same gi rder and share
common bol t holes, an erection safety hazard
exists. Thi s type of connecti on is shown in Detail
A below. Detail A---4 is shown on page 5 of
L
DOUBLE FRAMING ANGLES
SHOP AND FIELD BOLTED
, , , A - - 4 - - .
Relative Cost
1.05
Whereas the previous connections of this series have employed
single shearelements, A-4 is the standard connection consisting of
double framing angles which are both shop and field bolted. The
Relave Cost Index of A-4 is 5% above the single tab shear base
connection A-l, but large beam loading could influence the eco-
nomics and use this connection relative to A-1 because of an
increase in weld size. There is a safety hazard in erection when
using this connection. Placing pins and bolts while tnjing to align
two opposing beams through common holes may require the addi-
tion of seat angles on one or both sides of the girder to keep the
beam in position. Eliminating this hazard, as required by OSHA
laws, will add additional cost to this connection.
Steel Connections/Details and Relative Costs.
[19] Note: Both the detail and compl ete accompa-
nying comments are shown. Portions of the com-
ments regard relative costs for both shop and field,
may refer to a detail sequence, and may not apply
to the subj ect matter of this probl em. However, the
compl ete comments are shown because the rela-
tive cost s should be of interest to most readers.
Thi s note al so applies to details in subsequent
probl ems that are taken from Steel Connections/
Details and Relative Costs.
The hazard exi sts because the i ronworker must
remove the bol ts from the first beam connected, in
order to connect the second beam. Once the bolts
are removed, the connecti on no longer compl i es
wi th the requi rements of Secti on 1710(c)(1 ) of the
Construction Safely Orders. [13] This secti on
requires each end of a beam to be secured wi th not
less than two bolts before the hoi sti ng line is
released.
S o l u t i o n
To avoid this hazardous connecti on, design the
connecti on as shown in Detail A--1 on page 4 of
Steel Connections/Details and Relative Costs.
[19] Connecti on Detail A - - l , shown below, uses
single shear tabs (plates) shop-wel ded to the
carrying gi rder and fi el d-bol ted to the beams.
. - . f , .
t
SHOP WELDED TAB-FIELD H.S. BOLTED
= . . A - - 1 . . ,
Relative Cost
1.00
Connection A-1 is the most economical for this series of shear con-
nections and is assigned a Relative Cost Indexof 1.00. This connec-
tion employs a single shear tab shop welded to the carrying girder
and field bblted to the beam.
Detail A--1 provi des a more economi cal connec-
tion than Detail A--4, because it provi des for safer
erection, faster erection, and better i ronworker
morale. As stated byW. A. Thornton in Steel TIPS,
"As this example illustrates, single angles will work
even in heavy industrial appl i cati ons, and they are
much less expensive than double angl es, espe-
cially for erection." [20]
24
16. Double-Framed Beam Connections to Column Web
Problem
If two opposing beams, each with double framing
angles, connect to the same column web and
share common bolt holes, an erection safety haz-
ard exists. Additionally, beam erection and bolt
access is difficult. This type of connection is
shown in Details BW--4 and BW--5 below. These
details are shown on page 9 of SteelConnections/
Details and Relative Costs (Steel TIPS). [19]
1 RETURN
1
SHOP WELDED ANGLES TO BEAM
H.S. BOLTED TO COLUMN WEB
DOUBLE ANGLES
SHOP AND FIELD H.S. BOLTED
--- BW- - 4- - - ,,-,,BW--5 = =
Relative Cost Relative Cost
1.20 1.30
Double angle connections BW-4 and BW-5 have relative costs of 1.20 and 1.30. The shop-welded angles are
slightly less. Installation of these connections is hazardous because of the difficulty in placing pins or erection
bolts through common holes. Addition of angle seats under the beams may be necessary to keep the beams from
falling. The relative costs of BW-4 and BW-5 will then be even higher than those noted. Use of connections BW-
4 and BW-5 may not be possible at columns with moment connections to the flanges because continuity plates
or stiffeners, as shown in the "CF" series connections, would interfere with entry of the beam. A design engineer
may wish to use connections similar to BW-1 or BW-2 to avoid this problem as well as to take advantage of the
obvious economies.
25
The hazard exists because the ironworker must
remove the bolts from the first beam connected in
order to connect the second beam. Once the bolts
are removed, the connection no longer complies
with Section 1710(c)(1) of the Construction Safety
Orders. [13] This section requires each end of a
beam to be secured with not less than two bolts
before the hoisting line is released.
S o l u t i o n
To avoid this hazardous connection, design the
connection as shown in Detail BW--1 on page 8 of
Steel Connections/Details and Relative Costs(Steel
TIPS). [19] Connection Detail BW--1, shown be-
low, uses single shear tabs (plates) and horizontal
stiffener plates shop-welded to the column with the
single shear plates field-bolted to the beams.
Detail BW--1 provides a more economical connec-
tion than Details BW- - 4 or BW--5, because it
provides for safer erection, faster and easier erec-
tion, easy bolt accessibility, and better ironworker
morale.
COL. FLG.
SHOP WELDED TAB AND PLATES
FIELD H.S. BOLTED
, , ,BW--1 , , ,
Relative Cost
1.00
For simple shear connection to column web the base 1.00 index
connection BW-1 has a single vertical plate welded to the column
web with horizontal stiffener plates (normally 1/2" thick) welded at
its top and bottom. The bolt holes are located outside the toe of the
column flanges, which allows for easy erection entry of the beam as
well as accessibility for impacting the high strength bolts.
26
17. Mixed Bolts
Problem
The structural drawings show a mixed "bag of
bolts" throughout the structure. Different kinds of
bolts shown include:
A325 bearing bolts in single-plate shear connec-
tions for the connections of fill-in beams. Some
of the connections require a snug-tight condition
of the bolts to prevent moment transfer. Other
connections allow snug-tight or fully-tightened
bolts.
A325 slip-critical bolts and A490 slip-critical bolts
for beam-to-column web connections at the same
column work point.
A325 slip-critical bolts and A490 slip-critical bolts
for bracing connections at the same work point.
Bolt Desi gn on Most Proj ects. The majority of
projects are designed with only one kind of bolt--
fully-tightened A325 bolts. Designing for different
kinds of bolts requires additional quality control,
with resulting added cost, to prevent the erector
from installing the wrong kind of bolt. Additional
quality control includes the following actions:
The fabricator (or erector) must prepare an erec-
tion drawing that shows--i n addition to the bolt
diameter and length---whether the bolt is A325,
A490, slip-critical (fully-tightened), bearing (snug-
tight or fully-tightened), or bearing (snug-tight
only).
The erector must not only distribute A325 and
A490 bolts of the correct diameter and length to
the work points, but must make certain that the
A325 and A490 bolts are installed in the correct
connection at a work point.
The erector must set up procedures and checks
to make certain that bearing bolts required to be
snug-tight are not accidentally fully-tightened.
The inspector must set up procedures to deter-
mine that each kind of bolt is properly installed
and tightened.
S o l u t i o n
Bolt Desi gn. The possibility of the erector using
the wrong kind of bolt can be reduced or elimi-
nated, and costs reduced, if:
A325 and A490 bolts are not used at the same
connection point.
The use of A490 bolts is limited to similar connec-
tions throughout the job--say all 36-inch-deep
beams, or all bracing connections.
The single-plate shear connections are designed
to allow fully-tightened bearing bolts. This design
assumes some transfer of moment is allowed.
Fully-tightened bolts should be allowed because
allowable loads for single-plate shear connec-
tions as tabulated in Table X on page 4-52 in the
Manual of Steel Construction are based on fully-
tightened or snug-tight bearing bolts. [6]
Steel erectors have discovered that the difference
in cost of installing snug-tight bearing bolts, fully-
tightened bearing bolts, and slip-critical bolts (fully-
tightened) is not distinguishable. Let's look at the
installation requirements for bearing bolts. Sec-
tion 8(c), "Joint Assembly and Tightening of Shear/
Bearing Connections," in the "Specification for
Structural Joints Using ASTM A325 or A490 Bolts,"
in the Manual of Steel Construction states:
Bolts in connecti ons.., shall be installed in
properly aligned holes, but need only be
tightened to the snug tight condition. The
snug tight condition is defined as the tight-
ness that exists when all plies in a joint are
in firm contact. This may be attained by a
few impacts of an impact wrench or the full
effort of a man using an ordinary spud
wrench. [6]
So after figuring out which bolts are bearing bolts,
the ironworker now has a choice of using the full
effort of a spud wrench or a few impacts of an
impact wrench. The choice is obvious. The
ironworker will use the impact wrench, and prob-
ably fully tighten the bolts, whether or not the bolts
need full tightening.
Bolt Uniformity. As stated by David T. Ricker, in
"What Design Engineers Can Do to Reduce Fab-
rication Costs," in Steel TIPS:
Bolt Uniformity. Minimizing the number of
diameters and types of bolts on a given job
lessens the chance for a mixup in the shop
or field... [12]
27
18. Mixed Diameter Bolts
Problem
The structural drawings show various bolt diam-
eters. The different diameters of bolts increases
the chance for the wrong bolts to be supplied or
installed. Additionally, installation cost is increased
due to added quality control, more supervision,
more tools, and tool changes.
Solution
Bol t Uniformity. The structural designer should
be aware that different diameters of bolts will add
to the fabrication and bolting cost. As stated by
David T. Ricker in "What Design Engineers Can
Do to Reduce Fabrication Costs," in Steel TIPS:
Bolt Uniformity. Minimizing the number of
diameters and types of bolts on a given job
lessens the chance for a mixup in the shop
or field and allows more efficiency in drilling
or punching operations. [12]
Minimize Number of Diameters. If structural
drawings require several diameters of bolts, the
erector should work with the fabricator and struc-
tural designer to minimize the number of diam-
eters to be used. For example:
Replace large-diameter machine boits with
smaller-diameter A325 bolts to match other A325
bolt diameters.
Keep the A325 bolt diameters the same by using
either more or less bolts.
Limit the number of bolt diameters. Instead of
using 3/4-inch, 7/8-inch, 1-inch, and 11/B-inch diam-
eters, try to use just 7/8-inch and 1-inch diameters.
Avoid using large-diameter A490 bolts. 13/8-inch
and 11/2-inch diameter A490 bolts require bigger,
heavier, and more costly equipment to tighten
the bolts. Some erectors do not have this equip-
ment. Further, the ironworkers certainly don't
like to use the heavy equipment.
19. Reuse of High-Strength Bolts
Problem
To correct alignment of exterior beams connected
with A325 slip-critical bolts, the erector loosens
and retightens some bolts and loosens, removes,
reinstalls, and retightens other bolts. However,
the inspector and engineer claim that retightening
the bolts constitutes reuse of the bolts, and they
request that the bolts be replaced.
Solution
Reuse of A325 Bolts. The "Specification for
Structural Joints Using ASTM A325 or A490 Bolts"
(Specification) in Part 5 of the Manual of Steel
Construction, prohibits the reuse of A490 bolts
and galvanized A325 bolts, but allows the reuse of
other A325 bolts, if approved by the responsible
engineer. [6]
The steel erector should bring to the attention of
the inspector and engineer Section 8(e), page 5-
276, "Reuse of Bolts," in the Specification. The
Specification, along with the AISC recommenda-
tions on page 17 in Quality Criteria andlnspection
Standards (AISC publication S323), should allow
the erector to obtain approval from the engineer
for the reuse of A325 bolts. [6, 7]
The "Steel Interchange" feature in Modern Steel
Construction, March 1992, contains an excellent
discussion on the reuse of non-galvanized A325
bolts. [21]
28
20. Prequalified and Non-Prequalified Weld Joints
Problem
Both prequalified weld joints and non-prequalified
weld joints are used in the structure. Either the
structural designer designs a connection with a
non-prequalified weld joint that requires a quali-
fied-by-test weld joint, or the erector decides that,
for cost considerations, a qualified-by-test joint is
more appropriate than a prequalified weld joint.
When problems occur using the qualified-by-test
joint, or even the prequalified joint, a finger-point-
ing contest is sometimes generated, and correc-
tive action is required.
Solution
By Welding Code. The article, "Welded Joints -
Requirements," in Part 4, page 4-152, of the
Manual of Steel Construction states in part:
AWS prequalification of a weld joint is based
upon experience that sound weld metal with
appropriate mechanical properties can be
deposited, provided work is performed in
accordance with all applicable provisions of
the Structural Welding Code. [6]
Design with Prequalified Joints. So the first
essential step for a sound weld is to design con-
nections that can use prequalified weld joints.
These joints are shown in Part 4 of the Manual of
Steel Construction, and in Section 2 of the Struc-
tural Welding Code. [6, 22]
Use Prequal i fi ed Joints. The second essential
step is for the steel erector to use prequalified weld
joints at the connections, and to follow all the
required procedures.
Qual i fi ed-By-Test Joints. If prequalified weld
joints are not used, either by necessity or by
choice, the third essential step is to use a qualified-
by-test weld joint. The AWS Structural Welding
Code sets forth the requirements for testing and
qualifying non-prequalified weld joints. [22]
Take Precautions. The fourth essential step
requires the steel erector to take precautions while
welding, and not take the attitude that a qualified
joint--prequalified or non-prequalified--will pro-
duce a successful weld. As further stated in the
article quoted above, a successful weld also re-
quires attention to:
The magnitude, type, and distribution of forces to
be transmitted.
Accessibility.
Restraint to weld metal contraction. See the
problem, "Restrained Welded Joints."
Thickness of connected material.
Effect of residual welding stresses on connected
material.
Distortion.
The articles, "Avoiding Weld Defects," "Correcting
Weld Defects," "Nondestructive Testing (NDT),"
and "Projects Specifications," contained in "Struc-
tural Steel Construction in the '90s," in Steel TIPS,
contain much valuable information on producing
successful welds. [17]
Welding Procedure. The essential step, and
one that is often overlooked, is for the erector to
produce a complete and comprehensive welding
procedure for each project. The welding proce-
dure should include:
A weld sequence for both the complete frame
and the individual joint. The joint sequence
should include when beam-to-column web joints
are tightened, if the webs are bolted.
The prequalified and qualified-by-test joint weld-
ing procedures.
A requirement that only certified welders may be
used, and that they must be certified for the
process used and the weld position. A weld joint
may be properly designed, be prequalified, and
be thoroughly planned, but the success of the
weld produced depends on a certified and dedi-
cated ironworker making the weld.
29
21. Extending Continuity Plate for Back-up Bar
Problem
In certain beam-to-column web welded moment
connections, the back-up bar for the flange weld
fouls on the column flanges.
Solution
Plate Design. To provide adequate clearances
for back-up bars, design the connection with con-
tinuity plates extended beyond the column flanges.
See Detail DW--1, on page 12 of Steel Connec-
tions/Details and Relative Costs (Steel TIPS). [19]
Detail DW--1 and "Note" also discuss correct
welding of the continuity plate. For convenience,
a modified Detail DW--1 is shown below.
Extending the continuity plate is also recommended
on pages 4-11 and 6-55 in the AISC Manual of
Steel Construction, Volume II, Connections. [11]
Fabricator or Erector Requests. Ifthe structural
designer has not provided for an extended conti-
nuity plate, the fabricator or erector will probably
request the plate to be extended. The structural
designer should grant that request.
T&B
/
-T-
tE
FLANGE >
tx,
II
- - EXTENSION
WEB BOLTED - FLANGE BUTT WELDED
30
22. Welded Connections to Inside of Column
P r o b l e m
The structural drawings show beam-to-column
web connections made with field welds inside the
column flange areas. See Details BW--3 and
DW--4 below. Some of these welds are difficult to
make because of electrode positioning, equip-
ment access, welder access, and welder visibility.
NON-MOMENT CONNECTION M O M E N TC O N N E C T I O N
I t l
I
I
OPTIONAL
SHOP WELDED SEAT- FIELD WELDED TO BEAM
- - = B W - - 3 = = =
Relative Cost
1.09
The extra connection pieces as well as the drilling of holes through
the beam flange add to the cost of this connection. If the column
has moment connections to its flange with column stiffeners, the
use of this connection may be prohibited as in the cases of connec-
tions BW-4 and BW-5.
1" TYP.
, . - 1 IF REQ'D.
EA. SIDE
TO COL. FLG.
- NO TAPER
OPTIONAL
TRIM LINE
WELDED MOMENT PLATES WITH SEAT
. . . D W - - 4 . . =
Relative Cost
1.50
Connection 0W-4, which is all welded, is not popu-
lar because of its high relative cost compared with
the first two connections in this Plate prepa-
ration and the full penelTation welding of flange
plates to the column results in an increase of the
relative cost to 50% over base connection 0W-1.
31
Solution
To avoi d the above problems, make the beam-to-
column web connecti ons as shown in Detai l s
B W - - 1 and D W - - 1 below. The fabri cator and
erector should work with the structural designer to
change the undesirable detai l s to the desi rabl e
detai l s.
N O N - M O M E N T C O N N E C T I O N M O M E N T C O N N E C T I O N
) COL. FLG.
SHOP WELDED TAB AND PLATES
FIELD H.S. BOLTED
,,,,. BW--1 ,,-,,
Relative Cost
1.00
For simple shear connecUon to column web the base 1.00 index
connection BW-1 has a single vertical plate welded to the column
web with horizontal stiffener plates (normally 1/2" thick) welded at
its top and bottom. The bolt holes are located outside the toe of the
columnflanges, which allows for easyerection entry of the beam as
well as accessibility for impacting the high strength bolts.
T. & B. FLANGE
- E
'1
I
i
!
M
I
I
/+
1
WEB BOLTED- FLANGE BUTT WELDED
. . . DW--1 . . .
Relative Cost
1.00
DW-1 is the base 1.00 index connection and
employs a bolted vertical web extension plate.
Note that only fillet welds are necessary for the
ver'dcal web plate. Flanges are folly welded to the
continuity plates.
Al l the above detai l s are from Steel Connecti ons/Detai l s and Rel ati ve Costs (Steel TIPS). [19]
32
23. Restrained Welded Joints
Problem
In beam-to-column flange moment connections,
the most economical and most common design
uses welded flanges and a high-strength bolted
web with ' bolts. This connection is
shown in Detail CF--1 on page 10 of Steel Con-
nections/Details and Relative Costs (Steel TIPS).
[19] Detail CF--1 is shown below.
Problems may occur on large beams with thick
flanges and deep webs. If the web bolts are
tightened before the welds are made (the most
desired erection sequence), then the welds will be
restrained by the bolts while cooling, which could
result in lamellar tearing of the column flange, or
cracked welds.
CF--1
T. & B. FLG. F--2
WEB BOLTED- FLANGE BUTT WELDED
==CF--1 =,, , , =CF--2 , , ,
Relative Cost Relative Cost
1.00 1.06
For this category of connecUon, the beam-to-
column moment connection CF-1 is the baseRela-
live Cost Index 1.00 connection, with a single
shear plate being fillet welded to the column
flange. Beam flanges are fully welded to the
column flange, providing a very ductile and eco-
nomical momentconnection. Attachingthe shear
tab to the column with a full penetration weld
rather than a double fillet weld increases the rela-
ve cost 6%.
Sol uti on
Acceptable Procedure. For most beam flanges,
the structural designer should allow the web bolts
to be tightened before the welds are made. This
procedure is acceptable because:
While the welds are cooling, the shrinkage force
in the weld will overcome the allowable load on
the bolts. The bolts will slip horizontally and go
into bearing. After the weld has cooled, the bolts
will not slip again.
After the weld is made, the bolts will still act as
slip-critical bolts--as designed.
Special Procedures. For beams with too many
web bolts, the weld shrinkage force will not be able
to overcome the allowable bolt load. Special
design and erection procedures should be fol-
lowed, because the weld area must be allowed to
shrink. Additionally, "snugged-up" bolts may not
be able to be tightened after welding because the
bolts will bind as the weld shrinks and prevent
proper tightening. Two methods can be used to
solve the problem:
Keep the design of bolted webs. Provide hori-
zontal slotted holes in the column shear plate for
weld shrinkage. Bolts are then fully-tensioned
after the flange welds are made.
Change the design to welded webs. Some
erectors use horizontal slotted holes in the col-
umn shear plate for standard bolts. Other erec-
tors use standard holes in the shear plate and
use erection bolts. After the flange welds are
made, the web is then welded to the column as
shown in Detail CF---4, on page 11 of Steel
Connections/Details and Relative Costs (Steel
TIPS). [19] The web weld is restrained by the
welded flanges. However, since the weld size is
much smaller than the flange welds and distrib-
uted over a larger area of the column flange,
lamellar tearing or a cracked weld should not
occur if proper welding techniques are used.
Welding Techniques. Proper welding techniques
including preheat, peening, postheat, controlled
cooling, and electrode selection will help to avoid
defects in restrained welds. For information on
restrained welded joints see:
33
"Avoiding Weld Defects," on page 14 of "Struc-
tural Steel Construction in the '90s," in Steel
TIPS. [17]
Page 4-152 in the Manual of Steel Construction.
[6]
"Commentary on Highly Restrained Welded Con-
nections,'' AISC Engineering Journal. [23]
Records. The structural designer may require the
erector to provide proof that webs can be bolted
before successful flange welds are made. Records
of past experiences will be helpful to provide the
required proof. The records will be available if the
erector has made welding procedures for prior
projects that include a welding sequence where
the webs are bolted before the flange welds are
made.
34
24. Field-Welded Curb Angles
P r o b l e m
The structural drawings show curb angles (or bent
plates) field-welded to periphery beams with over-
head welds. See design detail below. These
overhead welds are costly and require the welder
to work on the exterior of the building--a safety
hazard not only to the welder, but to workers and
others below the welder.
I
I
V
' A325 SC
HOLES
IN BEAM
@P
Design Detail
Solution
Tolerances. Field-adjustment of curb angles is
necessary when the alignment of the angles re-
quires limits closer than the normal steel frame
alignment tolerances specified in Sections 7.11.3.1
and 7.11.3.2 in the Code of Standard Practice. [1]
When alignment of the angles is allowed to follow
the normal steel frame alignment, then the angles
are shop-welded. Tolerances for adjustable items
are specified in Section 7.11.3.3 of the Code. Do
not expect the steel erector to adjust the angles to
a "zero tolerance."
The alignment of these adjustable items requires
an adjustable connection to accommodate mill,
fabrication, and erection tolerances. See the last
paragraph on page 48 of the Commentaryon the
Code of Standard Practice. [2]
Field Attachment. Two methods of field-attach-
ing the angles to provide adjustment and to avoid
the overhead welds are:
. Field-bolting. Space bolts as required.
Field-welding with slotted plug welds near the
toe of the beam flange. Space welds as required.
See details below for these suggested two attach-
ment methods.
J
/L VOP
!
iN PL Or ANGLE
NOTE: P=CENTER TO CENTER
SPACING
Suggested Bolting Detail
@p
1 . L VOP
' t '
! I
Suggested Welding Detail
35
25. Steel Floor Deck Spanning Uneven Surfaces
Problem
While placing the steel floor deck, the steel deck
contractor can not make the deck bear on adjacent
supports. This condition exists when:
Fill-in beams or trusses with large camber are
adjacent to column line beams or trusses with
much smaller camber.
Beams, trusses, or joists with large cambers are
adjacent to deck shelf angles attached to con-
crete walls.
The elevation differential of adjacent supports is
too great to allow the steel deck to deflect and bear
on each support. See following elevation:
The saw cut is made to the top surface and vertical
surfaces of the ribs, but not the bottom surface
bearing on the beam. The resulting gap is taped
to contain the wet concrete. As the concrete is
poured, the fill-in beams will deflect and the gap
may close.
Note: The structural designer usually designs the
deck to span continuous over at least two supports
to take advantage of deck continuity over multiple
supports. This continuity reduces moment and
deflections in the deck. Before the deck is cut, the
structural designer must be notified.
FILL IN
STEEL FLOOR DECK 2
Deck Not Bearing
J GAP
, /
-F
Sol uti on
Practical Sol uti on. A practical, and probably the
only solution, is to saw-cut the deck at the support(s)
adjacent to the support that is not bearing. The
deck will then change from a cantilevered to a
simple span. See following elevation:
The taping of gaps at butted ends is found on
page 16, Article 4.3, "Lapped and Butted
Ends," in the Design Manual for
Composite Decks, Form
Decks and Roof
Decks. [24]
COLUMN LINE BEAM OR
CONCRETE WALL WITH Roof Deck.
SHELF ANGLE The installa-
tion of steel
roof deck on
di f f er ent i al
warped sur-
faces is discussed in the January 1992 issue of
Steel TIPS, titled, "Steel Deck Construction." [25]
FILL
STEEL FLOOR
DECK2 SAW CUT
Deck Bearing
COLUMN LINE BEAM OR
CONCRETE WALL WITH
SHELF ANGLE
36
26. Project Specifications
Problem
At times the specifications may:
Be vague.
Include implied statements.
Include requirements inappropriate to the project.
Be more restrictivethan necessaryfor the project.
For example, plumbing requirements that are
more restrictive than specified in the Code of
Standard Practice. [1]
Require fabricator to complete design in order to
make a bid. If so, the erector must also make
assumptions.
Conflict with the drawings or with notes and
specifications on the drawings. Note: In Califor-
nia, the structural designers typically place speci-
fication-type notes on the drawings.
Not be written for the specific project.
Assign work to the steel fabricator, erector, mis-
cellaneous metal contractor, etc.
Solution
Avoiding Specification Problems. The fabrica-
tor and erector must live with and comply with
specifications and drawings developed by the
structural designer. To avoid specification prob-
lems, the structural designer should:
Either prepare the specifications following the
Construction Specifications Institute's (CSI)
Manual of Practice, or coordinate structural steel
requirements with the specification writer when
the project has a specification writer. [26] Hope-
fully, the specification writer will follow the CSI
format.
Make certain specification-type notes placed on
the structural drawings agree with the structural
steel specification section.
Follow the specification requirements set forth in
Section 3, "Plans and Specifications," in the
Code of Standard Practice, and the checklist
contained in Section 3 in the Commentary on the
Code of Standard Practice. [1,2]
Review the structural steel specification sugges-
tions in "What Design Engineers Can Do to
Reduce Fabrication Costs," and in "Value Engi-
neering and Steel Economy," in Steel TIPS. [12,
18] However, items 11 and 12 in "Value Engi-
neering and Steel Economy" may be a little
misleading. Specification writers should not as-
sign work to subcontractors because the specifi-
cations are normally directed to the general con-
tractor. Instead, all of the required work and
items to be furnished should be specified in the
appropriate specification section.
Make certain the drawings show items required
by the specifications. For example, if the speci-
fications state, "Construction limits for erection
equipment are shown on Drawing S-6," then the
construction limits must be shown on that draw-
ing.
Include Charpy requirements for Groups 4 and 5
rolled shapes that require full penetration welds.
See "Heavy Structural Shapes in Tension Appli-
cations,'' in Steel TIPS. [27]
Specifications on Jobsite. Specification writers
and structural designers are sometimes disturbed
to discover the steel erector is not using the
specifications or structural drawings to erect the
steel. The specification writer and structural de-
signer should realize the shop drawings, erection
drawings, bolt lists, welding procedures, and some-
times erection equipment used by the steel erector
are all developed from, are based on, and are
extensions of the specifications and drawings.
The steel erector's field crews will use these
documents to erect the steel, and not the specifi-
cations and drawings prepared by the structural
designer.
37
Reference List
1. Code of Standard Practice for Steel Buildings
and Bridges, AISC, Chicago, June 10, 1992.
. Commentary on the Code of Standard Practice
for Steel Buildings and Bridges, AISC, Chicago,
June 10, 1992.
3. Detailing for Steel Construction, AISC, Chicago,
1983.
. David T. Ricker, "Steel Interchange," Modern
Steel Construction, AISC, Chicago, January
p. 9.
. David T. Ricker, "Some Practical Aspects of
Column Base Selection," Steel Design Guide
Series 1: Column Base Plates, AISC, Chicago,
1990, 43-51.
6. Manual of Steel Construction: Allowable Stress
Design, 9th ed., AISC, Chicago, 1989.
7. Quality Criteria and Inspection Standards, 3d
ed., AISC, Chicago, 1988.
. Vijay P. Khasat, "Steel Interchange," Modern
Steel Construction, AISC, Chicago, May 1993,
p. 10.
. David T. Ricker, "Steel Interchange," Modern
Steel Construction, AISC, Chicago, July 1993,
p. 9.
10. Thomas C. Powell, "Steel Interchange," Modern
Steel Construction, AISC, Chicago, December
1992, p. 12.
11. Manual of Steel Construction, Vol. II,
Connections, AISC, Chicago, 1992.
12. "What Design Engineers Can Do to Reduce
Fabrication Costs," Steel TIPS, Structural Steel
Educational Council, Moraga, California, June
1992 (Printed from Modern Steel Construction,
February 1992).
and
"What Design Engineers Can Do to Reduce
Fabrication Costs," Modern Steel Construction,
AISC, Chicago, February 1992, 28-33.
13.
1 4 .
1 5 .
1 6 .
17.
1 8 .
1 9 .
20.
21.
Article 29, "Erection and Construction," CALl
OSHA State of California Construction Safety
Orders, BNI Books, Los Angeles, June 1989, p.
244.
Lawrence A. Kloiber, "Designing Architecturally
Exposed Steel Tubes," Modern Steel
Construction, AISC, Chicago, March 1993, 30-
38.
Catalog of Standard Specifications Load Tables
and Weight Tables for Steel Joists and Joist
Girders, Steel Joist Institute, Myrtle Beach,
South Carolina, 1992.
Technical Digest, No. 9, Handling and Erection
of Steel Joists and Joist Girders, Steel Joist
Institute, Myrtle Beach, South Carolina, 1992.
F. Robert Preece and Alvaro L. Collin,
"Structural Steel Construction in the '90s," Steel
TIPS, Structural Steel Education Council,
Walnut Creek, California, September, 1991.
David T. Ricker, "Value Engineering and Steel
Economy," Steel TIPS, Structural Steel
Educational Council, Moraga, California, August
1992 (Printed from Modern Steel Construction,
February 1992).
and
David T. Ricker, "Value Engineering and Steel
Economy," Modern Steel Construction, AISC,
Chicago, February 1992, 22-26.
Steel Connections/Details and Relative Costs,
(Steel TIPS), The Steel Committee of California,
Walnut Creek, California, 1986.
W. A. Thornton, Ph.D., P.E., "Designing for Cost
Efficient Fabrication," Steel TIPS, Structural
Steel Educational Council, Moraga, California,
April 1992 (Printed from Modern Steel
Construction, February 1992).
and
W. A. Thornton, Ph.D., P.E., "Designing for Cost
Efficient Fabrication," Modern Steel
Construction, AISC, Chicago, February 1992,
12-20.
"Steel Interchange," Modern Steel Construction,
AISC, Chicago, March 1992, p. 10.
38
22. Structural Welding Code--Steel (D1.1), AWS,
Miami, 1992.
23. "Commentary on Highly Restrained Welded
Connections," AISC Engineering Journal,
Vol. 10, No. 3, 3d Quarter 1973, 61-73.
24. "SDI Specifications and Commentary for
Composite Steel Floor Deck," Design Manual
for Composite Decks, Form Decks and Roof
Decks (Publication No.26), Steel Deck Institute,
Inc., Canton, Ohio, 1987, 14-21.
25. "Steel Deck Construction," Steel TIPS,
Structural Steel Educational Council, Moraga,
California, January 1992.
26. Manual of Practice, The Construction
Specifications Institute, Alexandria, Virginia.
27. "Heavy Structural Shapes in Tension
Applications," Steel TIPS, Structural Steel
Educational Council, Moraga, California,
October 1993.
About the author
James J. Putkey is a consulting civil engineer in Orinda, California.
He received a BCE degree from the University of Santa Clara in
1954. After two years in the U.S. Army, 19 years with the Erection
Department of Bethlehem Steel Corporation--Pacific Coast Divi-
sion, and seven years with the University of Cal i forni a--Offi ce of the
President, he started his own consulting business. He has provided
consulting services to owners, contractors, attorneys, and steel
erectors for the past 12 years.
39
470 Fernwood Drive
Moraga, CA 94556
(510) 631-9570
SPONSORS
Adams & Smith
G.M. Iron Works Co.
Hoertig Iron Works
Hogan Mfg., Inc.
Junior Steel Co.
Lee & Daniel
McLean Steel, Inc.
Mid West Steel Erection
Oregon Steel Mills
PDM Strocal, Inc.
Reno Iron Works
H.H. Robertson Co.
Stockton Steel
- The- local structural Steel industry (above sponsors) stands ready to assist You in - [
I
-. determining the most economical solution for your products. Our assistance can
I
range from budget prices and estimated tonnage to cost comparisons, fabrication
Funding for this publication provided by the California Iron Workers Administrative Trust.
STEEL COMMITTEE OF CALIFORNIA
TECHNICAL INFORMATION & PRODUCT SERVICE
MAY1989
The Economies of LRFD
in Composite Floor Beams
by Mark C. Zahn
Reprinted with permission of the American Institute of Steel Construction from the
AISC Engineering Journal, Second Quart'er, 1987.
The Economies of. LRFD in Composite
Floor Beams
MARK C. ZAHN
The AISC Load and Resistance Factor Design (LRFD)
Specification represents the state-of-the-art approach in the
U.S. for routine structural steel design. Based on a reliabil-
ity, ultimate strength theory, use of the LRFD Specification
may have an economic advantage over the current 8th
Edition AISC Manual o f Steel Construction, 2 Allowable
Stress Design (ASD) Specification. An economic advan-
tage of reduced weight of building elements, while meeting
all strength and serviceability requirements, is sure to exist.
One major building element where the full economy of
LRFD could be realized is in composite floor beam con-
struction. A comprehensive study is thus warranted.
The purpose of this paper is to report the findings of an
in-depth comparative design 'study between the recently
published LRFD and the established ASD Specifications
for simply supported composite floor beams. The results
are based on a direct beam weight comparison of over 2,500
composite beam designs and shear stud requirements using
A36 steel.
DESIGN CONSIDERATIONS
The design of any structural member must meet both
strength and serviceability requirements. Strength criteria
is met by allowable stress limits using the ASD approach
and by member resistance, or capacity, using the LRFD
approach. The differences in design methodologies will be
illustrated by applicable specification sections and equa-
tions.
Serviceability requirements in this study include: (1)
limiting live load deflections to 1/360 of the beam span L
and (2) limiting the floor vibration induced by a heel drop
impact load. The upper limit range of slightly perceptible to
distinctly perceptible as reported by Murray3'4was selected
for this investigation. This corresponds to a response rating
R of 2.75. Floor vibration perceptibility was of particular
interest. With a savings in beam weight expected using the
Mark C Zahn ts a Structural Engineer w;th Chris P Stefanos
Associates, Inc , Consult;ng Structural Engineers, Oak Lawn,
I/hnots
LRFD Specification, a fundamental question arises. With
all design parameters and serviceability requirements being
equal, will an LRFD composite beam have both an accept-
able floor vibration response rating and an econom,c
advantage over an ASD composite beam design? A direct
economic comparison, using dollars and cents, is outside
the scope of this paper because of many varying factors
However, by making a comparison of beam weight and
shear stud requirements of each design, an indirect cost
comparison on a quantitative basis can be made.
DESIGN PARAMETERS
Several basic assumptions are common to each design
method: (1) only wide flange sections which meet the com-
pact section criteria of ASD Specification 1.5.1.4.1 and
LRFD Specification Sect. B5.1 are used, (2) beams are
considered to have continuous lateral support of their com-
pression flange during construction, (3) metal deck ribs run
perpendicular to the beam span, (4) construction live loads
are included for unshored construction and (5) dead load
deflections are not limited; it is assumed that beam camber-
ing during fabrication could be provided for in either design
method and would not be a significant factor in this study.
Loading, type of construction, spacing, and span param-
eters were selected to envelope each controlling design
criteria of strength and serviceability. Spans range from 10
ft to 45 ft; beam spacings range from five ft to 10 ft.
Two load conditions have been studied with both shored
and unshored construction types considered. Each load
condition uses minimum slab and metal deck depths m
order to investigate the minimum potential savings be-
tween the ASD and LRFD design approaches. Represent-
ing a "common" load case a live load of 100 psf with a dead
load of 50 psf was selected. This load case has a total slab
thickness of 4 in,, 1-1/2 in. metal deck, 3,000 psi concrete
strength and a live load to dead load ratio of two. On the
heavier side a live load of 250 psf, dead load of 60 psf, total
slab thickness of 6 in., 3-in. metal deck, and 4,000 psi
concrete strength is used. This second loading condition has
a live load to dead load ratio greater than four,
Table 1. ASD/LRFD Shape List
W 8xlO W 8x24 W21x44 W27x 94 W36x160
WlOx12 W16x26 W18x46 W30x 99 36x170
W 8x13 W 8x28 W21x50 W27x102 W36x182
WlOx15 W14x30 W24x55 W30x108 W36x194
W12x16 W16x31 W21x57 W30x116 W33x201
WlOx17 W14x34 W18x60 W33x118 W36x210
W 8x18 W18x35 W24x62 W33x130 W33x221
W12x19 W16x36 W24x68 W36x135 W36x230
W 8x21 W14x38 W24x76 W33x141 W36x245
W14x22 W18x40 W27x84 W36x150 W36x260
Wide flange sections range from W8 x 10 to W36 x 300
for this investigation. Considering all of the shapes com-
used for beam members, a reduced list of 50 shapes
was selected and arranged according to increasing weight.
reduced data base is necessary to eliminate duplicate
weight shapes and assure an economical design. For exam-
pie, W14 x 30 and W12 x 30 are of equal weight; however,
because of the greater depth, the W14 x 30 section is a part
of the design section data base. Depth of beam was not an
imposed limitation on beam selection.
DESIGN METHODOLOGY
Each composite beam shape is chosen from respective com-
puter data files for ASD and LRFD design using an IBM
PC/AT computer. Each file contains the same shapes and
order of selection (Table 1). Properties required for calcu-
lating transformed sections are taken from the 8th Edition
AISC Manual. The LRFD data file contains moment ca-
pacity values OPM,' and OPMp, neutral axis and resultant
compressive force locations YI and Y2 respectively, and
shear connector strength XQ,' associated with the above
values. This information is from Part 4 of the 1st Edition
LRFD Manual, Composite Beam Selection Tables.
Three areas of composite beam design have undergone
noteworthy changes in developing the LRFD approach
from the ASD approach. The first area which has been
revised is the determination of width. LRFD Sect.
I3.1 defines bq12 for interior beams as the minimum of: (a)
one-eighth of the beam span center line to center line of
supports, or (b) one-haft the distance to the center of
adjacent beams. ASD Section 1.11.1, on the other hand,
gives, for interior beams, the following definition for be/g12
as the minimum of: (a) one-eighth of the beam span, (b)
one-half the distance to the center line of adjacent beams or
(c) eight times the slab thickness plus one-haft the beam
flange width. By reviewing these criteria it is seen that for
intoner beam spacing effective widths will be equal with
both methods. However, as the spacings increase the slab
thickness criterion of the ASD approach will control and
givea smaller effective width for ASD beams. This requires
a larger beam shape to meet ASD requirements. Recent
studies$'6 have indicated the established ASD slab thick-
ness criterion is conservative as applied to interior beams
and that the LRFD effective width definition is much more
realistic.
The second, and most striking, change is in bending of
both shored and unshored construction. Based on the
assumptions previously stated, the ASD Specification
limits the allowable steel fiberstress for shored construction
to 0.66 Fy(Sect. 1.5.1.4.1) and the allowable concrete com-
pressive stress to 0.45 (Sect. 1.11.2.2). For unshored
construction an allowable steel flange tension stress of 0.89
Fy is indirectly applied? by ASD Specification Sect.
1.11.2.2: S = (1.35 + 0.35 MLL/MDOS,. Ss is the section
modulus of the steel beam alone at its bottom flange, MLL
and MoL are the live load and dead load moments respec-
tively, and S is the maximum transformed section modulus
for which unshored construction is permitted. The LRFD
approach relies on load factors, 1.6 live load and 1.2 dead
load and resistance factors q>. The criteria of factored re-
quired strength by analysis must be less than or equal to the
member design resistance, more simply R < dpR. Using
compact sections and full lateral support composite beams
may develop the maximum resistance of the steel section
from a plastic stress distribution (LRFD Sects. I3.2 and
Fl.3). By Sect. Fl.2 and I3.2 and Part 4 of the LRFD
Manual OPis equal to 0.9 for noncomposite action (prior to
concrete set during unshored construction) and 0.85 for
composite action. Having factored moments from analysis,
the engineer selects a section which provides a resistance
OPM,' equal to or exceeding that required. Composite beam
design aids are provided in the LRFD Manual. Part 4 gives
OPMt,, OPM,,, ILB, Y1, Y2, and ZQ,' values as defined pre-
viously. For a complete explanation of the development
and use of this design aid the reader is referred to Refs. 1
and 8.
The third noteworthy difference in design methods in-
volves the effective moment of inertia used for deflection
calculations. As noted, the LRFD composite beam design
aid in Part 4 of the Manual contains values of lower bound
moment of inertia (/LB). These values represent a trans-
formed moment of inertia using an equivalent area of con-
crete which is effective in compression and based on the
transfer of horizontal shear force? It is obvious deflection
calculations usingILs may differ considerably from those of
ASD which permits the full concrete thickness for moment
of inertia values.
Shear requirements on the web area Aw have not been
revised significantly. The familiar ASD Specification Sect.
1.5.1.2.1 gives the allowable shear as Vtt. = O.4 FyA w. This
same format applies to the LRFD Specification Sect. F2
defining the shear resistance as OPR, = O 0.6 Fy Aw for
compact sections with no stiffeners, where b equals 0.9.
RESULTS
The analysis comparing composite beam designs and the
potential savings between the LRFD and ASD Specifica-
tions focuses on four major areas: (1) strength, (2) floor
vibration response, (3) deflecuon, and (4) shear connector
requirements. The results are presented as a series of
graphs with accompanying discussion To include graphs
for each load case, type of construction, and beam spacing
would mean including over 70 graphs----a total well beyond
the intent of this report. Therefore, graphs have been
selected which best represent the overall results.
Consider first the criteria of strength only. With all other
limitations removed, designs by each method can be com-
pared for beam weight savings in its most basic form. Figure
la illustrates a plot of beam weight, based on strength only,
for each approach vs. span length. This graph represents an
eight foot beam spacing, shored construction, and the
heavy loading condition as described previously and gives a
reasonable representation of the overall results. With only
minor modifications, this graph could represent any load
case, construction type and beam spacing. In all cases,
there is a region in the lower span lengths which has very
little or no difference in beam weight by either specifica-
tion. As span length increases to 20 or 22 fi, the margin of
difference between sizes increases slightly and, in general,
will decrease at the 20- or 22-fi span length. This area, 13 to
22 fi, exhibits a reduction in beam weight using LRFD
ranging from two to seven percent. Past the 22-fi span,
LRFD. consistently gives reduced weights over the ASD
design approach. Increased strength associated with the
LRFD approach is illustrated in Fig. lb. It can be seen that
a greater strength is developed when considering the full
section as effective with a plastic stress distribution. Thus,
under the LRFD approach, strength requirements can be
met with a lighter, more economical section. In only a few
cases, design beam weights were equal at random span
lengths and in no case is an LRFD beam design heavier than
an ASD beam design. Figure la shows weight reductions as
high as 15%. Overall, based on strength only, reductions
reached a maximum of 45%. This is only the first indication
of the savings possible during the LRFD Specification
The addition of floor vibration response rating and
deflection' limitations gives very noteworthy results. All
calculations, for each design method, follow the procedure
by Murray3'4 with a damping of 4% and a weight term, for
the purpose of this investigation, which includes dead load
plus full live load. Interesting results occurred for the load
case using 100 psf live load and 50 psf dead load (having
minimum slab and deck depths). For span lengths less than
23 feet, vibration rating was of great significance (due to the
reduced total weight and reduced moment of inema terms)
and is the controlling factor for beam selection. With this in
mind, it is easily understood that there will be no difference
in beam weights using the LRFD or ASD approaches up to
23 foot spans. Beyond 23 feet, however, strength becomes
the controlling factor and a substantial savings in weight is
seen. Figures 2a and 3a show the relationships between the
two design methods for the load case of 100 p,sf LL and 50
psf DL. Spacing and type of construction are noted on each
figure. Figures 2b and 3b are plots of the percent reduction
AED LRF0
LOAD PLASTIC
8TRESS DISTRIBUTION STRESSDISTRI8UTION
Figure lb
10Q
S0
80
70
SO
SO
40
30
20
10
Fig. I a.
AEO DESIGN - J
..... . . . / / --LRFOOEO,OR
_ , . ,DS0 p.i
tO 16 20 8 3 3 40 45
BEAMSPAN, FI
ASD/LRFD composite beam design comparison
SO
SO
e 4O
m 30
l
' 20
/ -
VI BRATI ON STRENGTHCONTROLLING / . / /
i=R1CEABiLiTY
CONTROLLING i
i /__..--1
/_._/'
ASD DESIGN r-J'
BEAM SPACING9'-0' otc
../' SHORED CONSTRUCTION
' LIVE LOAD lOS pl)l
61
' I L R
1 I/2' DECK
t'G 3000 pet
10
Fig. 2a.
tE 20 25 30 36 40 4S
BEAM SPAN. FT
30
20
tO
Fig. 2b.
wTASO'" WTL.D
x tOO
W?ABD
BEAM SPAN FT
ASDILRFD composite beam weight savmgs
REAM EPACINO E'-D' c/o
SHORED CONSTRUCTION
LIVE LOAD IDS 9e
DEAD LOAD 50 pi /
2 1/2' BLAB
I 112' DECK
I*c ,3000 pll
in beam weight, or savings, versus span length associated
with Figs. 2a and 3a. Savings in weight reach as high as 30%
for the shored condition (Fig. 2) and as high as 37% for the
unshored condition (Fig. 3). Savings are, therefore, not
limited to one type of construction. Figures 4a and 4b give a
similar result for the heavier load case. However, in the
shorter span region vibration is not a controlling factor and
equal weights in this area are seldom encountered. In fact,
for the span lengths between 10 and 16 Et, 'savings are the
greatest and approach 40%.
The third item in this study is deflection. With the intro-
duction in the LRFD approach of the Lower Bound Mo-
ment of Inertia (ILs) concept, a direct comparison of live
load deflection is inconsequential. ILS values can be as
ED
TO
lO
EO
40
BO
20
10
0
-ID
-20
-30
-40
-60
so:
BO,
ID'
SPACING i ' -O' O14 BEAM
SNORED CONSTRUCTION 1
LIVE LOAD 100 pll.
DEAD LOAD 60 piE.
30
2 tiE' 8LAR
I tiE' DECK w
f'g 3ODD pll.
6D
o
I I I I I t i REAM SPAN. FT.
10 la 20 25 30 35 4&
O
8TUDABD --STUDLRFD
BTUDABD
Fig. 2c. ASD/LRFD shear stud comparison
VIRRATION
8ENVlGEARiLiTY
STRENGTH CONTROLLINO
.
BEAM EPACING Il'-(}' OtC
DNBHONED CONSTRUCTION
LIVE LOAD 100 pll
DEAD LOAD 50 pi /
2 112 SLAB
I t12' DECK
I c 30BO pi t
tO
3a.
,5 :5 :6 s' . , ,;
BEAM SPAN, FT
40
WTASO'-WTLRFo
16SAVINGS = X IOO
WTASD
Fig. 3b.
BEAM SPACING 5 -0'
UNBHOSED CONSTRUCTtON
LIVE LOAD 1o0 psi
DEAD LOAD 50 pi t
2 I/2' SLAB
I 112' DECK
I'D 3000 psi
'' ! I
. . . . ;o ', ,'o ,',
BEAMSPAN, FT
ASDILRFD composite beam weight savings
l O
70
BO
40
EO
ED
ID
O
-10
-ED
-BO
-40
-BO
V---
,'o'" , ; 60
REAM ESAOIN6 6' -O' oil
UNEHORED CONSTRUCTION
LIVE LOAD IDD met.
DEAD LOAD ED pll.
112' 8LA8
I ti2' DECK
1 rc3OOO pll.
PB J BEAM EPAN. FT
2
BD SB 4D .B
BTUB D --BTUDLRFD
REDUCTION X IDB
8TUDA8D
Fig. 3c. ASD/LRFD shear stud comparison
30
20
lO
ED
10
eD
ED
40
30
20
$O . . . . .'
10
Ftg. 4a.
I
1o
ABODESIGN
.,' BEAM SPACINGe'-o
..--/--' So BLAB
/ ' 3' DECK
I'G 4000 prat
' : ' ; ,
15 20 S 30 3 40 45
BEAM SPAN FT
ASD/ LRFD compostte beam destgn comparison
w s o - w n p o
- - %SAVINGS= T
BEAM SPAN. FT
Ftg. 4b. Compostte beam weight savmgs
BEAM8PACINGE'-O*
UNEHOREO CONBTRUCTION
LIVE LOAD250 pll
GEAOLOADOBplf
3' BLAB
3' DECK
rc 4000 IlL
E0
T0
eo
60
40
30
20
10
O 104"' ti6
-tO
-20
%REDUCTION
-30
-40
-S0
Fig. 4c.
BEAM BPACINEIB*-B' cig
UNBHOREDCONSTRUCTION
LIVE LOAD EEODIf.
DEADLOAD eo psi.
a' BLAB
3' DECK
f'c 4S00 pII
ID 215 3% 31E 20 5 BEAMBPAN. ET
STUDAsD --BTUDLRFO
X tOO
BTUDASD
ASD/ LRFD shear stud comparison
much as 50% !ess thanthe transformed moment of inertia
for the same size section by the ASD approach. It is worth
noting, however, in no case did the live load deflection
criteria of fi/360 cause a heavier section to be selected.
And, studies by Valleni!!a and Bjorhovde have suggested
the ASD Specification procedure underestimates actual
deflections, 6 Furthermore, Ref. 6 suggests the LRFD con-
cept gives more realistic results for interior beams. Live
load deflections which correspond to Fig. 3 range from 0.01
in. to 1.4 in. by LRFD and 0.01 in. to 1.02 in. by ASD.
Dead load deflections range from 0.01 to 1.55 in. by LRFD
and 0.01 to 1.29 in. by ASD.
The fourth area of interest is shear-stud connector re-
quirements. The established ASD Specification defines the
horizontal shear force for simply supported beams as the
smaller of Vh = 0.85A/2 (Formula 1.11-3) or Va = As
Fy/2 (Formula 1.11-4). In these formulas, As is the area of
the steel section and Ac is the area of concrete within the
effective width. LRFD is quite similar in format to ASD,
but extends the approach of ASD Formula 111-3 to an
ultimate state and considers only the effective area of con-
crete which is in compression. This area may not include the
full concrete depth. LRFD horizontal shear force is cov-
ered in Sect. I5.2 of the LRFD Manual. Composite Beam
Selection Tables in Part 4 tabulate horizontal shear force
requirements [Q based on neutral axis and compressive
force locations Y1 and }/2. Typical results of the percent
change in connector requirements from ASD to LRFD are
shown m Figs. 2c. 3c and 4c. For the 100-psf live load with
50-psf dead load cases of Figs. 2c and 3c, three distinct
ranges of span length are worth noting. For span lengths
between 10 and 20 ft, the differences in connector require-
ments are primarily due to the differences in design meth-
odologies. For this minimum deck and slab depth condi-
tion, the controlling ASD shear force is Formula 1.11-3.
The ASD specification permits using the full depth of con-
crete above the metal deck for calculating Ac. On the other
hand, because vibration is the controlling factor in this
region, the LRFD beam is relatively inefficient for compos-
ite action and develops only a very small effective concrete
depth. This explains the wide margin of difference between
the ASD and LRFD shear-stud requirements. Type of
construction is not a factor. As the span length increases to
40 ft there is a consistent reduction in shear studs required
by LRFD over ASD beams. These reductions reach as high
as 40% for closely spaced beams (Fig. 3c) and as high as
30% for larger spacings (Fig. 2c). From 40 to 45 ft, the ASD
approach requires increasingly fewer studs over the LRFD
approach, as beam spacing increases from 5 to 10 ft. In this
upper range of spans, the sharp decline in margin of differ-
ence is again attributed to methodology. Two factors are
important on the ASD side; the effective depth and the
effective width b,/f. As stated previously, the ASD criteria
of eight times the total slab depth criteria will control in Fig.
2c. Therefore, ASD formula 1.11-3 controls. Stud require-
ments for LRFD, however, will be controlled by LRFD
Sect. I5.2 Eq. 2: A, F,. Thus, the margin of difference will
decrease as span and spacing increases. In some cases, as
shown in Fig. 2c, ASD may require fewer studs over
LRFD. It should be noted, however, that a lighter section
may be used by the LRFD approach over the ASD method
and an overall economy may be realized.
The heavier load case of Fig. 4 does not follow this same
pattern. With few exceptions, span lengths of 10 to 20 ft
show a sign,ficant reducnon (up to 60%) in shear studs
using LRFD. Between 20 and 45 ft, the reduction varies
from 0 to 33%. Greatest reductions are shown in the 24- to
27-ft, 30- to 33-ft and 38- to 41-ft areas; Fig. 4c represents
this behavior. In no case did the ASD require fewer studs
than the LRFD approach for this load condition.
Further investigation has been completed using a 3-in.
metal deck, 2-in. slab, and the (100 psf LL)/(50 psf DL)
condition. In both the shored and unshored cases, the
behavior exhibited follows that shown in Fig. 4. As in Figs.
4a and 4b, the greatest savings in beam weight occur over
spans ranging from 10 to 25 ft. In spans from 25 to 45 fi, a
consistent savings of 15% occurs. Shear stud reductions
peak in the shorter spans and level off to 20% for spans
ranging from 25 to 35 ft.
The full economies of LRFD in composite floor beam
construction cannot be realized without addressing the rel-
ative cost differential between the ASD and LRFD
methods on the final composite beam design. Preliminary
results of this study show that, when using minimum slab
and deck parameters without regard to fire protection rat-
ing by the structure, there is an average savings of up to 6%
for span lengths between 10 and 18 ft, 15% for span lengths
between 18 and 30 ft and 14% for span lengths between 30
and 45 ft. These findings are based on a fabricated and
erected cost per pound of steel to cost per shear stud
connector ratio of 1:3.75.
CONCLUSION
This investigation has shown the recent 1st Edition Load
and Resistance Factor Design Manual of Steel Construction
does have an economic advantage over the 8th Edition
Manual of Steel Construction in composite floor beam con-
struction. For span lengths of 10 to 20 ft with a 100-psf live
load and 50-psf dead load condition, minimum slab and
deck depths, vibration serviceability is the controlling fac-
tor regardless of the design method used. As spans increase
to 38 ft, designs by the LRFD approach consistently give
lighter beam weights and require fewer shear connectors
for full composite action. Beyond 38 ft, lighter beam
weights are evident with a moderate increase in shear con-
nectors. Preliminary. results of a cost comparison study
indicate savings average 6% to 15% for span lengths rang-
ing from 10 to 45 ft. Serviceabihty has not been compro-
mised using the LRFD approach. All designs meet the
L/360 live load deflection limitation commonly used in
design. LRFD deflections are slightly higher than those
using the ASD method of transformed section properties.
It should be noted, however, previous studies indicate the
ASD approach currently underestimates actual composite
beam deflections.
Based on this investigation, the Load and Resistance
Factor Design approach to composite floor beam construc-
tion will have a substantial overall savings in material costs
without compromising serviceability. In addition, the intro-
duction of ILB in the LRFD Manual for deflection calcula-
tions gives a much more realistic and reliable account of live
load deflections.
ACKNOWLEDGMENTS
Appreciation is due Elizabeth Reardon, Structural En-
gineer with Sargent & Lundy Engineering, for her pro-
gramming contribution and floor vibration criteria input.
Thanks also to Lynn Echlin for typing this manuscript.
REFERENCES
1. American Institute of Steel Construction, Inc. Load
and Resistance Factor Design Manual of Steel Con-
struction 1st Ed., 1986, Chtcago, Ill.
2. American Institute of Steel Construction, Inc. Manual
of Steel Construction 8th Ed., 1980, Chicago, Ill.
3. Murray, Thomas M. Acceptability Criterion For Oc-
cupancy-Induced Floor Vibrations AISC Engineering
Journal, 2nd Qtr., 1981, Chicago, Ill. (pp. 62--69)'.
4. Murray, Thomas M. Design to Prevent Floor Vibra-
tions AISC Engineering Journal, 3rd Qtr., 1975, New
York, N.Y. (pp. 82--87).
5. American Society of Civil Engineers Structural Design
of Tall Buildings Vol. SB, 1979 (pp. 620--627).
6. Vallenilla, Cesar R. and R. Bjorhovde Effective Width
Criteria for Composite Beams AISC Engineering
Journal, 4th Qtr., 1985 (pp. 169-175).
7. Gaylord, Edwin H. and C. N. Gaylord Design Of
Steel Structures 2nd Ed., McGraw-Hill, Inc., 1972
(pp. 346--347).
8. Zahn, Cynthia J. LRFD Design Aids: Plate Girders
and Composite Beams AISC National Engineering
Conference Proceedings, 1986, Chicago, Ill. (pp. 37-8
through 37-15).
THE STEEL COMMITTEE OF CALIFORNIA
Ace & Stewart Detailing, Inc.
Artimex Iron Co., Inc.
Baresel Corporation
C. A. Buchen Corporation
Central Industrial Engineering
Co., Inc.
Cochran-lzant & Co., Inc.
Dovell Engineering, Inc.
Hoertig Iron Works
Hogan Mfg. Inc.
INland Steel Company
Junior Steel Co.
Lee & Daniel
McLean Steel, Inc.
PDM Strocal, Inc.
Reno Iron Works
Riverside Steel Construction
H. H. Robertson Co.
Stott Erection, Inc.
Northern California
43 Quail Court, # 206
Walnut Creek, CA 94596
(415) 932-0909
Southern California
9420 Telstar Ave.
El Monte, CA 91731
(818) 444-4519
EEL COUNCIL
&,, IOD'dCTSERViCE
OCTOBER 1992
Economical Use of Cambered Steel Beams
Accommodating Dead Load Deflections:
here are four methods of accommodating beam
dead load deflections during concrete place-
ment and creating an acceptably level floor slab:
1) Let beams deflect and pour a varying thick-
ness slab; 2) Overdesign beams to minimize deflections;
3) Camber beams to compensate for anticipated deflec-
tions; 4) Shore beams prior to concrete placement.
Example: An eco-
nomic anal ysi s is
helpful in selecting
the best approach. As
an example, consider
a 30'x30' bay in a typ-
ical office building.
If beams are .
allowed to sag, the
cost of additional con-
crete to produce a
level slab would be
$0.19/SF, assuming
$60/cu yd for the concrete.
W16 x 26(20)
A572-50
mY
k
-do-
-do-
, +
-do-
Increasing the size of the steel beams in order to
Cost Savings By Cambering:
For the same typical 30'x30' bay example, camber-
ing would cost $0.13/SF. It would eliminate the need for
additional concrete to obtain a level floor; and therefore,
When it comes to cam-
bering beams, more is not
better.
would save $0.06/SF. Plus, the cost of cambering can be
accurately determined with no additional hidden costs.
As bay sizes increase and deflections become
larger, the savings potential of cambering becomes more
dramatic.
For shoring to be economical its cost would have to
be less than the cambering cost of $0.13/SF, including
crack control slab reinforcement over girders. In addition,
there is the added expense caused by the shoring's inter-
ference with subsequent operations such as fire protec-
tion and mechanical systems installation.
. . . . . . . . . . . . _ . _ {i?:'"':";':':"?'::"' :E:""'"""' ' ';::[
DEFLECTEDBEAM
DUETOWEIGHTOFDECK
INDUCEDCAMBER
reduce their deflections, and thus the excess concrete
requirement, would not produce a more economical solu-
tion. The increase in cost for the heavier steel beams
would exceed the cost of concrete and shear studs
saved.
Guidelines for Specifying Camber:
Specifying camber properly is crucial to obtaining an
economical, level floor with the proper slab thickness.
Several factors that influence camber are identified
below.
Calculated Dead Load Deflections - - Ideally, for most
buildings, the finished floor slab should be both level and
of constant thickness. Thus the beam must be level after
the concrete is placed. Only the weight of the beams,
metal deck, and wet concrete should be included in the
dead load deflection calculations. Additional items, such
as partitions, mechanicals, ceiling and any live load
should be excluded.
Connection End Restraint Connections on the
beams provide some degree of end restraint. Therefore,
the full calculated dead load deflection will probably not
occur. The amount of camber specified can be reduced
to minimize the effect of connection end restraint. Many
engineers reportedly specify camber amounts in the
range of 2/3 to 3/4 of the calculated simple span dead
load deflection to acount for this effect.
Mill Tolerances and CamberLosses The tolerance
for mill camber of members 50 ft or less is minus 0" and
plus 1/2". Over 50 ft, the plus tolerance increases 1/8" for
each 10 ft in excessd of 50 ft. There will be additional
camber induced at the mill to assure that it is within toler-
ance.
The minimum amount
of camber is dependent on
both physical and economic
considerations.
desired slab thickness. It is usually easier and more eco-
nomical to accommodate under-cambered than over-
cambered beams.
Mill C a m b e r L i m i t s :
Shapes to be cambered are cold-worked to produce
desired curves subject to limitations shown below.
Cambering other than wide flange or standard beams is
subject to inquiry.
Cambers less than minimums outlined can be fur-
nished, but no guarantee can be given with respect to
their permanency.
Order must specify a single minimum value within
the ranges shown below for the length ordered.
Camber will approximate a simple regular curve
nearly the full length of the beam, or between any two
points as specified. Reverse or other compound curves
are available but subject to negotiation and customer
approval before shipment. Camber shall be specified by
the ordinate at mid length of the portion of the beam to be
curved. Ordinates at other points can be specified but
require negotiation.
However, the camber induced at the mill may not
necessari l y be present in the same amount when
received. The AISC states that "In general, 75% of the
specified camber is likely to remain." But, there is no
W24x 62 W21 x57 W18 x46 W16x 31
guarantee that some mill camber will be "lost" during 55 50 40 26
shipment, fabrication, and erection. 44 35
The effects of mill tolerances and camber losses
tend to offset each other; although, the net effect may be
W12x22 Wl Ox19 W 8 x 1 5 W6x16
slightly more actual camber than specified. 19 17 13 12
"The More is Better Syndrome m When it comes to
16 15 10 9
cambering beams, "more is not better." Excess camber
14 12
can result in difficulties in achieving level floors with the
* Inquire All Sections 300 lb per ft
* For grades other than A36 or lengths 60 ft 0 in., and
longer maximuns are available on inquiry for the follow-
ing sections:
W14 x 26
22
Minimum and Maximum Camber Inches
Nominal
Depth Over 20 to 30 Over 30 to 40 Over 40 to 52 Over 52 to 65 Over 65 to 85 Over 85 to100
(Inches) incl incl incl incl incl incl
24 and over* Inquire 1/2 to 1-1/2 incl 1/2 to 2-1/ 2 incl 1 to 4 incl 1 to 5 incl 1 to 6 incl
14 to 21 incl* Inquire 1/2 to 2 incl 1/2 to 3 incl 1 to 4 incl 1 to 5 incl Inquire
4 to 13 incl* Inquire 1/2 to 2 incl 1/2 to 3 incl Inquire Inquire Inquire
The preceding table provides reasonable guidelines affected by the actual elevation of the erected steel and
for minimum and maximum induced cambers. Obtaining are more seriously affected by high points in the steel
larger cambers on lighter weight beams with shorter than by Iow ones.
lengths, particularly for grades other than A36, is more In the constant thickness method, the finished floor
difficult. Therefore, it is prudent to consult the producer follows the steel below. High points in the steel cause
prior to specifying cambers near these extremes, high points in the finished slab, which may hinder the
installation of interior finishes.
In the constant elevation method, the finished floor is
The maximum amount set using a level. High points in the steel can result in
inadequate slab thickness.
of Camber t h at Can be p u t Therefore, prior to placing concrete slabs, beam ele-
vations should be verified. Then, if the expected floor
into a member is limite and profile is not satisfactory, the finishing approach can be
modified.
is dependent on its cross sec-
Summary:
tion, le' h and material Cambering is often the most economical method of
handling dead load deflections in beams. It saves money
graue, by reducing the excess concrete that may be required.
The cost of cambering can be accurately determined with
no additional hidden costs to consider.
The minimum amount of camber is dependent on
The amount of camber should be specified only after
both physical and economic considerations. Cambers of considering the following items:
1/2' or less should probably be avoided. At 1/2' the cost 1. Calculated dead load deflection.
of cambering usually exceeds the potential savings in
2. Connection end restraint.
concrete, especially since natural mill camber will proba- 3. Mill tolerances and camber losses.
bly be present. Also, below 1/2' the permanency may not
be assured. 4. The "More is Better" Syndrome.
5. Camber limits.
The maximum amount of camber that can be put into
member is limited, and is dependent on its cross section, f
length and material grade. Cambering is often the
Availability:
most economical method of
Cambering is available from the producing mills and
generally adds onlytwo weeks to the delivery of material, handling dead load deec-
Many fabricators also have the capability and expertise
to offer cambering, tions in beams.
Cost:
Most mills published price book offers cambering for Reference:
$0.03/Ib on beams up to 50 lbs/fi, and $0.02/Ib for beams
over 50 lbs/ff. "Economical Use of Cambered Steel Beams," by J.W.
Larson and R.K. Huzzard, Bethlehem Steel Corporation,
E f f e c t of Construction Methods: presented at the AISC National Steel Construction
Conference, March 1990.
Both methods used in the finishing of concrete slabs,
constant thickness and constant elevation, are greatly
Credit:
This TIPS is reprinted from a Bethlehem Steel Corp. Technical Bulletin titled
"Economical Use of Cambered Steel Beams."
TECHNICALINFORMATION& PRODUCTSERVICE
' t _ /
i j I . 7 i : : : : '
! . ,:- '3
-)
':-- " .C..
c.....
' ]
APRIL 1992
,:'% . = j
/
{ ?r'-'j ::'" :' i"
.... :'", i'" ] : " : : : : i ! 9 : : :
By W.A. Thornton, Ph.D., P.E.
specially in today's climate of reduced
construction activity, it is important to
do everything possible to reduce costs.
Through the careful design of structural
connections, fabrication and erection costs can
be reduced.
Bracing connections constitute an area in
which there has been much disagreement con-
cerning a proper method for design. Research
conducted during the past decade is just now
being distilled into a consistent method of de-
signing connections based on equilibrium mod-
els for the gusset, beam, and column that re-
quire that yield not be exceeded globally on any
gusset edge or section, and also on any section
in the column or beam.
Careful selection of. connections
can substantially reduce the fabri-
cation and erection costs on many
steel construction projects
While there are many possible equilibrium
models, three are presented here and then
applied to the design of a connection to deter-
mine their cost-effectiveness.
Mo dsl I . This is the most common and
simplest of all equilibrium models. The force
distribution on the gusset, beam, and col-
umn are shown in Figure 1. As with all
equilibrium models, this model guarantees
that the gusset, beam, and column are in
equilibrium under the brace load P. If the
work point coincides with the gravity axes of
V p
H
W.P.
/
, , / IR
Mc .
V p
',,._.,. Mil-Hee
EQUILIBRIUM MODEL
I
Fig u rc Z: Model 1, the simplest equilibrium model
V V
e P p
J EQUILIBRIUM
ec ,, 2 _
Figure 2: .Model 2, one of several z4.[SC.Models
r ' , I I II mi I lira
w.P
R-'N
e
; , l
R
H,- H-H V,-O
Vc-V Hc= 'cV
e
c =
Figure 3: Force Distributions for Model 2
i c - E Q U I L I B R I U M M O D E L
-, 5
Fb;u;'c 4: Model 3, an equilibrium model with no couples
V p
I [ J
w.P
the members, equilibrium is achieved with
'no connection induced couples in the beam,
column, or brace. Model 1 has been referred
to sarcastically as the "KISS" method (Keep
It Simple, Stupid).
,, [,odel 2 (AISC Model). This model is oneof
several adopted by AISC based on the re-
search during the past decade. The force
distributions for the gusset, beam, and col-
umn are shown in Figures 2 and 3, and as
with Model 1, these force distributions guar-
antee that the gusset, beam, and column are
in equilibrium under the brace load P. If the
work point coincides with the member grav-
ity axes, equilibrium is achieved with no
connection induced couples in the beam,
column, or brace. Model 2 is a little more
complex for calculations than Model 1, but it
yields less expensive designs.
Model 3. This model is the result of the
author's search for an equilibrium model for
W.P
R'\
I R v'-h
--F---; vZ' J i
Hs--P Ye-
r-/(,+ q:.,8.e,F.
Figure 5: Force distributions for Model 3
WlS,,106 -"
352__...
.;',,'"' J; w,4,6o5
Figure 6 Datafor illustrative example connection
bracing connections that achieves equilib-
rium for all components of the connection
with linear forces only, i.e., no couples. It is
the most efficientit yields the least expen-
sive designs---of the three models presented
here but is also the most complex in terms of
calculations required. Note, however, that
this is not a serious problem because a
computer program makes the calculation
aspect of all three models of little impor-
tance. The force distributions for the gusset,
beam, and column of Model 3 are given in
Figures 4 and 5.
The beam shear R in Figures 4 and 5 is
shown applied to the beam-to-column connec-
tion, If the shear is large, it may be desirable to
distribute it to the gusset-to-column connec-
tion as well. In this case the gusset serves as
a haunch and the gusset-to-beam forces must
be adjusted to effect the desired distribution of
R.
2 Steel Tips Apri l 1992
2'/. 4 -t z
WiSx 106 3 s,ots *WI 4x605
2- / . 6x 6 I I -
Fi.,;:tn' 7: ;olution to example connection using Model I
2- / Bx 6' 1 '--, 8
.
BOLTS 9 _,,,
2-t 4x4xSed
Figure 9.' Solution to example connection using Model 3
2- / . 6x6 x I ---,
I Wl 8xl O 6 3 sots
BOLTS,- *';,'f% "' /
HOLES:STD 3 s , o . s - ' s,.__,/
I d " 2-L6x4x,,
i
k
F
I 4 ' 6 0 5
I I I
Fig:ire 8: Solution to example connection using Model 2
Figure 6 provides an example. The column
is a W14x605, the beam a W18x106, and the
brace a W12x87 with 450 kips.
Figures 7, 8, and 9 give the completed
designs for Models 1,2, and 3, respectively. A
cost comparison shows that Model 3 gives the
most economical design, while Model 1, the
"KISS" method, gives a design that costs ap-
proximately 28-30% more and Model 2 gives a
design that costs approximately 13% more.
Using a lighter column section, a W18x119, to
assess the effect of drilling the heavy flange of
the W14x605 reveals similar results.
To see the effect on a project of using Model
1 rather than Model 3, consider a 40-story build-
ing with eight bracing connections per story. If all
these connections were similar to those shown
in Figure 6, the cost of using Model 1 rather than
Model 3 would be (840-658) x 8 x 40 = $58,240.
i i i i i i i i
Fisurt' 11): Same cost stiffened and unstiffened column
No transverse beams
iCJ
i i ,
., 0
WI4IO9(A36) W14x90 ( A36)
(Fillet welds)
This is assuming one bay of bracing on each of
the four faces. If two bays per face were used,
the extra cost of Model 1 would be about
$116,0O0.
Columns, when part of an unbraced frame,
are designed for bending moment as well as
axial force. The designer uses a rigid frame
analysis computer program, which also possi-
bly does a code check using the beam column
interaction equations or he performs the latter
operation manually. What the designer gener-
ally does not consider in his column design is
the "panel zone" between the column and the
transverse framing beams and this can be a
costly oversight.
Figure 10 shows a W14x90 column 34'-long
with fillet welded stiffeners and a same cost
W14x109 with no stiffeners. However, if a
Wl 4x99 column will work, a less expensive job
will result. The W14x109 also may be less
expensive if extra erection costs associated
with beams framing to the weak axis of the
W14x90 due to the stiffeners are considered.
Figure 11 shows the same W14'x90 column as
Figure 10, but here the designer has specified
full penetration groove welds of the stiffeners to
the column. This doubles the cost of the stiffen-
ers and means that an unstiffened W14x132
will cost about the same as the stiffened W14x90.
Now, looking at the sections between W14x90
and W14x132, we see that we have available a
W14x99, a W14x109, and a W14x120, all of
which will yield a less expensive design if they
satisfy the beam column design equations.
Figl.,'c i : Same cost stiffened and unstiffened columns
No transverse beams
P L zx7xl'-oSe
WI490 (A36) WI4xt32(A36)
(Full penetration welds)
Fief, rc I2: Rules of thumb: same cost columns
no transverse beams
- Pti:,x7xI:OSe
PL'4xll xa'-IO
S' CTION A-A
SECTION B-E}
Wl4x90 (A36)
I ll.
W14x145(A36)
4 doubler plates + 4 pairs of stiffeners = 1900 lbs steel
(Fillet welds)
Figure 12 shows the "fabricators nightmare"
of stiffeners and doublers. A clean W14x145
costs no more than the stiffened and doubled
W14x90, and all of the W14 Sections in be-
tween will give less expensive designs if they
satisfy the beam-column equations.
For the convenience of designers, Figure 13
gives the cost in lbs. of steel, as well as the cost
of column splices. Column weights can be
increased by approximately the amounts shown
here without increasing costs because, as pre-
viously mentioned, the stiffeners and the dou-
blers will tend to increase erection costs. (Note
that erection costs are not included in Figures
10 through 13.)
Figure 14 takes a different view. here the
connection with the stiffeners and doublers is
given per tributary length of column. As an
example, Figure 15 presents a W24x55 framing
to a column flange. The design moment is M =
212k-ft, which is just slightly less than the full
strength moment of the W24x55(A36), which is
226k-ft. The W14x90 column, which is deter-
mined to be adequate for M = 212k-ft and the
design axial load, requires stiffeners and dou-
blers. The W14x120, which is also adequate for
the design moment and axial force, requires no
stiffeners or doublers. Since 120- 90 = 30 lbs,
which is less than the 79 lbs from figure 14, the
W14x120 is the more economical choice. As
Figure 15 shows, $180 is saved per connection.
If there were 1,000 similar connections on the
job, savings would be approximately $180,000.
The stiffeners and doublers of the column
cost studies previously discussed are the result
( f i l l e t welded)
(full penetration welded)
One Pai r
Doubler Pl at es = 550 lbs steel
One Doubler Pl at e = 280 lbs steel
One Col umn Spl i ce = 500 lbs steel
Length of Increasein weight per foot with
column no increase in cost of "clean"
tributary to column
d connection StiffenersFillet Stiffeners
i with stiffeners Wel ded GrooveWelded
and doublers
._._.._ 10 95 135
12 79 113
4 Si' ]I:FENERS
14 68 96
2 DOUa.ES
16 59 84
NOWEAKAXiS
B Co.Ec.o 18 53 75
20 48 68
Column Selection Design Aid
of requirements for beam-to-column moment
connections, especially when full-strength mo-
ment connections are specified, as in Figure 16
for doublers. Since stiffeners and doublers can
add significant costs to a job, design engineers
should not specify full-strength moment con-
nections unless they are required by loads or
codes, e.g., ductile moment resisting frames for
high seismic loads.
For wind loads and for conventional mo-
ment frames where beams and columns are
sized for stiffness (drift control) as much as for
strength, full strength moment connections are
not required. Even so, many design engineers
will specify full strength moment connections,
adding to the cost of a structure.
Designing for actual loads has the potential,
without any increase in column weight, to dras-
tically reduce the stiffener and doubler require-
ments. On one recent 30-story building, a
change from full moment connections to a de-
sign for actual loads combined with using Fig-
ure 17 for doublers reduced the number of
locations where stiffeners and doublers were
required to several dozen from 4,500 locations
with an estimated cost savings of approximately
$50O,OO0.
The uniform design load (UDL) is a great
crutch of the engineer because it allows him to
issue design drawings without putting the beam
reactions on the drawings. Instead, often the
fabricator is told to design the beam end con-
: Example of use of column selection design
STORY HEIGHT 12'0
CokJrnn
wi cok.
Since 120 - 90 = 30 lbs. < 79 lbs.,
Saved 79 - 30 = 49 lbs./fi, x $.30/Ib. = $15/ft.
Therefore, per 12' of column, $15 x 12 = $180 saved
Building with 1,000 locations = $180,000 saved
:: . . . . . : Design for the full strength of the beams
db
M,
L / I M I
dc _
t =
WIEQ'D ' c
M2 FULL STRENGTH MOMENTS
Doubler Plates
Commonly Seen Requirements
nections for one-half UDL, or some other per-
centage to account for composite design, un-
less greater reactions are shown. Unless con-
centrated loads are located very near the beam
ends, UDL reactions are generally very conser-
vative. Because the reactions are too large,
extremely strong connections, such as double
framing angles, will often be required.
Single angles, because the bolts are in
single shear, will have about half the strength of
double clips for the same number of rows of
bolts. But if actual reactions are given, it will
almost always be found that a single angle
connection will work, perhaps with a couple of
extra rows of bolts.
Figure 18 is part of an industrial building with
dead Icad of 140 psf and live Icad of 250 psf.
Beam 1 of Figure 18 is shown in Figure 19. The
total Icad on Beam 1 is 82 kips and the actual
reactions are thus 41 kips. The one-half UDL
reaction is 45 kips, which is pretty close. Now
look at the connections. The minimum double
clip connection on this coped beam has four
rows and is good for 81 kips, almost twice the
actual reaction. Many designers routinely re-
quire "full depth" connections, i.e. six rows. The
six row double clip connection is good for 116
kips, almost three times the actual reaction.
However, a five row single angle is good for 52
kips, which is okay for the actual and the one-
half UDL reactions.
As this example illustrates, single angles
will work even in heavy industrial applications,
and they are much less expensive than double
clips, especially for erection. In Figure 20, the
connections for this W24x55 beam have the
same strength and have a differential cost of
$10 for fabrication. But, including erection, the
single angle beam costs approximately $25
less than the double clip beam. For a 30-story
building 200' x 200' with 25' bays and 200
beams per floor with tabs, there is a savings of
200 x 30 x 25 = $150,000.
Returning to Figure 18, suppose Beam 1 is
subjected to the same Icad of 82 kips total, but
32 of the 82 is a concentrated Icad located at
mid-span, such as from a vessel. Figure 21
shows the actual reaction of the beam, now a
W24x76, is still 41 kips, while the one-half UDL
reaction is 56 kips--which is 37% greater than
the actual reaction. This means while a five row
Figure 17: Design for the actual loads
Doubler Plates
' -Vt = Couu x,,4o S.eAR
t = 7 32 .95cl b
Will often eliminate doubler requirement Figure 18: Partial plan o[ industrial building fioor
Figure 19: Comparisons W21 x68 / - -3.25k/ft
for Beam I
k 25'-0 )
min.# CAP max# CAP # of CAP
kips/ft, kips kips kips kips of rows (kips) of rows (ki ps) rows (kips)
1 W21x68 3.25 0 82 41 45 4 81 6 116 5 52
Bolts 7/8 A325N, Clips 4 x 31/2 x /8, Welds 1/4" fillet
J".C,.=..;;:.-Z;...Tz:;';,='TT-"L=;--=-;'-= LJ--/ ~: . . . . .' 7 = - . ; = i i - : : :7. V: :%2: !7.,>.T:/i'J..'. :!??JUF::`.J;`-*va`7;` . . . . . .
. , :, : . Cost of same strength single and double clips
SINGLE CLIPS
W24,55
L 0' -0
I
V
DOUBLE CLIPS
24,55
k
I
Fabrication - $10 per beam less for single clips; Erection - $15 per beam less for single clips
Total Cost Reduct i on - $25 per beam usi ng single clips
Fi,urc 21: Comparisons
for Beam 1 (prime)
32K
t F
min.# CAP max.# CAP # of CAP
,1 (prime) W24X76 2 32 82 41 56 4 83 7 137 5 52
Bolts 7/8 A325N, Clips 4 x 31/2 x %'8, Welds t / 4 " fillet
Figm'e 22: Comparisons
for Beam 2
82 K 82K
I - I
Uniform Conc. Total Actual 1/2UDL Double Clips Single Clip
min# CAP max,# CAP #of CAP
2 W33x118 0 82 164 82 114 6 150 9 210 8 92
Bolts 7/80 A325N, Clips 4 x 31/2 x 3/8, Welds ,5'16" fillet
single angle connection is okay for the actual
reaction, a six row connection with a capacity of
66 kips would be required for the one-half UDL
reaction.
Figure 22 shows the disparity between ac-
tual and one-half UDL reactions for Beam 2.
Again, single angles are sufficient.
peared in the AISC magazine "Modem Steel
Construction. "A complimentary subscription to
"Modem Steel Construction" may be obtained
by contacting AISC, Chicago.
Steel Tips April 1992 7
S'7 TM ' '-'-I " :
auL, 1 J R A L S T E E L E D U ' " "'-:'"'x' "'
470 Fernwood Drive
Moraga, CA 94556
(510) 631-9570
SPONSORS
Adams & Smith
G.M. Iron Works Co.
Hoertig Iron Works
Hogan Mfg., Inc.
Junior Steel Co.
Lee & Daniel
McLean Steel, Inc.
Oregon Steel Mills
pDM Strocal, Inc.
Reno Iron Works
H.H. Robertson Co.
Stockton Steel
Stott, Inc.
The local structural steel industry (abovesponsors) stands ready to assist you in
determining the most economical solution for your products. Our assistance can
range from budgetprices andestimated tonnageto cost comparisons, fabrication
TECHNICAL INFORMATION& PRODUCTSERVICE
Value Engineering
And Steel Economy
By David T. Ricker, P.E.
e've all heard the aphorism, "There's

more than one way to skin a cat." This
expression is especially applicable to steel
design and the steel construction indus-
try, where many alternate approaches will result in
a safe and economical building. Designers often
have several choices as to how to approach a given
design task. Likewise, steel fabricators/erectors
have choices of such items as connection types,
fastening methods and fabricating processes.
Some methods appeal more to one group than
the other, and it is up to the design engineer to apply
common sense and engineering judgement in choos-
ing the best solution for a specific project. This
article includes tips and suggestions for engineers
to design safe and economical steel structures.
1. Keep abreast of current costs of various
steel product s used in structural design. A steel
Figure 1: Select the optimum bay size
! l l l i F L i
Z Z
w
.&.
AUGUST 1992
An experienced engineer
and fabricator offers 35 tips
on reducing fabrication costs
during the design stage
i i i
fabricator can supply current base prices upon
request. Designers should also be aware of where
the money is spent on steel construction: one-third
on material; one-third on fabrication; and one-third
on erection.
2, Take advantage of al l owabl e stress in-
creases permi tted by AISC Speci fi cati on A5.2
for temporary loads such as earthquake and wind.
3. Take advantage of live-load reducti ons if
governing codes permit.
4. Sel ect a proper mi x of A36 and hi gh-
strength steel s. High strength steels are advanta-
geous when strength is the major design criteria.
While A572 Grade 50 is more expensive than A36
steel, it also is more than 35% stronger. However,
when deflection, stiffness or some other service-
ability criteria governs, the nod still may go to A36
because the heavier sections required will generally
have a higher moment of inertia. With the gradual
narrowing of the cost difference between A572 and
A36, the added strength often will outweigh the
modest cost premium.
I I I I I I
Figure 2: Snow-drift loading
i
.._L__ .... / .
Figure 3: Column splice location
i
[ _ _I L I c E
L , I
L. OIt
%
CPRESIOkl
MOM.NI'
5. Avoid odd sections that may not be readily
avai l abl e or which are seldom rolled. These could
result in costly delays. Fabricators can assist in
identifying troublesome sizes and shapes.
6. Consideration should be gi ven t o the use
of parti al composite design of floor beams--
something in the range of 50% to 75%. Full compos-
ite design is inefficient. The cost of one shear stud
in place equals the cost of approximately 7 lbs. of
steel. Unless this ratio can be attained, the addition
of more studs will prove uneconomical.
7. Sel ect optimum bay lengths. An exhaustive
study by John Ruddy, P.E. (AISC Engineering Jour-
nal, Vol. 20, #3, 1983) indicated that a rectangular
bay with a length-to-width ratio of approximately
1.25 to 1.50 and a bay area of about 1,000 sq. ft. was
the most efficient. The filler members should span in
the long direction (see figure 1).
8. Space floor beams so as to avoid the
necessity for shoring the deck during the concrete
pour. The cost of shoring is relatively expensive and
can be easily offset by varying the span, gage, or
depth of the floor deck.
9. Tai l or the surface preparation and pai nt-
i ng requirements to the project condi ti ons--do not
under-do nor overdo the coating requirements. Also
consider the appropriate treatment of connections.
10. Show all necessary loads on the design
drawing to avoid costly over-designing of connec-
tions or dangerous under-designing. This is an
AISC specification requirement. In addition to grav-
ity loads, torsional loads, axial loads, and moments
should be given, and when columns are influenced
by gravity and lateral moments, these moments
should both be given.
11. Indicate who is responsible for "grey
area" items such as loose lintels, masonry an-
chors, elevator sill angles, elevator sheeve beams,
fastenings for precast concrete spandrel panels,
etc. Unl ess the responsibility is specifically del-
egated, it is likely that the cost of these items will be
included in the bids of two or more trades, meaning
the client may pay more than once for the same article.
12. Don't require the steel subcontractor t o
perform other work that should be done by other
trades such as installing masonry anchors, ceiling
hangers, lateral bracing for interior wall, toilet parti-
tion supports, window wall supports, and the like.
Information required to perform this work often is
slow to develop, resulting in needless delay for the
steel fabricator, who should be allowed to proceed
without del ay and unnecessary encumbrances.
13. Consider the use of cantilevered rafters
and purlins to save wei ght on roof design.
14. In areas of roof t hat are subject to snow
dr i f t l oadi ng, arrange the purlins parallel to the drift
and vary the spacing of the purlins so the same gage
of roof deck and same size purlins canbe used
throughout the area (figure 2).
2 Steel Tips August 1992
15. Do not desi gn for minimum wei ght alone.
Such a design may require more pieces and more
connections and will be more labor intensive in both
the shop and the field, and in all likelihood will be
more expensive.
16. Excessively stri ngent mill fabrication and
erection tolerances, beyond state-of-the-art prac-
tices, will probably reduce the number of bidders
and raise the cost to the owner. ASTM A6 toler-
ances and those established by the ANSI, AWS,
and AISC have served the industry well and should
be adhered to except under extraordinary circum-
stances.
17. Designate the proper type of high strength
bol t value. The correct application of each type is
well documented in the current bolt specifications.
Do not specify slip-critical values for the purpose of
obtaining an extra factor of safety.
18. Al l ow the use of tensi on control (twist-
off) hi gh strength bolts. These bolts are as reli-
able as are other methods of measuring bolt tension
and reduce labor costs.
19. Speci fy fi l l et wel ds rather than groove
welds, wherever possible. Groove welds are gen-
erally more costly because of the joint preparation
required and the greater volume of weld deposited.
20. Indi cate fi l l et wel ds that can be made in
one pass, wherever possible. When using the
shielded metal arc welding process in the horizontal
and flat positions, the maximum practical one-pass
fillet weld is 5/16".
21. Favor the horizontal and flat wel di ng
positions. These welds are easier and quicker to
make and are generally of a higher quality.
22. Don' t call for more weld than is neces-
sary. Over-welding creates excessive heat, which
may contribute to warping and shrinkage of the
members resulting in costly straightening expense.
23. Grant the fabricator the opti on of elimi-
nati ng some column splices, wherever possible.
The cost of one column splice equals the cost of
about 600 lbs. of A36 steel. However, the fabricator
Figure 4: Heavier column shaft eliminates stiffeners
STIFFENER,
bEAi ,....-
e.S --
f Cot..u,4 N
should study the situation carefully before deciding
to omit the column splice and running the heavier
shaft up to the next splice--the resulting column
may be too long for erection.
24. Avoid designing column splices at mid-
st ory height. These are often too high for the
erector to reach without rigging a float or scaffold. If
the splice can be located no higher than 5' above the
tops of the steel beams, it will save the expense of
the extra rigging and still be in a region of the column
where bending forces are relatively Iow (figure 3).
25. Do not design column splices t o "de-
vel op the full bending strength of the smaller
shaft." Seldom is the splice located at the point of
maximum bending and seldom do the bending
stresses result in a condition that would require a full
strength splice. The column has axial compressive
stresses. The excess capacity is allotted to bending
stresses that occur as compression in one flange
and tension in the other flange. The compression
forces are added to each other at one flange while
at the other flange the tension force is subtracted
from the compression force. Seldom does this other
side of the column ever go into tension and never
into full allowable tension of the magnitude that
would require a full strength splice. Once in a while,
lateral loads on a structure will result in some small
Steel Tips August 1992 3
. . . . . . . . . . I I I II
Figure 5: Stiffener Location
colu n J
A B C D E
-.-. 7, tE
I
LOAD
.F
Figure 6: Bracing Work Points
web
A E5
~ .
net tension stress, but not to the extent to justify a full
strength splice (figure 3).
26. Consider using a heavier column shaft or
high strength steel to eliminate the need for web
doubler plates and/or column stiffeners opposite
the flanges of moment connected beams. One pair
of stiffeners installed costs approximately the same
as 250 lbs. of A36 steel if the stiffeners are fillet
welded. If they must be groove welded, the cost
skyrockets to the equivalent of 1000 lbs. of A36
4 Steel Tips August 1992
steel. The cost of one installed doubler plate is about
the same as 500 lbs. of A36 steel (figure 4). Consid-
ering that for an average two-floor column there
could be as many as four pair of stiffeners and two
doubler plates, at least 2,000 lbs. of A36 steel (about
1,900 lbs. of A572 Grade 50) could be sacrificed in
order to save the time and expense of making the
lighter shaft work.
27. Avoid designing heavyor awkward mem-
bers in remote hard-to-reach portions of the struc-
ture. This may eliminate the need for larger, more
expensive hoisting equipment.
28. Reinforce beam web penetrations only
where necessary. Several software programs, in-
cluding AISC's Webopen, help identify and design
web penetrations that do not require reinforcing.
Also, it may be less costly to use a beam with a
heavier web, to move the opening to a less critical
location, or to change the proportions of the opening
to something less demanding.
29. For heavy bracing, where advantageous
allow the fabricator to translate the bracing work
points so that they lie on the intersection of the flange
faces rather than the centerline of members (figure
6). Generally, this will result in a more compact,
efficient connection (see AISC Engineering Jour-
hal, Vol. 21, 3rd Quarter 1984).
30. Al l ow the prudent use of oversized holes
and sl ots to facilitate fit-up costs and erection.
They may eliminate or reduce the need for costly
reaming of holes or re-fabrication.
31. Avoid the i ndi scri mi nate use of stiffen-
ers. Stiffeners are required to prevent local defor-
marion and/or to transfer load from one part of a
member to another (figure 5). If the main members
are capable of taking care of themselves then the
cost of stiffeners can be saved. Use partial depth
stiffeners where possible; full depth stiffeners can
cost twice as much as partial depth stiffeners.
32. Avoid the catch-all specification that reads
something like this: "Fabricate and erect all steel
shown or implied necessary to complete the steel
framework." The bids will undoubtedly be inflated to
cover whatever might be "implied". This is unfair to
the client.
33. Avoid the nebul ous speci fi cati on that
calls for stiffeners as required, roof frames as re-
quired, reinforcing of beam web penetrations as
required, etc. During the estimating and pricing
phases, the fabricator/erector rarely has enough
time to determine what is and what is not required,
and will therefore include in the bid an allowance for
the questionable items whether or not they are
eventually needed. The client may end up paying
for something that is not supplied. Spell out known
stability and tolerance problems applicable to the
specific project.
34. Avoid the overl y restrictive specifica-
tion. The more restrictions listed in the steel speci-
fications the greater the chances that no one will be
able to meet them all. This will eliminate some of the
competition and often result in higher bids.
35. Prepare complete desi gns. A goodly por-
tion of the fabricator's overhead expense is spent
on estimating. For every job he gets he may make
10 to 20 unsuccessful attemps. Some project speci-
fications are written in such a way as to require the
fabricator to complete significant portions of the
steel design in order to prepare an accurate cost
estimate. But requiring fabricators to complete sig-
nificant portions of the steel design in order to
prepare an accurate cost estimate wastes time and
boosts estimating costs, which will in turn be passed
on to the client. If eight fabricators are pursuing a
project, this design work must be duplicated eight
times. The fabricator who elects to bypass the
necessary design investigations may end up in
deep trouble with a bid and job that are less than
they should be. But if he loads his bid to cover the
"worst case" conditions, the chances are he will not
get the job. When a fabricator has to determine
loads via analysis of the member, the results will
often be unrealistic. Whether the member is de-
signed based on stiffness, aesthetics, minimum
thickness requirements, or deflection is not nor-
mally known by the fabricator.
A complete design is the best assurance that
those who must use that design will accurately
interpret the intent of the designer. There will be far
less chance for ambiguities, misinterpretations, er-
rors and/or omissions. Design shortcuts can only
hurt the other members of the construction team. A
complete design benefits everyone in the long run,
including the designer and the client.
David T. Ricker, P.E., is a retired vice president
of engineering with AiSC-member The Berlin Steel
Construction Co., Inc, and is now living in Ari-
zona.
peared in the AISC magazine "Modem Steel Con-
struction.' A complimentary subscription to "Mod-
ern Steel Construction" may be obtained by contact-
ing AISC, Chicago.
Steel Tips August 1992 5
470 Fernwood Drive
Moraga, CA 94556
(510) 631-9570
SPONSORS
Adams & Smith
G.M. Iron Works Co.
Hoertig Iron Works
Hogan Mfg., Inc.
Junior Steel Co.
Lee & Daniel
McLean Steel, Inc.
Oregon Steel Mills
PDM Strocal, Inc.
Reno Iron Works
H.H. Robertson Co.
Stockton Steel
Stott, Inc.
Vuicraft Sales Corp.
<.. . ,- :, , , . , - . ..?4s-..--c::..'?t,:: '
': The local:structuratsteeI industry'?v'i61sPonsors)sands 'ready to assist You in,'
rComparisons,; fabncatiOn ','.
>
STEELCOMMITTEE OFCALIFORNIA
TECHNICAL INFORMATION & PRODUCT SERVICE
FEBRUARY 1986
UN FIRE PROTECTED EXPOSED STEEL
PARKING STRUCTURES
INTRODUCTION
The key to construction economy in a
mul ti story parking deck is its structural
framing system. From 55 to 65 percent of the
total cost of most parking structures is the
cost of structural components -- framing,
fl oor slabs, and foundations. It is apparent,
therefore, that the selection of the structural
framing system is oneof t he most i mportant
economic decisions involved in the design
of a parking structure.
Un Fire Protected exposed steel was pro-
hibited by restrictive building code provi-
sions, ori gi nal l y developed for enclosed
garage buildings, that required fire-rated
constructi on in all parking structures. Every
steel frami ng member had to be "fi re-
proofed", at extra cost, even in open parking
decks.
The 1985 Uniform Building Code recog-
nizes the results of several recent fire tests
in open parking garages so that limitation
on area and height of non-fire protected
steel parking structures has been relaxed.
Changes have come about largely due to a
test sponsored by A.I.S.I. in cooperation
with the Scranton, Penna. Parking Authority,
the Scranton, Penna. Fire Bureau and Under-
writers Laboratory. The new code now allows
constructi on of Un Fire Protected steel
framed open parking structures of unlimited
area with certain restrictions noted later.
The elimination of fire-protective encase-
ment has reduced the cost of open-deck
steel parking structures by 7-10 percent. No
similar reduction has occurred in the cost of
concrete parking structures, because con-
crete constructi on is essentially the same
under both new and old codes. As a result,
prestressed and reinforced concrete parking
structures that once had been competitive
with fireproofed steel structures are now
significantly more expensive than today's
exposed steel structures.
No owner or designer can afford to ignore
the potential savings inherent in steel con-
struction f orparki ng structures. In 1986 it's
a new deal with a new structure -- an
exposed steel structure -- that offers Iow
cost and greater value than any other
parking structure.
GENERAL
An open parking garage is a structure of
Type I or Type II construction with openings
on two or more of its sides. Area and height
limits for open parking garages are given in
the table below, re-printed from the 1985
Uniform Building Code,
Type of Area Per Height
Construction Tier Ramp Access
I Unlimited Unlimited
II-FR 125,000 12 Tiers
I1-1 HR 50,000 10 Tiers
II-N 30,000 8 Tiers
In order to be considered "open" the
structure must meet the following:
1. It must be open on two or more sides
having an open wall area on each tier of 20%
or more of the total perimeter wall area of
each tier.
2. The total length of the openings must
be at least 40% of the perimeter.
AREA&HEIGHTINCREASES
Area & height increases over that allowed
by the above table are permitted if:
1. The structure is open on three fourths
of the perimeter, a 25% increase in area per
ti er and one ti er increase in height is per-
mitted.
2. The structure is open around the entire
perimeter, a 50% increase in area per ti er
and one ti er increase in hei ght is permitted.
These area & hei ght increases allowed
when the structure is "open" on three or four
sides result in the fol l owi ng table for Type
I1-1 Hr & Type II-N constructi on. Al so in-
cl uded in the tabl e is the parking capaci ty
based on a3 area of 300 sq. ft. per car.
Type o! Area Per Gross No. of
Construction Tier Height Area Cars
II-N (Open 3 Sides) 37,500 9 Tiers 337,500 1,125
II-N (Open 4 Sfdes) 45,000 9 Tiers 405,000 1,350
I1-1 HR (Open 3 Sides) 62,500 11 Tiers 687,500 2,292
H-1 HR (Open 4 SJdeS) 75,000 11 TJer$ 825,000 2,750
If the structure is constructed to heights
less than the number of tiers shown in the
above table the fol l owi ng applies:
1. The area per tier may be increased
provided the gross area does not exceed
that allowed for the hi gher structure.
2. At least three sides of each larger tier
must have openings at least 30" high for at
least 80% of the length of the side.
3. No part of the larger tier can be more
than 200 feet f rom an opening.
4. Each opening must face a street oryard
accessible to a street wi th a width at least 30
ft. for the full length of the opening.
5. Standpipes must be installed on each
tier.
Si nce publ i cati on of the 1982 UBC the
code has permi tted constructi on of open
parki ng structures of Type I I-FR, Type I I-lH
or Type II-N of unlimited area. The fol l owi ng
restructi ons apply.
1. The hei ght cannot exceed 75 feet.
2. All sides must be open for at least 50%
of the exteri or area of each tier.
3. The openings must be equally dis-
tri buted al ong each side.
4. No part ofatiercan be more than 200 ft.
f rom an openi ng.
RECENTREVISIONSTOTHE1985UBC
At the 1985 ICBO Annual Business Meet-
lng the fol l owi ng changes relative to open
parki ng structures were approved:
1. An open parki ng structure of Un Fire
Protected steel structure may be constructed
above an enclosed basement or first story
used for the parki ng or storage of motor
vehicles.
2. The di stance to adjacent property lines
f rom exteri or non fire rated walls has been
reduced f rom 20 ft. to 10 ft. and allows a one
hour fire rated wall less than 10 ft. and more
than 5 ft. Thi s will increase the instances
when the designer can get more sides open
thus al l owi ng greater use of unprotected
steel.
Both of these changes will appear in the
1986 suppl ement to the Uniform Building
Code.
THE STEEL COMMITTEE OF CALIFORNIA
Ace & Stewart Detailing, Inc.
Artimex Iron Co., Inc,
Baresel Corporation
C A Buchen Corporation
Butler Manufacturing Co,
Central Industrial Engineering Co.,
Inc.
Cochran-lzant & Co., Inc.
Dovell Engineering, Inc.
Hoertig Iron Works
Hogan Mfg., Inc.
Industrial Steel Corp.
Inland Steel Company
Inryco, Inc.
Junior Steel Co.
Kaiser Steel Corporation
Lee & Daniel
McLean Steel, Inc.
Pascoe Steel Corporation
Pittsburgh-Des Moines Corporation
Riverside Steel Construction
H.H. Robertson Co.
Ross-Carter Corporation
Stockton Steel
Stott Erection, Inc.
USS Fabrications
United States Steel Corporation
Western States Steel, Inc.
Northern California X Southern California
1250 Pine St. No. 301 [ 9440 Telstar Ave. No. 103
Walnut Creek, CA 94596 El Monte, CA 91731
(415) 932-0909 (818) 444-4519
Fundi ng for this publication provided by the California Field Iron Workers Administrative Trust.
STEEL HIGH-RISE BUILDING FIRE
OCCIDENTAL CENTERTOWER BUILDING, LOS ANGELES, CALIFORNIA
OCCIDENTALTOWERONNIGHTOFFIRE
November 19, 1976
COVER
On November 19, 1976, a fire broke out on the 20th floor of the 32-story Occidental Center
Tower. The building s a structural steel frame with floor beams supporting a composite steel
deck floor system and partbons of steel studs covered with steel lath and plaster.
The fire was detected shortly after 3'00 a m. and was under control within an hour Approx-
tmately 40% of the 20th floor was Involved in the fire, and almost half of this space was
burned out.
The structural steel frame and the steel deck suffered no damage. There was no loss of hfe,
although five firemen sustained minor injuries in the fire. Damage m the fare area to partlhons,
cedmgs, mechamcal and electrical systems, exterior aluminum panels and fire protechon
coating on structural members was estimated to be in excess of $1,200,000.
THE STEEL COMMITTEE
OF
s t e 171 Second Street, Room 402, San Francisco, CA 94105
NORTHERN CALIFORNIA
(415) 781-2803
ERRATA -- "Steel Hi,h-Rise Butldin Fire"
Pages 10 and 11: The photographs in Figures 6, 7 and 8 should be rotated
one position clockwise with respect to the captions. In other words,
the photo over Fi. 8 should be over Fig. 7, that over Fig. 7 should
be over Fig. 6 and the photo over Fig. 6 should be over Fig. 8.
Back Cover: Typographic errors should be corrected to read:
Ross-Carter Corporation
Pittsburgh-Des Moines Steel Co.
The structural steel industry (see back cover) stands ready to assist you in determining
the most economical solution for your problems. Our assistance can include budget
prices, estimated tonnage, cost comparisons, fabrication details, and delivery
schedules.
COVER PHOTO: LOS ANGELES TIMES
PREFACE
Adequate fire protection of structural members is an important consideration in the
design of modern steel buildings. Properly designed structural systems and construc-
tion details greatly reduce life hazards and limit fire damage.
Recent developments in fire protection engineering include analytical techniques
for prediction of fire characteristics and structural fire resistance. Some of these
developments are reflected in recent changes in Building Codes and Standards. In
1978, ICBO (International Conference of Building Officials) modified its provisions with
regard to fire resistive classification, allowing building officials to accept structural
calculations in lieu of tests.
One of the objectives of this case study is to encourage a better understanding of
the response of structural systems to real building fires. It is also intended to illustrate
some of the results of fire response predictions and to compare these predictions with
observed performance of a high-rise steel frame building exposed to a major fire.
Recent technical advances in fire modeling and response analysis now make it
possible to predict building fire temperatures, fire severity, and structural behavior.
The application of such analytical tools would result in more economical use of steel
construction consistent with the safety requirements of real fires.
TABLE OF CONTENTS
PREFACE
TRENDS IN THE DESIGN OF STRUCTURAL FIRE PROTECTION
THE BUILDING
Architectural
Structural
Mechanical and Electrical
Fire Protection
THE FIRE
General
Fire Spread
Fire Suppression
EXTENT OF DAMAGE
Structural
Architectural
Fire Protection
REHABILITATION
PREDICTION OF RESPONSE
Fire Modeling
Fire Severity
Calculated Temperatures in Structural Members
SUMMARY
REFERENCES
ACKNOWLEDGMENTS
2
TRENDS IN THE DESIGN OF STRUCTURAL FIRE PROTECTION
FIRE SAFETY PROVISIONS IN CURRENT BUILDING DESIGN CODES HAVE RECENTLY
COME UNDER CLOSE SCRUTINY. In ths regard, the 1973 Report of the National Commis-
sion on Fire Prevention and Control (NCFPC), commented as follows:
Fire safety analysis is lagging behind innovation in building design. For example,
there is an understandable trend toward ever lighter structural members which
reduce the cost without significantly reducing strength. Building designers in-
troduce these innovations while two important questions go unanswered. First,
are the structural members adequately protected from fire for the entire life of
the building, as well as during a fire that may occur tomorrow? Second, are ex-
isting tests for fire safety adequate for measuring the fire protection afforded by
the particular innovation?
Other problems identified by the NCFPC were:
1. Test results and actual performance in fires can be widely divergent.
2. Building code requirements are often inconsistent. Some safety-related provisions are
excessive, while others are virtually ignored.
3. The body of knowledge on which fire safety standards are based is deficient.
In recent years, the fire services have received wide recognition for significant im-
provements in fire detection and suppression. THE DEVELOPMENT OF ANALYTICAL
TECHNIQUES FOR REALISTIC EVALUATION OF STRUCTURAL FIRE RESISTANCE HAS
ALSO PROGRESSED SUBSTANTIALLY. THE FOLLOWING DEVELOPMENTS HAVE
RESULTED FROM THE COMBINED EFFORTS OF RESEARCHERS, DESIGNERS, CON-
SULTANTS, MATERIAL PRODUCERS AND PUBLIC AGENCIES:
1. Modeling of fire environments which accounts for different fire loads, ventilation, and
fire characteristics. 2 3 4
2. Structural analysis computer programs for predicting performance of frames and floor
systems exposed to different fire environments, s
3. The publication of design guidelines for fire-safe construction.6
The progress in analytical prediction of fire response of structures is reflected in several
proposals for revision of model building codes and standards. IN 1978, THE INTERNATIONAL
CONFERENCE OF BUILDING OFFICIALS (ICBO) ADOPTED THE FOLLOWING MODIFICA-
TIONS IN SEC. 4302(b):
As an alternate to Tables Nos. 43-A, B anc C, fire resistive construction may be
approved by the building official on the basis of evidence submitted by the per-
son responsible for the structural design showing that the construction meets
the required fire-resistive classification
The reason cited by ICBO for this modification is:
. . to encourage development of engineering solutions based on fire research
into the structural performance of floor or roof systems under fire exposure con-
ditions contemplated by. the code.
WITH THE INTRODUCTION OF THIS FLEXIBILITY INTO THE UNIFORM BUILDING CODE
FIRE RESISTIVE STANDARDS, DEVELOPMENT IN STRUCTURAL DESIGN FOR FIRE SAFETY
CAN BE EXPECTED. For example, FIRE SAFE STRUCTURAL STEEL,6 published by the
American Iron and Steel Institute, contains a procedure for calculating effects of fire on ex-
posed exterior steel as a function of fire load compartment geometry and ventilation. THESE
AND OTHER SIMILAR DEVELOPMENTS WILL UNDOUBTEDLY LEAD TO GREATER
ECONOMY IN THE COST OF BUILDING FIRE PROTECTION WITH GREATER RELIABILITY IN
STRUCTURAL FIRE RESISTANCE.
The following case study is intended to illustrate some of the results of the fire response
predictions and to compare these results with observed performance of a high-rise steel
frame building exposed to severe fire.
3
THE BUILDING
The Occidental Center Tower Building is located at 1150 Olive Street in downtown Los
Angeles. The building (see cover) is 438 ft. high with 32 stories above grade and 2 below, and
the plan dimensions are 200 ft. by 95 ft., Fig. 1. Structural floor framing and typmal detaJIs are
shown in Figs. 2 and 3.
The exterior wall construction consists of a glazed, terrazzo tde wall, with tempered glass
windows. Hollow aluminum vertical fins run continuously up the building face, and hollow
aluminum eyebrows at the floor level, attached flush with the exterior wall face, protrude from
the face of the building.
Structural framing for the building consists of steel shapes with bolted, rigid girder-column
connections. The floors consist of 11/2 m. deep, 0.0945 in. thick steel deck with Ightweight
reinforced concrete slab.
The building was designed as Type I, non-combustible construction, according to the 1963
Los Angeles Building Code. Except for a few isolated areas, the building is not equipped wtth
sprinkler systems.
The architectural, structural, electrical, mechanical, and fire protection systems of the Oc-
cidental Tower Building are described below:
ELEVATORS
- - - - T - - - ' " - ' " : : " ,"!'i',ii"i', i?,i,i;';';i. . -I ' :, , , -. . "?, :' !L. . . ' 7 _ _ 7 ,,,,,,,,? i;:ii;,T:? ' ' ' ' ' ' '
, . . . . ?...,
[ .
200'
95'
N -
LEGEND
..X- AREAS OF ORIGIN (3)
__HEAVY FIRE DAMAGE
!?-' ::' J.;:....?i HEAT AND/OR SMOKE DAMAGE
'4B LOCATION OF DOOR SHOWN IN FIG. 4
FIG. 1
ARCHITECTURAL PLAN AT 20TH FLOOR
4
Architectural
The Occi dental Center Tower Building serves as the home offi ce for the Occidental Life Insurance Company. A
typtcal floor plan and structural framing and details are shown in Figs. 1, 2, and 3. Construction of the building
began Jn 1962, following the 1963 edition of the Los Angeles Building Code and was compl eted in 1964.
Forty percent of the 20th and 21st floors is used for executive offices, typically furnished wi th wood desks,
chairs, sofas, bookcases, and hie cabinets. Approxi matel y 30% of the 20th and 21st floors consists of secretarial
and crculaton space, with the usual furnishings of desks and general files. The remaining space is used for
general purposes, such as mechani cal and storage rooms, lavatories, elevators and stairwells.
The typical executive offices on the 20th floor are small compartments, 200-300 sq. ft. in area, l ocated around
the permeter of the building, wi th large wi ndow openings which, during the fire, provided ampl e ventilation.
Prior to the fire, the offices were separated by full height partitions whi ch were constructed using 31/2 in., 0.0625
in. thick perforated steel studs, with expanded steel lath and 1 in. plaster on each face. Interior wall finish was vinyl
f abri c wallpaper. Doors were 13,4 m. solid core wood. The 1/2 in. carpet i ng was underlaid by a 3/4 in. pad. The
acousti cal tile ceihng was suspended 8 ft. 9 in. above the floor; approxi matel y 16% of the ceiling area was taken up
wth recessed plastic light panels.
Interior architectural features of the 20th floor were not charactestic of the entire building. Most of the other
floors are used as general offices, wi th large compartments of up to 2400 sq. ft. Partitions within these compart-
ments are not full height and the furnishings vary.
The building' s exterior wall construction consists of glazed, terrazzo tile wall extending 4 ft. 3 in. bel ow and 2 ft. 9 in.
above the floor tine. The wi ndows are approximately 6 ft. high, grey tempered glass, 1%, in. thick. Hol l ow aluminum
vertmal fins, 2% Bn. wde at 62 in. centers are attached wi th 11/2 in. cl earance from the exterior face of the wall.
These run continuously up the building face. Hollow aluminum eyebrows at the floor level, attached flush with the
exteri or wal l face, protrude 30 in. from the face of the building.
Structural
Structural framing for the building consists of wi de flange steel columns, spaced at 30 ft. on center in each direc-
tion. Wi de flange steel girders span between the col umns at each floor level, and bolted, rigid girder-column con-
nections provide for frame action in the east-west and north-south directions to resist lateral loads. Secondary floor
beams, spaced at 7.5 ft. on center, span between the east-west girders. This framing supports an electrified com-
postte steel deck floor of 1.5 in. deep, 0.0405 in. thick steel deck and 2.5 in. lightweight concrete rei nforced with
6 x 6 in. - - 10 x 10 welded wi re steel fabric.
The air conditioning system consists of galvanized steel duct s for the mai n supply system, wi th 5.5 in. di ameter
fiberglass wire-reinforced supply duct s connecting the metal duct s to the ceiling diffusers. The ceiling space is
utihzed as a return air plenum. The supply ducts are insulated wi th fiberglass covered by aluminum foil.
Air is supplied from a central shaft. One fan coil for heating and one for cooling is provided for each floor.
Capaci ty at each floor is approxi matel y 15,000 cfm. About 6% of the return air is exhausted to the outside of each
level.
Metallic floor inserts activate the electrical cells in the raceways in the metal deck, while steel condui t and flexi-
ble steel cabl e supply power to ceiling lights and wall switches.
There are three banks of elevators. One provides an express servi ce to the 20th floor, stopping at each suc-
ceedi ng floor thereafter to the 30th. A second bank provides express servi ce to the 11th floor and then services
each floor from the 12th to the 20th. The third bank provides local servi ce from the ground to the 11th floor. The
freight el evator stops at each floor through the 30th.
Fire Protection
The building was designed as Type I, non-combustible const ruct i on accordi ng to the Los Angeles Building Code.
A 3-hour fire rating is required for the mai n structural frame, and a 2-hour rating is required for the fl oor system.
Girders are protected by 13.4 in. mi ni mum of spray.applied vermi cul i t e fireproofing, and col umns are f urred to
30 x 30 in. fimshed dimensions and are protected by l y, in. mi ni mum vermi cul i t e gypsum plaster applied to expand-
ed steel lath. The bottom of the electrified steel deck composi te fl oor is protected by in. minimum thickness of
spray-applied vermiculite fireproofing, and secondary floor beams are protected by 1 in. minimum similar spray-
applied fireproofing.
Combination fixed-temperature and rate-of-rise detectors are provided for each 200 sq. ft. of floor area. These
devi ces acti vate an alarm at the lobby security desk when the temperature rises faster than 15F per minute, or
when a temperature of 135 F is attained. The al arm indicates only the floor and not the specific location of the ac-
tivated detector.
Each floor is served by two 6 in. di ameter combination standpipes wi th 21/2 in. diameter outlets, l ocated in each
encl osed stairway. There are three 11/2 in. outlets on each floor adj acent to the building core. This system is sup*
plied by domesti c cty water and pressurized by a 750 gpm, 300 psi el ectri c pump without emergency power. Provi-
sion s made for connection of the system to fire department pumps.
Except for a few isolated areas, the building does not have sprinklers.
THE FIRE
General
At 3:15 a.m., the ADT central station notified the security desk that a detector had been ac-
tivated on the 20th floor. Upon discovering the fire, the security guard used the elevator inter-
com to notify the officer at the lobby security desk, who in turn called the fire department be-
tween 3:18 and 3:20 a.m. Coincidentally, a driver of a passing fire department ambulance
noticed the flames in the windows of the Occidental Tower Building and notified the fire
department. The fire department was on the scene by 3:24 a.m. and by 4:50 a.m. the fire was
"knocked down."
O
O
.r-
Od
STAIRWELL STAI RWELL
I
AA A B C I D E F
I
o 'b'l. , J 6 SPCESi 361 ' 0 ' = 1861 ' o ' , o'o'_
44
i
II 5 = , i x 5 I 5
o 11 2 ,_2 o 2 2 5 1 5 1
__03 4 4 3 3 4
'" = z ' " Z l =1 =1 12 J ' / -.'7'
co 4 4 13 4 3 [3 j4 3 3 4-]?-/7"-j-
" " " j ' " /""'f . . . . .
t
ELEVATOR
SHAFTS N <
CJ
(TYP. FLOOR BEAMS)
V2 IN. STEEL DECK
1/2 IN. CONCRETE
FLOOR SLAB
GIRDER SIZES
1 W27x102
2 W27x94
3 W21x73
4 W21x68
5 W21x62
COLUMN SIZES (BETWEEN 20TH AND 21ST FLOOR)
I W14x398
2 W14x342
3 W14x287
4 W14x237
FIG. 2
STRUCTURAL FRAMING AT 21ST FLOOR
= 6x6 -- 10x10 WELDED WIRE FABRIC
21/2 IN. LIGHTWEIGHT CONCRETE
. . . . ../- .... 11/2 IN. STEEL DECK
-'1 ], .' IN. MIN. CEMENTITIOUS FIREPROOFING
1 IN MIN. COVER AT SECONDARY FRAMING
(FLOOR BEAMS)
CEMENTITIOUS FIREPROOFING
DECK AND FRAMING
I t IL
-II . . . . . . . . L.
30"
COLUMN
STEEL COLUMN
=4 EXPANDED STEEL LATH
-- 1} IN. VERMICULITE GYPSUM PLASTER
--FURRING
31/2 IN. STEEL STUDS
EXPANDED STEEL LATH
1 IN. PLASTER
PARTITION
FIG. 3
TYPICAL FIRE PROTECTION DETAILS
7
Fire Spread
It is believed that the fire was ignited with flammable liquid in several locations in the south-
west corner of the 20th floor, Fig. 1. It spread rapidly through the south and west corridors
and into the offices at the perimeter of the building. The doors to the executive offices were
generally left open, thus enabling the fire to spread rapidly through the hallways and into the
offices. The fire stopped at the northwest and southwest ends of the corridor where the doors
were closed. THE EFFECTIVENESS OF COMPARTMENTATION IN LIMITING FIRE SPREAD
WAS CLEARLY DEMONSTRATED BY THE DOOR SHOWN IN FIG. 4. This door was badly
charred on the fire side, but was not damaged on the unexposed side. It contained the fire,
preventing its spread to the north wing of the building on the 20th floor.
By the time the fire department arrived (3:24 a.m.), the 20th floor was fully involved, with
flames issuing from broken windows. Radiation from the flames broke windows in the south-
west corner of the 21st floor, and a small fire broke out in this area. Radiation also cracked
some windows on the 22nd floor, but the fire did not'spread to this floor. THE FIRE RESISTIVE
FLOOR SYSTEM WAS FULLY EFFECTIVE IN PREVENTING SPREAD OF FIRE ABOVE THE
20TH FLOOR.
Fire Suppression
The first company of fire fighters to arrive found the 20th floor ablaze and flames threaten-
ing the 21st floor. Additional units were immediately summoned. Control of the lobby security
center was established and pumping was initiated into the standpipe system.
The battalion chief in command dispatched two attack teams, consisting of a captain and
four or five men, to the 21 st floor to suppress the vertical spread of the fire. They found that
flames had entered the 21 st floor, a small area of which was burning. Two attack teams were
sent to the 20th floor to fight the main fire. They encountered intense heat and smoke and in-
itially had to work from a prone position. A salvage crew proceeded to the 19th floor.
The elevators were used to transport men and equipment to the 18th floor, and stairways
were used above the 18th floor. Although the elevators seemed to malfunction for a brief
period of time, they remained essentially operative throughout the fire.
Firemen characterized the fire as being very smoky. This problem was not alleviated by turn-
ing on the HVAC system, which had been turned off to conserve energy. Firemen had to break
windows to vent the smoke.
Additional fire fighting forces were assigned to various locations in the fire area. A
resources pool was established, and a heliport was set up two blocks away. Two helicopters
were used for external lighting and observation.
The fire was "knocked down" at approximately 4:50 a.m., almost 11/2 hours after the initial
alarm. A total of 58 fire companies responded, providing a force of 300 men and associated
equipment. Five firemen were injured, four from exhaustion, and one from falling glass.
8
FIG. 4 DOOR EFFECTIVELY LIMITED FIRE SPREAD
EXTENT OF DAMAGE
The 19th floor sustained no fire damage due to effective containment by the fire-resistive
floors, shafts and exterior wall eyebrows. Approximately 10% of the 21 st floor was damaged
by a small fire which broke out in the area of the southwest corner, and plastic diffusers on
ceiling lights were damaged by radiation emitted by flames outside the adjacent windows. A
few windows on the 22nd floor were cracked, and some minor smoke damage was conse-
quently sustained, Fig. 5. Operations on the 22nd floor were back to normal ten days after the
fire. Minor interior damage was noted on floors 22 through 32, caused by smoke which spread
partly through the elevator shafts and partly through the air-conditioning system.
Approximately 30% of the area on the 20th floor suffered heavy fire damage, characteriz-
ed by virtually complete consumption of all combustible materials, Fig. 6. This damage was
concentrated in the southwest corner and the central portion of.the floor area.
An additional 25% of the 20th floor was subjected to heat and smoke damage, Fig. 7, in-
cluding the southeast corner, several of the offices along the western edge of the building,
and the area near the eastern elevator banks. The northern end of the building was relatively
free of damage because of effective containment by the closed fire-resistant doors.
Structural
NO DAMAGE TO ANY STEEL BEAMS, GIRDERS, COLUMNS, oR DECK WAS OBSERVED.
FIRE PROTECTION FOR THESE ELEMENTS SERVED TO HOLD TEMPERATURES WELL
BELOW ALLOWABLE LIMITS. (See Prediction of Response - - Calculated Temperatures in
Structural Members, below.)
Architectural
The exterior aluminum eyebrows and fins melted where directly exposed to flames. None-
theless, vertical spread of the fire was initially inhibited by these elements. Soot formed on the
building exterior above the fire floor, and exterior window glass was shattered, as noted
above.
Interior furnishings directly exposed to fire were destroyed. The suspended ceiling burned,
and its suspension wires and T-bar framing collapsed. Although plaster partitions cracked ex-
tensively and began to separate from the metal studs, Fig. 8, the partitions acted as fire com-
partment walls, effectively deterring the spread of the fire.
9
Areas adjacent to the burned out areas suffered substantially less damage. Heat blistered
the interior finishes and soot covered large areas. Water damage was also sustained.
..
%
m
FIG. 5 DAMAGE TO THE EXTERIOR FACE
FLOOR
NOS.
,, 22ND
21ST
20TH
19TH
aa
,f
/
i mm i
HEAVY FIRE DAMAGE
ON 20th FLOOR
iiii i i i !
FIG. 7 HEAT AND SMOKE DAMAGE
ON 20th FLOOR
10
Air-conditioning ducts in areas of heavy fire exposure collapsed, and non-metallic connec-
tors in these areas were consumed, Fig. 6.
The 300 amp/480 volt buss ducts, that provided electricity for the entire floor was severely
damaged by heat and was short-circuited. Aisc, the electrical distribution system in the parti-
lions and ceilings suffered extensive damage, and a large percentage of the light fixtures and
utilization equipment was destroyed in the fire area. Because the steel deck raceways were
not damaged, new wiring could easily be replaced where necessary.
Fire Protection
Spray-applied fire protection at a beam column joint near the southwest corner of the 20th
floor cracked, and the fire protection on the underside of the deck and framing spalled locally.
It is likely that some of this material was knocked off by fire hose stream or spalled as a result
of rapid cooling. Nevertheless, the fire protection satisfactorily performed its function of pre-
venting any structural damage.
REHABILITATION
Building rehabilitation was subdivided into five categories: clean-up, electrical repair, re-
construction of the 20th and 21st floors, and general repair, which included replacement of
glass, exterior panels, painting, and replacement of filters in the air-conditioning system
remote from fire zones. NO REPAIR OR REPLACEMENT OF ANY STRUCTURAL STEEL
FRAME OR FLOOR SYSTEM COMPONENTS WAS REQUIRED. Clean-up was completed in
about six weeks, and all rehabilitated spaces in the building on the 20th and 21 st floors were
re-occupied within seven months. The total cost of rehabilitation was in excess of $1,200,000,
exclusive of cost of furnishings and other indirect costs.
FIG. 8 PARTITIONS AND SUSPENDED
CEILING DAMAGE ON 20th FLOOR
11
PREDICTION OF RESPONSE
Fire Modeling
Fire in an enclosure can develop in a number of different ways, depending on such factors
as fuel load, enclosure ventilation, enclosure volume, fuel surface area, relative position of
fire load components, enclosure thermal characteristics, and flame spread rate. When all of
these variables are precisely defined, it is possible to construct a time history of fire develop-
ment in terms of flame temperature within the compartment
The AISt publication, FIRE SAFE STRUCTURAL STEEL,6 describes a method, developed by
Margaret Law, for determining maximum fire temperature as a function of fire load, compart-
ment geometry, and ventilation. Harmathy,2 Babrauskas and Williamson,3 and Coulbert, 4
among others, described methods of modeling encl osure fires. Magnusson and Thelan-
dersson, in a reporF published in 1970 in ACTA POLYTECHNICA SCANDINAVICA, presented
a set of tables and curves defining a variety of possl e fires for given fuel loads, ventilation
factors, and enclosure characteristics.
Using Magnusson and Thelanderssons' approach, two time*temperature curves have been
cal cul ated for a typical office on the 20th floor of the Occidental Center Tower. The fuel load
used for determining these fire curves was estimated at 5.5 lbs. of combustibles per sq. ft. of
floor area, including wood furniture, doors, wool carpet, plastic wall covering, window shades,
lighting panels in the ceiling, acoustical ceiling tiles, as well as contents of filing cabinets. The
wood equivalent of this load was estimated between 6 psf (lower bound) and 12 psf (upper
bound). Using 7500 BTU/lb. for wood, two fire curves were determined and are shown in Fig.
9. Maxi mum temperatures have also been cal cul ated using equations given in the AISI
publication, FIRE SAFESTRUCTURAL STEEL. These values, 1600F and 2140F, respectively,
for the upper and lower bounds on fuel load, bracket the maximum values of 2010F and
2060F obtained by the Magnusson and Thelandersson analysis.
The curves shown in Fig. 9 reflect a typical fuel-controlled fire of short duration. With ample
ventilation and without intervention by the fire services, all the combustible materials in the
compartment would be consumed in about 1.25 to 1.5 hours. The peak temperature of about
2000 F is reached in about 0.1 hours after flashover and would last from 0.05 to 0.2 hours,
depending on fuel load. The curves do not reflect the early pre-fiashover fire growth, whi ch
depends on the source of ignition and on flammability of contents surrounding the source.
While this phase of fire growth may be of considerable significance for life safety and fire con-
trol, it is of little importance for prediction of structural performance.
Several interesting observations of physical evidence at the Occidental Center Tower fire
lend credibility to the theoretical approach:
1. Estimates of peak temperatures around 2000F, based on melting points of various
materials and characteristics of the observed damage, correspond closely with the
calculated maximum.
2. Assuming that the 20th floor alarm was activated at the time of initial flashover, maxi-
mum temperature and corresponding peak development would have occurred within
six to twelve minutes. The fire department arrived approximately nine minutes after
the first alarm and found the fire fully developed.
3. Initially, intense heat and smoke prevented fully effective intervention to extinguish
the existing fire and confined the fire fighters' efforts to limitation of the fire spread.
The observed damage suggests that most of the combustible materials in the intense
fire areas were consumed. Thus, the 11/2 hr. knock-down time corresponds well with
the fire duration obtained from the theoretical curves.
A comparison of the theoretical fire curves wi th the ASTM E-119 standard, Fig. 9, is of in-
terest. It can be seen that the maximum fire temperatures and the initial rate of temperature
rise during the first 0.25 hour of the ASTM curve is too Iow for the type of fire considered.
Also, the temperature drop is not reflected in the ASTM E-119 curve. Thus, results of standard
tests during the time of interest in the actual fire may not reflect performance of components
in real fire.
12
Pre-fire interior archi tectural features of the 20th and 21 st floors were not characteri sti c of
the enti re building. The partitions on the other floors were not all full height. Compartment
sizes and typical furnishings on the other floors varied. It is i mportant to note that such ar-
chi tectural features must be taken into account when model i ng real fire behavior.
2100
18O0
,,,iiii:
E.119 FIRE
1200
q = 35 MCAL/m2
900 q 35 M(AL/m 2
600
300
I i
0 0.2 0.4 0.6 0.8 1.0 1.2
TI ME (HOURS)
NOTES:
a. F, and F2 CALCULATED USING REF. NO. 7
b. OPENING FACTOR: (A/AT) = 0.12
c. FULE LOAD q = MCAL/BOUNDING SURFACE AREA
FIG. 9
C A L C U L A T E D TIME - - T E M P E R A T U R E CURVES
A
1.6 - o
I
u)
1 . 2 -
i-
0.8 -
I.U
(/)
w 0.4 -
IL.
ASTM E-119 F I R E .
FIRE F2
, , , i t i ,
0.2 0.4 0.6 0.8 1.0 1.2
TI ME (HRS)
FIG. 10
SEVERITIES O F ASTM E-119, Fl, F2 FIRES
13
Fire Severity
When the ASTM E-119 fire curve was adopted in 1922, it was recognized that actual fires
vary from the standard curve. It was suggested that the concept of fire severity, defined as
the area under the fire curve, would account for the differences and that the required en-
durance for a component may be established as the time, t, at which the area under the stan-
dard ASTM E-119 fire curve has the same value (severity, S) as the total area under the actual
fire curve.
Areas under the fire temperature curves, as a function of time, have been calculated and
plotted in Fig. 10. At about 1.0 to 1.25 hrs., all combustibles in ventilated enclosures would be
consumed without the intervention of the fire services. It is seen that the ultimate severity, S1
and S2, of the estimated fires would be achieved at approximately 0.7 to 0.9 hrs. of the stan-
dard fire. With active intervention (fire suppression) by the firemen, duration and consequently
severity of the fire should be significantly less.
It can be seen from Fig. 10 that, at the upper bound fuel load, the maximum fire severity is
about 1,300 HRS-F. The ASTM E-119 fire severities are approximately 1,500 HRS-F at 1-hr.,
2300 HRS-F at 2 hrs., and 4500 HRS-F at 3 hrs. duration. THUS, BASED ON THE FIRE.
SEVERITY CONCEPT, THE REQUIRED FIRE RATINGS IN THE CURRENT CODES ARE 2 TO 3
TIMES AS SEVERE AS THOSE EXPECTED UNDER A REPRESENTATIVE REAL FIRE.
Calculated Temperatures in Structural Members
Temperatures in the steel columns and floor deck were calculated using computer pro-
gram FIRES T3,8 modified for radiation heat transfer in the interior cavities. Results of these
calculations are shown in Figs. 11 and 12.
For the steel column shown in Fig. 11, the exterior face of the fire protection exposed to the
fire models, F and F2, experiences about the same maximum temperature as that developed
during an ASTM E-119 standard 3-hr. test. However, in a realistic fire, such as F and F2, high
temperature is developed much faster (.4 hr. vs. 3 hrs.) and is maintained for a very brief
period of time (10-15 mins. vs. 2 hrs.). The interior face of the fire protection exposed to fires
F and F2 develops only about 45% of the maximum temperature developed in the ASTM
E-119 3-hr. test, and the steel column web or flange develops only about 20% of the max-
imum temperature in the standard 3-hr. test. Columns in the Occidental Tower Building had
furred fire protection enclsoures, Fig. 3, and consequently the steel temperatures would be
even lower than those shown in Fig. 11.
For the steel deck, Fig. 12, with , in. vermiculite cover, the maximum temperature of the
exterior face of the fire protection is slightly higher (1950F vs. 1820F) for fires F and F2,
compared to the ASTM E-119 2-hr. test. However, the duration of high temperature exposure
is much shorter (Y2 hr. or less vs..1 /2 hrs.). THE BOTTOM SHEET OF THE FIRE-PROOFED
ELECTRIFIED STEEL DECK, EXPOSED TO MODEL FIRE F OR F2, EXPERIENCES ONLY
ABOUT 45% OF THE MAXIMUM TEMPERATURE DEVELOPED IN AN ASTM E-119 2-HR.
FIRE (400F VS. 900F).
14
(a) ASTM E-119 FIRE
(b) F FIRE MODEL
(c) F2 FIRE MODEL
T i T, T2 T3
W14x343
With 1} IN. VERMICULITE PLASTER COVER
2020F
1600
1oo l / \ x(c)
800 y b)".,.,..635OF
400 1 -r,,609OF
0 t i i
0.5 1.0 1.5
(a)
I I I
2.0 2.5 3.0
T I
FIRE PROTECTION
EXTERIOR
A
U.,
O
LU
I--
I,U
12.
UJ
I--
1200 -J 561 F
800J' ( b ) , L ( c ) _ _ / _(a)
400
0 478F
I I I I I
0.5 1.0 1.5 2.0 2.5
94F
800 [( /)r 123F
400 b)
0 I I I
0.5 1.0 1.5
1027 F T2
FIRE PROTECTION
INTERIOR
I
3.0
(a)..__ 429oF T3
WEB OF COLUMN
I I I
2.0 2.5 3.0
- - 1 2 0 F
800-I / .18OOF
400 ( b /
0 i i i
0.5 1.0 1.5
(a) 485 F T,
FLANGE OF COLUMN
I I I
2.0 2.5 3.0
TIME(HOURS)
FIG. 11
CALCULATED TEMPERATURES FOR INTERIOR COLUMN ON 20TH FLOOR
15
0.15 hr
0.30 hr
060 hr
0.15 hr
0.30 hr
0.60 hr
I I I I [ I
I I I [ I I I I
o o o a o o o IJ,.
0
, I I I I I I I
I I I I I I !
I
0.25 hr
0.5 hr
1.0 hr
2.0 hr
0.25 hr
0.5 hr
1.0 hr
2.0 hr
O
O
O
Od
I I I I I I 1
I I I I I I
8 8
0
I l l l i t l
I I I I l l l
21/2 IN. LIGHTWEIGHT
CONCRETE SLAB
I
1
} IN. FIREPROOFING
FIRE F, ASTM E-119 SECTION
FIG. 12
TEMPERATURE GRADIENTS FOR THE FLOOR DECK ABOVE 20TH FLOOR
SUMMARY
The following observations summarize the findings of this case study:
1. The structural steel frame and steel deck floor suffered no damage, even though the
combustibles within about 9000 sq. ft. of the 20th floor were completely burned out.
2. Fire-resistive floor systems and shaft enclosures confined the fire to one floor. On the
floor involved, closed fire doors effectively prevented fire spread to about one-half of
the floor area.
3. Fire models predicted for the office spaces of the 20th floor agreed with observed
thermal characteristics of the fire:
a. predicted peak fire temperatures agreed with those observed by fire investigators,
and
b. duration of the predicted fire agreed with actual duration.
4. Calculated temperatures in structural steel exposed to predicted models of fire are
substantially less (about 50% less) than those developed under ASTM E-119 standard
test exposures.
THE AGREEMENT BETWEEN PREDICTED AND OBSERVED PERFORMANCE OF STRUC-
TURAL STEEL WITH FIRE PROTECTION INDICATES THAT MORE ECONOMICAL DESIGN OF
STEEL BUILDINGS MAY BE POSSIBLE BASED ON REALISTIC FIRE MODELS.
16
REFERENCES
1. National Commission on Fire Prevention and Control, America Burning, Washington, D.C.,
1973, p. 71.
2. Harmathy, T.Z., "A New Look at Compartment Fires - - Parts I and I1," Fire Technology,
Vol.8, Nos. 3 and 4, 1972.
3. Babrauskas, V. and Williamson, R.B., "Post-Flashover Compartment Fires: Basis of a
Theoretical Model," Fire and Materials, Vol. 2, No. 2, 1978, London.
4. Coulbert, C.D., "Energy Release Criteria for Enclosure Fire Hazard Analysis -- Parts I and
I1," Fire Technology, Vol. 13, Nos. 3 and 4, 1977.
5. Bresler, B., "Behavior of Structures in Fire Environments," Proceedings of a Symposium
Honoring E.P. Popov, Prentice-Hall, 1979.
6. FIRE SAFE STRUCTURAL STEEL A DESIGN GUIDE, American Iron and Steel Institute,
Washington, D.C., 1979.
7. Magnusson, S.E. and Thelandersson, S., "Temperature -Time Curves of Complete Process
of Fire Development," Acta Polytechnica Scandinavica, Civil Engineering and Building
Construction Series No. 65, Lund Institute of Technology, Stockholm, 1970.
8. Iding, R., Bresler, B., and Nizamuddin, Z., "FIRES - - T3, A Computer Program for the Fire
Response of Structures - - Thermal," Report No. UCB FRG 77-15, Department of Civil
Engineering, University of California, Berkeley, October 1977.
A CASE STUDY
prepared by
Wiss, Janney, Elstner and Associates, Inc.
485 Baybridge Office Plaza
5801 Christie Avenue
Emeryville, California 94608
ACKNOWLEDGMENTS:
The assistance and cooperation of the Occidental Life Insurance Company of California,
the Los Angeles Department of Building and Safety, the Los Angeles City Fire Department,
and the American Iron and Steel Institute is gratefully acknowledged.
Occidental Center Tower Building
Architect: William Pereira Associates
Rehabilitation Project
Architect: Gin Wong Associates
General Contractor: William Simpson Construction Co.
LAYOUTAND DESIGN: NEOGRAPHIC
December 2003
Steel Construction in the
New Millennium
By
Patrick M. Hassett, S.E.
Principal
Hassett Engineering, Inc.
____________________________________________________________________________
(A copy of this report can be downloaded for personal use from www.aisc.org)
ii
This report rewrites a Steel TIPS, Steel in the 90s, that was written by Robert Preece, and Alvaro L.
Collin, which was very popular with structural engineers, fabricators, erectors, and inspectors. The
SSEC wished to update this publication to include a discussion of new steels, new welding issues, some
lessons learned from the Northridge earthquake, and pertinent developments derived from the SAC
investigations. Some sections are unchanged, others are new, while others have been updated,
amended, or completely revised.
First Printing, December, 2003
__________________________________________________________________________________
Patrick M. Hassett, S.E., Hassett Engineering, Inc, 3558 Castro Valley Blvd., Castro Valley, CA 94546
Disclaimer: The information presented in this publication has been prepared in accordance with recognized engineering
principles and is for general information only. While it is believed to be accurate, this information should not be used or
relied upon for any specific application without competent professional examination and verification of its accuracy,
suitability, and applicability by a licensed professional engineer, designer or architect. The publication of the material
contained herein is not intended as a representation or warranty on the part of the Structural Steel Educational Council or
of any other person named herein, that this information is suitable for any general or particular use or of freedom from
infringement of any patent or patents. Anyone making use of this information assumes all liability arising from such use.
Caution must be exercised when relying upon specifications and codes developed by others and incorporated by reference
herein since such material may be modified or amended from time to time subsequent to the printing of this document. The
Structural Steel Educational Council or the authors bears no responsibility for such material other than to refer to it and
incorporate it by reference at the time of the initial publication of this document.
Steel Construction in the New Millennium
By: Patrick M. Hassett, S.E.
v
STRUCTURAL STEEL CONSTRUCTION IN
THE NEWMILLENNIUM
By:
PATRICK M. HASSETT, S.E.
Principal
Hassett Engineering, Inc.
ACKNOWLEDGEMENTS...............................................................................................iii
PREFACE ..........................................................................................................................1
INTRODUCTION..............................................................................................................1
EARLY BUILDING DESIGN...........................................................................................1
Riveting, a Lost Art..............................................................................................2
High Strength Bolting..........................................................................................2
ALLOYING, KEY TO STRENGTH AND WELDABILITY ...........................................3
CARBON EQUIVALENT AND PREHEAT ....................................................................3
QUENCHING AND TEMPERING FOR HIGH STRENGTH..........................................4
TODAYS STRUCTURAL STEELS................................................................................4
Material Specification..........................................................................................5
Special Material Considerations, Supplementary Requirements..........................5
ASTM A36 (as rolled) .........................................................................................8
ASTM A992 (as rolled) ...................................................................................... 9
ASTM A572 (as rolled) ...................................................................................... 9
ASTM A529 (as rolled) ...................................................................................... 9
ASTM A913 (quenched and self-tempered) ......................................................10
ASTM A242 (as rolled) .....................................................................................10
ASTM A514 (quenched and tempered) .............................................................10
ASTM A709 (as rolled and quenched and tempered)........................................10
ASTM A759 (controlled cooled) .......................................................................11
ASTM A852 (quenched and tempered) .............................................................11
ASTM A53 (as rolled) .......................................................................................11
ASTM A500 (cold-formed structural tubing) ....................................................11
ASTM A501 (hot-formed structural tubing)......................................................12
ASTM A618 (hot-formed structural tubing)......................................................12
ASTM A847 (cold formed structural tubing).....................................................12
ASTM A27 (steel for castings) ..........................................................................12
THE WELDING PROCESS............................................................................................13
Electrodes...........................................................................................................13
Shielded Metal Arc Welding (SMAW)..............................................................14
Flux-Cored Arc Welding (FCAW).....................................................................14
Gas Metal Arc Welding (GMAW).....................................................................16
Submerged Arc Welding (SAW) .......................................................................16
Electroslag Welding (ESW)...............................................................................17
AVOIDING WELD DEFECTS.......................................................................................17
Weld Cracking...................................................................................................18
Hydrogen Embrittlement....................................................................................19
Lamellar Tearing ...............................................................................................19
ASTM A770......................................................................................................20
vi
ENGINEERS ROLE IN MINIMIZING WELD DEFECTS...........................................21
CORRECTING WELD DEFECTS..................................................................................24
CORRECTING WELD DISTORTION...........................................................................25
NONDESTRUCTIVE TESTING (NDT).........................................................................27
Visual Inspection (VI)........................................................................................27
Magnetic Particle Testing (MT).........................................................................27
Ultrasonic Testing (UT).....................................................................................28
Radiographic Inspection(RI).............................................................................29
Liquid Dye Penetrant Testing (PT)....................................................................29
PROJ ECT SPECIFICATIONS.........................................................................................29
FEMA 353 QUALITY RECOMMENDATIONS............................................................30
PROJ ECT SPECIFICATIONS CHECK LIST.................................................................30
WELDING PROCEDURE SPECIFICATIONS (WPSs)................................................30
SHOP DETAIL DRAWINGS..........................................................................................31
ELECTRONIC DATA INTERCHANGE........................................................................31
CONCLUSIONS..............................................................................................................31
REFERENCES.................................................................................................................32
FIGURES
Figure 1: Carbon Equivalent............................................................................................4
Figure 2: Stress-Strain Curves.........................................................................................5
Figure 3: Schematic Illustration of Shielded Metal-ArcWelding..................................14
Figure 4: Schematic Illustration of Flux-Cored Arc Welding.......................................14
Figure 5: Fillet Welds by SMAW and FCAW..............................................................16
Figure 6: Gas Metal Arc Welding-spray arc mode........................................................16
Figure 7: Schematic Illustration of Submerged Arc Welding.......................................16
Figure 8: Schematic of Electroslag Process for Welding Typical Butt J oints...............17
Figure 9: Weld Macrostructure......................................................................................18
Figure 10: Directional Nomenclature for Describing Anisotropy in Rolled Plates.........19
Figure 11: Relation Between Tensile Properties and Angle of Specimen ......................20
Figure 12: Development of Lamellar Tearing Crack Under a T-weld............................20
Figure 13: Avoiding Lamellar Tearing in Column Splices.............................................23
Figure 14: Avoiding Lamellar Tearing in Beam Splices.................................................23
Figure 15: Avoiding High Restraint in Beam-Column Details.......................................23
Figure 16: Full Member Restraint in Welded Frames.....................................................23
Figure 17: Internal Restraint in Weldments.....................................................................25
Figure 18: Welded Corner J oints.....................................................................................25
Figure 19: Angular Distortion Resulting from Weld Shrinkage......................................25
Figure 20: Rotation in a Butt Weld.................................................................................26
Figure 21: Transverse Shrinkage in a Butt Weld.............................................................26
Figure 22: Longitudinal Bowing in a Welded Beam.......................................................26
Figure 23: Variation in Yield Strength with Temperature...............................................26
Figure 24: Variation in Modulus of Elasticity with Temperature....................................27
Figure 25: Ultrasonic Testing of Weld............................................................................28
Figure 26: Typical CRT display of ultrasonically tested weld with indications..............28
TABLES
Table A: Historical Background of Structural Steels......................................................6
Table B: ASTM Specifications........................................................................................7
Table C: Filler Metal Requirements..............................................................................15
Table D: Prequalified Minimum Preheat and Interpass Temperature...........................22
ABOUT THE AUTHOR..................................................................................................33
1
Preface
Previous publications on this subject for the
Structural Steel Educational Council (SSEC),
Structural Steel in the 80s Materials,
Fastening and Testing, by F. Robert Preece, and
Structural Steel Construction in the 90s, by F.
Robert Preece, and Alvaro L. Collin, were very
popular with structural engineers, fabricators,
erectors, and inspectors. The SSEC wished to
update this publication to include a discussion of
new steels, new welding issues, some lessons
learned from the Northridge earthquake, and
pertinent developments derived from the SAC
investigations. Some sections are unchanged,
others are new, while others have been updated,
amended, or completely revised.
This publication is intended to be a quick
reference guideline. See the reference list for
more detailed information.
Introduction
Todays engineer has available a wide variety of
structural shapes, grades of materials and joining
techniques with which to design steel-frame
buildings and similar structures. This
proliferation of choices has come about through
rapid technological advances in steelmaking and
related industries over the past four decades.
The development of todays high-strength, low
alloy steels has achieved major material and
energy savings in the steel fabricating process,
while making possible a variety of structural
shapes available to the designer. Care should be
given to specifying high strength steels,
especially with regard to deflection and
vibration, as shall be explained herein.
This paper will review the trends in structural
steel design, from massive all-riveted and bolted
structures of the 40s and 50s to todays
simpler, lighter and more cost-efficient welded
designs. As an aid to the architect and engineer,
this paper will also describe the most commonly
available construction steels, their features and
important applications. Of particular importance
will be the special considerations for matching
electrodes to base metals, the influence the
design has on joint performance, and the quality
control procedures used to assure structurally
sound joints.
Also of interest to the young and experienced
structural engineer is an historical perspective of
the structural steel building industry and the
important developments in the art and science of
steel welding and the methods of minimizing
weld defects.
Early Building Design
The use of steel in building construction has its
roots at the end of the last century with the
construction of such towering edifices as the 14-
story Tacoma Building in Chicago and New
Yorks Tower Building, rising to a height of 129
feet.
As early designers recognized the greater
strength and ductility of mild carbon steel over
cast and wrought iron, office buildings grew
taller and taller. However, it remained until the
late 1920s before high-rise building construction
became truly economical. This economy was
due largely to technological advances made in
the rolling process that permitted the forming of
large structural shapes, coincident with the
development of high-speed elevators.
Until then, steel mills rolled so-called standard
shapes: angles, channels, I-beams and plates,
which were then riveted together to make the
larger columns and girders required for high-rise
construction. Built-up shapes such as these are
inordinately expensive by todays standards due
to the tremendous cost of detailing, drilling,
riveting and assembling. Over the years, some
loosely defined terms have been used to describe
carbon steels. Generally, the agreed terms used
in reference manuals are as follows:
Low Carbon Steel C <= .20%
Medium Carbon Steel 0.20 < C < .050%
High Carbon Steel C >= .50%
The built-up shapes used in early building
construction were necessary due to the inherent
nature of the rolling process. At the time, steel
mills used a single set of rolls for each structural
shape. The single-roll process prevented
increasing the thickness of a structural shape in
one direction (e.g., the flange) without
correspondingly increasing its thickness in
another (web). This represented a very
inefficient way of producing large shapes.
By the late 1920s, however, the (Gray) wide
flange mill had arrived on the scene. These mills
used several sets of rolls arranged in tandem.
Operators could now roll relatively wide-flange
2
shapes to 36 inches deep, while keeping the web
sections relatively thin.
These wide flange shapes revolutionized the
steel building industry and made possible the
construction of the Empire State Building and
the 30 or so other skyscrapers that dotted New
Yorks skyline in the late 1930s.
Meanwhile, during this same period, some far-
sighted designers were advocating welding, and
fabricators recognized the advantage of using
fewer materials with a method that would gain
100 percent elastic continuity. This, coupled
with the development of the Structural Welding
Code by AWS, led to the demise of riveting and
the development of the all-welded joint.
Riveting, A Lost Art
The earliest methods of joining steel were by
rivets and bolts. Rivets were used before 1850 in
boiler work and shipbuilding. Mild steel bolts
became popular around the same period for
joining heavy machine parts.
During the early building industry, before the
development of todays high-strength bolts,
rivets were considered stronger than bolts
because they filled the holes and prevented any
slippage of the joint. Consequently, they
enjoyed almost exclusive use in the first steel-
framed buildings. The exception generally
occurred in the connection of beams to girders,
where the loads were relatively small and the set-
up time to drive a few rivets per connection was
too expensive.
Existing riveted bridges have been retrofitted by
the addition of plates and the replacement of
rivets with high strength bolts. This process
requires a rigorous survey locating the center of
each existing rivet. These rivet locations are
used in the detailing and fabrication of new
plates. Rivets are then removed and holes are
generally reamed for new bolt installation.
Reaming is necessary because rivets were very
forgiving to hole tolerances, and the existing
holes on all plies are not necessarily aligned.
High Strength Bolting
By 1950, high strength bolts were being given
strong consideration as a replacement for rivets
in high-rise buildings. A development of the
railroad industry and used in heavy machinery,
these bolts have a tensile strength on the order of
approximately twice their mild steel
counterparts. For high strength bolts, yield or
proof strength is important. A325 is 105 min ksi
tensile, and A490 is 150 ksi min tensile. The
proof value of A325 is 2x that of carbon steel
A307 bolts, and the proof value of A490 is more
than 3x that of the A307 bolts. High strength
bolts are torqued so tightly that a tremendous
clamping force is produced. Load is initially
transferred between members by friction
between the two mating surfaces, rather than by
bolt bearing. This friction produces a stiffer and
more rigid joint, and hence there is less
distortion of the frame. If the slip load is
reached, the bolts go into bearing. For seismic
frames, this bolt slippage is considered a benefit
as it dissipates energy. The Structural Engineer
is cautioned, however, to consider the global
distortion of the entire frame in the event that
slippage occurs. Currently, the AISC Seismic
provisions allow for use of bearing values to
develop seismic forces, but bolts shall be
tightened to slip critical specifications.
Today, high strength bolts are again becoming
popular for field connecting moment frame
connections and lateral bracing of heavy and tall
buildings. Prior to the Northridge earthquake,
high welding deposition rates allowed the
welding on brace frames and moment frames to
keep up with the erection of buildings. The
lessons learned from the Northridge earthquake
taught us that better joint details, high notch-
toughness electrodes, along with more rigorous
visual inspection, would be required to ensure
seismic connection performance. These higher
notch toughness electrodes have relatively
slower deposition rates. Hence, the field welding
process is a more time consuming operation due
to these developments. Welding can often be the
critical path task on completion of the
construction of a steel frame. Due to these new
developments, and depending on the regional
demand and supply of qualified welders, some
erectors would prefer bolted connections on a
moment frame or braced frame building.
The AISC Seismic Provision currently requires
SC, or slip critical tightening on seismic joints,
but allows engineers to use bearing values for
design. Some erectors prefer to tighten by the
Turn of the Nut Method and will fully impact
tighten all bolts regardless of the SC
specification. Other erectors prefer the use of
tension control (TC or twist off) bolts or Direct
Tension Indicating (DTI Washers). Regardless
3
of the method, field bolt pretension inspection
should be limited to those connections that are
specified to be slip critical. Other appropriate
applications for slip critical bolting include
bridge connections, wind resisting connections,
and connections for mechanical equipment or
any connections where loading can be cyclic,
dynamic or vibratory.
Bearing bolts should be used wherever SC is
not required. The most efficient use of bearing
bolts is the X specification, which requires the
threads excluded from the shear plane. The
AISC Manual of Steel Construction, LRFD
Third Edition, has a Table 7-2 providing
parameters for threads excluded from the shear
plane, based on grip and ply thickness. The
erector should be able to describe their quality
control procedure in writing, to ensure the use of
the proper length bolt for a given grip.
Alloying The Key to
Strength and Weldability
All structural steels derive their strength
characteristics from the addition of various
alloys, especially carbon, and in the case of
A514, by heat treatment, in addition to alloying.
Carbon, in the form of iron carbide, or in
solution, is the basic alloying agent for hardening
steel and increasing its tensile strength. This is
done at some sacrifice to its ductility (its ability
to stretch without failure), and its weldability.
Manganese is also a powerful hardening agent,
serving as a ferrite strengthener. Molybdenum,
chromium, vanadium and columbium are also
added in small amounts, and they increase
strength and toughness.
The hardness of structural steels is important as
it relates not only to strength, but also
weldability. Too hard a steel leads to difficulties
in welding and to weld defects, such as brittle
fracture and shrinkage cracking. Hardness does
not cause weld defects but higher strength
decreases the tolerance for flaws and increases
the level of restraint. Thus, more precaution is
needed for higher strength steel. To ensure good
weldability, ASTM standards limit the amount of
alloying elements added. Carbon Equivalent
discussed below, is a quantitative representation
of the alloying elements in a particular steel,
useful for judging weldability.
Following is a quick reference of various
alloying elements and their roles in structural
steel:
Carbon, Manganese:
Basic Hardenability, Strengtheners
Vanadium, Columbium (also known as
Niobium): Carbide & Carbo-Nitride
Stabilization, Strengthening, Grain size control
Nitrogen: Carbide & Carbo-Nitride
Stabilization, Strengthening,
Chromium Molybdenum, Nickel, Boron:
Hardenability Agents, Tempering Control
Nickel:
Improves Notch Toughness
Silicon: Deoxidizer
Aluminum: Deoxidizer, and grain refiner
Copper: Helps resist atmospheric corrosion and
is a precipitation strengthener
Carbon Equivalent and Preheat
Dr. Winterton of the British Welding Institute,
while studying the effects of various elements in
the chemical composition of steels on their
mechanical properties, noticed their effect on
hardness. Knowing that hardness is related to
weldability and susceptibility to cracking, he
developed the relationship of the chemical
elements in steel to its hardness through a
Carbon Equivalent formula, basically because
carbon has the greatest effect on hardness,
strength and weldability in steels. Others in the
USA during WWII and immediately after also
developed carbon equivalent formulas.
Many different Carbon Equivalent (CE) formulas
are used as a guide for pre-heat requirements and
in welding procedures. The AWS D1.1-2002
Appendix XI5.1 Formula is used for structural
steels:
CE= C + (Mn + Si) + (Cr +Mo +V) + (Ni + Cu)
6 5 15
In % to determine the Zone requirements for pre-
heat.
Given the Carbon Content, one can refer to
Figure 1, to determine the zone classification as
a guide for the selection of either the hardness
control or the hydrogen control method of
4
determining preheat. See the AWS D1.1 for
more details.
Fig. 1 Zone Classifications of Steel
Note: Chart based on AWS D1.1 2002 AnnexXI
Fig. XI-1. This method does not take thickness
into effect.
Zone I Cracking is unlikely, but may occur with
high hydrogen or high restraint. Use hydrogen
control method to determine preheat.
Zone II The hardness control method and
selected hardness shall be used to determine
minimum energy input for single pass fillet
welds without preheat.
For groove welds, the hydrogen control method
shall be used to determine preheat.
For steels with high carbon, a minimum energy
to control hardness and preheat to control
hydrogen may be required for both types of
welds.
Zone III The hydrogen control method shall be
used.
Quenching and Tempering
for High Strength
To obtain a high-strength steel other than by
alloying it, it is possible to heat-treat certain steel
formulations by quenching and tempering.
Quenching hardens or strengthens steel.
Tempering increases ductility but lowers tensile
strength. Temperature and time are important.
This procedure requires the steel to be heated
initially to an elevated temperature above the
upper critical to form a crystalline structure
known as austenite. Rapid cooling in water to
produce martensite or a partially martensitic
microstructure follows this. When tempered to
precipitate a fine dispersion of carbides, this
structure has good ductility and fracture
toughness, along with high strength.
Of the many grades of structural steel, only
A514, A852, A709 and A-913 are quenched and
tempered. Yield strengths are on the order of 90
to 100 ksi, almost three times that of A36 steel
and twice that of other grades. Of course, the
cost of this special treatment is reflected in a
significantly higher price.
Todays Structural Steels
At the turn of the century, there were only two
grades of steel being produced: A-7 for bridges
and A-9 for buildings (See Table A). These
materials were so similar in chemical
composition and mechanical properties that in
1939 they were consolidated by ASTM into one
standard, A-7, which was used for both types of
applications until 1960. When a higher strength
was needed, primarily for bridges, silicon steel
was specified. It was very difficult to weld so it
was primarily used in riveted and bolted
construction.
By the 1950s, the strength and economics of
welding were becoming apparent. Despite
earlier bridge failures in the railroad industry,
which were more related to poor workmanship
and technique than to materials, welding was
being tried in some high-rise structures on the
Gulf Coast. Here, structural engineers applied
the experience gained in the welding of refineries
and oil pipelines to high-rise buildings.
Designers soon noticed that while most of the A-
7 steel being used showed good welding
characteristics, there were instances in which
some A-7 steel presented welding problems
because of the limited chemistry requirements.
Thus, extensive research began into the
metallurgy of steel to develop structural
materials that were uniformly weldable.
By 1964, the American Institute of Steel
Construction (AISC) had adopted five grades of
steel for structural application. Table B shows
the chemical composition and some of the
mechanical properties of these high strength, low
alloy steels. The high elongation property of
todays steel (up to 25 percent) permits large
overstress, due to welding and deformation
during construction, without losing its ultimate
strength. Figure 2 compares typical tensile
stress-strain curves for these steels.
5
Today, the ASTM Specifications for Structural
Steel cover many carbon steels, high-strength,
low-alloy steels and some quenched and
tempered construction alloys. Structural steels
include plates, bars, shapes, pipes, and structural
tubing. There are many acceptable processes for
welding these steels.
Materials Specification
The following tables and sections discuss various
structural materials, their properties, their
availability in shapes and plates, and
applicability for use. The reader should also
refer to the AISC Manual of Steel Construction,
LRFD Third Edition, page 2-20 through 2-29 for
a complete listing and discussion of applicable
and available materials.
Fig. 2 Engineering stress-strain tensile curves
for ASTM structural steel grades A-514, A-572
GR.60, A-441, A-588, A-572 GR.50 and A-36.
Special Material Considerations,
Supplementary Requirements
In certain circumstances, Charpy V notch
toughness testing of base material is required. It
is required to be specified in the contract
documents when applicable, per the AISC Third
Edition LRFD Specification Section A3.1c.
The general requirement currently applies when
ASTM A6 Group 4 and 5 rolled shapes (see
Table 2-4 page 2-27 of AISC Manual of Steel
Construction, LRFD Third Edition for Group
definitions), or plates exceeding 2 in. thick, are
used as members subject to primary tensile
stresses due to tension or flexure, and are spliced
using complete-joint-penetration groove welds.
The impact test shall meet a minimum average
value of 20 ft-lbs. absorbed energy at 70 degrees
F. See the AISC specification Section A3.1c and
corresponding commentary for a more complete
discussion.
For Seismic applications, the 2002 AISC
Seismic Provisions for Structural Steel
Buildings, Section 6.3 specifies that for members
of the Seismic Load Resisting System, ASTM
A6 Groups 3,4 and 5 shapes with flanges 1 in.
thick and thicker, and plates that are 2 in. thick
or thicker shall have a minimum Charpy V-
Notch toughness of 20 ft-lbs. at 70 degrees F,
determined as specified in LRFD Specification
Section A3.1c. An important consideration is
the frequency of testing. AISC specifies P, Piece
frequency for thick plate. Alternatively, heat
testing frequency of thick shapes substantially
reduces testing costs and merits consideration.
The toughness requirement is intended to ensure
a reasonable toughness of the base metal to
preclude cracking. The steel must be able to
withstand the tremendous tensile strains due to
weld shrinkage generated during complete
penetration welding of these thick members.
The web to flange intersection of these heavy
hot-rolled shapes, as well as interior portions of
heavy plates may contain a coarser grain
structure and/or a lower toughness due to a
slower cooling rate, as well as other factors.
6
Table A
Historical Background of Structural Steels
ASTM REQUIREMENTS
Tensile Yield Point Elongation
Dates Specification Definition Strength ksi ksi Minimum & Chemistry
1900- ASTM-A7 Soft to medium 52-70 32-35 No req. but
1904 steels usually listed

ASTM-A9 Medium Steel 60-70 35 No req. but
Buildings usually listed
NOTE- Basic Unit stresses recommended by mgf: bldgs.-16.0 ksi, bridges, 12.5 ksi
_________________________________________________________________________________________________________
1905- ASTM-A7 Structural 60 No req. But Not req. but
1913 Bridges steel desired Reported usually reported

ASTM-A9 Medium & 55-65 1/2 x T.S. Not req. but
Buildings Structural steel usually reported
_________________________________________________________________________________________________________
1914- ASTM-A7 Structural 55-65 1/2 x T.S. No req. but
1934 Bridges steel usually reported

ASTM-A9 Structural steel 55-65 1/2 x T.S. No req. but
Buildings steel 30 min. usually reported
NOTE- AISC 1923 Basic Unit Stress =18.0 ksi
_________________________________________________________________________________________________________
1934- ASTM-A7 Structural 60-72 1/2 x T.S. Usually reported
1938 Bridges steel 33 min.

ASTM-A9 Structural 60-72 1/2 x T.S.
Buildings steel 33 min.
NOTE- AISC 1936 Basic Stress Unit =20.0 ksi
_________________________________________________________________________________________________________
1938- ASTM -A7 Structural 60 -72 1/2 x T.S.
1949 Buildings & Bridges steel 33 min.
NOTE- ASTM-A7 & A9 consolidated into one spec (ASTM A7)
_________________________________________________________________________________________________________
1954 ASTM-A373 Structural 58- 72 32 EI.=24% ( 8" Ga.)
steel
NOTE:Revised 1958; phased out when A-36 was issued (1960)
1957 ASTM -A572 Structural 60-80 42-65 15-20% (8"Ga.)
steel 17-24% (2" Ga.)
NOTE- Structural steel for riveted, bolted or welded construction
1960 ASTM-A36 Structural 60-80 36 ksi 20% (8" Ga.)
Buildings & Bridges steel min. 23% (2"Ga.)
NOTE- Issued 1960, revised 1961 (called out as weldable)
C =.28, Mn =.80-1.20, P =.04, S =.05, Si=.15-.30, Cu =.20 if specified
1961-A7: Tensile ksi - 60-75, Fy =33 ksi, Elong (8") =21.0%
C =Not specified, Mn =N.S., Phos =.04, S=.05, Cu =20 min. when specified
_________________________________________________________________________________________________________
1988 ASTM-A36 Structural 58-80 36 ksi 20% (8" Ga.)
Buildings & Bridges steel min. 23% (2" Ga.)
NOTE- Structural steel for riveted, bolted or welded construction
Nominal chemistry % ( Refer to ASTM specs for detail)
C =0.26-0.29, Mn =.80-1.20, Phos =0.04, S =0.05, Si =0.15-0.40, Cu (when specified) =0.20 min.
_________________________________________________________________________________________________________
1991 ASTM-A913 Structural 65-90 50-70 14-18% (8" Ga.)
steel 16-21% (2" Ga.)
NOTE- High strength quenched and self tempered Structural Steel
_________________________________________________________________________________________________________
1999 ASTM- A992 Structural 65 50-65 18% (8" Ga)
Steel 21% (2" Ga)
NOTE- Structural steel developed for seismic applications; killed, max carbon equiv =0.45 to .47%
Max yield point =65, max yield to tensile ratio=0.85
_________________________________________________________________________________________________________
NOTE- based on data from AISC: Iron & Steel Beams- 1873 to 1952; past issues of ASTM Specs, AWS Codes
(including First Edition - 1928), AISC Specs and other publications on steel.
7
Table B
ASTM Specifications
ASTM SPEC.# Mechanical Properties Chemical Requirements %
Tensile Yield Str. Elong % 8" uno C Mn Phos. Sulf. Si Cu Other & Notes
ksi ksi Ksi
Plates and Shapes

A36
58-80 36 min. 20 0.26 0.80-1.20 0.04 0.05 0.15-
0.40
0.20 min
Plate to 8", channels and angles

A913 65-90 50-70 14-18 0.12-0.16 1.60 0.03-0.04 0.03 0.40 0.35-0.45 Special Order- Quenched and self tempered

A992 65 50-65 18 0.23 0.5-1.5 0.035 0.045 0.40 0.6 Wide Flange- common specificaton

A572 60-80 42, 65 15- 20 0.21-0.26 1.35 0.04 0.05 0.15-
0.40
Wide flange, Plate Gr. 42 To 6", Channel
and angle in large special orders only

A588
63-70 42,46,50 18 0.10-0.19 0.50-1.35 0.04 0.05
0.15-
0.90
0.20-0.50
Plate to 8" and all Shapes-
Corrosion Resistant - special order

A709 58- 130 36-100 18-20 Large variations by grade, see ASTM spec Bridge steels to 4" PL- special order
Weathering steel in Grade 50, 70 & 100
A759 varies varies -- 0.67-0.84 0.7-1.1 0.04 0.05
0.10-
0.50
-- Crane rails- design by manufacturers

A852
90-110 70 19 0.19 0.8-1.35 0.035 0.04
0.20-
0.65
0.20-0.40
Corrosion resistant Bridge Steel -special
order

HSS Hollow Sections
Round Rect
Round and Rectangular tubes:
A500 Grade A
45 33 39 (2) 25 0.26 -- 0.035 0.035
--
0.20

A500 Grade B
58 42 46 (2) 23 0.26 -- 0.035 0.035
--
0.20 Most available and commonly used HSS
A500 Grade C
62 46 50 (2) 21 0.23 1.35 0.035 0.035
--
0.20
A500 Grade D
58 36 36 (2) 23 0.26 -- 0.035 0.035
--
0.20

A501
58 36 (2) 23 .
0.26 -- 0.04 0.05 0.20
A618 65-70 50 (2) 22 0.15 1.0 0.07-0.15 0.025 -- 0.20
A847 70 50 (2) 19 0.20 1.35 0.15 0.05 -- 0.20 High strength and corrosion resistant
Steel Pipe
A53 Grade B 60 35 See ASTM 0.25 0.95 0.05 0.045 -- 0.4
Most available and commonly used
structural steel pipe
NOTE - This data is selected information intended for a guideline reference; refer to ASTM for more information.

8
Thicker members have a greater chance that
inclusions and discontinuities will be larger and
more prevalent. This is due to the fact that the
steel is produced with less working of the hot
metal during rolling. Also, thick sections cool
more slowly and this also adversely affects
strength and toughness. Therefore, steel ordered
from the mill with specified notch toughness
requirements will most likely be fine grain-
killed steel, as discussed briefly under ASTM
A992.
Other Supplementary Requirements are listed in
ASTM A6 and are for use at the option of the
purchaser. These requirements must be specified
in the contract documents. Those supplementary
requirements that are considered suitable for
each different material are listed in the ASTM
under each specification, and have been included
in this publication for the convenience of the
reader. It should be noted that a cost might be
associated with each supplementary requirement.
Thus, they should be used sparingly and with a
clear understanding of the costs and benefits of
their use.
A major change in steel production that has
occurred during the last decade or so, has been
the use of continuous cast preformed slabs and
shapes rather than casting of ingots. Mills
continuously cast wide slabs to be rolled to plate.
The slabs are several times thicker than the final
plate for thinner plates. For thick plates the slabs
may be only a few times thicker than the final
plate. Similarly, WF shapes are rolled from a
preformed cast shape similar to an H.
The continuous casting process is subject to
center shrinkage problems, as are other castings
and ingots, if the continuous process is not
carefully controlled. The center or mid-thickness
shrinkage defect can persist to the finished plate
or shape as a plane of little or no strength. This
mid-thickness problem appears to be more likely
if the steel is thick and not severely hot worked.
Supplementary requirement S8, Ultrasonic
Testing, using standard testing procedures and
acceptance criteria such as A435 and A898 will
not cause rejection of steel with mid thickness
defects, even those that cause the steel to have
almost no strength and ductility in the through
thickness direction. The mid-thickness defects
are similar to laminations. UT testing
procedures may identify the mid-thickness
reflectors as flaws but they may not be
rejectable.
Better assurance of through thickness properties
can be achieved by specifying that the steel shall
have through thickness tensile tests in
accordance with A770. This is a costly item and
should be used only for conditions with through
thickness loads and in areas of greatest concern.
Schedule impacts need also to be considered.
ASTM A36 (as rolled)
Once the most commonly used steel in building
construction, A36 is a material that has seen its
use change quite drastically in the past 10 years.
Due to the use of recycled steel in the production
of new wide flange shapes in the modern mini-
mills, most A36 can also qualify under ASTM as
A575 Grade 50. Since A36 has no upper limit
on yield strength, the term dual grade was
termed and used to represent this structural steel.
A36 was once the most economical steel to be
used in floor systems, but soon engineers would
take advantage of the inherent over-strength and
specify A572 Grade 50 for the same material
cost. When used in the design of moment
frames, this dual grade material posed a problem
in that yielding would occur at a higher force
level than anticipated in the design.
Furthermore, the Northridge earthquake brought
to light other physical and chemical properties
that were unfavorable to welded connections that
were expected to go past their elastic limit.
These developments brought about A992 (see
section following). Plate, angle, and channel,
however, are still produced. typically as A36,
and their use as such is still the most economical.
9
Of the standardized supplementary requirements
(that may be specified in the contract documents)
listed in ASTM A6, the following is considered
suitable for use with ASTM A36.
S5. Charpy V-Notch Impact Test
A992 is relatively new steel developed from the
lessons learned after the Northridge earthquake.
This specification provides improved properties
for wide flange shapes in welded moment frame
and brace frame construction, by giving
engineers a more reliable limited range of force
levels for design, with no significant additional
cost. Furthermore, its chemical properties
provide for excellent weldability. Derived from
A572 Gr. 50 with special requirements as per
AISC Technical Bulletin #3, March 1997, this
steel is specified to provide a minimum yield
strength of 50 ksi, a maximum yield strength of
65 ksi, and a minimum tensile strength of 65 ksi.
It has a maximum yield to tensile ratio of 0.85,
and a maximum carbon equivalent of 0.47% (not
required of A572 Gr. 50). The carbon equivalent
is defined and discussed in a subsequent section
of this publication.
This steel is specified to be killed. Killed steel is
steel that is deoxidized, either by the addition of
strong deoxidizing agents, or by vacuum
treatment, to reduce the oxygen content to such a
level that no reaction occurs between carbon and
oxygen during solidification. As such, the steel
shall be affirmed in the test report by a statement
of killed steel, a value of 0.10% or more for the
silicon content, or a value of 0.015% or more for
the total aluminum content.
A992 is also specified to contain no greater than
0.012% nitrogen, or it shall be made to a practice
producing nitrogen no greater than 0.015% and
contain one or more nitrogen binding elements.
listed in ASTM A6, the following are considered
suitable for use with ASTM A992:
S1. Vacuum Treatment
S2. Product Analysis
S8. Ultrasonic Examination
A572 is a carbon manganese steel, augmented
for strengthening by columbium, vanadium and
nitrogen (optional) additions. A572 covers five
grades ranging from 42-to 65-ksi minimum yield
strengths, depending on plate thickness and
product size, Grade 50 being the most common
in building construction.
It is available in rolled shapes, plates, sheet
piling, and bars. Grades 42, 50, and 55 are
intended for bolted, or welded structures. Grades
60 and 65 are intended for bolted construction of
bridges, or for bolted, or welded construction in
other applications. A572 has a constant
minimum yield strength within any one grade.
For example, A572 grade 42 has 42-ksi
minimum yield strength for all plate thicknesses
to 6. Increases of minimum yield strengths for
A572 grades to 65 ksi are accomplished by
increasing maximum carbon content from .21%
(grade 42) to .26% (grade 65), plus other
chemistry adjustments within the specifications.
Limits for carbon and manganese, plus grain size
control, provide good weldability for this grade.
Note, the maximum thicknesses for A572 grades
are: Grade 42: 6 in, Grade 50: 4 in, Grade 55: 2
in, Grade 60: 1-1/4 in; and Grade 65: 1-1/4 in or
less, subject to composition.
S18. Maximum Tensile Strength
ASTM A529
A529 is a carbon manganese, killed steel
available in 50 and 55 ksi yield strengths in
shapes, plates, and bars for use in bolted or
welded construction of buildings and general
structural purposes.
It is available in both grades for rolled shapes of
Groups 1 and 2, and plates 1 inch thick and 12
inches wide. Bars are available in grade 50 to 2
inches thick, and in grade 55 to 1 inches
thick. A common use for A529 in building
construction are bars (standard width flat bars)
used for shear plates and stiffeners.
10
S78. Maximum Carbon Equivalent
ASTM A913 (quenched and
self-tempered)
This specification covers high strength, low alloy
steel shapes in Grades 50,60,65, and 70 produced
by the quenching and self-tempering (QST)
process. The QST process, evolved from the
thermo-mechanical control processes (TMCP),
produces fine-grained steel by a combination of
chemical composition and integrated controls of
manufacturing processes from ingot or bloom
reheating to in-line interrupted quenching and
self-tempering.
The members may not be heated to temperatures
exceeding 1100 deg. F. during or after welding
or other fabrication processes. This steel is ideal
for use in large axially loaded columns or truss
chords. Its weldability is excellent due to
relatively low carbon equivalent of between
0.38% to 0.45%, depending on the Grade.
Preheat is not required by AWS, however good
practice would dictate enough preheat to remove
the moisture from the steel prior to welding.
S3. Simulated Post-Weld Heat Treatment of
Mechanical Test Coupons
This steel has essentially been
superseded by ASTM A709, Grade 50W
and by A588.
A high strength, low-alloy steel with enhanced
atmospheric corrosion resistance of
approximately two times that of carbon structural
steel with copper, the same as A588 (or four
times carbon structural steel without copper). It
has three grades of yield strengths; 42,46 and 50
ksi, depending on thickness, and is weldable
with proper welding procedures, but is limited to
material up to 4 inches in thickness. It is
available in shapes, plates and bars for welded or
bolted construction intended primarily for use as
structural members where savings in weight or
added durability are important.
tempered)
This steel has essentially been
superseded by ASTM A709 Grade 100
for bridges.
A514 covers a number of low-alloy grades,
(basically manufacturers recipes), with a
variety of alloying elements. Since A514 is a
quenched and tempered steel with 90-100 ksi
minimum yield strength, depending on plate
thickness. Alloying elements are balanced to
provide hardenability, tempering and notch
toughness controls:
Tempering at the mill is performed at rather high
temperatures (1200-1250 degrees F) for steels
included in this grade. However, any
temperatures in excess of the tempering
temperature will reduce the strength. Therefore,
exposures to heating such as in welding cycles
must be controlled in order not to soften the
hardened structure, thereby lowering the
strength. Similarly, overheating may transform
the steel structure and result in a structure that is
too hard and brittle. Controlled welding
techniques can be expected to produce
consistently good results for this steel.
A-514 is quite suitable as a structural material,
but is intended mainly for welded bridges. This
steel is not available in wide-flange or hot-rolled
shapes, and therefore is not commonly used in
building construction. Aside from cost
considerations, the resulting lightweight
structures may produce higher deflections in
long span members with consequent undesirable
vibration characteristics.
ASTM A709 (as rolled and
quenched and tempered)
These structural steel specifications cover carbon
and high-strength, low-alloy and quenched and
tempered steels in seven grades with four yield
strengths, available in plates and shapes for use
in bridges. Grade names are 36,50,50W,HPS,
11
50W,HPS 70W, 100,100W. These steels are
basically A36, A572, A588, or A514, but with
minimum impact test requirements for non-
fracture critical and fracture critical applications
in Zones 1,2&3 as set by the American
Association of State Highway and
Transportation Officials. Grades 36 and 50 are
semi-killed or killed. Grades 50W, HPS 50W
and HPS 70W are made to fine grain practice.
Grades HPS 50Wand HPS 70W are made using
a low hydrogen practice. Grade 100 and 100W
meet the requirements for fine austenitic grain
size per A6. Grades HPS 50W, HPS 70W, may
be furnished as rolled, controlled rolled, thermo-
mechanical control processed, or quenched and
tempered. Grades, 100 and 100W are heat
treated and quenched and tempered.
S60. Frequency of Tension Tests
S83. Non-Fracture-Critical, T, Material;
Toughness Tests and Marking
S84. Fracture-Critical, F, Material; Toughness
Testing and Marking
S92. Atmospheric Corrosion Resistance
S93. Limitation on Weld Repair (Fracture
Critical Material Only)
ASTM A759 (controlled
cooled)
This specification covers carbon steel crane rails
of special designs only for crane runway use.
Design details for the special crane rails are
given in the crane rail catalogs of individual
manufacturers and are referred to in the AISC
manual as well as in other publications. This
steel is not listed in AWS D1.1. The high carbon
content indicates that careful consideration be
given to alternatives and potential problems
before specifying that these rails be arc welded.
S1. End Hardening
S2. High strength rails (heat treated, head
hardened, or alloy rails)
S3. Chamfering (ends)
S4. Ends prepared for electric arc welding
tempered) This is A588 steel with
Q&T and is similar to A709 70W
This specification covers quenched and tempered
high strength, low-alloy structural steel plates for
welded, riveted, or bolted construction. It is
intended primarily for use in welded bridges and
buildings where savings in weight, added
durability, and good notch toughness are
important. The atmospheric corrosion resistance
of this steel in most environments is substantially
better than most carbon steels. Other properties
include impact toughness meeting 20 ft-lbs at 50
degrees F, and a fine austenitic grain size
produced by fine grain practice and heat
treatment.
S3. Simulated Post-Weld Heat Treatment of
Mechanical Test Coupons
S6. Drop-Weight Test
This specification covers seamless and welded
black and hot-dipped galvanized carbon steel
pipe in nominal pipe size (NPS) 1/8 to 26
diameter with nominal wall thickness from .068
to 2.344 depending on diameter. Pipe weight
class is Standard (STD), Extra Strong (XS), and
Double Extra Strong (XXS). This pipe is very
popular for structural use as its chemistry and
mechanical properties for Types E and S, Grade
B, are very similar to ASTM A36. Grade B, the
most common structural pipe used, has a
minimum yield strength of 35 ksi, and a tensile
strength of 60 ksi.
ASTM A500 (cold-formed
structural tubing)
This specification covers cold-formed welded
and seamless carbon steel structural tubing in
round, square and rectangular shapes. It is
available up to a maximum periphery of 64
inches (20-3/8 diameter and 16 x16 square)
12
with a maximum wall thickness of 0.625, and
yield strength grades A, B and C of from 33 to
50 ksi, depending on shape and grade. Its yield
strengths and chemistry make it compatible with
A36 and HSLA, high strength low alloy steels.
Normally stocked in local steel service centers.
ASTM A500 Grade B, the most common
specification for rectangular tubing has a
minimum yield strength of 46 ksi and a tensile
strength of 58 ksi. The use of ASTM A500
Grade C would most likely be a special order.
ASTM A501 (hot-formed
structural tubing)
This specification covers hot-formed welded and
seamless carbon steel structural tubing in round,
square, rectangular, or special shapes for general
structural purposes. Round tubing is furnished in
nominal pipe-size (NPS) to24 diameters, in
wall thicknesses of 0.109 to 1.000 depending
on size. Square and rectangular tubing is
available in sizes 1 to 10 across the flat sides,
in wall thicknesses of 0.095 to 1.000 and may
be furnished with hot-dipped galvanized coating.
It has one yield strength, 36 ksi, and chemistry
comparable to A36. Check for availability.
ASTM A618 (hot-formed
structural tubing)
This specification covers hot-formed welded and
seamless high-strength low alloy steel structural
tubing in round, square, rectangular, or special
shapes for general structural purposes. It has
yield strengths of 46 to 50 ksi. Check for
availability.
ASTM A847 (cold formed
structural tubing)
This specification covers cold-formed, welded
and seamless high strength, low-alloy round,
square, rectangular, or special shaped structural
tubing for welded, or bolted construction of
bridges or buildings and for general structural
purposes where high strength and enhanced
atmospheric corrosion resistance are required.
Generally, tubing is available in welded sizes
with a maximum periphery of 64 inches, a
maximum wall thickness of 0.625 inches, and in
seamless with a maximum periphery of 32
inches and a wall thickness of 0.5 inches.
Tubing in other dimensions and special shapes
may be available by inquiry and negotiation with
the manufacturer. Yield strength is 50 ksi, and
the tensile strength is 70 ksi.
ASTM A27 (steel for
castings)
This specification covers carbon steel castings
for general applications that require up to 70 ksi
minimum tensile strength. Castings can be an
effective solution to highly restrained welded
joints when the amount of weld or complexity of
the joint becomes extreme. Castings allow the
designer to tailor the geometry of the node and
thereby directly design the stress state. The
following paragraphs are excerpts from a paper
on Steel Castings presented at the 2003 North
American Steel Construction Conference (please
see reference list):
Many believe that cast steel is brittle because the
cast iron that is commonly used in automotive
and household goods, like cookware, easily
cracks. However, the properties of steel are very
different from iron. Steel castings can meet or
exceed the ductility, toughness, or weldability of
rolled steels. Technically, all steel is cast.
Designers generally think of design requirements
in terms of strength, but the design is commonly
constrained by modulus of elasticity, fatigue,
toughness or ductility. Increasing the strength of
steel normally reduces the ductility, toughness,
and weldability. It is often more desirable in
steel casting design to use a lower strength
grade and increase the section size or modify the
shape. The design freedom makes castings an
attractive way to obtain the best fabrication
material performance and the needed component
stiffness and strength.
Rolled sections of steel have their structure
elongated in the direction of rolling. The
strength and ductility is improved in that
direction but they are reduced across the rolling
direction. The lack of a rolling direction in steel
castings gives them uniform properties in all
directions. Rolling steel cold can also
strengthen the steel but reduces ductility and
toughness. Cast steel grades achieve the same
trade off by alloying and heat treatment.
Steel castings are used in demanding
applications that are safety critical, highly
specified, and performance demanding. A
railroad coupler is a good example of a common
application that is critical. Castings are used in
13
high-pressure service in nuclear power plants.
The use of steel castings in pressure containing
systems is common and specified in the ASME
Boiler and Pressure Vessel Code. One aspect of
the ASME code is the requirement that suppliers
develop and demonstrate a weld procedure
including welded properties for the components
and materials they supply. The cast carbon
steels that would be used in building
construction are already well known and
established in the Code, including their design
requirements and welded properties.
The biggest advantage in quality that forged or
rolled shapes have over steel castings is their
ability to begin with a simple optimal casting.
The ingot or bar can be easily inspected prior to
rolling or forging. The use of casting processes
to make uniquely designed shapes requires
inspection that is correlated to the casting
process, part design, and performance
requirements. Often the purchaser of steel
castings uses nondestructive examination,
mechanical testing, and engineering analysis to
ensure the desired reliability.
Steel casting producers routinely test each heat
of steel to make sure it meets the mechanical
properties required in the material specification.
The heat is also analyzed chemically to certify
that it meets the standard. Other specialized
tests can be required like low temperature
impact testing when service performance
requirements dictate. The dominant material
used in building construction is carbon steel
because of its reliable properties, low cost, and
ease of fabrication. One common grade used for
building construction in rolled sections is ASTM
Specifications A36. The use of steel castings is
permitted in building construction, using
material from either ASTM A27 Grade 65-35 or
ASTM A148 Grade 80-50 (AISC, 1998). The
properties of common steel depend on the
composition and heat treatment. Because
designers use yield strength as a basic property
in design, often material is ordered to higher
strength without considering the advantage in
castings of using a lower strength material with
optimum ductility and weldability. Since the
load-carrying cross-section can be increased to
accommodate lower strengths, the casting can be
supplied in the highest ductility with strength
levels that are compatible with the rolled
structural shapes. This use of cast carbon steel
in its optimal condition makes sure that the
casting will perform safely and reliably and that
excessive loads will cause failure to occur first in
the rolled section familiar to the designer. The
use of ASTM A27 Grade 65-35 in the normalized
and tempered condition will give a strong,
ductile, weldable steel.
Traditionally, nondestructive testing has been
used to certify casting quality. Soundness is
verified through the use of radiographic
inspection. Surface quality is evaluated using
magnetic particle inspection. More recently, the
use of computer simulation of solidification of
the casting integrated with finite element
analysis of its performance has been used to
design optimal casting configurations. The
development of these tools allows the designer to
ensure that critical areas of the part meet
requirements while ensuring the most
economical means of manufacturing the whole
part. Additional information covering steel
castings is available from the Steel Founders
Society of America at http://www.sfsa.org.
The Welding Processes
Along with the development of better structural
materials came improved methods of joining
steel. Of the methods described below, some are
restricted to shop use only; others find
widespread use in both the shop and the field.
There are four pre-qualified welding processes:
Shielded Metal Arc Welding (SMAW)
Flux Cored Arc Welding (FCAW)
Gas Metal Arc Welding (GMAW in the
spray mode)
Submerged Arc Welding (SAW)
Other welding processes such as Electro Slag
Welding (ESW), Electro Gas Welding and Gas
Metal Arc Weldingshort circuit mode can be
qualified-by-the test.
Electrodes
There are many different types and brands of
electrodes used by the industry. The structural
engineer needs only to specify the design
strength and the notch toughness requirement for
his or her design. The design strength is based
on matching filler metals to the base metals, (see
Table C), and the notch toughness demands
depend on the use of the structure. For seismic
design, Section 7.3a of the AISC Seismic
Provisions requires electrodes to meet 20 ft-lbs
@ minus 20 degrees F. Given these
specifications, it is then the responsibility of the
steel fabricator and erector, to choose the
electrode that best suits the skills of their
14
welders, accommodates their equipment, and is
appropriate for the weld type and position in
which it is used. Electrode storage and exposure
limitations vary by electrode. FEMA 353
Appendix D recommends a testing protocol, to
be performed by the manufacturer, to establish
acceptable limitations on electrode exposure to
the atmosphere. The fabricator or erector must
then work within the limitations of these test
results after packaging has been broken. The
engineer should request electrode product data
sheets for review. This data provides
recommended welding parameters along with the
electrode exposure limitations. Welding
procedures should be reviewed to work within
the parameters of the product data sheets.
Shielded Metal Arc Welding (SMAW)
Commonly referred to as stick or manual
welding, this is the oldest and was the most
popular method of structural welding and
involves the use of flux-coated electrode (stick),
which is consumed in the process, see Fig. 3.
SMAW has a long and successful history. As
compared to other processes, it is ideal for weld
joint repairs and light field applications. The
equipment is less costly; it is more portable;
electrodes may be purchased off the shelf in
most locations. SMAW is particularly
advantageous when the job involves repetitive
starting and stopping, as in short fillet welds.
Stick welding, however is slower and more
costly than other methods of welding, and is
more dependent on operator skill for high quality
welds.
Fig. 3 Schematic Illustration of Shielded Metal-
Arc Welding (SMAW). Reverse polarity is shown
(D.C. electrode positive)
Developments through the years have been
directed at minimizing the formation of
hydrogen gas by removing water from the
electrodes (low hydrogen). Therefore common
field practice calls for the electrodes to be dry
before use. The low hydrogen electrodes are
normally shipped in hermetically sealed
containers. Low hydrogen electrodes have these
designations: EXX15, 16, 18, 28 and 48. The
XX indicates the tensile strength in ksi. After a
container is opened, and electrodes are exposed
to the atmosphere beyond the manufacturer or
AWS specified time limit, electrodes are dried
with the use of on-site ovens, or discarded. Low
hydrogen electrodes that are shipped in other
than hermetically sealed containers are required
to be baked in special high temperature ovens
prior to use. The important exceptions to the
electrode drying/ baking are cellulose-coated
electrodes (for example, E6010 and E6011 that
give good penetration but are not low hydrogen
electrodes), which are compounded to contain 3
to 7 percent moisture. Redrying can actually
impair the quality of these electrodes. It is
imperative that AWS D1.1-2002 requirements
for storage handling of low hydrogen electrodes
and the electrode manufacturers
recommendations for all electrodes be followed.
D1.1 does not address electrodes other than low
hydrogen. Improper storage can lead to
hydrogen intrusion creating underbead cracking,
transverse cracking, and or porosity.
SMAW is normally used with a constant current,
drooping characteristic power source.
Flux-Cored Arc Welding (FCAW)
This process, shown in Figure 4, employs a
tubular electrode with the flux contained within
the core of the tube.
Fig. 4 Schematic illustration of Flux-Cored Arc
Welding (FCAW). Reverse polarity is shown
(D.C. electrode positive).
15

TABLE C FILLER METAL REQUIREMENTS

AWS ASTM Steel Electrode Specification & Classification
Group Spec. & Grade

SMAW AWS A5.1 E60XX, E70XX
A5.5* E70XX-X
I
A36 <=3/4
A53 Grade B A500
Grades A, B
SAW AWS A5.17 F6XX-EXXX, F6XX-ECXXX
F7XX-EXXX, F7XX-ECXXX
A5.23* F7XX-EXXX-XX, F7XX-ECXXX-XX
A501
A709 Grade 36<=3/4"
GMAW AWS A5.18 ER70S-X, E70C-XC, E70C-XM*
A5.28* E70S-XXX,E70C-XXX

FCAW AWS A5.20* E6XT-X , E6XT-XM
E7XT-X , E7XT-XM
A5.29* E6XTX-X, E6XT-XM
E7XTX-X, E7XTX-XM

A36 >3/4
A572 Grades 42,50,55
SMAW AWS A5.1 E7015, E7016, E7018, E7028
A5.5* E7015-X, E7016-X, E7018-X
II A709 Grade 36>3/4
SAW AWS A5.17 F7XX-EXXX, F7XX-ECXXX
A5.23* F7XX-EXXX-XX, F7XX-ECXXX-XX
A709 Grades 50, 50W
GMAW AWS A5.18 ER70S-X, E70C-XC, E70C-XM*
A5.28* ER70S-XXX, E70C-XXX
A913 Grade 50
A992
FCAW AWS A5.20 E7XT-X *, E7XT-XM*
A5.29* E7XTX-X, E7XTX-XM

SMAW AWS A5.5* E8015-X, E8016-X, E8018-X
III A572 Grades 60,65 SAW AWS A5.23* F8XX-EXXX-XX, F8XX-ECXXX-XX
A913 Grades 60,65 GMAW AWS A5.28* ER80S-XXX,E80C-XXX
FCAW AWS A5.29* E8XTX-X, E8XTX-XM

SMAW AWS A5.5* E9015-X, E9016-X, E9018-X, E9018-M
IV A709 Grade 70W SAW AWS A5.23* F9XX-EXXX-XX, F9XX-ECXXX-XX
A852 GMAW AWS A5.28* ER90S-XXX, E90C-XXX
FCAW AWS A5.29* E9XTX-X, E9XTX-XM

This is an abbreviated table for selected steels based on AWS D1.1:
2002, Table 3.1. See the AWS table for more detail.
* Some exclusions apply, see AWS D1.1: 2002, Table 3.1

16
There are two versions: The self shielded type
uses flux compounds alone to protect the weld
from oxidation during cooling. The gas-shielded
type uses flux compounds, plus an auxiliary
shielding gas (usually carbon dioxide) for weld
protection. To minimize spatter, a mixture of
75% CO2+25%A is becoming popular. In
general, only the self-shielded type is used for
field application due to the effect of wind. Both
are semi-automatic, high production methods.
Although equipment is bulky, FCAW is the
method of choice for high production, deep
penetration welding on low-rise and high-rise
structures. Typical deposition rates (about 8 to
12 pounds per hour) are about twice that
obtained from normal stick welding. Figure 5
compares the depth of penetration of a fillet weld
produced by SMAW and FCAW. The power
source should be a constant voltage type.
Fig. 5 Fillet welds by SMAW, left, and FCAW,
right, in A36 steel. Note the increased
penetration of FCAW.
Gas Metal Arc Welding (GMAW)
This welding process is pre-qualified when used
in the spray transfer mode; it must be qualified
by test when used in the short-circuiting transfer
mode. The electrode is a solid welding wire or
metal cored wire, and the shielding gas is Argon,
Helium, or CO2 or a combination of these gases,
and is semi-automatic (see Figure 6). The power
source should be a constant voltage type, direct
current, and is normally used in reverse polarity
(DCEP). It produces a very clean weld and
deposit rates are very good, comparable to
FCAW. Though developed primarily for the
aircraft industry, it is now very popular in
structural steel fabrication shops, but is not
practical for outside or field welding due to wind
effects.
Fig. 6 Gas Metal Arc Welding- spray arc mode.
Most structural sections for buildings and
bridges, welded in prefabrication plants or
temporary fabrication plants, use the SAW
process with a fully automatic setup. This
process, as shown in Figure 7, deposits a flux
powder in advance of the electrode, so that the
resulting arc produced is submerged in the flux
and is not visible to the operator. It is the
workhorse of the structural shop for built up
members. SAW is particularly well suited to
long welding runs of thirty feet or more. It can
be used on thin or thick sections of metal and is
capable of producing high quality fillet, partial,
and complete joint penetration welds at typically
high deposition rates, but is restricted to flat or
horizontal welding positions.
Fig. 7 Schematic illustration of Submerged Arc
Welding (SAW. Reverse polarity is shown
(DCEP).
17
Electroslag Welding (ESW)
Introduced to construction use in the 1960s,
electroslag welding is the newest production
welding process. Its chief feature is its
unsurpassed production capacity, depositing
filler metal at 35 to 50 pounds per hour, while
producing a clean, high quality joint on a
continuous casting basis. This is not a pre-
qualified process, but must be qualified by test,
and the test results submitted. As illustrated in
Figure 8, there is no arc. The slag, heated by
electrical resistance, melts the filler electrode and
parent metal. Melting of the electrode in the slag
cleans the metal by providing excellent slag to
metal contact. The parent metal, surrounding the
molten pool, is heated deeply and the resultant
slow cooling allows time for gas bubbles to
escape, keeping porosity to a minimum.
Although the process is used primarily in the
shop for butt welding of plates and for the final
closure welds of interior stiffeners used in box
columns using the key-hole weld technique, it
has been adapted for field welding of solid
prismatic members of between 4 and 6 inches
thick.
In spite of cleanliness, welded sections using the
ESW process will often show lower fracture
toughness than the parent metal, especially at
temperatures below 0 degrees F. The
mechanical properties are equivalent in all other
respects. Special ESW techniques have been
developed recently to improve weld toughness
by grain size control. A process known as
Narrow Gap Improved Electroslag (NGIE)
welding has been developed intended for use on
bridge welding. This process, still being
perfected, utilizes a narrow gap to reduce weld
consumables and corresponding heat input, as
well as a Nickel-molybdenum alloy electrode
wire to produce improved toughness.
In shops so equipped, the fabricator can also
obtain further improvements by normalizing
the weld zone. This involves reheating the
welded area to between 1650 and 1700 degrees F
to form austenite, followed by air-cooling.
Normalizing produces a uniform, refined grain
structure with improved fracture toughness. This
process is not generally used in construction
because the high cost cannot be justified.
Fig. 8 Schematic of electroslag process for
welding typical butt joint.
Note: Single or multiple electrodes may be used:
with or without oscillation for either will depend
on plate thickness. Width of gap is 1 to 1-1/2.
Techniques are being developed to eliminate the
starting sump.
Avoiding Weld Defects
Although welding has been with us a long time,
its application to structural use was impeded by
early failures in the bridge and shipbuilding
industries. Early failures had many causes
including lack of understanding in the
engineering community, poor quality steel,
inferior welding electrodes, poor details, and
workmanship. WWII era steels were less
weldable than current steels. Low hydrogen
electrodes did not become common until after
WWII. Also, there were poor designs that
caused stress risers which contributed to some
failures.
Much research has been done over the past five
decades, not only in the development of better
base materials, but also in providing electrode
filler metals that better fit the metallurgical
properties of todays structural steels. Table C
shows a variety of filler materials available for
the type of steel and welding processes being
used. Many of the weld defects found in
structural applications are caused less by the
18
quality of the base and filler metals than by poor
joint design in combination with improper
welding practice. These factors lead to a variety
of weld defects including those described in the
following sections.
Weld Cracking
Figure 9 illustrates the metallurgical features of a
typical weld. The weld area is divided into a
fusion zone and a heat affected zone (HAZ).
Fig. 9 Weld macrostructure (at 1.2X size), A36 steel, P.M1, joined to heat treated steel, P/M2. Good fusion
is indicated, accompanied by prominent heat affected zones (HAZ), which show darkest etched structures;
however, there is lack of penetration at the root of the partial penetration groove weld. Arrows designate
cracks in HAZ of P/M2. The indentations are Rockwell hardness test impressions. Joint was fabricated by
SMAW (stick welding), with E7018 electrode. (Cracking due to lack of preheat and unbalance of weld.)
The HAZ microstructure reflects changes in the
grain structure of the parent metal produced by
heat from the adjacent molten metal. In figure 9,
good fusion is indicated in the weld itself.
However, the lack of preheat allowed the joint to
cool too quickly. This in turn produced hard,
brittle martensite in the HAZ and led to cracking
in the HAZ and the parent metal. Preheat is
especially critical in thicker sections due to the
fact that more heat will be lost through
conduction than through convection.
All structural steels experience some degree of
hardening in the HAZ due to high heat inputs
during welding. The hardness increase is
proportional to the carbon content and is affected
by the alloying elements and weld cooling rates.
Some modern structural steels have rather low
carbon contents and the HAZ do not get very
hard. Q&T steels require that the HAZ have a
hardened martensite structure, otherwise the
strength and toughness will deteriorate.
Hardened HAZ are susceptible to hydrogen
related cracking and under bead cracking.
19
Specific welding controls that consider the
hydrogen potential of weld filler metal should be
used. Preheat, post heat, and slow cooling such
as with insulating blankets are sometimes used.
AWS D1.1 has prequalified preheat requirements
for non-low hydrogen covered electrodes that
require higher temperatures than for other
electrodes.
With proper weld procedures and good
workmanship, such cracking can be eliminated.
Other types of cracks can be encountered in
structural welding, such as under bead cracking,
cold cracks and hot cracks.
Hydrogen Embrittlement
When steel is melted during welding, hydrogen
may dissolve in the molten metal and diffuse into
the parent metal HAZ. Molten metal has a great
affinity for atomic hydrogen, but low affinity as
cooling takes place and hydrogen is rejected.
Some hydrogen gas may become trapped in the
weld metal and create high internal pressure
which can induce micro-cracks in the steel.
Hydrogen cracking is controlled or eliminated by
the use of low hydrogen electrodes and proper
preheat. The source of hydrogen is water, which
may originate from the weld environment, the
steel base material, the shielding of a SMAW or
FCAW electrode or the flux of the SAW process.
In the welding arc, water will break down into
hydrogen and oxygen. Hydrogen in the fusion
zone deposited metals or in the HAZ of the
parent metal can cause embrittlement. If the
HAZ has become hardened during welding, the
sensitivity to hydrogen embrittlement is even
greater.
Preheating of the metal by AWS standards and
use of low-hydrogen electrodes are the best
means of avoiding hydrogen embrittlement. See
Table D. Preheating dries the steel surface,
slows the cooling rate to limit HAZ hardening,
and retards cooling so as to permit hydrogen to
diffuse out of the steel. As mentioned earlier,
proper storage, as well as heat drying of
electrodes is desirable for removing excessive
moisture.
Lamellar Tearing
One of the most disconcerting welding defects
caused by poor joint design in combination with
bad welding practice is lamellar tearing. This
occurs in highly restrained joints because the
designer may not fully understand the
anisotropic properties of the base metal, and the
fabricator has not undertaken adequate
preventive measures during welding. Design
information through AISC is available to aid the
engineer, detailer and fabricator to reduce the
occurrence of these defects.
Fig. 10 Directional nomenclature for describing
anisotropy in rolled plates.
In structural steels, mill test reports primarily
address the longitudinal properties. Figure 10
and 11 illustrate the orientation of a steel plate
relative to the direction of rolling, or its
longitudinal (X) direction. Transverse (Y)
direction properties, while usually lower in
ductility and toughness than longitudinal
properties, are nonetheless predictably good.
Strength and ductility are the lowest in the
through-thickness (Z) direction
Certain non-metallic compounds are formed
during the alloying process and become trapped
in the steel. These so-called inclusions are
round in shape when the steel is cast into ingots,
but become flattened and elongated during
rolling. They are usually microscopic in size and
difficult at best to detect even with ultrasonic
testing.
Due to the through-thickness tensions developed
during welding of a highly restrained joint, these
flattened inclusions, usually silicates, aluminates,
sulfides and oxides, can link up to form micro-
fissures and eventually form visible cracks.
Figure 12 illustrates how lamellar tearing can
develop in a T-weld section. Following welding,
20
contraction strains are generated as the heated
metal cools. When the assembly is highly
restrained, i.e., when portions or all of the
assembly are kept from contracting during
cooling, and when the welding sequence is such
that the strain demand cannot be accommodated
by plastic deformation in some element, the
micro-fissures grow into tearing cracks.
Fig. 11 Relation between tensile properties and
angle of specimen form the plate surface to
demonstrate loss of properties in the Z direction.
The worst case of lamellar tearing is
delamination, or complete internal separation of
the steel, as shown in Figure 12. This is more
likely to occur in larger weldments using thicker
base materials because more non-metallic
inclusions collect with size and because of the
greater restraint derived from rigidity.
ASTM A770
This delamination effect occurs very rarely if the
steel is clean and virtually free of fine non-
metallic inclusions. A mill order specifying mill
testing in conformance to ASTM A770 can
provide such a steel. Although costly, this
specification will normally instigate a mill to use
special procedures such as Calcium Argon
blowing, shape control additions, and a reduced
sulfur content to satisfy a through thickness
tensile test with a minimum 20% reduction in
area.
Other precautionary measureswould include:
Careful detailing of weld joint
preparations and weld root openings to
minimize weld volume
Proper preheat of the entire joint per
AWS D1.1 requirements
Following an approved WPS
Careful sequencing of welds to allow
for contraction during cooling
Controlled cooling if warranted, for the
specific structural restraint and
environmental conditions
Use of fine grain killed steel for
members with through-thickness
stresses.
Specify through thickness mill testing
of the steel in accordance with ASTM
A770.
Fig. 12 Lamellar tearing crack developing under
a T-Weld. Microfissures initiate at flattened non-
metallic particles which grow to cracks by
interconnecting tears, note stepped ruptures; can
be picked up by UT.
21
Engineers Role in
Minimizing Weld Defects
In general, weld defects are most likely to occur
when using higher-strength carbon steels, when
joining thicker sections of steel over 1-1/2
inches, and when applying large amounts of
weld metal.
The higher carbon steels by their chemical nature
are less ductile, and therefore cannot as easily
accommodate the strain demand accompanying
weld shrinkage. Thus, low alloy steels are the
structural steels used today.
The tendency to produce weld defects when
thicker sections are being used arises from the
fact that for most steels to maintain a constant
minimum yield strength as the thickness is
increased, there is a corresponding increase in
the amount of carbon and manganese
requirements. Therefore, as a section gets
thicker the steel becomes less ductile.
Furthermore, thicker steels receive less working
of the hot metal during the rolling operations,
and therefore have a greater chance that
inclusions and discontinuities will be larger and
more prevalent, with less chance of being
forged-welded together during rolling.
The Structural Engineer is in the best position to
avoid situations that lead to weld defects by
designing joints that are not highly restrained.
Figures 13 through 18 show some typical joints
that are highly restrained, as compared with
alternative joint configurations designed to
minimize restraint.
Equally important in the design of weldments is
not to overweld. Often, if a joint is difficult to
analyze, the Engineer specifies that every
available edge is to be welded, and then the joint
is considered conservatively designed. Not only
is weld electrode material expensive, but also the
cost to deposit it is about 20 to 30 times the cost
of the base material. Therefore, the designer is
obligated to minimize the amount of welding.
Savings in weld metal also means less distortion,
less tendency for restraint, and consequently, less
cracking, less shrinkage, and less residual stress.
Do not specify CJ P welds when unnecessary.
In cases where the Engineer cannot avoid
designing a highly restrained joint, there are
compensating techniques available to the
fabricator of which he should be aware. Among
the techniques are the use of preheat, post-heat,
controlled cooling and the selection of more
ductile electrodes. Peening is helpful when
performed under knowledgeable and close
supervision. The sequence of welding is also
important. For most applications, welding
should begin at the center of the mass of the
weldment, where restraint is likely to be
concentrated, and proceed outward in block steps
with the electrode travel directed toward the
center of the mass. Wire shims (called softies)
may be used at critical points to provide the
necessary air gap within which shrinkage can
occur. Weld shrinkage and related strains are
directly related to weld volume and heating
cycles. Many tests have shown that a few larger
weld passes result in less shrinkage than many
stringer beads. Stringer beads should not be
specified if control of weld shrinkage, distortion
and minimizing residual stress are important.
AWS D1.1 has adequate controls on weld pass
size. Under restrained conditions adding more
heat to the weld area can increase the total
shrinkage strain.
The Engineer of record should not try to dictate
restraint compensating techniques or special
welding sequences or procedures to the steel
fabricator or erector. Rather, where warranted,
the Engineer should require a special submittal
by the fabricator or erector, outlining special
sequences or procedures for particularly
restrained elements of the structure. Most
experienced fabricators and erectors have a good
understanding of potential problems due to weld
shrinkage. Simply requiring the
fabricator/erector to create a written plan, and
stimulating thought on the issue often averts
potential problems. If the Engineer is not
comfortable performing the review of the
submittal, he or she may opt for review by a
welding or erection expert.
The Engineer should also call out for the
standard submittals by the fabricator of welding
procedures as per the most current AWS D1.1
and of shop detail drawings as per AISC
Specification for Structural Steel Buildings and
Code of Standard Practice.
22
Table D
Prequalified Minimum Preheat and Interpass Temperature
This table is intended as a convenient reference for selected steels based on Table 3.2 AWS D1.1:2002.
See the AWS Table for additional materials, cross references and footnotes.
Steel Specifications Welding Process Thickness Minimum Preheat &
(thickest part) Int. Temp. (deg F)
(inches)
1/8 to 3/4 incl. 32*

ASTM A 36 <=3/4 in. SMAW with other Over 3/4 thru 150
ASTM A 53 Grade B than low hydrogen 1-1/2 incl.
A ASTM A 139 Grade B electrodes
ASTM A 500 Grade A,B Over 1-1/2 thru 225
ASTM A 501 2-1/2 incl.
ASTM A 709 Grades 36 <=3/4 in.
Over 2- 1/2 300

ASTM A 36 ( >3/4 in.) 1/8 to 3/4 incl. 32
ASTM A 53 Grade B
ASTM A500 Grade A,B 0ver 3/4 thru 50
ASTM A501 SMAW with low- 1-1/2 incl.
ASTM A529 Grade 50,55 hydrogen electrode
B ASTM A 572 Grades 42, 50, 55 SAW, GMAW, Over 1- 1/2 150
ASTM A 588 FCAW thru 2-1/2 incl.
ASTM A 709 Grade 36 (>3/4in.)
Grade 50 Over 2-1/2 thru 225
ASTM A 913 Grade 50
ASTM A 992

1/8 to 3/4 incl. 50

SMAW with low- 0ver 3/4 thru 150
ASTM A 572 Grades 60,65 hydrogen electrode 1-1/2 incl.
C ASTM A 709 Grade 70W SAW, GMAW,
ASTM A 913 Grade 60,65 GTAW Over 1- 1/2 225
thru 2-1/2 incl.

Over 2-1/2 thru 300

SMAW,SAW All thicknesses
D ASTM A913 Grade 50,60,65 GMAW, and >=1/8 in. 32*
FCAW**

*When the base metal temperature is below 32 deg. F, the base metal shall be preheated to 70 deg. F
**With electrode/flux system capable of depositing weld metal with max diffusible hydrogen content
of 8ml/100g when tested according to AWS A4.3.
23
Fig. 13 Column Splices
Fig. 14 Beam Splices
Fig. 15 Beam-Column Details- (note also that bottom
continuity plates could be sloped at unequal beam
depths)
Fig. 16 Full Member Restraint
24
Correcting Weld Defects
Welding codes in general prohibit cracks of any type
in the completed weldments. When cracks are
detected, the AWS D1.1 requires that they be repaired
by removal and rewelding. The repair of such defects
is a normally encountered process during welding
operations, and AWS procedures are applicable for
repair by the fabricator and approval by the
knowledgeable Engineer. There are occasions,
however, when repeated attempts at repair are met
with repeated failure, and the Structural Engineer is
brought in for consultation either by the inspector in
support of his rejection or by the fabricator because he
may believe the design is contributing to the problem.
It is in this type of adversary situation that the
Engineer must prove his worth as a diplomat and
mediator as well as a good engineering technician.
It is important that the Structural Engineer retains his
composure and makes every effort to determine the
facts without letting the people problems outshine
the welding problems. Keep in mind that repeated
repairs are costly to the fabricator who wants to
produce an acceptable product, preferably without
flaws. Try to determine whether there is an actual
rejection of AWS D1.1:2002 requirements and if the
inspector is being fair and reasonable in his demands
or whether it is a case of punitive reprisal for past, real
or imagined, grievances.
It is wise to resist taking over and directing the
fabricator how to perform the repairs. However, the
designer can be helpful by asking for review of the
proposed repair procedure and by following a
formalized checklist to determine that all possible
sources for trouble have been considered. Among
some of the basic questions to consider are the
following:
1. Is the exact chemistry as well as mechanical
properties of the base metal known? A rough check of
carbon equivalent would be helpful in checking
weldability.
2. Do the electrodes and other joining materials
comply with AWS and ASTM standards? Ask for
certification or, if in doubt, have them tested and
check storage conditions.
3. Are the electrodes and base metal compatible as
called for under AWS matching standards (see
Table B)?
4. Has the extent of the crack or defect been
determined? Where is it located with respect to the
weld?
5. Do the welder, his supervisor and the inspector all
agree as to location and extent of the indicated defect?
Perhaps a third party may be helpful to settle
arguments of this nature.
6. Can the design be revised to minimize restraint?
7. Is the fabricator using more than minimum
required preheat to help slow the cooling down after
weld completion? Sometimes post heating and/or
insulation blankets will help prevent cracking.
8. Are the welder and inspector using a heat indicator
to determine preheat and interpass temperature?
Guessing is not sufficient.
9. Prior to beginning the replacement weld, was
magnetic particle testing (MT) used to make sure the
entire defect was removed?
10. The entire repair procedure for important
weldments should be written out and reviewed prior to
starting repairs and should include:
Size, type and AWS designation of electrode
material.
ASTM designation of base metal.
Sketch of defect showing size, extent, and
location in weldment.
Procedures followed for detection (NDT).
Preheat and interpass temperatures to be
used.
If necessary, post heat treatment or methods
to slow the cooling rate, such as asbestos
insulating blankets. Post heating is rarely
required and electric blankets are a major
issue.
Procedures for re-inspection after completion
of repairs.
Keep in mind that once the welding repair has started
it is mandatory to complete the repair without
interruption. Repeated heating and cooling invites
repeated cracking due to increased potential of
contamination.
If the fabricator has not already done so, it may be
helpful to suggest that a welding engineer or a
metallurgist knowledgeable in practical welding
problems be called in for consultation. Having faced
such problems many times previously, he may be able
to point out the technical cause of the problem
immediately.
Usual practice calls for the inspector to make a daily
inspection report and the fabricator is given a copy
with the original to the engineer. If the fabricator
disagrees with the inspectors report, the inspector
sends a non-conformance report (NCR) to the
Engineer, copy to the fabricator, for resolution of the
dispute.
25
Fig. 17 Internal Restraint in Weldments
Fig. 18 Welded Corner Joints
A fair, open-minded approach with a desire to work
cooperatively with the fabricator can have the best
chance of successfully correcting the problem and
keep it in its place, out of court.
Correcting Weld Distortion
Weld metal shrinks as it changes from liquid to solid
in contrast to water which expands as ice forms. The
solidification shrinkage is in addition to the thermal
contraction that occurs on cooling from approximately
2,800F. The volume change must be absorbed as
internal elastic and plastic strains, as movement in
some element, or as a rupture. A lack of fabrication
skill in coping with these movements is evidenced by
distortion of the finished structure or by cracking.
Fig. 19 Angular Distortion has resulted from weld
shrinkage. Compensating tilt of vertical member
and/or use of strong-back arc methods to control
vertical alignment to vertical position when welded.
26
Two thick plates fit up, tacked at 90 degrees and
welded together without fixtures can create angular
distortion, as shown in Figure 19. The accumulative
angular distortion of three weld passes is shown
schematically in Figure 20.
Fig. 20 Rotation in a butt weld. The rotation equals
the cumulative shrinkage from each weld layer.
Techniques have been developed which will minimize
this effort.
Distortion from transverse weld strain demand with
the welds shown balanced and the plate flat is shown
in Figure 21, below. The fabrication should position
the plates to account for the changes caused by the
first and second welds.
Fig. 21 Transverse shrinkage in a butt weld
Distortion from longitudinal weld strain demand is
shown in Figure 22. The position of the weld relative
to the center of gravity of the cross section produces
the bow. Supplementary weld beads are sometimes
used to achieve the desired camber. Sequence,
technique, and peening will minimize distortion.
Fig. 22 Longitudinal bowing in a welded beam may
produce either positive or negative camber (in X-Z
plane). Lateral box in X-Y plane) can occur.
A method of straightening or curving steel, known as
Heat Straightening or Flame Bending, uses
thermal upsetting to plastically deform the material.
Localized heating of steel causes thermal expansion
and a reduction of yield strength in the heated section,
while the material is still hot. The expansion is
inhibited by the cold, stronger, surrounding metal,
forcing the heated portion to yield plastically to accept
its own demand for increased volume. (See Figure 23
for yield strength at elevated temperatures and Figure
24 for variations in modulus of elasticity.) After
cooling, the shape of the steel piece is changed, and
the heated zone recovers its strength. Several heating
and cooling cycles may be required to complete an
operation.
Fig. 23 Variation in yield strength with temperature.
The flame bending technique is used in the shop to
flatten web plates, to camber beams, or to straighten
work distorted by welding. The maximum
temperature recommended for this operation is 1,200
degrees Fahrenheit for as rolled structural steels, and
27
1,100 degrees Fahrenheit for quenched and tempered
steels such as those of A709 but not higher than the
tempering temperature. Cooling may be in air, or by
water spray or wet rags for more rapid cooling.
Fig. 24 Variation in modulus of elasticity
Localized heating should not exceed the critical
temperature of 1,333 degrees Fahrenheit, or
undesirable changes in mechanical properties may
result. Heat-treated steels such as those of A709 must
not be heated above the tempering temperature, which
should be obtained from technical information
furnished by the steel manufacturer.
Nondestructive Testing (NDT)
One of the main reasons for the success of all welded
structures in the building industry has been the
development of fast and accurate methods of
examining welded joints without destroying or
impairing their actual usefulness. Currently, there are
a variety of techniques being employed by the
fabricator and independent inspection agencies to
assess the reliability of weldments. Used properly,
these methods can reveal practically all of the
common surface and internal defects that normally
occur with improper welding procedures and
practices, and will result in a quality level consistent
with Project Specification requirements. As in all
inspection methods, the experience and skill of the
technician and an inspection procedure developed by a
Quality Control Engineer are very important criteria
for reliable nondestructive testing. Users should
become familiar with all the limitations of NDT
methods.
The engineer may also request written documentation
as to the type of quality assurance program established
by the fabricator. Many fabricators have their own
quality assurance program meeting the nondestructive
testing specifications established by both local
building codes and the American Welding Society.
AISC has also developed a set of standards for quality
certification and has designated member firms
meeting these standards under the following
categories: Conventional Buildings, Complex
Buildings, Simple Bridges, Major Bridges. In
addition, there are several supplemental endorsements
to up-grade the basic categories.
The engineer who realizes the high degree of
protection afforded by these various organizations
through their time-tested standards and specifications
can do much to simplify his own design specifications
related to the welding of structural steel. Streamlined
specifications referring to accepted industry-wide
standards and avoiding unnecessary abstruse verbiage
would do more to assure that the specification will be
read and followed by the fabricator and contractor.
Visual Inspection (VI)
This is a requirement of the AWS D1.1-90 Structural
Welding Code wherein the duties are detailed. In-
progress visual (edge preparation, fit-up, root pass and
fill-in-passes) by a qualified and experienced inspector
is considered the most reliable method and most cost
effective. By far, an alert welder or inspector detects
most cracks in weldments visually. Sometimes
detection is made hours or days after completion of
the weld. This has been termed delayed cracking
when, in most cases, the cracks were probably there at
the completion of welding, but merely opened up wide
enough to see when the entire weldments cooled.
Proper visual inspection requires careful examination
in the areas outside the weldments, particularly along
plate edges and parallel to the weld where cracking
and lamellar tearing can occur. AWS D1.1 requires
NDT of high strength Q&T steels A514, A709 grades
100 and 100W to be performed 48 hours after
completion of welding. Delayed inspection and
delayed NDT is generally not necessary for the more
common structural steel.
Magnetic Particle Testing (MT)
This method is primarily for detecting surface cracks
or defects on or just below the surface of the metal.
MT is currently being used per FEMA 353
28
recommendations on some fillet welds and some
groove welds of seismic connections. MT is
particularly applicable to crack-sensitive material and
especially useful in detecting fatigue cracking. During
the test, a very strong magnetic field is applied to the
weld area, and the surface covered with a suspension
of ferro-magnetic particles. Defects such as cracks,
inclusions, etc., interrupt lines of force, causing the
particles to concentrate around these areas. Often, the
residual magnetic properties created by welding are
sufficient to allow the use of magnetic particles
without the application of a magnetic field. Because
this method is simple, easy to read, and the equipment
is portable, it is preferred for examination of welds
and adjacent areas for surface cracking caused by
weld shrinkage. The magnetic particle method is also
very useful during repairs to see if the defect has been
completely removed and to examine individual weld
passes and layers for hot cracking.
Fundamentally MT is an enhanced surface
examination and as such, the cost and schedule
impacts need to be balanced with the perceived
benefits.
Ultrasonic Testing (UT)
Since the development in the 1960s, ultrasonic testing
has grown to become the most important tool in
nondestructive testing of structural welded joints. In
this method, high-frequency sound waves are used to
locate and measure discontinuities in welded joints
and base metals prior to welding. This method is very
sensitive in detecting both surface and subsurface
discontinuities. During testing, a sound wave is
directed towards the weld joint and reflected back
from the discontinuity and shown on a calibrated
screen of an oscilloscope as shown in Figure 25 and
26. This method is highly sensitive in detecting planar
defects, such as incomplete weld fusion, delamination,
or cracks; however, orientation is very important. As
the wave strikes the defect, the time-distance
relationship will locate this interception. This is
shown on the oscilloscope and indicates the location
of the defect in the weld joint.
Fig 25. Ultrasonic testing of weld
Fig. 26 Typical CRT (cathode ray tube) display of
ultrasonically tested weld with indications.
This ultrasonic method can detect internal planar
defects in sections of practically unlimited thickness.
AWS D1.1, sets procedures for steel thickness from
5/16 to 8, but other thicknesses may be tested by
qualified procedures. It is portable and relatively fast.
Most importantly, it requires access to only one side
of a test section.
There are some limitations to ultrasonic testing.
Rough surfaces reduce its sensitivity and reliability.
Also, the method does not produce a permanent record
of the tested weld joint. In addition, globular defects,
such as gas bubbles and other porosities are not easily
detected. Because of the spherical nature of these
defects, ultrasonic waves tend to pass around them
rather than reflect back as with planar defects.
However, this deficiency of the ultrasonic methods is
not considered serious. Ultrasonic techniques, as
practiced, are normally limited to joints with plate
thicknesses above 5/16 inch, and are very sensitive to
flaw orientation and geometry. Most building codes
(and FEMA 353 recommendations) require ultrasonic
testing of complete joint penetration groove welds.
The ultrasonic method is highly dependent on the skill
and integrity of the operating technician for proper
interpretation of the results, and therein arises a major
weakness. An operator can quickly lose credibility if
he calls for a joint to be completely gouged out for a
defect that cannot be found. Consequently, it is easier
for the operator, unless technically competent, to say
nothing rather than risk being found wrong and then
subsequently challenged repeatedly by the fabricator
or contractor on the project.
On all special inspections calling for ultrasonic testing
of welded joints by an independent testing agency, it
29
is important for the Engineer to seek evidence as to
the qualifications of the ultrasonic technicians
involved. UT technicians are usually qualified in
accordance with the American Society for
Nondestructive Testing Recommended Practice SNT-
TC-1A.
In order to improve the reliability of UT in seismic
and other selected connections, future training and
qualification of UT operators will utilize some of the
techniques and provisions outlined in Annex K of
AWS D1.1. Annex K outlines a more rigorous
procedure for UT. It was originally developed by use
on off-shore structures that undergo extreme fatigue
loading. There is still some debate within the industry
as to the supply of qualified personnel capable of
using these procedures, and therefore, it is intended to
adopt the requirements gradually.
Radiographic Testing (RT)
Radiography relies on the use of electromagnetic
radiation to determine the soundness of a weld. X-
rays and Gamma rays are the two types of waves used
to penetrate solid materials such as a welded joint. A
permanent record of the weld structure is obtained by
placing a sensitized film at the back of the weldment.
As the rays pass through the weld material, they fall
on the sensitized film and produce a negative of
varying intensity. If the rays pass through gas bubbles,
slag inclusions or cracks, more rays will pass through
these less dense areas and will register on the film as
dark areas. Orientation of the discontinuity is very
important, especially for planar discontinuity.
Although radiography is a superior method of
detecting porosity defects and slag inclusions, for
most steel building construction, RT is impractical.
This is because the film must be placed opposite the
source of radiation to graphically record the defects,
and the actual geometry of completed joints,
particularly T-joints, generally prohibits proper
placement of the film. RT is not commonly applied to
structural fabrication because of the cost and schedule
impacts associated with the need to shield personnel
from radiation.
Liquid Dye Penetrant Testing (PT)
This method relies upon surface tension and capillary
action of certain dye-carrying liquids to penetrate
small surface flaws such as cracks. Subsequent
application of suitable developer brings out the dye
and outlines the defect. During the test, the surface
weld is cleaned and dried, then coated with a thin film
of the penetrant. After waiting a short time for the dye
to flow into the flaw, the surface is wiped clean and
the developer applied. The liquid penetrant will then
bleed out onto the surface to react with the developer
and sharply outline the flaw so it can be seen or
photographed. The use of dye penetrants in multi-
pass welding has been limited to investigative use
only because of the interruptions to welding process
and consequent cost to fabricator.
Project Specifications
It has been the writers experience in reviewing
hundreds of project specifications through the years
that there has been a needless waste of effort on the
part of the designers in writing and rewriting portions
of all of the standard AISC and AWS specifications.
Sometimes, needless litigation has ensued as a direct
result of rewriting nationally accepted standards to
include the personal bias of the designer, albeit arising
from previous bitter experience.
For the most part, fabricators, welders and inspectors
are well aware of the national standards and keep up
with them. The designers are well advised to do
likewise. They will get a better product with less
confusion and discord if they adopt them by reference
and omit any attempt to elaborate, clarify or otherwise
tamper with the nationally accepted standards unless
there is specific conflict with the project
specifications. There have been many large,
successful projects completed with a one line
specification limit that merely states, All materials
and workmanship shall be in accordance with the
latest revised edition of the AISC manual of Steel
Construction, which includes the specification for
Structural Steel Buildings, the code of Standard
Practice and the AWS Structural Welding Code.
However, for those who feel a project specification is
only sufficient when it has a few pages under each
section, a recommended list of items is included as a
checklist and reference in the following section,
Project Specifications Check List.
Normally, the nondestructive testing section of a
project specification is more detailed than other
sections devoted to structural steel. The AISC code of
Standard Practices requires that, When
nondestructive testing is required, the process,
locations, extent, technique and standards of
acceptance are clearly defined in the contract
documents. This is also in the AWS D1.1:2002
Structural Welding Code-Steel.
It is advised to keep in mind that inspection
requirements may vary between local, state and
federal building regulatory jurisdictions. Standard
inspection requirements should satisfy most
jurisdictions because normal practice requires
30
continuous inspection by a qualified inspection agency
paid for by the owner, for whose benefit the inspection
is being performed. However, the designer is
cautioned to determine for himself what differences, if
any, are required by the governing agency for each
project.
It is suggested that the structural steel designer obtains
a copy of the AISC publication Quality Criteria and
Inspection Standards and studies it carefully,
particularly as it relates to dimensional tolerances. If
the structure being designed requires closer
construction tolerances than allowed, either change
the design to accommodate them or put a large sign on
your drawings to the effect that care must be exercised
by the fabricator and erector to meet closer than
normal tolerances, and then spell them out so there is
no room for misunderstanding.
If the structure is tied into or otherwise supported on a
masonry or concrete structure built prior to erection of
the steelwork, dont expect the anchor bolts to be in
exact position. Make provisions in your design of the
connections for misalignment vertically and
horizontally of such anchors and/or make a field check
mandatory. A review of normal construction
tolerances for such construction will be enlightening,
to say the least.
If the building structure will not resist wind or
earthquake forces until materials other than structural
steel are in place, it is recommended practice to notify
the contractor with a note on the drawings or in the
specifications which clearly state that fact. Such a
requirement is contained in the AISC Code of
Practice; Section 7.10 entitled Temporary Support of
Structural Steel Frames.
FEMA 353 Quality
Recommendations
Although there were no collapses of steel structures,
serious concern grew out of the Northridge
Earthquakes moment frame beam flange weld
fractures (refer to FEMA 350 for background
information). In general, the typical projects steel
specifications have changed significantly since
research and recommendations were released by the
SAC J oint Venture, which was the most
comprehensive steel research program ever
undertaken. Out of this project, among other
documents, came FEMA 353, which recommends
various quality programs that could be applied to a
steel construction project. The document was
intended to be used as a guide that would be used
judiciously. Engineers should not broadly specify
FEMA 353 in job specifications. Rather, an
engineering firm should review the document and
decide which provisions are prudent and necessary for
their project. The cost and benefit of each
requirement should be evaluated. Engineering firms
with experience in steel structures have merely
updated their specification using FEMA 353 just as a
guide. Other engineers may not have the knowledge
or the expertise to do this on their own. Fortunately,
the next issue of the AISC Seismic Provisions will be
a consensus document which contains quality
guidelines for steel construction. These will outline
performance and documentation of everything from
visual inspection to non destructive testing, submittals
of contractor quality control programs, and owner
quality assurance requirements for seismic resisting
systems.
Project Specifications Check List
The following list of items is recommended as a
reference and checklist to help develop the project
specifications:
Scope of work
References to National Standards (AISC,
AWS, AST, UBC, RCSC, SSPC, etc.)
Shop Detail Drawings submittals
Welding Procedure Specifications submittals
Materials (List ASTM Specifications,
Structural Steel, Pipe, Structural Tubing,
High-Strength Bolts, Std. Bolts, Nuts and
Washers)
Welding Processes: Shop and Field:
Prequalified and qualified-by-test.
Filler Metal Specifications and
Classifications
Quality Control and Quality Assurance
Fabrication
Erection
Galvanizing, Painting: Shop and Field
Inspection: Shop and Field, including
verification of welder certifications
Specific requirements for the Seismic Force
Resisting Frame.
Welding Procedure
Specifications (WPSs)
Confusion may exist among some structural engineers
and fabricators regarding written welding procedures.
The AWS D1.1, clearly states in section 3.6 that All
pre-qualified WPSs to be used shall be prepared by
the manufacturer, fabricator or Contractor as written
pre-qualified WPSs, and shall be available to those
authorized to use or examine them. AWS D1.1
requires any WPS qualified by test to be approved by
31
the Engineer of record. In spite of this mandatory
requirement, some Engineers do not require their
submittal. Some fabricators always submit them for
review while others neglect to do so. Some engineers
review the WPS submittal in house, while others hire
a consulting engineer that is familiar with the welding
issues of steel fabrication and erection.
The Engineer reviewer can compare the submittals
with the sample forms shown in Annex E and
checklists in Annex H. These Annexes cover the
mandatory code requirements of a (WPS). Where
welds are not pre-qualified, they must be qualified by
test. For these welds that are qualified by test, a
Procedure Qualification Record (PQR) is required to
accompany the WPS. WPSs should be submitted with
the electrode manufacturers product data sheets that
outline the recommended parameters for voltage,
amperage, electrical stick-out, polarity, and other
pertinent variables. WPSs should show that the
procedure falls within the manufacturers
recommended parameters.
Shop Detail Drawings
Shop drawings have been the subject of much debate
for many years, and yet there still remains an absence
of a uniform understanding within the design
professions, legal professions, and the construction
industry. There has been a wide variation in the
manner in which shop drawings have been used,
leading to a great deal of confusion.
Shop drawings are necessary to facilitate steel
fabrication, erection and installation of various
elements of the work. Their very nature is such that
they are required to comply with the contract
documents. Review of shop drawings is simply to
confirm compliance and to facilitate progress of the
work. It is the position of the author of this paper that
much of the confusion that exists has come about
because of the use of shop drawings as design
documents. The misconception that shop drawings
are part of the design process must be eliminated.
Shop drawings are not part of the contract documents
and must not be used as such. If changes are proposed
or made by the contractor or engineer, they must be
done through a change order process. When shop
drawings are used as an instrument of change, they
can only lead to confusion. Clearly, changes proposed
by the contractor, or the submittal of details or
systems based upon performance-type specifications,
must be reviewed and approved by the engineer of
record. This process must take place through
documents other than shop drawings, such as change
order or supplemental design details.
Electronic Data Interchange
(EDI)
The steel industry has been working to produce a set
of standards for sharing electronic data. CIMsteel, or
CIS/2 Integrated Standard, has been incorporated into
many design, detailing, and fabrication packages. The
standard allows a single electronic model to be carried
through the entire project, from design, to material
orders, detailing, fabrication, and erection drawings.
Engineers may have concern over liability of releasing
their models, but with the appropriate agreements and
understandings with contractors on the accuracy of
models, and the methods of handling revisions, all
parties can benefit. The industry has yet to scratch the
surface of the capabilities and future application of
EDI. Numerous articles on EDI may be found at
AISCs website, www.aisc.org, upon searching for
EDI.
Conclusions
Steel has been and will continue to be a very versatile
structural material. From the manufacture of steel
plate and shapes, through structural design, detailing,
fabrication and erection, the steel industry continues to
evolve along with technology to meet the demands of
modern structures. It is the privilege and the
responsibility of engineers to combine the latest
technology with the state of the art and the state of the
practice, along with lessons learned from earthquakes
and research, to produce a sound and practical design.
Design engineers and plan check engineers can benefit
from a reminder that no material or process is perfect,
and that it is the engineers job to manage those
imperfections. Quoting the first paragraph from
Salmon and J ohnsons STEEL STRUCTURES
Design and Behavior, 2
nd
edition: Structural design
may be defined as a mixture of art and science,
combining the experienced engineers intuitive feeling
for the behavior of a structure with a sound
knowledge of the principals of statics, dynamics,
mechanics of materials, and structural analysis, to
produce a safe economical structure which will serve
its intended purpose.Computations involving
scientific principles should serve as a guide to
decision making and not be followed blindly. The art
or intuitive ability of the experienced engineer is
utilized to make the decisions, guided by the
computational results.
REVISED MARCH1993
Structural Steel Construction
in the '90s
by
F. Robert Preece
and
Alvaro L. Coilin
Table of Contents
ACKNOWLEDGEMENTS ............................................................................................................................... i
PREFACE ....................................................................................................................................................... 1
INTRODUCTION ............................................................................................................................................ 1
EARLY BUILDING DESIGN ........................................................................................................................... 1
Riveting, A Lost Art .................................................................................................................................... 2
High Strength Bolting ................................................................................................................................. 2
TODAY'S STRUCTURAL STEELS ................................................................................................................ 2
Special Materials Considerations ............................................................................................................... 8
ASTM A-36 (as rolled) ............................................................................................................................... 8
ASTM A-242 (as rolled) ............................................................................................................................. 8
ASTM A-441 (as rolled) ............................................................................................................................. 8
ASTM A-572 (as rolled) ............................................................................................................................. 8
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
A-588 (as rolled) ............................................................................................................................. 10
A-514 (quenched and tempered) .................................................................................................... 10
A-709 (as rolled and Q&T for bridges) ............................................................................................ 10
A-53 (as rolled) ............................................................................................................................... 10
A-500 (cold formed structural tubing) .............................................................................................. 11
A-501 (hot-formed structural tubing) ............................................................................................... 11
ASTM A-139 (welded steel pipe) ............................................................................................................... 11
ALLOYING - KEY TO STRENGTH AND WELDABILITY ............................................................................... 11
CARBON EQUIVALENT ................................................................................................................................. 11
QUENCHING AND TEMPERING FOR HIGH STRENGTH ............................................................................ 12
THE WELDING PROCESSES ........................................................................................................................ 12
Shielded Metal Arc Welding (SMAW) ........................................................................................................ 12
Flux-Cored ARC Welding (FCAW) ............................................................................................................. 12
Gas Metal ARC Welding (GMAW) ............................................................................................................. 13
Submerged ARC Welding (SAW) .............................................................................................................. 13
Electroslag Welding (ESW) ........................................................................................................................ 13
AVOIDING WELD DEFECTS ......................................................................................................................... 14
Weld Crocking ............................................................................................................................................ 14
Hydrogen Embrittlement ............................................................................................................................ 14
Lamellar Tearing ........................................................................................................................................ 16
Engineers' Role in Minimizing Weld Defects ............................................................................................. 18
CORRECTING WELD DEFECTS ................................................................................................................... 21
CORRECTING WELD DISTORTION ............................................................................................................. 22
Flame Bending and Straightening .............................................................................................................. 22
NONDESTRUCTIVE TESTING (ND'F) ........................................................................................................... 23
Visual Inspection (VI) ................................................................................................................................. 24
Magnetic Particle Testing (MT) .................................................................................................................. 24
Ultrasonic Testing (UT) .............................................................................................................................. 24
Radiographic Inspection (RI) ..................................................................................................................... 25
Liquid Dye Penetrant Testing (PT) ............................................................................................................. 25
PROJECT SPECIFICATIONS ........................................................................................................................ 25
Project Specifications Check List ............................................................................................................... 26
Welding Procedure Specifications ............................................................................................................. 26
Shop Detail Drawings ................................................................................................................................ 26
THE AUTHORS .............................................................................................................................................. 27
Tables
TABLE A ASTM SPECIFICATIONS - STRUCTURAL STEELS ........................................................... 9
TABLE B FI LLERMETAL REQUIREMENTS ........................................................................................ 15
TABLE C MINIMUM PREHEAT AND INTERPASS TEMP OF ............................................................... 17
Charts
HISTORICAL BACKGROUND OF STRUCTURAL STEELS ................................................. 7
CARBON EQUIVALENT ......................................................................................................... 11
CHART 1
CHART 2
ii
Structural Steel
Education Council
Structural Steel
Construction in the '90s
Preface
The first publication on this subject for the Struc-
tural Steel Education Council, "Structural Steel in the
'80s - Materials, Fastening and Testing," by F. Robert
Preece and edited by Robert L. Ray and A. L. Coffin,
was a very popular paper among structural designers,
fabricators, erectors and inspectors.
Rather than re-issue this paper, it was determined
by the Structural Steel Education Council that it be re-
written so as to include materials information changes
in the 1989 AISC "Specification for Structural Steel
Buildings (ASD)" and the 1990 AWS D1.1 "Structural
Welding Code - Steel."
While the basic purpose is to review and comment
on the structural steeis in building construction, it is
helpful to include comments on related items such as
welding processes and procedures, inspection and
connection details.
It must be recognized that each of these related
items could be, by itself, the subject of an in-depth
study and write-up.
The information given in this publication is a quick
reference guideline; refer to AISC, AWS D1.1 and
ASTM Codes & Specifications for specific data.
F. Robert Preece
Ak, am L. Coffin
Introduction
Today's engineer has available a wide variety of
structural shapes, grades of materials and joining
techniques with which to design steel-frame buildings
and similar structures. This proliferation of choices
has come about through rapid technological ad-
vances in steelmaking and related industries over the
past two decades.
The trend away from rivets and bolts to all-welded
joints has produced an integrally stronger, fully elastic
structure that has made possible the use of the com-
puter in steel frame design and increased its effi-
ciency.
The development of today's high-strength, Iow-
alloy steels has also achieved major material and en-
ergy savings in the steel fabricating process, while
making possible a variety of structural shapes avail-
able to the designer. Care should be given to specify-
ing high strength steels, especially with regard to de-
flection and vibration, as shall be explained later
herein.
This paper will review the trends in structural steel
design, from the massive all-riveted and bolted struc-
tures of the '40s and '50s to today's simpler, lighter
and more cost-efficient welded designs.
As an aid to the architect and engineer, this paper
will also describe the most commonly available con-
struction steels, their features and important applica-
tions. Of particular importance will be the special con-
siderations for "matching''l electrodes to base metals,
the influence design has on joint performance, and
the quality control procedures used to assure struc-
turally sound joints.
This paper is written also for the young structural
engineer to provide an historical perspective of the
structural steel building industry and to relate the im-
portant developments in the art of steel welding and
the methods of minimizing weld defects.
Early Building Design
The use of steel in building construction had its
roots at the end of the last century with the construc-
tion of such towering edifices as the 14-story Tacoma
Building in Chicago and New York's Tower Building,
rising to a dizzying height of 129 feet.
As early designers recognized the greater strength
and ductility of mild carbon steel over cast and
wrought iron, office buildings grew taller and taller.
However, it remained until the late 1920s before high-
rise building construction became truly economical.
Thi s was due largely to technological advances made
in the rolling process that permitted the forming of
large structural shapes, coincident with the develop-
ment of high-speed elevators.
Until then, steel mills rolled so-called "standard"
shapes: angles, channels, I-beams and plates, which
were then riveted together to make the larger col-
umns and girders required for high-rise construction.
Built-up shapes such as these are inordinately expen-
sive by today's standards due to the tremendous cost
of detailing, drilling, riveting and assembling. Figures
1, 2, 3 and 4 (pages 3-6) illustrate the gradual transi-
tion from all-riveted standard channels, angles and
plates (Figure 1) to today's all-welded variety (Figure
4). Of course, there is no direct method available to
compare their relative costs in today's dollars, but the
material savings is directly evident; see Chart 1 (page
7), Historical Background of Structural Steels.
Over the years, some loosely defined terms have
been used to describe carbon steels. Generally, the
t' Matchi ng' aa UKI here le the ad)itrary mslgnmeat gl etala electrode materlela
to beammelab by AW$ oode. It b undm'dood that while they perform satlacioly
l a aL,ucfixes, they mayno( be truly ' matched' In metallurgical mam ' match- 1
I ultk&l ten,,le Idreltl'lll rathe' than ylatd Idreroth$ is oatroveralal.
agreed terms used in reference manuals are as fol-
lows:
Low Carbon Steel C <_.20%
Medium Carbon Steel 0.20 __.C ; .050%
High Carbon Steel C 2.50%
The built-up shapes used in early building con-
struction were necessary due to the inherent nature
of the rolling process. At the time, steel mills used a
single set of rolls for each structural shape. The
single-roll process prevented increasing the thickness
of a structural shape in one direction (e.g., the flange)
without correspondingly increasing its thickness in an-
other (web). This represented a very inefficient way of
producing large shapes.
By the late 1920s, however, the (Gray) wide flange
mill had arrived on the scene. These mills used sev-
eral sets of rolls arranged in tandem. Operators could
now roil relatively wide-flange shapes to 36 inches
deep, while keeping the web sections relatively thin.
These wide flange shapes revolutionized the steel
building industry and made possible the construction
of the Empire State Building and the 30 or so other
skyscrapers that dotted New York's skyline in the late
1930s.
Meanwhile, during this same period, welding was
being advocated by some farsighted designers and
fabricators who recognized the advantages of using
less materials with a method that would gain 100 per-
cent elastic continuity. This, coupled with the develop-
ment of the Structural Welding Cede by AWS, led to
the demise of riveting and the development of the all-
welded joint.
Riveting, A Lost Art
The earliest methods of joining steel were by rivets
and bolts. Rivets were used before 1850 in boiler
work and shipbuilding. Mild steel bolts became popu-
lar around the same period for joining heavy machine
ports.
During the early building industry, before the devel-
opment of today's high-strength bolts, rivets were
considered stronger than bolts because they filled the
holes and prevented any slippage of the joint. Conse-
quently, they enjoyed almost exclusive use in the first
steel-framed buildings. The exception generally oc-
curred in the connection of beams to girders, where
the loads were relatively small and the set-up time to
drive a few rivets per connection was too expensive.
High Strength Bolting
By 1950, high-strength bolts were being given
strong consideration as a replacement for rivets in
high-rise buildings. A development of the railroad in-
dustry and used in heavy machinery, these bolts have
a tensile strength on the order of 120 ksi, approxi-
mately twice their mild steel counterparts. The impor-
tance of the high-strength bolt to structural applica-
tions is that it can be torqued up so tightly that it pre-
stresses the joint with its tremendous clamping force.
Load could now be transferred from beam to column
or from beam to beam by the friction of the two mat-
ing surfaces rather than the shear strength of the rivet
or bolt. This produced a stiffer, more rigid joint with
less movement of the joint, hence less distortion of
the frame.
Today, high-strength bolts are still commonly used
for fit up prior to welding, for fastening secondary
members or for all-bolted structures in remote areas
where conditions may not warrant the use of field
welding. For high-rise construction, however, the all-
welded joint has supplanted the use of bolts except
for beam web connections. There are several rea-
sons: first, the variety of uniformly weldable steels
available to the designer; second, an increasing un-
derstanding of the metallurgical process of welding;
third, the number of welding processes available;
fourth, the development industry-wide of comprehen-
sive quality assurance programs, including the use of
nondestructive testing (NDT) methods, such as ultra-
sonics and magnetic particle inspection; and fifth, the
continuing development of the AWS Structural Weld-
ing Code.
Today's Structural Steels
At the turn of the century, there were only two
grades of steel being produced: A-7 for bridges and
A-9 for buildings (See Chart 1, page 7). These mated-
als were so similar in chemical composition and
mechanical properties that in 1939 they were consoli-
dated by ASTM into one standard, A-7, which was
used until 1960 for both types of applications. In
effect, for 50 years there was basically one grade of
steel being used in the building industry - A-7 - until
ASTM-A36 appeared in 1960. When a higher
strength steel was needed, primarily for bridges,
silicon steel was specified, basically for riveted and
bolted construction, because it was very difficult to
weld.
By the 1950s, the strength and economics of weld-
ing were becoming apparent. Despite earlier bridge
failures in the railroad industry, which were more
related to poor workmanship and technique than to
materials, welding was being tried in some of the high-
rise structures being built on the Gulf Coast. Here,
structural engineers applied the experience gained in
the welding of refineries and oil pipelines to high-rise
buildings. Designers soon noticed that while most of
the A-7 steel being used showed good welding charac-
teristics, there were instances in which some A-7 steel
presented welding problems because of the limited
-z.--4',,,4',=e,,'""3 P,]VET5 s.s.
- - - - ' - I .......... ', . . . . . . .h--;==z==.= :=v='i="- =-;-----;---; 1.,,_ J..-
T I / [4' ' / . - x 3 x 8 %-/ 4 x / I . /-'ir
" " L , 5 " ' "
, 1 / I l Fz7. ''-''' '
I
'i 171- I
F I G . : ] . P L , N
: i i
- II I I!ItlI I I Ii!
N !ii j
. . . . . .
l i l t i i !i
ijll I I Ii!
j, F I
,,, J--l-
ii / I i
.11 t
l i
Jii _
HAND HOLE
?7 . . . . . . . . . . . . . . : -
,- 44,' -
I T l ] R I V E T , i P.S.
? I
,)
u = " r -1-"
,-- RIVET5
TO COLUMN
ALL RIVETS ARE ' 4 _ _
, o _ % , 4 , '=-
F:IlYET ..
j G U s S F - T I-"34
. S E C T I O N B - I?,
S E C T I O N A - A
ESTIMATE OF I,qEICHT,.5 - LISS.
COL.L.IMN ,HAFT 2" _6
ll:)F__ .4-0O0 4 I
I<:)O I
DETAIL 1410 16
TOTAL =17C,0 IOO
EST. COST FACTOR = 2.00
Fig. 1 Pre- 1920s era - all riveted standard channels, angles andplates.
3
A
----h-7/a" . .
J ),Obq-, 'T"rP
FIG. 2 I::::'L.,A..N
,t,.-
: itl
F-'
-- L4 4- .. '/- 0'- 8"
,' WI8 55
' t
?
S E / N - T
I 111 I I -"L,'4'"'"''-8 ''
I 11ll4,1--l--t----P L L B & I
I I I t ',,v".o'-"
S E C T I O N A-A
'I:LP l BObT ? [ - q t - - - - - - - l - ..--q'C. IM PLI',NG Th'It,
'-LG F L o'L'r--IT- - - JJ / y_WT,8 ,,. 7 ,,. _4."
? ,,.'.4 ,,-,.
. , -
S I C : : T I O N B - Eb
EST. IM/',,q'E OF W'IIrHT,.5- LBS. %
COLUMN -sH,d=T 3 I., 3,7
GI DE I 900 JO
BI,AM I 1 18
DT,,I L. 91& 15
i o i / NL f ,x:) o o
F:$T COST F,.,cTo = 1.50
F i g . 2 1930s - riveted and bolted rolled sections.
4
i
=
/k
i _ I
_LU J
FIG,. 3 PL/N
i
f ' I ' l H J ' t ' '
1
, i & I H I , x L ,
SECTION A-A
T'm T , Fb
VEK' IE :sNBAff.. It
SECTION B - B
ESTIM,,TEE OF WEIG,HT - LBS. '/,
COLUMN SHAFT 2200 39
GIRDER I 700 aO
BEAM Iooo 18
DETAIL 70 13
TOTAL 5 5o ioo
ET COST FACTOR. = 1.25
F i g . 3 1950-1960-shop welded-field
5
TYP.>-
B
COLUMN
NI4x17
Ix)LTS, T'FP,
TIFE
x ' I'-o$_'
ESTIMATE O F IAIEIqHTS- LBS.
COLUMN SHAFT 211_ 4'2-
QIRDEF I?ex') 94
6EAM Iooo 2O
DETAIL 15,'5 '4-
TOTAL 5000 IO0
EST. COST FACTOR = 1.00
FIG. 4 PLAN
18
(T6 FL(5. T(I FLG.
I 'r-
7
<50 X r-. <
[:C' ';,
x /
/
IP,
IAI24-x C
SECTION A-A SECTION B-B
Fig. 4 1980s - all weldedconstruction.
6
Chart 1
Historical Background of Structural Steels
ASTM Requirements
Tensile Yield Point
D a t e s Specification Definition Strength b i ksi Minimum
Rev. Jan., 1991
Elongation
& Chemistry
1900- ASTM-A7 Soft to medium 52-70
1900 Bridges steels
ASTM-A9 Medium steel 60-70
Buildings
NOTE - Basic Unit stresses recommended by mfg: bldgs.-16.0 ksi, bridges, 12.5 ksi
32-35
35
No req. but
usually listed
No req. but
usually listed
1905- ASTM-A7 Structural 60
1913 Bridges steel desired
ASTM-A9 Medium & 55-65
Buildings structural steel
No req. but
reported
1/2 x T.S.
No req. but
usually reported
No req. but
usually reported
1914- ASTM-A7 Structural 55-65
1934 Bridges steel
ASTM-A9 Structural
Buildings steel
NOTE - AISC 1923 Basic Unit Stress .. 18.0 ksi
55-65
x T.S.
1/2 x T.S.
30 min.
No req. but
usually reported
No req. but
usually reported
1934- ASTM-A7 Structural
1938 Bridges steel
ASTM-A9 Structural
Buildings steel
NOTE - AISC 1936 Basic Unit Stress = 20.0 ksi
60-72
60-72
1/2 x T.S.
33 min.
x T.S.
33 min.
Usually reported
1939- ASTM-A7 Structural
1949 Buildings & Bridges steel
NOTE - ASTM-A7 & A9 consolidated into one spec (ASTM A7)
60-72 1/2 x T.S.
33 min.
1954 ASTM-A373 Structural
steel
NOTE - Revised 1958; phased out when A-36 was issued (1960)
58-72 32 El. = 24% (8' Ga.)
1960 ASTM-A36 Structural 60-80
Buildings & Bridges steel
NOTE - Issued 1960, revised 1961 (caJled out as weldable)
C., .28, Mn = .80-1.20, P = .00, S = .05, Si = .15-.30, Cu = .20 if specified
1961-A7: Tensile ksi - 60-75, Fy = 33 ksi, Elong (8') = 21.0%
C = Not specified, Mn = N.S., Phos =, .04, S = .05, Cu = 20 min. when spec.
36 ksi
min.
20% (8' Ga.)
23% (2' Ga.)
1988 ASTM-A36 Structural 58-80 36 ksi
Buildings & Bridges steel min.
NOTE - Structural steel for riveted, bolted or welded construction
Nominal chemistry % (Refer to ASTM specs for detail)
C =, 0.26-0.29, Mn = .80-1.20, Phos = 0.04, S = 0.05, Si = 0.15-0.40, Cu (when specified) ,, 0.20 min.
20% (8' Ga.)
23% (2' Ga.)
- Based on data from AISC: Iron & Steel Beams - 1873 to 1952; pest issues of ASTM Specs, AWS Codes (including
First Edition - 1928), AISC Specs and other publications on steel.
GENERAL TERMS:
Low Carbon Steel: C =; 0.20%
Medium Carbon Steel: 0.20 ; C < 0.50%
High Carbon Steel: C > 0.50%
Reference:
ASM Handbook,
USS - Steel Making
AISC Publications
7
chemistry requirements. Thus began extensive re-
search into the metallurgy of steel to develop structural
materials that were uniformly weldable.
By 1964, the American Institute of Steel Construc-
tion (AISC) had adopted five grades of steel for struc-
tural application. Table A (page 9) shows the chemical
composition and some of the mechanical properties of
these high-strength, Iow-alloy steels. The high elonga-
tion property of today's steel (up to 25 percent) permits
large overstress due to welding and deformation dur-
ing construction without losing its ultimate strength.
Figure 5 (below) compares typical tensile stress-strain
curves for these steels.
e3
0.2% 4 s e l
/ r 115,ooo . i .
'x
80, 000
../' / ..-- -. , A441iA541,AS' /
--- . . .
'F7
0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
STRAIN - INCHES PER INCH
Fig. 5 Engineering stress-strain tensile curves for ASTM
structural steel grades A-514, A-572, GR. 60, A-441, A-588
and A-572, GR. 50 and A-36.
Today, the ASTM Specifications for Structural
Steels cover many carbon steels, high-strength, Iow-
alloy steels, and some quenched and tempered con-
struction alloys. Structural steels include plates, bars,
shapes, pipes and structural tubing. And there are
many processes for welding these steels, four pre-
qualified and others that are qualified-by-test.
Actually, there is no reason for problems in welding
structural steels if:
* A good welding procedure is written and followed.
A technique based on good welding practice is used.
Good workmanship by quar[r welders is performed.
Quality control through use of qualified inspectors is
specified and implemented.
Special Materials Considerations
The AISC-ASD Ninth Edition (1989), Chapter A,
Paragraph A3.1.C. (Heavy Shapes) calls for ASTM
A6, Groups 4 & 5, Rolled Shapes (also known as
Jumbo Sections) to have Charpy V-Notch impact test
minimum values of 20 ft-lbs at 70F. This requirement
applies when these shapes are primarily subject to
tensile stresses due to tension or flexure and are
spliced using complete joint penetration groove
welds. The same requirement applies to built-up
members with plate 2' or greater in thickness.
ASTM A-36 (as rolled)
A-36 represents the majority of tonnage used for
general structural application. Because of its strength
and economy and availability, it is the steel of choice
and should be specified whenever possible, espe-
cially where deflection or vibration govern. When
strength requirements exceed the normal strength
range of A-36 steel, other higher strength steels may
be specified within the limits of availability and cost.
Steel mills often require a minimum order of five tons
per section size for rolling.
A high-strength, Iow-alloy steel with enhanced at-
mospheric corrosion resistance of approximately two
times that of carbon structural steel with copper, the
same as A588 (or four times carbon structural steel
without copper). It has three grades of yield strengths
- 42, 46 and 50 ksi, depending on thickness, and is
weldable with proper welding procedures, but is lim-
ited to material up to 4' in thickness.
A high-strength, Iow-alloy (manganese vanadium)
steel with yield strength limits from 40,000-50,000 psi,
depending on the product and thickness. In addition
to the strengthening by carbon and manganeses,
A-441 benefits by a small vanadium addition (.02%
minimum) and by copper (.20% minimum).
Element Function
Carbon & Manganese Basic Strengthening
Vanadium Toughness & Strengthening
Copper Corrosion Resistance
Vanadium, a powerful deoxidizer, acts to increase
toughness and enhance carbide stabilization plus im-
prove ferrite strengthening. Vanadium also is effective
in assisting grain size control. The .20% copper mini-
mum addition to A-441 provides improved atmo-
spheric corrosion resistance over structural carbon
steels, such as A-36. The restrictions on carbon
(.22% maximum) and manganese (1.25% maximum
for plates) help to provide improved weldabillty for this
grade over A-36.
A-572 is a carbon manganese steel, augmented
for strengthening by columbium, vanadium and nitro-
gen (optional) additions.
!
- I
.
i -
, .
r u g g --. .
d g
+* >X
0
it)
O"
8
.$
0
z
9
Element Function
Carbon & Manganese Basic Strengthening
Columbium, Vanadium, Carbide& Carbo-nitride
Nitrogen Stabilization
A-572 covers four grades ranging from 42 to 65 ksi
minimum yield strengths, depending on plate thick-
ness and product size and is available in four types.
Unlike A-441, which varies in minimum yield strength
limits according to section thickness, A-572 has a
constant minimum yield strength within any one
grade. For example, A-572 grade 42 has 42 ksi mini-
mum yield strength for all plate thickness to 6
whereas A-441 varies from 40 ksi to 50 ksi minimum
yield strength for the range of plate thickness to 6'.
This indicates that the ferrite strengthening and car-
bide stabilization effects of A-572 are less sensitive to
section thickness than the leaner alloying of A-441.
Increases of minimum yield strengths for A-572
grades to 65 ksi are accomplished by increasing max-
imum carbon content from .21% (grade 42) to .26%
(grade 65), plus other chemistry adjustments within
the specifications. Limits for carbon and manganese,
plus grain size control, provide good weldability for
this grade.
ASTM A-588 (as rol l ed)
A-588 covers multiple Iow-alloy grades (basically
manufacturers' "recipes"), which contain a variety of
elements in various combinations of the following:
Element
Carbon, Manganese
Molybdenum,
Chromium, Nickel
Titanium, Vanadium
Copper
Function
Basic Strengthening
Ferrite Strengthening
Corrosion Resistance
Carbide Stabilizers &
Toughness
Corrosion Resistance
This steel, with yield strength levels of 42,000-
50,000 psi minimum, depending on plate thickness, is
well known for its weathering characteristics. Under
proper atmospheric conditions, formation of a protec-
th/e, tightly adhering oxide surface is developed with
time, with atmospheric corrosion resistance twice car-
bon structural steel with copper.
ASTM A-514 (quenched and t empered)
A-514 covers a number of Iow-alloy grades, (basi-
cally manufacturers' Urecipes"), with a variety of alloy-
ing elements. Since A-514 is a quenched and tem-
pered steel with 90,000-100,000 psi minimum yield
strength, depending on plate thickness, the following
elements are balanced to provide hardenability, tem-
pering and notch toughness controls:
10
Element
Carbon, Manganese
Chromium, Molybdenum,
Nickel, Boron
Nickel
Function
Basic Hardenability,
Strengtheners
Hardenability Agents,
Tempering Control
Notch Toughness Control
Tempering at the mill is performed at rather high
temperatures (1,200-1,250F) for steels included in
this grade. However, any temperatures in excess of
the tempering temperature will reduce the strength.
Therefore, exposures to heating such as in welding
cycles must be controlled in order not to soften the
hardened structure, thereby lowering the strength.
Similarly, overheating may transform the steel struc-
ture and result in a structure that is too hard and
brittle. Controlled welding techniques can be ex-
pected to produce consistently good results for this
steel.
Although A-514 is quite suitable as a structural ma-
terial, since it is not available in wide-flange or hot-
rolled shapes, it is not commonly used in building
construction, but is intended for welded bridges.
Aside from cost considerations, the resulting light-
weight structures may produce higher deflections in
long span members with consequent undesirable vi-
bration characteristics. There are also special precau-
tions that must be taken during welding in order to
prevent destroying the special heat-induced proper-
ties of this steel.
ASTM A709 (as r ol l ed and Q&T for bri dges)
These structural steel specifications cover carbon
and high-strength, Iow-alloy and quenched and tem-
pered steels available in six grades with four yield
strengths - 36, 50, 70 and 100 ksi - and are plates
and shapes for use in bridges. These steels are basi-
cally A36, A572, A588 or A514, but with minimum im-
pact test requirements for non-fracture critical and
fracture critical applications in Zones 1, 2 & 3 as set
by the American Association of State Highway and
Transportation Officials.
ASTM A53 (as rol l ed)
This specification covers seamless and welded
black and hot-dipped galvanized carbon steel pipe in
nominal pipe size (NPS) 1/8' to 26' diam. with nomi-
nal wall thickness from .068' to 2.344' depending on
diameter. Pipe weight class is Standard (STD), Extra
Strong (XS) and Double Extra Strong (XXS). This
pipe is very popular for structural use as its chemistry
and mechanical properties for Types E and S, Grade
B, are very similar to ASTM A36.
ASTM A500 (cold-formed structural tubing)
Thi s specification covers cold-formed welded and
square and rectangular shapes. It is available up to a
maximum periphery of 64 inches (20-3/8' diam. and
16'x 16' square) with a maximum wall thickness of
0.625', and yield strength grades A, B & C of from 33
to 50 ksi, depending on shape and grade. Its yield
strengths and chemistry make it compatible with A36
and HS-LA steels. Normally stocked in local steel ser-
vi ce centers.
ASTM A501 (hot-formed structural tubing)
Thi s specification covers hot-formed welded and
square and rectangular shapes. Round tubing is fur-
nished in nominal pipe-size (NPS) 1/2'to 24' di am-
eters, in wall thicknesses of.109' to 1.000' depend-
ing on size. Square and rectangular tubing is avail-
able in sizes 1 ' to 10' across the flat sides, in wall
thicknesses of .095'to 1.000' and may be furnished
with hot-clipped galvanized coating. It has one yield
strength, 36 ksi, and chemistry comparable to A36.
Check for availability.
ASTM A139 (welded steel pipe)
Thi s specification covers arc welded steel pipe (pri-
marily for use in transmission of fluids) in five grades
(A, B, C, D & E with respective yield strengths of 30,
35, 42, 46 & 52 ksi) and is available in nominal pipe-
sizes of 4 'to 92' diameters wi th wall thickness of
1/16'to 1.0', depending on size.
Thi s specification is practically identical to API-SL -
X42 & X52, is not listed by AISC nor AWS D1.1-90,
but can be qualified by test for applications that ex-
ceed normal sizes and strengths of standard pipe and
structural tubing.
Alloying - - Key to
Strength and Weldability
All structural steels derive their strength character-
istics from the addition of various alloys, especially
carbon, and in the case of A-514, by heat treatment,
in addition to alloying.
Carbon, in the form of iron carbide or in solution, is
the basic alloying agent for hardening steel and in-
creasing its tensile strength. Thi s is done at some
sacrifice to its ductility, or its ability to stretch without
failure, and in weldability.
Manganese is also a powerful hardening agent,
serving as a ferrite strengthener. Molybdenum, chro-
mium, vanadium and columbium are also added in
small amounts, and they function in a similar manner,
especially with regard to atmospheric corrosion resis-
tance and toughness.
The hardness of structural steels is important as it
relates not only to strength, but also weldability. Too
hard a steel leads to difficulties in welding and to weld
defects, such as brittle fracture and shrinkage crack-
ing. To ensure good weldability, ASTM standards limit
the amount of alloying elements added.
Carbon Equivalent
Dr. Winterton of the British Welding Institute, while
studying the effects of various elements in the chemi-
cal composition of steels on their mechanical proper-
ties noticed their effect on hardness. Knowing that
hardness is related to weldability and susceptibility to
cracking, he developed the relationship of the chemi-
cal elements in steel to its hardness through a Car-
bon Equivalent formula, basically because carbon
has the greatest effect on hardness, strength and
weldability in steels.
Many different Carbon Equivalent (CE) formulas
are used as a guide for pre-heat requirements and in
welding procedures. The AWS D1.1-90 Apendix XI
Formula is used for structural steels.
CE=C+(Mn+Si) + (Cr+Mo+V) + (Ni +Cu)
6 5 15
(in % to determine the Zone requirements for pre-heat.)
C = Carbon; Si = Silicon; Mo = Molybdenum; Ni = Nickel;
Mn = Maganese; Cr = Chromium; V = Vanadium;
Cu = Copper.
.4o
IO
.00
UOT AL 5TE /
/ _ l KBO K.IIGE
I .WJ; KEG.UI&E MEklT.5 /AW.. RUIF, E.I,4' :NT
] hOUATE ExOE.'T l"o / MVST E INCKEh',E
] &lOlP5 4' '' 5, 5H/,.pE5 ' / SEE I , 0 ' 5p.%5"
7.ONE I
.J_7
.IF
.o .o . .5o .4o .
cE
chart 2. carbon Equivalent
Note: chart based on AWS D1.1-90 Appendix Xl, Fig. Xl-1.
This method does not take thickness into effect.
11
Quenching and Tempering
For High Strength
To obtain a high-strength steel other than by alloy-
ing, it is possible to heat-treat certain steel formula-
tions by quenching and tempering. Tempering in-
creases ductility but lowers tensile strength. Tem-
perature and time are important. This procedure re-
quires the steel to be heated initially to an elevated
temperature above the upper critical to form a crystal-
line structure known as austenite. This is followed by
rapid cooling in water to produce martensite or a par-
tially martensitic microstructure. When tempered to
precipitate a fine dispersion of carbides, this structure
has good ductility and fracture toughness, along with
high strength.
Of the many grades of structural steel, only A-514
and A-852 are quenched and tempered. Yield
strengths are on the order of 90 to 100 ksi, almost
three times that of A-36 steel and twice that of other
grades. Of course, the cost of this special treatment is
reflected in a significantly higher price.
The Welding Processes
Along with the development of better structural ma-
terials came irnlxoved methods of joining steel. Of the
methods described below, some are restricted to
shop use only; others find widespread use in both the
shop and the field. There are four pre-qualified weld-
ing processes:
1. Shielded Metal Arc Welding (SMAW)
2. Flux Cored Arc Welding (FCAW)
3. Gas Metal Arc Welding (GMAW -
in the spray mode)
4. Submerged Arc Welding (SAW)
Other welding processes such as Electro Slag
Welding (ESW), Electro Gas Welding and Gas Metal
Arc Welding -- short circuit mode can be qualified-by-
test.
DIRECTION OF TRAVEL m,
ROD CORE
FLUX CO
~
ROD CORE . . . . ] : CUEn
COVERED ELECTRODE
(conlul,mlll )
PO
MOLTEN FLUX GASEOUS SHIELD
FUSED SLAG
ARC PLASMA
SOLIDIFIED WELD METAL-- --MOLTENWELD METAL
+
ER
Fig. 6 Schematic illustration of Shielded Metal-Ars Welding
(SMAW). Reverse polarity is shown (D.C. electrode
positive).
Shi el ded Metal Arc Wel di ng (SMAW)
Commonly referred to as "stick' or manual welding,
this is the oldest and most popular method of struc-
tural welding and involves the use of flux-coated elec-
trode (stick), which is consumed in the process,
Figure 6.
SMAW welding has a long and successful history.
As compared to other processes, it is ideal for field
application. The equipment is less costly; it is more
portable; electrodes may be purchased "off the shel f
in most locations; and there are also mare people
qualified to perform this type of welding. Therefore, it
is relatively easy to find welders who have the exper-
tise.
SMAW welding is particularly advantageous when
the job involves repetitive starting and stopping, as in
short fillet welds. It is also the preferred method for
joint repairs, following automated welding. Develop-
ments through the years have been directed at mini-
mizing the formation of hydrogen gas by removing
water from the electrodes (Iow hydrogen). Therefore,
common field practice calls for the electrodes to be
dry before use. The exceptions are the cellulose-
coated electrodes (E6010 and E6011 that give good
penetration but are not Iow hydrogen electrodes)
which are compounded to contain 2 to 5 percent
moisture. Redrying can actually impair the quality of
these electrodes. AWS D1.1-90 requirements for stor-
age of all electrodes should be followed.
Fl ux-Cored Arc Wel di ng (FCAW)
This process, Figure 7, employs a tubular electrode
with the flux contained within the core of the tube.
DIRECTION OF TRAVEL z,
Fig. 7 Schematic illustration of Flux-Cored Arc Welding
(FCAW). Reverse polarity is shown (D.C. electrode
positive).
There are two versions: the self-shielded type uses
flux compounds alone to protect the weld from oxida-
tion during cooling. The gas-shielded type uses flux
compounds, plus an auxiliary shielding gas (usually
carbon dioxide) for weld protection, 75% CO2 + 25%A
is becoming popular to minimize spatter. In general,
only the self-shielded type is used for field applica-
12
tion. Both are semi-automatic, high-production
methods. Although equipment is bulky, FLAW is the
method of choice for high-production, deep penetra-
tion welding on high-rise structures. Typical deposi-
tion rates (about 8 to 12 pounds an hour) are about
twice that obtained from normal stick welding. Figure
8 compares the depth penetration of a fillet weld pro-
duced by FCAW and SMAW.
FLAW. Though developed primarily for the aircraft in-
dustry, it is becoming popular in structural shops; it is
not practical for outside or field application.
Most structural sections for buildings and bridges,
welded in prefabrication plants or temporary fabrica-
tion plants, use the SAW process with a fully auto-
matic setup. This process, Figure 10, deposits a flux
powder in advance of the electrode, so that the result-
ing arc produced is submerged in the flux and not vis-
ible to the operator. It is the workhorse of the struc-
tural shop but is losing ground to FCAW, GMAW and
ESW.
Fig.8 Fillet welds by left, and FLAW, right, in A36
steel Note the increased penetration of FLAW.
Gas Metal Arc Welding (GMAW)
This welding process is pre-qualified when used in
the spray transfer mode; it must be qualfied-by-test
for the short circuiting transfer mode. The electrode is
a solid welding wire and the shielding gas is either Ar-
gon, Helium or CO2 or a combination of these gases,
and is semi-automatic. The power source is direct
current and reverse polarity is the usual polarity
(DCEP). It produces a very clean weld and deposit
rates (lbs. per hr.) are very good, comparable to
SHIELDINGGAS]
SOLIDWIREELECTRODE J
CURRENTCONDUCTOR - / / "
Pw
TRAVEL
ARC /
I
-- MOLTENWELDMETAL J
Fig. 9 Gas Welding - spray arc mode.
WIRE FEEOROLLS
ELECTRODEWIRE
(nmibl'} +
CONTACT TUBE
;ED FLUX
SLAG FLUX
0 METAL ARC PLASMA
DIRECTION OF TRAVEL I,
UNFUSED FLUX
SLAG
SOLIDIFIED WELO METAL
Fig. 10 Schematic illustration of SubmergedArc Welding
(SAW). Reverse polarity i s shown (DCEP).
SAW is particularly well-suited to long welding runs
(30 feet or more). It can be used on thin or thick sec-
tions of metal and is capable of producing high-quality
fillet, partial penetration or full penetration welds at
typical high deposition rates, but is restricted to flat
and horizontal welding positions.
All of the above processes, when properly qualified
in accordance with the standards established by the
American Welding Society, will yield satisfactory weld
joints for structural applications.
Electroslag Welding (ESW)
Introduced to construction use in the 1960s,
electroslag welding is the newest production welding
process. Its chief feature is its unsurpassed produc-
tion capacity, depositing filler at 35 to 50 pounds per
hour, while producing a clean, high quality joint on a
continuous casting basis. This is not a pre-qualified
process, but must be qualified-by-test and the test re-
sults submitted.
As illustrated in Figure 11 (see next page), there is
no arc. The slag, heated by electrical resistance,
melts the filler electrode and parent metal. Melting of
the electrode in the slag cleans the metal by providing
13
r
I i !
GUIDETUBEANDWIREFEED [
TOP VIEW
1 COPPERDAMS
I
l
RUN
RUN-OFFBLOCKS
(removedaftwwed ismade) J
WIREELECTRODE
(mng m'Jagpoo
SOLIOIFIEDWELDMETAL
I
t
o,
-'5
WIRE ELECTRODE
GUIDE TUBE
--I
I ORECTI ON OFTRAVEL
MOLTENSLAGPOOL
MOLTENWELD
T j r STARTINGSUMP
(removed afterweld ma d e )
FRONT VIEW
PLATE
Fig. 11 Schematic of electroslag process for welding
O/pica/butt joint.
Note: Single or multiple electrodes may be used; with or
without oscillation for either will depend on plate thickness.
of gap is 1" to 1-1/2'. Techniques are being devel-
oped to eliminate the starting sump.
excellent slag-metal contact. The parent metal, sur-
rounding the molten pool, is heated deeply and the
resultant slow cooling allows time for gas bubbles to
escape, keeping porosity to a minimum.
Although the process is used primarily in the shop
for butt welding of plates and for the final closure
welds of interior stiffeners used in box columns using
the key-hole weld technique, it has been adapted for
field welding of solid prismatic members (approxi-
mately 6 inches thick).
In spite of cleanliness, welded sections using the
ESW process will often show lower fracture tough-
ness than the parent metal, especially at tempera-
tures below 0F. The mechanical properties are
equivalent in all other respects. Special ESW tech-
niques have been developed recently to improve weld
toughness through grain size control. In shops so
equipped, the fabricator can also obtain further im-
provements by "normalizing" the weld zone. This in-
volves reheating the welded area to 1,650F to
1,700F to form austenite, followed by air cooling.
1 4
Normalizing produces a uniform, refined grain struc-
ture, with improved fracture toughness. This process
is not generally used in construction because the high
cost cannot be justified.
Avoiding Weld Defects
Although welding has been with us a long time, its
application to structural use was impeded by early
failures in the bridge and shipbuilding industries due
more to workmanship and lack of proper welding pro-
cedures than to materials. Much research has been
done over the past four decades, not only in the de-
velopment of better base materials, but also in provid-
ing electrode filler metals that better fit the metallurgi-
cal properties of today's structural steels. Table B
(see next page) shows the variety of filler materials
available for the type of steel and welding process be-
ing used.
Many of the weld defects found in structural appli-
cations are caused less by the quality of base and
filler metals than by poor joint design in combination
with improper welding practice. These factors lead to
a variety of weld defects, including the following:
Wel d Cracki ng
Figure 12 (see page 16) illustrates the metallurgical
features of a typical weld. The weld area is divided
into a fusion zone and a heat affected zone (HAZ).
The HAZ microstructure reflects changes in the
grain structure of the parent metal produced by heat
from the adjacent molten metal. In Figure 12, good fu-
sion is indicated in the weld itself. However, the lack
of preheat allowed the joint to cool too quickly. This in
turn produced hard, brittle martensite in the HAZ and
led to cracking in the HAZ and the parent metal.
All structural steels experience some degree of
hardening in the HAZ due to high heat inputs during
welding. The hardness increase is minimal for Iow-
strength structural steels, such as A-36, but for higher
strength and for thick sections of all structural grades,
the hardness increase can be embrittling. Conse-
quently, specific welding controls, such as heating be-
fore welding (pre-heat) and post heating, slow the
cooling rate and prevent the formation of HAZ mar-
tensite.
With proper weld procedures and good workman-
ship, such cracking can be eliminated. Other types of
cracks can be encountered in structural welding, i.e.,
under bead cracking, cold cracks, hot cracks, etc.
Hydrogen Embr i t t l ement
When steel is melted during welding, hydrogen
may dissolve in the molten metal or the parent metal
HAZ. Molten metal has a great affinity for atomic hy-
AWS
Group
ASTU Steel
Specification
& Grade 1
Table B
Filler Metal Requirements
Filler Metal Requirements
Tensile Yield
Electrode Specification Strength SIzength
& Classification ksi Minimum ksi Minimum
A36
A53 Grade B
A139 Grade B
A500 Grades A, B
A501
A709 Grade 36
SMAWAWS A5.1 or A5.5
E60XX & E70XX 62/72 50/60
FCAW AWS A5.20
E6XT-X}
E7XT-X} except -2, -3, -10, -GS
62/72 50/60
GMAW AWS A5.18
ER70s-X 72 60
SAW AWS A5.17 or A5.23
F6X-EXXX F7X-EXXX 62/72 50/60
A139 Grade B
A,?.42
A441
A500 Grades C, D
A572 Grades 42, 50
A588
A709 Grades 50, 50W
SMAW AWS A5.1 orA5
E70XX (Iow hydrogen) 72 60
FCAW AWS'A5.20
E7XTX-X except -2, -3,-10 -GS 72 60
AWS A5.18
ER70S-X 72 60
SAW AWS A5.17 or A5.23
F7X-EXXX 70-90 50
III
A572 Grades 60, 65 SMAW AWS ),5.5
E80XX (Iow hydrogen) 80 67
FCAWAWS A5.20
E7XTX-X except -2, -3, -10, -GS 72 60
GMAW AWS A5,?.8
ERSOS-X 80 65
SAW AWS A5,23
F8X-EXXX 80-100 65
IV
A514 (over 2-1/2')
A709 Grades 100
lOOW (2-1/2' to 4')
SMAW AWS A5.5
E100XX (low hydrogen) 100 87
FCAW AWS A5.29
E10XTX-X 100-120 65
AWS A5.28
ER100S-X 100 90
SAW AWS A5.23
F10X-EXXX 100-130 88
V
A514 (under 2-1/2')
A709 Grades 100
100W (under 2-1/2')
SMAW AWS A5.5
E100XX (Iow hydro<3en) 110 97
FCAWAWS ,6,5.29
E1lXTX-X 110 98
GMAW AWS A5.23
ER110S-X 110 98
SAW AWS A5.23
F11X-EXXX 110-130 98
In joints involving base metals of two different yield points or strengths, filler metal electrodes applicable to the lower strength base metal
may be used, except that if the higher strength base metal requires Iow hydrogen electrodes, they shall be used.
15
Fig. 12 Weld macrostructure (at 1.2X size), A-36 steel, P/M1, j oi ned to heat treatedsteel, P/M2. Good fusion is
indicated, accompanied by prominent heat affected zones (HAZ), which show darkest etched structures; however, there
is lack of penetration at the root of the partial penetration groove weld. Arrows designate cracks in of P/M2. The
indentations are Rockwell hardness test impressions. Joint was fabricated by SMAW (ticld' welding), with E7018
electrode. (Cracldng due to lack of preheat and unbalance of weld.)
drogen, but Iow affinity as cooling takes place and hy-
drogen is rejected. Some hydrogen gas may become
trapped in the weld metal and create high internal
pressure which can induce micro cracks in the steel.
Hydrogen cracking is not a serious problem in struc-
tural steels, especially if Iow hydrogen electrodes and
preheat are used. The source of the hydrogen is
water, which may originate from the shielding of a
SMAWor FCAW electrode, the flux of the SAW pro-
cess, the steel base material or the weld environ-
ment. In the welding arc, water will break down into
hydrogen and oxygen. Hydrogen in the fusion zone
deposited metals or in the HAZ of the parent metal
can cause embrittlement. If the HAZ has become
hardened during welding, the sensitivity to hydrogen
embr'lement is even greater.
Preheating of the metal by AWS standards and
use of Iow-hydrogen electrodes are the best means of
avoiding hydrogen embrittlement. See Table C (page
17). Preheating expels water before welding and al-
lows residual hydrogen to escape by slowing the
cooling rate and thus decreases hardening especially
in the critical heat affected zone (HAZ). As mentioned
earlier, heat drying of electrodes is also desirable for
removing excessive moisture.
Lamellar Tearing
One of the most disconcerting welding defects
caused by poor joint design in combination with bad
welding practice is fiamellar tearing." This occurs in
highly restrained joints because the designer may not
fully understand the anisotropic properties of the base
metal, and the fabricator has not undertaken ad-
equate preventive measures during welding. Design
information through AISC is available to aid the engi-
neer, detailer and fabricator to reduce the occurrence
of these defects.
Figures 13 and 14 (see page 18) illustrate the ori-
entation of a steel plate relative to the direction of roll-
ing, or its longitudinal (X) direction. (In structural
steels, mill test reports are concerned primarily with
longitudinal properties.) Transverse (Y-direction)
16
Table C
Minimum Preheat & Interpass Temp F
AWS
Category
Steel Specification=
(For AISC-89 & AWSDI.I-90)
Thickness
Welding Process (inches)
AWS D1.1.90 Rec. Min.
MinimumTemp. F Temp.OF
A ASTM -A36
-A53 Grade S
-A139 Grade B
-A500 Grades A, B
-A501
-A709 Grade 36
Shielded Metal Am
Welding with other
than low hydroaen
electrodes (SMAW)
Up to 3/4 ...................... None ............................... 50
>3/4 thru 1-1/2.............. 150. .............................. 150
>1-1/2 thru 2-1/2........... 225 .............................. .250
Over 2-1/2 ..................... 300 - -
> 2-1/2 thru 3-1/2 ............................................... 300
Over 3-1/2 .......................................................... 350
or greater
See AISC J2.7
B ASTM -A36
-A53 Grade B
-A242
-A441
-A500 Grades C, D
-A501
-A572 Grades 42, 50
-A588
-A70g Grades 36,
60, 50W
Shielded Metal Arc
Welding with Iow
hydrogen electrodes
Rux Core Am Welding
Gas Metal Arc Welding
Submeq;:l Am Welding
Up to 3/4 ...................... None ............................... 75
>3/4 thru 1-1/2 ............... 50 ................................. 100
>1-1/2 thru 2-1/2........... 150 ............................... .200
Over 2-1/2 .................... 225 - -
> 2-1/2 thru 3-1/2 ................................................ 300
Over 3-1/2 .......................................................... 350
or greater
See AISC J2.7
C ASTM -A572
Grades 60,65
SMAW (low hydrogen
electrodes)
FCAW
GMAW
SAW
Up to 3/4 ......................... 50 ................................. 75
>3/4 thru 1-1/2 .............. 150 ............................... 200
>1-1/2 thru 2-1/2 ...........225 ............................... 250
Over 2-1/2 .................... 300 - -
2-1/2 thru 3-1/2 ............................................... 300
Over 3-1/2 ......................................................... 350
or greater
See AISC J2.7
D ASTM -A514
ASTM -A709
Grades 100 & 100W
SMAW (low hydrogen
electrodes)
FCAW
GMAW
SAW
Up to 3/4 ......................... 50 ................................. 75
>3/4 thru 1-1/2 .............. 125 ............................... 150
>1-1/2 thru 2-1/2........... 175 ............................... 250
Over 2-1/2 ..................... 225 - -
> 2-1/2 thru 3-1/2 ............................................... 300
Over .........................................................350
or greater
See AISC J2.7
NOTE 1: Table C based on data from AWS D1.1-90 and 1989 AISC Specifications for Steel Buildings.
17
I '3 t ' ...... ' ' ' ' / Rolling 10
110
Angleofipecimenfrom
u 100 --- platesurface.
8
o 9O
80
; % YieldStrength
L (Transverse Direction) 3o ,o
Fig. 13 Directional nomenclature for describing anisotropy I .. R u
in rolledplates. 20
10
properties, while usually lower in ductility and tough-
01 I I I I I I t I
ness than longitudinal properties, are nonetheless
predictably good. Strength and ductility are the lowest
in the through-thickness (Z) direction.
The reason is that certain non-metallic compounds
are formed during the alloying process and become
trapped in the steel. These so-called "inclusions" are
round in shape when the steel is cast into ingots, but
become flattened and elongated during rolling. They
are usually microscopic in size and difficult at best to
detect even with ultrasonic testing.
During welding of a highly restrained joint and be-
cause of the resultant through-thickness tension,
these flatened inclusions, usually silicates, alumi-
nates, sulfides and oxides, can link up to form micro-
fissures and, eventually, visible cracks. Figure 15 (on
next page) illustrates how lamellar tearing can de-
velop in a T-weld section. Following welding, contrac-
tion strains are generated as the heated metal cools.
When the assembly is highly restrained (i.e., when
portions or all of the assembly are kept from contract-
ing during cooling) and when the welding procedures
are so sequenced that the strain demand cannot be
accommodated by plastic deformation in some ele-
ment, the microfissures grow into tearing cracks.
The worst case of lamellar tearing is delamination,
or complete internal separation of the steel (Fig. 15).
This is more likely to occur in larger weldments using
thicker base materials because more non-metallic in-
clusions collect with size and because of the greater
0 10 20 30 40 50 60 70 80
ANGLEOFSPECIMENFROMPLATESURFACE-{ Z
Fig. 14 Relation between tensileproperties and angle of
specimen from theplate surface to demonstrate loss of
properties in the Z direction.
restraint derived from rigidity. This delamination effect
occurs very rarely, however, if proper procedures,
techniques and good workmanship are followed.
Engineer's Role in M i n i m i z i n g Weld Defects
In general, weld defects are most likely to occur
when using higher-strength carbon steels, when join-
ing thicker sections of steel (over 1-1/2 inches) and
when applying large amounts of weld metal. The
higher carbon steels, by their chemical nature, are
less ductile and, therefore, cannot as easily accom-
modate the strain demand accompanying weld
shrinkage. For these reasons they are no longer used
in building structures, the preference being Iow-alloy
steels.
The tendency to produce weld defects when
thicker sections are being used arises from the fact
that for most steels to maintain a constant minimum
yield strength as the thickness is increased, there is a
corresponding increase in the amount of carbon and
manganese requirements. Therefore, the thicker the
section, the less ductile the steel. Also, for thicker
steels, there is less working of the hot metal during
90
18
/
/
/
/
/
// /
/ / \
/ / \
, / \
/
%.
FLATTENED NON-METALLICS
_ _
- _
/ INTERCONNECTINGSHEAR
MICROCRACKS
Fig. 15 Lamellar tearing cracks developing under a T-weld.
Microfissures initiate at flattened non-metallic particles
which grow to cracks by interconnecting tears, note
stepped ruptures; can be picked up by UT.
A p
ERECTN
80 L'r5 - (
POTNTL'
LANELLAR
TIBAK
+
+--
F
Z
SECTION ,A-A
QL,IE5-[IONABLE DETAIL
,,Dp . . . iL_.
FOR IBE/,,RING, J
IF REUIR
IMPROVED DETAIL
Fig. 16 Column splices.
TIlL
LAHELL-AR TEAR-
"x,x
CLIP'
SECTION C-C
QUF__-STIONAISLE DETAIL.
SECTI ON D-D
IMPROVED DETAIL
(SHOP NELDEDmFIELD tSOLTED kN[)/dR FIEL[7
Fig. 17 Beam splices.
-HIH ESTRAINT ROt] , j I I NE, A5 NELL
' - ' - P CONTINUITT r-/r AS pLANES,
. p( 2F EA, DEAN, 5 E C T f ON E - E
CUESTIONAELE DETAI L
i
, [ - -
; i ; i
51MP!.EE PAN ) -
fEIRE5 NOMELD- :
IN OF Pt.-,.
1 5 RESULT J t J TO cox-: _.o
,,..,o,.,,_.-,' 5 [ ? Z
GONI'INUITT /Ir., I f51BL' '
ECTIOH P -
IMPROVED DETAIL_
Fig. 18 Beam- Column Details.
19
]
L ev,c.o ^Y i f=,LL.-I.OAY 'i ^E ..,V
C)I41'RIk:TION MON COOLIII [.J,T I- EITHER DY'
yIIL.OIN Qq[ G I N I Oral eL.OHC./TIOFI P' TNE FULl. M'MBER.
Fig. 19 Full Restraint.
I. ELD CONTRACTION VS. TIHT FIT-UP
POTENT/hE
. . . . _ f- -- ! . LfXHELL/& T
2. IIELD CONTRACTION V.5. PREVIOUSLY DEPOSITED IIELD METAL
Fig. 20 Internal Restraint in Weldments.
Fig. 21
rolling operations and, therefore, greater chance that
inclusions and discontinuities will be larger and more
prevalent with less chance of being forged-welded
together during rolling.
POTENTIAL
TEAR "m: "' "l
QUETIOHABLE
[:)ET.Al l -
(b)
QUESTIONABLE
DETAIL-
IMPROVED
DP--TAIL
IMP!OVF=O
D E T A I L
D---TAIL
m p o v o
D E T A I l -
--T
Welded Comer Joints.
The structural engineer is in the best position to
avoid situations that lead to weld defects by designing
joints that are not highly restrained. Figures 16
through 21 show some typical joints that are highly
restrained, as compared with alternative joint configu-
rations designed to minimize restraint.
Equally important in the design of weldments is the
requirement not to 'overweld." Often, if a joint is diffi-
cult to analyze, the designer specifies that every
available edge is to be welded, and the joint is thus
thought to be conservatively designed. Not only is
weld electrode material expensive, its cost to put in
place is about 20 to 30 ti mes that of the base material
cost. Therefore, the designer is obligated to consider
carefully the amount of welding. Savings in weld
metal al so means less distortion, less tendency for re-
straint and, consequently, less cracking.
In cases where the engineer cannot avoid design-
ing a highly restrained joint, there are compensating
techniques available to the fabricator of which he
should be aware. Among the techniques are the use
of preheat, post heat, controlled cooling and the se-
lection of more ductile electrodes. Peening, when per-
formed under knowledgeable and close supervision,
is helpful.
The sequence of welding is al so important. For
most applications, welding should begin at the center
of the mass of the weldment, where restraint is likely
to be concentrated, and proceed outward in block
Steps with the electrode travel directed toward the
center of the mass. Wire shims (called "softies") may
be used at critical points to provide the necessary air
gap within whi ch shrinkage can occur.
The engineer should call out for submittals by the
20
fabricator of welding procedures as per AWS D1.1-90
and of shop detail drawings as per AISC Specifica-
tions for Structural Steel Buildings and Code of Stan-
dard Practice.
Correcting Weld Defects
Welding codes in general prohibit cracks of any
type in the completed weldments. When cracks are
detected, the inspector will require that they be
repaired by removal and rewelding. The repair of
such defects is a normally encountered process
during welding operations, and AWS procedures are
applicable for repair by the fabricator and approval by
the knowledgeable inspector. There are occasions,
however, when repeated attempts at repair are met
with repeated failure, and the structural designer is
brought in for consultation either by the inspector in
support of his rejection or by the fabricator because
he may believe the design is contributing to the prob-
lem. It is in this type of adversary situation that the
designer must prove his worth as a diplomat and
mediator as well as a good engineering technician.
It is important that the structural designer retain his
composure and make every effort to determine the
facts without letting the "people problems" outshine
the welding problems. Keep in mind that repeated
repairs are costly to the fabricator who wants to pro-
duce an acceptable product, preferably without flaws.
Try to determine whether there is an actual rejection
by AWS D1.1-90 requirements and if the inspector is
being fair and reasonable in his demands or whether
it is a case of punitive reprisal for past, real or imag-
ined, grievances.
It is wise to resist taking over and directing the fab-
ricator how to perform the repairs. However, the de-
signer can be helpful by asking for review of the pro-
posed repair procedure and by following a formalized
check list to determine that all possible sources for
trouble have been considered. Among some of the
basic questions to consider are the following:
1. Is the exact chemistry as well as mechanical
properties of the base metal known? A rough
check of carbon equivalent would be helpful in
checking weldability.
2. Do the electrodes and other joining materials
comply with AWS and ASTM standards? Ask for
certification or if in doubt have them tested and
check storage conditions.
3. Are the electrodes and base metal compatible as
called for under AWS "matching" standards (see
Table B)?
4. Has the extent of the crack or defect been deter-
mined? Where is it located with respect to the
weld?
5. Do the welder, his supervisor and the inspector
all agree as to location and extent of the indi-
cated defect? Perhaps a third party may be help-
ful to settle arguments of this nature.
6. Can the design be revised to minimize restraint?
Can weld size and amount be reduced without
jeopardizing safety? Is the joint over welded?
Sometimes the fabricator can make suggestions
which can solve welding problems without reduc-
ing the design safety of the joint.
7. Is the fabricator using more than minimum re-
quired preheat to help slow the cooling down af-
ter weld completion? Sometimes post heating
and/or insulation blankets will help prevent crack-
ing.
8. Are the welder and inspector using a heat indica-
tor to determine preheat and interpass tempera-
ture? Guessing is not sufficient.
9. Prior to beginning the replacement weld, was
magnetic particle testing (MT) used to make sure
the entire defect was removed?
10. The entire repair procedure for important
weldments should be written out and reviewed
prior to starting the repairs and should include:
a. Size, type and AWS designation of electrode
material.
b. ASTM designation of base metal.
c. Sketch of defect showing size extent and lo-
cation in weldment.
d. Procedures followed for detection (NDT).
e. Preheat and interpass temperatures to be
used.
f. Post heat treatment or method to be used to
slow the cooling rate, such as asbestos
blankets or electric hot pads.
g. Procedures for reinspection after completion
of repairs.
11. Keep in mind that once the welding repair has
started it is mandatory to complete the repair
without interruption. Repeated heating and cool-
ing invites repeated cracking due to increased
potential of contamination.
12. If the fabricator has not already done so, it may
be heiful to suggest that a welding engineer or a
metallurgist knowledgeable in practical welding
problems be called in for consultation. Having
faced such problems many times previously, he
may be able to point out the technical cause of
the problem immediately.
13. Usual practice calls for the inspector to make a
daily inspection report and the fabricator is given
a copy with the original to the engineer. If the
fabricator disagrees with the inspector's report,
the inspector sends a non-conformance report
(NCR) to the engineer, copy to the fabricator, for
resolution of the dispute.
14. Remember, a fair, open-minded approach with a
desire to work cooperatively with the fabricator
can have the best chance of successfully
21
correcting the problem and keep it in its place,
out of court.
Correcting Weld Distortion
The discussion of strain demand in the section on
lamellar tearing described the volume changes of the
weld as it cools. The volume change must be ab-
sorbed as internal elastic and plastic strains, as
movement in some element, or as rupture. A lack of
fabrication skill in coping with these movements is
evidenced by distortion of the finished structure or by
cracking.
Two thick plates fit up, tacked at 900 and welded
together without fixtures can create angular distortion
(Figure 22). The accumulative angular distortion of
three weld passes is shown schematically in
Figure 23.
Fig. 23 Rotation in a butt weld. The rotation equals the
cumulative shrinkage from each weld layer. Techniques
have been developed which will minimize this effort.
Distortion fromtransverse weld strain demand with
the welds shown balanced and the plate fiat is shown
in Figure 24 below. The fabricator should position the
plates to account for the changes caused by the first
and second welds.
I
[r,S/SS/ZSZSZZZZSHHSJSSSZSS/SSZZSZSSSZH.' FINAL
Fig. 24 Transverse shrinkage in a butt weld.
Fig. 22 Angular distortion has resulted from weld shrinkage.
Compensating tilt of vertical member and/or use of strong-back
arc methods to control alignment to vertical position when
welded.
Distortion from longitudinal weld strain demand is
shown in Figure 25 (top of next page). The position of
the weld relative to the center of gravity of the cross
section produces the bow. Supplementary weld
beads are sometimes used to achieve the desired
camber. Sequence, technique and peening will mini-
mize distortion.
Fl ame Bendi ng & St rai ght eni ng
The flame bending technique, using thermal
upsettings, is used to straighten or curve members.
Localized heating of steel causes thermal expansion
and a reduction of yield strength in the heated sec-
tion. The expansion is inhibited by the cold, stronger
surrounding metal forcing the heated portion to yield
plastically to accept its own demand for increased
22
Z
r
Fig. 25 Longitudinal bowing in a welded beam may
produce either positive or negative camber (in X.Z plane).
Lateral bow (in X-Yplane) can occur.
volume. (See Figure 26 for yield strength at elevated
temperatures and Figure 27 for variations in modulus
of elasticity.) After cooling, the shape of the steel
piece is changed, and the heated zone recovers its
strength. Several heating and cooling cycles may be
required to complete an operation.
The "flame bending" technique is used in the shop
to flatten web plates, to camber beams, or to
straighten work distorted by welding. The maximum
temperature recommended for this operation is
1,200F for as rolled structural steels and 1,100F for
quenched and tempered steels (A514) but not higher
than the tempering temperature. Cooling may be in
air, or by water spray or wet rags for more rapid cool-
ing.
The temperature of this localized heating should
not exceed the critical (1,33301=) or undesirable
changes in mechanical properties may result. Heat
treated steels such as A-514 must not be heated
above the tempering temperature specified by the
steel manufacturer. This tempering temperature
should be obtained from technical information fur-
nished by the supplier.
140
120
100
80
60
40
20
0
1
0 200 400 600 800 1000 12001400 1600 18002000
TEMPERATURE--OF
Fig. 26 Variation in yield strength with temperature.
A-36 STEEL
\
\
\
\
\
0 200 400 600 800 1000 1200
TEMPERATURE -- OF
Fig. 27 Variation in modulus of elasticity with temperature.
Nondestructive TestinD (NDT)
One of the main reasons for the success of all
welded structures in the building industry has been
the development of fast and accurate methods of ex-
amining welded joints without destroying or impairing
their actual usefulness. Currently, there is a variety of
techniques being employed by the fabricator and in-
dependent inspection agencies to assess the reliabil-
ity of a weldment. Used properly, these methods can
reveal practically all of the common surface and inter-
nal defects that normally occur with improper welding
procedures and practices and will result in a quality
level consistent with Project Specification require-
ments. As in all inspection methods, the experience
and skill of the technician and an inspection proce-
dure developed by a Quality Control Engineer are
very important criteria for reliable nondestructive test-
ing and users should become familiar with all the limi-
tations of NDT methods.
The engineer may also request written documenta-
tion as to the type of quality assurance program es-
tablished by the fabricator. Many fabricators have
their own quality assurance program meeting the
nondestructive testing specifications established by
both local building codes and the American Welding
Society. AISC has also developed a set of standards
for quality certification and has designated member
firms meeting these standards as 'Category I, II or III
Certification."
23
The engineer who realizes the high degree of pro-
tection afforded by these various organizations
through their time-tested standards and specifications
can do much to simplify his own design specifications
related to the welding of structural steel. Streamlined
specifications referring to accepted industry-wide
standards and avoiding unnecessary abstruse ver-
biage will do more to assure that the specification will
be read and followed by the fabricator and contractor.
Visual Inspection (VI)
This is a requirement of the AWS D1.1-90 Struc-
tural Welding Code wherein the duties are detailed.
In-progress visual (edge preparation, fit-up, root pass
and fill-in-passes) by a qualified and experienced in-
spector is considered the most reliable method and
most cost effective. By far, most cracks in weldments
are detected visually by an alert wel der or inspector.
Sometimes detection is made hours or days after
completion of the weld. This has been termed "de-
layed cracking" when, in most cases, the cracks were
probably there at the completion of welding, but
merely opened up wide enough to see when the
entire weldment cooled. Proper visual inspection re-
quires careful examination in the areas outside the
weldment, particularly along plate edges and parallel
to the weld where cracking and lamellar tearing can
occur.
Magnetic Particle Testing (MT)
Thi s method is primarily for detecting surface
cracks or defects on or just below the surface of the
metal. Thi s method is particularly applicable to crack-
sensitive material and especially useful in detecting
fatigue cracking. During the test, a very strong mag-
netic field is applied to the weld area, and the surface
covered with a suspension of ferro magnetic particles.
Defects such as cracks, inclusions, etc., interrupt
lines of force, causing the particles to concentrate
around these areas. Often, the residual magnetic
properties created by welding is sufficient to allow the
use of magnetic particles without the application of a
magnetic field. Because this method is simple, easy
to read and the equipment portable, it is preferred for
examination of welds and adjacent areas for surface
cracking caused by weld shrinkage. The magnetic
particle method is also very useful during repairs to
see if the defect has been completely removed and to
examine individual weld passes and layers for hot
cracking.
Ultrasonic Testing (UT)
Since the development in the 1960s, ultrasonic
testing has grown to become the most important tool
in nondestructive testing of structural welded joints. In
this method, high-frequency sound waves are used to
locate and measure discontinuities in welded joints
24
and base metals prior to welding. Thi s method is very
sensitive in detecting both surface and subsurface
discontinuities. During testing, a sound wave is di-
rected towards the weld joint and reflected back from
the discontinuity and shown on a calibrated screen of
an oscilloscope (Figure 28). Thi s method is highly
sensitive in detecting planar defects, such as incom-
plete weld fusion, delamination, or cracks; however,
orientation is very important. As the wave strikes the
defect, the time-distance relationship will locate this
interception. This is shown on the oscilloscope and
indicates the location of the defect in the weld joint.
This ultrasonic method can detect internal planar
defects in sections of practically unlimited thickness.
AWS D1.1-90, Section 6, Inspection 6.13.1, sets pro-
cedures for steel thickness from 5/16'to 8' , but other
thicknesses may be tested by qualified procedures. It
is relatively portable and relatively fast. Most impor-
tantly, it requires access to only one side of a test
section.
There are some limitations to ultrasonic testing.
Rough surfaces reduce its sensitivity and reliability.
Also, the method does not produce a permanent
record of the tested weld joint. In addition, globular
defects, such as gas bubbles and other porosities,
are not easily detected. Because of the spherical na-
ture of these defects, ultrasonic waves tend to pass
around them rather than reflect back as with planar
defects. However, this deficiency of the ultrasonic
methods is not considered serious. Ultrasonic tech-
niques, as practiced, are normally limited to joints
with plate thicknesses above 5/16' and are very sen-
sitive to orientation and geometry. Most building
Fig. 28 Utltrasonic Testing of Weld.
codes require ultrasonic testing of complete joint pon-
etration groove welds.
The ultrasonic method is highly dependent on the
skill and integrity of the operating technician for
proper interpretation of the results and therein arises
a major weakness. An operator can quickly lose cred-
ibility if he calls for a joint to be completely gouged
out for a defect that cannot be found. Consequently, it
is easier for the operator, unless technically compe-
tent, to say nothing rather than risk being found
wrong and then subsequently challenged repeatedly
by the fabricator or contractor on the project.
On all special inspection calling for ultrasonic test-
ing of welded joints by an independent testing
agency, it is important for the engineer to seek evi-
dence as to the qualifications of the ultrasonic techni-
cians involved. In particular, the engineer should
verify that the National Bureau of Standards has in-
spected the agency and qualified the NDT techni-
cians per "Recommended Practice for Determining
the Qualifications of Nondestructive Testing Agen-
cies." UT technicians are usually qualified by ASNT
Recommended Practice SNT-TC-IA.
Radiographic Inspection (RI)
Radiography relies on the use of electromagnetic
radiation to determine the soundness of a weld. X-rays
and Gamma rays are the two types of waves used to
penetrate solid materials such as a welded joint. A per-
manent record of the weld structure is obtained by
placing a sensitized film at the back of the weldment.
As the rays pass through the weld material, they fall on
the sensitized film and produce a negative of varying
intensity. If the rays pass through gas bubbles, slag in-
clusions or cracks, more rays will pass through these
less dense areas and will register on the film as dark
areas. Orientation of the discontinuity is very irnpor-
rant, especially for planar discontinuity.
Although radiography is a superior method of de-
tecting porosity defects and slag inclusions, for practi-
cal reasons it is not a suitable method of examining
some welded joints. This is because the film must be
placed opposite the source of radiation to graphically
record the defects, and the actual geometry of com-
pleted joints, particularly T-joints, generally prohibits
proper placement of the film.
Liquid Dye Penetrant Testing (PT)
This method relies upon surface tension and capil-
lary action of certain dye-carrying liquids to penetrate
small surface cracks. Subsequent application of a suit-
able developer brings out the and outlines the de-
fect. During the test, the surface weld is cleaned and
dried, then coated with a thin film of the penetrant. Af-
ter waiting a short time for the dye to flow into the
cracks, the surface is wiped clean and the developer
applied. The liquid ponetrant will then bleed out onto
the surface to react with the developer and sharply
outline the crack so it can be seen or photographed.
The use of dye penetrants in multi-pass welding has
been limited to investigative use only because of the
interruptions to welding process and consequent cost
to fabricator. Also, there is a possible health hazard to
welders.
Project Specifications
It has been the writers' experience in reviewing hun-
dreds of project specifications through the years that
there has been a needless waste of effort on the part
of the designers in writing and rewriting portions of all
of the standard AISC and AWS specifications. Some-
times, needless litigation has ensued as a direct result
of rewriting nationally accepted standards to include
the personal bias of the designer, albeit arising from
previous bitter experience.
For the most part, fabricators, welders and inspec-
tors are well aware of the national standards and keep
up with them. The designers are well-advised to do
likewise. They will get a better product with less confu-
sion and discord if they adopt them by reference and
omit any attempt to elaborate, clarify or otherwise
tamper with the nationally accepted standards unless
there is specific conflict with the project specif'mations.
There have been many large, successful projects
completed with a one line specification item that
merely states, "All materials and workmanship shall be
in accordance with the latest revised edition of the
AISC Manual of Steel Construction, which includes the
Specification for Structural Steel Buildings, the Code of
Standard Practice and the AWS Structural Welding
Code."
However, for those who feel a project specirmation
is only sufficient when it has a few pages under each
section, a recommended list of items is included as a
check list and reference in the following section,
"Project Specifications Check Est."
Normally, the nondestructive testing section of proj-
ect specifications is more detailed than other sections
devoted to structural steel. This is because the AISC
Code of Standard Practices requires that, "When
nondestructive testing is required, the process, loca-
tions, extent, technique and standards of acceptance
are clearly defined in the contract documents." This is
also in the AWS Dl .l -9OStructural Welding Code-
Steel.
It is well to keep in mind that inspection require-
ments may vary between local, state and federal build-
ing regulatory jurisdictions. Standard inspection re-
quirements should satisfy most jurisdictions because
normal practice requires continuous inspection by a
qualified inspection agency paid for by the owner, for
25
whose benefit the inspection is being performed. How-
ever, the designer is cautioned to determine for himself
what differences, if any, are required by the governing
agency for each project.
It is suggested that the structural steel designer ob-
tain a copy of the AISC publication "Quality Criteria
and Inspection Standards" and study it carefully, par-
ticularly as it relates to dimensional tolerances. If the
structure being designed requires closer construction
tolerances than allowed, either change the design to
accommodate them or put a large sign on your draw-
ings to the effect that care must be exercised by the
fabricator and erector to meet closer than normal toler-
ances, and then spell them out so there is no room for
misunderstanding.
If the structure is tied into or otherwise supported on
a masonry or concrete structure built prior to erection
of the steel work, don't expect the anchor belts to be in
exact position. Make provisions in your design of the
connections for misalignment vertically and horizon-
tally of such anchors and/or make a field check man-
datory. A review of normal construction tolerances for
such construction will be enlightening, to say the least.
If the building structure will not resist wind or earth-
quake forces until materials other than structural steel
are in place, it is recommended practice to notify the
contractor with a note on the drawings or in the specifi-
cations which clearly state that fact. Such a require-
ment is contained in the AISC Code of Practice, Sec-
tion 7.9 and particularly 7.9.3 entitled "Non-Self-Sup-
porting Steel Frames."
Project Specifications C h e c k List
The following list of items is recommended as a ref-
erence and check list to help develop the project speci-
fications:
1. Scope of work
2. References to National Standards (AISC, AWS,
ASTM, UBC, RCSC, SSPC, etc.)
3. Shop detail drawings submittals
4. Welding procedure specifications submittals
5. Materials (List ASTM Specifications, Structural
Steel, Pipe & Structural Tubing, High-Strength
Bolts, Std. Bolts, Nuts & Washers)
6. Welding Processes - Shop & Field, Pre-Qualified
and Qualified-by-Test
7. Filler Metal Specifications & Classifications
8. Quality Control and Assurance
9. Fabrication
10. Erection
11. Painting - Shop and Field
12. Inspection - Shop and Field, including verification
of welder's certification.
Welding Procedure Specifications
Confusion still persists among some structural engi-
neers and fabricators regarding written welding proce-
dures.
26
AWS D1.1.-90 Structural Welding Code - Steel
clearly states in Section 5.1.2 that "All pre-qualified
joint welding procedures to be used shall be prepared
by the manufacturer, fabricator or contractor as written
pre-qualified welding procedure specifications, and
shall be available to those authorized to use or exam-
ine them."
In spite of this mandatory requirement, some engi-
neers do not require their submittal. Some fabricators
always submit them for review while others neglect to
do so.
The engineer reviewer can easily check to see that
the AWS Code is followed by comparing the submittals
with the sample forms shown in Appendix E and check
lists in Appendix H. These cover the mandatory code
requirements of written procedure specifications
(WPS).
Alternately, the engineer may require that they be
reviewed by a qualified welding engineer employed by
an inspection agency.
In addition to all pre-qualified joints, all other joints
must be qualified prior to use by tests as prescribed by
Part B of Section 5 of the same code.
Shop Detail D r a w i n g s
Shop drawings have been the subject of much de-
bate for many years, and yet there still remains an ab-
sence of a uniform understanding within the design
and legal professions and the construction industry.
There has been wide variation in the manner in which
shopdrawi ngs have beenused, leading to a great deal
of confusion.
Shop drawings are necessary to facilitate steel fabri-
cation and erection, and installation of various ele-
ments of the work. Their very nature is such that they
are required to comply with the contract documents.
Review of shop drawings is simply to confirm compli-
ance and to facilitate progress of the work.
It is the position of the authors of this paper that
much of the confusion that exists has come about be-
cause of the use of shop drawings as design docu-
ments. We believe that the concept that shop drawings
are part of the design process must be eliminated.
Shop drawings are not part of the contract docu-
ments and must not be used as such. If changes are
proposed, or made by the contractor (or the engineer),
they must be done through the change order process
(or equivalent). When shop drawings are used as an
instrument of change, it can only lead to confusion.
Clearly, changes to the contract documents or the
submittal of details or systems based upon perfor-
mance-type specifications must be reviewed and ap-
proved by the engineer of record. Thi s process must
take place through documents other than shop draw-
ings, such as change orders or "supplemental design
details."
About the Authors
F. Robert Preece is co-founder and President of Preece/Goudie & Associates,
a civil and structural engineering firm engaged in the design and seismic
analysis of buildings and special-purpose structures. Mr. Preece has had more
than 40 years of experience in civil and structural engineering, including special
expertise developed in the failure analysis of buildings and materials. His experi-
ence enables him to write from the viewpoint of the structural designer, the
structural steel contractor and the engineering materials testing laboratory.
Mr. Preece holds a BS degree from the University of Nevada and an MS degree
from Stanford University. His memberships include the American Institute of
Steel Construction, American Welding Society, American Society of Civil
Engineers, Consulting Engineers Association of California (past president),
International Conference of Building Officials, Structural Engineers Association
of California (past president), Structural Engineers Association of Northern
California (past president), American Concrete Institute, Applied Technology
Council (past president) and the Earthquake Engineering Research Institute.
Alvaro L. Collin is a Consulting Engineer with California registration in Civil
Engineering and Metallurgical Engineering. He received a BS degree from the
University of California, Berkeley, in 1941 as a Civil Engineering major and a
Mechanical Engineering minor. After 24 years with Kaiser Steel Corporation as
Manager of Engineering of the Fabrication Division, Southern California, and
Senior Development Engineer, Steel Manufacturing Division, Oakland, CA, he
has been consulting the past 10 years on welded construction, heavy equipment
design and material handling systems.
Mr. Coffin is a life member of the Structural Engineers Association of Northern
California. He has been a member of the Board of Directors and the Steel and
Seismic committees of SEAONC. He is a long-time member of the American
Welding Society, having served on the National Board of Directors, on the
National Qualification & Certification Committee and as chairman of the Los
Angeles and San Francisco sections. Al has been awarded the National, District
and Section Meritorious Awards of AWS. He has served on AISC and AISI
Code Committee Task Groups and is a member of the Earthquake Engineering
Research Institute.
27
STRUCTURAL STEEL EDUCATION COUNCIL
470 Fernwood Drive
Moraga, CA 94556
(510) 631-9570
SPONSORS
Adams & Smith
Walker Div. Butler Manufacturing Co.
C. A. Buchen Corporation
Hoertig Iron Works
Hogan Mfg., Inc.
Junior Steel Co.
Lee & Daniel
McLean Steel, Inc.
Oregon Steel Mills
PDM Strocal, Inc.
Reno Iron Works
H. H. Robertson Co.
Stockton Steel
The local structural steel industry (above sponsors) stands ready to assist you in
determining the most economical soluOon for your products. Our assistance can
range from budget prices and estimated tonnage to cost comparisons, fabrication

September 2001

Notes on Design
Of Steel Parking Structures
Including Seismic Effects

By

Lanny J. Flynn, P.E., S.E. and Abolhassan Astaneh-Asl, Ph.D., P.E.
Principal and Professor
Vice President of Design-Build Services Department of Civil and Env. Engineering
Chalker Putnam Collins & Scott, Tacoma, WA University of California, Berkeley, CA

_________________________________________________________________
Copyright 2001, by Lanny J. Flynn and Abolhassan Astaneh-Asl, All rights reserved.
Notes on Design of Steel Parking Structures-Including Seismic Effects, L.J. Flynn and A. Astaneh-Asl, 2001, All rights reserved.

1
Notes on Design of Steel Parking Structures-Including Seismic Effects

By Lanny J. Flynn and Abolhassan Astaneh-Asl

This report presents information and tips on the design and construction of steel parking structures
including information related to seismic behavior and design of such parking structures. Steel parking
structures have been used throughout the world particularly in seismic regions such as Japan. This report is
prepared to provide the state of the art knowledge of design of steel parking structures in general and
particularly design of such structures in seismic regions. First, a summary of issues related to design of
parking structures is provided. Then issues specific to design of steel parking structures such as design of
deck systems, painting information, and fire resistance are discussed. Finally, notes on seismic design of
steel parking structure are presented.

First Printing, September 2001,
COPYRIGHT 2001 by Lanny J. Flynn and Abolhassan Astaneh-Asl. All rights reserved.
_____________________________________________________________________________
Lanny J. Flynn, P.E., S.E., Principal and Vice President of Design Build Services, Chalker Putnam Collins
& Scott, 950 Pacific Avenue, Suite 200, Tacoma, WA 98402,
Phone: (253) 383-2797, Fax: (253) 383-1557,
E-mail: [email protected] Web page: www.cpcsengineers.com
______________________________________________________________________________

Abolhassan Astaneh-Asl, Ph.D., P.E., Professor, 781 Davis Hall, University of California, Berkeley, CA
94720-1710,
Phone: (510) 642-4528, Fax: (925) 946-0903,
E-mail: [email protected], Web page: www.ce.berkeley.edu/~astaneh

Disclaimer: The information presented in this publication has been prepared in accordance
with recognized engineering principles and is for general information only. While it is believed
to be accurate, this information should not be used or relied upon for any specific application
without competent professional examination and verification of its accuracy, suitability, and
applicability by a licensed professional engineer, designer or architect. The publication of the
material contained herein is not intended as a representation or warranty on the part of the
Structural Steel Educational Council or of any other person named herein, that this
information is suitable for any general or particular use or of freedom from infringement of
any patent or patents. Anyone making use of this information assumes all liability arising
from such use.

Caution must be exercised when relying upon specifications and codes developed by others
and incorporated by reference herein since such material may be modified or amended from
time to time subsequent to the printing of this document. The Structural Steel Educational
Council or the authors bears no responsibility for such material other than to refer to it and
incorporate it by reference at the time of the initial publication of this document.


2

ACKNOWLEDGMENTS

The publication of this report was made possible in part by the support of the Structural
Steel Educational Council (SSEC). The authors wish to thank all SSEC members for their
valuable comments. Particularly, special thanks are due to Fred Boettler, Jeff Eandi, Pat Hassett
and James Putkey for their valuable and detailed review comments. Chuck Whittaker, formerly of
the Skyline Steel Corporation, TRADEARBED Inc., encouraged the authors to develop this
report and provided valuable information on the design and construction of steel parking
structures and European practices. Professor Brady Williamson of UC-Berkeley provided
valuable information, publications and comments on fire-resistance. Special thanks to Andy
Johnson and John Cross of the AISC Marketing, Inc. for their valuable input on open parking
structures.
The opinions expressed in this report are solely those of the authors and do not necessarily
reflect the views of the Chalker Putnam Collins & Scott, where the first author is a Principal and
Vice President for Design-Build Services, the University of California, Berkeley where the second
author is a Professor, the Structural Steel Educational Council, the American Institute of Steel
Construction or other agencies and individuals whose names appear in this report.

This report is dedicated to the memories
of the firefighters and rescue workers who
heroically sacrificed their lives on
September 11, 2001 at the World Trade
Center to save others and to the
memories of all victims of this horrifying
act of violence against innocents.

Lanny J. Flynn and Abolhassan Astaneh-Asl


3

NOTES ON DESIGN OF STEEL
PARKING STRUCTURES-
INCLUDING SEISMIC EFFECTS

By:

LANNY J. FLYNN, P.E., S.E.
Principal
Chalker Putnam Collins & Scott, Tacoma, Washington

And

ABOLHASSAN ASTANEH-ASL, Ph.D., P.E.
Professor
Department of Civil and Environmental Engineering, University of California, Berkeley

____________________________________________

TABLE OF CONTENTS

ACKNOWLEDGMENTS / Page 2

TABLE OF CONTENTS / Page 3

CHAPTER 1. INTRODUCTION / Page 4

CHAPTER 2. PAINTING GUIDE / Page 7

CHAPTER 3. FIRE CODE REQUIREMENTS / Page 12

CHAPTER 4. SLAB DESIGN CONSIDERATIONS/ Page 14

CHAPTER 5. NOTES ON SEISMIC DESIGN / Page 24

BIBLIOGRAPHY/ Page 33

LIST OF Steel TIPS REPORTS ON THE WEB/ Page 38

ABOUT THE AUTHORS OF THIS REPORT / Page 39

4

1. Introduction

1.1. Introduction

The need for multi-story parking structures has grown considerably over the years and will
continue to grow as metropolitan densities increase. There are several key issues, which need to
be addressed in the design of multi-story parking structures. They are:

1. Site considerations, environmental and neighborhood impacts and traffic access
2. Number of parking spaces, car circulations, ramps and other architectural aspects
3. Security and safety
4. Structural aspects (particularly in highly seismic areas, seismic design aspects)
5. Cost and speed of construction
6. Life cycle cost of maintenance
7. Fire resistance and/or need for fireproofing.

The first three items in the above list, to great extent, are impacted by the decisions of
architects. Items 4 to 6 in above list, also are impacted by architectural aspects, however, these
three items are primarily impacted by the structural design and decisions made by the structural
engineers. Today, structural steel provides viable systems that address the above key issues. In
the past, a large percentage of parking structures throughout the country were designed and built
using reinforced concrete structures. However, since 1980s in many regions of the US including
seismic areas such as California, more and more steel parking structures have been designed and
built. According to Emile Troup (1989), nearly three out of every five car parks for which
contracts were awarded in 1987 in New England were steel. He attributes this increase in use of
steel structures in open parking structures to the fact that as a result of research and testing done
in 1970s the issue of fire-proofing of steel structures in car parks was put to rest and the use of
unprotected steel in parking structures was accepted (Troup, 1989). Because of extensive
research and testing of bare steel structures subjected to fire, the fire codes no longer have very
stringent requirement for fire protection of steel car parking structures. This development, along
with education and dissemination of information on viability and economy of using steel structures
in car parking, may have been instrumental in visible increase in design and construction of steel
parking structures.

5
1.2. General Aspects of Design of Parking Structures

Design of car parking involves good combination of information on not only building
design but also bridge design. Like bridges, in many cases, especially in open car parks, there are
very few non-structural elements and the car park building, as a bridge is primarily a bare
structure with minimal mechanical and non-structural elements. According to Emile Troup: In
many cases the structure the deck and frame is the car park. The concept and design of the
deck and frame will largely determine the success of the facility: its cost and its ability to perform,
relatively problem free, for the design life expectancy. Therefore, it is recommended that the
structural engineer for the car park share the lead role as building designer, in close association
with others charged with developing the optimum parking concept. The April 2001 issue of the
Modern Steel Construction magazine (MSC, 2001) featured six articles on various aspects of
steel parking structures. In almost all case studies, the prominent role of structural engineer and
the impact of structural engineering decisions on making the projects highly successful are very
clear. The reader is urged to refer to the articles for very useful information and case studies on
efficient design and construction of modern steel parking structures.

As an introduction to design concepts for parking structures, the following briefly lists the
important requirements:

Since floor loadings are relatively light, floor plans usually need large, simply framed
areas ordinarily consisting of easy-to-design structural elements.

The size and number of columns in parking structures is critical since closely spaced and
large columns quite often reduce the useful width of the traffic lanes as well as reducing
width and number of the parking spaces in a given floor. Therefore, parking structures
normally have clear spans of about 60 feet at least in one direction.

Because both framing and floors are atmospherically exposed, this exposure may create
a condition of standing water and in some areas exposure to de-icing salts. Hence,
long-term structural maintenance should be given appropriate care and consideration.
Joints in the floor decks can result in leakage, corrosion and chloride attacks. Floor
joints should be avoided if possible and if they are absolutely needed, the number of
joints in the floor deck should be kept to a minimum.

Although gasoline and other combustible elements are invariably present and thereby
suggest fire hazards, this is not the case in open deck parking structures. Tests have
determined that this building type needs no fire protection since 1) fire-spread risk is
minimal, and 2) if an incendiary situation does start it is easily accessible to fire-fighting
devices. It should be noted that very useful information on this and other items
regarding fire safety of steel structures could be found in just published book by A.
Buchanan (2001).


6
For a self-standing parking structure, foundation, architectural, and mechanical costs are
relatively minimal, and largely, structural elements are approximately two-thirds of the
total construction cost of this building type.

Low construction and life-cycle costs and speed of construction are primary objectives
of this building type.

In many urban parking structures, the architects and owners demand a quality product
that fits the upscale architecture of their surrounding areas and adjacent buildings. Since
in most parking structures, the structures are mostly exposed, the close coordination of
the architectural and structural aspects becomes a necessity to achieve an aesthetically
pleasing, structurally sound and economical parking structure.

Quite often in urban areas, the lots available for parking structures are tight in space and
have limited construction-launching space. Therefore, the structural system should lend
itself to relatively small amount of on site construction activity and more to shop pre-
fabrication.

One of the primary goals of this publication is to provide information to architects,
engineers, and owners, on the design and construction of steel frame open deck parking structures
in general and particularly in seismic regions. After introduction in Chapter 1, since in many open
deck-parking structures, the steel structure is exposed, Chapter 2 is devoted to painting issues.
Chapter 3 summarizes current fire code requirements for steel open deck parking structures.
Chapter 4 of the report discusses issues related to design, construction and maintenance of
parking floor slabs. Chapter 5 is devoted to discussion of issues related to seismic design of
parking structures and pros and cons of currently used steel structural systems when used in a
steel parking structure. A list of references is provided at the end of the report.


7

2. PAINTING GUIDE

2.1. Introduction

This chapter is intended to provide guidance for architects, engineers, owners, or
specifiers that will assist them in making proper choices in selecting a protective coating system
for the structural steel for a parking structure.

2.2. Factors That Affect Cost and Performance

When selecting a coating system, the system sought should provide maximum
performance at the lowest cost. In making the proper choice, a number of factors should be
considered:

Functional requirements;
Service life of coating and structure;
Quality of coating system;
Quality of surface preparation and application;
Maintenance program; and
Determination of coating cost.

2.2.a. Functional Requirements

In most environments, coatings are a requisite for the protection of steel from corrosion.
Usually exposed steel in parking structures is quite visible to the public; hence, maintenance of its
appearance the gloss and color retention is an important requisite.
2.2.b. Service Life of Both Coatings and Structures

One of the dependencies that influence the selection of a coating system is the length of
time the coating provides the corrosion protection and the maintenance required. With present-
day coating systems, the usual expectation for paint life is from 20 to 25 years.

8
2.2.c. Coating System Quality

As previously noted, the type of coating selected is an important factor for both its
performance and cost. Normally, the material is from 15% to 20% of the systems total cost.
Thus, merely saving a few dollars-per-gallon for lesser quality materials may not be a wise
decision.
2.2.d. Quality of Surface Preparation and Application

In virtually all systems that use high-technology coatings (e.g., ethyl silicate; zinc-rich,
epoxy-polyamide polyurethanes), their most costly portion is surface preparation. The degree of
surface preparation that is reached is a critical factor in determining ultimate performance of the
coating system. Table 2.1 summarizes methods of surface cleaning.

A recognized necessity for high-technology coating is blast cleaning. Hence, by initially
investing in a superior surface preparation, the result will usually be a lifetime increase. Usually
an SSPC-SP6 commercial blast cleaning, or an SSPC-SP10 near-white metal blast cleaning, is
recommended for use in parking structures.

The following is a brief of SSPC-SP-6 and SSPC-SP-10 blast cleaning:

a) SSPC-SP 6 Commercial Blast Cleaning

This method defines a more thorough, but not perfect, degree of blast cleaning. It is a
minimum specification that is used with coating systems of higher performance, yet less
forgiving of surface imperfections.

During cleaning, all rust, mill scale, and other detrimental matter is removed; however,
staining that resulted from previously existing rust and mill scale, is permitted on 33% of each
square inch of surface. The advantage of commercial blast cleaning lies in the lower cost for
adequate surface preparation for a majority of cases where blast cleaning is deemed
appropriate.

Note that certain paint systems (e.g., inorganic zinc-rich), may not be able to tolerate
placement over a surface that has been prepared in this manner.

b) SSPC-SP 10 Near-White Metal Blast Cleaning

While this specifications price is higher than the Commercial, it only permits staining on 5%
of each square inch of the previously described surface. Generally used, only when the
expense of this higher cleaning level is justified by the chosen paint materials, and the severity
of the anticipated service environment; Near-White Metal Blast Cleaning is frequently
specified in combination with inorganic zinc-rich coatings.


9
Unless the anticipated service environment is extremely severe unlikely in the case of
parking structures the advantage of this type of cleaning can be considered as optimum
performance achieved at 10% to 35% savings in surface preparation costs over that of SSPC-
SP 5 White Metal Blast Cleaning.

It is not anticipated that any parking structure will require the use of a surface preparation that
is more stringent than the Near-White Blast Cleaning.
2.2.e. Maintenance Program

The magnitude of maintenance expenditure and the interval between such expenditures
depends on the initial coating choice and the established type of maintenance program. A well-
established maintenance program will help create a substantial increase in the life of the initial
coating system.
2.2.f. Determining Coating Costs

To assist in making an informed decision, designers, specifiers, and owners of parking
structures, should require information on comparative costs and lifetime extents of alternative
coating systems. Shop-application coating costs are normally divided as follows: material,
surface preparation, application, inspection, and overhead. For precise estimates, individual shops
should be contacted in order to determine the costs of labor, materials, and other items for
specific coating systems.

2.3. Recommended Coating Systems for Parking Structures

1. SSPC-SP 6, 2-pack epoxy polyamide zinc-rich with high-build epoxy topcoat.
2. SSPC-SP 6, followed by moisture-cured polyurethane zinc-rich primer and Aliphatic
polyurethane acrylic topcoat.
3. SSPC-SP 10, followed by ethyl silicate inorganic zinc primer and epoxy topcoat.
4. SSPC-SP 10 / epoxy-polyamide zinc-rich, high-build epoxy topcoat.

NOTE:
1. For these systems, an Aliphatic polyester polyurethane topcoat may be substituted in order
to attain improved a) durability, b) abrasion resistance, and c) easy removal of graffiti.
2. For slip critical connections surfaces an AISC Class B surface, conforming to SSPC-PS-
12.01 can be provide by most paint manufactures.

LOW-VOC ALTERNATI VES VOC =2.8 lbs / gal (340 g / liter)

To meet future 2.8 lbs / gal VOC (340 g / liter) requirements the above-listed Systems 1
through 4, plus the alternate Aliphatic polyurethane topcoats are available for commercial use at
this level.


10
NOTE: Low-VOC versions of these coatings do not have the long service life that has
been documented for their high-VOC counterparts, therefore, the manufacturer/supplier should be
required to furnish evidence of both their field performance and application properties.

Where water-borne coating systems are required, the following can be specified:

SSPC-SP 10 followed by water-borne inorganic zinc alkali silicate primer with 100% acrylic
topcoat.

NOTE: While this system has demonstrated a good long-term service life, the
manufacturer/supplier must demonstrate the suitability of shop application properties, as well as
citing the products specific field-usage.

2.4. Specifying Coating Systems

System 1: Epoxy Polyamide

Zinc-rich epoxy primer; SSPC-Paint 20, Type II
Epoxy intermediate or topcoat: SSPC-Paint 22.
Polyurethane topcoat (optional). SSPC specification is not available for this. Request
supplier to submit laboratory and field-test data. This topcoat must consist of two-
component Aliphatic isocyanate polyurethane.

System 2: Polyurethane/Polyurethane

Moisture-cured polyurethane zinc-rich primer. SSPC-Paint 20, Type II. Request
paint supplier to submit exterior exposure test panels or service) data for at least three
years; names of the facility owners should be given to verify the performance.
Epoxy intermediate is optional.
Polyurethane topcoat.

System 3: Inorganic Zinc-Epoxy

Ethyl silicate inorganic zinc-rich primer; SPC-Paint 20, Type I.
Epoxy intermediate or topcoat.
Polyurethane topcoat (optional).

System 4: Epoxy/Epoxy

Epoxy polyamide zinc-rich primer SSPC-Paint 20, Type II. Request paint supplier to
submit exterior exposure (test panels or service) data for at least three years; names of
the facility owners should be given to verify the performance.
Epoxy intermediate or topcoat (see System 3).
Polyurethane topcoat (optional; see System 3).

11

Table 2.1

SUMMARY OF SURFACE PREPARATION SPECIFICATIONS

SSPC
Specification

SSPC-Vis 1-89
Photograph

Description

SP 1, Solvent Cleaning Removal of oil, grease, dirt, soil, salts, and contaminants
by cleaning with solvent, vapor, alkali, emulsion, or
steam.

SP 2, Hand Tool Cleaning Removal of loose rust, loose mill scale, and loose paint to
degree specified, by hand chipping, scraping, sanding,
wire brushing, and grinding.

SP 5, White Metal Blast
Cleaning
A, B, C, D, SP 5 Removal of all visible rust, mill scale, paint, and foreign
matter by blast cleaning by wheel or nozzle (dry or wet)
using sand, grit, or shot. (For very corrosive atmospheres
where high cost of cleaning is warranted.)

SP6, Commercial Blast
Cleaning
B, C, D, SP 6 Blast cleaning until at least two-thirds of the surface area
is free of all visible residues. (For rather severe
conditions of exposure.)

SP 7, Brush-Off Blast Cleaning B, C, D, SP 7 Blast Cleaning of all except tightly adhering residues of
mill scale, rust, and coatings, exposing numerous evenly
distributed flecks of underlying metal.

SP 8, Pickling Complete removal of rust and mill scale by acid pickling,
duplex pickling, or electrolytic pickling.

SP 10, Near-White Blast
Cleaning
B, C, D, SP 10 Blast cleaning nearly to White Metal Cleanliness, until at
least 95% of the surface area is free of all visible residues.
(For high humidity, chemical atmosphere, marine, or
other corrosive environments.)

SP-11-89T, Power Tool
Cleaning to Bare Metal
Complete removal of all rust, scale, and paint by power
tools, with resultant surface profile.

Vis 1-89, Visual Standard for
Abrasive Blast Cleaned Steel
Standard reference photographs; optional supplement to
SSPC Surface Preparation Specification SSPC-SP 5, 6, 7,
and 10.

Vis 2, Standard Method of
Evaluating Degree of Rusting on
Painted Steel Surfaces
A geometric numerical scale for evaluating degree of
rusting of painted steel, illustrated by color photographs
and black and white dot diagrams.


12

3. FIRE CODE
REQUIREMENTS

3.1. Fire Code Requirements

Recent years have witnessed appearance of an increasing number of open-deck, multi-level
parking structures that have unprotected steel framing. This growth of unprotected steel framed
open-deck parking structures is in recognition that fire severity in this type of structure is actually
quite low.

The American Iron and Steel institute (AISI) and the Municipal Parking Congress
conducted research in order to provide a new, statistically reliable basis for evaluating the fire
protection requirements and the insurance rates of parking structures. To accurately document an
actual fire severity and its effects on parking decks, an intensive study was conducted in Scranton,
PA on 15 October 1972. In this study, AISI sponsored a full-scale fire test using a newly erected,
multi-level parking structure. This full-scale fire test was conducted while the facility was in
normal daytime operation.

The principal objective of AISI was to make a comprehensive and totally objective
determination of the effects of a burning auto on bare structural steel framing. During this 50
minute test period the maximum recorded steel temperature on a steel girder, located directly
above the burning auto, was 440 F. Within the same test period, this girder showed a maximum
deflection of 1 5/8 and a maximum elongation of 1/8. After the completion of the test, both the
deflection and elongation readings returned to zero.

All the results and findings of this full-scale test were documented in detail. The resulting
data confirmed the fact that bare-steel framing in open-deck parking structures faces little danger
from automobile fires.

Model building codes reflect these carefully observed findings by allowing the use of
structural steel without fireproofing or specific fire projection assemblies for open-deck parking
structures. Table 3.1 shows a short summary of the requirements of two model-building codes
commonly used along the west coast. The allowable height, number of tiers and area are a

13
function of many factors. The applicable building code should be consulted for detailed
requirements.
Special considerations should be given to open parking structure applications where the
building code requirements mandate that some or all of the structural steel have fire-resistance
ratings. Most manufacturers of steel fireproofing materials have products for this application.
Following the specific recommendations of the manufacturers is essential for determining the
proper product application. Factors that should be considered when selecting a fireproofing
system are:

a) Climate/Exposure: The very nature of an open parking structure requires the fire-
protection material to have a measured resistance to environmental effects. This would
include: freeze-thaw cycling, direct rain exposure, corrosion protection and wind erosion.
Most manufacturers have specific tests that document their products performance under
these conditions. Additionally, UL fire testing often lists particular products for exterior
exposure applications.

b) Durability: In addition to environmental exposure, open parking structure applications are
susceptible to the normal activities of the buildings use. This may involve such things as
human contact and vehicular impact. Typically, manufacturers can provide density,
hardness, and impact-resistance testing, in order to verify their products ability to stay in
place when abused.

Table 3.1

HEIGHT AND AREA LIMITS IN MODEL BUILDING CODES FOR OPEN DECK PARKING
STRUCTURES OF UNPROTECTED NONCOMBUSTIBLE CONSTRUCTION

Code

Edition

Number of Tiers and Area per Tier Allowed by Codes
IBC International
Building Code
2000 Eight Tiers @ 50,000 ft
2
without an automatic sprinkler system. Additional
tiers and area per tier are allowed by the IBC if certain provisions are met. For
example, a Type II-B parking structure with all sides open may be unlimited
in area when the height does not exceed 75 ft. See Note (a) below.
ICBO Uniform
Building code
1997 Eight Tiers @ 30,000 ft
2
without an automatic sprinkler system. Additional
tiers and area per tier are allowed by the UBC if certain provisions are met.
For example, a Type II-N parking structure with all sides open may be
unlimited in area when the height does not exceed 75 ft. See Note 9a) below.
(a) For more information on these code provisions and their proper use, the reader needs to refer to the actual code
((ICC, 2000) and (ICBO, 1997).


14

4. SLAB DESIGN
CONSIDERATIONS

4.1. Introduction

Structural steel is the basic framing structure material for many open deck-parking
buildings. Concrete is usually used as structural floor material. Many concrete floors supported
by both structural steel and concrete frames in the past often have required either complete
removal or extensive repair. Therefore, special attention needs to be paid to design and
construction of floor system in an open parking structure. The floors in an open parking structure
are expected to be subjected to wet and corrosive environment. Unlike closed human occupancy
floors where minor hair cracks in the floor are normally tolerated, in car parking, such cracks can
be the source of leakage as well as corrosion. In recent years, through development and use of
epoxy coated rebars, galvanized composite steel decks, post-tensioned concrete floors and special
deck coating, the life expectancy of parking floors have been extended significantly and the
maintenance problems have been reduced to normal expected maintenance. In addition, unlike
building structures where flat floors are most desirable, in a parking structure, quite often, the
floors need to have slope to rapidly drain the water.

Today, one of the most economical floor systems for steel parking structures is typical
cast-in-place composite steel deck/concrete slab system connected to the floor steel beams and
girders with shear studs. The resulting floor makes floor beams and girders composite members as
well. Pre-cast or cast in place concrete slabs have also been used for parking floors. In case of
cast in place slabs, post-tensioning of slab has been used (Monroe and Baum, 2001) to have a
relatively large joint-free and crack free floor avoiding leakage problems that can occur in pre-cast
concrete floors. To ensure long term durability of reinforced concrete decks and to avoid
corrosion, the use of epoxy-coated.

In seismic areas, the use of steel deck and concrete slab is preferred over cast-in-place or
precast systems. Recent tests by Astaneh-Asl et al. (2001) has indicated that the steel deck
provides a very ductile secondary system to carry the floor loads, in a Catenary manner, even if a
column has collapsed or removed.


15
In recently completed parking, structures in New York and New Jersey designers Englot
and Davidson (2001) have used precast/prestressed double tees for flat portions of the structure.
The designers indicate that the lighter weight double tee floor construction along with steel
framing system has resulted in overall lighter structure, longer spans, less number of columns and
foundations and less seismic forces. The reduction of seismic forces was very important for this
structure since the site had a thick layer of seismically liquefiable organic soil. In addition, the
double tee floor system has resulted in less overall depth of floors providing generous headroom
of 8-2 to the bottom of the pre-cast concrete double tee beams. For ramps, which were double
helical, external ramps, high-strength (7000 psi in 28 days) cast in place reinforced concrete slabs
were used Englot and Davidson (2001). Fly ash, silica fume, a corrosion inhibitor and a high
range water reducer were included in the mix to ensure long life for the ramps.

The concrete in the slab can be normal weight or lightweight concrete. From seismic point
of view, of course lightweight concrete is preferred. This is because of lightweight concretes
lower weight (mass) resulting in smaller seismic forces to be dealt with in design. Performance of
lightweight concrete and structural members using lightweight concrete has been studied and
tested in recent years and significant information on this subject is currently available in ACI and
other publications.

One of the early applications of lightweight concrete deck was in the upper deck of the
San Francisco Oakland Bay Bridge in 1936. The lightweight slab has been under heavy traffic
(currently with a daily traffic of about 50,000 trucks and 250,000 cars). After the 1989 Loma
Prieta quake, as part of comprehensive studies of the Bay Bridge, (Astaneh-Asl et al., 1990-1992)
the second author led a study to assess the condition of the light weight concrete deck (Astaneh-
Asl and Mori, 1990). The studies indicated that the lightweight concrete deck has performed well
and the concrete did not show any sign of deterioration, even though it was exposed to over-
water environment and heavy traffic.

4.2. Slab Construction for Open Deck Parking Structures

4.2.a. Slab Design Characteristics

Typically floors in steel-frame, open deck parking structures consist of reinforced concrete
slabs. In recent years, a number of modern parking structures have been constructed in California
and other areas using typical steel deck concrete slab system. A parking structure deck is less
dependent on its supporting structural frame, than on the slab materials and the construction
details. Over economizing floor slabs should not be a controlling design factor in parking
structures. The designer should remember that the parking slab is the structural element that is
most frequently subjected to wear and most often exposed to harsh atmospheric elements.
Virtually every square foot of parking structure surface is a source of income and owners are
usually unhappy when they have to shut down for slab repairs.

Durability is a primary design consideration for parking structures. Surveys have revealed
that structural slabs and their topping deteriorate a good deal prior to their supporting structural

16
frames therefore it is strongly recommended that the designer provide sufficient attention in the
design and detailing of the parking slab decks. This chapter describes current practice that will
help obtain satisfactory concrete slabs for decks that are supported on steel frames.

4.2.b. Service Loads

1. Dead Loads

Normal weight stone concrete with a density of about 150 pcf, is the most common and is
often recommended for its high level of durability. Lightweight aggregate concrete, with a
density of about 110 pcf, have also been used successfully in the past and if properly designed and
detailed can meet the durability requirements for parking decks. More information on design and
details are provided in Section 4.3.

2. Live Loads

Most national building codes specify a uniformly distributed, minimum live load of 50
pounds per square foot (psf), or a minimum concentrated load of 2,000 pounds that is placed
anywhere it will produce greater stress.

3. Snow and Other Roof Loads

If the top parking deck is not covered by a roof structure snow loading should be
accounted in applicable geographic locations. Snow loads should include drifts along parapet
walls, sides of exposed autos, etc. to the extent required by the local code. Deliberate snow piling
to clear parking spaces is a common practice that should be investigated. Such snow loads may
be considerably higher than those in the uniformly applied code requirements.

4. Internally-Induced Stresses

Volume changes that result from a) thermal, b) shrinkage, and c) creep effects can cause
indirect forms of service loads on all slabs. These loads must be accounted for in the design of
any rigidly attached steel frame element. In addition, if a post-tensioned slab system is used effects
from elastic shortening of the slab must be investigated.

Internal stresses are reduced and better managed by utilization of appropriately spaced
expansion joints, and construction joints such as pour strips. However, some degree of stress is
always present.

5. Loading Variations

Usually, concrete slab design and construction is regulated by the ACI Code. Pre-stressed
concrete must be designed by following the procedures required by the ACI Codes strength
design method (formally referred to as ultimate strength design); the working-stress design
method of ACI Codes Appendix A, may only be used for non-prestressed concrete.

17

4.2.c. Concrete Qualities

1. Basic Materials

Concretes quality is dependent on four aspects: a) its materials, b) mixture, c) placement
procedures and d) the curing process. In building a slab system, if care is used in all steps except
the use of high-quality concrete, the parking deck slab will present a problem. It is not costly to
achieve concrete of high quality. It does however, require effort since several choices and project
participants are involved in its preparation and placement.

The first step in the design of the slabs is to specify the quality of the requisite materials
that make up the concrete.

a) Stone aggregate concrete of normal weight is a desirable material for parking
structures in consideration for their weather exposure however, lightweight aggregate
can be used.

b) The ACI Code (Reference 6) specifies acceptable basic materials for concrete, and
refers to its Commentary, the Code of Practice, and the national standards of the
American Society for Testing and materials (ASTM).

ACI Code Chapter 3 lists materials, Chapter 4 stipulates durability requirements, and
Chapter 5 specifies the means that will assure concretes proper quality.

In order to obtain the correct concrete for every planned project, using these code
provisions is a prerequisite.

2. Externally-Applied Chlorides

The nations highways and city streets are often kept free from accumulation of snow and
ice by extensive use of chemicals such as de-icing salts. Chlorides can readily penetrate and
damage the contacting concrete structure if both concrete and its steel reinforcement are
improperly designed and constructed.

Parking structures that do not employ chlorides on interior floors are still subjected to
roadway salts by autos entering for parking after being driven on chloride-treated roads.
Therefore, parking structure floors that are not directly exposed to the weather are also subjected
to the problem of roadway salts.

When externally applied chlorides are placed in direct contact with the slab, three added
constraints are required by the ACI Codes:

A maximum water/cement ratio is stipulated

18

A minimum entrained-air content is required

A minimum clear, concrete cover on reinforcing is specified.

3. Permeability Reduction by Minimizing Water-Cement Ratio

Concrete should be made as impermeable as is possible when the parking structures are
frequently exposed to de-icing salts. Consideration should be given to the ratio between water
and the cement materials of the concrete. This should be decreased as much as practicable. The
ACI Code stipulates a maximum water-cement (w/c) ratio for different conditions and exposures.

4. Admixtures with Normal Concrete

Concrete ingredients that are beyond the basic aggregates cement and water are
classified as admixtures that require particular formulation under the supervision and approval of
the structures design engineer.

Some admixtures can be very beneficial; or are mandatory for certain structures to meet
the ACI Code. Others, even though offering convenience during construction, can be harmful to
the concrete in the long run, and are either prohibited or not recommended. An admixture should
be avoided unless it serves a specific purpose in a particular structure. The indiscriminate use of
admixtures is not advisable. .

5. Air-Entraining Admixtures

A simple description of air entrainment is the process of chemical capture and maintenance
of microscopic air bubbles within a fluid concrete mixture, as well as after the concrete has set.
These bubbles behave as tiny pressure relief valves in the process of carrying out their major
functions:

a) Enabling the concrete to help withstand freezing-thawing cycles
b) Enabling the concrete to help withstand scaling action of de-icing salts on its surface.

Air entrainment is recommended for all exposures of parking structure slabs since it
improves both workability and surface finish.

For areas subjected to freeze-thawing conditions, and/or where de-icing salts may reach
the slab, the ACI Code (Section 4.1.1) requires air-entrainment.

Air-entrainment is required in varied amounts ranging from 4 to 7%; a particular
amount is given for each of the exposures listed in ACI Code Table 4.2.1 and described in the
ACI Code Commentary.


19
Air entrainment makes a significant lessening of bleed water. As a result, concrete
finishing can therefore be started sooner, and the long-term quality of the surface is improved.

6. Water-Reducing Admixtures

By reducing the water content necessary for the production of a workable concrete, such
admixtures can radically alter both fresh and the post-hardening properties of the concrete. The
following are achievable goals that may be gained from this admixture type:

Reduction of water/cement ratio

Increase slump and workability

Reduction of water amount needed to produce a particular slump

Normal water-reducing admixtures provide a 5 to 10% water reduction. A high-range
extension of this admixture type, called a super-plasticizer, can reduce water content by 12 to
30%, producing a flowing concrete, which remains cohesive. When using a super-plasticizer,
another means of specifying a water/cement ratio standard of quality will be necessary.

The possible advantage of using any water-reducing admixture for parking structure slabs
is the ability to reduce the concretes water/cement ratio and permeability while still providing a
workable mixture. Careful vibration and consolidation of the concrete particularly around the
reinforcing steel is required and should not be neglected when any water reducer is used.

7. Mineral Admixtures

Over a period of many years, several classes of finely divided admixtures have been used
in producing concrete. They consist of blast furnace slags and other minerals that have
cementitious properties and pozzolans (siliceous materials) that can improve certain concrete
qualities and become cementitious. Fly ash, a by-product of coal-burning power plants, is an
example of the latter effect.

8. Microsilica Admixture

Microsilica is a brief substitute as a name for condensed silica fume, a by-product of alloy
steel production. The particles of this element are ultra-fine in size and are marketed in either
powder or liquid form.

Microsilica benefits parking structure slabs as compared to ordinary concrete by reducing
the permeability and reducing the chloride intrusion.

Fresh concrete that contains microsilica, will behave in a manner quite different from that
of ordinary concrete. In adding microsilica, contractors should be given notice for radical changes

20
in normal mixing, slump, placing and finishing habits, and the need for strict adherence to the
microsilicia manufacturers recommendations for handling this type of concrete.

In order to place a microsilica-enhanced concrete slab, it is usually needed to utilize a
compatible super-plasticizer, i.e., a high-range water-reducing admixture.

9. Corrosion-Inhibiting Admixtures (Inhibitors)

Because of its very high, natural alkalinity, concrete imparts a protective oxide film on the
embedded steel. This film is penetrated or broken down by certain aggressive chemicals such as
chlorides which can start steel corrosion.

Admixtures such as calcium nitrite can stabilize and reinforce the protective film when it is
attacked. The amount inhibitor will vary directly with the amount of attacking chloride present in
the concrete. Steel corrosion will begin when the level of chloride exceeds the inhibitors ability
to maintain the film. These admixtures will delay both the start and rate of steel corrosion and are
effective if the chloride intrusion does not increase more rapidly than originally estimated.

10. Placing, Finishing, and Curing Concrete Slabs

In order to attain quality concrete at the job site the following is a listing of dos and donts:

a) Use the lowest possible slump consistent with conditions of the given job; plan for a
water-reducing admixture if needed for maintaining the specified low water/cement ratio.

b) Do not add any water in the field beyond the design mix dosage. Reject watered-down
mixes.

c) Despite the use of plasticizers, mechanically consolidate the concrete particularly around
all embedded materials. Avoid excessive vibration, which among other effects can cause
segregation of aggregates, surfacing of water and the reduction of air entrainment.

d) Make certain that the concrete is well compacted under the top layer of reinforcing steel
to avoid steel settlement during the concretes hardening. Concrete placed over settled
bars is weak and can crack as a result.

e) When using a microsilica admixture, be certain that you have learned the proper
techniques for handling this concrete type.

f) Do not begin trowelling the concrete surface until the bleed water has evaporated. A high
water/cement ratio at the surface with less durable concrete can result.

g) If possible, use wet curing. A curing compound should not be employed unless there are
some very good supportive reasons.


21

4.3. Slab Joint Details

Open deck parking structures are more directly affected by weather-temperature changes
than any other occupied building type. Since the parking structure has no roof insulation, and no
exterior wall, its entirety is subjected to the full range of exterior temperatures.

Expansion joints should be built into parking structures in order to accommodate the
thermal changes. Control and construction joints may also be needed in slabs between the
expansion joints in order to control either cracking or the size of a days work.

All of the joints in the slab should be designed to withstand the effects of vehicular traffic.
Since the parking structure is not subjected to highway or bridge stresses, its joints can be
designed simpler and more economical. Unless completely drained, every slab joint of this
structure should be watertight, and capable of long-term maintenance.

Requirements for slab expansion joints are related to volume change therefore the shorter
the distance between slab joints, the less movement or intermediate cracking will occur. Weighed
against frequent joints is the cost to properly build and maintain them.

1. Joint Types

Every parking structure requires three types of slab joints:

a) Expansion Joint

This joint will adjust to temperature changes that surround the building throughout its
existence. Expansion joints accommodate the overall structures expansion and contraction
and divide it into distinct parts; sometimes theyre called isolation joints.

In locating expansion joints, the designer must recognize the fact that every parking structure
has a set of site conditions, a shape and size that are unique. If the structure is permitted to
move at points of natural change and reasonable intervals along its length, the design is
adequate.

Because expansion joints are constantly subjected to wide movement, they should be fitted
with a prefabricated assembly that is regularly produced for this special purpose. Many
manufacturers who produce several product types are available.

b) The Control Joint

It is needed prior to the beginning of the concrete slab shrinkage, when the slab assumes its
deflected shape under load, or moves over the particular structural frame that supports it.


22
Control joints should be located at places where cracking strains can build up. Only a shallow
slot in a slab surface is usually needed to direct concrete shrinkage to manageable places. The
cuts must be made prior to shrinkage in the concrete.

c) The Construction Joint

It is needed at the completion of a given days concrete slab pour; or provides a desirable
post-tensioning interval. This type of joint is where a great number of leaks typically occur
and is often not given much attention by the designers.

2. Joint Sealants

The basic ingredients of most modern sealants are urethanes. Only high-quality products
capable of withstanding traffic abuse should be considered. Every component, including the
backing rod should have a long successful history in similar joints, be non-absorbent of moisture
and should not be attacked by de-icing chemicals or ultra-violet rays.

The most important requisite for the particular products that are to be used is their
preliminary selection prior to pouring the concrete so that the right-shape groove for sealing
materials can be left in the concrete. Unfortunately, the sealant often fails because it was placed in
a joint that had the wrong shape. The important bond-breaker at the joints bottom should not be
overlooked. The sealant manufacturer is a proper source for advice in joint design, and to supply
the appropriate requirement data for contact drawings.

4.4. Drainage System

Concrete is not a waterproof or watertight material. While the concrete permeability can
be reduced by quality control and admixtures, it cannot be eliminated. When water is allowed to
lie on concrete surfaces it will eventually find minute, virtually invisible cracks that exist in all
concrete.

As part of their drainage systems, parking deck slabs should have a minimum specified
slope all over, including relatively flat, transition areas. It is essential that all levels have a well-
designed inlet and piping plan for carrying off the water from a low point.

Good drainage reduces concentrations of de-icing chemicals that can attack reinforcing
steel within the slab. Good drainage will increase the useful life of a parking structure.

4.5. Surface Sealers

Sealers are liquids that are applied to the concrete surface for the purpose of either curing
or resisting water penetration. Many of the available sealers are suitable for parking structures,
but a number are not. None of them should be considered capable of producing complete

23
permanent waterproofing. They cannot act as a substitute for properly constructed systems of
both jointing and drainage.

Since surface sealers are relatively clear and hardly noticeable liquids, their maintenance is
often overlooked. To remain effective, sealers require periodic re-application every few years, in
heavy traffic areas.

1. Types of Sealers

Two groups of synthetic sealers, particularly suitable for parking structure decks, are
classified by the amount of penetration and surface film they can provide. As long as they are
chemically compatible, both may be used in the same building and tailored to exposure.

a) Deep Penetration, No Appreciable Film

These chemicals react with the concretes cementitious elements, and thereby cause the
repelling of water. Compatibility with concrete aggregates should be verified. Silanes and
siloxanes are the basic chemicals of these penetrants; while they are more expensive, they
provide good performance.

A newer class of water based penetrating sealers is now available containing
fluorocarbons. These provide oil and grease resistance in addition to water resistance
(Adams, 2001).

b) Slight Penetration; Continuous, Visible Surface Film

This group as compared to the Deep Penetration, No Appreciable Film, is more
extensively harmed by traffic wear and natural deterioration. While it is less expensive, its
life is shorter. In these coatings, the basic chemicals are urethanes, epoxies, acrylics, and
other polymer resin blends.

These sealers deposit a durable clear or pigmented film on the surface. It acts as a physical
barrier to resist water, oils and even acid. The film may change the surface appearance and
will have to be resealed occasionally due to traffic abrasion. This type of sealer may be
acrylic, urethane, epoxy or a blend of two or more resins (Adams, 2001).

Sealer manufacturers use-instructions should be carefully followed for both preparation
and application. Surface preparation is very important; even a new concrete slab must be clean
and sound prior to application of a sealer. A test patch using the selected sealer should be done to
determine suitability before beginning a large project (Adams, 2001)


24

5. NOTES ON
SEISMIC DESIGN

5.1. Seismic Performance Criteria for Parking Structures

Current seismic design of typical residential or office buildings strives to satisfy the
following performance criteria:

1. The building should survive small and moderate earthquakes that occur frequently with
minimal damage that can easily be repaired. In addition, in recent years, in many cases,
seismic design is done such that the building remains functional and occupied after a small
or moderate earthquake, and;

2. The building should survive major earthquake without collapse, loss of life or major
injuries.

Application of the above performance criteria is quite justified for modern residential and
office buildings. However, there are a number of differences between a parking structure and a
typical residential or office buildings that may warrant a slightly different seismic performance
criteria for parking structures. The main differences between a parking structure and a typical
residential or office building are:

Unlike residential and office buildings where considerable non-structural elements such
as partitions, walls, claddings and windows exist, in a parking structure, usually there
are very limited amount of non-structural elements and the bulk of the building is
primarily a bare structure.

Unlike residential and office buildings, mechanical equipment and lifelines in a parking
structure are very limited.

Unlike residential and office buildings where the structure and particularly connections
are usually covered by fireproofing and non-structural elements, in parking structures,
almost all structure is exposed and any damage can easily be detected and repaired.

25
Considering above differences, it appears that there is a need for a different drift limitations
for parking structures than the limitations currently prescribed in building seismic design codes
International Building Code (ICC, 2000), Uniform Building Code (ICBO, 1997), and SEAONC
Blue Book (SEAOC, 1999). It appears that the drift limitations in current codes have been
established to prevent excessive damage to non-structural elements and mechanical systems of a
typical residential or office buildings. Of course, excessive drifts can result in increased P-delta
effects and considerable reduction in load carrying capacity of columns. However, such excessive
drifts, in the order of a few percentage of floor height are well beyond the current code drift
limitations. It seems that parking structures are facilities that are closer to a bridge than a
residential or office building. In both parking structures and bridges, cars are the main load and
main occupier and in both cases, the facility is almost a bare structure with minimal non-structural
and mechanical components. Therefore, it seems reasonable that a more realistic limit for drift
limitations of parking structures be established. It seems that such a drift limit should be related to
structural performance and not preventing damage to non-structural elements.

Based on above discussion, for open parking structures, where there are very few non-
structural and mechanical elements, the following seismic performance criteria are suggested. The
main difference between this proposed criteria for parking structures and the criteria presented for
buildings at the beginning of this section is in the first criterion regarding serviceability and
damageability during small and moderate earthquakes. The proposed criteria is:

1. The parking structure should survive small and moderate earthquakes that occur frequently
with minimal damage to its structure that can easily be repaired. In addition, the structure
needs to return to its plumb position, and;

2. The parking structure should survive major earthquake without collapse, loss of life or
major injuries.

5.2. Gravity and Seismic Loads for Parking Structures

The information in this section is based on current International Building Code (ICC,
2000). For actual loading of a parking structure for design purposes, the reader needs to refer to
the actual governing code.

The dead load of a parking structure is established the same way as any other structure
and current codes do not have any provisions specific to parking structures. However, since in a
parking structure, the bulk of dead load is due to the weight of structural elements and the
dimensions of these elements usually involve less uncertainty than the non-structural elements, it
appears that a dead load combination factor of lower than 1.2 (the current code value) may be
justified for parking structure. Until further research on this item is conducted, the load factor of
1.2 as specified by current codes should be used.

The live load specified by current codes for parking structures (i.e. parking structures with
passenger cars only) is 50 pounds per square feet in IBC-2000 (ICC, 2000) and UBC (1997). The

26
IBC-2000 also specifies that: Floors in garages or portions of building used for the storage of
motor vehicles shall be designed for the uniformly distributed loads of Table 1607.1 (which is 50
psf for parking structures) or the following concentrated load: (1) for passenger cars
accommodating not more than nine passengers, 2,000 pounds acting on an area of 20 square
inches; (2) mechanical parking structures without slab or deck, passenger cars only, 1,500 pounds
per wheel.

Section 1607.9.1.2 of the IBC-2000 specifies live load reduction for passenger car
garages. It states: The live loads shall not be reduced in passenger car garages except the live
loads for members supporting two or more floors are permitted to be reduced by a maximum of
20 percent, but the live load shall not be less than L as calculated in Section 1607.9.1 (ICC,
2000a. For more information, the reader is referred to the actual code (ICC, 2000).

In seismic design and in establishing total dead load of the building, W, to be used to
calculate base shear, according to IBC-2000, (ICC, 2000) floor live load in public garages and
open parking structures need not be included.

5.3. Lateral Load Resisting Systems for Parking

Common lateral load resisting systems used in steel structures today are:

a. Concentrically braced frames
b. Eccentrically braced frames
c. Moment frames (Fully Restrained, FR)
d. Semi-rigid frames (Partially Restrained, PR)
e. Steel shear walls
f. Dual systems combining moment frames with either braced frames or shear walls
g. Composite (steel and reinforced concrete) systems

All of the above structural systems can be used in parking structures with some being
more economical than others are park. In general, in order to achieve better economy in steel
parking structures the following basic principles suggested 30 years ago in a US Steel publication
seems still valid:

Employ shop labor and prefabrication as much a possible.
Employ standard AISC connections when possible.
Avoid full penetration welding especially in the field.
Use braced frames, as opposed to achieving lateral stability through moment
connections
(Excerpt from Ref. (USS, 1971))

In the following sections, some advantages and disadvantages of using each structural
system in a parking structure are discussed.

27

5.3.a. Concentrically braced frames

The concentrically braced steel frames are one of the most economical, if not the most
economical, lateral load resisting systems. In most parking structures, it is possible to use this
system in at least one direction. The X-braces might be more economical than the V-brace or
Chevron braces that sometimes are used in buildings to accommodate door and window openings.
The beam-to-column connections in this system can be the more common shear tabs (Astaneh-Asl
et al, 1989) or any other standard AISC shear connections such as seat angles, stiffened seat or
web angles. The bracing members can be single angle, double angles, single channels, double
channels, tubes, pipes or wide flanges. The end connections of bracing members are usually gusset
plates.

Current design codes (AISC, 1997), include information on two types of eccentrically
braced frames: (a) Special Eccentrically Braced Frames and (b) Ordinary Eccentrically Braced
Frames. More design-oriented information on concentrically braced frames and their connections
can be found in (LA-AISC-PMC and Flynn, 2000) and (Astaneh-Asl, 1998).

5.3.b. Eccentrically braced frames

Eccentrically braced frames can also be used in steel parking structures. However,
compared to concentrically braced frames, eccentrically braced frames may not be as economical
for this application.

5.3.c. Moment frames

Steel moment frames, especially frames with field full penetration welds, can be quite
costly compared to concentrically braced frames. In many applications, because of interference
with the ramps or driving lanes, braced frames cannot be used. In these cases, better economy of
design can be achieved if at early stages of design, the fabricator is also involved in deciding the
type and details of the moment connections. One of the best sources of information on seismic
design of steel moment frames is the SAC steel Joint Venture publication FEMA-350 and four
other reports in the series FEMA 351 through FEMA 354 (FEMA, 2000).

In addition to moment connections discussed in FEMA documents (FEMA, 2000), Collin
and Putkey (1999) have proposed a welded connection that minimizes residual stresses. Details of
this connection are discussed in the Steel TIPS report (Collin and Putkey, 1999). Studies of
failures of field-welded connections during Northridge earthquake have indicated that relatively
large residual stresses in full-penetration field welds may have been one of the parameters
contributing to fracture of welds. According to Putkey (2001), Improper root openings were a
probable cause of weld failures in the Northridge earthquake. The connection proposed by
Collin and Putkey (1999) appears to be a clever solution to avoid residual stresses. Collin and

28
Putkey state that: Our suggested connection avoids direct beam flange to column flange welds
and restrained cover plate to column flange welds. It eliminates medium or high residual stress
that occurs when welding these joints to a column flange because connection restraint or
member restraint is not present (Collin and Putkey, 1999). Using the connection proposed by
Collin and Putkey one culprit, the residual, stresses can be effectively kept out.

Shop-welded field-bolted column-tree moment frames can also be very efficient system for
parking structures. In a column-tree system, a short length of girder is welded to the column in
the shop. During erection of the frame, after columns are erected, the girders are placed between
the column short girders and are spliced to them, Figure 5.1. By placing the splice point near the
point of inflection of beam under gravity load, the girder splice can be designed to carry shear due
to gravity combined with shear and moment due to seismic load. As a result, the splice will be
reasonably small. In fact, one can take advantage of this splice and design moment capacity of the
splice less than the capacity of the girder and the beam-to-column welded connection. By doing
so one can make the splice to be the weakest link in the bending moment diagram with the highest
demand to capacity ratio as shown in Figure 5.2 forcing plastic hinge in the girder to form in the
splice. The plastic hinge formed in the splice will act as a fuse and protect the welded connection
at the face of column from fracture as well as the girder flanges from local buckling. More
information on seismic behavior and design of column-tree moment frames can be found in
Astaneh-Asl (1997). In designing steel moment frames, bolted moment connections can also be
very economical.

Figure 5.1. Column-Tree Moment Resisting Frame
Since the 1994 Northridge earthquake and the increased cost of making full penetration
filed-welded moment connections more ductile, the bolted moment connections such as top &
Column-Tree
Moment Frame
Field-bolted Splices
Brace
Frame

29
bottom bolted plate connections have become quite economical and have been used in many
structures in various seismic regions of the West Coast. More information on seismic design of
bolted moment frames can be found in (Astaneh-Asl, 1995), (Astaneh-Asl, 1998) and FEMA 350
Report (FEMA, 2000). The fact that in a parking structure most connections are exposed makes
bolted connections more desirable since in the aftermath of a major earthquake all one has to do is
inspect the connection bolts and if any bolt was found sheared off or loosened simply replace
and/or tighten them.

Figure 5.2. Suggested Details for Column-Tree Moment Frame (Astaneh-Asl, 1997)

Short Girder Stub

M

C.P.
T+B
Erection Clip

M

C.P.

M

C.P.

T+B
SHOP AND FIELD WELDED

One or Two Web Splice Plate(s)

Flange Splice Plate

One or Two Rows of

Short Girder Stub

Shop Fillet Welds

High Strength Bolts

Field Welds (Fillet Weld)

Shop Fllet Welds

Field Welds (Fillet Weld)

on the Sides

Field Welds

M

C.P.

T+B
SHOP AND FIELD WELDED

(c)

(d)

(Full Penetration Shop Weld)

SHOP WELDED AND FIELD BOLTED
Web Splice Plate

H.S. Bolts

Short Girder Stub

Shim as Required

to Adjust Elevation

M

C.P.
T+B
(b)

Shims

H.S. Field Bolts

Flange Splice Plate

(Full Penetration Shop Weld

H.S. Field Bolts

One or Two Web Plates

Flange Splice Plate

One or Two Rows of
High-Strength Bolts
In Slotted or Round Holes
Short Girder Stub


Shop Fillet Welds

SHOP-WELDED AND FIELD-BOLTED
Shim as Required
to Adjust Elevetaion
Shop Welds
M

C.P.
T+B
(a)

Shims


30
5.3.d. Semi-rigid (Partially Restrained) Frames

Numerous analytical studies, laboratory tests and observations in the aftermath of actual
earthquakes, all have indicated that semi-rigid (PR) steel frames have the best balance of stiffness,
strength, damping and ductility to resist seismic forces with economy and efficiency. The reader
unfamiliar with seismic behavior and design of semi-rigid frames is referred to references (Nader
and Astaneh-Asl, 1989), (Nader and Astaneh-Asl, 1992) and (Astaneh-Asl, 1994).
The benefits of using semi-rigid (PR) steel frames to carry gravity and wind loads have
been recognized during the last 20 years and a number of semi-rigid steel structures have been
designed and constructed in U.S. However, even though steel semi-rigid frames are perhaps one
of the most suitable structural systems for seismic areas, they are one of the least utilized systems
to resist seismic loads. The main reason for this lack of usage may be lack of explicit seismic code
provisions for semi-rigid steel frames in current codes. Particularly, many structural engineers are
concerned about large drifts that in their opinion a semi-rigid frame might develop.

Perhaps this feeling about large drift to be developed in semi-rigid frames has its roots in
long-practiced equivalent static load given in current codes. In this method, seismic forces
(which are actually dynamic inertia forces) are applied to the structure at floor levels as static
forces. Then the structure is analyzed and member forces and story drifts are established.
Following this method, if one uses semi-rigid frame in place of similar but rigid frame, the drift of
semi-rigid frame would be larger. However, during an earthquake, the inertia forces developed in
a structure are not static forces but they are dynamic forces and their magnitude depends on
stiffness, damping, ductility, energy dissipation and dynamic properties of the structure and the
ground motion shaking it. As a result, as shown by the research and shaking table tests conducted
by Nader and Astaneh-Asl (1992) the drift and seismic forces developed in a semi-rigid frame can
be in fact less than a similar but rigid frame. In the following, a brief summary of behavior and
seismic design of semi-rigid frames is provided.

It appears that for low- and mid-rise structures and particularly for parking structures
where there are very few non-structural brittle elements, semi-rigid steel frames can be the best
and most economical lateral load-resisting system.

Steel rigid as well as semi-rigid moment frames resist seismic effects primarily by bending
and forming plastic hinges within the moment connection area. The main difference between a
rigid and semi-rigid steel moment frames is in the bending strength and rotational stiffness of the
beam-to-column connections relative to the connected beams. In rigid frames, the connections are
designed to be stronger and stiffer than the beam and are expected to remain essentially elastic
during earthquakes. In semi-rigid frames, the connections are intentionally designed to have less
bending capacity and stiffness than the connected beams so that the bulk of yielding and rotational
ductility is in the connection elements and not in the girder.

Almost all semi-rigid connections used today are shop-welded field bolted or entirely field
bolted. Rigid moment connections typically show bi-linear moment-rotation response with two
distinct regimes of behavior: (1) the initial elastic behavior and (2) the post-yielding non-linear
behavior. In steel semi-rigid connections, the moment-rotation behavior in general has four

31
distinct regions: (1) the initial elastic region, (2) the first stage of softening due to yielding or
friction slippage of the connection elements, (3) the secondary stiffening mostly due to kinematic
hardening, and (4) the final yielding. In rigid connections with a given moment capacity, it is very
difficult to control the initial rigidity which is generally very high. However, in today's common
semi-rigid connections, all parameters of behavior such as initial stiffness, secondary stiffness,
initial yield or slip moment, and final moment capacity can be controlled by choosing appropriate
connection geometry and material properties (Shen and Astaneh-Asl, 1993; Nader and Astaneh-
Asl, 1989 and 1992). Other studies, some of which are listed in the references, have shown
similar behavior.

A comparison of the seismic behavior of rigid and semi-rigid steel moment frames reveals
that the seismic forces generated in semi-rigid frames are generally less or on the same order as
forces in comparable rigid frames. The lateral displacements of semi-rigid frames are usually
slightly more than rigid frames. The decrease of forces and some increases in displacement in
semi-rigid frames is attributed to elongation of period, increase in damping, decrease of stiffness
at early stages of behavior, and the 'isolation effects' due to gap opening and closing in semi-rigid
frames. If a semi-rigid steel structure has connections with sufficient ductility, the studies done so
far, indicate that the behavior of bolted semi-rigid steel frames is superior to the behavior of
welded rigid frames.

Currently most seismic design codes permit the use of semi-rigid steel building frames.
However, the codes have very limited guidelines and provisions on how these structures should
actually be designed. In the Uniform Building Code (ICBO, 1997), the AISC seismic Provisions
(AISC, 1997) and the International Building Code (ICC, 2000) semi-rigid steel frames are
categorized as "Ordinary Moment Frames" with a response modification factor of R equal to four.
However, in the current codes, the composite partially restrained moment frames are placed in a
separate category and are assigned an R factor of 6.0 which makes composite PR (semi-rigid)
frames very competitive economically with other systems in seismic zones 1,2 and 3 and most
likely competitive for low rise parking structures in seismic zone 4 as well.

5.3.e. Steel Shear Walls

Steel shear walls are being used more and more in tall buildings. Although current US
seismic codes do not have specific provisions for steel shear walls, there is considerable
information on seismic behavior of steel shear wall buildings and their seismic design that one can
use and design safe and economical steel shear walls. A recent publication by second author
(Astaneh-Asl, 2001) summarizes the available information on behavior of steel shear walls during
actual earthquakes and in the laboratories, discusses code provisions for steel plate shear walls,
provides information on how to design these systems and present suggested steel shear wall
systems and details. It appears that steel plate shear walls, compared to braced frames, are more
economical when used in high-rise buildings. Therefore, for low-rise parking structures, braced
frames are preferred. In addition, unless shear walls are used within the perimeter frame or around
the elevator shaft, they may result in obstruction of open view that is a desirable factor in parking
structures.

32

5.3.f. Dual systems combining moment frames with either braced frames or shear walls

For low and mid-rise parking structures, it seems that there may be no need to use dual
system, which compared to braced frame systems such as concentric braced frames may not be as
economical.

5.3.g. Composite (steel and reinforced concrete) systems

Composite systems can be very economical systems in seismic areas for all structures
particularly parking structures. By using composite columns and beams, one can economically
optimize the cost and save on painting and fireproofing (for closed parking) costs. Current seismic
codes and specifications have specific provisions on seismic design of composite structures
(nternational Building Code (ICC, 2000) and AISC Seismic Provisions (AISC, 1997).)


33
________________________________________________________________________

BIBLIOGRAPHY
_________________________________________________________________________

AASHTO, (1998) LRFD Bridge Design Specifications, 2nd Edition, American Association of
State Highway and Transportation Officials, Washington D.C.

ACI (1989), Building Code Requirements for Reinforced Concrete (ACI 318-89) and
Commentary (ACI 318R-89), American Concrete Institute.

ACI, Manual of Standard Practice for Detailing Reinforced Concrete Structures, ACI 315,
American Concrete Institute.

ACI (1985), State-of-the-Art Report for Parking Structures, American Concrete Institute, p. 362-
385.

ACI, Structural Design Guide to ACI Building Code, Second Edition, Van Nostrand Reinhold
Co., p. 417.

ACI-ASCE (1989), Committee 423, Recommendations for Concrete Members Pre-stressed with
Unbonded Tendons, ACI Structural Journal, Vol. 86, No. 3, May-June 1989, pp. 301-318.

Adams, B.G., (2001), Private Communication with Authors: L. Flynn and A. Astaneh-Asl,
GlazeN Seal, Quality Products for Concrete, Tile and Masonry, Irvine, CA.

AISC (1989), Specification for Structural Steel Buildings, Allowable Stress Design and Plastic
Design, with Commentary American Institute of Steel Construction Inc., Chicago.

AISC (1989), Manual of Steel Construction, Allowable Stress Design, Ninth Edition, American
Institute of Steel Construction Inc., Chicago.

AISC (1999), Manual of Steel Construction-LRFD,3
rd
edition AmericanInstitute of Steel
Construction Inc., Chicago

AISC (1997), Seismic Provisions for Structural Steel Buildings, American Institute of Steel

AISC (1999), Load and Resistance Factor Design Specification, American Institute of Steel

ANSI (1982. (ANSI), American National Standard Minimum Design Loads for Buildings and
Other Structures A581-1982, American National Standards Institute.


34
Astaneh-Asl, A., Call, S.M., and McMullin, K.M. (1989), "Design of Single Plate Shear
Connections, Engineering Journal, AISC, Vol. 26, 1st Quarter, pp. 21-32.

Astaneh-Asl, A., Nader, M. N. and Malik, L., (1989),"Cyclic Behavior of Double Angle
Connections," Journal of Structural Engineering, ASCE, Vol. 115, No. 5.

Astaneh-Asl, A., Mori, H., (1990) "Tests of Material of San Francisco Oakland Bay Bridge,
Technical Report, Department of Civil and Environmental Engineering, Univ. of California,
Berkeley.
(This Report can be ordered by sending an e-mail to [email protected])

Astaneh-Asl, A., Nader, M. N. and Harriott, (1991) "Seismic Behavior and Design
Considerations in Semi-Rigid Frames, " Proceedings, AISC, 1991 National Steel Construction
Conference, Washington, D.C., June.

Astaneh-Asl, A., (1994), Seismic Behavior and Design of Steel Semi-Rigid Structures,
Proceedings, 1
st
International Workshop and Seminar on Behavior of Steel Structures in Seismic
Regions STESSA, Timisoira, Romania, June.

Astaneh-Asl, A., (1995), Seismic Behavior and Design of Bolted Steel Moment-Resisting
Frames, Steel TIPS Report, Structural Steel Educational Council, California, July.
(This report can be downloaded free for personal use from www.aisc.org web page.)

Astaneh-Asl, A. (1997). Seismic Design of Steel Column-tree Moment-resisting Frames, Steel
TIPS Report, Structural Steel Educational Council, California, CA, April.
(This report can be downloaded, free for personal use, from www.aisc.org web site.)

Astaneh-Asl, A. and Ravat, S. (1998). Cyclic Behavior and Seismic Design of Steel H-piles,
Report No. UCB/CEE-Steel- 98/01, Department of Civil and Environmental Engineering,
University of California, Berkeley, May.
(This Report can be ordered by sending E-mail to [email protected])

Astaneh-Asl, A., (1998), Seismic Behavior and Design of Gusset Plates, Steel TIPS Report,
Structural Steel Educational Council, California, December.

Astaneh-Asl, A., (2000), Seismic Behavior and Design of Steel Shear Walls, Steel TIPS
Report, Structural Steel Educational Council, California, November.

Astaneh-Asl, A., Jones, B. and Zhao, Y. (2001), Progressive Collapse Resistance of Steel
Building Floors, Report No. UCB/CEE-Steel-2001/03, Department of Civil and Environmental
Engineering, University of California, Berkeley, May.
(This Report can be ordered by sending an E-mail to [email protected])


35
Bakota, J.F., (1988), Parking Structure with a Post-tensioned Deck, Engineering Journal,
AISC, 3
rd
Quarter 1988, Volume 25, No. 3, p. 199.

Becker, R., (Editor), (1999), Connections for Use in Special Moment Resisting Steel Frames,
Steel TIPS Report, Structural Steel Educational Council, California, July.
(This report can be downloaded free from www.aisc.org web site.)

Burkowsky, B., and Englot, J. (1988), Analyzing Good Deck Performance on Port Authority
Bridges, Concrete International, November, p. 25.

BOCA, (1990), The Basic Building Code-1990, Building Officials & Code Administrators
International, Inc.

Clear, K.C. and Virmani, Y.P., (1983), Corrosion of Non-specification Epoxy-Coated Rebars in
Salty Concrete, Public Roads, Transportation Research Board, National Research Council, Vol.
47, No. 1, June.

Collin, A. A. and Putkey, J. J. (1999), Welded Moment Frame Connections with Minimal
Residual Stress, Steel TIPS, Structural Steel Educational Council, California, October.
(This report can be downloaded free from www.aisc.org web site.)

CRSI, Suggested Project Specifications Provisions for Epoxy-coated Reinforcing Bars,
Engineering Data, Report Number 19, Concrete Reinforcing Steel Institute.

Englot, J.M. and Dividson, R.I. (2001), Steel-Frames Parking Garages Take Off at JFK and
Newark International Airports, Modern Steel Construction, April 2001, p. 26-33.

FEMA (2000), Recommended Seismic Design Criteria for New Steel Moment-Frame Buildings,
Report No. FEMA 350, Federal Emergency Management Agency, Washington D.C.

Gewain, R.G., (1973), Fire Experience and Fire Test in Automobile Parking Structures, Fire
Journal, National Fire Protection Association, July.

CRSI (1987), CRSI Handbook, Concrete Reinforcing Steel Institute.

ICBO, (1998), California State Fire Code-Title 24. Part 9.

ICBO, (1997), Uniform Building Code, International Conference of Building Officials.

ICC, (2000a), The International Building Code, IBC-2000, International Code Council, Falls
Church, VA.

ICC, (2000b), "The International Fire Code, IFC-2000, " International Code Council, Falls
Church, VA.


36
Kosmatka, S. H. and Panarese, W.C., (1988) , Design and Control of Concrete Mixtures, 13
th

Edition, Portland Cement Association.

LA-AISC-PMC and Flynn, L.J., (2000), Design of Special Concentric Braced Frames,
Technical Report and Presentation, Los Angeles AISC Professional Member Council, October.

Liu, J. and Astaneh-Asl, A., (2000), Cyclic Tests on Simple Connections Including Slab Effects,
Proceedings, North Am. Steel Construction Conference, AISC, Las Vegas.

Moore, K.S., Malley, J.O., and Englehardt, M.D., (1999), Design of Reduced Beam Section
(RBS) Moment Frame Connections, Steel TIPS, Structural Steel Educational Council,
California, August. (This report can be downloaded free from www.aisc.org web site.)

MSC (2001), April 2001 Issue of Modern Steel Construction, American Institute of Steel
Construction, Chicago.

Munger, C.G., (1984), Corrosion Prevention by Protective Coatings, National Association of
Corrosion Engineers.

Nader, M. and Astaneh-Asl, A. (1992). Seismic Behavior and Design of Semi-rigid Frames.
Report No. UCB/EERC - 92/06, Dept. of Civil Engrg. Univ. of California, Berkeley, May.

Nader, M. and Astaneh-Asl, A. (1989). Experimental Studies of a Single Story Steel Structure
with Fixed, Semi-rigid and Flexible Connections. Report No. UCB/EERC - 89/15, Department of
Civil Engineering, University of California, Berkeley, August.

NPA (1980)) Recommended Building Code Provisions for Open Parking Structures National
Parking Association, Parking Consultants Council, July 1980 (Revised).

PCI (1985), Specification for Unbonded Single Strand Tendons, Journal Prestressed Concrete
Institute, Vol. 30, No. 2, March-April 1985, pp. 22-39.

Pfeiffer, D.W. and Scali, M.J., (1981) , Concrete Sealers for Protection of Bridge Structure,
NCHRP Report No. 244, Transportation Research Board, Washington, D.C.

Putkey, J. (2001). Personal Communication with L. Flynn and A. Astaneh-Asl, August.

SBC (1991), Standard Building Code, Southern Building Code Congress International, Inc.

SDI (1989), LRFD Design manual for Composite Beams and Girders with Steel Deck, No.
LRFD1, the Steel Deck Institute.

SEAOC, (1999), Recommended Lateral Force Requirements and Commentary, Seventh Ed.,
Structural Engineers Association of California, Sacramento, CA.


37
Shen, J. H. and Astaneh-Asl, A. (1993). Hysteresis Behavior and Modeling of Double-angle
Semi-rigid Connections. Vol. 8 of Seismic Condition Assessment of the East Bay Crossing of
SFOBB. Report No. UCB/CE-Steel-93/10. Department of Civil and Environmental Engineering,
University of California, Berkeley,
(This Report can be ordered by sending an E-mail to [email protected])

SSPC (1984) Journal of Protective Coatings & Linings, issued monthly by SSPC. First Edition,
June 1984. Automatically distributed to all SSPC members, or available by subscription.

SSPC (1986), Problem Solving Forum, from Journal of Protective Coatings & Linings, Vol. 1
(June 1984-December 1986), SSPC Pub. No. 87-10; Vol. 2 (January 1987-June 1989), SSPC
Pub. No. 89-06.

SSPC-AISC (1972), A Guide to the Shop Painting of Structural Steel, jointly prepared by SSPC
and the American Institute of Steel Construction, 16pp.

Steel Structures Painting Manual, Volume 1, Good Painting Practice, Second Edition, 1982,
SSPC Report 82-01, 585 p., ed. J. Keane et al. al.

Steel Structures Painting Manual, Volume 2, Systems & Specifications, Fifth Edition, 1989,

TradeARBED INC., (1991), Car Parks in Structural Steel, Information Brochure published by
TradeARBED.

TRB (1987), Evaluation of Bridge Deck Protective Strategies, National Cooperative Highway
Research Program, Report 297, September 1987.

Troup, E.W.J., (1989), Steel frame car parks-New England style, Steel Construction Annual,
Structural Steel Fabricators of New England.

USS (1971), Technical Report on Steel Frames Parking Structures, Published and Distributed by
United States Steel Corporation Pittsburgh, PA.

Wallace, Jr., W.J., Shop Painting of Steel in Fabricating Plants, Chapter 9 of SSPC Painting
Manual, Volume 1, pp. 242-262.

Wiss, Janney, Elstner Association, Inc (1987), Protective Systems for New Prestressed and
Substructure Concrete, prepared for Federal Highway Administration, FHWA/RD-86/193,
April 1987.

38

Following Steel Technical Information and Product Services Reports (Steel TIPS)
are available at AISC website: www.aisc.org and can be downloaded free
for personal use courtesy of the California Iron Workers Administrative Trust.
-------------------------------------------------------------------------------------------------------------------------------------------
-----

Sept. 01: Notes on Design of Steel Parking Structures Including Seismic Effects, by Lanny J. Flynn , and
Abolhassan Astaneh-Asl.
Jun '01: Metal Roof Construction On Large Warehouses or Distribution Centers, by John L. Mayo.
Mar. '01: Large Seismic Steel Beam-to-Column Connections, by Egor P. Popov and Shakhzod M.Takhirov.
Jan 01: Seismic Behavior and Design of Steel Shear Walls, by Abolhassan Astaneh-Asl.
Oct. '99: Welded Moment Frame Connections With Minimal Residual Stress, by Alvaro L. Collin and James J.
Putkey.
Aug. '99: Design of Reduced Beam Section (RBS) Moment Frame Connections, by Kevin S. Moore, James O.
Malley and Michael D. Engelhardt.
Jul. '99: Practical Design and Detailing of Steel Column Base Plates, by William C. Honeck & Derek Westphal.
Dec. '98: Seismic Behavior and Design of Gusset Plates, by Abolhassan Astaneh-Asl.
Mar. '98: Compatibility of Mixed Weld Metal, by Alvaro L. Collin & James J. Putkey.
Aug. '97: Dynamic Tension Tests of Simulated Moment Resisting Frame Weld Joints, by Eric J. Kaufmann.
Apr. '97: Seismic Design of Steel Column-Tree Moment-Resisting Frames, by Abolhassan Astaneh-Asl.
Jan. '97: Reference Guide for Structural Steel Welding Practices.
Dec. '96: Seismic Design Practice for Eccentrically Braced Frames (Based on the 1994 UBC), by Roy Becker &
Michael Ishler.
Nov. '95: Seismic Design of Special Concentrically Braced Steel Frames, by Roy Becker.
Jul. '95: Seismic Design of Bolted Steel Moment-Resisting Frames, by Abolhassan Astaneh-Asl.
Apr. '95: Structural Details to Increase Ductility of Connections, by Omer W. Blodgett.
Dec. '94: Use of Steel in the Seismic Retrofit of Historic Oakland City Hall, by William Honeck & Mason
Walters.
Dec '93: Common Steel Erection Problems and Suggested Solutions, by James J. Putkey.
Oct. '93: Heavy Structural Shapes in Tension Applications.
Mar. '93: Structural Steel Construction in the '90s, by F. Robert Preece & Alvaro L. Collin.
Aug. '92: Value Engineering and Steel Economy, by David T. Ricker.
Oct. '92: Economical Use of Cambered Steel Beams.
Jul. '92: Slotted Bolted Connection Energy Dissipaters, by Carl E. Grigorian, Tzong-Shuoh Yang & Egor P.
Popov.
Jun. '92: What Design Engineers Can Do to Reduce Fabrication Costs, by Bill Dyker & John D. Smith.
Apr. '92: Designing for Cost Efficient Fabrication, by W.A. Thornton.
Jan. '92: Steel Deck Construction.
Sep. '91: Design Practice to Prevent Floor Vibrations, by Farzad Naeim.
Mar. '91: LRFD-Composite Beam Design with Metal Deck, by Ron Vogel.
Dec. '90: Design of Single Plate Shear Connections, by Abolhassan Astaneh-Asl, Steven M. Call and Kurt M.
McMullin.
Nov. '90: Design of Small Base Plates for Wide Flange Columns, by W.A. Thornton.
May '89: The Economies of LRFD in Composite Floor Beams, by Mark C. Zahn.
Jan. '87: Composite Beam Design with Metal Deck.
Feb. '86: UN Fire Protected Exposed Steel Parking Structures.
Sep. '85: Fireproofing Open-Web Joists & Girders.
Nov. '76: Steel High-Rise Building Fire.

39
About the authors.

Lanny J . Flynn, P.E., S.E., is Principal and Vice
President of Design-Build Services for Chalker
Putnam Collins & Scott, a structural engineering
consulting firm with offices in Tacoma and Seattle
Washington.

He has worked in the structural consulting field for
several prominent engineering firms as well as his own
private practice. His experience is broad based
working on a variety of challenging projects both
domestic and international, ranging from high rise
towers to single story projects.

He has served as the American I nstitute of Steel
Construction, AI SC, Regional Engineer for the
Western United States providing technical assistance
to structural engineers, architects, steel fabricators,
contractors and owners.

He currently serves on AI SCs, Technical Advisory
Committee, Manual and Textbook Committee and the
Committee on Specifications, TC-9 Seismic.

He can be reached at:

Lanny J . Flynn, P.E., S.E.
Chalker Putnam Collins & Scott
950 Pacific Avenue, Suite 1100
Tacoma, WA 98402
Phone: (253) 383 2797, Fax: (253) 383 1557
E-mail: [email protected]
Web page: www.cpcsengineers.com

Abolhassan Astaneh-Asl, Ph.D., P.E., is a professor of
structural engineering at the University of California,
Berkeley. He is the winner of the 1998 AI SC, T.R.
Higgins Award.

Dr. Astaneh-Asl received a master of science in civil
engineering from Tehran Polytechnic (now Amir
Kabir University) in I ran in 1968. He was a structural
engineer and construction manager from 1968 to 1978
in Tehran designing and constructing buildings and
other structures. I n 1979, he received an M.S. and in
1982 a Ph.D. degree, from the University of
Michigan.

Since 1982, he has been involved in teaching,
research and design of steel structures. I n recent years,
he has conducted several major projects on seismic
design and retrofit of steel long span bridges and tall
buildings. Since1995, he has also been studying
behavior of steel structures subjected to blast loads and
has been involved in testing and further development
of a cable- based mechanism to prevent progressive
collapse of steel structures. The original concept of the
system was suggested by Dr. J oseph Penzien in 1996
and in the aftermath of terrorist attack on Murrah
bulding in Oklahoma City.

Since September 11, 2001, he has been heavily
involved in conducting research, funded by the
National Science Foundation, on the collapse of the
World Trade Center due to terrorist attack.

He can be reached at:

Abolhassan Astaneh-Asl, Ph.D., P.E.,
781 Davis Hall, University of California,
Berkeley, CA 94720-1710
Phone: (510) 642 4528, Fax: (510) 643 5258
Home office Phone and Fax: (925) 946-0903
Cell Phone for Urgent Calls: (925) 699-3902
E-mail: [email protected],
Web page: www.ce.berkeley.edu/~astaneh

1
P.O. Box 6190
Moraga, CA 94570
Tel. (925) 631-1313
Fax. (925) 631-1112

Fred Boettler, Administrator

Steel TIPS may be viewed and downloaded at www.aisc.org

S P O N S O R S

Adams & Smith Four Star Erectors Plas-Tal Manufacturing Co.
Bannister Steel, Inc. Gayle Manufacturing Reno Iron Works
Baresel Corp The Herrick Corporation SME Industries
Bethlehem Steel Corporation Hoertig Iron Works Schollenbarger-Borello, Inc.
Bickerton Industries, Inc Junior Steel Company Strocal Inc.
Bostrum Bergen. Martin Iron Works Inc. Templeton Steel Fabrication
California Erectors McLean Steel Inc. Trade Arbed
Eagle Iron Construction Nelson Stud Welding Co. Verco Manufacturing, Inc
Eandi Metal Works Oregon Steel Mills Vulcraft Sales Corp.
Western Steel & Metals, Inc.
The local structural steel industry (above sponsors) stands ready to assist you in determining the most
economical solution for your products. Our assistance can range from budget prices and estimated
tonnage to cost comparisons, fabrication details and delivery schedules.

Funding for this publication provided by the California Iron Workers Administrative Trust.
Steel

Steel Tips Committee of California Parte 3

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Steel Tips Committee of California Parte 3

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Steel Tips Committee of California Parte 3

Uploaded by

Copyright:

Available Formats

STRUCTURAL STEELEDUCATIONALCOUNCIL

TECHNICAL INFORMATION& PRODUCTSERVICE

Tests shall be conducted by the

e've all heard the aphorism, "There's

You might also like