JOINT DESIGN FOR

REINFORCED CONCRETE BUILDINGS

by

Michael J. P~ei~~er
David Darwin

A Report on Research Sponsored by

The University o~ Kansas
Structural Engineering and Materials Laboratory

UNIVERSITY OF KANSAS
LAWRENCE, KANSAS
December 1987
 ii

Abstract

JOINT DESIGN FOR

REINFORCED CONCRETE BUILDINGS

This report discusses construction, contraction and expansion joints in

reinforced concrete buildings. The report addresses the purpose of each

type of joint and emphasizes the selection of joint locations and joint

spacings. Some aspects of joint configuration and construction are also

covered. Empirical and analytical design techniques are presented.

 iii

Acknowledgements

This report is based on a project report submitted by Michael J.

Pfeiffer to the Civil Engineering Department of the University of Kansas in

partial fulfillment of the requirements for the MSCE degree.

Support for publication of the report is provided by the University of

Kansas Structural Engineering and Materials Laboratory and the Department of

Civil Engineering.
 iv

Table or Contents
Page
ABSTRACT . • • • 11

ACKNOWLEDGEMENTS iii

LIST OF TABLES • v

LIST OF FIGURES. vi

INTRODUCTION

THE NEED FOR JOINTS. 2

CONSTRUCTION JOINTS. 4

Joint Construction. 5

Joint Location. 6

Summary • • . 9
CONTRACTION JOINTS 9
Joint Configuration 10

Joint Location. 10

EXPANSION JOINTS .• 11

Single Story Buildings: Martin and Acosta 13

Single and Multi-Story Buildings: Varyani and Radhaji 18

Single and Multi-Story Buildings: National Academy of Sciences 24

REFERENCES. 30

TABLES. 33
FIGURES 45
APPENDIX A - NOTATION 53
APPENDIX B - EXPANSION JOINT EXAMPLES 55

Example 1: Single Story-Multi Bay Building. 55

Example 2: Multi Story-Multi Bay Building . 65
 v

LIST OF TABLES

Page

Table Contraction joint spacings .. .... . . ..• 33

Table 2 - Contraction joint spacings for sanitary
engineering structures based on reinforcement
percentage (Rice 1984) . 33

Table 3 - Expansion joint spacings 34

Table 4 - Beam-column frame geometry used for temperature and

vertical load analysis (Martin and Acosta 1970). • • • • • 34

Table 5 - Maximum and minimum daily temperatures for given

locations (Martin and Acosta 1970) • • • • • • • • 35

Table 6 - Temperatures used for calculation of 6T

(National Academy of Sciences 1974). • • • • • • • • • • • 36

Table 7 - Results of Temperature Analysis

(National Academy of Sciences 1974) • • • • • • • • • • • • • • 44
 vi

LIST OF FIGURES

Figure Page

Wall expansion joint cover

(courtesy Architectural Art Mfg., Inc.) • • • • • • • • • • • • 45

2 Fire rated filled expansion joint

(courtesy Architectural Art Mfg., Inc.) •• • • • • 45

3 Length between expansion joints vs. design temperature

change, ~T (Martin & Acosta 1970)
(1ft= 0.305 m.; 1°F = ~°C) • • • • • • • • • 46

Multi-bay frame and one bay substitute frame

(after Varyani & Radhaji 1978) • • • • .. • • • • • • • 47

5 Moments at base of corner columns due to gravity

one bay substitute frames (after Varyani & Radhaji 1978) • • • 48

6 Moments at base of corner columns due to temperature

change using one bay substitute frames
L. =total length between expansion joints
(after Varyani & Radhaji 1978) • • • • • • • • • • • • • • • • 49

7 Expansion joint criteria of the Federal Construction

Council (National Academy of Sciences 1974)
(1 ft = 0.305 m; 1°F =~·c) • • • • . • • • • • • • • • • • • • 50

8 Expansion joint criteria of one federal agency

(National Academy of Sciences 1974)
(1ft= 0.305 m; 1•F = ;•c) • • • • • • • • • • • • • • • • • 51

9 Frames subjected to a uniform temperature change

(National Academy of Sciences 1974) • • • • • • • • • • • • • • 52
 INTRODUCTION

Volume changes caused by changes in moisture and temperature should be

accounted for in the design of reinforced concrete buildings. The magnitude

of the forces developed and the amount of movement caused by these volume

changes is directly related to building length. Contraction and expansion

joints limit the magnitude of forces and movements and cracking induced by

moisture or temperature change by dividing buildings into individual

segments. Joints can be planes of weakness to control the location of

cracks (contraction joints), or lines of total separation between segments

(expansion joints).

There is currently no universally accepted design approach to accom-

modate building movements caused by temperature or moisture changes. Many

designers use "rules of thumb" that set limits on the maximum length between

building joints.

Although widely used, rules of thumb have the drawback that they do not

account for the many variables which control volume changes in reinforced

concrete buildings. For example, variables which affect the amount of ther-

mally induced movement include the percentage of reinforcement, which limits

the amount of movement and cracking in the concrete; the restraint provided

at the foundation, which limits the movement of the lower stories; the

geometry of the structure, which can cause stress concentrations to develop,

especially at abrupt changes in plan or elevation; and provisions for in-

sulation, cooling, and heating, which affect the ability of a building to

dampen the severity of outside temperature changes.

In addition to these variables, the amount of movement in a building is

directly related to the type of aggregate, cement, mix proportions, admix-

tures, humidity, construction sequence, and curing procedures used. While

 2

these variables can be addressed quantitatively, their consideration is

usually beyond the scope of a typical design sequence and will not be con-

sidered here. A number of these parameters are addressed by Mann (1970).

The purpose of this report is to provide guidance for the placement of

contraction and expansion joints in reinforced concrete buildings. A sec-

tion is included on construction joints. Isolation joints on slabs on grade

within the buildings are not covered.

The following section provides a brief overview, outlining the need for

joints. The next section is devoted to construction joints, reviewing cur-

rent procedures for locating and detailing these joints. A section on

contraction joints follows, reviewing current recommendations for contrac-

tion joint spacing. In the final section, three different approaches to

expansion joint placement are presented: Martin and Acosta (1970), Varyani

and Radhaji (1978), and the National Academy of Sciences (1974). Design ex-

amples illustrate the application of the three methods. For additional

information, the reader is directed to an annotated bibliography by Gray and

Darwin (1984).

THE NEED FOR JOINTS

Due to the low tensile capacity of concrete, some cracking in rein-

forced concrete is unavoidable. Contraction joints provide a weakened plane

for cracks to form. Through the use of architectural details, these joints

can be located so that cracks will occur in less conspicious locations

within a building and possibly be eliminated from view.

Expansion joints allow thermally induced movements to occur with a min-

imum build-up of stress. The greater the spacing between joints, the

greater the stresses. Typically, these joints isolate a frame into a series
 3

of segments with enough joint width to allow the building to expand with in-

creasing temperature. By isolating the segments, expansion joints also

provide relief from cracking due to contraction, and therefore act in a dual

role.
Crack control in reinforced concrete buildings is needed for two

reasons. The obvious reason is aesthetics. Where cast-in-place concrete is

to be the finished product, cracks are unsightly. Cracks in major framing

elements such as girders and columns tend to promote questions concerning

the structural adequacy of the structure. They may, in fact, pose no struc-

tural problems, but to the average person without structural knowledge,

they can be cause for alarm. Secondly, cracks of substantial width invite

air and moisture into the framework of the structure, possibly having

deleterious effects. Two examples illustrate the magnitude of potential

cracks.

Lewerenz (1907) cites a plain concrete retaining wall located at the

U.S. Navy Yard in Pudget Sound, Washington. This wall was built with expan-

sion joints spaced every 70ft (21.4 m). After being subjected to four

complete cycles of summer-winter temperature changes [8 to 95 °F (-13 to 35

°C)], the joints had opened as much as 3/16 in. (4.8 mm).

Hunter (1953) describes a four story bakery, 200 ft (61 m) long by 50

ft (15 m) tall, built in 1937. An expansion joint placed at mid-length

opened as much as 3/4 in. (19 mm) at the roof level. The width of the joint

tapered to zero at the basement level. The magnitude of this movement is

directly attributed to thermal strains caused by the heat generated by the

ovens, coupled with outside temperature effects.

 4

As demonstrated by these two examples, the need for crack control in

reinforced concrete structures is real. The key questions are: How to con-

trol the amount of cracking (through the use of contraction joints), and how

to limit stresses in members to an acceptable level (through the use of ex-

pansion joints)? In the sections that follow, recommendations are presented

for contraction joint spacing, and specific procedures are presented for the
placement of expansion joints.

Once joint locations are selected, the joint must be constructed so

that it will act as intended. The weakened section at a contraction joint

may be formed or sawed, either with no reinforcement or a portion of the to-

tal reinforcement passing through the joint. The expansion or isolation

joint is a discontinuity in both reinforcement and concrete. Therefore, an

expansion joint is effective for both shrinkage and temperature variations.

Both joints can be used as construction joints, as described in the follow-

ing section.

CONSTRUCTION JOINTS

Except for very small structures, it is impractical to place concrete

in a continuous operation. Construction joints are needed in order to ac-

commodate the construction sequence for placing the concrete. The amount of

concrete that can be placed at one time is governed by batching and mixing

capacity, crew size, and the amount of time allotted. Correctly sited and

properly executed construction joints provide limits for successive concrete

placements, without adversely affecting the structure.

For monolithic concrete, a good construction joint provides a well-

bonded watertight surface, which allows for flexural and shear continuity

through the joint. Without this continuity, a weakened region results,

which may serve as a contraction or expansion joint. A contraction joint is

 5

formed by limiting the percentage of reinforcement through the joint, thus

creating a plane of weakness. An expansion joint is formed by leaving a gap

in the structure of sufficient width to remain open under extreme tempera-

ture conditions. If possible, construction joints should coincide with

contraction or expansion joints, which are discussed in the following

sections. The balance of this section is devoted to construction joints in

regions of monolithic concrete.

Joint Construction

To achieve a well-bonded watertight joint, a few conditions must be met

prior to placement of the fresh concrete. The hardened concrete must be

clean and free of all laitance (ACI Committee 311 1981).

If only a few hours elapse between successive placements, a visual

check is needed to be sure that all loose particles, dirt, and laitance are

removed. The new concrete will be adequately bonded to the hardened green

concrete, provided that the new concrete is vibrated thoroughly over the

area.

Older joints need a little more surface preparation. Cleaning by means

of an air-water jet or wire brooming can be done when the concrete is still

soft enough that any laitance can be removed, but hard enough to prevent ag-

gregate from loosening. Concrete that has set should be prepared using a

wet sand blast or ultra-high pressure water jet (ACI Committee 311 1981).

ACI 318 states that existing concrete should be moistened thoroughly

prior to placement of fresh concrete. Green concrete will not require any

additional water, but concrete that has dried out may require saturation for

a day or more. No pools of water should be left standing on the wetted sur-

face at the time of placement.

 6

Form construction plays an important role in the quality of a joint.

It is essential to minimize the leakage of grout from under stop-end boards

(Hunter 1953). If the placement is deeper than 6 in., the possibility of

leakage is even greater due to the increase in the pressure head of the wet

concrete. Grout which escapes under the form will form a thin wedge of

material, which must be cut away prior to the next placement. If not

removed, this wedge will not adhere to the fresh concrete, and under load,

deflection in the element will cause this joint to open.

Joint Location

The final consideration is placing the construction joint in the right

place. Assuming an adequate production capacity, construction joints should

be located where they will least affect the structural integrity of the ele-

ment under consideration, while at the same time being compatible with the

building's appearance. Placement of joints varies, depending on the type of

element under construction. For this reason, beams and slabs will be ad-

dressed separately from columns and walls.

Beams and Slabs--From the point of view of strength in beam and slab

floor systems, desirable locations for joints placed perpendicular to the

main reinforcement are at points of minimum shear or at points of

contraflexure. Typically, joints are located at mid-span or in the middle

third of the span, but locations should be verified by the engineer before

placement is shown on the drawings. In beam and girder construction, where

a beam intersects a girder at the point of minimum shear, ACI 318 states

that the construction joint in the girder should be offset a distance equal

to twice the width of the incident beam.

Horizontal construction joints in beams and girders are usually not

recommended. Common practice is to place beams and girders monolithically

 7

with the slab. In the case of beam and girder construction where the mem-

bers are of considerable depth, Hunter (1953) recommends placing concrete in

the beam section up to the soffit of the slab, then placing the slab in a

separate operation. The reasoning behind this is that cracking of the top

surface may result due to vertical shrinkage in a deep member. With this

procedure, there is a possibility that the two surfaces will slip due to
horizontal shear in the member. In this case, adequate shear transfer must

be provided (ACI 318).

Construction joints parallel to the slab span can be placed anywhere,

except those locations in T-beam construction that rely on a portion of the

slab to act with the beam in resisting flexure.

The main concern in joint placement is to provide adequate shear trans-

fer and flexural continuity through the joint. Flexural continuity is

achieved by continuing the reinforcement through the joint with enough

length past the joint to insure an adequate splice length for the

reinforcement. Shear transfer is provided by shear friction between the old

and new concrete, and/or dowel action in the reinforcement through the

joint. Shear keys are usually undesirable (Fintel 1974), since keyways are

possible locations for spalling of the concrete. If proper concreting pro-

cedures are followed, the bond between the old and new concrete, plus the

effect of the reinforcement crossing the joint, are adequate to provide the

necessary shear transfer.

Columns and Valls--It is general practice to limit concrete placements

to a height of one story. Construction joints in columns and bearing walls

should be located at the undersides of floor slabs and beams, and at the top

of floor slabs for columns continuing to the next floor. Column capitals,

haunches, drop panels, and brackets, should be placed monolithically with

 8

the slab. Depending on the architecture of the structure, the construction

joint may be used as an architectural detail, or located to blend in without

being noticeable. Quality form construction is of paramount importance in

order to provide the visual detail required (PCA 1982).

The placement of fresh concrete on a horizontal surface can affect the

joint. Common practice has been to provide a bedding layer of mortar, of

the same proportions as that in the concrete, prior to placement of new con-

crete above the joint. The ACI Manual of Concrete Inspection (ACI Committee

311 1981) recommends using a bedding layer of concrete with somewhat more

cement, sand, and water than the design mix for the structure. Aggregate

less than 3/4 in. can be left in the bedding layer, but all aggregate larger

than 3/4 in. should be removed. This mix should be placed 4 to 6 in. deep

and thoroughly vibrated with the regular mix placed above. To avoid settle-

ment cracks in slabs and beams due to vertical shrinkage of previously

placed columns and walls, the concrete in the columns and walls should be

allowed to stand for at least two hours prior to placement of subsequent

floors.

Placement of vertical construction joints in walls also needs to be

compatible with the architectural flavor of the structure. Construction

joints are often located near reentrant corners of walls, alongside columns,

or other locations where they become an architectural feature of the

structure. If the building architecture does not dictate where the joints

should be placed, placement considerations, such as production capacity of

the crew or whether or not one set of forms will be reused along the length

of the pour may limit the length between joints. This criteria will usually

limit the maximum horizontal length to 40 ft between joints in most build-

ings (PCA 1982). Due to the critical nature of building corners, it is best
 9

to avoid vertical construction joints at or near a corner, so that the

corner will be tied together adequately.

Shear transfer and bending at joints in walls and columns should be ad-

dressed in much the same way it is for beams and slabs. The reinforcement

should continue through the joint, with adequate length to insure a complete

splice. If the lateral shears are high, the joint must be capable of trans-

fering the load by shear friction or dowel action.

Summary

Construction joints are necessary in most reinforced concrete

construction. Due to their critical nature, they should be located by the

designer, and indicated on the design drawings to insure adequate force

transfer and aesthetic acceptability at the joint. If concrete placement is

stopped involuntarily for a time longer than the initial setting time of the

concrete, the joint should be treated as a construction joint, with advance

input from the designer as to any additional requirements needed to insure

the structural integrity of the element being placed.

CONTRACTION JOINTS

Drying shrinkage and decreases in temperature cause tensile stresses in

concrete, if the material is restrained. Cracks will occur when the tensile

stress reaches the tensile strength of the concrete. Due to the relatively

low tensile capacity of concrete (ftc= 4.0- 7.5 ~for normal weight con-

crete, f' and ft in psi), cracking is likely to occur. Contraction joints

c c
provide planes of weakness for cracks to form, without marring the ap-

pearance of a structure. Contraction joints are used primarily in walls

and in slabs-on-grade.
 10

The greater the distance between contraction joints, the greater will

be the forces in a structure due to volume change. To resist these forces

and minimize the amount of cracking in the concrete, greater amounts of

reinforcement are required.

Joint Configuration

Contraction joints consist of a region with a reduced concrete cross

section and reduced reinforcement. The concrete cross section should be

reduced by a minimum of 25 percent to insure that the section is weak enough

for a crack to form. In terms of reinforcement, there are two types of con-

traction joints currently in use, known as "full" and "partial" contraction

joints (ACI 350R). Full contraction joints, preferred for most building

construction, are constructed with a complete discontinuity in reinforcement

at the joint. All reinforcement is terminated approximately 2 in. (51 mm)

from the joint and a bond breaker placed between successive placements, if

the joint is a construction joint. Partial contraction joints are con-

structed with not more than 50 percent of the reinforcement passing through

the joint. Partial contraction joints are used in liquid containment

structures. In both types of joint, waterstops may be used to insure

watertightness.

Joint Location

Once the decision is made to use contraction joints, the question

remains: What spacing is needed to limit the amount of cracking between the

joints? As shown in Table 1, a number of recommendations are given for con-

traction joint spacing. Recommended spacings vary from 15 to 30 ft (4.6 to

9.2 m) and from one to three times the wall height. For sanitary struc-

tures, Rice (1984) prescribes contraction joint spacings for given

reinforcement percentages (Table 2).

 11

The limits prescribed by Rice in Table 2 are extensions of the limits

recommended in ACI 350R, accounting for reinforcement grade and minimum bar

size. A graphical representation of the same information is given by Gogate

( 1984). It should be noted that if a "partial" contraction joint is used,

the joint spacing should be approximately 2/3 of the full contraction joint

spacing (ACI 350R).

Wood (1981) suggests that any joint within a structure should go

through the entire structure in one plane. If the joints are not aligned,

movement at a joint may induce cracking in an unjointed portion of the

structure until the crack intercepts another joint.

EXPANSION JOINTS

Temperature changes will induce stresses in a structure, if the struc-

ture is restrained. Without restraint, no stresses result. In practice,

all buildings are restrained to some degree. Temperature induced stresses

vary with the magnitude of the temperature change; large temperature varia-

tions can result in substantial stresses that must be accounted for in

design, while low temperature changes may result in negligible stresses.

Temperature induced stresses are the direct result of volume changes

within a structure between restrained points. A rough indication of the

amount of elongation caused by temperature increases is obtained by multi-

plying the coefficient of expansion of concrete [about 5.5 X 10- 6 in./in.

per 0
f ( 9. 9 x 10
-6 mm/mm per °C)] by the length of the structure and the

temperature change. A building 200 ft long subjected to a temperature in-

crease of 25 °f (14 °C) will elongate about 3/8 in. (9.5 mm).

Expansion joints are used to limit member forces caused by thermally

induced volume changes. Expansion joints permit separate segments of a

 12

building to expand or contract without adversely affecting structural in-

tegrity or serviceablity. Expansion joints should be wide enough to prevent

portions of the building on either side of the joint from coming in contact,

when the structure is subjected to the maximum expected temperature rise.

Joints vary in width from 1 to 6 in. (25 to 152 mm) or more, with 2 in. (51

mm) being typical. The wider joints are used to accommodate additional dif-

ferential building movement that may be caused by settlement or siesmic

loading. Joints should pass through the entire structure above the level of

the foundation. Expansion joints should be covered (Fig. 1) and may be

empty or filled (Fig. 2). Filled joints are required for fire rated

structures.

Expansion joint spacing is dictated by the amount of movement that can

be tolerated, plus the allowable stresses and/or capacity of the members.

As with contraction joints, rules of thumb have been developed (Table 3).

These range from 30 to 400 ft (9 to 122 m) depending on the type of

structure. In addition to the rules of thumb, a number of methods have been

developed to calculate expansion joint spacing. This section presents three

of these methods. The three methods are based on the work of Martin and

Acosta (1970), Varyani and Radhaji (1978), and the National Academy of

Sciences (1974).

Martin and Acosta (1970) present an expression for the maximum spacing

of expansion joints in one story reinforced concrete buildings. The expres-

sion is developed based on the results of a study of frame structures

subjected to temperature change. Joint spacing is a function of the length

and stiffness of frame members, and seasonal temperature changes that occur

at the building site. The design temperature change is based on the dif-

ference between the extreme values of the normal daily maximum and minimum
 13

temperatures plus an additional drop in temperature of about 30 •F (17 °C)

to account for drying shrinkage in the concrete. Martin (1970) provides

site specific values of shrinkage equivalent temperature drop. Because of

the additional volume change due to drying shrinkage, joint spacing is

governed by contraction instead of expansion.

Varyani and Radhaji (1978) present a method for calculating expansion

joint spacing based on the design of corner columns in a symmetrical frame

to resist a combination of gravity load plus thermally induced load. Joint

spacing is a function of the extreme daily temperature changes, including an

allowance for shrinkage, and the relative stiffnesses of the first floor

beams and columns. The procedure is applicable to rectangular column

layouts for single and multi-story construction. This method is somewhat

more general but is similar in approach to the method presented by Martin

and Acosta (1970).

The National Academy of Sciences (1974) present an empirical procedure

for selecting expansion joint spacing. A graph relates joint spacing to

yearly extremes in temperature at the building site. The procedure is ap-

plicable to single and multi-story frame structures, and accounts for

variables such as building stiffness and configuration, the type of column

connection to the foundation, and the use of heating and air conditioning

systems.

Example calculations and a discussion of the relative merits of these

methods are presented in Appendix B.

Single Story Buildings: Martin and Acosta

Martin and Acosta (1970) present a method for calculating the maximum

spacing of expansion joints in one story beam-column frames with ap-

proximately equal spans. This method is based on the premise that with
 14

adequate joint spacing, the factor of safety for vertical loading will also

provide an adequate safety factor for the effects of temperature change. By

applying this premise to a number of frame structures, Martin and Acosta

develop a single expression for expansion joint spacing.

To remove the need for specific temperature calculations in design,

Martin and Acosta feel that a building should have an adequate factor of

safety to withstand the applied lateral, vertical, and thermally induced

loads. Due to the short-term nature of thermal loading, Martin and Acosta

address this type of load in the same way as ACI 318-63 addressed wind

loads. Accordingly, the structure should be designed to meet the following

criteria:

Dead and live loads:

u 1 • 50 + 1 • 8L ( 1)

Dead, live, and temperature loads:

U=1.25(D+L+T) (2)

in which U = required strength of the element under consideration: D = dead

load; L = live load, and T = thermal load.

Setting Eq. (1) equal to Eq. (2), yields an expression for the maximum

temperature induced load, T, that is allowed in order for thermal effects to

be omitted in the design calculations. This maximum equivalent load is rep-

resented by:

T = 0.20 + 0.44L ( 3)
 15

This means that temperature loads should not exceed the effect of 20 percent

of the dead load plus 44 percent of the live load.

Beam-column frames with the geometries given in Table 4 were analyzed

for temperature variations obtained at a number of geographical locations.

The number of spans (and consequently the length of the buildings) were in-

creased to model varying lengths between expansion joints.

The magnitude of the design temperature change, AT, is taken as two-

thirds of the difference between the extreme values of the normal daily

maximum and minimum temperatures, T and T . , at these locations. Martin

max m1n
and Acosta arbitrarily chose the two-thirds factor to account for the fact

that the temperature at which the building is completed would statistically

not be at the maximum or minimum daily temperature, but somewhere between

6
the two. They used a coefficient of thermal expansion, a, of 5.5 X 10-

in./in. per °F (9.9 x 10

-6 mm/mm per •c). To account for concrete

shrinkage, Martin and Acosta used an additional equivalent temperature dif-

ferential, which averages about 30°F, in the temperature analysis to obtain

the total design temperature change.

2
AT = CT - T . ) + 30°F (4)
3 max m1n

Although AT is used to proportion expansion joints, AT in Eq. (4) actually

represents the combination of a temperature drop with shrinkage.

The shears and moments due to temperature and shrinkage were compared

to the shears and moments due to vertical loads applied to the same frame.

The loads considered in the vertical analysis were the self weight of the

structure plus a roof live load of 30 psf (1.44 kPa), resulting in a

uniformly distributed dead load of 2 kips per foot (29 kN/m), and a
 16

uniformly distributed live load of 0.5 kips per foot. The critical tempera-

ture effect for the longer beam spans (structures Type E through L of Table

4) was the bending moment in the exterior columns. Structures with shorter

beam spans (Type A through D of Table 4) were governed by the beam moment at

the exterior face of the first interior column.

Using the structural analyses and Eq. (3) as the basis to determine the

maximum expansion joint spacing, Martin and Acosta developed an expression

for maximum expansion joint spacing, Lj in ft.

112,000
L. = R llT (5)
J
in which:
I
c r )
R 144h2 (11 ++ 2r (6)

r = ratio of stiffness factor of column to stiffness factor of beam = Kc/Kb;

3
K = column stiffness factor = I /h, in. ; Kb = beam stiffness factor =
0 0

I /1, in.•; h =column height, in.; 1 = beam length, in.; I = moment of

c c
inertia of the column, in.'; Ib = moment of inertia of the beam, in.' To

calculate llT, Tmax and Tmin are obtained from the Environmental Data
Service for a particular location (see Table 5 for a partial listing). The

resulting length between joints, Lj' given by Eq. (5) for typical values of

R is given in Fig. 3.
In order to avoid damage to exterior walls, the Martin and Acosta

propose an additional criteria for L. to limit the maximum allowable lateral

J
deflection, o, to 1/180 of the column height, h. The maximum lateral

deflection imposed on a column is taken as

 17

( 7)

This leads to the limitation on Lj of

Lj <- 2000h
t;T
t;T in °F. ( 8)

Eq. ( 8) is based on the assumption that the lateral deflection of a

floor system caused by a temperature change is not significantly restrained

by the columns. This assumption is realistic since the in-plane stiffness

of a floor system is generally much greater than the lateral stiffness of

the supporting columns. Thus, the columns have little effect on o.

Discussion--Martin and Acosta state that Eq. (5) yields conservative

results in all cases studied, but is very conservative for very rigid

structures. Due to changes in ACI 318 since 1963, Eq. (5) can be revised to

account for current load factors (ACI 318-83). Eq. (1) and (2) become:

u 1 • 4D + 1 • 7L (9)

u 0.75(1.40 + 1.7L + 1.4T) ( 1 0)

Setting Eq. (9) equal to Eq. (10) yields the maximum allowable temperature

induced load.

T = 0.33D + 0.41L ( 11)

In this case, temperature effects should not exceed the effect of 33 percent

of the dead load plus 41 percent of the live load.

 18

To obtain an updated version of Eq. (5) requires a reanalysis of the

original data of Martin and Acosta. In lieu of a reanalysis, Eq. (5)

remains as a conservative guideline for expansion joint spacing, since Eq.

(3) is more conservative than Eq. (11) in practical applications.

Single and Multi-Story Buildings: Varyani and Radhaji

Varyani and Radhaji (1978) present a procedure to calculate the maximum

distance between expansion joints for symmetrical, single and multi-story

beam-column frames. Thermal loads are based on a temperature differential

similar to that used by Martin and Acosta (1970), two-thirds of the dif-

ference between maximum and minimum daily temperatures, except Varyani and

Radhaji select temperatures from the single day on which the difference be-

tween the maximum and minimum temperatures is the greatest. Moments due to

thermal loading are calculated at the corner columns, accounting for maximum

biaxial bending in the columns. Varyani and Radhaji find that only the

first story columns and the beams supported by these columns are substan-

tially affected by temperature change. Above the first story, the effect of

temperature is dissipated. Therefore, only the first two stories of multi-

story buildings are assumed to be critical in the analysis.

The corner columns are designed for the factored axial load plus

biaxial bending due to gravity load. Trial values of expansion joint spac-

ing, L., are used to calculate temperature induced column moments. The
J
final length between joints is selected so that the column design obtained

under gravity load [Eq. (9)] is adequate under combined gravity and thermal

loading [Eq. (10)]. Alternatively, column reinforcement can be increased to

withstand increased moments for a selected value of Lj.

 19

The thermally induced moments at the base and top of a corner column

can be calculated using analysis techniques such as matrix analysis or mo-

ment distribution. Due to the tedious nature of these calculations for long

structures, Varyani and Radhaji recommend the use of a substitute one-bay

open frame, shown in Fig. 4. The one-bay frames replace the first interior

column with a fixed point at the beam-column joint. Varyani and Radhaji

compared the results obtained with this approximation to results obtained

with more accurate methods (Reynolds 1960) for up to 4 bays, and compared

the substitute bay analysis to frame analyses for buildings up to 10 bays.

They considered the 0 to 10.5 percent margin of error to be satisfactory.

Varyani and Radhaji present expressions for the moment at the base of

corner columns, Mg(base), in a number of different single bay frames under

the effect of gravity. These are given here and in Fig. 5:

Single story, single bay:

Mf
M (base ) = -;-;::--=-;-, ( 12)
g (2+r')

Single story, multi-bay:

M (base) ( 1 3)
g 2(1 + r')

Multi-story, single bay:

Mf
M (base ) = ,.....--,!'--;-, ( 1 4)
g (4 + r')
 20

Multi-story, multi-bay:

Mf
M ( b a s e ) = -::-:--::---=--...,..,- ( 15)
g 2(2 + r')

in which Mf = fixed-end moment at the beam-column joint due to gravity

-1
loading; and r' = r = Kb/Kc
Varyani and Radhaji also present expressions for the base moment of

corner columns in frames under the effect of temperature loading, MT(base).

These expressions are given here and in Fig. 6.

Single story, single bay:

1 r')
MT(base) = 6E K
c c §.
h (2 +
+ r' ( 16)

Single story, multi-bay:

MT(base) 3E K §_ (1 + 2r') (17)

c c h 1 + r'

Multi -story, single bay:

( 1 8)

Multi-story, multi-bay:

~(base) 3 E K §. ( 3 + 2r' ) ( 1 9)
c c h 2 + r'
 21

in which Ec =modulus of elasticity of concrete; and 6 is given in Eq. (7).

Like Martin and Acosta (1970), Varyani and Radhaji use a 2/3 reduction

factor for the maximum temperature change. The maximum fall in temperature

(contraction) is adjusted by an equivalent temperature drop due to shrinkage

of 15 •c (27 °F). The maximum rise in temperature (expansion) is adjusted

by an equivalent temperature drop due to shrinkage of 8 •c (14 °F). Varyani

and Radhaji recommend that these temperature changes be reduced by one-half

to account for creep, duration of load, and loss of fixity due to soil

movement.

l [- ~ (T - T . ) - 27 oF) (20a)
2 3 max m1n

l[~(T - T . ) - 14 •F) (20b)

2 3 max m1n

Substituting Eq. (7) for o into Eq. (16) - (19) results in expressions for

expansion joint spacing, Lj, in terms of~= MtCbase).

Single story, single bay:

(2 + r') ( 21 )
3E K allT 1 + r'
c c
 22

Single story, multi-bay:

(22)

Multi-story, single bay:

l-l.rh r' )
Lj = 3E K ClbT ( 34 +
+ r'
(23)
c c

Multi-story, multi-bay:

(2 + r' ) (24)
3E c Kc ClbT 3 + 2r'

Discussion.--Although Varyani and Radhaji do not develop a "general"

expression for Lj, such as Eq. (5) by Martin and Acosta, the two methods are

similar in approach. Both methods base the expansion joint spacing on the

ability of the first level beams and columns to resist the thermally induced

loads.

Varyani and Radhaji equate the temperature induced moment at the base

of a column with the moment due to gravity at the same point. However, a

column design based on gravity loading is normally governed by the moment at

the top of the column. Therefore, it makes more sense to compare the maxi-

mum combined factored moment at the top or bottom of a column [Eq. ( 1 0)]

with the factored gravity moment [Eq. (9) ]. This approach is demonstrated

in the examples in Appendix B.

 23

The following equations provide the values of Mg and MT at the top of

columns for the single bay frames shown in Fig. 5 and 6.

Single story, single bay:

M (top) (25)
g (2 + r')

6E K .§. ( r' ) (26)

c c h 2 + r'

Single story, multi-bay:

( 27)

MT(top) = h ( 1 r'
c c .§.
6E K
+ r'
) (28)

Multi-story, single bay:

M (top) ( 29)
g (4 + r')

6E K .§_ (~ + r') (30)

c c h + r'

Multi-story, multi-bay:

M (top) =
(2 + r')
( 31 )
g

= 6E K ~ (1 + r • )
~(top) c c h 2 + r' ( 32)
 2~

Varyani and Radhaji recommend the use of the maximum temperature dif-

ferential for a single day, rather than the difference in the extreme value

of normal maximum and minimum daily temperatures used by Martin and Acosta.

However, logic suggests that seasonal changes in temperature are more ap-

propriate for calculating dimensional changes of structures. For this

reason, the temperature range suggested by Martin and Acosta should be ap-

plied with both methods.

Varyani and Radhaji do not specifically consider the effects of tem-

perature change on interior beams and columns, and the results obtained with

the method are very sensitive to the assumed beam and column stiffnesses.

Although Varyani and Radhaji include a 50 percent reduction in ~T [Eq. (20)]

to account for creep and duration of load, the calculated expansion joint

spacings obtained with this method are considerably less than expansion

joint spacings which are used and have performed well in existing

structures. This point is amply demonstrated in Appendix B. The results

can be improved somewhat by using cracked, rather than uncracked section

properties. Treating the members as cracked has relatively little effect on

the gravity moments but reduces the calculated value of MT. The result is

an increased, and somewhat more realistic, value for Lj.

The relative complexity, time consuming nature, and overconservative

results obtained with the method compared to the methods of Martin and
Acosta (1970) and the National Academy of Sciences (197~) are likely to

render this procedure unattractive to designers.

Single and Multi-Story Buildings: National Academy of Sciences

The lack of nationally accepted design procedures for locating expan-

sion joints prompted the Federal Construction Council to undertake the task

of developing more definitive criteria for locating expansion joints in

 25

federal construction. As a consequence, the Council directed its Standing

Committee on Structural Engineering to develop a procedure for expansion

joint design to be used by federal agencies in building design. The

criteria formulated by the Committee was published by the National Academy

of Sciences (1974) in the form of a graph (Fig. 7) which expresses the al-

lowable building length as a function of a design temperature change. The

relationships shown in Fig. 7 are directly applicable to beam-column frames,

with columns hinged at the base and heated interiors. In order for the

graph to be adaptable to a wide range of buildings, the Committee provided

modification factors to reflect building stiffness and configuration, heat-

ing and cooling, and.the type of column connection to the foundation.

Fig. 7 is based on the Committee's investigation of the procedures in

use by federal agencies to select joint spacing, plus an analytical study

comparing the theoretical effects of temperature change on two-dimensional

elastic frames to the actual movements recorded during a one-year study by

the Public Buildings Administration (1943-1944).

The Committee first examined procedures used by federal agencies to

select expansion joint spacing. Although no significant quantitative data

was found to support their criteria, most federal agencies relied on rules

(Fig. 8) that provided maximum building dimensions for heated and unheated

buildings as a function of the change in the exterior temperature.

Temperature change is taken as the maximum difference between the mean tem-

perature during the normal construction season and either the summer or

winter extremes.

The criteria illustrated in Fig. 8 reflect two assumptions: that the

maximum allowable building length between joints must be decreased as the

maximum difference between the mean annual temperature and the

 26

maximum/minimum temperature increases; and that dimensions between joints

can be increased for heated structures, which have the ability to dampen the

severity of the outside temperature changes through temperature control in

the building. The upper and lower bounds of 600 and 200 ft were felt to

reflect a consensus within the engineering profession and have no experimen-

tal or theoretical justification.

Additional information was drawn from an unpublished report by struc-

tural engineers of the Public Buildings Administration (1943-1944). This

report documents an investigation of expansion joint movement over a period

of one year (September 1943 to August 1944) in nine federal buildings.

Based on the report, the conclusions drawn by the Committee and implemented

in their design recommendations are as follows:

A considerable time lag (2 to 12 hours) exists between the

maximum dimensional change and the peak temperature associated

with this change. The investigators attributed this time lag to

three factors: the temperature gradient between the outside

and inside temperatures, the resistance to the heat transfer due

to insulation, and the duration of the ambient temperature at its

extreme levels.

The coefficient of thermal expansion of the first floor level

is approximately one-third to two-thirds that of the upper floors.

The dimensional changes in the upper levels of the building cor-
-6
respond to a coefficient of thermal expansion between 2 X 10 and
5 X 10- 6 in./in. per •F. The investigation seems to confirm that

the upper levels of a building undergo dimensional changes cor-

responding to the coefficient of thermal expansion of the

principal material of which each is constructed.

 27

To interpret the observed dimensional changes reported by the Public

Buildings Administration (1943-1944), and enhance their understanding of the

distribution of stresses and resulting frame deformations, the Committee

analytically studied the effects of a uniform temperature change on typical

two-dimensional frames. This study was based on an analysis of nine beam-

column frames subjected to a design temperature change, 6T, of 100° F. In

practice, 6T, is calculated for a specific site as the larger of:

6T T - T ( 33a)
w m

6T T - T ( 33b)
m c

in which T = temperature during the normal construction season in the

m
locality of the building, assumed to be the continuous period in a year

during which the minimum daily temperature equals or exceeds 32 °F (0 °C);

Tw = temperature exceeded, on average, only 1 percent of the time during the

summer months of June through September; and T = temperature equaled or ex-

c
ceeded, on average, 99 percent of the time during the winter months of

December, January, and February.

Observed values for T , T , and T for locations throughout the United

m w 0

States are given in Table 6 (ASHRAE Handbook of Fundamentals, 1972).

For the analysis, member forces, F, associated with 6T = 100°F were ap-

plied at beam-column joints in buildings consisting of bay spacings of 25 ft

(7.6 m), with the first story height assumed to be 13 ft (4 m) and the upper

stories spaced at 10 ft (3m). F is calculated according to the following:

F = aAE 6T (34)
c
 28

in which F the force resulting in a restrained member due to a temperature

change ~T; a = 6.0 X 10

-6 in./in. per •F (10.8 x 10 ~ mm/mm per •c) (assumed

in the Committee's study); and A= cross-sectional area of the member. The

column sizes, beam sizes, number of stories, and type of connection to the

foundation (hinged versus fixed) were varied to help determine the effects

of temperature changes on a wide range of buildings.

The results of the analysis are given in Table 7 and summarized as

follows:

The intensity of the horizontal shear in the 1st story

columns (Fig. 9) is greatest at the ends of the frame and ap-

preaches zero at the center.

The beams near the center of the frame in Fig. 9 are sub-

jected to maximum axial forces, while the columns at the ends of

the frame are subjected to maximum bending moments and shears at

the beam-column joint.

Shears, axial forces, and bending moments at critical sec-

tions within the lowest story are almost twice as high for fixed-

column buildings as the forces at corresponding locations in

hinged-column buildings.

The horizontal displacement of one side of the upper floors,

o, is approximately equal to the assumed displacement that would

develop in an unrestrained frame if both ends of the frame are

equally free to displace= ~aL.~T [Eq.(7)].

J
The horizontal displacement of a frame that is restricted

from side displacement at one end (non-symmetrical stiffness,

analysis M-1) results in a total horizontal displacement of the

other end of approximately aLj~T.

 29

An increase in the relative cross-sectional area of the beams

(not associated with a simultaneous increase in the moment of in-

ertia of the beams), as illustrated in analysis A-1 versus A-2,

results in a considerable increase in the controlling design

forces. This occurs because the magnitude of the thermally in-

duced force, F, is proportional to the cross-sectional area of the

element [Eq. (34)].

Hinges placed at the top and bottom of the exterior columns

of the frame result in a reduction of the maximum stresses that

develop (analysis 1-1 versus M-2). These hinges, however, allow

an increase in the horizontal expansion of the first floor.

As a result of the Committee's investigation, Fig. 7 was developed.

This graphical representation of the maximum expansion joint spacing is

based on the correlation of the observed dimensional changes given by the

Public Buildings Administration (1943-1944) with the Committee's analytical

study, plus the current practices of federal agencies (Fig. 8). The

Committee rationalized that the step function of Fig. 8 could not represent

the behavior of a physical phenomenon such as thermal effects and therefore

assumed a linearly varying function for a 30 to 70 °F (17 to 39 °C) tempera-

ture change. The upper and lower bounds are based on Fig. 8.

The limits prescribed in Fig. 7 are directly applicable to buildings of

beam-column construction (including structures with interior shear walls or

perimeter base walls), hinged at the foundation, and heated. If these con-

ditions are not met, the Committee recommends the following conservative

modification factors which reflect the collective experience and judgement

of the Committee.
 30

If the building will be heated only and have hinged column

bases, use the length specified.

If the building will be air-conditioned as well as heated, in-

crease the allowable length by 15 percent.

If the building will be unheated, decrease the allowable

length by 33 percent.

If the building will have fixed column bases, decrease the al-

lowable length by 15 percent.

If the building will have substantially greater stiffness

against lateral displacement at one end of the structure, decrease

the allowable length by 25 percent.

When one or more of these conditions occur, the total modification factor is

the algebraic sum of the individual adjustment factors that apply.

Discussion--The Federal Construction Council Committee does not recom-

mend their procedure for all situations. Particularly, when a generic

representation is not adequate for a unique structure or when the empirical

approach provides a solution that professional judgement indicates is too

conservative, a detailed analysis should be performed. This analysis should

recognize the amount of lateral deformation that can be tolerated, and the

structure should be designed so that this limit is not exceeded.

REFERENCES

ACI Committee 22~, 198~, Control of Cracking in Concrete Structures (ACI

22~R-80)
(Revised 198~), American Concrete Institute, Detroit, Michigan, ~3
pp.

ACI Committee 311, 1981, ACI Manual of Concrete Inspection, Sixth Edition,
ACI Publication SP-2, Detroit, Michigan, 508 pp.

ACI Committee 318, 1983, Building Code Requirements for Reinforced Concrete
(ACI 318-83), American Concrete Institute, Detroit, Michigan, 111 pp.
 31

ACI Committee 350, 1983, "Concrete Sanitary Engineering Structures (ACI

350R-83)," ACI Journal, Proceedings, v. 80, No.6, Nov.-Dec., pp. 467-486.

Billig, Kurt, 1960, "Expansion Joints," Structural Concrete, London,

McMillan and Co., Ltd., pp. 962-965.

Crane, T., 1956, Architectural Construction, The Choice of Structural

Design, Second Edition, John Wiley and Sons, Inc., New York, 433 pp.

Fintel, Mark, 1974, "Joints in Buildings," Handbook of Concrete Engineering,

New York, New York, VanNostrand Reinhold Company, pp. 94-110.

Gray, David C., and Darwin, David, 1984, "Expansion and Contraction Joints
in Reinforced Concrete Buildings - An Annotated Bibliography," SM Report No.
14, University of Kansas Center for Research, Lawrence, KS, December, 43 pp.

Gogate, Anand B., 1984, "An Analysis of ACI Committee 350's Recommended
Design Standards," Concrete International: Design and Construction, V. 6,
No. 10, Oct., pp. 17-20.

Hunter, L. E., 1953, "Construction and Expansion Joints for Concrete, 11 Civil
Engineering and Public Works Review, V. 48, No. 560, February, pp. 157-158,
and V. 48, No. 561, Mar., pp. 263-265.

Indian Standards Institution, 1964, IS 456-1964, Code of Practice for Plain

and Reinforced Concrete, Second Revision, New Delhi.

Lewerenz, A.C., 1907, "Notes on Expansion and Contraction of Concrete

Structures," Engineering News, V. 57, No. 19, May 9, pp. 512-514.

Mann, 0. Clarke, 1970, "Expansion-Contraction Joint Locations in Concrete

Structures," Designing for the Effects of Creep, Shrinkage, and Temperature
in Concrete Structures, SP-27, American Concrete Institute, Detroit,
Michigan, pp. 301-322.
Martin, Ignacio, 1970, "Effect of Environmental Conditions on Thermal
Variations and Shrinkage of Concrete Structures in the United States,"
Designing for the Effects of Creep, Shrinkage, and Temperature in Concrete
Structures, SP-27, American Concrete Institute, Detroit, Michigan, pp. 279-
300.
Martin, Ignacio and Acosta, Jose, 1970, "Effect of Thermal Variations and
Shrinkage on One Story Reinforced Concrete Buildings," Designing for the
Effects of Creep, Shrinkage, and Temperature in Concrete Structures, SP-27,
American Concrete Institute, Detroit, Michigan, pp. 229-240.
Merril, W.S., 1943, "Prevention and Control of Cracking in Reinforced
Concrete Buildings," Engineering News Record, V. 131, No. 23, Dec., 16, pp.
91-93.
National Academy of Sciences, 1974, "Expansion Joints in Buildings,"
Technical Report No. 65, Washington, D.C., 43 pp.
 32

Portland Cement Association, 1982, Building Movements and Joints, Skokie,

Illinois, 64 pp.

Reynolds, C.E., 1960, Reinforced Concrete Designer's Handbook, Sixth

Edition, Concrete Publications Limited, London.

Rice, Paul F., 1984, "Structural Design of Concrete Sanitary Structures,"

Concrete International: Design and Construction, V. 6, No. 10, Oct., pp. 14-
16.
Varyani, V.H., and Radhaji, A., 1978, "Analysis of Long Concrete Buildings
for Temperature and Shrinkage Effects," Journal of the Institution of
Engineers (India), V. 59, Part CI 1, July, pp. 20-30.
Wood, Roger H., 1981, "Joints in Sanitary Engineering Structures," Concrete
International: Design and Construction, V. 3, No. 4, Apr., pp. 53-56.
 33

Table 1 - Contraction joint spacings

Author Contraction Joint Spacing

Merrill (1943) 20 ft for walls with frequent openings, 25 ft.

in solid walls

Fintel (1974) * 15 to 20 ft for walls and slabs on grade

Wood (1981) 20 to 30 ft for walls

PCA (1982) 20 to 25 ft for walls depending on number of

openings

ACI 350R-83 (1983) 30 ft in sanitary structures

ACI 224R-80 (Revised 1984) one to three times the height of the wall in
solid walls

* Recommends joint placement at abrupt changes in plan and at changes in

building height to account for potential stress concentrations.
1 ft = 0. 305 m

Table 2 - Contraction joint spacings for sanitary engineering structures

based on reinforcement percentage (Rice 1984)

Contraction joint spacing Minimum percentage of shrinkage

in ft. and temperature reinforcement*
Grade 40 Grade 60
less than 30 .30 .25

30 - 40 .40 .30
40 - 50 .50 .38
greater than 50 .60 .45
* minimum temperature and shrinkage reinforcement should be #5 (16 mm) bars
or D-31 (200 mm 2 ) wire, 12 in. (0.305 m) on center each face.
1 ft = 0.305 m; Grade 40 = 276 MPa; Grade 60 = 413 MPa.
 34

Table 3-Expansion joint spacings

Author Expansion Joint Spacing

Lewerenz (1907) 75 ft for walls

Hunter (1953) 80 ft for walls and insulated roofs
30 to 40 ft for uninsulated roofs
Billig (1960) * 100 ft maximum building length without joints
Wood (1981) 100 to 120 ft for walls

Indian Standards (1964) 45 m (approx. 148 ft) maximum building length

Institution between joints

PCA (1982) 200 ft maximum building length without joints

ACI 350R-83 120 ft in sanitary structures partially filled

with liquid (closer) spacings required when no
liquid present)

* Recommends joint placement at abrupt changes in plan and at changes in

building height to account for potential stress concentrations.
1 ft = 0.305 m

Table 4-Beam-column frame geometry used for temperature and

vertical load analysis (Hartin and Acosta 1970)

Beam Square column

Width Depth Span Side Height
Structure Type in. in. ft in. ft
----------------------------------------------------------------
A 14 20 20 14 12
B 14 20 20 14 20
c 14 20 20 18 12
D 14 20 20 18 20

E 14 20 30 14 12
F 14 20 30 14 20
G 14 20 30 18 12
H 14 20 30 18 20

I 14 30 30 14 12
J 14 30 30 14 20
K 14 30 30 18 12
L 14 30 30 18 20

in. = 25.4 mm; 1 ft 0.305 m

 35

Table 5-Maximum and minimum daily temperatures for

given locations (Martin and Acosta 1970)

Normal daily temperatures °F

Location Maximum Minimum

Anchorage, Alaska 66.0 4.3

Atlanta, Ga. 87.0 37.1
Boston, Mass. 81.9 23.0
Chicago, IlL 84. 1 19.0
Dallas, Tex. 95.0 36.0
Denver, Colo. 88.4 1 4.8
Detroit, Mich. 84.7 19. 1
Honolulu, Hawaii 85.2 65.8
Jacksonville, Fla. 92.0 45.0
Los Angeles, Calif. 75.9 45;0
Miami, Fla. 89.7 57.9
Milwaukee, ·wis. 78.9 12.8
New Orleans, La. 90.7 44.8
New York, N.Y. 85.3 26.4
Phoenix, Ariz. 104.6 35.3
Pittsburgh, Pa. 83.3 20.7
San Fransisco, ·calif. 73.8 41.7
San Juan, P.R. 85.5 70.0
Seattle, Wash. 75;6 33.0
St. Louis, Mo. 89;2 23.5
Tulsa, Okla. 93. 1 26.5

Temperature in C = i(Temperature in F-32)

Table 6-Temperatures used for calculation of AT
(National Academy of Sciences 1974)

Location Temperature °F
T T T
--------------------------------~------~------2------
Alabama
Birmingham 97 63 19
Huntsville 97 61 13
Mobile 96 68 28
Montgomery 98 66 22

Alaska
Anchorage 73 51 -25
Barrow 58 38 -45
Fairbanks 82 50 -53
Juneau 75 48 -7
Nome 66 45 -32
Arizona
Flagstaff 84 58 0
Phoenix 108 70 31
Prescott 96 64 15
Tuscon 105 67 29
Winslow •97 67 9
Yuma 111 72 37
Arkansas
Ft. Smith 101 65 15
Little Rock 99 65 19
Texarkana 99 65 22

T ~ temperature exceeded, on the average, only 1 per-

cent of the time during the summer months of June
through September

T = temperature during the normal construction season

T = temperature equaled or exceeded, on the average,
99 percent of the time during the winter months of
December through February

Temperature in C = ~(Temperature in F-32)

 37

Table 6, continued

Location Temperature °F
T T T
--------------------------------~------~------2 ______
California
Bakersfield 103 65 31
Burbank 97 64 36
Eureka/ Arcata 67 52 32
Fresno 101 63 28
Long Beach 87 63 41
Los Angeles 94 62 41
Oakland 85 57 35
Sacramento 100 60 30
San Diego 86 62 42
San Fransisco 83 56 35
Santa Maria 85 57 32
Colorado
Alamosa 84 60 -17
Colorado Springs 90 61 •1
Denver 92 62 -2
Grand Junction 96 64 8
Pueblo 96 64 -5
Connecticut
Bridgeport 90 60 4
Hartford 90 61 1
New Haven 88 59 5
Delaware
Wilmington 93 62 12
Florida
Daytona Beach 94 70 32
Ft. Myers 94 74 38
Jacksonville 96 68 29
Key West 90 77 55
Lakeland 95 72 35
Miami 92 75 44
Miami Beach 91 75 45
Orlando 96 72 33
Pensacola 92 68 29
Tallahassee 96 68 25
Tampa 92 72 36
West Palm Beach 92 75 40
Georgia
Athens 96 61 17
Atlanta 95 62 18
Augusta 98 64 20
Columbus 98 65 23
 38

Tab1e 6, continued

Location Temperature °F
T T T
--------------------------------~------~------2 ______
Georgia (cont'd.)
Macon 98 65 23
Rome 97 62 16
Savannah/Travis 96 67 24
Hawaii
Hilo 85 73 59
Honolulu 87 76 60
Idaho
Boise 96 61 4
Idaho Falls 91 61 -12
Lewiston 98 60 6
Pocatello 94 60 -8
Illinois
Chicago 95 60 -3
Moline 94 63 -7
Peoria 94 61 -2
Rockford 92 62 -7
Springfield 95 62 -1
Indiana
Evansville 96 65 6
Fort Wayne 93 62 0
Indianapolis 93 63 0
South Bend 92 61 -2
Iowa
Burlington 95 64 -4
Des Moines 95 64 -7
Dubuque 62 63 -11
Sioux City 96 64 -10
Waterloo 91 63 -12
Kansas
Dodge City 99 64 3
Goodland 99 65 -2
Topeka 99 69 3
Wichita 102 68 5
Kentucky
Covington 93 63 3
Lexington 94 63 6
Louisville 96 64 8
 39

Table 6, continued
Location Temperature °F
T T T
--------------------------------~------~------2 _____ _
Louisiana
Baton Rouge 96 68 25
Lake Charles 95 68 29
New Orleans 93 69 32
Shreveport 99 66 22

Maine
Caribou 85 56 -18
Portland 88 58 .;.5

Maryland
Baltimore 94 63 12
Frederick 94 63 7

Massachusetts
Boston 91 58 6
Pittsfield 86 58 -5
Worcester 89 58 -3
Michigan
Alpena 87 57 -5
Detroit-Metro 92 58 4
Escanaba 82 55 -7
Flint 89 60 -1
Grand Rapids 91 62 2
Lansing 89 59 2
Marquette 88 55 -8
Muskegon 87 59 4
Sault Ste Marie 83 55 -12

Minnesota
Duluth 85 55 -19
International Falls 86 57 -29
Minneapolis/St. Paul 92 62 -14
Rochester 90 60 -17
St. Cloud 90 60 -20

Mississippi
Jackson 98 66 21
Meridian 97 65 20
Vicksburg 97 66 23
Missouri
Columbia 97 65 2
Kansas City 100 65 4
St. Joseph 97 66 -1
St. Louis 98 65 4
Springfield 97 64 5
 40

Table 6. continued
Location Temperature °F
T T T
--------------------------------~------~------2 ______
Montana
Billings 94 60 -10
Glasgow 96 60 -25
Great Falls 91 58 -20
Havre 91 58 -22
Helena 90 58 -17
Kalispell 88 56 -7
Miles City 97 62 -19
Missoula 92 58 -7

Nebraska
Grand Island 98 65 -6
Lincoln 100 64 -4
Norfolk 97 64 -11
North Platte 97 64 -6
Omaha 97 64 -5
Scottsbluff 96 62 -8

Nevada
Elko 94 61 -13
Ely 90 59 '-6
Las Vegas 108 66 23
Reno 95 62 2
Winnemucca 97 63 1

New Hampshire
Concord 91 60 -11

New Jersey
Atlantic City 91 61 14
Newark 94 62 11
Trenton 92 61 12

New Mexico
Albuquerque 96 64 14
Raton 92 64 -2
Roswell 101 70 16
New York
Albany 91 61 -5
Binghampton 91 67 -2
Buffalo 88 59 3
New York 94 59 11
Rochester 91 59 2
Syracuse 90 59 -2
 41

Table 6, continued
Location Temperature °F
T T T
--------------------------------~------~------2------
North Carolina
Asheville 91 60 13
Charlotte 96 60 18
Greensboro 94 64 14
Raleigh/Durham 95 62 16
Wilmington 93 63 23
Winston/Salem 94 63 14
North Dakota
Bismarck 95 60 -24
Devils Lake 93 58 -23
Fargo 92 59 -22
Minot 91 -24
Williston 94 59 -21
Ohio
Akron/Canton 89 60 1
Cincinnati 94 62 8
Cleveland 91 61 2
Columbus 92 61 2
Dayton 92 61 0
Mansfield 91 61 1
Sandusky 92 60 4
Toledo 92 61 1
Youngstown 89 59 1
Oklahoma
Oklahoma City 100 64 11
Tulsa 102 65 12
Oregon
Astoria 79 50 27
Eugene 91 52 22
Medford 98 56 21
Pendleton 97 58 3
Portland 91 52 21
Roseburg 93 54 25
Salem 92 52 21
Pennsylvania
Allentown 92 61 3
Erie 88 59 7
Harrisburg 92 61 9
Philadelphia 93 63 11
Pittsburgh 90 63 5
Reading 92 61 6
 42

Table 6, continued
Location Temperature °F

-------------------------------~~-----~~-----:2 _____ _
Pennsylvania, cont'd.
Scranton/Wilkes-Barre 89 61 2
Williamsport 91 61 1

Rhode Island
Providence 89 60 6

South Carolina
Charleston 95 66 23
Columbia 98 64 20
Florence 96 64 21
Greenville 95 61 19
Spartanburg 95 60 18

South Dakota
Huron 97 62 -16
Rappid City 96 61 -9
Souix Falls 95 62 -14

Tennessee
Bristol/Tri City 92 63 11
Chattanooga 97 60 15
Knoxville 95 60 13
Memphis 98 62 17
Nashville 97 62 12

Texas
Abilene 101 65 17
Amarillo 98 66 8
Austin 101 68 25
Brownsville 94 74 36
Corpus Christi 95 71 32
Dallas 101 66 19
El Paso 100 65 21
Fort Worth 102 66 20
Galveston 91 70 32
Houston 96 68 28
Laredo AFB 103 74 32
Lubbock 99 67 11
Midland 100 66 19
Port Arthur 94 69 29
San Angelo 101 65 20
San Antonio 99 69 25
Victoria 98 71 28
Waco 101 67 21
Wichita Falls 103 66 15
 43

Table 6, continued
Location Temperature °F
T T T
--------------------------------~------~------2------
Utah
Salt Lake City 97 63 5
Vermont
Burlington 88 57 -12
Virginia
Lynchburg 94 62 15
Norfolk 94 60 20
Richmond 96 64 14
Roanoke 94 63 15
Wahington, D. c.
National Airport 94 63 16

Washington
Olympia 85 51 21
Seattle 85 51 20
Spokane 93 58 -2
Walla Walla 98 57 12
Yakima 94 62 6
West Virginia
Charleston 92 63 9
Huntington 95 63 10
Parkersburg 93 62 8

Wisconsin
Green Bay 88 59 -12
La Crosse 90 62 -12
Madison 92 61 -9
Milwaukee 90 60 -6

Wyoming
Casper 92 59 -11
Cheyenne 89 58 -6
Lander 92 58 -16
Sheridan 95 59 -12
 44

Table 7-Results of Temperature Analysis

(National Academy of Sciences 1974)

(1) ( 2) Beam Column

Anal. Col. Beam Base Total No. Max. Max. Max. Max.
No.· Size Size Type Length Flrs. Moment Axial Load Moment Shear
in. in. 2 ft. in.-kips kips in.-kips kips
----------------------------------------------------------------------------
1-1 24 280 F 200 3 70 135 600 69
2-1 24 280 H 200 3 45 77 250 19

3-1 16 280 F 200 3 53 70 170 22

4-1 16 280 H 200 3 25 25 75 6

A-1 24 280 F 200 2 182 570 67

A-2 24 1000 F 200 2 70 321 750 95

B-1 24 280 F 400 2 128 280 840 90
M-1 24 280 F 200 3 100 253 740 83

M-2 24 280 F 200 3 50 138 500 60

(1) Column size= one side of a square column.

(2) Beam size = cross-sectional area. Moment of inertia about the vertical
axis= 4667 in.• for all cases.

M-1 analysis includes columns at one end of the frame substantially stiffer
than the rest of the columns.

M-2 analysis includes hinges placed at the top and bottom of the exterior
columns.

F columns fixed at the foundation.

H columns hinged at the foundation.

1 in. = 25.4 mm; 1 ft = 0.305 in.; 1 in.-kip 113 m-kN; 1 kip= 4.45 kN
 45

Fig. 1 -Wall expansion joint cover (courtesy Architectural Art Mfg., Inc.)

Fig. 2- Fire rated filled expansion joint (courtesy Architectural Art Mfg.,
Inc.)
 46

112,000
Lj =
RAT

--
~

...J
400

-..,
tf)•

c
0 300
c
0
·c;;
c
cu
a.
X
UJ
200
c
CD

-CD
:=
CD
(l)

-
..r::
0)
c
CD
....1
100
R=4
R=5
A

0
0 20 40 60 80 100
Design Temperature Change, T (°F)

Fig. 3 - Length between expansion joints vs. design temperature change, ~T

(Martin & Acosta 1970)(1 ft = 0.305 m.; 1"F = t•c)
 47

~Frame

1-l+l+ltl+l+l-1 1--J----1
Corner
Column

MULTI-BAY FRAME ONE BAY

SUBSTITUTE
FRAME

Fig. 4- Multi-bay frame and one bay substitute frame (after Varyani &
Radhaj i 1978)
 48

•I

Kc K__....
c

Mt Mt
Mg = (2+r' l Mg = 2(1+r')

SINGLE-STORY SINGLE BAY SINGLE-STORY MULTI-BAY

I• ·I
h

h h

Mt Mt
Mg • (4+r'l Mg = 2(2+r')

MULTI-STORY SINGLE BAY MULTI-STORY MULTI-BAY

M f = Beam fixed-end moment

Kb
r'- r-1 - K
c

Fig. 5- Moments at base of corner columns due to gravity using one bay
substitute frame::: (after Varyani & Radhaji 1978)
 49

I
\
------------ IKb
-1
\ I h
\ I
Kc Kc

Mr<base) = 6E 0 K0 h6 ( 21+r'
+r'
)

SINGLE-STORY SINGLE BAY SINGLE-STORY MULTI-BAY

6 6 6 },
•I
•I I• •I I• · I I·
I I I
I I
I I
I
I I h II
I
I
\(-
\
-------- -...,.
Kb
/
I
I I
'("
'' Kb
\ I h h \
Kc

Mr<basel = 6E0 Kc h6 (3+r')

4 +r' Mr<basel = 3E 0 K 0 h6 { 3+2r'
2 +r'
)
MULTI-STORY SINGLE BAY MULTI-STORY MULTI-BAY

Kb
r' = r-1 = -
Kc

6 = ~ aLJ'~~T

Fig. 6 - Moments at base of corner columns due to temperature change using

one bay sub-stitute frame, L. = total length between expansion
joints (after Varyani & Radhaj1 1978).
 50

Rectangular Multi-Framed

-·-
~

...J
600 Configuration with
Symmetrical Stiffness

-Ol
•
.t::

cQ) 400
Non-Rectangular
Steel
Concrete
...J
Ol Configuration
c
J2
·:;
co 200 Any Material
Q)
.0
CCI
3:
0
<
0
0 20 40 60 80
Design Temperature Change, T (°F)

Fig. 7- Expansion joint criteria of the Federal Construction Council

(National Academy of Sciences 1974)(1 ft = 0.305 m; 1°F = ~°C)
 51

c:
G)
600
G)

........
S::;:;
a>~ -
all/)
Q)
....c:
g·o 40o /Heated Buildings
-
~Unheated
cu..,
I I) c:
i5 .2
II)

§ ~ Buildings/
E
·x w
e- 200 -
cu
:2 I I I I I I I I I
0 20 40 60 80
Temperature Change (°F)

Fig. 8- Expansion joint criteria of one federal agency (National Academy of

Sciences 1974) (1 ft = 0.305 m; 1 °F = i°C)
 52

~
Location of Maximum Axial Forces

nr nr 7/T nr TJT
. 7Jr 7) r .
Locat1on of Max1mum Bending
nr nr ~r
Moments and Maximum Shears
FRAME FIXED AT FOUNDATION
; ILocafon
I of Ma XI·m urn A x1a
. I Forces

7!>r r.>r 7'r -r.>r r>r -r>r -r. ~ 7l >r ~ 'r >r
Location of Maximum Banding;,.
Moments and Maximum Shears
FRAME HINGED AT FOUNDATION

Fig. 9 - Frames ,subjected to a uniform temperature change (National Academy

of Sciences 1974)
 53

APPENDIX A - NOTATION

A cross-sectional area of the element under consideration;

D unfactored dead load;

Ec modulus of elasticity of concrete;

F force resulting in a restrained member due to a temperature change,

aT;
f I specified compressive strength of concrete;
c
tensile capacity of concrete;

column height;

moment of inertia of the beam;

moment of inertia of the column;

beam stiffness factor;

column stiffness factor;

L unfactored live load;

1 beam length;

total length of building between expansion joints;

fixed-end moments at the beam-column joints due to gravity loading,

unfactored;
column moment due to gravity, unfactored;

column moment due to temperature, unfactored;

r ratio of the stiffness factor of the column to the stiffness factor of
the beam Kc/Kb;
r'

T unfactored thermal load;

temperature equaled or exceeded, on the average, 99 percent of the

time during the winter months of December, January, and February;

temperature during the normal construction season in the locality of

the building;
 54

APPENDIX - NOTATION, cont'd.

Tmax extreme normal daily maximum temperature;

extreme normal daily minimum temperature;

temperature exceeded, on the average, only 1 percent of the time

during the summer months of June through September;

u required strength of the element under consideration;

w uniformly distributed load;

coefficient of thermal expansion for concrete;

design temperature change; and

lateral frame deflection between expansion joints due to temperature

change ~T.
 55

APPENDIX B - EXPANSION JOINT EXAMPLES

Example 1: Single Story-Multi Bay Building

Given:

A one-story, 20 ft high warehouse is to be constructed in St. Louis,

Missouri. The building is to be a reinforced concrete frame with plan

dimensions of 210ft (east-west) by 600ft (north-south). Columns are

spaced at 30 ft in each direction.

Design Parameters:

Roof dead load is 5 psf (not including slab) and roof live load is 20

psf. Concrete design strength, f' = 6 ksi

c
Assumptions:

Architectural and preliminary design considerations result in the fol-

lowing member sizes: 5 in. roof slab thickness, 12 in. x 20 in.

interior beams, 15 in. x 20 in. perimeter beams, 18 in. x 20 in. gir-

ders, and 14 in. x 14 in. columns.

Typical Corner Section:

...... "IT" """ -

f===)?=======n=======j]======¢ 1'1 "''i
II II II 11 COLUMN
II II II II
II II II II A
II II II II 1r
'oil
ll 11 o
lfrc
11 2
11.
II11!'1o N
I"> II II~ II !:! II.;
II II- II II-
II11 II
II
II
11
II
II
tt II II II
II II II II
F--=6=-----.JL...!§..•A-ll-----:::6
-- ------3;,(ii07": 30·---
 56

Required:

Using each of the three methods described in the text, determine the

maximum length between expansion joints. Base calculations on un-

cracked gross section properties.

Martin and Acosta (1970)

Step 1: Calculate member properties:

a) Column and beam moments of inertia, I:

14"
'1'2 = 3201 in. •

( 15)( 20).
= 1 0, 000 in."
12

3
(18)(20)
E-W: Ib = 12
= 12,000 in.•

b) Column and beam stiffness factors, K:

I 3201 in. •
c
K =-
c h = 20 ft X 12 = 13.34 in. '

Ib 10,000 in. •
N-S:
~ -i = 30 ft X 12
= 27.78 in. '

Ib 12,000 in. •
30 ft X 12 = 33.33 in. '
E-W: Kb = r-=

c) Ratio of column to beam stiffness factor, r:

Kc 13.34 in.•
N-S: r = -Kb = ~~--7'-'-'T =
27.78 in.'
0. 480
 57

13.34 in.'
E-W: r = .;,33""'.~3;:::-3...:1;..:_n'-".•• = 0 • 400

Step 2: Calculate design temperature changes, 6T, using Eq. (4):

6T = g(T - T . ) + 30°F ( 4)
3 max m1n

Substituting from Table 5 for St. Louis, MO,

T = 89.2°F
max

Tmin = 23.25°F

= 73.8°F

Step 3: Calculate value of R, using Eq. (6):

1
c (1 + r ) ( 6)
R = 1 44 tiT 1 + 2r

Substituting h =20ft x 12 in., and Ic and r from Step 1,

N-S: R
= (144)(3201 in.') ( 1 + 0.480 ) = 6.043
(20 ft X 12) 2 1 + (2)(0.480)

E-W: R (144)(3201 in.') ( 1 + 0.400 )

(20ft X 12) 2 1 + (2)(0.400) 6.224

Step 4: Calculate expansion joint spacing, Lj, using Eq. (5):

112, 000
Lj = R 6T ( 5)
 58

Substituting R (Step 3) and ~T (Step 2),

112,000
(6.043)(73.8) - 251.1 ft

112' 000
E-W: Lj = ( 6 • 224 )(
73 .B) = 243.8 ft

Step 5: Compare expansion joint spacing from Step 4 to the limitation ex-

pressed in Eq. (8):

.- 2000h (8)
L
j " ~T

L :; 2000(20) = 542.0 ft
j 73.8

The values obtained in Step 4 control.

Varyani and Radhaji (1978)

Step 1: Calculate member properties:

See Step 1 of Martin and Acosta

Step 2: Calculate r•:

1
r' r

Substituting r from Step 1 of Martin and Acosta

1
N-S: r' = 0.480 = 2.083

1
E-W: r• = ~'-::-:- = 2.500
0.400
 59

Step 3: Calculate design temperature change, 6T, using Eq. (20a) and

(20b):

6T = 1[-
2
~(T
3 max
- T . ) - 27°F]
mln
(20a)

Substituting values for Tmax and Tmin (see Step 2 of Martin and

Acosta)

1 2
6T = 2[- 3(89.2 - 23.5) - 27°F]

= -35.4°F for contraction

6T = l[~(T
2 3 max
- T . ) - 14°F]
mln
(20b)

= 29.1°F for expansion

Note that although the absolute value of 6T is greater for contraction,

the value obtained for expansion will usually control since it adds to
the negative gravity moments at exterior beam-column connections.

Step 4: Calculate factored fixed-end moments, Mf:

Design parameters result in the following factored loads.

N-S: W (dead)
u
= 0.857 kips/ft
 60

W (live) = 0. 151 kips/ft

u

E-W: W (dead) = 0. 525 kips/ft

u
Wu(live) = 0

R (dead) = 1 8. 0 kips (beam reactions)

u
R (live) = 5. 1 kips (beam reactions)
u

(0.857)(30) 2
N-S: Mf(dead) = = 64.3 ft-kips
12

(0.151)(30) 2
Mf(live) = 11.3 ft-kips
12

= (0.525)(30) 2 + _,_(~18~)~(~1~0)~(~2~0~)__ +
2
(18)(10)(20) 2
12 ~ 30 30 2

159.4 ft-kips

2
= (5.1)(10)(20)
30 2
+
(5.1) (10) (20) 2
30 2

34.0 ft-kips

Step 5: Calculate factored gravity moments, M , at the base and top of the
g
column using Eq. (13) and (27):
Base

( 1 3)

Substituting Mf (Step 4) and r• (Step 2),

N-S: Mg ( base ) ( 13 + +2 11
= 264. • 3) = 1 2 .2 6 f t-kips
• 083
 61

E-W: M (base) 159.4 + 34.0 27.63 ft-kips

=
g 2( 1 + 2.500)
Top
Mf
Mg(top) =
( 1 + r') = 2 Mg(base) (27)

N-S: M (top)
g
= 24.52 ft-kips

E-W: M (top)
g
= 55.26 ft-kips

Step 6: Using Eq. (9) and (10), calculate the unfactored temperature

induced moment, MT, at base and top of the column, assuming column

design is governed by the factored gravity moment obtained at the

top of the column, Mg(top). This procedure is based on the as-

sumption that the entire column, including the connection at the

base, is proportioned based on the maximum gravity moment, which

occurs at the top of the column.

U 1.4D + 1.7L ( 9)

U = .75(1.4D + 1.7L + 1.4T) (10)

Substituting U = 1.4D + 1.7L = Mg and T = ~·

M (top) = 0.75 [M (base or top) + 1.4MT(base or top)]
g g
Base

N-S: 24.52 = 0.75(12.26 + 1.4 MT)

~(base) = 14.60 ft-kips

E-W: 55.26 = 0.75(27.63 + 1.4 MT)

~(base) = 32.89 ft-kips

 62

Top

N-S: 24.52 = 0.75(24.52 + 1.4 MT)

My(top) = 5.84 ft-kips

E-W: 55.26 = 0.75(55.26 ft-kip + 1.4 MT)

My(top) = 13.17 ft-kips

Step 7: Using Eq. (22), calculate expansion joint spacing, Lj, based on

the column moments at the base of the column:

2
MT (1 + r' ) ( 22)
Lj = ""3-:::E"""K~ct-:b""'T 1 + 2r '
c c

Substituting MT(base) from Step 6, h = 20 ft x 12, r' (Step 2), Ec

= 57,000/6000 = 4,415,201 psi (ACI 318), Kc (Step 1), ct= 5.5 x

10
-6 , and bT = 29.1°F (Step 3),

( 1+2.083 )
~(2~)~(~1~4~.6~0~x~1~2,~0~0~0~)(~2~0~x~12~)~1_+(~2~) (~2.~0~8~3~)
=
7
(3)(4,415,201)(13.34)(5.5x10- 6 )(29.1)

= 1770 in. = 147.5 ft

( 1 + 2. 500 )
(2)(32.89x12,000)(20x12) 1+(2)(2.500)
(3)(4,415,201)(13.34)(5.5x10- 6 )(29.1)

= 3904 in. = 325 ft

Step 8: Using Eq. (28), calculate expansion joint spacing, Lj, based on the

column moments at the top of the column:

 63

MT(top) = 6E K .§. ( r' ) (28)

c c h 1 + r'

Substituting Eq. (7) for o into Eq. (28) and solving for Lj
results in the equation for calculating Lj directly,

( 1 + r' )
3E K a~T 1 + 2r'
c c

Substituting MT(top), h, r', E, K, and ~T gives,

c c

( 1 + 2. 083)
= _ _;('-'<5:.:..·.:::.84:.:.x::.. :1.:::2L.,0::.:0:.::0:.:. )-'-'(2::.:0:.::x:..:.1.::.2:.. )_.::.2:..:.O'-i8;.:e3_ __
6
(3)(4,415,201)(13.34)(5.5x10- )(29.1)

= 879.5 in. = 73.3 ft

( 1 + 2. 500)
= _,_(1~3~·_.:_1-"7x,_1:..::2:.!•..:::0.:::.0.:::.0):.._(:..::2:..::0"'x-'-12::.:l:___::2:..c·5::_,0;::0:___ _
(3)(4,415,201)(13.34)(5.5x10- 6 )(29.1)

= 1878.3 in. = 156.5 ft

Step 9: Summary of expansion joint spacings, Lj:

a) Based on base of column moment calculations (Step 7)

N-S: L. = 1 47.5 ft
J
E-W: L. = 325.0 ft
J
b) Based on top of column moment calculations (Step 8)

N-S: Lj = 73.3 ft
E-W: Lj = 156.5 ft
 64

National Academy of Sciences (1974)

Step 1: Calculate design temperature changes, liT, using Eq. (33a) and

(33b):

liT = T - T ( 33a)
w m

Substituting from Table 6 for St. Louis, MO:

T = 65°F
m
liT = 98°F - 65°F

liT = T - T (33b)
m c
T = 65°F
m
T = 4°F
c
liT = 65°F - 4°F

= 61 °F

Therefore, use liT= 61°F

Step 2: Using Fig. 5, compute the allowable building length assuming liT =

61 °F:
From Figure 5, allowable building length is approximately 400 ft

Step 3: Apply modification factors:

Fixed column bases, decrease length by 15%.

Lj =400ft- (400)(0.15)
= 340 ft
 65

Example 2: Multi-Story, Multi-Bay Building

Given:

A four-story, 56 ft high business complex is to be constructed in St.

Louis, Missouri. The building is to be a reinforced concrete frame

with plan dimensions of 210 ft (east-west) by 600 ft (north-south).

Columns are spaced at 30 ft in each direction. Each story is 14 ft

high.

Design parameters:

Roof dead load is 5 psf, roof live load is 20 psf, and floor live load

is 125 psf. Concrete design strength, f' = 6 ksi.

c
Assumptions:

Architectural and preliminary design considerations result in the fol-

lowing member sizes: 5 in. slab thickness (roof and floors), 12 in. x

24 in. interior beams, 18 in. x 24 in. perimeter beams, 24 in. x 24 in.

girders, and 24 in. x 24 in. columns.

Typical Corner Section:

l. ,. ...... ""'
r:=y=====ty=====~=====~
II II II II
II II II 11
II II II 11
II
II
II
II
II
II
II
II :to ~
. II II~ II~ II';; N
~~ II~ II ~ lloo
-,r II~ II!:! rr-
ll II II II
ll11 II
II
II
II
II
II
II II II II
II II II II
fo=~====dk====~=====6
- 3 $P e 10' • 30'
 66

Required:

Determine the maximum length between joints.

Varyani and Radhaji (1978)

Step 1: Calculate member properties;

a) Column and beam moment of inertia, I,

24 4
Ic = ~ = 27,648 in.'

(18)(24) 3
12
= 20,736 in.~~.

E-W: Ib = ( 24 )~~ 4 )' = 27,648 in. 4

b) Column and beam stiffness factors, K:
1
c 27,648
Kc = ~ = 14x 12 = 164.57 in.•

Ib
N-S: Kb = -= 20,736 = 57.6 in. •
t 30x12

Ib
E-W: Kb = - = 27,648 = 76.8 in. •
t 30x12

Step 2: Calculate r' :

Kb 57.6
N-S: r• = i<= 164.57 = 0. 350
c

Kb 76.8
E-W: r' = -K = 164.57 = 0.467
c
 67

Step 3: Calculate design temperature change, 6T, using Eq. (20a) and

(20b): See Example 1.

6T = 29.1 °F

Step 4: Calculate fixed-end moments, Mf:

Design parameters result in the following factored loads,

N-S: W (dead) = 1 • 01 3 kips/ft

u
W (live) = 0.929 kips/ft
u

E-W: W (dead) = 0.840 kips/ft

u
W (live)
u
= o kips/ft
R (dead) = 1 8. 1 kips
u
R (live) = 31.9 kips
u

= ( 1 " 01 ~~( 3 0) = 75.98

2

N-S: Mf(dead) ft-kips

(0.929)(30) 2
12
= 69.68 ft-kips

2
- 0:..:•.. :.8_;.;40~)-=("'30:..:.)_ +
2
( 18. 1 ) ( 10 )( 20) (18.1)(10)(20) 2
E-W: Mf(dead) = .:...< 2 +
12 30 30 2

= 183.75 ft-kips

( 31 • 9) ( 10) ( 20) 2 2
+ _,_(3"'1c.::•..::.9.:,.)::-;<2;:..::0"-)_,_(1.:..;0:;.!)_
Mf(live) =
302 30 2

= 212.5 ft-kips
 68

Step 5: Calculate factored gravity moments, Mg' at the base and top of the

column using Eq. (15)'and (31):

Base
Mf
M (base) = "'27( 2-:c'+'-r-;'') ( 15)
g

Substituting Mf (Step 4) and r' (Step 2),

75.98 + 69.68
N-S: Mg(base) =
2(2 + .350) = 30.99 ft-kips

183.75 + 212.5
E-W: M (base) =
2(2 + .467) = 80.31 ft-kips
g
Top
Mf
M (top) =
( 2+r' ) 2M g (base) (31)
g

N-S: M (top) 61.98 ft-kips

g

E-W: M (top)
g
= 160.62 ft-kips

Step 6: Using Eq. (9) and (10), calculate temperature induced moment, MT,

at the base and top of the column assuming column design to be

governed by gravity moment obtained at the top of the column,

M (top). Follow the same procedure as used in Step 6 of Varyani
g
and Radhaji in Example 1.

Base

N-S: 61.98 = 0.75(30.99 + 1.4 MT)

~(base) = 36.89 ft-kips

E-W: 160.62 = 0.75(80.31 + 1.4 MT)
 69

MT(base) = 95.61 ft-kips

Top

N-S: 61.98 ft-kip = 0.75(61.98 + 1.4 MT)

~(top) = 14.76 ft-kips

E-W: 160.62 ft-kip = 0.75(160.62 + 1.4 MT)

~(top) = 38.24 ft-kips

Step 7: Using Eq. (24), calculate expansion joint spacing, Lj, based on

the column moments at the base of the column.

2MTh (2 + r' )
(24)
3E c Kc a~T 3 + 2r'

14 ft x 12, r' (Step 2), E

Substituting MT(base) from Step 6, h =
c
-6
= 4,415,201 psi, K (Step 1), a= 5.5 x 10 , and ~T = 29.1°F
0

(Step 3),

( 2+0. 350 )
(2)(36.89x12,000)(14x12) 3+(2)(0.350)
(3)(4,415,201)(164.57)(5.5x10- 6 )(29.1)

= 222.4 in. = 18.5 ft

( 2+0.467 )
(2)(95.61x12,000)(14x12) 3+(2)(0.467)
=
6
(3)(4,415,201)(13.34)(5.5x10- )(29.1)

= 570.0 in. = 47.5 ft

 70

Step 8: Using Eq. (32), calculate expansion joint spacing, Lj, based on

the column moments at the top of the column,

M = 6E K .§. ( 1 + r') (32)

T c c h 2 + r'

Substituting Eq. (7) for o into Eq. (32) and solving for Lj
results in an equation for calculating Lj directly.

L = MT (2 + r')
j 3E K aAT 1 + r'
c c

Substituting MT(top), h, r', Ec' Kc' and AT gives

(2+0.350)
= --~(~1~4~.7~6~x~12~·~0~0~0~)(~1~4~x~12~)~1~+0~·~3 5~0_____
7
(3)(4,415,201)(164.57)(5.5x10- 6 )(29.1)

= 148 in. = 12.4 ft

(2+0.467)
= --~(~3~8~.2~4~x~12~·~0~0~0~)(~1~4~x~12~)~1~+0~.~476~7_____
6
(3)(4,415,201)(164.57)(5.5x10- )(29.1)

= 371 in. 30.9 ft

Step 9: Summary of expansion joint spacings, Lj:

a) Based on base of column moment calculations

N-S: Lj = 18.5 ft
E-W: Lj = 47.5 ft

b) Based on top of column moment calculations

 71

N-S: Lj = 12.4 ft
E-W: Lj = 30.9 ft

National Academy o£ Science (1974)

This method does not distinguish between single and multistory

structures. Thus, since the temperature changes and degree of column fixity

are the same for this structure as for the single story warehouse in Example

1, the required value of expansion joint spacing is the same.

Lj = 340 ft

Comparison and Discussion

Single Story Multi-Story

Martin and Acosta (1970)

N-S: Lj = 251. 1 ft Not Applicable

E-W: L. = 243.8 ft Not Applicable

J
Varyani and Radhaji (1978)

(Based on base of column) N-S: L. = 147.5 ft L. = 18.5 ft

J J
E-W: L. = 325.0 ft L. = 47.5 ft
J J
(Based on top of column) N-S: L. = 73.3 ft L. = 12.4 ft
J J
E-W: Lj = 156.5 ft Lj = 30.9 ft
National Academy of Sciences (1974}
N-S: L. 340 ft L. = 340 ft
J J
E-W: Lj = 340 ft L. = 340 ft
J

As demonstrated by these examples, the resulting expansion joint spa-

cings vary greatly between methods. In some cases, the answers are

ridiculously low. All three methods utilize variables that require the
 72

designer to make certain assumptions. These assumptions are critical and

may be a partial cause of the disparity between the results.

Martin and Acosta's (1970) method requires that the designer to decide

whether to use a cracked or uncracked moment of inertia. Using uncracked

moments of inertia results in lower values of Lj" As pointed out, Eq. (5)

is based on load factors [Eq. ( 1) and (2)] that are more conservative than

those now in use [Eq. (9) and (10)]. Conservative load factors will also

decrease the calculated value of L.. Assumptions concerning which modifica-

J
tion factors to use affect the results obtained with the National Academy of

Sciences (1974) method. Revising the original assumptions and modifying Eq.

(5) to account for the current more liberal load factors will lead to a

closer correlation between Martin and Acosta (1970) and the National Academy

(1974).

Varyani and Radhaji's (1978) method provides the most unrealistic

values of L. and is clearly sensitive to the design assumptions. Using a

J
cracked, instead of uncracked moment of inertia will increase the calculated

values of Lj, but the extremely short joint spacings are primarilY the

result of the premise that the combined factored gravity and thermal moments

[Eq. (10)] should not exceed the factored gravity moments [Eq. (10)].

Typical design practice often results in columns that have a structural

capacity that far exceeds the factored moments. If this is the case, a more

appropriate way of calculating the "usable" temperature induced moment, MT'

would be to equate the combined gravity and thermal effects to the actual

capacity of the column, rather than the factored gravity moment. This is

obtained by modifying Eq. (10) as follows:

$M (B1)
n
 73

Eq. (81) will yield longer, more realistic joint spacings than Eq. (10).

For a fixed building geometry (and Lj), MT is also fixed. In this case, ~M
n
can be increased to satisfy Eq. (81).

The preceding examples illustrate the strengths and weaknesses of the

proposed methods. The method requiring the least effort (National Academy

of Sciences 1974) produces joint spacings that are in line with current

practice. The most time-consuming method (Varyani and Radhaji 1978) can

produce unreasonably low values of Lj and appears to need additional

modifications before an acceptable version is available. The final deter-

mination of which method to use rests with the designer and must provide an

expansion joint spacing that will limit member forces without adversely af-

fecting structural integrity or serviceability.