A Review of Headed-Stud Design

Criteria in the Sixth Edition of the

PCI Design Handbook
The Precast/Prestressed Concrete Institute (PCI) is sponsoring a com-
prehensive research program to assess the shear capacities of groups
of headed-stud anchors. This program was initiated in response to new
provisions introduced in ACI 318–02, which were based on an exten-
sive database dominated by results of post-installed anchor tests. Tests
of headed-stud anchors loaded in shear, as used in precast concrete
construction, are not prevalent in the literature. This test program, con-
ducted by Wiss, Janney, Elstner Associates Inc., examines headed-stud
connections in several geometric configurations and edge conditions.
Neal S. Anderson, P.E., S.E. This paper provides a summary of the background studies and the
Consultant
Wiss, Janney, Elstner Associates Inc. research work that culminated in the design equations presented in
Northbrook, Ill. Section 6.5 of the sixth edition of the PCI Design Handbook.

I
n the precast concrete industry, pre- tural Concrete (ACI 318) Appendix
cast components are typically con- D2,3 have raised questions about the
nected by use of an embedded plate, older design models. Specifically, it
a majority of which are anchored with raises questions about the PCI design
welded-headed studs. Welded-headed model for headed-stud anchors, which
studs have been used to connect con- has been used successfully since 1971.
crete components to other structural el- The PCI design model was adopted by
ements for decades. In fact, formal de- ACI Committee 349, Concrete Nuclear
sign concepts for headed-stud anchors Structures, in its publication.4,5 This
have existed in the Precast/Prestressed model is collectively known as the 45-
Donald F. Meinheit, Ph.D., Concrete Institute’s (PCI) PCI Design degree-cone model. Testing and ana-
P.E., S.E. Handbook since the early 1970s.1 lytical studies in Germany in the 1980s
Senior Consultant Although concepts related to headed- led to the development of a design pro-
Wiss, Janney, Elstner Associates Inc. stud design that account for multiple cedure for headed-stud anchors known
Chicago, Ill.
anchors, variable spacing, or anchors as the Kappa method.6 Additional re-
close to free edges have existed for finements of the Kappa method pro-
years, recent design provisions in the duced the concrete capacity design
American Concrete Institute’s (ACI) (CCD) method, which is the basis of
Building Code Requirements for Struc- the ACI 318 Appendix D provisions.
 PCI JOURNAL
 In the mid-1990s, PCI initiated a headed-stud research installed anchors were generally designed using manufac-
program to create a capacity database for headed-stud group turers’ catalogs and/or procedures and an accepted factor of
connections. This research program responded to industry safety, usually between 3 and 4. The tabular design values in
concerns that the then-proposed provisions for ACI 318 Ap- the catalogs were based on standard concrete strengths and
pendix D on headed-stud connection design were more con- prescribed anchor patterns. In addition, some manufacturers
servative than the 45-degree model approach used in the PCI provided additional design guidelines for edge distance and
Design Handbook. Because the CCD method’s design pro- spacing effects; however, there was no uniform treatment of
cedures for tension and shear loading of headed studs were the design approach throughout the anchor industry on such
based on a limited amount of research data, the purpose of topics. Likewise, designers could not easily make numerical
PCI’s industry-sponsored research project was to satisfy two comparisons among different manufacturers’ designs. Visu-
primary points: alizing the tension or shear behavior of an anchor was also
• Provide justification for the PCI design procedures not readily apparent by looking in a table of ultimate and safe
used in the past, which through ACI 318 implementa- working capacities.
tion and adoption were now considered unconserva- The codification of the anchorage requirements was in-
tive; and tended primarily for the post-installed anchorage market,
• Create a database of test results to justify (a) accepting but cast-in anchors, including headed studs and bolts, were
and conforming to the provisions of ACI 318-02 Ap- also incorporated into the design provisions. These new pro-
pendix D, (b) modifying the ACI 318 procedures, (c) visions present a simple physical concrete breakout model,
refining the design procedures as published in the fifth which can readily accommodate the effects of anchor spacing
edition of the PCI Design Handbook,7 or (d) writing a in two directions and the effects of edge conditions. Hence,
new design procedure independent of ACI 318, which the model and procedures provide the design engineer with
is permitted in the code. the tools to design an anchor and readily consider many
This paper examines the background of the ACI 318 Ap- of the geometry and member influences that can affect its
pendix D design provisions and how the provisions apply to capacity.
headed-stud connections. It also reviews the evolution of the The ACI 318 Appendix D provisions are not necessar-
PCI design method through its various editions. Finally, the ily a one-size-fits-all design method; the behavior of post-
research work sponsored by PCI is summarized and the de- installed and cast-in-place anchors can be different, however,
sign provisions for headed-stud connections in the new sixth especially given the individual actions of shear, tension, or
edition of the PCI Design Handbook8 are presented. some combination thereof, coupled with the variability of
The testing program and experimental work in deriving the field installation versus plant production conditions. Because
equations has been presented in three papers to date.9–11 Ad- of the nature of past anchorage research in the United States
ditional background information and test data on the various and Europe, the CCD method was primarily based on a da-
influences, to substantiate the PCI Design Handbook provi- tabase dominated by post-installed-anchor test results. The
sions, will be presented in detailed research papers in future work sponsored by PCI was an effort to expand the database
issues of the PCI Journal. of headed-stud test results in order to confirm the CCD meth-
od’s applicability or provide new design guidelines that bet-
ter fit the plant-cast headed-stud anchor behavior. Continuing
research will provide additional information to improve the
Origins of ACI 318 Appendix D reliability of headed-stud design.

In ACI 318-02, the design provisions for anchorage were

finally codified into one document. The ACI 318 Appendix
D provisions are based on the CCD method,6 which was an PCI’s Research Initiative
adaptation of the original European Kappa method proposed
by Eligehausen and Fuchs in the late 1980s. The refined CCD The ACI design method was calibrated using an extensive
method was summarized in a paper by Eligenhausen, Fuchs, database of post-installed anchor test results. Logical rea-
and Breen in the ACI Structural Journal.12 This method was sons created this one-sided bias. Upon thorough review of
molded into ACI 318 due to successful use of the design pro- the published literature on headed studs and embedded bolts,
visions in Europe. The European code had used CCD con- the authors found many more results from tests performed
cepts for about 10 years prior to their incorporation into ACI on post-installed anchors loaded in shear compared with the
318-02. number of tests performed on headed studs and embedded
Anchorage design requirements were placed in ACI 318 bolts. Thus, there is a potential for the codified shear design
because a rational design method for post-installed anchors method to be biased toward post-installed anchors. The data-
was needed. Post-installed anchors are those that are installed base containing results from tension tests is more substantial
in hardened concrete. The need for these requirements was with regard to headed studs and embedded bolts. Although
obvious: Several anchor manufacturers do business in the there are no known systemic design problems with anchors
United States and, traditionally, the only source for the ca- designed using the pre-2002 procedures, it is fair to question
pacity of a given anchor was the manufacturers’ catalogs. the earlier design procedures given that the new and old de-
Prior to ACI’s codification of the design provisions, post- sign provisions yield different solutions. Consequently, the
January–February 2007 
Table 1. Review of Past PCI Design Handbook Requirements for Concrete Breakout Capacity

Handbook Edition Basic Concrete Capacity Equation Modification Factors

(Year Published)

First (1971) None

(
Vu' = φ 2500de − 3500 )

Second (1978) None

f c'
(
φVc = 3250φ de − 1 λ ) 5000
and

φVc ≤ φ Pc = φ Ao 4λ f c'

Third (1985) Away from an edge None

φVc = φ 800 Abλ f c'
Near a free edge
2
φVc = φ 2π de λ fc'

Fourth (1992) Away from an edge See below

⎛ ⎞
φVc = ⎜ φ 800 Abλ fc' ⎟ n
⎝ ⎠
Near a free edge
φVc = φVc' CwCt Cc

φVc' = φ12.5de1.5λ f c'

Fifth (1999) φVc = φVc' CwCt Cc with Thickness

⎛ h ⎞
Ct = ⎜ ⎟ ≤ 1.0
φVc' = φ12.5de1.5λ f c' ⎝ 1.3de ⎠

Group width
⎛ b ⎞
Cw = ⎜ 1+ ⎟ ≤ ns
⎝ 3.5d e⎠

Corner
⎛ d ⎞
Cc = ⎜ 0.4 + 0.7 c ⎟ ≤ 1.0
⎝ de ⎠

Notes: b = center-to-center distance between the outermost studs in the back row of the group (in.); dc = the distance, measured perpendicular to the direction of the
load, from the free edge of concrete to the center line of the nearest stud (in.); de = distance from free edge of concrete to back row of studs in direction of load (in.); h =
thickness of the concrete member (in.); ns = number of studs in the back row; λ = concrete unit weight factor (1.0 for normal weight, 0.85 for sand lightweight, 0.75 for
all lightweight); and φ = 0.85 (strength reduction factor).

 PCI JOURNAL
 same provisions for cast-in anchorages and post-installed an-
chorages may not be appropriate.
Comments PCI initiated research work on a large-scale testing pro-
gram to expand the database of headed-stud test results.
Phase 1 of this study examined headed-stud anchors loaded
in shear only. The Phase 1 testing program included 364 tests
of headed studs loaded in shear. In the comprehensive pro-
gram, the shear research variables included front-edge dis-
tance, corner conditions, side-edge distance, rear-edge dis-
tance, and in-the-field-type connections. Phase 2 of this study
Because a shear cone failure has been observed in shear tests: included a literature review of the tension-only test result da-
φVc = φPc
Vu
tabase compiled for headed-stud connections. In addition to
tension behavior, Phase 2 also included some experimental
studies to evaluate the effects of combined tension and shear
de

on a headed-stud connection. In the Phase 2 testing, 69 tests

were conducted that subjected headed-stud groups to differ-
@e
Note: Reinforcement ent combinations of tension and shear. Wiss, Janney, Elstner
omitted for clarity
Associates Inc. (WJE) conducted both of these project phases
in its Northbrook, Ill., structural laboratories.
While there have been no failures or serious design pro-
From the PCI Design Handbook: For groups of studs, the cedure problems reported that can be attributed to the use of
design shear strength, based on concrete strength, should be the design methods published in previous editions of the PCI
taken as the least of: Design Handbook, modifications to the welded, headed-stud
• Strength of the weakest stud, based on the above equa-
design procedures were made in the sixth edition in response
tions, times the number of studs,
to the new test data available through this research program.
• Strength based on the de of the weakest row of studs
times the number of rows, or
• Strength based on the de of the row of studs farthest
from the free edge. Background
Note: These are based on 9 “normal” arrangement of studs.
For arrangements that are very unsymmetrical or unusual, a
separate analysis, which considers the “zipper” effect, should Historical Perspective
be made. Concrete breakout behavior, as a function of edge dis-
tance, has been part of anchorage design for a number of
years. Reference should be made to Fig. 1, which illustrates
a typical headed-stud anchorage connection and the geo-
metric terminology used. For a majority of cast-in-place an-
chorage designs subject to shear, the front-edge distance de3
is the most critical factor governing the connection behavior
and strength. However, in precast concrete frame connec-
tions, consideration of the corner and side-edge distances is
also necessary.
Discussions of the equations used to predict concrete
breakout capacity over the past 30 years correspond well
to the provisions contained in the first five editions of the
The 800 equation was eliminated from the fifth edition. PCI Design Handbook. The progressive development of the
concrete breakout equations for headed-stud anchors in the
b United States and Europe is reflected in the various PCI
h Design Handbook editions. Thus, the handbook provisions
form an excellent basis to introduce and review the previ-
de
ous work, research, and design philosophies for concrete
breakout capacities.
dc V Since its inception in 1971, the PCI Design Handbook has
recognized the need for headed-stud connections in precast
concrete design. The first five editions of the handbook incor-
porated design guidelines for headed-stud connections, with
improvements to the design equations made in successive
editions. Aside from design information published by manu-
facturers, the information contained in the handbook repre-
sents some of the only recognized and published informa-
January–February 2007 
 Superior Concrete Products produced for the precast concrete
industry. As shown in Table 1, this early equation was quite
y simple. The only variable was the distance to the free edge.
Concrete strength, headed-stud area, and diameter were not
x
de4
included as variables in this equation.
(back-edge
distance)
Shear (V) direction The concrete-breakout equation contained in the second
edition of the PCI Design Handbook originated from work at
Lehigh University that was sponsored by Nelson Stud Weld-
y-row 2
ing.17 This equation is only a slightly modified version of the
y one developed by McMackin, Slutter, and Fisher (the PCI
y-row 1
equation has added a lightweight concrete factor λ).18 Also,
de3 the McMackin, Slutter, and Fisher equation was apparently
(front-edge
distance)
x-row 1 x-row 1 modified or normalized with respect to a 5000 psi concrete.
The f c' term was omitted from the Vc equation.
de1 x de2 In addition to checking the basic concrete shear capacity
(side-edge
distance)
(side-edge
distance)
defined by φVc, a secondary check was required in the second
edition of the handbook to ensure a tensile pullout (actually
Plan meaning tension concrete breakout) failure did not occur.
Therefore, φVc was to be checked against the tensile capacity
φPc. This edition of the handbook notes that this check was
instituted because cone-type failures had been observed in
hef
some shear tests, probably meaning the shear failure was a
h d pryout failure.
dh The third edition of the PCI Design Handbook saw the in-
troduction of the 45-degree–cone breakout model incorpo-
rated into the shear design provisions.19 This model assumed
Section thru slab a semi-conical failure surface defined by a height equal to the
edge distance and a 45-degree projected concrete breakout
surface to the free edge. The design equation included the
Fig. 1. Geometric terminology used to define a headed-stud
anchorage. effect of edge distance squared and the square root of the
concrete compressive strength. Originally proposed by the
Tennessee Valley Authority (TVA) in a design standard,20
tion on this subject during the 1970s and 1980s. During this this equation was adopted by ACI Committee 349 in the late
time, design equations for concrete breakout capacity were 1970s. In 1982, Klingner, Mendonca, and Malik reviewed the
included in an ACI Committee 349 report and in the Uni- literature on shear capacity of short anchor bolts and headed
form Building Code.13 As addressed earlier, however, only studs.21 Their work confirmed the applicability of this equa-
recently has a comprehensive design method for anchorage to tion. These three sources also note that the equation is appli-
concrete been incorporated into the widely accepted ACI 318 cable for fully embedded studs.
Building Code Requirements for Structural Concrete.3 The third edition of the PCI Design Handbook contained
Table 1 presents the concrete shear breakout design verbiage on computing the capacity of stud groups, which
equations provided in the first five editions of the PCI De- is reproduced in Table 1. Guidelines were presented in this
sign Handbook. These equations are applicable for concrete edition for determining the stud-group capacity based on
breakout when the anchorage is loaded in shear perpendicular the weakest stud, weakest row, or row farthest from the free
to a free edge. As shown in Table 1, the first three editions edge. Also, caution was given that unusual arrangements
of the handbook provided no modifiers to the basic concrete of studs should be analyzed to prevent zipper-type failures.
capacity for groups or spacing because there was a lack of re- Although this edition did not give guidance as to the nature
search on these influences at that time. Research at Oklahoma of this type of failure, the verbiage implies a sequential-type
State University (OSU) and in Europe brought about three failure based on the critical studs in an unusual arrangement.
modification factors that were included in the fourth and fifth The design engineer could garner additional guidance on the
editions of the handbook.14,15 These modification factors to zipper effect from the construction/design booklet Embed-
the basic concrete breakout capacity accounted for corner, ment Properties of Headed Studs,22 published by TRW Nel-
thickness, and group-spacing effects. son. Figures 2 and 3 reproduce the relevant figures from this
In the first edition of the PCI Design Handbook, the basic publication, showing the interaction of a stud group.
equation used to calculate concrete breakout when subject to For anchors away from an edge,23 the third edition intro-
shear was an extrapolation of a breakout equation developed duced the 800 coefficient shear capacity equation originally
by Superior Concrete Products in the late 1960s.16 Superior formulated by Shaikh and Yi. This equation was a modifica-
was a predecessor to Dayton Superior, previously known as tion of an original relationship developed by Ollgaard, Slut-
Dayton Richmond. This equation was developed from in- ter, and Fisher with proposed simplifications by Martin and
house research on cast-in-place specialty concrete inserts that Korkosz and later by Shaikh and Yi.24,25
 PCI JOURNAL
 Free edge cone overlap

su X
A
su

X
B

4 A

3
su

Steel plate
X

X
2(Des + Ds/2)

Fig. 2. The influence of overlapping failure cones in shear.21

The fourth edition of the PCI Design Handbook did not

change the third edition’s concrete shear breakout capac-
ity equation when the anchorage was far from a free edge.26
That is, the equation with the 800 coefficient was reserved
for studs located far from a free edge. This equation was in-
tended to apply specifically to edge distances greater than or Fig. 3. A zipper-type breakout for a stud-group anchorage.21
equal to 15 times the stud diameter, that is de > 15db. The fifth
edition did not include this equation, under the premise that
the equation may not fully represent concrete capacity when free edge de. This equation was also based on the analysis
the edge distance was greater than 15db and because of an ap- of headed stud and cast-in, ASTM A307,anchor bolt test
parently incomplete understanding of the behavioral origins data. 27
of the equation. For an anchor group, the edge distance term de was based
Both the fourth and fifth editions of the PCI Design Hand- on the rear stud row, as shown schematically in the fifth edi-
book reflected new information, studied by Shaikh and Yi, tion row entry of Table 1. Research conducted at OSU by
on the concrete shear breakout capacity near a free edge. Cruz and Wong on groups of stud anchors confirmed that
Researchers Shaikh and Yi proposed the basic, empirically the rear stud row,14,15 or stud row farthest from the free edge,
derived equation for calculating concrete shear breakout ca- controlled the ultimate breakout surface of the assembly and,
pacity φVc near an edge as: hence, the capacity. Rong and Fafitis also observed this be-
havior for headed studs in PCI-sponsored research work at
1.5 Arizona State University.28
( )
φVc = φ12.5 de f c' (lb) (Eq. 1) Prior to the fourth edition of the PCI Design Handbook,
German researchers Eligenhausen and Fuchs developed the
This empirical equation represented a lower bound capac- Kappa ( κ ) method to define the concrete shear breakout ca-
ity equation, derived from test data with concrete compres- pacity near a free edge. This equation was derived from the
sive strengths f c' in the 4000 psi to 5300 psi (27.6 MPa to analysis of data on headed, expansion, and bonded anchors.
36.5 MPa) range, and accounted for the distance to the front The latter two anchor groups were the post-installed type.

January–February 2007 
Their original average concrete shear breakout capacity equa- tribution. The front-row studs are ineffective because of the
tion Vc in SI units was: anchorage plate rigidity. Consequently, the back-row studs
dictate the strength in concrete breakout. The rigidly attached
1.5 connection plate distributes the shear load in accordance with
Vc = 1.3 db ( )
f c' c1 (N) (Eq. 2) the relative stiffness of each headed stud, though at ultimate,
the breakout crack is concentrated at the back row. This de-
where sign philosophy has been reflected as early as the fourth edi-
db = diameter of the anchor (mm) tion of the PCI Design Handbook.
f c' = concrete cube compressive strength (N/mm2)
c1 = edge distance (mm) Anchorage Design Guidelines
The authors note that this equation is limited to anchors An important factor in the performance of headed studs,
with embedment depths ranging from 4db to 8db and to those when their design is governed by concrete capacity, is the
embedded in slabs where there were no thickness effects, as confinement of the failure area with reinforcement. In shear,
addressed in the state-of-the-art report from the CEB.29 Simi- design capacity and ductility can be increased with such rein-
lar limits are associated with the current code language of forcement, likewise in tension. It has been recommended in
ACI 318-05 Appendix D. Moreover, the equation is based the fourth, fifth, and sixth editions of the PCI Design Hand-
on concrete compressive strengths in the range of 1740 psi to book that reinforcement be placed to cross failure planes
6960 psi (15 MPa to 60 MPa) cube strength. around headed-stud anchors. However, the design provisions
The Kappa equation (Eq. 2) originally had a dimension- presented in the handbook represent a lower bound on capac-
less term of (hef /db)0.2 as a multiplier on the right side. It was ity, determined by the capacity at first cracking in an unre-
simplified to its form in Eq. 2 by assuming hef (the effective inforced member. Providing reinforcement can augment the
stud embedment length) is approximately 4db, thus including anchorage capacity; however, this load-carrying mechanism
the (hef /db)0.2 term as the 1.32 in the constant coefficient. If requires a separate design that develops reinforcement be-
the hef /db term is included in the Kappa equation, limits are yond postulated failure planes. In some cases, that is difficult
placed on the embedment depth such that 4db < hef < 8db. to detail properly.
The conversion of this average equation to English units Welded, headed studs are designed to resist direct tension,
(and with concrete cube strengths assumed to be about 1.25 shear, or a combination of the two. The design equations
times the concrete cylinder strengths) becomes:30 given in the sixth edition of the PCI Design Handbook are
applicable to studs that are welded to steel plates or other
1.5
Vc = 17.5 db ( )
f c' c1 (lb) (Eq. 3) structural members and embedded in unconfined concrete. It
is assumed that the steel plates are of sufficient thickness to
where prevent significant plate deformation and to adequately trans-
db = diameter of the anchor (in.) fer applied load to and between the studs.
f c' = concrete cylinder compressive strength (psi) Where feasible, headed-stud connections should be de-
signed and detailed such that the connection failure is pre-
c1 = edge distance (in.)
cipitated by failure (typically defined as yielding) of the stud
By substituting a 1/2-in.-diameter (12.7 mm) stud into Eq. 3 material rather than failure of the surrounding concrete (un-
for db, the resultant average equation becomes similar to the less reinforcement crosses the concrete failure surface). Gen-
original PCI equation (Eq. 1). erally, the in-place strength of the anchor group should be
There has been considerable debate by ACI committees taken as the smaller of the design values based on concrete
as to the proper anchor row to consider when examining a and steel. This requirement necessitates the computation of
multiple-row, headed-stud connection. Early versions of the individual steel and concrete capacities in all cases. Unfor-
CCD method used the front row to compute the area breakout tunately, with so many variables affecting concrete capacity,
factors. The shear capacity computed was then doubled for a each connection type and configuration will have a unique
second, back row. However, this computational model con- capacity. For this reason, it is impossible to globally define
flicts with actual observed behavior and is not always a good the edge distance where an anchor group failure mode transi-
capacity predictor for headed-stud connection groups. tions from concrete to steel.
The OSU test results showed that multirow (front and back
rows) anchor group connections loaded in shear exhibit a The New Section 6.5
behavior consistently indicating that the front stud row was
ineffective and not part of the concrete failure surface. This Anchorage design provisions in the sixth edition of the
behavior was repeated in testing at the University of Wiscon- PCI Design Handbook are the result of a combination of the
sin–Milwaukee and Arizona State University.31 The failure WJE research, provisions included in past editions, and the
crack surface always propagated through the back stud row ACI 318 provisions. In light of the end user of the PCI De-
and then forward at an angle toward the free edge. sign Handbook, the provisions contained therein are geared
All of these test results, and in particular the OSU stud toward headed-stud design. Caution should be exercised in
group tests, show that a headed-stud group connection has the application of these provisions to post-installed anchor
the ability to redistribute the applied load through plastic dis- design. The new provisions are postulated to uncracked
 PCI JOURNAL
concrete, which is typical in precast concrete products, with rated into the present PCI Design Handbook provisions, as
reduction modifiers for instances of cracked concrete. This Table 6.5.1.1. These material property values have slightly
philosophy is opposite that of ACI, where cracked concrete increased from those listed in the fifth edition.
is considered typical. Currently, AWS classifies Type B studs as those that are
The information provided in this paper details the back- headed, bent, or of other configuration; in 1/2, 5/8, 3/4, 7/8, and 1
ground information used to develop the sixth edition of the in. diameters (13, 16, 19, 22, 25 mm); and used as an essen-
PCI Design Handbook provisions, as outlined in the follow- tial component in composite beam design and construction.
ing sections. Specific individual section background is also These stud diameters represent the majority of those also
provided in an effort to offer explanatory information (com- used in precast concrete construction.
mentary) to the handbook provision philosophy. Type A studs cover the 1/4- and 3/8-in.-diameter (6 and 9
mm) stud sizes used occasionally in precast concrete con-
struction. As shown in Table 2, Type A studs currently have
Steel Materials a 61 ksi (420 MPa) minimum tensile strength Fut and a 49 ksi
(340 MPa) minimum yield strength Fy. AWS defines Type
Minimum Plate Thickness A studs as “general purpose of any type and size used for
Conventional carbon steel used for anchorages should con- purposes other than shear transfer in composite beam design
form to the minimum requirements of ASTM A36 for plates and construction.”
or ASTM A992 for shapes.32,33 Stainless steel plates shall Stainless steel studs can be welded to either stainless or
conform to the minimum requirements of ASTM A666,34 mild carbon steel. Fully annealed stainless steel studs are
Type 304 or 316. Other steel types can be used, but their recommended when welding stainless-steel studs to a mild
applicability to the stud welding process should be verified. carbon steel base metal. Using annealed stainless-steel studs
The minimum plate thickness tp to which studs are attached has been shown to be imperative for welding to carbon steel
should be: plates subject to repetitive or cyclic loads. In such cases,
stress corrosion failure in the weld can occur,40 and use of
1 the annealed stud minimizes the chance of weld cracking and
tp ≥ d (Eq. 4, handbook Eq. 6.5.1.18) failure. Consult the headed-stud supplier to obtain additional
2 0
information on stainless-steel stud use and availability.
where
d0 = the stud diameter (in.) Steel Stud Capacity
This provision is a carryover from several past editions of As presented in an earlier paper,9 the design ultimate shear
the PCI Design Handbook and is based on the research of or tensile strength governed by steel failure can be expressed
Goble at Case Western Reserve University.35 Increased plate by:
thickness may be required for bending resistance or to ensure
a more uniform load distribution to the attached studs. Perry, φVs = φNs = φsnstudsAsFut (Eq. 5, handbook Eq. 6.5.2.18)
Funk, and Burdette provide more information on plate stiff-
ness.36
where
Headed-Stud Properties φ = 0.65 (steel capacity reduction factor for studs in
shear)
The Structural Welding Code, AWS D1.1-04, has recog- = 0.75 (steel capacity reduction factor for stud in ten-
nized that mild steels conforming to ASTM A108 (Grades sion)
1010 through 1020) and used for headed studs have increased Ns = nominal tensile strength of an anchorage based on
material properties.37,38 Table 2, adapted from Table 7.1 in steel capacity (kip)
AWS D1.1-02, shows the current minimum tensile strength Vs = nominal shear strength of an anchorage based on
Fut and yield strength Fy for Type B studs to be 65 ksi and 51 steel capacity (kip)
ksi (450 MPa and 350 MPa), respectively, which is incorpo- nstuds = number of headed studs in the anchorage

Table 2. Minimum Mechanical Property Requirements for Headed Studs

Property Type A Type B
(Diameters) (1/4 to 3/8 in.) (1/2 to 1 in.)

Tensile strength Fut (min.) 61,000 psi 65,000 psi

Yield strength Fy (0.2% offset) 49,000 psi 51,000 psi

Elongation (min. % elongation in 2 in.) 17% 20%

Reduction of area (min.) 50% 50%

Note: 1 in. = 25.4 mm; 1 psi = 6.895 KPa.

January–February 2007 
 As = nominal area of the headed stud (in.2) λ = concrete density factor
Fut = minimum design ultimate tensile strength of the = 1.0 for normalweight concrete
stud steel (ksi) = 0.85 for sand lightweight concrete
= 65 ksi for normal Type B headed studs used in = 0.75 for all lightweight concrete
precast concrete anchorages
For anchors in tension, the single anchor capacity is modi-
ACI 318 Appendix D provisions place a lower steel capac- fied by several factors to account for the effects of edges,
ity reduction factor (φ = 0.65) on the steel shear strength than spacing, and cracking.
when loaded in tension (φ = 0.75). Section RD.4.4 of ACI
318 states: Ncb = Ncbg = f(φ,Nu,s1,s2,de1,de2,de3,de4,cracking) (Eq. 7)
“The φ factors for steel strength are based on using futa to
determine the nominal strength of the anchor (see D.5.1 and The factors are a combination of modification and correc-
D.6.1) rather than fya as used in the design of reinforced con- tion factors because the CCD model was used as the basis of
crete members. Although the φ factors for use with futa appear the design procedure. In some respects, the PCI design proce-
low, they result in a level of safety consistent with the use of dure follows the philosophy first used in the Kappa method.
higher φ factors applied to fya. The smaller φ factors for shear However, because the already-accepted CCD method in ACI
than for tension do not reflect basic material differences but 318 is fundamentally based on the Kappa method, returning
rather account for the possibility of a non-uniform distribu- to the Kappa method approach in the sixth edition of the PCI
tion of shear in connections with multiple anchors.” Design Handbook did not seem appropriate.
The authors believe that this factor is too restrictive for For simplification, the group equation is presented in the
headed studs welded to a plate, and a factor φ = 0.75 would following form in the handbook:
be more appropriate. This is based on the fact that the steel
plate can plastically redistribute the shear to headed studs Ncb = Ncbg = CbsANCcrbΨed,N (Eq. 8, handbook Eq. 6.5.4.18)
better than to post-installed anchors.
where the breakout strength coefficient is defined by:

Concrete Strength f c'

Cbs = 3.33λ (Eq. 9, handbook Eq. 6.5.4.28)
hef
Tension
The design tensile strength of a single anchor governed by Nb
concrete failure is given in ACI 318-02, which was confirmed Cbs is equivalent to the term Ψ in ACI 318-02:
where ANo 3
for headed-stud use by reviewing the existing database of
tension test results for cast-in anchors as part of the Phase 2 Nb = basic concrete breakout strength in tension of a
WJE-PCI research. From these test results, group factors in single headed-stud anchor in cracked concrete
tension are slight variations of the ACI model factors. The in- Ψ3 = modification factor, for strength in tension, to ac-
place strength of the headed-stud anchorage should be taken count for cracking
as the minimum value based on computing both the concrete ANo = projected concrete failure area of one anchor, for
and steel capacity for a given anchorage configuration. calculation of strength in tension when not limited
by edge distance or spacing
Single Anchor Tension Capacity Cracking—Cracking in the vicinity of an anchor will reduce
its concrete tensile breakout capacity. The amount of reduc-
The single stud concrete breakout prediction equation is tion is dependent on the width of the crack, as derived from
given by: research. In accordance with ACI 318, the cracking coeffi-
1.5
cient Ccrb is:
( )
φ Nu = φ 30 λ f c' hef (Eq. 6) Ccrb = 1 .0 for uncracked concrete (ft < fr) at
service loads, and
Ccrb = 0.80 for cracked concrete at service loads.
where
φ = concrete strength reduction factor = 0.75 (for Edge distance factor—Anchors are affected by edge dis-
tension) tance because, when close to a free edge, the concrete break-
Nu = concrete tensile breakout capacity for a single stud out surface is not fully developed. In simplistic terms, the
hef = effective embedment = the effective steel stud loss of available concrete for the tension breakout surface
height, after welding burn off, defined to the base of reduces the overall capacity of the connection. The edge
the head distance factor is effectively a model correction factor ac-
=L  (stud length after welding) - A (head height) + the counting for differences between the model prediction and
thickness of the plate to which the studs are attached the test results in the database. The edge distance modifica-
(See Table 6.5.1.2 in PCI Design Handbook.)8 tion factor φed,N only needs to be applied once, even if there
burnoff = amount of stud length lost in the welding process are edge distances within 1.5 hef on more than one side of
10 PCI JOURNAL
the anchors. If there are three or more edges, special rules Side-Face Blowout
are given in ACI 318 Appendix D to reduce the effective
embedment depth of the anchors. See the commentary in For a single headed stud located close to a free edge (de1 <
ACI 318-05 for further details. 0.4 hef), the side-face blowout capacity is defined as:

⎛d ⎞ N sb = 160de1 Abrg f c'

Ψ ed ,N = 0.7 + 0.3⎜ e,min ⎟ ≤ 1.0
⎝ 1.5hef ⎠ (Eq. 13, handbook Eq. 6.5.4.68)
(Eq. 10, handbook Eq. 6.5.4.38)
This side-face blowout capacity Nsb must be multiplied by
where a factor if the single headed stud is located near an edge such
dmin = minimum dimension of de1, de2, de3, or de4 (in.); refer- that de2 < 3de1. This factor is defined as:
ence should be made to Fig. 6.5.4.2 in the hand-
book8  d 
1+ e2 
Eccentricity factor—Any eccentricity of the applied load  de1 
(Eq. 14, handbook Eq. 6.5.4.78)
relative to the geometric centroid of the anchor group causes 4
a non-uniform distribution of resisting forces in the anchors.
The eccentricity modification factor φec,N accounts for the where
non-uniform load distribution in the anchor group.
de2
1≤ ≤3
1 de1
φ ec,N = ≤ 1.0
 2e' 
1+ N  For multiple headed studs located close to a free edge (de1 <
 3h 
 ef  0.4 hef), the side-face blowout capacity is further modified as:
(Eq. 11, handbook Eq. 6.5.4.48)
where  s 
N sbg = 1+ 0  N sb
e'N = distance between resultant tension load on a group  6de1 
of anchors loaded in tension and the centroid of
(Eq. 15, handbook Eq. 6.5.4.88)
the group of anchors loaded in tension (in.); e'N is
always positive
where
e'N < s /2, where s = minimum individual x or y spacing,
de1 = distance to closest edge (in.)
depending on the eccentricity direction being
Abrg = bearing area of the head of stud or anchor bolt (in.2)
evaluated
s0 = center-to-center spacing of transverse reinforcement
within the length of l0 (in.)
Pull-Out Strength Nsb = side-face blowout strength of a single anchor (lb),
see Eq. 13
Pull-out capacity Npn is dictated by a failure of the concrete
around the head of the headed stud. When the bearing area of Side-face blowout failures are unique to embedded, headed
the head is small, concrete crushing occurs at the head and the anchors. This failure type is affected by an edge condition,
anchor can pull out and crush the concrete without forming a but not the same edge condition associated with a general
concrete breakout cone. Local crushing under the head of the concrete breakout failure. If the head of an anchor is close to
anchor significantly reduces the stiffness of the anchor and a free edge, the compression stress bulb at the head bearing
increased displacement is associated with crushing under the region can cause the concrete to spall because it is no longer
head of the anchor. confined. Note that this condition applies to very small edge
distances and relatively deeply embedded anchors.
N pn = 11.2 Abrg f c'Ccrp
Shear
(Eq. 12, handbook Eq. 6.5.4.5 )8

The procedure to determine the design shear strength gov-

where erned by concrete failure is based on the models and test re-
Abrg = bearing area of the stud head in tension (in.2) sults from the WJE/PCI research project. The results used
= area of the head − area of the shank (values are were those from experimental testing programs exclusively
shown in handbook Table 6.5.1.28) based on headed-stud anchors. The strength of an anchor
Ccrp = cracking coefficient (pullout) should be taken as the minimum value based on computing
= 1.0 for concrete assumed uncracked (most common) both the concrete and steel capacity for the characteristics of
= 0.7 for locations likely to become cracked the unique anchor configuration.
January–February 2007 11
 where the back-edge distance (BED) to the rear row of
Vn
Y = ‹y studs is defined as:

BED = de3 + ∑ Y =d
i e3
+Y
y
(Eq. 18, handbook Eq. 6.5.5.38)
BED

de3
where
x x y = individual Y-row spacing (center to center) (in.);
X reference Fig. 4
Y = the overall, out-to-out dimension of the column of
Front edge conditions studs on the side row of the anchorage
= Σy (in.), parallel to the applied shear force
Fig. 4. Conventional concrete breakout when the shear load is λ = ACI 318 lightweight concrete factor
applied perpendicular or normal to the free edge. Note: BED = de3 = front-edge distance parallel to the shear load appli-
back-edge distance. cation direction and y-axis, taken from the center of
a front-anchor shaft to the front concrete edge (in.)
Front Edge (de3) X spacing—The influence of the stud spacing in rows per-
pendicular to the applied shear force between adjacent col-
The front-edge condition represents the majority of shear umns of anchors (when two or more studs are in the back
loaded connections in design and is the condition that has row) requires a strength modification by the X-spacing coef-
typically yielded the smallest concrete breakout capacity. A ficient CX3:
shear force is applied perpendicular, or normal, to the front
edge of the concrete, as illustrated in Fig. 4. The design ca-
pacity, when governed by concrete front-edge breakout, is X
C X 3 = 0.85 + ≤ nstuds−back
given by: 3BED

φVc3 = φVco3 (CX3) (Ch3) (Cev3) (Cvcr) C X 3 = 1.0 when X = 0
(Eq. 16, handbook Eq. 6.5.5.1 ) 8

(Eq. 19, handbook Eq. 6.5.5.48)

where
φ = concrete strength reduction factor where
= 0.70 without confinement steel X = the overall, out-to-out dimension of the row of
= 0.75 with confinement steel studs in the back row of the anchorage
Vc3 = nominal concrete breakout capacity for a single or = Σx (in.), perpendicular to the applied shear
multiple stud connection, accounting for member force
and connection geometry = 1.0 for a connection with only one stud in the
Vco3 = nominal concrete breakout capacity for a single stud back row
connection unaffected by connection or member BED = back-edge distance, defined previously (in.)
geometry nstuds-back = number of studs in the back row
CX3 = coefficient for overall X spacing, spacing of anchors x = center-to-center spacing of stud rows in X
in rows parallel to the free edge (or spacing of direction
anchors perpendicular to the applied shear force), of
a connection with two or more columns, for the de3 Member thickness—The influence of the concrete mem-
type anchorage ber’s thickness is accounted for by the member thickness
Ch3 = coefficient for member thickness (h) for the de3 type coefficient Ch3:
anchorage
Cev3 = coefficient for eccentric shear force influences for a Ch3 = 1.0 for h > 1.75BED
de3 type anchorage
Cvcr = coefficient for cracking in the concrete h
Ch3 = 0.75 for h ≤ 1.75BED
BED
The prediction equation for single stud concrete breakout
capacity is given by: (Eq. 20, handbook Eq. 6.5.5.58)

where
(
Vco3 = 16.5λ f c' BED )1.33 h = member thickness (in.)
(Eq. 17, handbook Eq. 6.5.5.28) BED = back-edge distance, defined previously (in.)
12 PCI JOURNAL
Eccentricity—The location of the applied shear force is not
always concentric with the centroid of the resisting anchors. Vn
Critical
Effectively, this places the anchor group into a torsional-shear cattycorner
stud
state. To account for this eccentricity, the group capacity re-
quires modification by the eccentric load factor Cev3:
y
1
Cev3 = ≤ 1.0 BED
⎛ e' ⎞ de3
1+ 0.67 ⎜ v ⎟
⎝ BED ⎠
x x
X de1
for ev' ≤ X
2
SED

(Eq. 21, handbook Eq. 6.5.5.68) Corner conditions

where Fig. 5. Corner-concrete breakout when a headed-stud

ev' = eccentricity of shear force on a group of anchors anchorage is located near a member corner. Note: BED =
= the distance between the point of shear force applica- back-edge distance; SED = side-edge distance.
tion and the centroid of the group of anchors resist-
ing shear in the direction of the applied shear (in.) where the side-edge distance (SED) to the cattycorner (an-
Cracking—Cracking in the vicinity of, or through, the an- chor point) stud, as shown in Fig. 5, is defined as:
chors will reduce their shear capacity, as observed from re-
search. The influence of a crack parallel to the shear force SED = de1 + ∑x = d e1
+X
on the anchor capacity is not quite as dramatic as that from
cracking in tension. The cracking coefficients are of a differ- (Eq. 23, handbook Eq. 6.5.5.88)
ent magnitude than the ones shown in ACI 318. The basic
equation in the PCI Design Handbook is the uncracked ca-
The design shear capacity governed by concrete breakout
pacity, rather than the cracked capacity, as reflected in ACI
at the corner is thus given by:
318. The cracking coefficient Cvcr is:
for uncracked concrete (ft < fr):
φVc3 = φVco3 (Ch3) (Cc3) (Cvcr) (Cev3)
Cvcr = 1.0 (Eq. 27, handbook Eq. 6.5.5.98)

for cracked concrete at service load:

Cvcr = 0 .70 when no edge reinforcement or reinforcement where
is smaller than No. 4 (No. 13M) bar φ = concrete strength reduction factor
Cvcr = 0 .85 with supplementary reinforcement of No. 4 = 0.70 without confinement steel
(No. 13M) bar or greater between the anchor and = 0.75 with confinement steel
the edge Ch3 = coefficient for member thickness (h) for a de3 type
Cvcr = 1.0 with supplementary reinforcement between the anchorage (see handbook Eq. 6.5.5.68)
anchor and the edge of No. 4 (No. 13M) bar or great- Cvcr = coefficient for cracking in a member, loaded in
er and supplementary reinforcement enclosed within shear
stirrups with a spacing less than 4 in. (100 mm) Cc3 = coefficient for the corner influence for a de3 type
anchorage
Corners Cev3 = coefficient for eccentric shear force influences for a
de3 type anchorage
The corner condition is considered to be a special case of For stud anchors located near a corner, the corner coef-
the front-edge loaded anchor. Again, the shear force is ap- ficient Cc3 is:
plied perpendicular or normal to the concrete’s front edge
as illustrated in Fig. 5, but the anchor is located sufficiently
close to the corner (side edge) so that a different concrete
SED
Cc3 = 0.7 3 ≤ 1.0
breakout shape occurs. A corner condition results when: BED

SED (Eq. 25, handbook Eq. 6.5.5.108)

0.2 < ≤ 3.0
BED
For the special case when there is a large overall spacing
(Eq. 22, handbook Eq. 6.5.5.78) in an X-row (the row perpendicular to the applied shear) and
January–February 2007 13
 de1

de1

de3

Connection poitioned such that SED/BED > 3 but stud at de1 < BED
may dictate a corner breakout (Cc3 = 1.0 for corner breakout equation)

Fig. 6. Corner transition zone where a close-to-corner stud may induce a zipping type crack propagation to the corner (PCI
Design Handbook8 Figure 6.5.5.3). Note: BED = back-edge distance; SED = side-edge distance.

the stud anchor is located near a corner, such that SED/BED The design capacity governed by concrete breakout at the
> 3.0, but one of the stud rows remains fairly close to the side edge is given by:
corner, a corner type crack and breakout may still result. A
transition zone exists under these conditions, such that: φVc1 = φVco1 (CX1) (CY1) (Cev1) (Cvcr)

(Eq. 28, handbook Eq. 6.5.5.128)

Cc3 = 1.0 for SED > 3.0BED and
(Eq. 26)
de1 / BED ≤ 2.5 where
φ = concrete strength reduction factor
= 0.70 without confinement steel
This situation is shown in Fig. 6, reproduced from Fig. = 0.75 with confinement steel
6.5.5.3 in the sixth edition of the PCI Design Handbook. Given Vc1 = concrete breakout capacity for a single or multiple
the large, wide overall anchor spacing of the outside studs stud connection, accounting for member and con-
(large X) with the end stud fairly close to the corner, yet with nection geometry
SED/BED > 3.0, crack propagation can occur along the line of Vco1 = concrete breakout capacity for a single stud connec-
the rear stud row and be directed to the side edge, as shown in tion unaffected by connection or member geometry
Fig. 5. This corner-type failure was experimentally observed in Cvcr = coefficient for cracking in a member, loaded in
the WJE/PCI research for the conditions set forth in Eq. 26. shear
CX1 = coefficient for overall X spacing of the outside studs
Side Edge (de1 or de2) in a connection with two or more x rows (rows per-
pendicular to the shear force) for a de1 type anchor
A side-edge concrete breakout failure mode is significantly CY1 = coefficient for overall Y spacing of the outside studs
different from the traditional front-edge breakout mode. In in a connection with two or more y rows (rows par-
this case, the shear force is applied parallel to the side edge allel to the applied shear force) for a de1 type anchor
of the concrete member, as illustrated in Fig. 7. To determine Cev1 = coefficient for eccentric shear force influences for a
if a connection is a corner or a side-edge condition, the fol- de1 type anchor
lowing equation is evaluated. The anchors will behave in a
side-edge breakout mode if: The single stud concrete breakout prediction equation is
given by:
SED
≤ 0.2
BED 1.33
( ) ( d0 )
Vco1 = 87 λ f c' de1
0.75

(Eq. 27, handbook Eq. 6.5.5.118) (Eq. 29, handbook Eq. 6.5.5.138)
14 PCI JOURNAL
 where
de1 = side-edge distance to the first line of studs (in.); for
cases where two parallel sides exist and the anchor- Y = ‹y
age is off center, use the lesser of de1 or de2
d0 = stud diameter (in.) Vn
λ = ACI 318 lightweight concrete factor
f c' = concrete compressive strength (psi)

X spacing—The factor accounts for anchors in a row perpen- y

dicular to the shear force located near one edge of the con-
crete member and for anchors having studs in a row adjacent
to two parallel edges of the concrete member.
de3
Figure 8 reproduces Figure 6.5.5.4 from the sixth edition
x x
of the PCI Design Handbook and illustrates the side-edge- de1 X
distance conditions. The X-spacing factor for two or more
studs in a row perpendicular to the shear is given by:

nx x
Cx1 = + 2 − nsides
2.5de1
where 1 ≤ Cx ≤ nx
Cx1 = 1.0, when x = 0
Side edge conditions

(Eq. 30, handbook Eq. 6.5.5.148) Fig. 7. Corner concrete breakout when the shear load is
applied parallel to the free edge.

where
nx = number of X-row stud lines Cx1 = nx (Eq. 31, handbook Eq. 6.5.5.158)
x = individual X-row spacing (in.)
nsides = number of edges or sides
Y Spacing—The Y-spacing is a factor for spacing in a col-
For all anchorages with multiple rows perpendicular to the umn of anchors parallel to the load. This influence applies to
shear force and that are located adjacent to two parallel edges, two or more stud rows perpendicular to the shear force and
such as a column corbel connection, the X-spacing factor for adjacent to one edge of the concrete member or in the case
two or more studs in the row perpendicular to the shear is: of two parallel side edges, such as the vertical edges of a

de1 de1 de1 de1 de1

nxx
Cx1 nsides
de1 Cx1 nx

nx
nsides nx
ny nsides nx
ny ny

(one edge) (two edges) (two edges)

Fig. 8. Various definitions for side-edge distance factors for one or two side edges (PCI Design Handbook Figure 6.5.5.48).

January–February 2007 15
column with a shear force acting parallel to the height of the when used in thin-wythe members.
column: When a headed-stud anchor is sufficiently far from all
edges, termed in-the-field of the member, the anchorage ca-
pacity will be governed by the capacity of the steel stud(s).
CY1 = 1.0 for ny = 1 (one y row)
Pryout failure is a concrete breakout mode that may govern
0.25 when short, stocky studs are used (hef /d0 < 4.5). The pryout
CY1 =
(n Y )
y
+ 0.15 ≤ 1 for ny ≥ 1.0 capacity in lightweight and normal weight concrete has been
0.6de1 found to govern when hef /d0 < 4.5. If this condition exists:

(Eq. 32, handbook Eq. 6.5.5.168) φVcp = φVcp0

where 1.5 0.5

ny = number of Y-row stud lines (rows parallel to the Vcp0 = 215nstudsΨ y ( ) ( hef )
f c' d0 ≤ nstuds As Fut
shear)
Y = out-to-out Y-row spacing of anchors = Σy
(Eq. 34, handbook Eq. 6.5.7.18)

Author’s note: Please note the errata in this equation for

Cy1 in the PCI Design Handbook. The factor is capped at where
ny and not 1.0 as shown in the handbook.
y y
Ψy = for ≤ 20
Eccentric shear—When the shear load is applied eccentric 4d0 d0

to the anchorage’s centroid, the load eccentricity will reduce
the anchorage capacity by:
= 1.0 for y = 0

e 
CeV 1 = 1.0 −  V 1  ≤ 1.0
φ = strength reduction factor
 4de1 
= 0.70 without confinement reinforcement
Vcp = nominal pryout shear strength (lb)
(Eq. 33, handbook Eq. 6.5.5.178)
nstuds = number of studs in the connection group
As = effective cross-sectional area of the stud anchor (in.2)
where f c' = specified compressive strength of concrete (psi)
eV1 = the eccentricity from the shear load to the centroid y = center-to-center spacing of studs in direction of
of the anchorage (in.) load

Back Edge (de4)

Author’s note: Please note the errata in the above equa-
A back-edge condition exists when the shear force is ap- tion for ψy in the PCI Design Handbook. The factor
plied perpendicular, or normal, to the back edge of the con- ψy = 1.0 when y = 0 and should be on the next line, as
crete member, as illustrated in Fig. 1. Under a condition of presented in the handbook.
pure shear loading, the back-edge distance has been found,
through testing, to have no influence on the connection ca-
pacity when the studs are well developed. The pryout failure
mode should be checked, as applicable. Combined Tension and Shear

WJE is in the process of analyzing the research data and

In the Field formulating design recommendations for tension and shear as
part of Phase 2 of the headed-stud research project. The design
The WJE/PCI research identified a failure mode that oc- guidelines in the sixth edition of the PCI Design Handbook
curs with very short studs located far enough from an edge to allow for anchorage capacity determination with the tension-
preclude a concrete-edge breakout failure. This failure mode shear interaction permitted by ACI 318-02. In the ACI provi-
is known as pryout and was somewhat evaluated in the 800 sions, a simplified tri-linear relationship for the interaction of
equation in previous editions of the PCI Design Handbook. tension and shear is permitted. In the ACI 318 commentary,
The equation presented in the current PCI Design Handbook, the traditional five-thirds (5/3) power interaction is permit-
Chapter 6, is a refinement of this equation based on a review ted, which is in accordance with previous editions of the PCI
of existing literature and WJE and German test data.41 This Design Handbook. Both relationships are plotted in Fig. 9. It
design condition will influence a very small population of is noted in the sixth edition of the PCI Design Handbook that
headed-stud sizes, yet it may be a very common stud type both equations should be examined when determining the ca-
16 PCI JOURNAL
 Nn 5/3 5/3 Vn
N  V 
 u  +  u  = 1.0
 φNn   φVn  Contact bearing on
áNn Plate
rear of anchor (typ.)
 N  V 
 u  +  u  = 1.2
 φNn   φVn  Slight gap in front

Edge of concrete

Installation spacing of
anchors

Hole spacing in plate

0.2áNn (a) Spacing perpendicular to shear arrangement allowing bearing on both

anchors.

Vn
0.2áVn áVn Vn
Installation spacing
of anchors

Fig. 9. The interaction of tension and shear presented in the Hole spacing Plate Gap in back–no
contact bearing
PCI Design Handbook8 Figure 6.5.8.1. in plate

Contact bearing on
rear of anchor
pacity of headed-stud anchors, as the tri-linear relationship
may truncate some allowable combinations when the ten- Oversized holes in
plate for tolerance
sion and shear magnitudes are approximately equal (between Edge of concrete

about 30 and 60 degrees, or the shaded region of Fig. 9).

As mentioned, it is anticipated that the present analysis
of the interaction and the design recommendations will be (b) Spacing parallel to shear arrangement increases the likelihood of unequal load
on the anchors.
incorporated into the seventh edition of the PCI Design
Handbook.
Fig. 10. Load conditions inherent with cast-in-place anchor
Cast-In Anchor Bolts bolts or post-installed anchorages (a) X-spacing arrangement
allowing bearing on both anchors (b) Y-spacing arrangement
increases the likelihood of unequal load on the anchors.
The provisions for the front-edge distance de3 condition are
intended for use with headed-stud anchors, where the stud is
welded to an attachment plate. Cast-in anchor bolts behave Adhesive Anchors
somewhat similarly to studs in a majority of the concrete
breakout modes, but their behavior is highly dependent on Anchors used in concrete construction fall into two broad
the degree of fixity to the attachment plate and which of the categories: cast-in-place anchors and post-installed anchors.
bolts in the group are bearing on the attachment plate. Head- With increasing demand for more flexibility in the planning
ed studs are fully welded and, hence, fixed to the plate, so and construction of concrete structures, and for repair and
double curvature in the stud displacement can develop under retrofit applications, post-installed anchors have seen in-
lateral shear loading. Likewise, a more equal distribution of creased use. One popular form of the post-installed anchor is
the applied load to the individual studs can be assumed in a the adhesive anchor.
headed-stud anchor. An adhesive anchor is a threaded rod or a reinforcing bar
A cast-in anchor bolt may be placed into an oversized hole that is inserted in a hole drilled in hardened concrete. The di-
for tolerance purposes, such that some “slop” is introduced ameter of the hole is typically 15% to 25% larger than the di-
into the connection. Bolt rotation within the plate hole can ameter of the anchor. The annulus around the anchor is filled
occur, making full fixity difficult to achieve. In anchorages with an adhesive that bonds the steel anchor to the concrete,
with multiple cast-in bolts, the various oversized hole dimen- and the anchor is termed an adhesively bonded anchor.
sions coupled with the actual bolt location within the hole While there are generally accepted and more recently codi-
can give rise to conditions of uneven bearing among the bolts fied procedures for the design of cast-in-place anchors, such
and, consequently, uneven loading on individual bolts in the as headed studs, comparable information is not yet available
group. Figure 10 illustrates these issues with hole versus an- for adhesive anchors. Thus, in the interim, the designer must
chor size. rely on manufacturers’ recommendations to estimate the
The provisions of ACI 318 Appendix D are more appli- strength of these anchors.
cable to these anchorage situations. Moreover, post-installed During the past five years, research has been reported that
anchors should be designed in accordance with the ACI pro- suggests unified models for calculating tensile strength and
visions or those recommended by the anchor manufacturer. shear strength for single adhesive anchors.42,43 While the cal-
January–February 2007 17
culation models exist, the designer should correlate calculat- front edge de3 and side edge de1;
ed capacity with manufacturers’ listed capacities to assess the • Capacity of anchors in lightweight concrete; limited
actual factor of safety associated with adhesive anchors. research has been performed to verify the assumed
applicability of the λ factors from ACI 318; and
• Capacity of anchorages with large Y-spacing (spac-
Notes on Notation ing parallel to the shear force) to investigate shear lag
influences.
To simplify the notation used to define the planar geometry
of a headed-stud plate, the designer should notice the use of
a Cartesian coordinate system of the stud layout in the PCI Acknowledgments
Design Handbook. Figures 4, 5, and 7 illustrated the x- and
y-axes’ layout used in the design equations. In general, the The authors express their appreciation to PCI for spon-
y-axis is oriented parallel to the applied shear force, whereas soring this research project. In particular, the authors thank
the x-axis is then oriented perpendicular to the applied shear PCI’s Research & Development Committee (C. Douglas Sut-
force. Similar to a spreadsheet, it is sometimes convenient to ton, chair) and the members of the Projects Advisory Com-
view a multi-stud connection in terms of columns (parallel to mittee (Thomas J. D’Arcy, chair) for their constructive com-
the y-axis) and rows (parallel to the x-axis). ments during the course of this project. The thoughtful and
constructive review comments and suggestions from the PCI
Journal paper reviewers is acknowledged and appreciated.
Summary
References
The WJE/PCI headed-stud research program has produced
an alternate shear design procedure that better represents the 1. Industry Handbook Committee. 1971. PCI Design Handbook:
behavior of headed-stud anchors in precast concrete mem- Precast and Prestressed Concrete. 1st ed. Chicago, IL: Pre-
bers. This design procedure also conforms to the ACI 318 stressed Concrete Institute (PCI).
Appendix D requirements. The sixth edition of the PCI 2. American Concrete Institute (ACI) Committee 318. 2002.
Design Handbook now recognizes that there are different Building Code Requirements for Structural Concrete (ACI 318-
types of failure modes associated with headed-stud anchors 02) and Commentary (ACI 318R-02). Farmington Hills, MI:
depending on the type of edge condition, connection geom- ACI.
etry, and the edge distances in relation to the connection. The 3. ACI Committee 318. 2005. Building Code Requirements for
front-edge breakout mode was contained in previous editions Structural Concrete (ACI 318-05) and Commentary (ACI 318R-
05). Farmington Hills, MI: ACI.
of the PCI Design Handbook and has been refined through
4. ACI Committee 349. 1983. Proposed Revisions to: Code Re-
WJE/PCI research. In addition, the corner-concrete breakout quirements for Nuclear Safety Related Concrete Structures
mode has been found to have a greater influence, and revised (ACI 349-76). Journal Proceedings. V. 80, No. 2 (March): pp.
connection capacity equations are presented. In the sixth edi- 79–84.
tion of the PCI Design Handbook, the concept of a side-edge 5. ACI Committee 349. 1978. Addition to Commentary on Code
breakout was introduced, as it was part of the WJE/PCI re- Requirements for Nuclear Safety Related Concrete Structures
search program. Capacity equations for a connection adjacent (ACI 349-76). ACI Journal, V. 75, No. 8 (August): pp. 329–
to a side edge are presented for the first time in this edition of 347.
the handbook. 6. Eligenhausen, R., and W. Fuchs. 1988. Tragverhalten von
Tension design of headed-stud anchors follows the ACI Dübelbefestigungen bei Querzug-, Schrägzug-, und Biege-
318 Appendix D approach because the design model was beanspruchung [Load-Bearing Behavior of Anchor Fastenings
under Shear, Combined Tension and Shear or Flexural Load-
found by WJE to be a good representation of headed-stud
ing]. [In German and English.] Betonwerk + Fertigteil-Technik,
behavior in tension. V. 54, No. 2: pp. 48–56.
7. Industry Handbook Committee. 1999. PCI Design Handbook:
Precast and Prestressed Concrete. 5th ed. MNL-120-99. Chi-
Research Needs cago, IL: PCI.
8. Industry Handbook Committee. 2004. PCI Design Handbook:
Through the course of the WJE/PCI research endeavor, at- Precast and Prestressed Concrete. 6th ed. MNL-120-04. Chi-
tempts were made to isolate a number of variables to deter- cago, IL: PCI.
mine their influence. However, the nature of research is to un- 9. Anderson, N. S., and D. F. Meinheit. 2000. Design Criteria for
Headed Stud Groups in Shear: Part 1—Steel Capacity and Back
cover unanswered questions through extensive data analysis,
Edge Effects. PCI Journal, V. 45, No. 5 (September–October):
newly discovered behavior, lack of appropriate or relevant pp. 46–75.
tests, or new methodologies to review existing data from the 10. Anderson, N. S., and D. F. Meinheit. 2001. Steel Capacity of
literature. To this end, a number of conditions or behaviors Headed Studs Loaded in Shear. In Connections between Steel
were uncovered that could be addressed through future re- and Concrete: International RILEM Symposium, proceedings
search. Suggested research needs include the following: pro021, ed. R. Eligehausen, pp. 202–211. Cachan, France:
• Better definition of the transition region of corners to RILEM Publications S.A.R.L.

18 PCI JOURNAL
11. Anderson, N. S., and D. F. Meinheit. 2005. Pryout Capacity Worked Austenitic Stainless Steel Sheet, Strip, Plate, and Flat
of Cast-In Headed Stud Anchors. PCI Journal, V. 50, No. 2 Bar. ASTM A666-03. V. 01.03. West Conshohocken, PA:
(March–April): pp 90–112. ASTM.
12. Fuchs, W., R. Eligenhausen, and J. E. Breen. 1995. Concrete 35. Goble, G. G. 1968. Shear Strength of Thin Flange Composite
Capacity Design (CCD) Approach for Fastening to Concrete. ACI Specimens. Engineering Journal, V. 5, No. 2: pp. 62–65.
Structural Journal, V. 92, No. 1 (January–February): pp. 73–94. 36. Perry, T. C., R. R. Funk, and E. G. Burdette. 1983. Effect of
13. International Conference of Building Officials (ICBO). 1979. Plate Flexibility on Anchor Loads. Symposium on Anchorage to
Uniform Building Code. Whittier, CA: ICBO. Concrete. Farmington Hills, MI: ACI.
14. Cruz, R. D. 1987. Effect of Edge Distance on Stud Groups 37. American Welding Society (AWS). 2004. Structural Welding
Loaded in Shear and Torsion. Masters thesis. Oklahoma State Code—Steel. AWS D1.1/D1.1M. 20th ed. Miami, FL: AWS.
University, Stillwater, OK. 38. ASTM. 2003. Standard Specification for Steel Bars, Carbon,
15. Wong, T. L. 1988. Stud Groups Loaded in Shear. Masters the- Cold-Finished, Standard Quality. ASTM A108-03. V. 01.05.
sis. Oklahoma State University, Stillwater, OK. West Conshohocken, PA: ASTM.
16. Courtois, P. 1969. Industrial Research on Connections for Pre- 39. AWS. 2002. Structural Welding Code—Steel. AWS D1.1/
cast and In-Situ Concrete. In Mechanical Fasteners for Con- D1.1M. 18th ed. Miami, FL: AWS.
crete, SP-22, pp. 123–138. Detroit, MI: ACI. 40. Chambers, H. A. 2001. Principles and Practices of Stud Weld-
17. Industry Handbook Committee. 1978. PCI Design Handbook: ing. PCI Journal, V. 46, No. 5 (September–October): pp. 46–
Precast and Prestressed Concrete. 2nd ed. Chicago, IL: PCI. 58.
18. McMackin, P. J., R. G. Slutter, and J. W. Fisher. 1973. Head- 41. Zhao, G. 1994. Tragverhalten von randfernen Kopfbolzenver-
ed Steel Anchor under Combined Loading. AISC Engineering ankerungen bei Betonbruch [Load-Carrying Behavior of Hea-
Journal, V. 10, No. 2: pp. 43–52. ded Stud Anchors in Concrete Breakout Away from an Edge].
19. Industry Handbook Committee. 1985. PCI Design Handbook: [In German.] Report 1994/1. Stuttgart, Germany: Institut für
Precast and Prestressed Concrete. 3rd ed. Chicago, IL: PCI. Werkstoffe im Bauwesen, Universität of Stuttgart.
20. Cannon, R. W., E. G. Burdette, and R. R. Funk. 1975. Anchor- 42. Cook, R. A., J. Kanz, W. Fuchs, and R. C. Kanz. 1998. Behav-
age to Concrete. Report No. CEB 75-32. Knoxville, TN: Civil ior and Design of Single Adhesive Anchors under Tensile Load
Engineering Branch, Tennessee Valley Authority. in Uncracked Concrete. ACI Structural Journal, V. 95, No. 1
21. Klingner, R. E., J. A. Mendonca, and J. B. Malik. 1982. Ef- (January): pp. 9–26.
fect of Reinforcing Details on the Shear Resistance of Anchor 43. Bickel, T. S., and A. F. Shaikh. 2002. Shear Strength of Adhe-
Bolts under Reversed Cyclic Loading. ACI Journal, V. 79, No. sive Anchors. PCI Journal, V. 47, No. 5 (September–October):
1 (January–February): pp. 3–12. pp. 92–101.
22. TRW Inc. Nelson Stud Welding Division. 1988. Embedment
Properties of Headed Studs. Elyria, OH: TRW Inc.
23. Shaikh, A. F., and W. Yi. 1985. In Place Strength of Welded
Headed Studs. PCI Journal, V. 30, No. 2 (March–April): pp. Notation
56–81.
24. Ollgaard, J. G., R. G. Slutter, and J. W. Fisher. 1971. Shear
Strength of Stud Connectors in Lightweight and Normal-Weight As = effective cross-sectional area of a stud anchor (in.2)
Concrete. AISC Engineering Journal, V. 8, No. 2: pp. 55–64. Ase = effective cross-sectional area of a stud anchor
25. Martin, L. D., and W. J. Korkosz. 1982. Connections for Pre- (in.2), ACI 318 Appendix D notation
cast Prestressed Concrete Buildings, Including Earthquake Re- db = diameter of the anchor
sistance. Technical Report No. 2. Chicago, IL: PCI. de1 = side-edge distance normal to the shear load appli-
26. Industry Handbook Committee. 1992. PCI Design Handbook: cation direction, parallel to the x-axis, taken from
Precast and Prestressed Concrete. 4th ed. MNL-120-92. Chi- the center of an anchor shaft to the side concrete
cago, IL: PCI. edge (in.)
27. American Society for Testing and Materials (ASTM). 2004.
de2 = side-edge distance normal to the shear load appli-
Standard Specification for Carbon Steel Bolts and Studs. ASTM
cation direction, parallel to the x-axis, taken from
A307-04. Volume 01.08. West Conshohocken, PA: ASTM.
28. Rong, A. Y., and A. Fafitis. 1989. Tensile and Shear Strength of the center of an anchor shaft to the side concrete
Single and Group Studs. Tempe, AZ: Civil Engineering Depart- edge (in.); also, the side-edge distance opposite de1
ment, Arizona State University. de3 = front-edge distance parallel to the shear load
29. Comité Euro-International Du Béton (CEB). 1994. Fastenings application direction and y-axis, taken from the
to Concrete and Masonry Structures. Lausanne, Switzerland. center of a front-anchor shaft to the front concrete
30. Mindess, S., and J. F. Young. 1981. Concrete. Englewood edge (in.)
Cliffs, N.J.: Prentice-Hall Inc. de4 = back- or rear-edge distance parallel to the shear
31. Kuhn, D. P., and A. F. Shaikh. 1997. Pilot Study on Headed An- load application direction and y-axis, taken from
chor Studs: A Comparison between PCI and CCD. Milwaukee, the center of a back anchor shaft to the rear con-
WI: Department of Civil Engineering, University of Wiscon-
crete edge (in.)
sin–Milwaukee.
32. ASTM. 2005. Standard Specification for Carbon Structural
d0 = shaft diameter of a headed stud (in.)
Steel. ASTM A36-05. V. 01.04. West Conshohocken, PA: f c' = specified compressive strength of concrete (psi)
ASTM. Fut, fut = specified ultimate tensile strength of anchor steel
33. ASTM. 2004. Standard Specification for Structural Steel Shapes. in tension (psi)
ASTM A992-04a. V. 01.04. West Conshohocken, PA: ASTM. Fvy = shear yield strength of anchor steel (psi)
34. ASTM. 2003. Standard Specification for Annealed or Cold- Fy, fy = specified yield strength of anchor steel in tension (psi)

January–February 2007 19
h = thickness of a concrete member in which the λ = concrete density factor
anchors are embedded, measured parallel to the = 1.0 for normalweight concrete
anchor axis (in.) = 0.85 for sand lightweight concrete
hef = effective headed-stud embedment depth taken as the = 0.75 for all lightweight concrete
length under the head to the concrete surface (in.) κ = one-sided population limit (fractile) factor for a
n = number of anchors in a connection or group normal distribution
Ncb = nominal concrete breakout strength in tension of a µ = coefficient of friction
single anchor (lb), ACI 318 Appendix D notation φ = strength reduction factor
Nu = concrete tensile breakout capacity for a single stud
tp = thickness of the attachment plate (in.)
Vn = nominal shear strength (lb) SI Equivalents
Vs, Vsteel = nominal shear strength of a single headed stud or
group of headed studs governed by steel strength 1 in. = 25.4 mm
x = center-to-center spacing of stud anchors in the x 1 ft = 0.3048 m
direction of the Cartesian plane (in.) 1 lb = 4.448 N
X = out-to-out X-row spacing = Σx 1 kip = 4.448 kN
y = center-to-center spacing of stud anchors in the y 1 psi = 6.895 kPa
direction of the Cartesian plane (in.) 1 ksi = 6.895 MPa
Y = out-to-out Y-row spacing = Σy 1 yd3 = 0.7646 m3

20 PCI JOURNAL