Journal of Constructional Steel Research 67 (2011) 379388

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

journal homepage: www.elsevier.com/locate/jcsr

Bolted steel slip-critical connections with fillers: I. Performance

Daniel J. Borello a , Mark D. Denavit a , Jerome F. Hajjar b,
a

Department of Civil and Environmental Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, USA

Department of Civil and Environmental Engineering, Northeastern University, Boston, MA, USA

article

info

Article history:
Received 11 May 2010
Accepted 4 October 2010
Keywords:
Steel connections
Slip-critical bolted connections
Bolted bearing connections
Filler plates

abstract
This paper reports the results of sixteen experiments of bolted slip-critical connections with fillers. Fifteen
of the connections used oversized holes and one connection used standard holes. Such connections with
oversized holes are commonly fabricated for use with structures such as long-span trusses, since the
use of oversized holes allows erection in-place rather than first assessing fit-up on the ground. Filler
plates are used to connect members of different depths or widths. The sixteen experiments reported
herein highlight the behavior of bolted steel connections with oversized holes in which fillers are
included and are undeveloped, partially developed, or fully developed. Both single-ply and multi-ply
fillers are investigated, as are welded developed fillers, and specimens fabricated using either turn-ofthe-nut or tension control bolts to pretension the connection. The results augmented with previous
literature document the slip and shear strengths of these connections, propose formulas for assessing
these strengths for the different conditions investigated, and provide recommendations for design of these
types of connections.
2010 Elsevier Ltd. All rights reserved.

1. Introduction
Filler plates are used in bolted steel connections where hotrolled structural steel members of different depths are joined. Filler
plates are commonly found in long-span truss connections, steel
girder splices, and column splices. Typical filler plate thicknesses
range from 6 to 102 mm (1/4 to 4 in.) or larger. For long-span
trusses in particular, recent fabrication and erection practices have
favored the use of oversized holes in connections with fillers so
that the trusses may be erected in place without first checking
fit-up through a trial erection on the ground. Previous work has
concentrated on the behavior of connections with filler plates
thinner than 25 mm (1 in.). The 2005 Specification for Structural
Steel Buildings of the American Institute of Steel Construction
(AISC) [1] accounts for fillers in Section J5 by requiring a reduction
in the bolt shear strength, a requirement that the connection be
designed as slip-critical at strength load levels, or development
of the filler (or, equivalently, extension of the connection splice
plate to include bolts sufficient to develop the filler). Fillers are
considered developed when an additional number of bolts or
equivalent welds are provided to enable a more uniform stress
distribution throughout the combined section of the connecting

Corresponding address: Department of Civil and Environmental Engineering,

400 Snell Engineering Center, 360 Huntington Avenue, Northeastern University,
Boston, MA 02115, USA. Tel.: +1 617 373 3242; fax: +1 617 373 4419.
E-mail address: jf.hajjar@neu.edu (J.F. Hajjar).
0143-974X/$ see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jcsr.2010.10.003

member and the filler plates. The number of development bolts

is typically determined by multiplying the number of bolts in the
connection by the ratio of the filler cross-sectional area to the
combined cross-sectional area of the connected member and filler
plate. If the filler and connected member are constant width, this
ratio reduces to the thickness of the filler divided by the combined
thickness of the filler and connected member. The filler plate can
also be developed by a weld to the connected member with an
equivalent strength as the slip strength of the required number of
bolts.
Eurocode 3: Design of Steel StructuresParts 18: Design of
Joints [2] refers to filler plates as packing. In bolted connections,
fillers are accounted for in Section 3.6.1 by specifying a bolt shear
strength reduction factor as a function of filler thickness and bolt
nominal diameter. Development of the filler has not provided an
alternative to the bolt shear strength reduction. As in the 2005 AISC
Specification [1], a reduction is not required for very thin fillers,
specifically, those thinner than one-third the nominal diameter of
the bolt. Early research on the effect of fillers in bolted connections
included a series of tests conducted at Dorman Long and Company
in 1965 which used fillers in a bolted splice connection, reported by
Lee and Fisher [3]. The specimens were constructed using standard
size bolt holes with filler plate thicknesses ranged from 2 to 25 mm
(1/16 to 1 in.). In 1981 Frank and Yura [4] further studied the effect
of undeveloped fillers. The tests specimens had filler thicknesses
ranging from 2 to 19 mm (0.075 to 0.75 in.) using standard size
holes. The effect of multiple filler plates i.e., multi-ply fillers, was
investigated by testing a single 19 mm (0.75 in.) thick plate and
three 6 mm (0.25 in.) plates.

380

D.J. Borello et al. / Journal of Constructional Steel Research 67 (2011) 379388

Table 1
Specimen test matrix.
Specimen name

Upper column

Development

730-std
730-over
159f
159h
159n1
159n2
455f
455h
455n1
455n2
159n-2ply1
159n-2ply2
159h-TC
159n-TC
159f-weld
159h-weld

W14 730
W14 730
W14 159
W14 159
W14 159
W14 159
W14 455
W14 455
W14 455
W14 455
W14 159
W14 159
W14 159
W14 159
W14 159
W14 159

Undeveloped
Undeveloped
Fully developed: 4 bolt rows
Half developed: 2 bolt rows
Undeveloped
Undeveloped
Fully developed: 2 bolt rows
Half developed: 1 bolt row
Undeveloped
Undeveloped
Undeveloped
Undeveloped
Half developed: 2 bolt rows
Undeveloped
Fully developed: 406 mm (16 in.) of 13 mm (1/2 in.) fillet weld per edge of filler
Half developed: 330 mm (13 in.) of 8 mm (5/16 in.) fillet weld per edge of filler

Consistent with the observations of Lee and Fisher [3], Frank

and Yura [4] observed that the addition of a filler plate reduced
the slip resistance by approximately 17%. However, the addition
of multi-ply filler plates produced a more drastic reduction in
slip resistance, 46% below that of no fillers. The shear strength of
connections with fillers decreased with increasing filler thickness,
while the multi-ply fillers showed a small additional decrease in
shear strength relative to the connection with a single-ply filler.
Based on these experiments, an empirical equation for the shear
strength reduction was developed as a linear function of filler
thickness, which is utilized in the 2005 AISC Specification [1]. In
recent work by Dusicka and Lewis [5] experiments were performed
with high strength materials and fillers up to 51 mm (2 in.) thick.
Similar trends were found for thin fillers as in prior research.
However, the shear strength of the 51 mm (2 in.) thick fillers
was greater than that of the 25 mm (1 in.) thick fillers, indicating
that the detrimental effects do not monotonically increase with
filler thickness. Earlier work published in Japanese by Miyachi
and Koeda [6], Takizawa et al. [7], Sugiyama et al. [8], and Kanda
et al. [9] complements the studies summarized herein.
The present investigation explored the effect of thick fillers
[thicker than 25 mm (1 in.)] on the slip-critical strength and
shear strength of connections while also considering the effects
of development and multiple plies. This paper summarizes the
results of the experimental testing performed at the University
of Illinois at Urbana-Champaign. Details of the research program
are reported in [10]. Detailed analysis of the behavior of these
connections are discussed in [11].
2. Test setup
2.1. Test specimens
Sixteen specimens were tested, to replicate common bolted
splice connections between wide-flange members of different
depths, designed to explore the effect of filler thickness, filler
development, development method, and bolt pretension method.
The specimen test matrix is shown in Table 1. The specimen
name was based on the nominal weight of the top column (in lbs
per foot), the development (nnone, hhalf, ffull), and unique
details. Where duplicate specimens were tested, an additional
specimen number was added to the end of the designation. For
example, the second undeveloped specimen with a W14 159 top
column was identified as 159n2.
Each specimen, shown in Fig. 1, consisted of two wide-flange
members, connected by two 51 mm (2 in.) thick splice plates,
with filler plates provided where required to provide a constant
connection depth. The bottom column for all specimens was a

W14 730. The top column was a W14 159, W14 455, or
W14 730. The W14 159 and W14 455 top column specimens
required a filler plate of 95 mm (3 3/4 in.) and 41 mm (1 5/8 in.)
respectively. All surfaces of the specimens were blasted to a Class
B surface using a compressed air nozzle and G40 (type GL) steel
grit size steel shot. The resulting surface profile met SSPC-SP6 [12]
and measured 3.57 mils with a 0.43 mil standard deviation using
pressofilm tape. Two specimens (159n-2ply1 and 159-2ply2) were
tested with two undeveloped filler plates consisting of an 89 mm
(3 1/2 in.) and a 6 mm (1/4 in.) plate placed back-to-back, rather
than a single filler plate that was 95 mm (3 3/4 in.) thick, as
was used for the other W14 159 specimens. Two specimens
(159h-TC and 159n-TC) utilized ASTM F2280 Twist-Off tension
control structural bolts [13]. The remaining specimens utilized
ASTM A490 [14] structural bolts pretensioned using the turnof-nut method. For two specimens (159f-weld and 159h-weld)
the development was achieved by a fillet weld, instead of bolts,
between the filler plate and top column at each flange tip.
The splice plates were connected to the top and bottom
columns with twenty-four and sixty-four 229 mm (9 in.) long
bolts, respectively. Up to as many as sixteen 178 mm (7 in.) long
additional development bolts connected the filler plate to the
top column. The top column was assembled in a negative bearing
condition to provide the opportunity for the maximum possible
slip within the bolt holes between the top column, the filler
plate(s), and the splice plate. The bottom column was assembled
in a bearing condition between the column and splice plate to
effectively eliminate the possibility of slip in this location. All
bolts were 29 mm (1 1/8 in.) diameter with the threads excluded
from the shear plane. For the A490 bolts, the elongation of four
control bolts was monitored and if necessary re-torqued along
with neighboring bolts [10]. The bolt holes were oversized by
8 mm (5/16 in.) to a diameter of 37 mm (1 7/16 in.) for 15 of
the specimens; the bolt holes of the remaining specimen (730-std)
were standard [30 mm (1 3/16 in.) diameter]. An 89 mm (3 1/2 in.)
gap was provided between the top and bottom columns to allow
for movement between the columns.
The filler was designated as undeveloped, half developed,
or fully developed. Development was achieved using additional
178 mm (7 in.) long bolts or fillet welds between the filler plate
and top column. Full development was determined as the number
of bolts through the filler needed to uniformly distribute the load of
the connection across the filler plates and flange of the top column,
rounded to a whole number of bolt rows. For example, for the
specimens with W14 455 top columns [82 mm (3.21 in.) thick
flanges with 41 mm (1.625 in.) thick filler plates], the required
number of bolt rows for full development was 2 [e.g., (6 bolt rows)
(41 mm)/(82 + 41 mm)]. Half development was achieved using

D.J. Borello et al. / Journal of Constructional Steel Research 67 (2011) 379388

(a) Photograph of specimen.

(b) North and south schematic of specimen.

381

(c) East and west schematic of specimen.

Fig. 1. Typical test specimen.

Fig. 2. Strain gage instrumentation plan.

half of the number of bolt rows required for full development. Fillet
welds with strength equivalent to the slip strength of the bolts
was used to develop specimens 159f-weld and 159h-weld. If the
specimen was undeveloped no bolts were provided between the
filler plate and the top column. The specimens were instrumented
with strain gages on each component as shown in Fig. 2. With
the exception of the gages on the filler plates, the strain gages
were placed symmetrically about the strong and weak axes of the
columns. On the south side, seven strain gages were placed in a row
across the filler plate near the top edge of the splice plate (shown in
gray in Fig. 2). Strain gages were only applied to the filler plate if it
was developed by bolts. Several strain gages were damaged during

testing due to excessive strain or physical contact from breaking

bolts.
The specimens were instrumented with linear variable differential transformers (LVDTs), as shown in Fig. 3. The absolute displacement of the top column, filler plates and splice plates was
measured with respect to the floor of the testing machine. The difference between these absolute displacements could be used to assess the relative motion between the plates. LVDTs were also used
to directly measure the relative displacement between the filler
and splice plates as well as the top column and filler plates.
The applied load was measured using the testing machines
13.3 MN (3000 kip) hydraulic load cell. The stroke of the machine
crosshead was measured using the machines internal Yo-Yo gage.
The instrumentation was continuously recorded at 10 Hz, and
video (both low and high resolution) were recorded for each test.
The specimens were tested in axial compression due to the
fact that testing connections in tension at this scale would
have been prohibitive in cost. These types of connections are
generally subjected to compression as well as tension in the field;
consequently, investigation of compression forces are warranted.
Prior testing reported in the literature [15,16] on smaller-scale
specimens with potentially different eccentricities and stress
patterns from the specimens tested in this work has indicated that
specimens tested in tension may fail approximately 10% earlier
than specimens tested in compression due to prying of the lap
plates in these specimens. However, eccentricities, prying forces,
effects of Poissons ratio, and possible strength reductions of these
types of connections in compression versus in tension are complex,
with both types of loading causing potential detrimental effects,
and it is deemed that these tests in compression are appropriate
for comparison with prior research on bolted connections.
2.2. Testing procedure
The specimens were tested in Talbot Laboratory at the
University of Illinois at Urbana-Champaign in a 13.3 MN (3000 kip)
testing machine. The top loading platen contained a spherical head
that was locked into place after a load of 89 kN (20 kips) was

382

D.J. Borello et al. / Journal of Constructional Steel Research 67 (2011) 379388

(a) West side LVDTs.

(b) East side LVDTs.

Fig. 3. LVDT instrumentation plan.

Table 2
Summary of measured bolt properties.
Bolt property

A490 178 mm (7 in.) length

A490 229 mm (9 in.) length

F2280 178 mm (7 in.) length

F2280 229 mm (9 in.) length

Tb , pretension kN (kips)
Fv , shear strength MPa (ksi)
Fu , tensile strength MPa (ksi)

503 (113)
703 (102)
1103 (160)

512 (115)
683 (99)
1158 (168)

a
717 (104)
1186 (172)

a
745 (108)
1241 (180)

Pretension force exceeded machine capacity.

applied prior to testing to inhibit specimen rotation. Load was

applied by controlling the hydraulic oil pressure and volume. The
tests were generally carried out at an approximate loading rate of
4.5 kN (1 kip) per second to obtain a quasi-static loading protocol.
Typically, during sudden jumps in displacement, the load briefly
dropped and fluctuated as the hydraulic pressure stabilized. After
bolt shear failure of one side of the connection, the load was
immediately removed by the operator.
To verify instrumentation, the specimen was typically initially
loaded elastically to 890 kN (200 kips), then unloaded to zero.
Monotonic load was then applied to failure, with brief pauses to
observe behavior.
3. Predicted strength
3.1. Material properties
The column wide-flange shapes were specified as ASTM A992
steel [17]. The filler and splice plates were specified as ASTM A572
Grade 50 steel [18]. The plates and wide-flange shapes of the same
size came from the same heats (with exception of two heats for
the splice plates). Measured steel properties from the mill reports
are detailed in [10]. However, the strength of the specimens was
generally governed by the bolt properties. Therefore, these plate
and shape material properties did not control the ultimate strength
of the specimens, unless reported otherwise.
3.2. Ancillary tests
Ancillary tests were conducted to determine the bolt shear
strength, tensile strength, pretension force, and slip coefficient. The
bolt tension and shear strengths were determined as per ASTM
F606-06 [19] at the University of Cincinnati for both the A490

and F2280 bolts, with results summarized in Table 2. Bolts of the

same size and type (A490 or F2280) were from the same heat.
The relationship between the bolt pretension and bolt elongation
was determined using a SkidmoreWilhelm machine by W&W
Steel Corporation following standard procedures [4]. The number
of turns past snug-tight required for the A490 bolts to reach
the tension strength plateau, 5/6th of a turn, was subsequently
used for the installation procedure. The plateau force, Tb , was
assumed to be the clamping force per bolt due to pretension in
subsequent calculations of predicted slip strength. The capacity
of the SkidmoreWilhelm machine was reached prior to reaching
the pretension force of the F2280 bolts. Therefore insufficient
ancillary data was available to predict the slip strength of the
specimens utilizing F2280 bolts and their predicted slip strengths
are excluded from test-to-predicted comparisons in this work.
The slip coefficient was determined at the University of Illinois
at Urbana-Champaign using a procedure comparable to that
outlined in RCSC [20] Appendix A. The 2005 AISC Specification [1]
provides a nominal slip coefficient of 0.50 for Class B surfaces. In a
recent statistical study in which the results of 354 tests of blastcleaned surfaces were compiled, the mean slip coefficient was
determined to be 0.525 with a coefficient of variation of 0.193 [21].
Fourteen slip tests were performed, including combinations of
most of the faying surfaces present in the full-scale specimens,
utilizing plates from the same heats. The plate thickness was
not changed from the original plate thickness; therefore, the
thickness did not conform to the RCSC Specification [20]. To
mimic the RCSC Specification and bring the applied force closer
to the faying surface, 16 mm (5/8 in.) wide shims were placed
between the testing machine and specimen at a location closest
to each faying surface [10]. Table 3 summarizes the components
of slip test specimens and the experimental slip coefficient. The
average slip coefficient of the ancillary tests was 0.46. This value

D.J. Borello et al. / Journal of Constructional Steel Research 67 (2011) 379388

383

Table 3
Ancillary slip test matrix.
Specimen number

Outside plates

6B-3

Thickness

Splice

51 mm (2 in.)

6C-1
6C-2
6C-3

Splice

6D-2

W14 730

Inside plates

Type

51 mm (2 in.)

125 mm (4 15/16 in.)

Type

Thickness

Filler

41 mm (1 5/8 in.)

0.435

Filler

95 mm (3 3/4 in.)

0.58
0.55
0.445

Splice

51 mm (2 in.)

0.445

7B-1
7B-2
7B-3

W14 159

30 mm (1 3/16 in.)

Filler

89 mm (3 1/2 in.)

0.33
0.5
0.365

7C-1
7C-2
7C-3

W14 159

30 mm (1 3/16 in.)

Filler

95 mm (3 3/4 in.)

0.49
0.52
0.45

7D-1
7D-2
7D-3

W14 455

81 mm (3 3/16 in.)

Filler

41 mm (1 5/8 in.)

0.43
0.405
0.45

Average
Standard deviation
Coefficient of variation

0.46
0.07
0.15

Specimen 08: 455h

Load vs. Top Column Displacement

10000

10000

8000

8000

Load (kN)

Load (kN)

Specimen 08: 455h

Load vs. Time

6000

6000

4000

4000

2000

2000
08top-1e
08top-1w

0
0

10

20

30

Time (min)

10

20

30

40

Top Column Displacement (mm)

(a) Load versus time.

(b) Load versus top column displacement.

Fig. 4. Typical specimen response.

is within a standard deviation of the mean slip coefficient reported

in [21]. Therefore the ancillary tests were deemed as an accurate
representative of the slip coefficient present on the tested surfaces
and this value was used to predict the slip strength of each surface
in the large scale specimens.
3.3. Predicted strength
The predicted slip strength of the large connections utilizing
A490 bolts (the pretension force was unavailable for the F2280
bolt) was calculated as the product of the pretension force obtained
from the ancillary tests, the slip coefficient of 0.46 obtained from
the ancillary slip tests, the number of splice plate slip planes (equal
to 2, independent of the number of filler plates), and the number
of bolts in the splice plate per side of the connection (equal to 12).
The predicted shear strength was calculated as the product of the
measured bolt shear strength from ancillary bolt shear tests, and
the number of bolts between the splice plate and top column. The
predicted strength presented in this paper does not account for the
filler and hence represents the expected strength for an equivalent
connection without fillers; this is commonly the approach used in
practice. Predicted strengths were also developed using a similar
procedure for tests reported herein from the literature to provide
a consistent strength parameter for comparison.

4. Specimen behavior
The typical response of a typical specimen, including both load
versus time and load versus the displacement of the base of the
top column is shown in Fig. 4. Prior to slip, the loaddisplacement
relation exhibited elastic behavior, indicated by the stiff linear
response. Upon reaching the slip load, there was a sudden increase
in displacement, corresponding to the slip of at least one of the
faying surfaces (Fig. 5(a)). Since the machine could not maintain
force during this dynamic action (because the hydraulic oil volume
required time to restabilize), the load decreased after slip. As
the load was stabilized, slip continued. The load stabilization
typically took approximately 30 s, during which the specimens
often experienced additional slip events over a period of several
or tens of seconds following the initial event. The expected total
slip was twice the difference between the hole diameter and the
bolts shank diameter [3 mm (0.125 in.) for standard holes, 16 mm
(0.625 in.) for oversize holes], after which the bolts would be
expected to come into a bearing condition. Evidence of this is
seen in Fig. 4(b), indicated by a noticeable increase in stiffness
of the connection. As the load was further applied, the bearing
surfaces began to yield in compression and bolts began to yield
in shear, lowering the stiffness of the specimen. Ultimately, all
of the bolts going through the splice plate on one side of the

384

D.J. Borello et al. / Journal of Constructional Steel Research 67 (2011) 379388

(a) After slip.

(b) After shear.

Fig. 5. Specimens during testing.

specimen suddenly failed in shear (Fig. 5(b)), indicating that the

shear strength of the connection had been reached.
After initial slip between two surfaces, prior to shear failure of
the connection, six bolts (approximately 3%) prematurely failed
through the threads at the face of the nut (these bolt failures for
each specimen are itemized in [10]); no correlation was found with
overall specimen performance and these premature bolt failures,
as the shank of the bolt often stayed in the hole.
5. Slip strength
Slip was taken as the load at which relative displacement
between any two surfaces initiated. Slip was accompanied by a
loud noise and a violent vibration. Table 4 summarizes the design,
predicted (based on measured properties), and experimental slip
loads for each specimen. The average of all the test-to-predicted
ratios is greater than unity. This could be because of the uncertainty
of the slip coefficient, the randomness of bolt pretension despite
the use of control bolts, small eccentricities in the test specimens,
or mechanisms other than pure friction (such as gouging
between the plates) to resist some of the loading.
Consistent with previous work [4], there is only a slight
difference between the slip strengths of the 730 specimens with
standard and oversized holes. One of the specimens with two plies,
159n-2ply1, experienced very early slip between the two plies of
the filler plate on one side of the connection, at 52% of the predicted
load. The specimen was disassembled and there were no noticeable
irregularities on the surfaces that failed early. The duplicate of that
specimen, 159n-2ply2, experienced slip at 106% of the predicted
load, resulting in an average slip strength for the two specimens of
79% of the predicted value.
The slip test-to-predicted ratio of each specimen type is
summarized and disaggregated in Table 5. There is evidence that
each additional ply reduces the slip strength of the connection.
A linear regression analysis conducted on the slip strengths of
undeveloped fillers combined with previous literature is presented
in Fig. 6. In order to avoid a reduction for zero plies, the linear

regression excluded specimens without fillers. There is evidence

that addition of each ply reduces the slip strength of connections.
The mean test-to-predicted ratio for connections with 0 plies is
1.18; with 1 ply is 1.0; with 2 plies is 0.7; and with 3 plies is 0.51.
Based on a linear regression as shown in Fig. 6, the recommended
slip strength reduction factors are summarized in Table 6; it is
recommended to reduce the slip strength by 10% per ply of the filler
on each side of the connection. However, neglecting the reduction
for a single filler ply may be adequate and appropriate for design.
A statistical analysis of slip strength of connections with multiple
plies, including the effect of development is presented in [10,11].
The test-to-predicted ratios of slip strength versus filler
thickness are shown in Fig. 7 using the results of Lee and
Fisher [3], Frank and Yura [4], and this work. No discernable
trend can be identified, indicating that filler thickness does not
significantly influence the slip strength. The slip strengths for
the same set of tests separated by hole size are shown in Fig. 8.
No significant difference in strength is seen, indicating that it is
unnecessary to reduce the strength of connections for oversize
holes, except to account for the increased consequences of slip with
oversize holes [21]. Therefore, the slip strength of a connection is
predominantly seen to be a function of the number of filler plies
and independent of filler thickness and bolt hole size.
6. Shear strength
At failure, all bolts on one side of the specimen were
suddenly sheared into two pieces, often exiting the hole with
significant velocity. Table 7 summarizes the design, predicted and
experimental shear loads for each specimen. The shear strength
of specimens 159h-TC and 159n-TC exceeded the capacity of the
testing machine (13.3 MN (3000 kips)). The shear strength of
specimen 159f-weld was not achieved due to local buckling of the
top column prior to shear failure, but it exceeded its predicted yield
strength. For these three specimens, the resulting lower bound of
the measured shear strength was plotted but was neglected for
subsequent statistical calculations.

D.J. Borello et al. / Journal of Constructional Steel Research 67 (2011) 379388

385

Table 4
Slip strength experimental test results.
Name

Test kN (kips)

Pn kN (kips)a

Predicted kN (kips)

Test-to-predicted ratio

730-std
730-over
159f
159h
159n1
159n2
455f
455h
455n1
455n2
159n-2ply1
159n-2ply2
159h-TC
159n-TC
159f-weld
159h-weld

7549 (1697)
7268 (1634)
5445 (1224)
7549 (1697)
8358 (1879)
7580 (1704)
6090 (1369)
5227 (1175)
6174 (1388)
6374 (1433)
2927 (658)
5996 (1348)
7233 (1626)
5738 (1290)
7495 (1685)
7188 (1616)

4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)
4102 (922)

5649 (1270)
5649 (1270)
5649 (1270)
5649 (1270)
5649 (1270)
5649 (1270)
5649 (1270)
5649 (1270)
5649 (1270)
5649 (1270)
5649 (1270)
5649 (1270)
b
b
5649 (1270)
5649 (1270)

1.34
1.29
0.96
1.34
1.48
1.34
1.08
0.93
1.09
1.13
0.52
1.06

1.33
1.27

Average
Std. dev.

6512 (1464)
1317 (296)

a
b

1.15
0.25

Design values are from AISC 2005 [1].

Insufficient ancillary information available.

Table 5
Slip strength test-to-predicted ratio by specimen type.
Specimen type

Number of specimens

Slip strength test-to-predicted ratio

Mean value

Standard deviation

All
No filler
1 ply filler
2 ply filler
No development
1 ply filler, no development
1 ply filler, half and full development
1 ply filler, full development

14
2
10
2
6
4
6
3

1.15
1.31
1.19
0.79
1.10
1.26
1.15
1.12

0.25
0.04
0.18
0.38
0.33
0.18
0.19
0.19

Undeveloped Fillers Slip Test-to-Predicted Ratio

Table 6
Recommended multi-ply reduction factor for slip-critical connections with fillers.
Reduction factor
1.0
0.90a
0.80
0.70

A reduction factor of 1.0 may be adequate for design.

Specimens 730-std and 730-over achieved shear strengths close

to predicted values, with a test-to-predicted ratio of 1.04 and 1.01
respectively. All specimens with a W14 159 top column achieved
a shear strength at least 4% higher than the predicted value, with
an average test-to-predicted ratio of 1.13 with a standard deviation
of 0.05. There are several possible reasons for the test-to-predicted
ratios greater than unity, as discussed in [10,11].
All specimens with a W14 455 top column failed to achieve
the predicted shear strength, with the lowest specimen reaching
90% of the predicted strength. The mean and standard deviation
of the test-to-predicted ratios for these four columns were 0.93
and 0.05 respectively. They did, however, meet the 2005 AISC
Specification [1] design strength, presented in Table 7. Specimen
455f achieved the highest shear strength of the group with a testto-predicted ratio of 0.99.
The 159 specimens with one-ply fillers (159f, 159h, 159n1 and
159n2) demonstrated similar strength to the 159 specimens with
two-ply fillers (159n-2ply1 and 159n-2ply2). The two-ply filler
consisted of a thick ply and a relatively thin ply. Therefore, the influence of the thin plate was minimal. It is expected that multiple
plies of similar size would further reduce the shear strength of the
connection, since the bending restraint of the bolt within the bolt
hole of the thick filler would be reduced; these effects are investigated further in the mechanistic analysis presented in [10,11].

1.4
Slip Test-to-Predicted Ratio

Number of plies on one side of connection

0
1
2
3

1.6

Lee & Fisher

1.2

Frank & Yura

1.0
y = -0.11x + 1.00
R2 = 0.17

0.8

Bo rello,
Denavit &
Hajjar
Linear (All
Authors)

0.6

0.4
0

0.5

1.5

2.5

3.5

4.5

Number of Plies

Fig. 6. Undeveloped fillers slip test-to-predicted ratio versus number of plies.

The results from this study for the ultimate shear strength
test-to-predicted ratio versus filler thickness are combined with
the results from Frank and Yura [4] studies in Fig. 9, including
specimens without fillers and specimens with undeveloped singleply fillers. The bolt shear strength decreases as a function of
filler thickness for relatively thin fillers, but this reduction is
mitigated by having sufficiently large fillers, approximately 25 mm
(1 in.) or greater. The bolt shear strength reduction formula of
Section J5 of 2005 AISC [1] and two proposed reduction equations
(obtained from regression analyses based on single-ply fillers)
are also superimposed in Fig. 9. The two proposed equations are
identified as Reduction Equations (A) and (B), given by Eqs. (1) and
(2) respectively, where t is the thickness of the filler in millimeters
(or inches). To account for the presence of single-ply fillers the
bolt shear strength of the connection is then multiplied by the

386

D.J. Borello et al. / Journal of Constructional Steel Research 67 (2011) 379388

Table 7
Shear strength experimental test results.
Name

Test kN (kips)

Pn kN (kips)a

Predicted kN (kips)

Test-to-predicted ratio

730-std
730-over
159f
159h
159n1
159n2
455f
455h
455n1
455n2
159n-2ply1
159n-2ply2
159h-TCb
159n-TCb
159f-weldc
159h-weld

11,307 (2542)
10,938 (2459)
11,761 (2644)
12,931 (2907)
11,334 (2548)
11,637 (2616)
10,800 (2428)
9773 (2197)
9737 (2189)
10,000 (2248)
12,513 (2813)
13,038 (2931)
>13,345 (>3000)
>13,345 (>3000)
>12,099 (>2720)
12,250 (2754)

5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)
5968 (1342)

10,871 (2444)
10,871 (2444)
10,871 (2444)
10,871 (2444)
10,871 (2444)
10,871 (2444)
10,871 (2444)
10,871 (2444)
10,871 (2444)
10,871 (2444)
10,871 (2444)
10,871 (2444)
11,503 (2586)
11,503 (2586)
10,871 (2444)
10,871 (2444)

1.04
1.01
1.08
1.19
1.04
1.07
0.99
0.90
0.90
0.92
1.15
1.20
>1.16
>1.16
>1.10
1.13

Average
Std. Dev.

>11,386 (>2560)
>1123 (>252)

a
b
c

>1.07
>0.10

Design values are from AISC 2005 [1].

The shear strength of the TC specimens exceeded the capacity of the testing machine.
The top column of specimen 159f-weld experienced detrimental local buckling prior to the shear load.

reduction factor :

Slip Test-to-Predicted Ratio

1.6

(SI )
(2)

(US )

(SI )
(US )

Slip Test-to-Predicted Ratio

1.2

These formulas can be compared with the reduction formula in the

AISC Specification [1] (including a 6 mm (0.25 in.) offset to avoid a
reduction for thin fillers used for fit-up):

= 1 0.0154(t 6) 0.8
= 1 0.4(t 0.25) 0.8

1.4

(1)

1.0
Lee & Fisher
0.8

Frank & Yura

0.6

Borello, Denavit, &

Hajjar

0.4
0.2
0.0
0

(3)

As both proposed reduction equations presume that the shear

strength is influenced by the thickness of the fillers but is
independent of the number of plies, a further reduction for the
number of plies is discussed below. Eq. (1) (Reduction Equation (A))
does not account for the reduction mitigation provided by thick
fillers. The 2005 AISC filler reduction equation is calibrated for
6 mm (0.25 in.) connection deformation, as suggested by Frank and
Yura [4]. The proposed reduction equations are calibrated based
on ultimate shear strength, which requires significant connection
deformation to achieve. Calibration based on 6 mm (0.25 in.)
deformation is unreasonable for oversize holes due to the large
initial slip. A further alternative, with little loss in accuracy, would
be to retain the 2005 AISC [1] equation [Eq. (3)], which is presently
applicable for fillers 6 mm (0.25 in.) to 19 mm (0.75 in.) thick,
expand its applicability to all filler thicknesses, and increase the
minimum value of the reduction to 0.85 (which would occur
at a filler thickness of 16 mm (0.625 in.)), as shown in Fig. 9.
Increasing the minimum value of the reduction can also be justified
by noting that the ultimate shear strength is larger than the load at
6 mm (0.25 in.) deformation for which it was originally calibrated.
The bolt shear strength reduction in the Eurocode provisions [2]
is a function of the nominal bolt diameter; therefore, a direct
comparison between various tests is difficult. However, the above
reductions are still applicable.
The average shear strength test-to-predicted ratio for specimens with one-ply fillers from this and previous studies [3,4] was
approximately unity. The two-ply fillers from this study exhibited
an average shear strength 18% higher than predicted. However,

20

40

60

80

100

Filler Thickness (mm)

Fig. 7. Slip test-to-predicted ratio versus filler thickness.

Histogram of Slip Test-to-Predicted Ratio
6

4
Number of Tests

= 1 0.0051t 0.87 (SI )

= 1 0.13t 0.87
(US )

1 0.0051t
t 25 mm
=
0.87 + 0.0031(t 25) 25 mm < t 67 mm
1 0.13t
t 1 in.
=
0.87 + 0.08(t 1) 1 in. < t 2.625 in.

Standard
Oversize

0
0.4

0.5

0.6

0.7

0.8

0.9

1.1

1.2

1.3

1.4

1.5

1.6

Slip Test-to-Predicted Ratio

Fig. 8. Slip test-to-predicted ratio versus hole size.

since one of the plies was significantly thicker than the other, the
behavior of the connection was likely dominated by the thicker ply.
The fillers with three equal plies in the study by Frank and Yura [4]
exhibited an average shear strength 14% lower than predicted. The
mechanism associated with connection behavior and the influence
of multiple-ply fillers is explored in [10,11]. It is shown that the
shear strength of a connection may be reduced for connections

D.J. Borello et al. / Journal of Constructional Steel Research 67 (2011) 379388

Shear Test-to-Predicted Ratio
No Filler and Single-Ply Undeveloped Fillers
1.10
Frank & Yura

1.05
Borello,
Denavit, &
Hajjar

Test-to-Predicted Ratio

1.00

Reduction
Equation A

0.95

Reduction
Equation B

0.90

2005 AISC
(F&Y)

0.85

2005 AISC
(F&Y) with 0.85
Cap

0.80

0.75
0

20

40

60

80

100

Filler Thickness (mm)

Fig. 9. Shear test-to-predicted ratio versus filler thickness.

with fillers with greater than one ply to attain the best accuracy
in the test-to-predicted ratios, but that the value of the reduction
is modest (0.95 for two and three plies) and can likely be neglected
for design.
A direct comparison between different levels of development
in bolted connections is seen in this test series. For the specimens
with a W14 455 top column, a trend is seen in the effect
of development on the bolt shear strength. The fully developed
specimen, 455f, achieved 99% of its predicted shear strength while
the less than fully developed specimens, 455n1, 455n2, and 455h,
achieved 90%, 92% and 90% of their predicted shear strengths,
respectively. This data indicates that a connection with developed
fillers may not suffer the same detrimental effects in shear as
one with undeveloped fillers. It is seen in Table 7 that specimens
159f, 159h, 159f-weld, and 159h-weld achieved slightly higher
shear strengths than specimens 159n1 and 159n2. However,
the series of tests with a W14 159 top column all achieved
higher than predicted shear strengths, limiting the applicability
of a comparison based on development. Another option is to
consider part of the connection as effectively developing the rest
of the connection, dividing some of the bolts into development
bolts and connection bolts. A detailed analysis of developed and
undeveloped connections is outlined in [10].
7. Conclusions
The research presented in this paper and summarized in
detail in [10], augmented by previous studies from the literature,
demonstrates definitive trends regarding the influence of filler
plates on the slip and shear strengths of bolted connections,
including:

The connections tested in this work generally provided

excellent resistance to slip, with only three falling below
the predicted value, and with two of those having test-topredicted ratios above 0.93. When combined with assessment
of experiments reported in the literature on the behavior of
connections with fillers, the slip strength is seen to generally
be reduced by the introduction of filler plates, independent of
filler thickness and hole size. Recommended filler reductions
are presented, with a reduction of 0.8 recommended for twoply filler connections, and 0.7 recommended for three-ply filler
connections. The reduction in bolt shear strength for a singleply filler plates is modest (0.9) and could be neglected in design.
For connections with filler plates, the bolt hole oversize does
not affect the slip strength of the connection. Therefore, there

387

is no evidence to support a reduction in the design strength

for oversize holes based on experimental data. However, the
detrimental effect to the stability of the structure caused by slip
with oversize holes may warrant a more cautious treatment
in design than for connections with standard holes (e.g., by
designing slip-critical connections to have a lower reliability
against slip than for connections with standard holes).
The shear strength exhibited in these connections was consistently larger than the slip strength, and was larger than the
predicted value for all but four of the specimens. Those four
specimens all had test-to-predicted ratios larger than 0.90.
Therefore, it is recommended that the shear strength of connections with filler plates be reduced per Eq. (1), derived in this
work (rounded values in the equation may be adopted for design), or Eq. (3), consisting of the equation in the 2005 AISC Specification [1] but with a minimum value of the reduction of 0.85.
In addition, for these equations, a 6 mm (1/4 in.) offset, comparable to what is used in the 2005 AISC Specification [1], may
be used to avoid a reduction for thin fillers with little loss of
accuracy.
The specimens never failed in bolt shear immediately after bolt
slip (other than 12 occasional premature bolt failures [10]),
despite potentially large dynamic effects during slip; rather,
the specimens took on significant additional load prior to bolt
shear failure. Additional reductions for multiple plies may be
necessary, although the values are such that they become
significant only for connections with four plies or more per side
of the connection, which are rare.
In this work, a developed filler plate is seen to act quite integrally with the member to which it is connected. Developing or
extending the connection helps to mitigate reductions both in
slip strength due to multiple plies or in shear strength due to
thick fillers or multiple plies. However, for assessment of slip
in particular, the evidence is less definitive, because an inadequate number of tests have been completed, for example for
investigation of slip strength with multi-ply specimens that are
developed and also since the number additional bolts provided
by developing the connection depends on the size of the filler.
Specimens with filler plates welded to the columns also performed well. This research also shows that undeveloped connections generally perform well both for bolt slip and for bolt
shear. Reduction formulas and associated statistical assessment
are presented to account for the effect of the number of plies on
the slip strength and the filler thickness on the shear strength
for cases where the connection is not developed. Assessment
of the undeveloped connections as effectively developing a reduced number of bolts also shows that extending the connection works well to ensure the predicted slip and shear strengths
are reached in the connection.
Acknowledgements
This work was supported by the American Institute of Steel
Construction, W&W Steel Corporation, and the University of
Illinois at Urbana-Champaign. In-kind funding was provided by
Lohr Structural Fasteners. The authors thank G.A. Rassati and
J. Swanson of the University of Cincinnati for conducting the
ancillary bolt shear and tension tests; Prof. P. Dusicka of Oregon
State University and Prof. G. Grondin of the University of Alberta for
sharing data related to their research on bolted steel connections;
Prof. J. Philips, Director of the 3000,000 lb Testing Machine at
the University of Illinois at Urbana-Champaign, for his extensive
contributions to this research; Prof. J. Song of the University of
Illinois at Urbana-Champaign for his contributions to the statistical
assessment of slip strength; and K. Elam, D. Foley, G. Banas,
T. Prunkard, M. Bingham, and M. Parkolap of the University of

388

D.J. Borello et al. / Journal of Constructional Steel Research 67 (2011) 379388

Illinois at Urbana-Champaign for their pivotal contributions to

the execution of the experiments. The authors also thank the
members of the AISC Technical Advisory Panel (TAP) for their
excellent contributions to this research, including T. Schlafly (TAP
Chair and AISC Director of Research), C.J. Carter, and C.J. Duncan,
AISC; S. Armbrust, T. Winneberger, and W. Lindley, W&W Steel
Corporation; L.S. Muir and W.A. Thornton, Cives Engineering
Corporation; L.A. Kloiber, LeJeune Steel Company; Prof. G. Grondin,
University of Alberta; G. Heathcock, FabArc Steel Supply, Inc.; and
J.M. Fisher, Computerized Structural Design, S.C.
References
[1] American Institute of Steel Construction (AISC). ANSI/AISC360-05: specification for structural steel buildings. Chicago (IL): AISC; 2005.
[2] European Committee for Standardization (CEN). EN 1993-1-8: Eurocode 3:
design of steel structurespart 18: design of joints. Brussels (Belgium): CEN;
2003.
[3] Lee JH, Fisher JW. Bolted joints with rectangular of circular fillers. Report
no. 318.6. Fritz engineering laboratory. Bethlehem (PA): Department of Civil
Engineering, Lehigh University; June 1968.
[4] Frank KH, Yura JA. An experimental study of bolted shear connections. Report
no. FHWA/RD-81/148. Federal highway administration. Washington (DC): US
Department of Transportation; December 1981.
[5] Dusicka P, Lewis G. High strength steel bolted connections with filler plates.
Journal of Constructional Steel Research 2010;66:7584.
[6] Miyachi S, Koeda Y. Slip-behavior characteristics of bolted joint with a filler
plate. Kawada Technical Report 1999;18:689 [in Japanese].
[7] Takizawa S, Sawada H, Kuroda I, Abe Y. Slip test of friction type high
tension bolted joints according to guideline of steel road bridges. Bridge and
Foundation Engineering 1999;33(9):258 [in Japanese].
[8] Sugiyama N, Kubo M, Kondou A, Ito I, Oda H. Slip test on high-strength bolted
friction joints with filler plate. Reports of the Faculty of Science and Technology
2001;41:506 [in Japanese].

[9] Kanda K, Uemura A, Mori T. Slip resistance of high strength bolted single
friction joints with filler plates. Journal of Constructional Steel 2006;14:
63946 [in Japanese].
[10] Borello DJ, Denavit MD, Hajjar JF. Behavior of bolted steel slip-critical
connections with fillers. Report no. NSEL-017. Newmark structural laboratory
report series. Urbana (IL): Department of Civil and Enviromental Engineering,
University of Illinois at Urbana-Champaign; 2009.
[11] Denavit MD, Borello DJ, Hajjar JF. Bolted steel slip-critical connections with
fillers: II. Behavior. Journal of Constructional Steel Research 2011;67(3):
398406.
[12] The Society for Protective Coatings (SSPC). SSPC-SP 6/NACE no. 3: commercial
blast cleaning. Pittsburgh (PA): SSPC; 2007.
[13] American Society of Testing Materials (ASTM). Designation F2280-06:
standard specification for twist off type tension control structural
bolt/nut/washer assemblies, steel, heat treated, 150 ksi minimum 50 tensile
strength. West Conshohocken (PA): ASTM International; 2006.
[14] American Society of Testing Materials (ASTM). Designation A490-04: standard
specification for structural bolts, alloy steel, 53 heat treated, 150 ksi minimum
tensile strength. West Conshohocken (PA): ASTM International; 2004.
[15] Kulak GL, Fisher JW, Struik JHA. Guide to design criteria for bolted and riveted
joints. 2nd ed. New York (NY): John Wiley & Sons; 2001.
[16] Wallaert JJ, Fisher JW. Shear strength of high-strength bolts. Journal of the
Structural Division, ASCE 1965;91(ST3):99125.
[17] American Society of Testing Materials (ASTM). Designation A992/A992M-04:
standard specification for structural steel shapes. West Conshohocken (PA):
ASTM International; 2004.
[18] American Society of Testing Materials (ASTM). Designation A572/A572M04: standard specification for high-strength low-alloy columbium-vanadium
structural steel. West Conshohocken (PA): ASTM International; 2004.
[19] American Society of Testing Materials (ASTM). Designation F60606: standard
test methods for determining the mechanical properties of externally and
internally threaded fasters, washers, direct tension indicators, and rivets. West
Conshohocken (PA): ASTM International; 2006.
[20] Research Council on Structural Connections (RCSC). Specification for structural
joints using ASTM A325 or A490 bolts. Chicago (IL): RCSC; June 2004.
[21] Grondin G, Jin M, Josi G. Slip critical bolted connectiona reliability analysis for
design at the ultimate limit state. The American Institute of Steel Construction.
Edmonton (AB, Canada): Department of Civil and Environmental Engineering.
University of Alberta; April 2008.