HBRC Journal (2016) 12, 2532

Housing and Building National Research Center

HBRC Journal

http://ees.elsevier.com/hbrcj

Experimental study of prequalied status

of ush end plate connections
a,*
Sherif H.M. Hassanien , Hazem M. Ramadan b, M. Nabil Abdel-Salam c,
Sherif A. Mourad d

a
Structures and Metallic Construction Institute, Housing and Building National Research Center, HBRC, Egypt
b
Structural Engineering Department, Faculty of Engineering, Cairo University, Egypt
c
Structural Engineering, Structures and Metallic Construction Institute, Housing and Building National Research Center,
HBRC, Egypt
d
Faculty of Engineering, Cairo University, Egypt

Received 18 June 2014; revised 29 June 2014; accepted 30 June 2014

KEYWORDS
Flush end-plate connection;
Seismic performance;
Steel structures;
Experimental testing;
Prequalied connections
Abstract Seismic design of steel structures is an essential part of the design process. Egyptian load-
ing code development process continues in a high rate to catch up with emerging new concepts and
standards. Steel design codes (ASD and LRFD) are not developing in the same speed, which
prevents the full utilization and application of loading code. The above reason leads to the need
for evaluating ush end plate connections from prequalication point of view according to interna-
tional standards. Due to the lack of sufcient experimental data on ush end-plate connections, an
experimental program was conducted to investigate this topic. Six ush end-plate samples were
designed according to the Egyptian code for steel construction (ECP205 ASD) using different beam
and column sections, bolt diameters and grades. A cyclic loading pattern dened by international
standards was used in the testing process, and the performance was evaluated accordingly.
Evaluation of MU curves showed that in some cases ush end plate connections satisfy the strict
requirements for prequalication. However, beam sections having limited depth fail to achieve
prequalication criteria for the connections. Reduced web may be used to enhance the connection
status and is investigated in one of the samples to evaluate its impact on connection performance
and the failure mode. The proposed staggered hole conguration showed a promising performance.
2014 Production and hosting by Elsevier B.V. on behalf of Housing and Building National Research
Center.

* Corresponding author. Introduction

Peer review under responsibility of Housing and Building National
Research Center. Seismic design of steel structures has taken major steps since the
1994 Northridge earthquake Tremblay [1]. Seismic design codes
try to cope with the rapid changes in seismic design procedures
Production and hosting by Elsevier and concepts. Current codes utilize the design spectrum method

http://dx.doi.org/10.1016/j.hbrcj.2014.06.013
1687-4048 2014 Production and hosting by Elsevier B.V. on behalf of Housing and Building National Research Center.
26 S.H.M. Hassanien et al.

to estimate seismic forces on structures. One of the key aspects of carried out. The tests were conducted at the laboratory of the
the response spectrum analysis is the response modication fac- Housing and Building Research Center (HBRC), Egypt. The
tor (RMF), which reduces earthquake loads due to the nonlin- studied parameters were bolt diameters, bolt grades and
ear structural behavior. Two values of RMF can be used for arrangements. The connections were designed such that the
moment resisting frames. Using a high RMF value requires that head plate does not reach its yield stress, hence, the plate thick-
the frame possesses sufcient ductility. Ductility requirements ness was conservatively chosen to ensure such assumption and
for steel structures are dened in international codes such as match AISC requirements for thick end plates. IPE 160 and
AISC [2] and standards like FEMA350 [3]. The requirements HE120A beam sections were selected to study different bolt
vary from one country to another according to loads, factor of spacings and arrangements. Both beams have the same bend-
safety, design methods, and construction methods. To date, ing strength since their section moduli are almost equal to
Egyptian steel structure design codes ECP-ASD [4] and ECP- avoid its effect on connection strength. Fig. 1 shows a general
LRFD [5] do not include any seismic provisions to ensure the layout of the test setup. The specimen dimensions and compo-
ductility of steel structures. Lack of seismic provisions leads to nents are summarized in Tables 1 and 2. Beam length in all
two major problems; the rst is that the designer is unable to uti- specimens was 1000 mm from the point of load application
lize a high value of the RMF, resulting in an uneconomic design. to the column face. This distance was chosen as the least prob-
The second problem is that evaluation and retrotting of exist- able distance in order to decrease the moment applied on the
ing steel structures are limited to the use of the smaller RMF, specimens to the least value. Butt welds were used to join the
which constrains the assessment. Each country has its own end plate to the steel beam. To examine the effect of web holes
methods of erection and construction, which are reected on on connection behavior and plasticity, one sample, S4, was
the standards of the country. AISC [6] and RCSC [7] dene designed considering staggered holes in its web as shown in
the state of snug tightened connection and the installation pro- Fig. 1.
cedure for bolts. ECP-ASD [4], which is the most used code for
steel design in Egypt, does not address the installation, plate sur- Instrumentation
face condition and the denition of bearing type connection.
AISC [2] lists three major conditions that must be fullled The instrumentations were designed to measure the applied
in moment resisting frames in order to be considered special loads, deformations along the specimen and strains at specic
moment frames and to achieve high ductility. These conditions locations in the specimens as shown in Fig. 2. Electrical strain
are: members slenderness, column stress ratio and connection gauges, with 10 mm grid length, were glued to measure strains
ability to sustain predened percentage of beam moment at different locations at the connection zone during different
capacity combined with certain rotation angle. Connections stages of loading. Four linear variable displacement transduc-
fullling previous requirements are said to be prequalied. ers (LVDTs), with an accuracy of 0.01 mm, were used to
FEMA350 [3] demonstrates how to prequalify a connection measure displacements at different locations in the specimens.
or dene its status. The process requires performing quasi-sta-
tic test with predened loading pattern such as SAC97 [8] on
the joint. Researches in end plate moment connection have
been initiated in the last century, however, limited number of
experimental researches in the area of ush end plate connec-
tions have been conducted. Experimental work and proposed
basis of design were presented by Jenkins et al. [9] and Thom-
son and Broderick [10,11] have studied seismic response of
ush end plate connections experimentally, and concluded that
connection stiffness is overestimated while its moment capacity
is underestimated using a proposed Eurocode model. The
papers presented a test setup that was modied and used in this
current experimental program. Hedayat [12] presented a
reduced web technique to enhance post-Northridge connection
ductility, which triggered the idea of staggered web congura-
tion, which is tested in this paper.
This paper presents an experimental program, which is a
part of a comprehensive research conducted by Hassanien
[13] including six samples designed according to ECP [4] and
AISC [14] to evaluate its seismic performance. The test results
are used to verify the numerical models, which were developed
in that research. The design is based on thick plate assumption,
which limits the failure of the connection to bolt failure and
prevents the formation of any plastic hinges in plate.

Experimental program

A series of six tests designed to investigate the behavior, load

capacity and failure mode of ush end plate connections were Fig. 1 Experimental test set up (dimensions in mm).
Experimental study of prequalied status of ush end plate connections 27

Table 1 Sample dimensions.

Test tp (mm) H (mm) B (mm) e1 (mm) e2 (mm) e3 (mm) e4 (mm)
S1 25 142 120 14.5 35 48 0
S2 25 188 117 14 33.5 44 0
S3 33 137 120 11.5 35 38.5 0
S4 33 188 117 14 33.5 44 50
S5 22 190 118 15 34 45 50
S6 15 190 118 15 34 45 50

Table 2 Sample components.

Test Number of bolts Bolt grade Bolt diameter (mm) Beam Col. Web holes
S1 4 10.9 16 HE120A HE140A No
S2 4 10.9 20 IPE160 HE120A No
S3 4 8.8 20 HE120A HE140A No
S4 6 10.9 20 IPE160 HE140A Yes
S5 6 10.9 16 IPE160 HE140A No
S6 6 8.8 16 IPE160 HE120A No

Fig. 2 Instrumentation layout.

Two of them were attached to the column ange to obtain col-
umn rotation while the others were attached to the beam
anges to measure the whole connection rotation. The value
of the applied load was measured directly by a load cell
attached to the actuator.
Loading procedure
Column was supported laterally at both ends through existing
lateral frames. Axial constant load of 98.1 KN was applied on
the column top to support them vertically during the applica-
tion of cyclic vertical load at the beam tip. The load value is
estimated based on the maximum predicted cyclic load value
during test in order to keep the column under compression.
All tests are subjected to a loading history which includes
successive different displacement cycles according to the step-
wise loading protocol as shown in Fig. 3 (SAC97 [8]). The
applied displacement rate was chosen the slowest as possible
in order to preclude any impact or dynamic effects and
guarantee better observation of the testing procedure, while
preserving a reasonable testing period. The applied displace-
ment was gradually increased until the specimen showed large
deformations (one-tenth of beam length) or failure of one of
the connection components.
Results and discussion
All samples were tested till applied displacement limit of 1/10
of beam span except two samples S3 and S4. The test of sample
S3 was terminated after 40 mm displacement due to an error in
28 S.H.M. Hassanien et al.

Fig. 3 Sample of cyclic loading pattern.

the data logger device, whereas S4 test ended after 80 mm

displacement due to observation of beam instability. Data
extracted from the tests are used to obtain connection capacity
curves, hysteresis responses and level of ductility.

Moment-rotation curves

Figs. 49 show the resulting moment-rotation curves for all

samples. The moment values are calculated as the product of
the load times its lever arm of 1000 mm, which is the distance
from the loading point to the column ange. The rotation val-
ues are calculated using the relative displacements of the beam
top and bottom anges and the column to account for the col-
umn panel zone rotation and the end gap rotation caused by
the relative deformation between the end plate and the column
ange. Close examination of these curves highlights the follow-
ing notes:
Fig. 5 S2 moment-rotation curve.
 The hysteresis loops of the specimens are un-pinched and
very stable.
 Connections behave in a stiff linear manner at the begin- 45.0
Moment (KN.m)

ning of loading till reaching yield strength of either head

plate or bolts. Then the behavior changes to non-linear 30.0
for all values of head plate thicknesses and bolt diameters.

15.0

Rotation (m.rad)
0.0
-150 -100 -50 0 50 100 150

-15.0

-30.0

-45.0

Fig. 6 S3 moment-rotation curve.

 The initial rotational stiffness dened as the slope of

moment-rotation curve linear stage varies for the samples,
Fig. 4 S1 moment-rotation curve. as shown in Table 3.
Experimental study of prequalied status of ush end plate connections 29

45.0 45.0

Moment (KN.m)

Moment (KN.m)
30.0 30.0

15.0 15.0

Rotation (m.rad) Rotation (m.rad)

0.0 0.0
-150 -100 -50 0 50 100 150 -150 -100 -50 0 50 100 150

-15.0 -15.0

-30.0 -30.0

-45.0
-45.0

Fig. 7 S4 moment-rotation curve. Fig. 9 S6 moment-rotation curve.

connection improve remarkably. This is clear since the

moment capacity and stiffness of the ush end plate connec-
tion increase with thicker end plates. However, increase in
the endplate thickness did not result in an increase in con-
nection ductility.
 To examine the connection prequalication status in accor-
dance to FEMA 350 [3], the applied moments were calcu-
lated at different levels of rotation. As per AISC [2]
requirements, the connection should sustain moment with
a value higher than 80% of beam plastic capacity at
40 m rad rotation. The measured moment values are listed
in Table 4 which show that specimens S2, S4, S5 and S6
satisfy the code requirements while S1 and S3 fail.

Fig. 8 S5 moment-rotation curve.
 On comparing the samples having the same beam depth,
HE120A, sample S3 shows a higher ductility while sample
S1 develops the highest moment. For samples with beam
section IPE 160, values of resisting moments and ductility
are close.
 It is apparent that end plate thickness affects the behavior
of ush end plate connections signicantly, and when
it increases, the workability and deformability of this
Strain measurements
Strain measurements at different locations of the specimens are
plotted versus the lateral applied displacements in Figs. 1018.
The main observations are summarized as follows:
 Strain values measured at the top and bottom of the beam,
Gauges 8 and 9, reverse at the connection zone according to
the direction of loading (Figs. 10 and 13).
 Compared to the beam yield strain (1.42 10 3), which was
obtained from the coupon test results, it is clear that yield
strain was recorded at gauge 8 at high loading displacements

Table 3 Initial rotational stiffness.

Sample S1 S2 S3 S4 S5 S6
Initial rotational stiness (KN m/mrad) 1.14 0.94 0.59 0.77 1.61 1.37

Table 4 Measured moment values at 40 m rad.

Sample S1 S2 S3 S4 S5 S6
Measured moment (KN m) 21 32 18 31 34 28
Beam plastic capacity (KN m) 31.70 32.84 31.70 32.84 32.84 32.84
30 S.H.M. Hassanien et al.

120 120

Displacment (mm)

Displacment (mm)
80 80

40 40

Strain x 1000 Strain x 1000

0 0
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2 -1.2 -0.6 0 0.6 1.2

-40 -40

-80 -80

-120 -120

Fig. 10 Specimen S4 strains at gage-8. Fig. 13 Specimen S4 strains at gage-9.

120
Displacment (mm)

120

Displacment (mm)
80

80
40
40
0 Strain x 1000
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
0
Strain x 1000 -0.45 -0.35 -0.25 -0.15 -0.05 0.05 0.15
-40
-40
-80
-80

-120
-120
Fig. 11 Specimen S5 strains at gage-8.
Fig. 14 Specimen S4 strains at gage-6.

120
Displacment (mm)

120
Displacment (mm)

80
80

40
40
Strain x 1000
Strain x 1000
0
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2 0
-0.45 -0.35 -0.25 -0.15 -0.05 0.05 0.15

-40 -40

-80 -80

-120 -120

Fig. 12 Specimen S6 strains at gage-8. Fig. 15 Specimen S5 strains at gage-6.

Experimental study of prequalied status of ush end plate connections 31

120 120

Displacment (mm)
Displacment (mm)
80 80

40 40
Strain x 1000
0 0
-0.45 -0.35 -0.25 -0.15 -0.05 0.05 0.15 -0.3 -0.2 -0.1 0 0.1 0.2 0.3
Strain x 1000
-40 -40

-80 -80

-120
-120

Fig. 16 Specimen S6 strains at gage-6. Fig. 18 Specimen S4 strains at gage-25.

120
using different beam and column sections, bolt diameters and
Displacment (mm)

grades. A cyclic loading pattern dened by international stan-

80 dards was used in the testing process and the performance was
evaluated accordingly. Analyzing the obtained results, the
40 following could be concluded:

Strain x 1000
0
-0.3 -0.2 -0.1 0 0.1
-40
-80
-120
Fig. 17 Specimen S4 strains at gage-24.
for S5 and S6 (Figs. 11 and 12) and at a lower one for S4
(Fig. 10). This indicates that with web hole conguration,
the plastic hinge moves away from connection toward the
connecting beam. This is very desirable in seismic design
provisions where the formation of the plastic hinges at the
location of connection and its vicinity must be avoided.
 On comparing the direction of strains measured by gauge 6
(Figs. 1416), it is found that the presence of web holes in
test S4 changes the load transfer mechanism. Gauge 6
measures the strain along the beam web.
 Specimen S4 diagonal strains were measured through
gauges 24 and 25 (Figs. 17 and 18) to investigate its varia-
tion with loading between the holes. It is apparent that
gauge 25 shows higher strain values compared to gauge
24 readings at the same level of displacement or loading
which may be attributed to higher stresses at the location
of gauge 25.
Conclusions
An experimental program including six samples was conducted
to investigate the behavior of ush end-plate connections when
subjected to cyclic loading. The samples were designed accord-
ing to the Egyptian code of steel construction (ECP205-ASD)
(1) The end plate thickness affects the behavior of ush end-
0.2 0.3 plate connections signicantly. This is clear since the
moment capacity and stiffness of the ush end plate
connection increase with thicker end plates. However,
the increase in the end plate thickness did not result in
an increase in connection ductility.
(2) The initial rotational stiffness of ush end-plate connec-
tions depends signicantly on the connection parameters
such as end plate thickness, bolt diameter and grade.
(3) The designed ush end plate connections have satised
the strict requirements of FEMA 350 prequalication
criteria except for beam sections with limited depth.
(4) The presence of beam-web holes may change the load
transfer mechanism and increase the web strains.
(5) Beam web hole conguration in ush end plate connec-
tions allows the plastic hinge to move away from the
connection zone toward the connecting beam, which is
desirable in seismic design provisions.
Conict of interest
None.
References
[1] R. Tremblay, P. Timler, M. Bruneau, A. Filiatrault,
Performance of steel structures during the 1994 Northridge
earthquake, Can. J. Civ. Eng. 22 (1995) 338360.
[2] ANSI/AISC. Seismic Provisions for Structural Steel Buildings.
ANSI/AISC 34105. Chicago: The American Institute of Steel
Construction; 2002.
[3] FEMA. Recommended Seismic Design Criteria for New Steel
Moment-Frame Buildings. FEMA-350. Washington: Federal
Emergency Management Agency; 2000.
[4] ECP. Egyptian Code of Practice for Steel Construction and
Bridges (Allowable Stress Design). ECP-205 ASD. Egypt:
HBRC; 2007.
32
[5] ECP. Egyptian Code of Practice for Steel Construction and
Bridges (Load and Resistance Factor Design). ECP-205 LRFD.
Egypt: HBRC; 2012.
[6] ANSI/AISC 36005. Specication for Structural Steel Buildings.
American Institute for Steel Construction. ANSI/AISC 36005.
Chicago: The American Institute of Steel Construction; 2005.
[7] RCSC. Specication for Structural Joints Using ASTM A325 or
A490 Bolts. Chicago: Research Council on Structural
Connection; 2004.
[8] SAC97. Protocol for Fabrication, Inspection, Testing, and
Documentation of Beam-Column Connection Tests and other
Experimental Specimens. SAC/BD-97/02. California: Sac Joint
Venture and Its Organization; 1997.
[9] W.M. Jenkin, C.S. Tong, A.T. Prescott, Moment-transmitting
endplate connections in steel construction and a proposed basis
for ush endplate design, Struct. Eng. 5 (64A) (1986) 121132.
S.H.M. Hassanien et al.
[10] A.W. Thomson, B.M. Broderick, Seismic resistance of ush
endplate connections, Struct. Eng. 17 (78) (2000) 2833.
[11] B.M. Broderick, A.W. Thomson, The response of ush end-
plate joints under earthquake loading, J. Constr. Steel Res. 58
(2002) 11611175.
[12] A. Hedayat Amir, Celikag Murude, Post-Northridge connection
with modied beam end conguration to enhance strength and
ductility, J. Constr. Steel Res. 65 (2009) 14131430.
[13] S. H. Hassanien, Prequalication status of ECP designed ush
end plate connection and staggered web holes retrotting
technique. (Ph.D. dissertation), Cairo University, Faculty of
Engineering, Egypt (in progress).
[14] American Institute of Steel Construction. Flush & Extended
Multiple Row Moment End Connections. Steel Design Guide
No16. Chicago; 2002.