Академический Документы
Профессиональный Документы
Культура Документы
www.elsevier.com/locate/jcsr
Abstract
A new theoretical model to evaluate the moment–rotation (M–φ) relationship for stiffened and extended steel beam–column end-plate
connections has been derived in this paper. Based on a specific definition of the end-plate connection rotation, the end-plate connection is
decomposed into several components, including the panel zone, bolt, end-plate and column flange. The complete loading–deformation process of
each component is then analysed. Finally the loading–deformation process for the whole connection is obtained by superimposing the behaviour of
each component. In addition, 5 joint tests have been conducted to verify the proposed analytical model. By comparing it with the test results, it has
been concluded that this analytical model can evaluate the rotational behaviour of end-plate connections, as well as the moment–rotation (M–φ)
curve and the initial rotational stiffness accurately. Furthermore, it can analyse every contribution to the joint’s rotational deformation, such as the
shear deformation of the panel zone, the bolt extension, the bending deformation of the end-plate and column flange, etc. This analytical model
also provides moment–shear rotation (M–φs ) and moment–gap rotation (M–φep ) curves, which establish a reliable foundation for analysing the
detailed rotational behaviour of end-plate connections.
c 2006 Elsevier Ltd. All rights reserved.
1. Introduction in the global analysis and design of steel frames with semi-rigid
connections. Hence, much effort has been focused in recent
Conventional analysis and design of steel frames are usually years toward determining connection moment–rotation (M–φ)
carried out under the assumption that the connections joining relationships [7].
the beams to the columns are either fully rigid or ideally Certainly, full-scale and carefully conducted joint experi-
pinned. In fact, as is evident from experimental observations, ments are the most reliable sources and direct method of ob-
all connections used in current engineering practice possess taining M–φ relationships. While more than 800 tests of beam-
rigidities which fall between the extreme cases of fully rigid to-column connections have been performed around the world
and ideally pinned [1], i.e., the connection is actually semi- today, only about 300 of them have provided currently useful
rigid. The behaviour of semi-rigid connections significantly moment–rotation data [2]. Since the connection details consist
influences not only the internal force distribution, but also of a number of components, any changes in these connection
the deformation of steel structures [2]. For most connections, details may lead to significant variations in the connection char-
the axial and shearing deformations are usually low compared acteristics [8]. In addition, many other variable parameters, for
to the rotational deformation. Consequently, for practical example, details of the fabrication and assembly of the connec-
design, it is essential to determine the connection’s rotational tions, also vary enormously and can affect their behaviour [9].
deformation. Therefore, almost all the steel design codes of
It is impossible to test all of the connections that might be used
different countries from all over the world [3–6] require that the
in steel construction. Some researchers such as Goverdhan,
effect of connection deformations should be taken into account
Nethercot, Kishi and Chen, have collected the available experi-
mental results and constructed steel connection data banks that
∗ Corresponding author. Tel.: +86 10 6278 2012; fax: +86 10 6278 8623. provide the user with not only the test data, but also some pre-
E-mail address: shiyj@tsinghua.edu.cn (Y. Shi). dictive equations [1]. Where the connections detailing, beam,
unstiffened columns, which may not be suitable for all the end-
plate connection types. Furthermore, the last part of this curve
is the horizontal line and this cannot consider the hardening
effect, which leads to great discrepancies with the actual
behaviour of the connection [32]. Finite element analysis is
precise and reliable, and can analyze complicated connection
profiles that may be difficult to investigate by experiment.
However, such finite element analysis generally requires a large
commercial package and is not therefore feasible for many
practical applications. Most importantly, the effects of torque,
lack of fit, construction imperfections and defects, etc., may be
difficult to include in a finite element model. A well-defined
theoretical method that can be easily carried out and is only
based on the connection’s details is needed indeed for the
analysis of the end-plate connection M–φ relationship.
Fig. 1. Test specimen and loading arrangement.
Table 2
Material properties
Material Measured yield Measured tensile Measured elastic Design value of bolt Measured bolt average
strength (MPa) strength (MPa) modulus (MPa) pre-tension force (kN) pre-tension force (kN)
Steel (thickness ≤ 16 mm) 391 559 190 707 – –
Steel (thickness > 16 mm) 363 537 204 228 – –
Bolts (M20) 995 1160 – 155 185
Bolts (M24) 975 1188 – 225 251
Table 3
Test results
Specimen Loading Moment Moment resistance change Initial rotational Initial rotational Failure mode
number capacity (kN) resistance (kN m) compared with EPC-1 (%) stiffness stiffness change
S j,ini (kN m/rad) compared with
EPC-1 (%)
EPC-1 286.4 343.7 0.0 52 276 – Bolt fracture
EPC-2 268.4 322.1 −6.3 46 094 −11.8 Bolt fracture
EPC-3 325.3 390.3 13.6 46 066 −11.9 Buckling of beam flange and
web in compression
EPC-4 342.3 410.8 19.5 47 469 −9.2 Buckling of beam flange and
web in compression
EPC-5 296.1 355.4 3.4 41 634 −20.4 Bolt fracture and buckling of
end-plate rib stiffener in
compression
and the column from its original configuration [1]. In this paper,
the joint rotation φ of the beam-to-column end-plate connection
is defined as the relative rotation between the centre lines of the
beam top and bottom flanges at the beam end, and it usually
includes two parts: the shearing rotation φs , contributed by the
panel zone of the column, and the gap rotation φep , caused by
Fig. 7. M–φ curves of the specimens.
the relative deformation between the end-plate and the column
flange, including the bending deformation of the end-plate and
S j,ini , of the connections are defined as the secant rotational
column flange as well as the extension of the bolts (Fig. 6).
stiffnesses up to this bending moment. According to Eurocode
The shearing rotation φs is calculated by ∆/ h t , and the gap
3 [3], S j,ini is compared to the flexural stiffness of the connected
rotation φep is calculated by δ/ h t , giving φ = φs + φep , where
beam EIb /L b . The connection is rigid when S j,ini is larger than
∆ is the displacement difference of the panel zone at the centre 25EIb /L b for unbraced frames, nominally pinned when S j,ini
lines of the top and bottom beam flanges at the beam end which is less than 0.5EIb /L b , and semi-rigid when S j,ini is between
can be measured by displacement transducer Nos. 13 and 14; these two values. If we assume that the natural beam length of
δ is the gap between the end-plate and the column flange at specimens is 1.2 m × 2 = 2.4 m, then EIb /L b is 9751 kN m,
the beam tension flange centre line, which can be measured by and S j,ini for all the specimens are about 4.3–5.4 times EIb /L b ,
displacement transducer No. 4; and h t is the distance between so all the tested connections are semi-rigid. If the tested beam
the centre lines of the top and bottom beam flanges, and is section sizes are applied in 4–5 m span frames and EIb /L b is
288 mm. M–φ curves of all the specimens are shown in Fig. 7. 4681–5851 kN m, the Sj,ini of all the connections are about 7.1–
M–φs and M–φep curves of each specimen are also measured 11.2 times EIb /L b , and therefore all the connections are still
to verify the analytical model, and the comparison results are classified as semi-rigid.
shown in Figs. 12 and 13. The influence of connection details on the behaviour of end-
The loading capacity presented in Table 3 is the maximum plate connections can be analyzed from Table 3 and Fig. 7.
pushing load applied. The moment resistance is calculated by EPC-1 is a reference specimen, and the other connection
multiplying the load with the arm of the loading(1.2 m). It can specimens alter only one or two parameters from EPC-1. EPC-
be seen from Fig. 7 that the M–φ curves of all the end-plate 2 has increased the end-plate thickness compared with EPC-
connection specimens are almost linear when the moment is 1, but its moment resistance decreases remarkably and its
less than 60 kN m; therefore the initial rotational stiffnesses, initial rotational stiffness changes less; its rotational stiffness
1284 Y. Shi et al. / Journal of Constructional Steel Research 63 (2007) 1279–1293
is much larger than EPC-1 when nonlinearity of the M–φ end-plate is thin and its bending stiffness is smaller, leading
curve occurs. EPC-3 has increased the bolt diameter, which also to the bolt tension force distribution among the four bolts in
means a larger bolt pre-tension force; its moment resistance tension being more uniform. Meanwhile, its initial rotational
is much higher than EPC-1, which fails by bolt fracture. stiffness decreases significantly, with excellent ductility and
EPC-4 has increased both the end-plate thickness and bolt rotation capacity.
diameter; its moment resistance is the largest among all the
4.2. Bolt force and distribution
specimens; higher rotational stiffness is observed during its
loading process. EPC-5 has reduced the end-plate thickness, The bolt tension forces generated by the applied moments
but its moment resistance increases on the contrary, because the are shown in Fig. 8. In this paper, the bolt tension force
Y. Shi et al. / Journal of Constructional Steel Research 63 (2007) 1279–1293 1285
Table 4
Moment values of curve of Specimens 1, 2 and 3 (kN m)
Here the actual increments on the bolt tension force, the bolt After the tension strain of the bolts(Fig. 10) in the tension
tension strain, the total tension force and the extension of the zone has been obtained, the actual increment on the bolt tension
bolt in the first bolt-row can be calculated by Eqs. (9)–(12) force, the total tension force and the extension deformation of
respectively. this bolt can be calculated. The corresponding joint moment and
From the linear distribution assumption, the bolt tension the bolt extension rotation can be calculated as:
strain in tension zone εi can be calculated from X
M =2 (Ni · yi ) (23)
ε1
εi = · yi . (20) 1l 1l1 1l2 1li
y1 φb = = = = (24)
ht 2y1 2y2 2yi
The tension force of this bolt can be calculated from Eqs. (6) P
and (7), and can also be calculated as where, is summation for all the bolts in the tension zone.
From the above calculation procedure (a)–(e), we can get
N1
Ni = · yi . (21) some pivotal points for the joint moment and the corresponding
y1 bolt extension rotation. Linking these pivotal points by the
(b) The contact force between the end-plate and the column sequence of the magnitudes of the corresponding bolt extension
flange at the second bolt-row is reduced to zero, i.e., at the rotation, the moment–bolt extension rotation curve can be
moment the end-plate is separated from the column flange by obtained, extending the straight line segment by its own slope
applying moment at the second bolt-row. beyond the point (e).
Here the actual increment on the bolt tension force, the bolt (iii) Joint rotation contributed by the bending deformation of
tension strain, the total tension force and the extension of the the end-plate and column flange
bolt in the second bolt-row can be calculated by Eqs. (9)–(12) The joint rotations contributed by the bending deformation
respectively. of the end-plate and column flange are called the end-plate
Similarly, bolt tension strain in tension zone εi can be rotation and the column flange rotation respectively. The
calculated as bending deformation of the end-plate and column flange can be
ε2 calculated according to the bolt tension force obtained by the
εi = · yi . (22)
y2 above calculation procedure (a)–(e), taking into account only
(c) The first bolt-row yields. the deformation of the end-plate and column flange around one
Here the bolt tension strain and the total tension force of the bolt-row on both sides of the beam tension flange.
bolt in the first bolt-row can be calculated by Eqs. (14) and (15) Since the calculation procedure of the column flange
respectively. The bolt tension strain in tension zone εi can be deformation is the same for the end-plate, the end-plate rotation
calculated by Eq. (20). is given to introduce the calculation method.
(d) The second bolt-row yields. For the end-plate connection meeting the above standard
Here the bolt tension strain and the total tension force of the details requirements, the beam flange, the beam web and the
bolt in the second bolt-row can be calculated by Eqs. (14) and end-plate extension rib stiffener can be considered as the fixed
(15) respectively. The bolt tension strain in tension zone εi can restraints of the end-plate segment. The end-plate segment
be calculated by Eq. (22). around the bolt in tension can be separated and simplified into
(e) The bolt of the first bolt-row reaching ultimate tension a two-edge fixed plate [41] as shown in Fig. 11. The arrow is
strength. the bolt clamp force. As the pre-tension bolt clamps the plates
Here the bolt tension strain and the total tension force of the effectively, it can be assumed that the end-plate segment is fixed
bolt in the first bolt-row can be calculated by Eqs. (18) and (19) at the bolt centre line. Furthermore, this end-plate segment can
respectively. The bolt tension strain in tension zone εi can be be decomposed into two plates with opposite ends fixed. The
calculated by Eq. (20). stiffness of this end-plate segment k I or kII is equal to the
1288 Y. Shi et al. / Journal of Constructional Steel Research 63 (2007) 1279–1293
Fig. 12. Comparison of the M–φs curve from the analytical models and tests.
Fig. 13. Comparison of the M–φep curve from analytical model and tests.
1290 Y. Shi et al. / Journal of Constructional Steel Research 63 (2007) 1279–1293
mechanics, the k1 and k2 can be expressed as to the bending resistance of the end-plate segment is larger
than the bolt’s ultimate tension resistance, this indicates that
1 1
k1 = β1 = β1 (26) the end-plate is rather thick and it does not yield. Here, the
3
ew α s ew 3
ew α s ew
12EI
+ GA 3 + slope of the last straight line segment, i.e. the straight line
Eb1 tep Gb1 tep
segment after point (e), is taken to be equal to the slope of its
1 1 preceding straight line segment. This circumstance should be
k 2 = β2 = β2 (27)
e3f αs e f e3f αs e f avoided during the design. According to the relevant tests and
12EI
+ GA 3 +
Eb2 tep Gb2 tep investigation results [3,32,43], the failure of the end-plate in
where αs = 1.5, is a parameter taking into account the shear bending can provide the joint with higher deformation capacity,
deformation; β1 = 1 − A2 /(b1 ew ) and β2 = 1 − A1 /(b2 e f ) are while the failure of bolts in tension is much less dissipative and
the reduction coefficients, because the stiffness of the areas A1 less deformable. The seismic design rule of “strong connection,
and A2 in Fig. 11 have been calculated repeatedly. weak plate” should be adopted. It also suggested applying the
After yielding occurs at the section edge of the end- end-plate with moderate thickness and adequately stiffening.
plate segment and it reaches its bending resistance, according The bolt diameter should be adequate so that the bolt ultimate
to references [3,32], the end-plate segment stiffness is tension resistance is higher than the end-plate segment ultimate
approximate to 1/7 of its initial stiffness, i.e., k I /7 or kII /7. bending resistance, in order to assure the joint ductility and
After the total cross-section of the end-plate segment yields, energy dissipation capacity.
it reaches its ultimate bending resistance. Taking into account For the end-plate connection with a strong bolt and weak
the strain hardening, the end-plate segment stiffness is taken as end-plate, it may happen that the bending resistance of the end-
Eh Eh plate segment at the first bolt-row is below the bolt tension
E k I or E kII .
From Ref. [41], the bolt tension force corresponding to the force of the first bolt-row when the end-plate and column
bending resistance of a two-edge fixed end-plate segment is flange separate at this position, i.e. the end-plate segment yields
given by prior to the separation of the end-plate and column flange.
Here a pivotal point should be added to the moment–end-plate
2 f
bep tep y (e f + ew )tep
2 f
y rotation curve, which is the bolt tension force of the first bolt-
Ny = + . (28) row corresponding to the yielding resistance of the end-plate
6e f 3ew
segment at this position. The relevant bolt transferred tension
When this segment reaches its ultimate bending resistance, the force, the bolt extension, the bending deformation of the end-
bolt tension force is taken as plate and the corresponding end-plate rotation can then be
Nu = 1.5N y . (29) calculated.
(iii) Moment–gap rotation (M–φep ) curve
Based on the bolt tension force calculated from the above The moment–gap rotation (M–φep ) curve can be obtained
calculation procedure (a)–(e), the end-plate deformation at the by superposing the moment–bolt extension rotation curve, the
first bolt-row ∆ep1 and second bolt-row ∆ep2 can be calculated. moment–end plate rotation curve and the moment–column
The deformation at the centre line of the beam tension flange flange rotation curve.
∆ep , is taken as the average of ∆ep1 and ∆ep2 , i.e., ∆ep = The calculation method for the moment–column flange
(∆ep1 + ∆ep2 )/2. The corresponding end-plate rotation is given rotation curve is the same as that for the moment–end plate
by rotation curve. If the column flange is wider than the end-plate,
the values of b1 and b2 for the column flange should be taken
∆ep
φep = . (30) as the values of the end-plate.
ht
From the above calculation procedure (a)–(e), the joint 5.3. Moment–rotation (M–φ) curve
moment and the corresponding end-plate rotation can be
calculated. The moment–end-plate rotation curve can be The moment–rotation (M–φ) curve can be obtained by
obtained. After point (e), the line slope is taken as EEh K i , where superposing the moment-shearing rotation (M–φs ) curve and
K i is the initial stiffness of this curve, i.e. the slope of the first the moment–gap rotation (M–φep ) curve. During the process
straight line segment of this curve. of superposing, for the moment value of each pivotal point of
For the actual behaviour of all the end-plate connection the M–φs curve and M–φep curve, adding the corresponding
moment–rotation (M–φ) curves, its tangent stiffness is shearing rotation φs and gap rotation φep , the joint rotation φ
always decreasing. Accordingly, it is required to revise the is equal to φs + φep . Linking these pairs of M and φ, the M–φ
moment–end plate rotation curve obtained above. Whenever the curve of the end-plate connection can be obtained.
slope of any straight line segment is larger than its preceding
straight line segment, its slope is taken equal to its preceding 6. Comparison of analytical and test results and discussion
straight line segment, and the moments of the pivotal points do
not change. The M–φ curves for test specimens EPC-1–EPC-5 have
For the calculation procedure (e), if the edge of the end-plate been calculated using the above analytical model, and the
segment does not yield, i.e. the bolt tension force corresponding comparison between the M–φs curves, the M–φep curves and
Y. Shi et al. / Journal of Constructional Steel Research 63 (2007) 1279–1293 1291
Fig. 14. Comparison of the M–φ curves from the analytical model and tests.
the M–φ curves of the analytical models and test results are analytical model is taken as the slope of the first straight line
shown in Figs. 12–14. The comparison between the joint initial segment of its M–φ curve. In Figs. 12–14 and Table 5, Model
rotational stiffness, S j,ini , of the analytical models and test 1 denotes that the analytical results are calculated according to
results are listed in Table 5. The value of S j,ini from the the material property values specified in the current Chinese
1292 Y. Shi et al. / Journal of Constructional Steel Research 63 (2007) 1279–1293
Table 5
Comparison of S j,ini of specimens from analytical and test results
steel design code; the yield strength of Q345 steel is f y = final results of the M–φ curve are in accordance with the tests
345 MPa; the ultimate strength is f u = 470 MPa; the yield generally.
strength of grade 10.9 bolt is f by = 940 MPa and its ultimate
tension strength is f bu = 1040 MPa; the elastic modulus of 7. Conclusions
steel is E = 2.06 × 105 N/mm2 and its shear modulus G = (1) In this paper, the stiffened and extended beam–column
79 × 103 MPa. Model 2 denotes that the analytical results are end-plate connection has been recommended for end-plate
calculated according to the material actual properties obtained moment connections in multistory steel frames, and its
from coupon tests which are listed in Table 2. The steel’s standards details have been proposed. Five full-scale joint tests
hardening modulus is assumed to be E h = 0.04E; the bolt’s of this type of end-plate connection have been conducted to
hardening modulus is assumed to be E bh = 0.1E b ; Poisson’s investigate the influences of bolt size and end-plate thickness on
ratio is taken as µ = 0.3; and the elastic modulus of bolt is the joint behaviour. The rotational stiffness, moment resistance
taken as E b = 2.06 × 105 N/mm2 . and moment–rotation (M–φ) curves are obtained. A clear
The other connection parameters are taken as h bw = definition for the end-plate connection rotation has been
276 mm, h cw = 276 mm, tcw = 8 mm, Ae = 244.8 mm2 proposed. With a special method to measure the bolt strain, the
(M20) or 352.5 mm2 (M24); lb = 48 mm (EPC-1), or 58 mm distribution and the development of bolt tension force have been
(EPC-2), or 50 mm (EPC-3), or 60 mm (EPC-4), or 40 mm obtained during the tests.
(EPC-5); α = 10, d f = 288 mm, y1 = 200 mm, y2 = 88 mm, (2) A new analytical model to evaluate the moment–rotation
ew = 49 mm, e f = 50 mm, b1 = 100 mm, b2 = 95 mm, (M–φ) relationship of this type of end-plate connection has
β = 0.856. been proposed. The end-plate connection is decomposed into
From the comparison of the results in Figs. 12–14 and several components, including the panel zone, bolt, end-
Table 5, it can be concluded that: plate and column flange. The complete loading process of
each component is analysed. The moment–rotation curve of
(1) The M–φ curves, the M–φs curves and the joint
the whole connection is obtained by superimposing each
initial rotational stiffnesses obtained from the analytical model
component. Comparing with the test results, it has been verified
coincide well with the test results if the actual material
that this analytical model can sufficiently predict the rotational
properties are applied. The comparison verifies the accuracy
behaviour of end-plate connections, such as the initial rotational
of this analytical model. In practical structural design, where
stiffness and the moment–rotation (M–φ) curve. Furthermore,
the nominal material property specified in current code is
the contributions to the joint rotational deformation of each
applied, the analytical models also give satisfactory solutions component, such as shear deformation of the panel zone,
with adequate accuracy. the bolt extension, bending deformation of the end-plate and
(2) The M–φep curves obtained from the analytical results column flange etc. are provided. This analytical model can also
coincide well with the tests results at the initial loading stage. provide the moment–shear rotation (M–φs ) and moment–gap
After nonlinearity occurs in the loading curves, there are some rotation (M–φep ) curves, thus provide a reliable foundation
discrepancies between the analytical results and the test results. for analysing the detailed rotational behaviour of end-plate
The possible reasons behind these discrepancies may connections.
be explained as follows: The end-plate and column flange
deformation is evaluated under the assumption that the end- Acknowledgements
plate segment is fixed at the bolt centre line. The end- The writers gratefully acknowledge the support for this
plate and column flange contact closely due to the bolt pre- work, which was funded by the Tsinghua Basic Research
tension force at the initial loading stage, and this assumption Foundation (Grant No. JCqn2005006) and the Natural Science
is reasonable. With the joint moment increasing, the contact Foundation of China (No. 50578083).
force between the end-plate and column flange is significantly
reduced and even if they separate, this assumption will lead to References
some discrepancies, but the discrepancies do not considerably [1] Chen WF, Lui FM. Stability design of steel frames. Boca Raton (FL):
influence the evaluation of joint total rotation (Fig. 14). The CRC Press; 1991.
Y. Shi et al. / Journal of Constructional Steel Research 63 (2007) 1279–1293 1293
[2] Attiogbe E, Morris G. Moment–rotation functions for steel connections. [24] Tarpy TS, Cardinal JW. Behavior of semi-rigid beam-to-column end plate
Journal of Structural Engineering 1991;117(6):1703–18. connections. In: Howlett JH, Jenkins WM, Stainsby R, editors. Joints in
[3] CEN: prEN1993-1-8. Eurocode 3: Design of steel structures: Part 1.8 structural steelworks. London: Pentech Press; 1981. p. 2.3–2.25.
Design of joints. 2002. [25] Maxwell SM, Jeckins WM, Howlett JH. Theoretical approach to the
[4] American Institute of Steel Construction (AISC). Allowable stress design analysis of connection behavior. In: Howlett JH, Jenkins WM, Stainsby R,
specification for structural steel buildings. Chicago (IL); 1989. editors. Joints in structural steelworks. London: Pentech Press; 1981.
[5] American Institute of Steel Construction (AISC). Load and resistance p. 2.49–2.70.
factor design specification for structural steel buildings. Chicago (IL); [26] Sherbourne AN, Bahaari MR. 3D simulation of end-plate bolted
1999. connections. Journal of Structural Engineering 1994;120(11):3122–36.
[6] GB50017-2003. Code for design of steel structures. 2003 [in Chinese]. [27] Bahaari MR, Sherbourne AN. 3D simulation of bolted connections
[7] Chen WF. Practical analysis for semi-rigid frame design. Singapore: to unstiffened columns-II. Extended endplate connections. Journal of
World Scientific; 2000. Constructional Steel Research 1996;40(3):189–223.
[8] Shi YJ, Chan SL, Wong YL. Modeling for moment–rotation character- [28] Sherbourne AN, Bahaari MR. Finite element prediction of end plate
istics for end-plate connections. Journal of Structural Engineering 1996; bolted connection behavior I: Parametric study. Journal of Structural
122(11):1300–6. Engineering 1997;123(2):157–64.
[9] Shi G, Shi YJ, Wang YQ, Chen H. Numerical simulation and experimental [29] Bahaari MR, Sherbourne AN. Finite element prediction of end plate
study on bolted end-plate connections. In: The eighth international bolted connection behavior II: Analytic formulation. Journal of Structural
symposium on structural engineering for young experts, vol. 1. 2004. p. Engineering 1997;123(2):165–75.
137–42. [30] Shi G, Shi YJ, Wang YQ, Li S, Chen H. Finite element analysis and
[10] Kukreti AR, Murray TM, Abolmaali A. End-plate connection tests on bolted end-plate connections in steel portal frames. Advances in
moment–rotation relationship. Journal of Constructional Steel Re-
Structural Engineering 2004;7(3):245–56.
search 1987;8:137–57.
[31] Shi G, Shi YJ, Chen H. Finite element analysis of beam–column bolted
[11] Douty RT, McGuire W. High strength bolted moment connections.
end-plate connections in steel frames. In: Seventh pacific structural steel
Journal of the Structural Division 1965;91(ST2):101–28.
conference. 2004.
[12] Nair RS, Birkemoe PC, Munse WH. High strength bolts subject to tension
[32] Faella C, Piluso V, Rizzano G. Structural steel semirigid connections.
and prying. Journal of the Structural Division 1974;100(ST2):351–72.
Boca Raton (FL): CRC Press LLC; 2000.
[13] Kato B, McGuire W. Analysis of T-stub flange-to-column connections.
[33] Ghobarah A, Osman A, Korol RM. Behaviour of extended end-plate
Journal of the Structural Division 1973;99(ST5):865–88.
connections under cyclic loading. Engineering Structures 1990;12(1):
[14] Agerskov H. High-strength bolted connections subject to prying. Journal
15–27.
of the Structural Division 1976;102(ST1):161–75.
[34] Korol RM, Ghobarah A, Osman A. Extended end-plate connections under
[15] Zoetemeijer P. Semi-rigid bolted beam-to-column connections with
cyclic loading: Behaviour and design. Journal of Constructional Steel
stiffened column flanges and flush-end plates. In: Howlett JH,
Jenkins WM, Stainsby R, editors. Joints in structural steelworks. London: Research 1990;16:253–80.
Pentech Press; 1981. p. 2.99–2.118. [35] Ghobarah A, Korol RM, Osman A. Cyclic behaviour of extended end-
[16] Packer JA, Morris L. A limit state design method for the tension region of plate joints. Journal of Structural Engineering 1992;118(5):1333–53.
bolted beam–column connections. The Structural Engineer 1977;55(10): [36] Tsai KC, Popov EP. Cyclic behavior of end-plate moment connections.
446–58. Journal of Structural Engineering 1990;116(11):2917–30.
[17] Mann AP, Morris LJ. Limit design of extended end-plate connections. [37] Adey BT, Grondin GY, Cheng JJR. Cyclic loading of end plate moment
Journal of the Structural Division 1979;105(ST3):511–26. connections. Canadian Journal of Civil Engineering 2000;27(4):683–701.
[18] Witteveen J, Stark JWB, Bijlaard FSK, Zoetemeijer P. Welded and bolted [38] JGJ82-91. Specification for design, construction and acceptance of high
beam-to-column connections. Journal of the Structural Division 1982; strength bolt connections in steel structures. 1991 [in Chinese].
108(ST2):433–55. [39] CECS102:2002. Technical specification for steel structure of light-weight
[19] Stark JWB, Bijlaard FSK. Design rules for beam-to-column connections buildings with gabled frames. 2003 [in Chinese].
in Europe. Journal of Constructional Steel Research 1988;10:415–62. [40] Krawinkler H, Bertero VV, Popov EP. Shear behavior of steel frame
[20] Krishnamurthy N. Correlation between 2- and 3-dimensional finite joints. Journal of the Structural Division 1975;101(ST11):2317–36.
element analyses of steel bolted end-plate connections. Computers & [41] Li S, Shi G, Shi YJ, Wang YQ, Chen H. Research on end-plate bearing
Structures 1976;6:381–9. capacities under action of single bolt in the bolted end-plate connection of
[21] Krishnamurty N, Huang HT, Jefferey PK, Avery LK. Analytical M-θ steel structures. In: Second international symposium on steel structures.
curves for end-plate connections. Journal of the Structural Division 1979; 2002. p. 433–9.
105(ST1):133–45. [42] Krishnamurthy N. A fresh look at bolted end-plate behavior and design.
[22] Krishnamurthy N. Modeling and prediction of steel bolted connection Engineering Journal 1978;15(2):39–49.
behavior. Computers & Structures 1980;11:75–82. [43] Shi G, Shi YJ, Li S, Wang YQ. Cyclic tests of semirigid end-plate
[23] American Institute of Steel Construction. Manual of steel construction: connections in multi-story steel frames. Journal of Building Structures
Load and resistance factor design. 3rd ed. Chicago (IL); 2001. 2005;26(2):74–80 [in Chinese].