Engineering Structures 64 (2014) 2338

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Behaviour of beam-to-column high strength steel endplate connections

under re conditions Part 1: Experimental study
Xuhong Qiang a,b, Frans S.K. Bijlaard b, Henk Kolstein b, Xu Jiang c,
a

College of Civil Engineering, Shenzhen Graduate School, Harbin Institute of Technology, 518055 Shenzhen, China
Structural and Building Engineering, Department of Design and Construction, Faculty of Civil Engineering, Delft University of Technology, P.O. Box 5048, 2600 GA Delft,
The Netherlands
c
College of Civil Engineering, Tongji University, 200092 Shanghai, China
b

a r t i c l e

i n f o

Article history:
Received 26 April 2013
Revised 15 December 2013
Accepted 20 January 2014
Available online 19 February 2014
Keywords:
High strength steel
Endplate connection
Beam-to-column connections
Structural re performance
Experimental study

a b s t r a c t
In order to reveal more information and understanding on behaviour and failure mechanisms of high
strength steel endplate connections under re conditions, an experimental study has been carried out
and presented in this paper. Full-scale tests on beam-to-column high strength steel endplate connections
were conducted at elevated temperature 550 C under steady state re condition and at ambient temperature as reference. Further, their behaviour was compared with that of mild steel endplate connections.
Moreover, the provisions of Eurocode 3 were validated with test results of high strength steel endplate
connections. It is found that a proper thinner high strength steel endplate can enhance the connections
rotation capacity both at ambient temperature and in re (which guarantees the safety of an entire structure), and simultaneously achieve almost the same moment resistance with a mild steel endplate
connection.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
Fires in buildings often have enormous consequences on safety
and economy. Structural re safety is therefore a key consideration
in the design of buildings and is attracting worldwide attention.
Beam-to-column connections are important components of steel
framed structures, as they are supposed to resist resultant forces
at the end of the beam and transfer them into the columns and surrounding structural components. The fracture of a connection can
cause the collapse of the connected beam, which may lead to a progressive collapse of the entire building structure. Therefore the
behaviour of connections in a building is of extreme signicance,
not only at ambient temperature but also under re conditions.
In Europe, endplate connections are typical beam-to-column
connections for steel structures produced by welding at workshops
and erected by bolting in situ. The simplicity and economy associated with its fabrication make this type of connection popular in
steel structures. Rules for prediction of strength, stiffness and
deformation capacity of endplate connections at ambient temperature have been included in current leading design standards, such
as Eurocode 3 Part: 18 [1]. According to Eurocode 3, for structural
steels up to S460, plastic design of connections may be used.
Corresponding author. Tel.: +31 15 2784752; fax: +31 15 2783173.
E-mail addresses: X.Qiang@tudelft.nl (X. Qiang), F.S.K.Bijlaard@tudelft.nl
(F.S.K. Bijlaard), M.H.Kolstein@tudelft.nl (H. Kolstein), Xu.Jiang@tudelft.nl (X. Jiang).
http://dx.doi.org/10.1016/j.engstruct.2014.01.028
0141-0296/ 2014 Elsevier Ltd. All rights reserved.

However, for steel grades higher than S460 up to S700 only elastic
design of connections can be employed [2], which is very uneconomical for steel structures. This is due to the lack of experimental
and theoretical evidence that these high strength steel (HSS)
connections have sufcient deformation capacities. Coelho and
Bijlaard have found that the high strength steel S690 endplate
connections satisfy the design provisions for resistance and
achieve reasonable rotation demands at ambient temperature [3].
However, research results on performance of high strength steel
endplate connections under re conditions are not available in
the literature.
In this paper and its companion paper [4], a perspective of combining high strength steel endplate with mild steel beam and column in beam-to-column endplate connections is proposed and
investigated. The aim of this research is to reveal more information
and understanding on behaviour of high strength steel ush endplate connections under re conditions, for an effective application
of high strength structural steels in civil engineering as well as
enhancing the re safety of steel structures.
In the present paper, the behaviour of high strength steel endplate connections under re conditions is studied via full-scale
tests. Tests on seven beam-to-column endplate connections were
conducted at elevated temperature 550 C under steady state, to
observe their performances in re. As reference, the experimental
study on these seven endplate connections at ambient temperature
was carried out as well. In the connection specimens, three

24

X. Qiang et al. / Engineering Structures 64 (2014) 2338

endplate connections are made of mild steels for comparison with

high strength steel endplate connections. The parameters investigated herein are the endplate thickness and the endplate material.
All specimens are designed to conne failure to the connections
rather than the beam or column. The momentrotation characteristic, rotation capacity and failure mode of high strength steel endplate connections in re and at ambient temperature were
obtained through tests and compared with those of mild steel endplate connections. Further, the provisions of Eurocode 3 were
validated with experimental results.
2. Test programme
2.1. Test specimen
In this experimental study, the strengths of the endplate connections are designed according to Eurocode 3 Part: 18 Design of
joints [1]. In the endplate connections, the endplates are made of
high strength steels (S690 and S960) while the beam and column
are made of Q345 (mild structural steel, the nominal yield stress
of which is 345 MPa, similar to S355). The material properties of
various components at ambient and elevated temperatures are described in the companion paper [4] of this manuscript. The beam
sections used in this study are HW300  300 (comparable to European Section HE320A) while the column sections are HW400  400
(comparable to European Section HE300M); their detailed dimensions are presented in Fig. 1. For comparison, the connections with
endplates made of mild steels Q235 (mild structural steel, the nominal yield stress of which is 235 MPa, similar to S235) and Q345 are
also included herein. In this experimental study, there are two series of endplate connections, see Table 1. In each of them, the load
bearing capacities of the connections are designed to be similar,
while the endplate materials and thicknesses vary. In order to

compare the behaviour of endplate connections under re conditions with that at ambient temperature, the tests at ambient temperature on each concerned endplate connection were conducted
as well. The overall dimension of the endplate connection test specimen is shown in Fig. 1, while the endplate materials and thicknesses of the specimens are shown in Table 1. Two stiffeners
were welded on each side of the beam web at loading point for
applying load, as shown in Fig. 1(a). The welds between endplate
and beam are 45-llet welds, which were made in the workshop
in a down-hand position. The welding material used for specimens
with endplates made of mild steels Q235 and Q345 is ER50-6, while
that for specimens with endplates made of high strength steels
S690 and S960 is ER76-G. The chemical compositions of these
two welding electrodes are listed in Table 2, while their mechanical
properties are presented in Table 3. Since the yield strength of electrode ER50-6 is 439 MPa, the welds for specimens with endplates
made of mild steels Q235 and Q345 are overmatched. The yield
strength of electrode ER76-G is 735 MPa, similar to that of HSS
S690; hence the welds for specimens with endplate made of HSS
S690 are matched welding while those for specimens with endplate
made of HSS S960 are under matched. Therefore the weld sizes used
for endplates made of mild steels (Q235 and Q345) are 8 mm; while
those for endplates made of high strength steels (S690 and S960)
are 10 mm for compensation.
2.2. Test set-up
All re tests were conducted in a gas furnace (4.5 m  3.0 m 
1.7 m) as shown in Fig. 2. Since applying a tensile load under re
conditions is more stable than applying a compressive load, the
connection specimens were designed to be located upside down
in order to easily apply the tensile load from the outside of the
furnace, as shown in Fig. 3. The members below the furnace cover

(a) Endplate connection

(c) Beam section

(b) Endplate

(d) Column section

Fig. 1. Dimension of endplate connection specimen.

25

X. Qiang et al. / Engineering Structures 64 (2014) 2338

Table 1
Test specimens and re test conditions.
Connection ID

Endplate material

Endplate thickness (mm)

Electrode

Weld type

Weld size (mm)

1-1 A
1-2 A
1-3 A

Q235
S690
S960

20
12
10

ER50-6
ER76-G
ER76-G

Overmatched
Matched
Under matched

8
10
10

20
20
20

2-1
2-2
2-3
2-4

A
A
A
A

Q235
Q345
S690
S960

25
20
15
12

ER50-6
ER50-6
ER76-G
ER76-G

Overmatched
Overmatched
Matched
Under matched

8
8
10
10

20
20
20
20

1-1 E
1-2 E
1-3 E

Q235
S690
S960

20
12
10

ER50-6
ER76-G
ER76-G

Overmatched
Matched
Under matched

8
10
10

550
550
550

2-1
2-2
2-3
2-4

Q235
Q345
S690
S960

25
20
15
12

ER50-6
ER50-6
ER76-G
ER76-G

Overmatched
Overmatched
Matched
Under matched

8
8
10
10

550
550
550
550

E
E
E
E

Temperature of specimens (C)

Table 2
Chemical compositions of the electrodes (%).
Electrode

Mn

Si

Cr

Ni

Mo

Cu

ER50-6
ER76-G

0.08
0.08

0.015
0.009

1.45
1.66

0.018
0.01

0.96
0.56

0.003
0.09

0.02
1.32

0.002
0.46

0.004

0.10

Table 3
Mechanical properties of the electrodes.
Electrode

ER50-6
ER76-G

Yield strength (MPa)

439
735

Tensile strength (MPa)

510
790

Elongation (%)

26.40
21.50

Impact test
Temperature (C)

Impact value

30
40

88
123

106
123

98
132

column, the rotation of column can be calculated. In order to measure the displacement of endplate, one vertical displacement sensor
(DT4) and 4 horizontal displacement sensors (DT6DT9) were
placed, as shown in Fig. 4. According to the displacement of endplate, the rotation of endplate can be calculated. Based on the displacements of the aforementioned components, the rotation of
endplate connection in tests can be obtained. Another way of
obtaining the rotation of connection was also designed via using
four extra horizontal displacement sensors (DT11DT14) together
with the four horizontal displacement sensors DT6DT9. The two
methods of measuring rotation of endplate connection were used
to validate each other, which will be presented hereafter.
2.4. Test procedure

Fig. 2. Fire test furnace.

are in re during re tests, while those above the furnace cover are
out of the re eld at ambient temperature, see Fig. 3.
2.3. Displacement measurements
In the re tests, 3 vertical displacement sensors (DT1DT3) were
used to obtain the vertical displacement of the beam, as shown in
Fig. 4. According to the vertical displacements of beam, the rotation
of beam can be calculated. In order to record the displacement of
column, 2 vertical displacement sensors (DT5 and DT10) were
arranged as shown in Fig. 4. According to the displacement of

Before performing the full-scale tests, all equipment was calibrated by skilled technicians. After the installation of each test
specimen and measuring equipment, all test equipment was
checked by trial loading at ambient temperature. After everything
went well, the real tests were started. The full-scale re tests on
endplate connections were conducted under steady state re condition. The specimens were rstly heated to a pre-selected elevated temperature (550 C) at a constant heating rate 10 C/min,
which corresponds to a natural re to buildings. When the temperature of the concerned components reached the pre-selected elevated temperature, the mechanical load (moment for the
connection, see Fig. 5) was applied to the specimen at this constant
elevated temperature (550 C) until failure occurred. During loading, displacement control was used by controlling the displacement of piston of the hydraulic actuator at a constant rate

26

X. Qiang et al. / Engineering Structures 64 (2014) 2338

Jack

Load cel l

Fig. 3. Fire test set-up.

Fig. 5. Force application.

Fig. 4. Arrangement of displacement sensors.

10 mm/min. The displacements of beam, column and endplate and

the rotation of connection under re conditions were obtained and
recorded. After tests, the nal deformation state of connections can
be observed.
3. Experimental results
3.1. Deformation at the end of tests
For all endplate connections, both at ambient temperature and
in re, their nal deformation states were photographed at the end
of tests.
3.1.1. At ambient temperature
An overall description on components of all connections at the
end of the ambient-temperature tests is listed in Table 4. As an
example, the detailed illustrations on the nal deformation state

of specimen 1-1 A (Q235 20 mm) at ambient temperature are presented in Fig. 6. It can be seen that the failure was concentrated in
the endplate, and the bolts remained almost straight until very
large deformation on the endplate appeared. Damage to the endplate started as signicant bending; after very large plastic bending
deformation of the endplate, the two rows of bolts in tension tore
some bearing materials around the bolt holes. It corresponds to a
rapid drop in resistance of the connection. After very large bending
deformation of the endplate, some threads stripped off the bolts in
the top tensile row. The broken bolt in the second tensile row as
shown in Fig. 6(d) and (e) was due to the cut during removing
the specimen after tests. The detailed nal deformation states of
the other test specimens at ambient temperature can be found in
the literature [5].
3.1.2. At elevated temperatures
An overall description of the components of all connections at
the end of elevated-temperature tests is listed in Table 5. As an
example, the observed nal deformation state of connection 1-2
E (S690 12 mm) at elevated temperature 550 C is presented in

27

X. Qiang et al. / Engineering Structures 64 (2014) 2338

Table 4
Description of components at the end of ambient-temperature tests.
Test ID

Endplate
material

Endplate thickness
(mm)

Endplate
yielding

Fracture of bolts
in top tensile row

Nuts in top tensile

row stripped off

Weld failure in heat

affected zone

Bolts in compression
almost straight

1-1 A
1-2 A
1-3 A

Q235
S690
S960

20
12
10

Yes
Yes
Yes

No
No
No

No
Yes
No

No
No
Yes

Yes
Yes
Yes

2-1
2-2
2-3
2-4

Q235
Q345
S690
S960

25
20
15
12

Yes
Yes
Yes
Yes

Yes
Yes
Yes
Yes

No
No
No
No

No
No
No
No

Yes
Yes
Yes
Yes

A
A
A
A

Fig. 6. Final deformation state of connection 1-1 A (Q235 20 mm) at ambient temperature.

Table 5
Description of components at the end of elevated-temperature tests.
Test ID

Endplate
material

Endplate thickness
(mm)

Endplate
yielding

Fracture of bolts
in top tensile row

Nuts in top tensile

row stripped off

Weld failure in
heat affected zone

Bolts in compression
almost straight

1-1 E
1-2 E
1-3 E

Q235
S690
S960

20
12
10

Yes
Yes
Yes

Yes
Yes
Yes

No
No
No

No
No
No

Yes
Yes
Yes

2-1
2-2
2-3
2-4

Q235
Q345
S690
S960

25
20
15
12

Yes
Yes
Yes
Yes

Yes
Yes
Yes
Yes

No
No
No
No

No
No
No
No

Yes
Yes
Yes
Yes

E
E
E
E

Fig. 7. It can be seen that the failure started from the bending of the
endplate, and the bolts remained almost straight until very large
plastic deformation on the endplate appeared. After very large
bending deformation of endplate, the two rows of bolts in tension
were gradually pulled apart, with no obvious breaking point, as

shown in Fig. 7(d). It corresponds to the decline phase in resistance

of this connection. For the bolts in compression, they remained
straight in tests. The picture of bolts in compression after removing
from the specimens after tests was not taken, because they were
ruined during removing.

28

X. Qiang et al. / Engineering Structures 64 (2014) 2338

Fig. 7. Final deformation state of connection 1-2 E (S690 12 mm) at elevated temperature 550 C.

Fig. 9. Endplate connection rotation deformation.

u hb  hc

This deformation is only due to the force transferred by the

beam anges that are induced by the bending moment M
(M = F  L) acting on the beam.
The beam rotation hb can be calculated by the following
equations:
Fig. 8. Denition of joint and connection.

3.2. Momentrotation relationship of endplate connection

The behaviour of a steel beam-to-column joint is usually represented by Mu curve. The rotational deformation of joint results
from the in-plane bending moment M, and it includes the rotational
deformation of the connection as well as the shear deformation of
the column web panel zone. (The denition of joint and connection
is shown in Fig. 8.) The connection rotational deformation consists
of the deformation of the fastening elements (bolts, endplate, etc.)
and the deformation of the column ange. The rotational deformation of a connection u is dened as the relative rotation between
the beam and column axes, see Fig. 9, which is equal to:

jdDT1 j  jdDT4 j
jdDT2 j  jdDT4 j
arctan
900
600
jdDT3 j  jdDT4 j
jdDT6 j jdDT9 j
arctan
arctan
300
hb  t fb

hb arctan

where dDT13 are the vertical displacements of beam measured by

LVDT DTi, dDT4 is the vertical displacements of endplate, while
dDT6 and dDT9 are the horizontal displacements of beam anges,
see Fig. 4, hb is the height of beam section while tfb is the thickness
of beam ange. In Eq. (2), the deformation of the beam is
contributed not only by the deformation itself, but also by the
deformation of the column, the deformation of the panel zone and
the deformation of connection components. In this experimental
study, the column hardly deforms as it behaves almost like a rigid
element, which will be validated afterwards.

29

X. Qiang et al. / Engineering Structures 64 (2014) 2338

In tests, the endplate vertical displacement obtained from LVDT

DT4 was found very small and not in a range comparable with
those from other LVDTs. For illustration, the relationship of applied
mechanical load and vertical displacement readings of LVDTs
DT1-4 for connection specimen 1-3 A (S960 10 mm) at ambient
temperature is shown in Fig. 10 while the relationship of moment
and beam rotation computed according to Eq. (2) is presented in
Fig. 11. It can be seen that good agreements exist except the results
from DT3 and DT4, since the displacement sensor DT3 is located
closer to the endplate. In this region, the beam theory is not valid
and the stress distribution is not smooth. Therefore, the results
from displacement sensors DT1 and DT4 are used for further
analysis.
The column rotation hc can be calculated by the following
equation:

hc arctan

jdDT5 j jdDT10 j
jdDT11 j jdDT13 j
arctan
hc  t fc
hb  tfb

where hc is the height of column section and tfc is the thickness of

column ange, while hb is the height of beam section and tfb is
the thickness of beam ange; dDT5 anddDT10 are the vertical displacements of column while dDT11 and dDT13 are horizontal ones, see
Fig. 4. The relationship of moment and column rotation computed

according to Eq. (3) for connection specimen 1-3 A (S960 10 mm)

at ambient temperature is illustrated in Fig. 12, where good agreement can be observed.
What is more, the maximum rotation capacity /C of a connection, which is dened as the maximum rotation of the moment
rotation characteristic, is illustrated in Fig. 13, while /Mmax is the
rotation of connection corresponding to the maximum load level.
The rotations of various endplate connections obtained from this
experimental study are described hereafter in this section.
3.2.1. Effect of endplate material and thickness vary at ambient
temperature
The load bearing capacity and connection rotation of various
connections at ambient temperature obtained from experiments
are presented in Table 6. The momentrotation comparison of connections in Series 1 at ambient temperature is shown in Fig. 14. It
can be seen that the peak load of connection 1-2 A (S690 12 mm) is
similar to that of connection 1-1 A (Q235 20 mm) while the rotational capacity of the former is higher than the latter. In tests,
the connection 1-3 A (S960 10 mm) at ambient temperature failed
due to cracking of the endplate at the heat affected zone of welds
(which is presented in Fig. 22 of the companion paper [4]) after
moderate amount of plastic deformation of the endplate. Hence

250
200

200

moment (kNm)

force (kN)

150

100

D
DT1
D
DT2
D
DT3
D
DT4

50

DT5 and DT10

DT11 and DT13
150

100

50

20

40

60

8
80

100
0

120

20

40

dis
splac
ceme
ent (mm)
m
Fig. 10. Vertical displacement readings of LVDTs DT14 for specimen 1-3 A (S960
10 mm) at ambient temperature.

60

80

100

120

rotation (mrad)
Fig. 12. Column rotation computed from the displacement readings for specimen
1-3 A (S960 10 mm) at ambient temperature.

250

Mmax

250
200

moment (kN.m)

moment (kNm)

200
150

100

DT1 and DT4

DT2 and DT4
DT3 and DT4
DT9 and DT6

50

150

100

50

20

40

60

80

100

120

rotation (mrad)

max

0
0

50

100

150

200

rotation (mrad)
Fig. 11. Beam rotation computed from the displacement readings for specimen 1-3
A (S960 10 mm).

Fig. 13. Denition of connection rotation capacity.

250

30

X. Qiang et al. / Engineering Structures 64 (2014) 2338

Table 6
Characteristics of connections at ambient temperature.

300

Connection
ID

Endplate
Material

Thickness
(mm)

Moment
(kN m)

Force
(kN)

1-1 A
1-2 A
1-3 A

Q235
S690
S960

20
12
10

229.93
247.85
209.86

213.69
230.34
195.04

199
214
105

2-1
2-2
2-3
2-4

Q235
Q345
S690
S960

25
20
15
12

287.00
277.25
272.71
264.32

266.73
257.67
253.45
245.65

114
115
101
105

A
A
A
A

Peak load

Connection
rotation
/C (mrad)

250

moment (kNm)

Series

200

150

Q235 25mm
Q345 20mm
S690 15mm
S960 12mm

100

50

250

50

100

150

200

250

rotation (mrad)

moment (kNm)

200

Fig. 15. Momentrotation comparison of connections in Series 2 at ambient

temperature.

150

Table 7
Characteristics of connections at elevated temperature 550 C.
100

Series

Q235 20mm
S690 12mm
S960 10mm

50

Connection
ID

Endplate
Material

Thickness
(mm)

Moment
(kN m)

1-1 E
1-2 E
1-3 E

Q235
S690
S960

20
12
10

83.91
105.92
100.05

77.98
98.44
92.98

314
304
313

2-1
2-2
2-3
2-4

Q235
Q345
S690
S960

25
20
15
12

120.40
111.16
120.78
113.18

111.90
103.31
112.25
105.19

191
313
330
320

0
0

50

100

150

200

250

rotation (mrad)
Fig. 14. Momentrotation comparison of connections in Series 1 at ambient
temperature.

E
E
E
E

Peak load
Force
(kN)

Connection
rotation
/C (mrad)

120

its rotation capacity is relatively low and not included in the comparison. The momentrotation comparison of connections in Series
2 at ambient temperature is presented in Fig. 15.

100

moment (kNm)

3.2.2. Effect of endplate material and thickness vary at elevated

temperature
The test obtained load bearing capacity and connection rotation
of various connections at elevated temperature 550 C are illustrated in Table 7. The momentrotation comparison of connections
Series 1 at elevated temperature 550 C is presented in Fig. 16. It
can be found that the load bearing capacities of connection 1-2 E
(S690 12 mm) and connection 1-3 E (S960 10 mm) are higher than
that of connection 1-1 E (Q235 20 mm), but still in the comparable
range. The rotational capacities at elevated temperature 550 C of
the three connections in Series 1 are similar to each other. The momentrotation relationships of connections in Series 2 at elevated
temperature 550 C are shown in Fig. 17. It can be seen that the
load bearing capacities at elevated temperature 550 C of the four
connections in Series 2 are similar to each other, while the rotational capacities of connection 2-2 E (Q345 20 mm), connection
2-3 E (S690 15 mm) and connection 2-4 E (S960 12 mm) are comparable and much higher than that of connection 2-1 E (Q235
25 mm).
A conclusion can be drawn that in endplate connections a proper thinner endplate made of high strength steel can achieve similar
load bearing capacity and higher (or at least comparable) rotational capacity both at ambient temperature and at elevated temperatures in comparison to a thicker endplate made of mild steel.

Q235 20mm
S690 12mm
S960 10mm

80

60

40

20

0
0

50

100

150

200

250

300

350

400

rotation (mrad)
Fig. 16. Momentrotation comparison of connections Series 1 at elevated temperature 550 C.

3.2.3. Effect of endplate thickness varies at ambient temperature

To state the inuence of endplate thickness, the momentrotation curves of connections with the same endplate material (Q235,
S690 and S960) are shown in Figs. 1820. It can be observed that
for endplate connections of which the endplates are made of the
same steel material, with a thicker endplate the connection can

31

X. Qiang et al. / Engineering Structures 64 (2014) 2338

140

300

Q235 25mm
Q345 20mm
S690 15mm
S960 12mm

120

250

moment (kNm)

moment (kNm)

100

80

60

200

150

100

40

S690 12mm
S690 15mm
50

20

0
0

50

100

150

200

250

300

350

400

50

rotation (mrad)
Fig. 17. Momentrotation comparison of connections Series 2 at elevated temperature 550 C.

150

200

250

Fig. 19. Effect of endplate thickness in connections with endplate made of S690 at
ambient temperature.

300

Q235 20mm
Q235 25mm

300

100

rotation (mrad)

250

moment (kNm)

moment (kNm)

250

200

150

200

150

100

100

50

50

S960 10mm
S960 12mm

50

100

150

200

250

rotation (mrad)
Fig. 18. Effect of endplate thickness in connections with endplate made of Q235 at
ambient temperature.

50

100

150

200

250

rotation (mrad)
Fig. 20. Effect of endplate thickness in connections with endplate made of S960 at
ambient temperature.

3.3. Rotation capacity of endplate connections

achieve higher load bearing capacity but its rotational capacity is
reduced obviously at ambient temperature. The connections made
of S960 as shown in Fig. 20 are the exception, because of the early
cracking of the endplate at the heat affected zone of welds after
some moderate plastic deformations on the endplate for connection S960 10 mm, see Fig. 22 of the companion paper [4].

The connection rotations for all endplate connections at ambient and elevated temperatures obtained from tests are presented
in Tables 8 and 9 respectively. The comparison and discussion of
experimental results with the current provisions of Eurocode 3
[1] will be presented afterwards in Section 4.3.
4. Discussion

3.2.4. Effect of endplate thickness varies at elevated temperature

The momentrotation curves at elevated temperature 550 C of
connections with the same endplate material are presented in
Figs. 2123 for Q235, S690 and S960 respectively, to indicate the
inuence of endplate thickness on endplate connections under re
conditions. It can be found that for endplate connections of which
the endplates are made of the same steel material, with a thicker
endplate the connection can achieve higher load bearing capacity
at elevated temperature but its rotational capacity at elevated temperature is reduced, which is in line with the conclusion at ambient
temperature.

4.1. Failure modes

According to Eurocode 3 Part: 18 [1], there are 3 failure modes
for endplate connections. Mode 1 is complete yielding of endplate
or column ange, Mode 2 is bolt failure with yielding of endplate or
column ange, while Mode 3 is bolt failure. Mode 3 is considered
to be brittle and should be avoided in practical design. The failure
modes of all endplate connections can be obtained via theoretical
analysis based on the rules of Eurocode 3 Part: 18 [1]. The detailed analysis procedure is presented in the literature [5].

32

X. Qiang et al. / Engineering Structures 64 (2014) 2338

120

120

Q235 20mm
Q235 25mm

100

moment (kNm)

moment (kNm)

100

S960 10mm
S960 12mm

80

60

80

60

40

40

20

20

0
0

50

100

150

200

250

300

350

400

50

100

150

200

Fig. 21. Effect of endplate thickness in connections with endplate made of Q235 at
elevated temperature.

300

350

400

Fig. 23. Effect of endplate thickness in connections with endplate made of S960 at
elevated temperature.

Table 8
Connection rotation at ambient temperature obtained in tests.

140

S690 12mm
S690 15mm

120

Series

Connection ID

Endplate

Connection rotation

Material

Thickness
(mm)

/Mmax (mrad)

/C (mrad)

100

moment (kNm)

250

rotation (mrad)

rotation (mrad)

1-1 A
1-2 A
1-3 A

Q235
S690
S960

20
12
10

168
105
69

199
214
105

2-1
2-2
2-3
2-4

Q235
Q345
S690
S960

25
20
15
12

89
63
83
85

134
130
101
105

80

60

A
A
A
A

40

20

Table 9
Connection rotation at elevated temperature 550 C obtained in tests.

0
0

50

100

150

200

250

300

350

400

Series

Connection
ID

Endplate
Material

Thickness
(mm)

/Mmax
(mrad)

/C
(mrad)

1-1 E
1-2 E
1-3 E

Q235
S690
S960

20
12
10

106
69
76

314
304
313

2-1
2-2
2-3
2-4

Q235
Q345
S690
S960

25
20
15
12

84
90
58
64

191
313
330
320

rotation (mrad)
Fig. 22. Effect of endplate thickness in connections with endplate made of S690 at
elevated temperature.

4.1.1. At ambient temperature

The failure modes of endplate connections at ambient temperature obtained from theoretical analysis based on Eurocode 3 Part:
18 [1] and experimental study are listed in Table 10. It can be observed that good agreement exists between the theoretical analysis
and the experimental study.
4.1.2. At elevated temperatures
The failure modes of endplate connections at elevated temperature 550 C obtained from theoretical analysis based on Eurocode
3 Part: 18 [1] and experimental study are listed in Table 11. It can
be found that there is good agreement between the theoretical
analysis and the experimental study.
It is worthwhile to mention that the failure modes of all connections in series 1 turned from Mode 1 at ambient temperature to
Mode 2 at elevated temperature 550 C. This phenomenon attributes to the fact that the reduction of material properties of bolts
under re conditions is quicker than that of structural steels,
according to the reduction factors of material properties of bolts
and mild steels under re conditions recommended in Eurocode

E
E
E
E

Connection rotation

Table 10
Failure modes of endplate connections at ambient temperature.
Connection ID

Endplate

Failure mode

Material

Thickness (mm)

EC3

Test

1-1 A
1-2 A
1-3 A

Q235
S690
S960

20
12
10

Mode 1
Mode 1
Mode 1

Mode 1
Mode 1
Mode 1

2-1
2-2
2-3
2-4

Q235
Q345
S690
S960

25
20
15
12

Mode
Mode
Mode
Mode

Mode
Mode
Mode
Mode

A
A
A
A

2
2
2
2

2
2
2
2

3 Part: 12 [6] and the experimental results on the material properties reduction factors of high strength steels in re presented in
[711].

33

X. Qiang et al. / Engineering Structures 64 (2014) 2338

Table 11
Failure modes of connections at elevated temperature 550 C.
Connection ID

Endplate

Failure mode

Material

Thickness (mm)

EC3

Test

1-1 E
1-2 E
1-3 E

Q235
S690
S960

20
12
10

Mode 2
Mode 2
Mode 2

Mode 2
Mode 2
Mode 2

2-1
2-2
2-3
2-4

Q235
Q345
S690
S960

25
20
15
12

Mode
Mode
Mode
Mode

Mode
Mode
Mode
Mode

E
E
E
E

2
2
2
2

2
2
2
2

where this secant stiffness intersects with the experiment MU

curve, as illustrated in Fig. 26.
Eurocode 3 Part: 18 [1] includes Weynands proposal [15], as a
simplied method to obtain the plastic exural resistance of connections. In the simplied method, the initial rotational stiffness
of the connection (Sj,ini) and the momentrotation curve are used
to obtain the plastic exural resistance value. Firstly, the secant
stiffness (Sj) is calculated by Eq. (4), where the stiffness modication coefcient g is 2 for beam-to-column bolted endplate connections, as illustrated in Table 5.2 of Eurocode 3 Part: 18 [1]. Then

Further, it can be found that in each connection series both at

ambient and at elevated temperatures the failure modes of connections are the same. Thus all in all, in endplate connections an optimal design using a thinner high strength steel endplate can achieve
the same failure mode, similar load bearing capacity and comparable or even higher rotation capacity in comparison to an endplate
connection with thicker mild steel endplate. This experimental
study offers reliable understandings on performance of high
strength steel endplate connections both in re and at ambient
temperature. It probably opens a perspective using high strength
steels instead of mild steels in practical structural design to some
extent.

Fig. 25. Mj,R according to Zanon and Zandonini [14].

4.2. Plastic exural resistance of endplate connections

Eurocode 3 Part: 18 [1] gives quantitative rules for the prediction of plastic exural resistance (design moment resistance) Mj,Rd
of beam-to-column connections. The prediction based on Eurocode
3 was performed on all test connection specimens at ambient and
elevated temperatures.
For experimental study, a question arises on how to determine
the plastic exural resistance Mj,R from a specic test result. In the
literature, Steenhuis et al. [12] give an overview of approaches how
to determine the plastic exural resistance Mj,R from a test result.
Jaspart [13] proposes that the plastic exural resistance can be
read at the vertical axis from a test momentrotation curve by
drawing a secant line through the part of the test curve with
post-limit stiffness, see Fig. 24. Alternatively, Zanon and Zandonini
[14] give a denition of plastic exural resistance Mj,R as shown in
Fig. 25, where the plastic exural resistance Mj,R corresponds to the
intersection of the initial stiffness and the post-limit stiffness.
Weynand [15] suggests another denition. In his case, the initial
stiffness of the connection is assessed as the elastic stiffness of
the MU curve. Then, a secant stiffness is determined as the initial
stiffness divided by a xed stiffness modication coefcient g. The
plastic exural resistance Mj,R is dened as the moment level,

Fig. 26. Mj,R according to Weynand [15].

250

200

moment (kNm)

Mj,Rd, test,1
150

Mj,Rd, test,2

100

Q235 20mm test

Mj,Rd, EC3
Sj,ini,test

50

Sj,ini,test/

0
0

50

100

150

200

250

rotation (mrad)

Fig. 24. Mj,R according to Jaspart [13].

Fig. 27. Comparison on plastic exural resistance of 1-1 A (Q235 20 mm) at

ambient temperature.

34

X. Qiang et al. / Engineering Structures 64 (2014) 2338

300

250

Mj,Rd, test,2

Mj,Rd, test,2

200

Mj,Rd, test,1

Mj,Rd, test,1

200

moment (kNm)

moment (kNm)

250

150

100

S690 12mm test

Mj,Rd, EC3

50

150

Q345 20mm test

Mj,Rd, EC3

100

Sj,ini,test
Sj,ini,test/

50

Sj,ini,test
Sj,ini,test/

0
0

50

100

150

200

250

20

40

60

rotation (mrad)
Fig. 28. Comparison on plastic exural resistance of 1-2 A (S690 12 mm) at ambient
temperature.

100

120

140

160

Fig. 31. Comparison on plastic exural resistance of Q345 20 mm at ambient

temperature (2-2A).

250

300

250

Mj,Rd, test,2

200

Mj,Rd, test,2
Mj,Rd, test,1

moment (kNm)

200

moment (kNm)

80

rotation (mrad)

Mj,Rd, test,1

150

100

S960 10mm test

Mj,Rd, EC3

150

S690 15mm test

Mj,Rd, EC3

100

Sj,ini,test

Sj,ini,test

50

Sj,ini,test/

Sj,ini,test/

50

20

40

60

80

100

120

140

160

20

40

60

80

100

120

140

rotation (mrad)

rotation (mrad)
Fig. 29. Comparison on plastic exural resistance of 1-3 A (S960 10 mm) at ambient
temperature.

Fig. 32. Comparison on plastic exural resistance of S690 15 mm at ambient

temperature (2-3A).

300

300

Mj,Rd, test,2

250

250
200

150

moment (kNm)

moment (kNm)

Mj,Rd, test,1
200

Q235 25mm test

Mj,Rd, EC3

100

Sj,ini,test

150

S960 12mm test

Mj,Rd, EC3

100

Sj,ini,test

Sj,ini,test/

50

Mj,Rd, test,2
Mj,Rd, test,1

Sj,ini,test/

50

0
0

20

40

60

80

100

120

140

160

rotation (mrad)
Fig. 30. Comparison on plastic exural resistance of Q235 25 mm at ambient
temperature (2-1A).

20

40

60

80

100

120

140

rotation (mrad)
Fig. 33. Comparison on plastic exural resistance of S960 12 mm at ambient
temperature (2-4A).

35

X. Qiang et al. / Engineering Structures 64 (2014) 2338

100
100

Mj,Rd, test,2

80
80

moment (kNm)

moment (kNm)

Mj,Rd, test,2
Mj,Rd, test,1

60

40

Q235 20mm test

Mj,Rd, EC3
Sj,ini,test

20

Mj,Rd, test,1

60

40

S960 10mm test

Mj,Rd, EC3

20

Sj,ini,test/

Sj,ini,test
Sj,ini,test/

50

100

150

200

250

300

350

400

50

100

150

rotation (mrad)
Fig. 34. Comparison on plastic exural resistance of 1-1 E (Q235 20 mm) at
elevated temperature 550 C.

250

300

350

400

Fig. 36. Comparison on plastic exural resistance of 1-3 E (S960 10 mm) at elevated
temperature 550 C.

120

140

S690 12mm test

Mj,Rd, EC3

Mj,Rd, test,2

Sj,ini,test

Mj,Rd, test,1

80

120

Mj,Rd, test,1
Mj,Rd, test,2

100

Sj,ini,test/

moment (kNm)

100

moment (kNm)

200

rotation (mrad)

60

40

20

80
60
40

Q235 25mm test

Mj,Rd, EC3

20

Sj,ini,test/

Sj,ini,test

0
0

50

100

150

200

250

300

350

400

50

100

rotation (mrad)
Fig. 35. Comparison on plastic exural resistance of 1-2 E (S690 12 mm) at elevated
temperature 550 C.

200

250

300

Fig. 37. Comparison on plastic exural resistance of 2-1 E (Q235 25 mm) at

elevated temperature 550 C.

the plastic exural resistance can be obtained by the intersection

of the moment rotation curve and a straight line from the origin
with slope Sj.

120

In this investigation, Zanon and Zandoninis proposal is

employed as Method 1 to obtain the plastic exural resistance
Mj,Rd,test,1 of the endplate connection from an experimental
momentrotation curve. For comparison, the simplied method
recommended in Eurocode 3 Part: 18 [1] is used as Method 2 to
obtain the plastic exural resistance Mj,Rd,test,2 (for instance see
Fig. 27). The plastic exural resistances of endplate connections
obtained from this experimental study through these two methods
are compared with the predictions of Eurocode 3, as shown in
Figs. 2740.

Mj,Rd, test,2
Mj,Rd, test,1

100

moment (kNm)

Sj Sj;ini =g

150

rotation (mrad)

80

60

Q345 20mm test

Mj,Rd, EC3

40

Sj,ini,test

20

Sj,ini,test/

4.2.1. At ambient temperature

The plastic exural resistances of all endplate connections at
ambient temperature obtained from this experimental study are
compared with the theoretical predictions of Eurocode 3, as listed
in Table 12. It can be observed that reasonable agreements exist
between the theoretical predictions and the experimental results.
Via Ratio3, presented in Table 12, the observation can be found that
Mj,Rd,test,1 obtained based on Zanon and Zandoninis denition is

50

100

150

200

250

300

350

400

rotation (mrad)
Fig. 38. Comparison on plastic exural resistance of 2-2 E (Q345 20 mm) at
elevated temperature 550 C.

generally smaller than Mj,Rd,test,2, which is dened according to

Weynands proposal and the simplied method recommend by
Eurocode 3. By comparing Mj,Rd,test,1 with the predicted plastic

36

X. Qiang et al. / Engineering Structures 64 (2014) 2338

140

Mj,Rd, test,2

120

moment (kNm)

100

Mj,Rd, test,1

80
60

S690 15mm test

Mj,Rd, EC3

40

Sj,ini,test
Sj,ini,test/

20
0
0

50

100

150

200

250

300

350

400

rotation (mrad)
Fig. 39. Comparison on plastic exural resistance of 2-3 E (S690 15 mm) at elevated
temperature 550 C.

140

Mj,Rd, test,2

120

moment (kNm)

100

Mj,Rd, test,1

80
60
40

S960 12mm test

Mj,Rd, EC3

20

Sj,ini,test
Sj,ini,test/

0
0

50

100

150

200

250

300

350

400

rotation (mrad)
Fig. 40. Comparison on plastic exural resistance of 2-4 E (S960 12 mm) at elevated
temperature 550 C.

exural resistance according to Eurocode 3, it can be seen that the

predictions of Eurocode 3 are generally on the non-conservative
side when the test result is obtained based on Zanon and
Zandoninis denition. However, the comparison of Mj,Rd,test,2 with
the predicted plastic exural resistance according to Eurocode 3
shows that the predictions of Eurocode 3 are generally on the
conservative side.
Fig. 27 compares the plastic exural resistance Mj,Rd predicted
by Eurocode 3 with the tested momentrotation curve of connection 1-1 A (Q235 20 mm), and with the test obtained plastic

exural resistances Mj,Rd,test,1, Mj,Rd,test,2. It can be seen that Mj,Rd,EC3

is slightly higher than Mj,Rd,test,1 and Mj,Rd,test,2 for connection 1-1 A
(Q235 20 mm) at ambient temperature. The plastic exural resistances Mj,Rd of connection 1-2 A (S690 12 mm) and 1-3 A (S960
10 mm) predicted by Eurocode 3 are compared with the tested momentrotation curve and experimentally obtained plastic exural
resistances Mj,Rd,test,1 and Mj,Rd,test,2, as shown in Figs. 28 and 29. It
can be observed that for connection 1-2 A the plastic exural resistance Mj,Rd,test,1 is smaller than that of Mj,Rd,test,2, while the prediction of Eurocode 3 Mj,Rd,EC3 is conservative in comparison to
Mj,Rd,test,1 and Mj,Rd,test,2 for connection 1-2 A at ambient temperature. Similarly, the plastic exural resistance Mj,Rd,test,1 is smaller
than that of Mj,Rd,test,2 for connection 1-3 A. Moreover, the prediction of Eurocode 3 Mj,Rd,EC3 is between Mj,Rd,test,1 and Mj,Rd,test,2 for
connection 1-3 A at ambient temperature.
The comparison of the plastic exural resistance Mj,Rd predicted
by Eurocode 3 with the tested results by two denition methods
for the four connections in Series 2 at ambient temperature are
presented in Figs. 3033.
4.2.2. At elevated temperature 550 C
The plastic exural resistances of all endplate connections at
elevated temperature 550 C are compared with the theoretical
predictions of Eurocode 3, as listed in Table 13. It can be seen that
reasonable agreements exist between the theoretical predictions
and the experimental results.
By comparing Ratio3, presented in Table 13, it can be found that
Mj,Rd,test,1 obtained based on Zanon and Zandoninis denition is
generally smaller than Mj,Rd,test,2, which is dened according to
Weynands proposal as well as the simplied method recommended
by Eurocode 3. By comparing Mj,Rd,test,1 with the predicted plastic
exural resistance by Eurocode 3, see Ratio1 as presented in
Table 13, it can be seen that the predictions of Eurocode 3 are
generally non-conservative when the test result is obtained based
on Zanon and Zandoninis denition. However, the comparison of
Mj,Rd,test,2 with the predicted plastic exural resistance by Eurocode
3 shows that the predictions of Eurocode 3 are generally on the
conservative side, when the denition of the test result is based
on Weynands proposal and the simplied method recommended
by Eurocode 3.
The comparisons of plastic exural resistance Mj,Rd predicted by
Eurocode 3 with the experimentally obtained plastic exural resistances Mj,Rd,test,1 and Mj,Rd,test,2 for these seven endplate connections
at elevated temperature 550 C are presented in Figs. 3440.
4.3. Rotation capacity of endplate connections
Eurocode 3 Part: 18 [1] states that at ambient temperature a
bolted endplate connection may be assumed to have sufcient
rotation capacity for plastic analysis, provided that both of the following conditions are satised: (i) the moment resistance of the
connection is governed by the resistance of either the column

Table 12
Evaluation of plastic exural resistance of connections at ambient temperature.
Test ID

Connections

Mj,Rd,EC3 (kN m)

Mj,Rd,test,1 (kN m)

Mj,Rd,test,2 (kN m)

1-1 A
1-2 A
1-3 A

Q235 20 mm
S690 12 mm
S960 10 mm

167.07
176.04
170.62

155.61
207.18
159.98

2-1
2-2
2-3
2-4

Q235 25 mm
Q345 20 mm
S690 15 mm
S960 12 mm

245.14
239.79
248.54
234.17

243.15
235.79
225.92
199.65

A
A
A
A

j;Rd;EC3
Ratio1 Mj;Rd;test;1

j;Rd;EC3
Ratio2 Mj;Rd;test;2

Ratio3 Mj;Rd;test;1
j;Rd;test;2

153.20
222.73
181.98

1.074
0.850
1.067

1.091
0.790
0.938

1.016
0.930
0.879

259.41
255.04
233.22
202.987

1.008
1.017
1.100
1.173

0.945
0.940
1.066
1.154

0.937
0.925
0.969
0.984

Note: Mj,Rd,EC3 is the predicted plastic exural resistance according to Eurocode 3; Mj,Rd,test,1 is the test obtained plastic exural resistance according to Zanon and Zandoninis
evaluation method; Mj,Rd,test,2 is the test obtained plastic exural resistance according to Weynands evaluation method.

37

X. Qiang et al. / Engineering Structures 64 (2014) 2338

Table 13
Evaluation of plastic exural resistance of connections at elevated temperature 550 C.
Test ID

Connections

1-1 E
1-2 E
1-3 E

Q235 20 mm
S690 12 mm
S960 10 mm

76.77
89.90
95.26

68.13
89.99
84.68

2-1
2-2
2-3
2-4

Q235 25 mm
Q345 20 mm
S690 15 mm
S960 12 mm

109.29
107.37
105.94
105.41

106.50
104.68
101.25
104.83

E
E
E
E

Mj,Rd,EC3 (kN m)

Mj,Rd,test,1 (kN m)

Mj,Rd,test,2 (kN m)

j;Rd;EC3
Ratio1 Mj;Rd;test;1

j;Rd;EC3
Ratio2 Mj;Rd;test;2

Ratio3 Mj;Rd;test;1
j;Rd;test;2

72.94
105.38
93.74

1.127
0.999
1.125

1.053
0.853
1.016

0.934
0.854
0.903

108.79
105.86
120.02
111.90

1.026
1.026
1.046
1.006

1.005
1.014
0.883
0.942

0.979
0.989
0.844
0.937

Note: Mj,Rd,EC3 is the predicted plastic exural resistance according to Eurocode 3; Mj,Rd,test,1 is the test obtained plastic exural resistance according to Zanon and Zandoninis
evaluation method; Mj,Rd,test,2 is the test obtained plastic exural resistance according to Weynands evaluation method.

Table 14
Verication of the guidelines of EC3 on rotation capacity at ambient temperature.
Connection ID

Endplate
Material

Thickness
tep (mm)

Max thickness tep (mm)

according to EC3

1-1 A
1-2 A
1-3 A

Q235
S690
S960

20
12
10

18
11
9

2-1
2-2
2-3
2-4

Q235
Q345
S690
S960

25
20
15
12

18
15
11
9

A
A
A
A

5. Conclusion

ange in bending or the endplate in bending and (ii) the thickness t

of either the column ange or the endplate (not necessarily the
same basic component as in i) satises:

s
fu;b
fy

t 6 0:36d

elevated temperature 550 C the connection rotation at maximum

load /Mmax is higher than 4050 mrad, which is the minimum connection rotation to be considered to have sufcient rotation capacity for partial strength connections. Hence the rotation capacity of
high strength steel endplate connections in re is considered to be
sufcient, which guarantees the ductility and re safety of steel
structures using endplate connections made of high strength steel.

where d is the bolt diameter, fu,b is the tensile strength of the bolt
and fy is the yield stress of the relevant basic component (column
ange or the endplate). The verication of these guidelines in Eurocode 3 Part: 18 [1] with the characterization of the rotation capacity of high strength steel endplate connections obtained from tests
is carried out. It can be found that the rst condition is guaranteed
for all specimens (the moment resistance of connection is always
governed by the resistance of the endplate in bending), while the
second condition, Eq. (5) is not satised for all specimens, as listed
in Table 14.
In the literature it is generally accepted that a minimum of
4050 mrad ensures sufcient rotation capacity of a bolted
connection in a partial strength scenario [16,17]. Wilkinson et al.
suggested that a connection in steel moment resisting frames in a
seismic area must develop a minimum plastic rotation of 30 mrad
[18]. Table 8 shows that the maximum rotation capacities /C of all
specimens at ambient temperature in this experimental study are
much higher than 4050 mrad, even for all specimens the rotations
at maximum load /Mmax guarantee this condition as well. Therefore, the Eurocode 3 [1] current provisions seem too conservative
on rotation capacity of bolted endplate connections. In addition,
the current provisions on sufcient rotation capacity of connections in Eurocode 3 were mainly obtained from bolted connections
made of mild steels, and they need to be calibrated when applied
to connections made of high strength steels or very high strength
steels. In order to establish more accurate criteria for connections
made of high strength steels, further work such as more repetitive
tests as well as parametric study are still required in practice.
The maximum rotation capacity /C at elevated temperature
550 C of all specimens in this experimental study is higher in
comparison to that at ambient temperature. For all specimens at

Fire tests on seven endplate connections of which the endplates

are made of high strength steels (or mild steels for comparison)
were reported in this paper. As reference, the full scale tests on
these seven endplate connections at ambient temperature were
also conducted and presented herein. All specimens were designed
to trigger failure in the endplate connections rather than in beam
or column. The following conclusions can be drawn from this
experimental study:
1. The load bearing capacity as well as rotation capacity of endplate connections is dependent on the combination of endplate
material and endplate thickness.
2. For endplate connections the endplates of which are made of
the same steel material, with a thicker endplate the connection
can achieve higher load bearing capacity but its rotational
capacity is reduced obviously at both ambient temperature
and elevated temperatures. This is valid for both mild steels
(i.e. Q235) and high strength steels (i.e. S690 and S960).
3. The accuracy of Eurocode 3 for plastic exural resistance of endplate connections both at ambient temperature and under re
conditions is acceptable, no matter the endplate is made of mild
steels or high strength structural steels.
4. The plastic exural resistance of connections based on the definition of Zanon and Zandonini is generally smaller than that
according to Weynands proposal as well as the simplied
method recommend by Eurocode 3. This conclusion is valid
for all endplate connections in this experimental study (both
mild steel endplate connections and high strength steel endplate connections) at ambient and elevated temperatures.
5. The Eurocode 3 current provisions on rotation capacity of
bolted connections seem too conservative, which were mainly
obtained from connections made of mild steels. They need to
be calibrated when applied to connections made of high
strength steels and very high strength steels.
6. The rotation capacity of high strength steel endplate connections under re conditions is considered to be sufcient, which
guarantees the ductility and re safety of steel structures using
connections made of high strength steel.
7. In endplate connections, a proper design using a relatively thin
high strength steel endplate can achieve the same failure mode,
similar load bearing capacity and comparable or even higher

38

X. Qiang et al. / Engineering Structures 64 (2014) 2338

rotation capacity, both at ambient temperature and in re, in

comparison to a connection with relatively thick mild steel
endplate.
8. This experimental study opens a perspective of using high
strength steels to take place of mild steels in structural optimizing design. For practical applications and re-resistance design
of other types of connections and structural members, more
research work (such as tests on more types of connections and
parametric studies) should be further conducted on the behaviour of high strength steel structures under re conditions.

Acknowledgement
The preparation of test specimens by Shanghai Baoye Engineering Technology Co., Ltd. is gratefully acknowledged.
References
[1] CEN. Eurocode3: Design of steel structures. Part 18: Design of joints. Brussels;
2005.
[2] CEN. Eurocode 3: Design of steel structures. Part 112: additional rules for the
extension of EN 1993 up to steel grades S700. Brussels; 2004.
[3] Coelho AMG, Bijlaard FSK. Experimental behaviour of high strength steel endplate connections. J Constr Steel Res 2007;63:122840.
[4] Qiang X, Bijlaard FSK, Kolstein MH, Jiang X. Behaviour of beam-to-column high
strength steel endplate connections under re conditions Part 2: numerical
study. Eng Struct 2014;64:3951.
[5] Qiang X. Behaviour of high strength steel endplate connections in re and after
re. PhD thesis. Delft University of Technology, Delft, The Netherlands; 2013.

[6] CEN. Eurocode3: Design of steel structures. Part 12: general rules Structural
re design. Brussels; 2005.
[7] Lange J, Wohlfeil N. Examination of the mechanical properties of the
microalloyed grain rened steel S 460 at elevated temperatures. Bautechnik
2007;84:71120.
[8] Schneider R, Lange J. Constitutive equations of structural steel S460 at high
temperatures. In: Nordic steel construction conference 2009, Sweden; 2009. p.
20411.
[9] Qiang XH, Bijlaard FSK, Kolstein H. Elevated-temperature mechanical
properties of high strength structural steel S460N: experimental study and
recommendations for re-resistance design. Fire Saf J 2013;55:1521.
[10] Qiang XH, Bijlaard F, Kolstein H. Dependence of mechanical properties of high
strength steel S690 on elevated temperatures. Constr Build Mater
2012;30:739.
[11] Chen J, Young B, Uy B. Behavior of high strength structural steel at elevated
temperatures. J Struct Eng ASCE 2006;132:194854.
[12] Steenhuis CM, Vrouwenvelder ACWM, Herwijnen Fv, Snijder HH. Denitions
of resistance and deformation capacity for non-sway steel and composite
structures. Heron 2002:47.
[13] Jaspart JP. Study of the semi-rigidity of beam-to-column joints and its
inuence on the resistance and stability of steel buildings. PhD thesis. Lige
University, Lige, Belgium; 1991.
[14] Zanon P, Zandonini R. Experimental analysis of end plate connections. In:
Proceedings of the state of the art workshop on connections and the behaviour
of strength and design of steel structures, Cachan; 1988. p. 4151.
[15] Weynand K. Sicherheits- und Wirtschaftlichkeitsuntersuchungen zur
Anwendung nachgiebiger Anschlsse im Stahlbau. Heft 35, Shaker Verlag.
Aachen; 1997.
[16] Coelho AMG, da Silva LS, Bijlaard FSK. Ductility analysis of bolted extended end
plate beam-to-column connections in the framework of the component
method. Steel Compos Struct 2006;6:3353.
[17] Coelho AMG. Characterization of the ductility of bolted end plate beam-tocolumn steel connections. PhD thesis. University of Coimbra, Coimbra,
Portugal; 2004.
[18] Wilkinson S, Hurdman G, Crowther A. A moment resisting connection for
earthquake resistant structures. J Constr Steel Res 2006;62:295302.