Slanted Connection at Elevated Temperature Under Various Loadings

Engineering and Computational Mechanics
Volume 168 Issue EM1

Slanted connection at elevated temperature
under various loadings
Taghia, Zahmatkesh, Osman and Talebi
Proceedings of the Institution of Civil Engineers

Engineering and Computational Mechanics 168 March 2015 Issue EM1
Pages 316 http://dx.doi.org/10.1680/eacm.14.00009
Paper 1400009
Received 06/05/2014
Accepted 01/12/2014
Keywords: buildings, structures & design/computational mechanics/steel
structures
ICE Publishing: All rights reserved
Slanted connection at elevated

temperature under various
loadings
Pezhman Taghia MEng
Mohd Hanim Osman AUTM
Master of Engineering, Department of Civil Engineering, Universiti

Teknologi Malaysia, Johor Bahru, Malaysia
Associate Professor, Department of Civil Engineering, Universiti Teknologi

Malaysia, Johor Bahru, Malaysia
Farshad Zahmatkesh PhD
Elnaz Talebi PhD
PhD Student, Department of Civil Engineering, Universiti Teknologi

PhD Student, Department of Civil Engineering, Universiti Teknologi

This paper presents a detailed comparison of two types of slanted end-plate connections, under symmetric and nonsymmetric gravity loads, in terms of friction factors and angles of slanted end-plate connections. This study
contributes to better understanding of roll of slanted end-plate connections under non-symmetric gravity loads on
the value of the induced initial axial force due to elevated temperature in steel beams, and a comparison of the
effect of symmetric gravity loads on the behaviour of such connections, before and after elevated temperature. The
results imply that, before any thermal effects, a non-symmetric gravity load can be more critical than a symmetric
load; the average ratio of axial force of a non-symmetric load to a symmetric load was found to be 1 .491 .57. This
value was determined to be 1 .641 .90 when friction bolts instead of normal bolts were used. On the other hand,
when a steel beam is subjected to a temperature increase, the amount of reaction and initial axial forces in the steel
beam are not related to the type of loading. This can be considered an advantage of the slanted end-plate
connection compared with a conventional vertical end-plate connection, and it could be developed in structures
exposed to longitudinal movement due to expansion. The optimum angle of the slanted end-plate connection in
both cases ranged from 178 to 608.
Notation
A
a
E
I
L
Pi
Pt
Ptmax
W
T
Tm
s
1.
cross-section of beam column

vertical reaction factor of support due to nonsymmetric gravity load
Youngs modulus
moment of inertia
length of the beam column
compression axial load
axial load due to increase in temperature
movement axial load due to increase in temperature
distributed load (gravity load)
coefficient of thermal expansion
temperature increase
movement temperature
friction factor
slant angle
Introduction
When steel beams are exposed to elevated temperature, control of

expansion plays a vital role in the performance of the beams in
terms of induced axial force, yield and buckling. Indeed, expansion of steel beams due to uniform heating is one of the primary
causes of elastic failure (Mourao and Silva, 2007). In the past,
efforts have focused on the protection of isolated structural
members against fire and elevated temperatures (Dai et al.,
Downloaded by [] on [25/08/16]. Copyright ICE Publishing, all rights reserved.
2009). Moreover, the detail of the connections can play an

important role in the thermal behaviour of steel structures. The
performance of conventional vertical end-plate connections has
been investigated by a number of researchers. Vertical connections subjected to elevated temperatures have been numerically
simulated by Daz et al. (2011) and (Yang and Tan (2012), and
vertical types of connections have also been studied by
Heidarpour and Bradford (2009) and da Silva et al. (2001). The
performance of end-plate connections was tested experimentally
by Daryan and Bahrampoor (2009) and Qian et al. (2008).
Exploration of all previous methods shows that, generally, there
are two points of view regarding the strengthening of restrained
steel members subjected to heat one of them is yield stress
consideration and the other focuses on buckling of the members.
The strength reduction of a restrained member due to elevated
temperature is defined by critical temperature. The critical temperature for a steel beam is the temperature at which the beam
cannot safely carry a gravity load. Many building codes and
structural engineering standards define different critical temperatures, depending on the structural element type, configuration
orientation and loading characteristics (ECS, 1995; Zalosh,
2003), but none specify data for damping and reducing thermal
effects. The methods used for thermal control have also expanded
to different fields such as conventional steel frames and bridge
3


temperature under various loadings
girders to avoid undesirable movements. During the design of

steel structures, designers have to use a moment frame system or
a dual system (bracing or shear wall plus moment frame) to resist
lateral loads. Moreover, proper methods to minimise thermal
effects also need to be considered. Nowadays, engineering
solutions commonly used to decrease thermal influences include
increasing the section area, applying lateral supports, applying a
cooling system by means of air conditioning or water, system
covering with concrete or isolation and thermal breaks.
The evaluation of conventional vertical steel connections at
elevated temperatures has received much research attention, but
less attention has been devoted to slanted end-plate connections.
The current study was designed to fill this gap in the research.
An analytical model of a beam with slanted end-plate connection
subjected to symmetric and non-symmetric gravity loads and
temperature increase was formulated by Zahmatkesh and Talebi
(2010). It was demonstrated that an increase in temperature in a
beam with a conventional vertical bolted connection induces a
huge extra axial force, which decreases the beams capability of
carrying an external symmetric gravity load. By changing the
orientation of the connection to the slanted form, this additional
axial force will be dampened. In the analytical studies of
Zahmatkesh and Talebi (2010), a steel beam subjected to different
gravity loads (symmetric and non-symmetric) was assessed.
Direct stiffness finite-element modelling of a slanted end-plate
connection under non-symmetric load after thermal effect was
conducted to assess the elastic mechanical behaviour of such
connections, which was verified by experimental tests (Figures 1
and 2) (Zahmatkesh et al., 2014b). The effect of a bolted slanted
end-plate connection to reduce the extra thermal-induced axial
force in the elastic field was also analysed (Zahmatkesh et al.,
2014b).
Although the previous models described the behaviour of such
connections, an appropriate comparison between two case studies
would be useful to enable better understanding of the properties
of slanted end-plate connections. The comparative study de-
Gravity load
Force sensor
Beam
0688L
Figure 1. Experimental test setup (a 0 .688) (Zahmatkesh et al.,

2014b)
Figure 2. Experimental thermal test set-up (Zahmatkesh et al.,

2014b)
scribed here was thus conducted. It should be noted that most

methods to reduce the undesirable effects of conventional
connections are expensive. One of the main advantages of the
proposed connection (i.e. slanted end-plate) is cost savings compared with conventional end-plate connections. On the basis of
the obtained equations presented in the text, thermal axial forces
can be reduced when slanted end-plate connections are used
instead of conventional ones. Therefore, a steel beam member
can be designed for reduced thermal axial forces and, accordingly, member section sizes and the weight of the whole structure
can be reduced in comparison with a system using vertical endplate connections.
2.
Mechanical properties of a slanted endplate connection at elevated temperature
2.1
Beam subjected to symmetric gravity load (slanted

end-plate connection with frictionless support)
In the frictionless condition of slanted end-plate connections, an
idealised roller connection is assumed in which there is no
friction effect between two plates. The beam tends to move
downward when it is subjected to only a symmetric gravity load
W but an induced axial force Pi due to the gravity load resists the
downward motion to satisfy static equilibrium (Figure 3(a)).
Regarding both cases a beam under symmetric gravity load and
the existence of induced extra internal axial force in the beam
due to temperature increase (causing upward movement) the
following focuses on the frictionless condition of slanted endplate connections. The following equations show the behaviour of
these kinds of supports under symmetric gravity load and in
frictionless condition with various slant angles. According to
Figure 3(a), a uniform gravity load on the beam and the slope of
the connection are related as
1:
Pi
WL
cot
2
(frictionless)
where W is the distributed load, L is the length of the beam

column and Pi is the compression axial load.

Roller

Roller
L
(a)
Friction
connection
Friction
connection
Pi
s tan
N
Mi
(a)
W
Pt
Pt
Pi
M
(b)
N
Pb
s tan
Mi
(b)
Pb
Figure 4. Free body diagrams of beams with friction support:

(a) before thermal effect; (b) after temperature increase
Pt
Pt
(c)
Figure 3. Simplified model of a beam with slanted end-plate

connection: (a) frictionless support; (b) friction support; (c) bolted
slanted end-plate connection
2.1.1 Slanted end-plate connection with friction support

In the case shown in Figure 3(b), it is supposed that there is a
friction component on the support, acting opposite to the direction of movement; this implies that friction acts as the resistance
force against slanted end-plate connection movement. In this
case, the beam will be in static equilibrium before the thermal
effect (temperature increase) and the friction force resists against
downward movement caused by the gravity load. In contrast, after
a temperature increase, the beam would tend to expand in the
longitudinal direction, while there is an axial resistance force
against the elongation. Indeed, the tangential friction force on the
slanted surface resists upward sliding. According to static
equilibrium, before any thermal effect on the beam (Figure 4(a)),
the axial force can be derived as
2:
Pi
WL
cot( ) (with friction)
2
where is the angle between N WL=2 sin and the line

perpendicular to the end-plate.
After an increase in temperature, the friction force changes
direction. Referring to Figure 4(b)
3:
Ptmax
WL
cot( ) (with friction)
2
To describe how the friction force changes its direction, it should

be noted that the natural behaviour of the friction force tends to
resist against the movement direction. In a slanted end-plate
connection, there are two cases of movement direction before
the elevated temperature and after the elevated temperature.
Before elevated temperature, the steel beam is under gravity load
only. The beam thus tends to slide downward on the inclined
plane of the end-supports, but the generated friction force due to
gravity load tends to resist against the movement direction. As a
result (Figure 5(a)), the friction force presents upwards.
After elevated temperature (Figure 5(b)), the beam is subjected to
a combination of thermal expansion and gravity load. In this
case, thermal expansion tends to move the beam upward on the
inclined plane. On the other hand, similar to the previous case,
the gravity load tends to slide the beam downward. A rise in
temperature increases the axial force. At elevated temperatures,
the values of the upward movement forces due to thermal
expansion are higher than the downward sliding force due to the
gravity load. In this case, the friction force tends to resist against
the movement direction and hence the friction force changes its
direction to downward to resist the upward movement.
As shown by the well-known definition of induced force in a
beam due to thermal effects (Equation 4), the amount of induced
axial force in the beam depends on temperature increase T, the
coefficient of thermal expansion and material properties such
as the modulus of elasticity (Youngs modulus) E and the crosssection of the beam A
5


Pi
W
Y2
Y1
X1 L1 05L
1
Ry1
Ff
Pi
Y2
1 90
Element 2
Y3
Y1
Pi
X3
Y3
X2 L 05L
2
Element 1
90
Pi
X2
2 0
90
Ry3
Ff
X1
RM1
X3
3 1
RM3
(a)
W
Pi
Y1
X1 L1 05L
1
Ff
Ry1
W
Y2
Pi
Y2
Pi
Element 2
Y3
Y1
1 90
X3
Y3
X2 L 05L
2
Element 1
90
2 0
X2
Pi
90
Ff
Ry3
X1
RM1
X3
3 1
RM3
(b)
Figure 5. Force diagrams of beam with slanted end-plate

connections due to symmetric transverse load (friction supports:
(a) before elevated temperature; (b) after temperature increase
4:
Pt AET
Combining Equations 3 and 4, the movement elevated temperature is
5:
T m
WL
cot( )
2AE
upward and the friction force changes its direction and acts
against upward crawling (Figure 6(b)). In the case of the friction
bolt connection, there is an additional force due to the existence
of bolts, which helps to resist movement. By applying static
equilibrium on the connection, before thermal effects, Pi can be
derived as
(with friction)
6:
WL 2Pb sin
2 sin( )
Pi
WL 2Pb sin
cot( ) Pb cos
2
in which is the slant angle.

2.1.2 Slanted end-plate connection with friction bolt
support
In the third case (Figure 3(c)) it is supposed that the supports
reaction is affected by both friction force and bolt compression
forces. In the first step, before temperature increase, the beam is
in static equilibrium and the resultant frictional force resists
against downward movement, similar to friction support (see
Figure 6(a)). After a temperature increase, the beam would move
6
7:
The support reactions before and after thermal effect are given by
Equations 6 and 8, respectively. Equations 9 and 10 show the
relation between movement elevated temperatures and various


Pb
Pb
W
Pb
s tan
N
Pi
M
Mi
(a)
N
Pb
Pb
W
Pb
s tan
Pt
M
Mi
(b)
Figure 6. Free body diagrams of beams with bolted friction

support: (a) before thermal effect; (b) after temperature increase
friction factors and slant angles on the bolted slanted end-plate

connection
8:
9:
10:
WL 2Pb sin
2 sin( )
Ptmax
WL 2Pb sin
cot( ) Pb cos
2

1
WL 2Pb sin
cot( ) Pb cos
T m
AE
2
2.2.1 Slanted end-plate connection with frictionless

support
To establish frictionless behaviour, a roller support is considered
on the slanted end-plate in the steel beam under non-symmetric
gravity load, as shown in Figure 7(a). Equation 11 shows the
behaviour of this kind of support under a non-symmetric gravity
load in a frictionless condition with various slant angles. By
applying equilibrium, the amount of axial force in the beam
before and after thermal effect can be derived by Equation 11,
which implies that the roller supports behaviour of the beam due
to elevated temperature by crawling on the slant plane. For a nonsymmetric type of loading, according to Figure 7(a), the relation
between the force and the slant of the connection in the
frictionless condition is
11:
2.2 Beam subjected to non-symmetric gravity load
To simulate the same conditions (loading and section properties
of the steel beam with slanted end-plate connection) for both
cases (symmetric and non-symmetric gravity load), a non-symmetric gravity load (2W (L=2) WL Q) is considered.
Pi
WL
cot
2
(frictionless)
2.2.2 Slanted end-plate connection with friction support

To demonstrate the real condition of the connection, it is
supposed that there is a friction factor between the two faces of
7

Roller
2W

Roller
14:
RL
WL cos( )
sin 2
15:
RR
WL cos( )
sin 2
16:
Pi
L
L/2
L/2
(a)
Q WL
Pi
Pi
(b)
Friction bolts
05Pb Q
WL
05Pb
05Pb
WL(cos2 sin2 )
sin 2
05Pb
Pi
When the beam is subjected to a temperature increase, it tends to

slide upward, in left and right-hand sides supports similarly, to
damp the induced axial load due to expansion. So, the friction
component of the left-side connection resists against upward
sliding but, in the right-hand side, the connection acts similar to
the case before the thermal effect. Referring to Figure 8, the
reactions and elevated temperatures are
Pi
(c)
17:
RL RR
Figure 7. Simplified model of beam with non-symmetric gravity

load: (a) frictionless support; (b) friction support; (c) friction bolted
support
the joint plane. In this case, the beam is in static equilibrium

before any temperature increase (Figure 7(b)) to satisfy stability
through the whole length of the beam under the non-symmetric
gravity load Q WL. Indeed, the non-symmetric gravity load
causes the left-hand side support to move downward and, in
contrast, move the right-hand side support upward. According to
the nature of friction, it resists the direction of sliding. The initial
axial force Pi at the onset of sliding and without any temperature
influence is
12:
Pi
(aWL)(cos s sin )
(aWL) cot( )
sin s cos
In the equilibrium condition before sliding, the participation rate

of the gravity load on the support a is
18:
Ptmax
WL
cot( ) friction form
2
19:
T m
WL
cot( ) friction form
2AE
2.2.3 Slanted end-plate connection with friction bolt

support
In order to assess the mechanical behaviour of a beam with a
bolted slanted end-plate, it is supposed that the reactions of
supports depend on the friction factor and the compressive
induced forces exerted by the bolts normal to slant plane. In this
case, the beam is in a balanced condition before any thermal
effect (Figure 7(c)). Furthermore, the non-symmetric gravity load
Q WL causes the left and right-hand sides to move in opposition. Before the thermal effect
20:
13:
a<
cot( )
cot( ) cot( )
When the beam moves, the reaction of the two beam ends can be
derived by Equations 14 and 15 and the amount of initial axial
force Pi in such a condition is shown by Equation 16
8
WL
2 sin( )
N L (aQ) sin Pi cos Pb
21:
F inclinedlineleft 0
Pi (aWL) cot( )
Pb sin
sin( )


Q WL
(1 a)Q
(a)Q
(a)Q
90
FfL
90
Pi
Pi
NL
(a)Q 90
(1 a)Q
FfR
NR
90
(1 a)Q
Pi
Pi
s tan
RR
RL
Figure 8. Free body diagram of beam with non-symmetric gravity

load (friction support) before temperature increase
22:
N R (1 a)Q sin Pi cos Pb
F inclinedlineright 0
Pb sin
Pi (aWL) cot( )
sin( )
23:
24:
Pi
(WL 2Pb sin )(cos2 sin2 )

Pb cos
sin 2
After increase in temperature, the reactions in both sides and the

axial force in the beam are defined as
25:
26:
RL RR
Ptmax
WL 2Pb sin
2 sin( )
WL 2Pb sin
cot( ) Pb cos
2
Equation 27 shows the temperature needed for sliding in the case

of friction bolt support
T m
27:
1
AE

WL 2Pb sin
3
cot( ) Pb cos
2
3.
Analytical comparison of effect of

symmetric and non-symmetric gravity
load on slanted end-plate connection
responses
3.1 Frictionless support

Referring to Figures 3(a) and 5(a) and Equations 1 and 11, in
both conditions (symmetric and non-symmetric gravity load), the
amount of axial internal force in the beam, Pi, is the same. As
there is no friction effect, the induced axial load in the beam only
depends on the slant angle : Also, in both cases, by increasing
the angle of slant in the end-plate connection, the amount of Pi
will be decreased, which is governed by cot : This means that, in
this specific position, the distribution of loading does not have a
manifest effect on the amount of induced axial force in the beam.
Equally, in both cases, differences between the two types of
loading do not affect the manner of the sliding of the beam
member.
3.2 Friction support
To compare the action of the two different loads on a slanted
end-plate connection, in the case of friction support and in the
absence of a thermal effect, by considering a steel beam under
symmetric gravity load and referring to Equation 2, it is evident
that, by increasing the amount of friction, the value of Pi in the
beam will reduce due to the effect of the term cot( ): Also,
by enhancing the slant angle, (for a constant friction factor), the
internal axial force will decrease. By applying different inclination angles between 08 (conventional support) and 608 (practical
angle), it can be seen that the beam can slide with lower axial
force for a higher slant angle. Indeed, in this specific position, the
angle of slant and the friction factor act in reverse.
Referring to Equation 3, the term cot( ) represents the
behaviour of such a connection after a thermal effect.
9


Obviously, after the beam temperature increases, for constant

slope, the amount of Pt will increase by increasing the amount
of friction. In the case of non-symmetric loading and with the
existence of friction, by applying Equations 12 and 16, there
are two actions before and after sliding which indicates that
the load distribution can have a significant effect on the value
of Pi. Equation 12 confirms that, by increasing the slope and
friction (term ( )), the amount of axial force will reduce.
However, when the load increases, a larger force is required to
satisfy the balance condition in the connection. For constant
inclination, in the right-hand side, the same result can be
obtained. After a thermal effect, with respect to changes in
slant angle and friction factor and referring to Equation 16, the
term cot( ) represents the behaviour of such a connection.
Hence increasing the friction results in enhancement of the
maximum force Ptmax.
3.3 Friction bolt supports

By means of friction bolts (Figure 9), the behaviour can be
divided into four types for symmetric and non-symmetric types
of load and before and after applying temperature.
In another sense, friction directly affects the resistance against

upward movement of the beam. Comparison of Equations 5
and 19, which represent the temperature needed to start sliding
in symmetric and non-symmetric situations, respectively, leads
to the same result. When a frictional component does exist on
the surface, whether the gravity load is symmetric or nonsymmetric, it tends to move the beam downward. Equations 5
and 19 indicate that the friction component can act to resist
sliding, which means that more temperature should be applied
to have the desired action of the slanted end-plate connection
under such a condition. Furthermore, inspection of Equations 5
and 19 shows that the movement temperature Tm has the
same value in both cases and it varies by changing the slant
angle and friction factor. Indeed, symmetric and non-symmetric
types of loading have no effect on the time of sliding. This
flexible behaviour is another advantage of the slanted end-plate
connection.
Firstly, symmetric and non-symmetric loading of the connection

before application of temperature will be assessed. Equation 7
shows that bolts directly affect the amount of induced force Pi.
On the other hand, referring to Equations 21 and 24, for a nonsymmetric gravity load and before temperature increase, the
amount of induced axial force depends on the distribution of
load. Additionally, when the beam is subjected to a temperature
increase (for both types of loading), according to Equations 9 and
26, the beam reacts equally (similar to the case for friction
supports). Likewise, when the beam is under non-symmetric
gravity load, by applying the same W, , and Pb, the required
temperature (Tm) is the same as for symmetric loading (Equations 10 and 27). This means that, with regard to the axial force,
there is no significant difference between the two types of loading
after a temperature increase. Also, a beam with a bolted
connection behaves similarly to a beam with friction support,
except when there is additional force due to bolts, which
increases the resistance against movement. A higher temperature
is thus required to initiate sliding of a steel beam in a bolted
connection.
4.
Illustration
To demonstrate the different behaviours of the beam under the

two different types of loading (symmetric and non-symmetric),
applying numerical comparisons can be beneficial. Two steel
sections (IPE 300) of a beam with slanted fixed end-plate support
under symmetric and non-symmetric gravity loading are thus
compared. The properties of the sections were: A 5380 mm2,
E 210 kN/mm2, ambient temperature T0 208C, elevated tem-
Q WL
(a)Q
(a)Q
Pb
90
FfL
Pb
RL
Pb
Pt
s tan
Figure 9. Free body diagram of beam with non-symmetric gravity

load (friction bolted support) after temperature increase
10
Pt
NR
90
(1 a)Q
FfR
Pt
NL
90
Pt
(a)Q 90
(1 a)Q
Pb
(1 a)Q
RR


perature T 508C, length of beam L 6000 mm and coefficient of thermal expansion 1 .5 3 105/8C. The linear gravity
load for half-length of the span for the non-symmetric condition
was W 40 kN/m (Q 40 3 3 120 kN) and the uniformly
distributed symmetric gravity load for the symmetric condition
was W 20 kN/m, distributed over the entire span. The axial
applied load on the friction bolts Pb 50 kN. The slant angle
of the end-plate connection varied from 08 to 608 and the friction
coefficient factor s tan 0 .2, 0 .3 and 0 .5 (Zahmatkesh and
Talebi, 2010).
Referring to Figure 10, at T 0, there is a non-zero value for

the internal axial force, which proves that, before the thermal
effect, there is an axial force due to gravity load in case of
slanted end-plate connection. For the conventional connection,
there is no force due to gravity load at the same point. In other
words, the internal axial force in the steel beam due to the gravity
load will be gradually substituted by the axial force due to
heating.
As discussed earlier, when a steel beam with fixed conventional

supports is restrained, because of the elongation limit an internal
axial force is created in the steel beam, which increases with an
increase in temperature (Equation 4). Indeed, in the case of a
conventional connection, because there is no resistance force in
the X direction (idealised position, frictionless) the gravity load
does not cause an axial force through the X axis of the beam. In
this case, the axial force is directly proportional to temperature,
and it starts from zero until the steel beam collapses due to
buckling or yielding.
In the case of a slanted end-plate connection, the behaviour can
be more complicated. To gain better understanding, a beam with
friction support s 0 .2 and slant angle 458 under symmetric gravity loading is compared with a beam with a conventional connection in Figure 10. The graph shows comparisons
between two connections in terms of elevated temperature and
induced axial force. The salient points are as follows.
Point A11, when T 0 and there is only an internal axial
force due to the gravity load and it has a straight line.
j Point B11 is the equilibrium point before the thermal effect.
j Point C11 is called the sliding point, which shows the
equilibrium point just before sliding.
j Point D11 is the point after sliding (note that, in the graphs,
end points are not yield or buckling points of the members).
j
350
Axial force, Pi
300
250
s 02
11 45
21 55
Conventional connection
31 30
41 60
200
C31
400
350
300
D31
150
C11
100
B31 C21
A31
B11
50
A11
A21
B21
0
0
5
Referring to Figure 11, under the conditions s 0 .3 and

458, the magnitude of the internal axial force due to gravity
load in the steel beam is 32 kN (point A12). Whenever the steel
beam is subjected to heat, the same magnitude is required to
overcome this force. So, the graph has a straight line in this
portion and the amount of Pi is constant (point B12). With
continued heating, the axial force due to gravity load will be
completely substituted. The straight line with constant slope
Axial force, Pi
400
The substitution will continue until the induced axial force (due
to temperature increase) completely neutralises the gravity axial
force. Once the axial force due to the gravity load is eliminated,
movement of the steel beam begins, so a specific force will be
required to overcome the friction force between two surfaces,
which acts opposite to the sliding direction. By increasing the
temperature continuously, the friction force component will be
reduced until it reaches zero. Once the induced axial force
reaches zero, damping will occur and the steel beam will move
up (point C). After point C, the line is straight, which indicates
that, after sliding, by increasing the temperature, there will not be
an axial force in the beam because the beam can freely expand.
This can be assumed as an advantage of a slanted end-plate
connection over a conventional connection. To gain better understanding of the different behaviours of such connections in the
case of friction connection and under symmetric gravity load,
Figures 1012 compare three different friction factors (s 0 .2,
0 .3 and 0 .5) with five different slant angles ( 308, 408, 458,
558 and 608).
D11
D21
10
15
Elevated temperature: C
20
25
Figure 10. Axial forcetemperature diagram of slanted end-plate

connections for friction factor s 0 .2
250
s 03
12 45
22 55
Conventional connection
32 30
C32
42 60
D32
200
150
100
A32
50
A12 B12
A22
B22
0
0
C12
C22
D12
D22
10
15
20
25

11
Axial force, Pi: kN


s 05
1500
1400
13 30
1300
23 40
1200
Conventional connection C13
1100
1000
D13
33 60
900
800
700
600
500
400
300
D23
200 B33 C23
B13
100
B23 D33
A230
0
10
20
30
40
50
60
70
80
and 0 .5 are 124 .6 kN, 190 .22 kN and 907 .74 kN. With these
results it can be predicted that, by using higher friction factors,
the slanted end-plate connection plays a more effective role in
damping. Furthermore, the slant angle acts better for a higher
friction factor.
The difference between the second and third points of the graphs
regarding the effect of the friction factor can be calculated (Table
1). For 0 .2, 0 .3 and 0 .5, the values are sorted in ascending
s
90

(between points B12 and C12) represents the amount of force

needed to eliminate the friction component on the surface. At
point C12, the required force for sliding is 111 kN. The required
force to start damping is 111 kN and 76 kN for 458 and 558,
respectively. These results clearly indicate that, by increasing the
slope of the connection, less heat is required to move the beam.
Figures 1012 also imply that all four lines with different slant
angles have the same rate of change between points B and C
(before and after sliding).
The graph between points B and C has a linear shape with
constant slope. This part of the graph estimates the amount of
force required to overcome the friction component that acts to
oppose upward or downward movement depending on the case.
Comparison of the three friction factors (s 0 .2, 0 .3 and 0 .5)
for 458 shows that the slope of line does not change by
changing s :
Different friction factors (s 0 .2, 0 .3 and 0 .5) for slant angles
of 308 and 608 are also considered in Figures 1012. For 308,
with respect to the third points on the graphs, the required axial
force to incur damping is 177 .35 kN, 253 .93 kN and 998 .65 kN,
respectively.
For 608, considering the same friction factors, the required
axial force to start sliding is 52 .75 kN, 63 .71 kN and 90 .91 kN.
The differences between the two slant angles for s 0 .2, 0 .3
order. This clearly implies that, for a higher friction factor, a

higher force is required for resistance against the friction
component. The results in Table 1 reveal that the difference
between the second and third points is higher for 308 than for
608. It can thus be concluded that a lower temperature will
be needed to move the beam for a larger angle of connection.
This result was similar for all three friction factors considered.
Inspection of Figure 13 reveals that the same behaviour for roller
connections under symmetric and non-symmetric gravity loading
can be expected. The graph also implies that, before reaching the
axial force Pi 150 kN (the critical force), the steel beam can be
protected against buckling or yielding, whichever occurs earlier.
In other words, the practical range for slanted end-plate connections in roller conditions lies between 228 and the practical
slant angle of 608. Figure 13 shows that, for both types of
load, when the slant angle approaches zero (i.e. conventional
connection), the axial force is huge (infinite zone on graph). Once
the slant angle reaches 608, it will reduce.
Inspection of Figure 14 reveals that friction support under a
symmetric gravity load acts like a conventional (vertical) support,
and the maximum axial force in the beam is 120 kN. By applying
bolts to the connection, the axial force reduces to 60 kN. Bolts
thus play a significant role in reducing the axial force in the beam
before a temperature increase. After an increase in temperature,
the slope of the graph increases sharply, which implies that, for
slant angles of less than 358, a huge axial force is created (infinite
zone).
Referring to Figure 14 (s 0 .2) the amount of axial force Pi for
a non-symmetric gravity load before a temperature increase and
similar to the conventional condition ( 08) equals 450 kN.
Under the same conditions, the symmetric gravity loaded steel
beam has an axial force equal to 290 kN (in both cases the axial
s 0 .2
s 0 .3
s 0 .5
C31C41: kN
Slant
s 0 .2
angle, :
degrees Point B31 Point C31 Point B32 Point C32 Point B33 Point C33
30
60
68 .34
20 .35
177 .35
52 .75
56 .58
14 .22
253 .93
63 .71
39 .66
3 .85
Table 1. Comparison of axial forces for different friction factors

and different elevated temperatures
12
998 .65
90 .91
124 .6
C32C42: kN
s 0 .3
C13C33: kN
s 0 .5
190 .22
907 .74
Axial force in beamcolumn: kN

900
850
800
750
700
650
600
550
500
450
400
350
300
250
200
150
100
50
0

Pt A E T 84735 kN
Pi(Q)
Ptmax(Q T )
Pi(Q Pb)
Practical limitation
Ptmax(Q Pb T)
Py allowable
Pt
Py (fa /Fa) (fb /Fb) 1
12
17
22 27 32 37 42 47 52 57 62 67 72
Angle of inclination of slant connection: degrees
77
82
87
Figure 13. Comparison of symmetric and non-symmetric loading;

frictionless
700
650
600
550
500
450
400
350
300
250
200
150
100
50
0
Symmetric gravity load, Pi (W)

Symmetric gravity load, Ptmax (W T )
Symmetric gravity load, Pi (W Pb)
Symmetric gravity load, Ptmax (W Pb T )
Non-symmetric gravity load, Pi (W)
Non-symmetric gravity load, Ptmax (W T )
Non-symmetric gravity load, Pi (W Pb)
Non-symmetric gravity load, Ptmax (W Pb T )
10
15
20
25
30
35
40
45
50
55
60
Figure 14. Angle of slanted end-plate connection versus axial

force for friction factor s 0 .2 and symmetric and nonsymmetric loading
forces are due to gravity load only). Regarding the results, there
are significant differences in the behaviour of the left-hand side
and right-hand side slanted end-plate connections in terms of load
distribution, while to satisfy equilibrium the total amount of
axial force in the beam is the same for both cases. It is obvious
from Figure 14 that, by increasing the slant angle, differences
between the two types of loading are reduced. The behaviour of
the beam changes after a temperature increase. For the two
different types of loading, the amounts of induced axial forces
under various slant angles are the same (overlapping curves on
graph).
With the application of bolts, before a temperature increase and
for symmetric loading, Pi starts from 250 kN for 0 and
reduces to 20 kN for 608; in the non-symmetric case at the
same conditions, the axial force starts at 400 kN and decreases to

60 kN. These results imply that, in non-symmetric loading, the
left-hand side connection experiences larger axial force due to
the different distribution of load, and this should be considered in
the design of such connections. The same can be said for the
friction connection after a temperature increase both cases
behave similarly and in both cases the induced axial force starts
at 600 kN for 208. The effect of friction on the magnitude of
the axial force in the beam is significant. Figures 14, 15 and 16
show that, in the case of non-symmetric loading with friction
bolts, the bolts help to reduce the axial force before any temperature effect.
It can be predicted that, after a thermal effect in symmetric
loading, a bolted connection has a reverse effect on the magnitude
13


700
650
600
550
500
450
400
350
300
250
200
150
100
50
0

10
15
20
25
30
35
40
45
50
55
60

700
650
600
550
500
450
400
350
300
250
200
150
100
50
0

10
15
20
25
30
35
40
45
50
55
60

of the induced axial force, which is more significant for 608

with axial forces of 100 kN and 140 kN in frictional and bolted
connections, respectively. This means that, by applying bolts, the
amount of axial induced force will increase, which is not
desirable with regard to buckling or yielding of the steel beam.
However, the general behaviours of friction and bolted connections are similar. Figure 17 shows the relationship between
movement elevated temperature and slope angle of the slanted
connection. For s 0 and 108, Tm 208C for both types
of loading. Because of the absence of a friction component on the
slanted connection, symmetric and non-symmetric loadings have
the same configuration on the graph (overlapping curves).
14
Furthermore, for s 0 .2, the presence of frictional bolts does

not play a significant role. Indeed, after a temperature increase,
the bolts do not influence the damping of the slanted end-plate
connection. Figure 17 shows that the same movement elevated
temperature is required to start sliding in both symmetric and nonsymmetric connections. Inspection of Figure 17 also reveals that,
by increasing the friction factor, Tm will increase. So, by
applying weak contact between surfaces, damping can be incurred
at a lower temperature. Additionally, in the non-symmetric loading
case, after a temperature increase, the slope of the graph changes
sharply and the same behaviour as the reverse effect of bolts to
increase axial force in the steel beam under symmetric load can
Movement elevated temperature, Tm: C


100
Practical limitation
90
s 0
80
s 02
s 03
s 05
Maximum elastic moduleus of steel ST37 for
70
ambient temperature 20C
Tm (Q T ) s 02
Tm (Q T Pb) s 02
Tm (Q T ) s 03
Tm (Q T Pb) s 03
60
T 50C
50
Tm (Q T ) s 05
40
Tm (Q T Pb) s 05
30
Tm (Q T ) s 0
20
Tm (Q T Pb) s 0
10
0
0
10
15
20
25 30 35 40 45 50 55 60 65 70
75
80
85
90
Figure 17. Relation between movement elevated temperature

versus angle of slant connection for symmetric and nonsymmetric gravity loads; T 508C
be seen. So, by applying bolts, after a thermal effect, bolts cause

an increase in axial force, which can be dangerous with regard to
buckling. Comparison of the results of average ratio of axial force
for non-symmetric and symmetric loading confirms this. Before
applying bolts, the ratio was 1 .48, 1 .55 and 1 .57 for s 0 .2, 0 .3
and 0 .5, respectively; with the use of bolts the ratios are larger,
1 .64, 1 .72 and 1 .90 for s 0 .2, 0 .3 and 0 .5, respectively.
However, after a temperature increase the ratio is 1 for all cases.
5.
Conclusion
A comparison study of the mechanical behaviours of two different slanted end-plate connections under symmetric and nonsymmetric loading (Zahmatkesh and Talebi, 2010; Zahmatkesh et
al., 2014a, 2014b, 2014c) was carried out and the following
conclusions were deduced.
In the case of a non-symmetric gravity loaded slanted end-plate
connection, before elevated temperature there is a higher axial
force in the beam in comparison with a symmetric gravity loaded
slanted end-plate. The average ratio of axial force in nonsymmetric to symmetric cases was 1 .491 .57. These values
increased to 1 .641 .90 when friction bolts were applied. Before a
temperature increase and with no bolts, the average ratio of axial
force for non-symmetric and symmetric loading was 1 .48, 1 .55
and 1 .57 for s 0 .2, 0 .3 and 0 .5, respectively. After applying
bolts, the ratios were 1 .64, 1 .72 and 1 .90 for s 0 .2, 0 .3 and
0 .5, respectively. The ratio after the thermal effect was 1. In the
case of non-symmetric loading, the amount of axial force in the
steel beam before elevated temperature depends on the distribution of the gravity load.
The required temperatures to start sliding in both sides (left and
right) of the beam and consequently the time to start damping of
the induced axial force due to temperature in both loading cases
(symmetric and non-symmetric) were the same. Additionally, in

both cases, changing a conventional end connection to a slanted
end-plate type causes greater reduction in the magnitude of axial
force. Significant cost savings are associated with a reduction in
the weight of members.
By changing the angle of the slanted end-plate connection, an
optimum axial force (or optimum angle of slanted end-plate
connection) can be obtained for which the steel beam is in the
safe zone to avoid buckling. The safe zones of slanted end-plate
connections for friction factors of s 0 .2, 0 .3 and 0 .5 are slant
angles of 17608, 23608 and 33608, respectively.
Bolts have an antithetical behaviour in slanted end-plate connections. In other words, before elevated temperature, bolts help to
reduce the axial force in the beam but, after an increase in temperature, they can act to resist sliding of the beam, with consequently
less damping effect. Furthermore, the effect of bolts in reducing the
amount of axial force in a beam is larger for non-symmetric gravity
loading than for the symmetric gravity loading case.
REFERENCES
Dai XH, Wang YC and Bailey CG (2009) Effects of partial fire
protection on temperature developments in steel joints

protected by intumescent coating. Fire Safety Journal 44(3):
376386.
Daryan AS and Bahrampoor H (2009) Behavior of khorjini
connections in fire. Fire Safety Journal 44(4): 659664.
da Silva LS, Santiago A and Vila Real P (2001) A component
model for the behavior of steel joints at elevated
temperatures. Journal of Constructional Steel Research
57(11): 11691195.
Daz C, Victoria M, Mart P and Querin OM (2011) FE model of
15


beam-to-column extended end-plate joints. Journal of

Constructional Steel Research 67(10): 15781590.
ECS (European Committee for Standardization) (1995) ENV
1993-1-2: Design of steel structures, Part 1.2: General rules,
structural fire design. ECS, Brussels, Belgium.
Heidarpour A and Bradford MA (2009) Elastic behavior of panel
zone in steel moment resisting frames at elevated temperatures.
Journal of Constructional Steel Research 65(2): 489496.
Mourao HDR and Silva VPE (2007) On the behavior of singlespan steel beams under uniform heating. Journal of
the Brazilian Society of Mechanical Sciences and Engineering
XXIX(1): 115122.
Qian ZH, Tan KH and Burgess IW (2008) Behavior of steel beam-tocolumn joints at elevated temperature: experimental
investigation. Journal of Structural Engineering 134(5): 713
726.
Yang B and Tan KH (2012) Numerical analyses of steel beam
column joints subjected to catenary action. Journal of
Constructional Steel Research 70: 111.
Zahmatkesh F and Talebi E (2010) The performance of bolted
slant end plate connections subjected to temperature increase.

Proceedings of ASME 2010 10th Biennial Conference on
Engineering Systems Design and Analysis, Istanbul, Turkey,
vol. 2, pp. 207217.
Zahmatkesh FMO, Talebi E and Kueh ABH (2014a) Analytical
study of slant end-plate connection subjected to elevated
temperatures. Steel and Composite Structures 17(1):
4767.
Zahmatkesh F, Osman MH, Talebi E and Kueh AB (2014b) Direct
stiffness model of slant connection under thermal and nonsymmetric gravity load. Journal of Constructional Steel
Research 102: 2443.
Zahmatkesh F, Osman MH and Talebi E (2014c) Thermal
behaviour of beams with slant end-plate connection subjected
to nonsymmetric gravity load. Scientific World Journal 2014:
paper 323209, http://dx.doi.org/10.1155/2014/323206.
Zalosh RG (2003) Industrial Fire Protection Engineering. Wiley,
Chichester, UK.
WHAT DO YOU THINK?
To discuss this paper, please email up to 500 words to the

editor at journals@ice.org.uk. Your contribution will be
forwarded to the author(s) for a reply and, if considered
appropriate by the editorial panel, will be published as a
discussion in a future issue of the journal.
Proceedings journals rely entirely on contributions sent in
by civil engineering professionals, academics and students.
Papers should be 20005000 words long (briefing papers
should be 10002000 words long), with adequate illustrations and references. You can submit your paper online via
www.icevirtuallibrary.com/content/journals, where you
will also find detailed author guidelines.
16

Slanted Connection at Elevated Temperature Under Various Loadings

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Slanted Connection at Elevated Temperature Under Various Loadings

Caricato da

Copyright:

Formati disponibili

Engineering and Computational Mechanics

Volume 168 Issue EM1

Proceedings of the Institution of Civil Engineers