analysis of steel-concrete composite beam-to ... - CCVI Information

ANALYSIS OF STEEL-CONCRETE COMPOSITE BEAM-TO-COLUMN
JOINTS: BOLTED SOLUTIONS
Oreste S. Bursi
Department of Mechanical and Structural Engineering, University of Trento
Trento, Italy
oreste.bursi@ing.unitn.it
Fabio Ferrario
Department of Mechanical and Structural Engineering, University of Trento
Trento, Italy
fabio.ferrario@ing.unitn.it
Raffaele Pucinotti
Department of Mechanics and Materials, Mediterranean University of Reggio Calabria
Reggio Calabria, Italy
raffaele.pucinotti@unirc.it
Riccardo Zandonini
Department of Mechanical and Structural Engineering, University of Trento
Trento, Italy
riccardo.zandonini@ing.unitn.it
ABSTRACT
In this paper, a multi-objective advanced design methodology is proposed for steel-concrete
composite moment-resisting frames. The research activity mainly focused on the design of the
beam-to-column joints under seismic-induced fire loading together with the definition of
adequate structural details for composite columns. Thermal analyses of cross sections were
performed in order to obtain internal temperature distribution; structural analyses were then
performed on the whole frame to assess the global behavior under the combined action of static
and fire loadings. Besides, results of numerical analyses were used in order to derive
information about the mechanical and numerical behavior of joints. In this paper the
experimental program carried out on four beam-to-column joint specimens are described and
experimental results are presented and discussed together with the outcomes of numerical
simulations. Experimental tests demonstrated the adequacy of the seismic design. Numerical
simulations showed a satisfactory performance of joints under seismic-induced fire loading.
INTRODUCTION
Steel-concrete composite structures are becoming increasingly popular around the world due to
the favorable stiffness, strength and ductility performance of composite systems under seismic
loading, and also due to the speed and ease of erection. Moreover, such structures exhibit
better fire resistance characteristics compared to bare steel structures. Considering the high

probability of a fire after a seismic event, the use of steel–concrete composite structures in
seismic areas, potentially represents a fairly effective design solution. In fact, the adoption of
such a solution is in general more efficient from both a structural and a constructional viewpoint
when compared to bare steel structures. The benefit increases, if the high probability that an
earthquake and a fire can occur in sequence. In current Eurocodes for structural design,
seismic and fire safety are accounted for separately, and no sequence of seismic and fire
loading is taken into account. In reality, the risk of loss of life increases if a fire occurs in a
building after an earthquake. In the Kobe Earthquake (1995) many people died due to the
collapse of buildings exposed to a fire that followed an earthquake; in fact, large sections of the
city burned, greatly contributing to the loss of lives. It is obvious therefore that seismic-induced
fire is a design scenario that should be properly addressed in any performance-based seismic
design. In this situation the traditional single-objective design is not adequate, and a multiobjective
advanced design has to be adopted. This enables a proper account of: (i) seismic
safety with regard to accidental actions; (ii) fire safety with regard to accidental actions (iii) fire
safety on a structure characterized by stiffness deterioration and strength degradation due to
seismic actions. As a result, the fire design applied to a structure with reduced capacity due to
seismic damage will permit simultaneous fulfillment of the requirements associated with
structural, seismic and fire safety ,and with structural fire safety and structural seismic safety
separately taken as accidental actions. In order to characterize the seismic behaviour of
composite joints, a research project has been carried out by the University of Trento with the
objective of developing a design procedure for composite joints under the ‘combined’ action of
earthquake and post-earthquake fire. The research activity was mainly concerned with the
design of bolted beam-to-column joints with CFT columns with circular hollow steel sections.
This solution derives from a parent welded design solution developed in an European project
(Bursi et al. 2006) and is aimed at ensuring easiness of assembly and avoiding the problems
related with on site welding. The results of the experimental programme devoted to the
evaluation of the cyclic behavior of beam-to-column bolted joints are reported in the following. In
order to better understand the activation of all the transfer mechanisms proposed in the
Eurocode 8 (UNI EN 1998-1, 2005), a numerical finite element (FE) model of the slab has been
developed, together with FE model of the joint under fire.
SEISMIC AND FIRE DESIGN OF REFERENCE FRAMES
A reference composite steel-concrete office-building was considered, made up of three five
storeys moment resisting frames, placed at the distance of 7.5 m each in the longitudinal
direction and braced in the transverse direction. The storey height was equal to 3.5 m. Two
moment resisting frames, having the same structural typology but different slab systems, were
analyzed: i) a composite steel-concrete slab with structural profiled ribbed steel sheeting; ii) a
concrete slab composed of electro-welded lattice girders.
Since the two slab systems had different load bearing capacities, a different distance between
the secondary beams was adopted for the two solutions. The distance slightly affected the
frame geometry. The first solution (steel sheeting slab), is depicted in Figure 1a while the
second one, with the slab on prefabricated lattice elements, is shown in Figure 1b. The
elevation is shown in Figure 1c (Bursi et. al 2008). Two different types of composite beams were
hence designed according to Eurocode 4 (UNI EN 1994-1-1, 2005); the steel section was
maintained in both cases an IPE400 of steel grade S355. The slab, 150 mm deep, was
designed in accordance with the relevant specifications of Section 9 of Eurocode; moreover
design procedure indicated by the producer of the prefabricated lattice elements was followed in
the second case. All connections between the steel beam and the slab were made by Nelson
19 mm stud connectors made of steel with an ultimate tensile strength f u =450 MPa.

a) b) c)
Figure 1 - The configuration of the building and the longitudinal five-storey MR Frame; slab with
a) profiled ribbed steel sheeting; b) slab with prefabricated lattice girder; c) elevation
The frames were designed to have a dissipative structural behavior according to Eurocode 8
(UNI EN 1998-1, 2005). In detail two following structural types were considered.
1. Moment Resisting Frames: in which dissipative zones were mainly located to the end of the
beams, in the vicinity of beam-to-column connections, where plastic hinges could develop
(UNI EN 1998-1, 2005, 7.1.2 b);
2. Concentrically braced frames: in which the horizontal forces were mainly resisted by
members subjected to axial forces with dissipative zones located on the tension diagonals
only (UNI EN 1998-1, 2005, 7.1.2 c).
The effective width of the slab in composite beams was determined according to Eurocode 4
(UNI EN 1994-1-1, 2005, 5.4.1.2) for static and fire analyses, and to Eurocode 8 (UNI EN 1998-
1, 2005, 7.6.3) for seismic analyses. Beam-to-column connections resulted to be rigid according
to Eurocode criterion (UNI EN 1994-1-1, 5.1.2). The design of structural elements were
performed considering static and seismic design situations. Moreover, numerical simulations on
two-dimensional (2D) frames, were first performed by means of the SAFIR program (Franssen
J.-M. 2000), in order to study different fire scenarios acting in the reference buildings and to
evaluate the performance of different elements, i.e. composite beams, composite columns, and
beam-to-column joints, under fire load for different times of exposure to fire. Five fire scenarios
were considered. In the first one (FS1), fire acted only into a span in the first floor as illustrated
in Figure 3a.
a) FS1 b) FS2 c) FS3 d) FS4 e) FS5
Figure 3 - Fire scenarios considered in thermal analysis
In the second one (FS2), fire acted on the whole first floor; this means that the first level
columns and the first level beams were heated. In the third one (FS3), fire acted only into a
span in the last floor. In the fourth one (FS4), fire acted on the whole fifth floor. Finally, in the
fifth one (FS5) fire acted on the whole frame. The fire load followed the ISO 834 fire law. Figure
4 shows the evolution of the bending moment and of the axial force as a function of time at
various locations in the FE model for the case FS1. In detail the bending moment in the column

C shows an inversion in the sign when axial forces in the beams changes sign too. In fact, the
increase of the temperature cause first an increment of axial load in the beam (compression)
until to about 18 min owing the presence of the column restraint. Afterwards, the reduction of
stiffness of columns subject to fire prevails and the axial force in the beam turns to tension,
being similar to a “catenary” structure (with large displacement and deflection). The elongation
of the beam due to increase in temperature and the different “restraint” effect offered by the
column also causes a reversal of the sign of bending moment at mid-span of the beam which
from sagging becomes hogging and then sagging again. The results of the analysis shows that
the frame collapses because of the formation of a beam mechanism in the longest span
involving the formation of three plastic hinges located at mid-span and at both beam ends near
supports. Figure 5 shows column temperature distributions in the fire case FS1; it is possible to
notice that the steel tubular section offers a practically negligible protection to internal concrete;
therefore temperatures of steel elements directly exposed to fire are equal to that of the external
atmosphere.
1000.00
Column C - first floor
500.00
M (kNm)
0.00
0 5 10 15 20 25 30
-500.00
Axial Load (kN)
600
400
200
-400
-600
-800
151 155 160 161 165 170
A
B
Beams A & B - first Floor
110
C
101
0
-200 0 5 10 15 20 25 30
compressione
Time (min)
elem. 165
elem. 155
M (kNm)
-1000.00
-1500.00
400.00
0.00
-400.00
-800.00
Time (min)
Beam B - first floor
elem. 101
elem. 110
0 5 10 15 20 25 30
Time (min)
elem. 161
elem. 165
elem. 170
Figure 4 – Fire Case FS1: Bending Moment and Axial Load in beams and columns of the first
storey
10’ 30’ 40’
Figure 5 - Temperature distributions inside the column at different time. Fire Case FS1
The thermal distribution, for the composite slab with steel sheeting, evaluated after 10 min, 30
min and 40 min of ISO fire, by means of the thermal FE model created by SAFIR are presented
in Figure 6, while Figure 7 shows the thermal distribution for the case of composite slab with
prefabricated slab. As expected, the temperature increment was higher in the composite slab
with steel sheeting than in the slab with prefabricated R.C. elements. As a result, the moment
resistance of the former slab was lower than in the latter owing to the higher reduction in
materials strength.

10’ 30’ 40’
Figure 6 – Fire case FS1. Evolution of the temperature distribution inside beam B with a steel
sheeting composite slab.
10’ 30’ 40’
Figure 7 – Fire case FS1. Evolution of temperature distribution inside beam B with prefabricated
elements
BEAM-TO-COLUMN JOINTS DESIGN
The joint design aimed at ensuring to the joint between the column and the beam an
overstrength with respect to the beam. The proposed bolted solution derives from a parent
welded design (Bursi et. al 2008), and was conceived to guarantee easiness of assembly and to
avoid problems related to welding on site. The joint is comprised of two horizontal diaphragm
plates and a vertical through-column plate attached on the tube by groove welds as indicated in
Figure 8.
2x HORIZONTAL PLATES
VERTICAL PLATE
23°
23°
23°
23°
23°
16 168 16
= 320 x 1094 x 14
1094
36 75 75 75 572 75 75 75 36
2,3
18
18
14
R228,5
18
= 660 x 1304 x 16
124 36
Column
1304
320
124
660
200 230
502 300 502
= 320 x 1094 x 14
66 66 66 66 6636
46 108 46
36
Section A-A
1304
300
A
A
A
a=8 L=320
a=8 L=320
A
230
= 660 x 1304 x 16
660
14
R228,5
4,3
18
166
17
17
a=8 L=1475
a=8 L=320 a=8 L=320
Figure 8 - Steel-concrete composite beam-to-column joint: a) horizontal and vertical plates; b)
Nelson studs in columns
The flanges and the web of each beam were connected to the horizontal plates and the vertical
plate by two and three rows of bolts M27 10.9, respectively. A part the slab type, joints differed
owing to the presence of strengthening plates with holes required by Eurocode 8 to avoid brittle
fracture in net sections of bolted joints. Thus specimens JB-P1 and JB-S1, see table 1, where
endowed with extra plates; whereas JB-P2 and JB-S2 had no plates being beam-to-column
joints not dissipative. Beam-to-column joint design was developed using the component
method. The joint was simulated by a series of different components in agreement with
Eurocode 3 par 1-8 and Eurocode 8 (UNI EN 1993-1-8, 2005, UNI EN 1998-1, 2005), achieving
the necessary overstrength to the joint in respect to the connected composite beams.

Stiffness and strength of complex components, like top and bottom plates or concrete slab in
compression, were defined by means of refined Finite Element (FE) models of the joint including
friction between the slab and the column was developed. Depending on the level of friction, the
distribution of the compression forces in the slab under sagging bending moment was found to
be different: localized in front of the column for a friction coefficient equal to 0,35 as indicated in
Figure 9a; while it spreads over a more extended portion of the slab for increasing values of the
friction coefficient. In detail, the diffusion angle becomes greater than 80 degrees for a friction
coefficient of about 1. The results show that, in order to activate the transfer mechanisms
proposed in the Eurocode 8 (UNI EN 1998-1, 2005), i.e. Mechanism 1 front mechanism and
Mechanism 2 strut and tie mechanism, it is necessary to increase the level of friction between
the concrete slab and the composite column. As a result, Nelson 19 mm stud connectors
welded around the column were used in the tested specimens as indicated in Figure 8. As a
results beam-to-column joints were rigid full-strength joints satisfying the relation:
M
jRd ,
≥ s⋅γ
ov⋅ M
bplRd , ,
(1)
in which M j,Rd is the resisting moments of full-strength beam-to-column joints and M b,pl,Rd is the
resisting moment of the composite beam (UNI EN 1998-1, 2005). Moreover, the ductile
behaviour of joints was guaranteed by the following relationships:
≥ s ⋅γ R
(2)
R
d , bolt oν
⋅
pl,
Rd , beam
R
d bolt
Rpl,
Rd , plates
F
v Rd
Fb
, Rd
,
≥ (3)
,
≥ (4)
2
40°
80
1
40°
40 80°
2
a) b)
Figure 9 - Distribution of compression stresses in the slab for: (a) friction coefficient equal to
0,35; b) friction coefficient equal to 1
s = , γ o
= 1. 1, is the design strength of bolts, is the
where min{ f ;1.25
t
f
y
}
plastic design strength of the beam,
F
v , Rd
and
F b,
Rd
M 2
M 0
ν
R
d , bolt
R
pl,
Rd , plates
R pl , Rd , beam
is the plastic design strength of the plates, while
are the design shear strength and design bearing resistance of bolts,
respectively. The following conditions were also fulfilled for joints JB-P1 and JB-S1 to avoid
brittle fracture, i.e.:
0.90Anet
f Af
u y 0.90Anet
f Af
f
u
y
≥ and ≥ (5)
γ γ γ γ
where
A f
M 2
is the area of the tension flange. Nevertheless, being joints not dissipative, JB-P2 an
JB-S2 did not. A 3D finite model of the interior joint was implemented in the Abaqus 6.4.1 code
(Hibbitt et al. 2000). Then, fire design of beam-to-column joints subjected to fire load were
conducted. In particular, the component approach, exclusively applied to the moment-rotationtemperature
behaviour, in the absence of axial thrust owing to the thermal expansion restraint of
the beam, was adopted, and a simplified model was derived, which can predict the moment-
M 0

otation-temperature characteristic of the joint. As a results, at high temperatures, the joint were
designed to transfer shear forces owing to vertical loads from a beam to the other. Accordingly,
vertical through column plate, top horizontal plate together with Nelson stud connectors welded
around the column were arranged. In addition, two longitudinal steel rebars were added to
reduce the damage produced by the seismic actions before fire.
EXPERIMENTAL TESTS
The experimental programme consisted of 4 tests under cyclic loading of full-scale
substructures representing interior full-strength bolted beam-to-column joint (Figure 10).
Experimental tests were carried out at the Laboratory for Materials and Structures of the
University of Trento. Joint specimens were subjected to cyclic loadings up to collapse,
according to the ECCS stepwise increasing amplitude loading protocol, modified with the SAC
procedure (ECCS 1986, SAC 1997) by using e y =0.005h=17.5 mm where h represents the
storey height.
280
30
2850
2320
2290
5700
5140
500
2850
2320
280
5700
2850 2850
280 5140 280
30
2320 500 2320
2290
550
( 310 + 241.3 )
1200 1750
1470
A
A
IPE400
1344
CFT Ø 457 x 12
40 1230
IPE400
2590
2670
a)
b)
Figure 10: Bolted beam-to-column specimens a) slab with electro-welded lattice girders, b)
composite slab with profiled steel sheeting
The slab reinforcement, in the composite steel-concrete beams with steel sheeting, consisted of
3+3φ12 longitudinal steel rebars in order to carry the sagging moment, and of 4+4φ12@100 mm
and 7+7φ16@250 mm transverse steel rebars, in order to enable development of the seismic
slab-to-column transfer mechanism as well as the resistance to the shear force. A mesh
φ6@200x200 mm is also present as highlighted in Figure 11a. The concrete class was C30/37
while the steel grade S450 was adopted for reinforcing steel bars. In the case of the concrete
slab prefabricated R.C. elements, the slab reinforcement was made up of 3+3φ12 longitudinal
steel rebars and by 5+5φ12@100 mm plus 8+8φ16@200 mm transverse steel rebars. The same
mesh was adopted as for the composite slab (Figure 11b).
mesh Ø 6 / 200x200
160 300150 1500
pos.A 4 Ø 12 / 100
2000
pos.C 7 Ø 16 / 250
pos.B 3+3 Ø 12
560
720
880
mesh
HD 620
3Ø12
3Ø12
a)
b)
Figure 11 – Rehears layout in: a) steel sheeting slab; b) prefabricate lattice girder slab
1750 1750
1470
1200 550
80
( 310 + 241.3 )
mesh Ø 6 / 20x200
A
A
IPE400
pos.A 5 Ø 12 / 100
160 400 150 1400
2000
1344
CFT Ø 457 x 12
pos.C 8 Ø 16 / 200
pos.B 3+3 Ø 12
560
720
880
1230
40
HD 10/14/8 h=9,5cm
1Ø8
pos.E
mesh
HD 620
3Ø12 pos.B
IPE400
1Ø10 1Ø10
pos.H pos.D
1Ø10
pos.H
1Ø10
pos.D
2590
2670

The columns were concrete-filled columns with a circular hollow steel section with a diameter of
457 mm and a thickness of 12 mm. Steel grade is S355. The column reinforcement consisted of
8φ16 longitudinal steel rebars and of stirrups φ8@150 mm (Figure 12). The concrete class was
C30/37, while the steel grade S450 was adopted for the reinforcing steel bars. Actual values
better than nominal ones can be found in (Bursi et al. 2008). A scheme of the test set-up is
shown in Figure 13.
325
135
40
50
120 150
150 150
pos.E Ø 8 / 15 cm
2490
pos.D 8 Ø 16
90
95
2670
Figure 12 - Column and column reinforcement
L
L
4
I
L
L
I I
L L L I I
INCLINOMETERS
LVDTs
INFERIOR PLATE
SG
SG
SG
SG
SG
O
L
SG L L L
O SG
L SG
O O L
SG SG LL SGL
O
STRAING GAUGE
OMEGA STRAING GAUGE
LVDTs
SG
SUPERIOR PLATE
BEAM
SG
SG
BEAM
Figure 13 - Lateral view of the Test set-up and main instrumentation on specimens
No vertical actuator was imposed on the column to be consistent with tests in other labs. In the
tests, the following instrumentation were employed as illustrated in Figure 13: a) 5 inclinometers
measured the inclinations: of the column at the top and, in the zone adjacent to the joint, and of
the beams near the connection; b) 4 LVDTs detected the interface slip between the steel beam
and the concrete slab and between the bottom horizontal joint plate and the flange of beam; c) 2
LVDTs were employed in order to measure the bottom horizontal joint plate deformations; d) 10
LVDTs were utilized in order to measure concrete slab deformations in the zone around the
column; e) 4 Ω strain gauges detected the deformations of the concrete slab; e) 8 strain gauges
monitored axial deformations of the reinforcing bars to scrutinise the effective breadth of the
reinforcing bars at each loading stage; f) 4 strain gauges monitored deformations of top and
bottom horizontal plates; g) 4 strain gauges recorded flange strains in order to estimate internal
forces in steel beams; h) 2 load cell were set on the top of pendula and were utilized in order to
measure the horizontal and vertical components of the forces; i) 1 digital transducer
(Heidenhein DT500) in order to measure the top column displacement.
RESULTS AND SIMULATIONS
The observed response clearly indicated that all specimens exhibit a good performance in terms
of resistance, stiffness, energy dissipation and local ductility. Both the overall forcedisplacement
relationship and the moment-rotation relationships relevant to plastic hinges
formed in composite beams exhibited a hysteretic behaviour with large energy dissipation
without evident loss of resistance and stiffness as indicated in Figures 14-16. Hysteretic loops of
moment-rotation relationship are unsymmetrical owing the different flexural resistance of the
composite beam under hogging and sagging moments as can be observed in Figure 16. It was
evident that owing to local buckling effects, severe strength degradations appeared to exceed
SLAB
SG
SG

20 per cent of maximum strength values with rotations of about 40 mrad. The collapse of all
specimens was due to fracture of the bottom beam flange near the horizontal plate as illustrated
in Figure 17.
M [kNm]
Force [kN]
Specimen JB-P1
1000
750
500
250
0
-14 -10 -6 -250
2 6 10 14
-500
-750
-1000
Displacement [e/e y ]
Force [kN]
Specimen JB-P2
1000
750
500
250
0
-14 -10 -6
-250
-2 2 6 10 14
-500
-750
-1000
Displacement [e/e y]
Figure 14 - Specimens JB-P1 and JB-P2 – Force-Displacement relationship
Force[kN]
Specimen JB-S1
1000
750
500
250
0
-14 -10 -6
-250
-2 2 6 10 14
-500
-750
-1000
Displacement [e/e y ]
Force [kN]
Specimen JB-S2
1000
750
500
250
0
-14 -10 -6 -250
2 6 10 14
-500
-750
-1000
Displacement [e/e y ]
Figure 15 - Specimens JB-S1 and JB-S2 - Force-Displacement relationship
Specimen JB-S1
1000
750
500
250
0
-80 -60 -40 -20
-250
0 20 40 60 80
-500
-750
-1000
φ [mrad]
M (plastic hinge at dx)
M (plastic hinge at sx)
Figure 16 - Moment-rotation relationship of the
plastic hinges for the JB-S1
Figure 17 – Specimen JB-S1, plastic hinge in
the composite beam
Table 1 reports some key experimental data for each test: maximum applied displacement d,
maximum value F of the force, maximum values of both sagging and hogging moment M, total
number N tot of cycles performed during the tests. Joints behaviour is essentially symmetric in
terms of force, while, for all specimens, sagging moments are about 40% larger of hogging
moments. Figure 18 shows a comparison between Force-Interstorey drift curves for the
specimens with and without plate strengthening of the horizontal diaphragms in both
prefabricated slab and steel sheeting slab. These reinforcing plates were introduced in order to
satisfy the relationships in (5). In this case, in which dissipative zones were located to the end of
the beams, beam-to-column joints were rigid full-strength and the introduced reinforcing plates
did not influence the specimen behaviour.
Moreover, beam-to-column joints subjected to fire load were analyzed. In particular, the
component approach, exclusively applied to the moment-rotation-temperature behaviour, in the
absence of axial thrust owing to the thermal expansion restraint of the beam, was adopted, and
a simplified model was derived, which can predict the moment-rotation-temperature
characteristic of the joint. The method is capable of predicting the variation of failure modes
owing to change in the joint geometrical and material properties, as well as loading conditions.

Force [kN]
BJ-P1
BJ-P2
JB-P1 vs. JB-P2
800
400
0
-8 -6 -4 -2 0 2 4 6 8
-400
Force [kN]
BJ-S1
BJ-S2
JB-S1 vs. JB-S2
800
400
0
-8 -6 -4 -2 0 2 4 6 8
-400
Name
Test
Method
-800
Inter-Storey Drift [%]
-800
Inter-Storey Drift [%]
Figure 18 - Comparison between Force-Inter-storey drift curves
d
[mm]
JB-P1 Cyclic 210
JB-P2 Cyclic 210
JB-S1 Cyclic 210
JB-S2 Cyclic 175
Table 1: major experimental data
*M
N
[kNm] tot
F
[kN]
+655.09
-680.92
+666.70
-657.60
+647.12
-639.66
+627.19
-634.58
* + Sagging Moment; - Hogging Moment
+1047.2
-747.75
+892.60
-760.23
+760.09
-537.59
+887.46
-636.37
22
20
20
19
Type of Specimen
Specimen with electro-welded lattice
girders slab and Nelson connectors
around the column, with reinforcement
plates
Specimen with electro-welded lattice
girders slab and Nelson connectors
around the column, without
reinforcement plates
Specimen with profiled Steel Sheeting
slab and Nelson connectors around
the column with reinforcement plates
Specimen with profiled Steel Sheeting
slab and Nelson connectors around
the column, without reinforcement
plates
The temperature distribution into the composite joint subjected to three different timetemperature
curves, as depicted in Figure 19, was implemented by the ABAQUS 6.4.1 software
(Hibbitt et al., 2000) and a 3D FE model of joint was studied. Figure 20 shows the mean
temperature distribution on the slab as a function of the time of fire exposure obtained from the
application of the three inputs. It is evident how the application of the natural fire curve produces
a lower rise in temperature in the concrete slab than the application both of the ISO 834 curve
and parametric curve proposed in EC1 (UNI EN 1991-1-2 2004).
Temperature [°C]
1200
1000
800
600
400
200
0
ISO 834 EC1- Annex A OZone V2.2
0 10 20 30 40 50 60 70 80 90
Time [min]
Figure 19 - Different time-temperature curves
applied as a function of the time of fire
exposure
Temperature [°C]
300
250
200
150
100
50
0
Curve ISO 834
EC1-Annex A
OZone
0 5 10 15 20 25 30 35 40 45 50 55 60
Time [min]
Figure. 20 - Mean temperature distribution on
the slab as a function of the time of fire
exposure
As a results, Figure 21 shows that all the steel parts exposed to fire increase their temperature
very fast, reaching a very high temperature after only 15 minutes. Conversely, in the concrete
and in the steel component with a concrete cover rebars, horizontal plate and vertical plate

passing through the column the temperature does not increase so quickly, remaining near the
value of the ambient temperature.
Figure 21 - Temperature distribution in the beam-to-column composite joint with prefabricated
slab
On this basis, it was possible to incorporate degradation characteristics of the material into the
mechanical component model based on the “Strength Reduction Factor-SRF”, which is really a
strength retention factor present in the current European Design codes (CEN, Eurocode 3-1-2,
2003). This is basically the residual strength of the materials steel or concrete at a particular
temperature relative to its basic yield strength at room temperature. Figure 22 shows the
reduction of the design moment capacity of the joint as a function of the time of fire exposure.
M [kNm]
1400
1200
1000
800
600
400
200
0
Sagging Moment (Ozone)
t=20°C t=15 min t=30 min
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Rotation [rad]
M [kNm]
-1400
-1200
-1000
-800
-600
-400
-200
0
-0.2
-0.6
Hogging Moment (Ozone)
t=20°C t=15 min t=30 min
-1 -1.4 -1.8
Rotatio [rad]
Figure 22 - Moment-rotation relationship of the joint as a function of the time of fire exposition
It is possible to see how the hogging moment capacity reduces to approximately 90-95% of the
initial value after 15 minutes of exposure, while it approaches about 20 per cent of the initial
value after 30 minutes. The residual resistance quickly decreases owing to the loss of
resistance of the component exposed to fire. The degradation in resistance and stiffness is
observed both for sagging and hogging bending moment, respectively; nevertheless these
numerical results show that the beam-to-column joint is able to carry the internal action that the
beams transfer into the joint for a maximum time of 15 minutes without any passive fire
protection: this time is enough to evacuate the building after a severe earthquake.
CONCLUSIONS
The paper presented part of results of a numerical and experimental study aimed at developing
performed based design methods, realistically considering the structural performance under
seismic-induced fire loading. In particular, the paper discuss the experimental program carried
out on steel-concrete composite beam-to-column joints together with numerical simulations
regarding the fire behaviour. Experimental and numerical results showed how the joint details
influence the beam-column sub-assemblage response. All sub-assemblages exhibited rigid
behaviour for the designed composite joints and good performance in terms of resistance,
stiffness, energy dissipation and local ductility; as expected plastic hinges developed in beams
as a consequence of the capacity design. Moreover, a behaviour factor of about 4 has been
observed for the composite frames analysed. As a results, the joint can be used in Ductility
Class M structures. Numerical fire simulations have shown that the joint is able to carry the
-2.2
-2.6
-3

internal action for a maximum time of 15 minutes without any passive fire protection: this time is
enough to quit the building after a severe earthquake. Moreover, joints endowed with
prefabricated slab exhibits a better behavior compared to the joint with a composite slab.
ACKNOWLEDGEMENTS
The experimental programme was made possible by the grant provided under the RELUIS
National programme for seismic engineering: Task n.5 - "Development of innovative
approaches on the design of steel and steel-concrete composite structures.” The author
express their gratitude to Dr: Marco Molinari and the technicians of the Structural Testing
Laboratory of Trento University, who skilfully prepared and followed the tests.
REFERENCES
Bursi O.S., et al. Editors, Final Report PRECIOUS Project Contr. N. RFSCR-03034, 2008.
Prefabricated Composite Beam-to-Concrete Filled Tube or Partially Reinforced-Concrete-
Encased Column Connections for Severe Seismic and Fire Loadings;
ECCS 1986. Recommended Testing Procedure for Assessing the Behaviour of Structural Steel
Elements under Cyclic Loads. ECCS Publication n° 45;
Ferrario F., Pucinotti R., Bursi O. S., Zandonini R. 2007, Steel-Concrete Composite Full
Strength Joints with Concrete Filled Tubes. Part II: Experimental and Numerical Results,
Proceeding of 21st Conference of Steel Structures C.T.A., Catania, 1-3 October, pp. 397-404,
Dario Flaccovio Editor;
Franssen J.-M. 2000. “SAFIR; Non linear software for fire de-sign”. Univ. of Liege;
Hibbitt, Karlsson and Sorensen 2000. “ABAQUS User’s manuals”. 1080 Main Street, Pawtucket,
R.I. 02860;
RELUIS Task n.5: "Development of innovative approaches on the design of steel and steelconcrete
composite structures. Unità di Ricerca 8 - Buri O.S., Zandonini R.;
SAC 1997, Protocol for Fabrication, Inspection, Testing, and Documentation of Beam-Column
Connection Tests and Other Experimental Specimens, report n. SAC /BD-97/02;
UNI EN1991-1-2. 2004. “Eurocode 1: Actions on Structures. Part 1-2 : General Actions –
Actions on structures exposed to fire”;
UNI EN 1992-1-2. 2005. “Eurocode 2: Design of concrete structures. Part 1-2: General rules -
Structural fire design;
UNI EN 1993-1-1. 2005. “Eurocode 3: Design of steel Structures. Part 1: General rules and
rules for buildings”;
UNI EN 1993-1-8. 2005. “Eurocode 3: Design of steel Structures - Part 1-8: Design of joints”;
UNI EN 1993-1-2. 2005. “Eurocode 3: Design of steel Structures. Part 1-2: General rules -
Structural fire design”;
UNI EN 1994-1-1. 2005. “Eurocode 4: Design of composite steel and concrete Structures - Part
1.1: General rules and rules for buildings”;
UNI EN 1994-1-2. 2005. “Eurocode 4: Design of composite steel and concrete Structures - Part
1.2: General rules – Structural fire design”;
UNI EN 1998-1. 2005. “Eurocode 8: Design of structures for earthquake resistance - Part 1:
General rules, seismic actions and rules for buildings”.