Journal of Earthquake Engineering

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/ueqe20

A Simple Mechanical Model for Steel Beam –

Column Slab Subassembly Nonlinear Cyclic
Behaviour

Tushar Chaudhari, Gregory MacRae, Des Bull, G. Charles Clifton & Stephen J.
Hicks

To cite this article: Tushar Chaudhari, Gregory MacRae, Des Bull, G. Charles Clifton & Stephen J.
Hicks (2022): A Simple Mechanical Model for Steel Beam – Column Slab Subassembly Nonlinear
Cyclic Behaviour, Journal of Earthquake Engineering, DOI: 10.1080/13632469.2022.2139308

To link to this article: https://doi.org/10.1080/13632469.2022.2139308

Published online: 02 Nov 2022.

Submit your article to this journal

Article views: 2

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=ueqe20
JOURNAL OF EARTHQUAKE ENGINEERING
https://doi.org/10.1080/13632469.2022.2139308

A Simple Mechanical Model for Steel Beam – Column Slab

Subassembly Nonlinear Cyclic Behaviour
Tushar Chaudharia, Gregory MacRae b
, Des Bullb, G. Charles Clifton c
,
and Stephen J. Hicks d
a
Technical Manager - Structural Design, Neilsoft Limited, Pune, India; bDepartment of Civil and Natural Resources
Engineering, University of Canterbury, Christchurch, New Zealand; cDepartment of Civil and Environmental
Engineering, University of Auckland, Auckland, New Zealand; dSchool of Engineering, University of Warwick,
Coventry, UK

ABSTRACT ARTICLE HISTORY

A simple strut-and-tie macro-model has been developed to represent the Received 14 August 2021
seismic behaviour of a composite deck slab, acting as part of a steel moment- Accepted 18 October 2022
frame beam-column-slab joint. The moment frame is subject to large lateral KEYWORDS
deformations, representing those from a strong earthquake, and yielding Steel deck; earthquake
occurs at the beam ends. The frame sub-assemblies were modelled as elastic engineering; mechanical
line elements connected with non-linear/linear spring elements. The model model (strut and tie);
behaviour was compared to experimental test results. It was found that the moment frames; shear studs;
model developed satisfactorily represented the envelope cyclic behaviour of structural steel building with
the test frame sub-assembly with the different slab configurations. It is easy composite slab
to implement in practice using widely used commercial software available to
practising engineers.

1. Introduction
It is common practice to design composite deck floor slabs in steel buildings for gravity loads alone.
However, during earthquake shaking of buildings with moment-frames, the slab may also increase the
steel beam-column subassembly strength as a result of interaction between the slab and the steel
elements. The increase in composite beam and subassembly strength can change the deformation
mode from that anticipated into something undesirable, as noted by various researchers (Civjan 1998;
Civjan, Engelhardt, and Gross 2001; Doneux and Parung 1998; Lee and Lu 1989; MacRae et al. 2007;
Plumier and Doneux 2001). For example, in a subassembly designed for plasticity occurring at the
beam ends, the presence of the floor slab makes the composite beam flexural strength greater than that
of the steel beam alone, and column or panel zone yielding may occur instead of the anticipated beam
yielding (Braconi, Elamary, and Salvatore 2010; Green, Leon, and Rassati 2004; Lee and Lu 1989;
MacRae et al. 2013). This may lead to a higher probability of a soft-storey mechanism or other
undesirable behaviour. However, the subassembly strength increase may be lost at large deformations
due to slab degradation near the column. The increase in column demands due to slab interaction
depends on the slab type (thickness, strength, confinement) and the connection to the column
(touching outer flange, inner flange, or both). There are many different approaches to deal with the
influence of the slab in design. Some engineers ignore it, while others consider the strength increase for
column design as part of the overstrength considerations, but not for the design strength of the beam
alone (Chaudhari et al. 2019). There is no universally agreed method to assess the strength, and
provisions which do exist in Standards do not always provide a conservative result. Sometimes, in
order to minimise the slab contribution on the composite beam overstrength of the column demand,

CONTACT Tushar Chaudhari [email protected] Technical Manager - Structural Design, Neilsoft Limited, Pride Parmar
Galaxy, 8th Floor, 10/10 + A, Sadhu Vaswani Chowk, Pune, Maharashtra 411001, India
© 2022 Taylor & Francis Group, LLC
2 T. CHAUDHARI ET AL.

a gap is placed between the slab and the column even though this introduces other issues/weaknesses
into the system (MacRae and Clifton 2013). Given the lack of consistency of approaches and the non-
conservative nature of some approaches.
A method is needed to assess column demands to safely design the sub-assembly using simple
software readily available to engineers. The numerical model (finite element macro model) developed
should represent the slab behaviour and be used in static pushover and THA (time history analysis).
The model should consider:

(a) the slab interaction on column exterior and interior faces (called as force transfer Mechanism 1
and 2 in Eurocode (EN1998-1 2004)),
(b) shear stud characteristics and deck orientation, and
(c) slab confinement in the slab-column interaction zone.

The prime focus of this research paper is to address the need above by developing a simple
numerical model to predict column additional demands due to beam overstrength, caused by the slab-
beam-column interaction in the composite slab construction (MacRae et al. 2007, 2013; Mago and
Clifton 2008). In particular, answers are sought to the following questions:

(i) What are the previous considerations for slab-frame interaction in seismic moment frames?
(ii) Can a simple model be developed for readily available software which considers all likely
failure modes?
(iii) How well does the model work for slab detailing scenarios appropriate for practice?

2. Experimental Studies on Lateral Behaviour Composite Deck Slab and Steel Frame
Experimental tests of steel beam-to-column subassemblies (depicted in Fig. 1), such as those by Lee
and Lu (1989), Civjan, Engelhardt, and Gross (2001), Leon, Hajjar, and Gustafson (1998), Hobbs
(2014), Chaudhari (2015), have shown that the presence of a slab connected to the beam, may increase
the lateral strength to greater than that of the steel beam alone. In Fig. 1, it is assumed that the neutral
axis on the left side beam is below the slab, so no gapping occurs at the bottom of the slab, however
high neutral axis positions are also possible.
The major mechanisms contributing to this strength increase results from direct bearing on the
column outer flange, or on the inside of the column flanges. These have been called Mechanism-1 and
Mechanism-2 in EN1998–1 (2004), as shown in Fig. 2. An additional mode, caused by the slab pushing
on the tops of the out-of-plane beams (i.e. beams perpendicular to lateral frame) causing beam twist,

Figure 1. Frame sub-assembly subjected to lateral load.

 JOURNAL OF EARTHQUAKE ENGINEERING 3

Figure 2. Force transfer mechanisms between slab and column (EN1998–1 2004).

described as Mechanism-3 in EC8 (EN1998–1 2004), does not seem to have a significant influence on
the strength in most subassemblies (Webb et al. 2018).
Experimental tests also showed that the increase in strength was often around 30% for standard slab
details (Civjan, Engelhardt, and Gross 2001; Hobbs 2014; Chaudhari et al. 2015) and that it was
sensitive to the steel decking placement and details. For highly confined slabs beside the column,
lateral strength increases of up to 50% were obtained (Hobbs 2014; Chaudhari 2015). There is the
possibility that this increase in strength may cause yielding in a different, and possibly more undesir­
able, location than if there were no composite beam slab contribution (e.g. MacRae et al. 2013). This
may lead to a higher probability of a soft-storey mechanism (such as seen by Yamada et al. 2009) or
other unwanted behaviour. It has been observed that the post-peak strength degradation may occur
due to concrete slab fracture between the flange tips (Civjan, Engelhardt, and Gross 2001; Chaudhari
et al. 2019; Hobbs 2014), or slab concrete longitudinal shear on either side of, and parallel to, the
longitudinal beams (Hobbs 2014).
Hence, to avoid the strength increase (and possible subsequent decrease) in construction practice,
a bearing of the slab on the column, which is common in conventional construction (e.g. Doneux and
Parung 1998; Hobbs 2014; Chaudhari 2015) should be avoided. This may be accomplished by
providing a gap between the slab and the column (Leon, Hajjar, and Gustafson 1998; Hobbs 2014;
Chaudhari 2015). The gap may be filled with soft material (Chaudhari and MacRae 2016). However,
the presence of the gap may lead to a greater possibility of column instability and buckling (Uang and
Fan 2001; FEMA451 2006) and, as such, additional methods may be required to restrain this (Uang
and Fan 2001; MacRae and Clifton 2013).

3. Code Design Approaches

For design, slab in-plane effects are treated in different ways; some groups ignore the slab effect (e.g.
ANSI/AISC:341–10 2010). In New Zealand, the slab effect on the beam dependable strength is ignored
for the design of the beam itself, but must be considered in the beam overstrength values used for the
capacity design of the column and panel zone. The simple methodology, included in NZS3404:
Part1:1997(2007), based on work by MacRae et al. (2007), considers the tests of Civjan (1998), the
strut approach of Umarani and MacRae (2007), and the fibre element modelling of the beam end by
Kim, Stanton, and MacRae (2004). for reinforced concrete frames. It results in an increase in strength
4 T. CHAUDHARI ET AL.

of about 30% for some typical construction. This increase is less than the maximum obtained from
more recent steel building beam-column subassembly tests with composite deck flooring, where
increases of up to 50% have been obtained (Hobbs 2014; Chaudhari 2015). EN1998–1 (2004) requires
ductile reinforcement for the composite slab surrounding a column. To achieve steel section bottom
flange yielding without slab concrete crushing, EN1998–1 (2004) specifies that the total compressive
force developed in the EC8 force transfer mechanisms should be 20% greater than the total slab force
(due to positive and negative bending of the beams on either side of the column). However, an explicit
method for column design considering slab-column effects is not specified.
In summary, ANSI/AISC:341–10 (2010) ignores slab effects, while EN1998–1 (2004) considers
different force transfer mechanisms but does not explicitly describe the complete seismic design, and
NZS3404:Part1:1997(2007) uses a simple method which applies to slab overstrength but and not to the
beam design.

4. Numerical Approaches
Various numerical studies have also been conducted which consider slab-steel moment frame inter­
action effects for moment frames. For example, Kattner and Crisinel (2000) developed a two-
dimensional model using DIANA software to study the behaviour of semi-rigid composite joints,
where beam elements represented the steel beam, column, and the concrete slab. Shear studs were
modelled using translational springs. The slab-column interaction was realised through the horizontal
springs with a compression-only property acting only on the column outer flange without specific
consideration of slab reinforcing. A linear kinematic condition was used to link the slab with the
column, and the load-displacement behaviour of the shear stud translation springs was obtained from
experiments and codes. These authors concluded that the proposed model of the composite joint with
flush endplates could able to numerically simulate the global behaviour (i.e. moment-rotation) of the
semi-rigid joint.
Rassati, Leon, and Noè (2004) modelled partially restrained/semi-rigid composite connections
using the equations in Annex J of the ENV1993-1-1 (1992). This model is formulated using eight
different springs to account for the influence of the various deformation components, including a slip
in the bolts, partial interaction between the concrete slab and steel girder, shear deformation of the
panel zone, and cracking and crushing of the slab. The model needs a user-defined element developed
in ABAQUS software, but it is difficult for engineers to use in practice.
Umarani and MacRae (2007) proposed a beam-column joint model to study the slab effect on the
moment resting frames considering the beam growth caused by the gap opening as well as to study the
effect on beam overstrength. To consider the slab effect a “slab element” has been added; this element
was modelled as a strut element with bilinear hysteretic behaviour. The beam-column joint was treated
as a rigid joint; because of this, the shear deformation was ignored. Here, the slab element had some
limitations, such as: the slab’s bending effect was ignored; slab compressive resistance was considered
to be elastoplastic, and interaction with the column sides was ignored as the slab was not connected
with the column node. The proposed model developed using Ruaumoko − 2D (Carr 2004), and
calibrated with experimental results, captured the envelope behaviour for the models compared.
Elghazouli, Castro, and Izzuddin (2008) used a specialised non-linear finite element program
ADAPTEC to study the seismic performance of a composite moment frame designed to the
EN1998–1 (2004) provisions. They conducted time-history analyses and showed that the seismic
intensity and the panel zone parameters influence the behaviour. However, their main emphasis was
not the slab behaviour.
Sheet et al. (2013) improved the prediction by Umarani and MacRae (2007) by using a Ramberg-
Osgood hysteresis loop to better empirically predict the force-displacement unloading behaviour.
From the above, it may be seen that while various researchers have developed a beam-column joint
model composed of linear and non-linear springs (representing the components like bolts, end plate,
shear tabs, and panel zone), the modelling and calibration of these springs needs specialised numerical
 JOURNAL OF EARTHQUAKE ENGINEERING 5

software (like Opensees, Ruaumoko, ADAPTEC, and ABAQUS), which is not commonly used by
practitioners.

5. Proposed Macro Model

5.1. Geometry Idealisation
The moment frame internal beam-column sub-assembly with composite deck slab considered was
based on the experimental tests of Chaudhari (2015) as shown in Fig. 3a. This subassembly was
designed using capacity design principles to obtain strong column/connection – weak beam behaviour
so that the column remains elastic. The model, developed using SAP2000 (Elghazouli, Castro, and
Izzuddin 2008), considered an assemblage of non-linear springs, elastic beam-column elements, and
slab axial elements to capture the behaviour caused by EC8 Mechanism-1 and 2 (EN1998–1 2004). The
310UC158 column, 310UB32 beam, and 150 mm thick ComFlor80 composite deck slab (ComFlor80
2014) were modelled with line elements along the centrelines of these elements, as shown in Fig. 3b,c.
The various nomenclatures given in Fig. 3 are, “d” the depth of steel beam, “Hc” depth of steel column
and “Beff” the effective width of slab.
Each beam was divided into three parts; (i) a stiff beam element (representing the joint, endplate
and gusset plate connection), (ii) non-linear rotational springs (representing the beam plastic zone),
and (iii) elastic beam elements (representing the elastic component) as shown in Fig. 3b. Each shear
stud was modelled with a non-linear spring using the SAP2000 (CSI Berkeley, U. 2015) multilinear
plastic link element. In current MRSF (moment resisting steel frame) design practice as per NZS3404:
Part1:1997(2007) and AS/NZS:2327 (2017), shear studs are not permitted between the column face
and 1.5d from the column face, to minimise composite action in this zone thereby encouraging plastic
hinging there, and also to mitigate the possibility of shear studs initiating cracking in the beam plastic
hinge zone. Because of this, in the experiments and model, the first shear stud was located 1.5d from
the column face, as depicted in Fig. 3a.
The slab was divided into two zones, as shown in Fig. 3c. The first zone (Zone 1) represents the slab
near the shear studs. It was modelled as a line element connected by the shear stud springs to the beam
element. The second zone (Zone 2) considers the slab from the column centreline to the first shear
stud from the column. This was modelled as an assemblage of axial elements, as shown in Fig. 3c. In
the transition zone of Zone 2, stiff elements (shown as dashed) transfer force between the composite
slab section (i.e. Zone 1 slab element) and the elements in the Zone 2 braced rectangular box
arrangement shown in red and blue. The diagonal compression struts carry compression force from
bearing between the slab and the inside and outside of the column flanges to the transition zone
elements. The panel zone was assumed rigid based on an experimental study (Chaudhari 2015) and
previous numerical study observations.

5.2. Concrete Constitutive Relationship

The concrete uniaxial compression model of Aslani and Jowkarmeimandi (Aslani and
Jowkarmeimandi 2012) was used; here, the compression envelope used the Carreira and Chu
(Carreira and Chu 1985) concrete model, with exponential values for the ascending and descending
branches. The compressive stress is provided as a tabular function of the plastic strain. In the current
analysis, the uniaxial compressive stress–strain curve was assumed linear up to 0.4f’c. Thereafter it was
calculated according to Eqs. (1)–(3), as shown in Fig. 4a.
� �
f 0 c n εε0cc
fc ¼ � �n (1)
n 1 þ εε0cc
6 T. CHAUDHARI ET AL.

Figure 3. Schematic representation of numerical model.

0:74 0
n ¼ n1 ¼ ½1:02 1:17ðEsec =Ec Þ� ; if εc � εc (2)

0
n ¼ n2 ¼ n1 þ ða þ 28bÞ; if εc � εc (3)
 JOURNAL OF EARTHQUAKE ENGINEERING 7

45 2.5
40
Compressive Stress (MPa)

Tensile Stress (MPa)

35 2.0

30
1.5
25
20
1.0
15
10 0.5
5
0 0.0
0.000 0.001 0.002 0.003 0.004 0.005 0 0.0005 0.001 0.0015 0.002 0.0025
Compressive Strain Tensile Strain
a) Compression Envelope Curve b) Tension Envelope Curve
Figure 4. Uniaxial stress–strain curve of concrete under compression and tension loading.

where:
fc = Compressive stress of concrete (MPa)
Ec = Tangent modulus of elasticity of concrete (MPa)
Esec = Secant modulus of elasticity (MPa)
εc = Strain of concrete
fc 0 = Cylinder compressive strength of concrete (MPa) � � �
ε0c = Tensile strain corresponding to tensile strength = Efcc þ r r 1
n = Material parameter that depends on the shape of the stress – strain curve
n1 = Modified material parameter at ascending branch
n2 = Modified material parameter at descending � 0:46 branch
a = Constant = 3:5 12:4 0:0166f 0 c ðMP�a Þ
0
b = Constant = 0:85exp 911=f � c ðMPa Þ
0
r = Constant = f c ðMPa Þ=17 þ 0:8
The stress–strain relationship of concrete under tension was assumed to be linear up to the
maximum tensile strength of concrete, ftu, where ftu (MPa) = 0.36√(fʹc (MPa)) (NZS3101:1 2006).
Thereafter, the tensile strength decreased, as shown in Fig. 4b. The concrete tensile stress-strain model
is given in Eq. (4) (Aslani and Jowkarmeimandi 2012), where ft is the concrete tensile stress, ft is the
concrete tensile strain, ftu is concrete tensile strength, and εtu is the strain corresponding to concrete
maximum tensile strength.
�
Ec εt ifεt < εtu
ft ¼ (4)
ftu ðεtu =εt Þ0:85 ifεt > εtu

5.3. Coordinate System

In the global coordinate system used, the X-axis represents the longitudinal direction of beams,
and the global Y-axis is orthogonal to the beam in the horizontal plane (representing the
transverse direction). The global Z-axis defines the vertical direction of the frame sub-
assembly, which coincides with the longitudinal axis of the column, as shown in Fig. 5a.
The sectional and force-deformation properties for the various elements were provided in their
local co-ordinate system (CSI Berkeley, U. 2015). This is denoted using 1, 2 and 3 as shown in
Fig. 5b.
8 T. CHAUDHARI ET AL.

Figure 5. Numerical model co-ordinate system.

Figure 6. Beam element details.

5.4. Element Formulation

5.4.1. Beam Element with Rotational Spring
The idealised numerical model of the frame sub-assembly’s beam is shown in Fig. 6, which was
discretised in three parts (as mentioned before); stiff zone, plastic hinge zone, and the elastic beam.
 JOURNAL OF EARTHQUAKE ENGINEERING 9

The beam-to-column connection was considered as a stiff zone (based on the experimental observa­
tions). Also, to simplify the model, the endplate connection and gusset plates were treated as a stiff
beam element. The moment of inertia for the stiff beam element was calculated as the sum of the
moment of inertias of the bare beam and the gusset plates (i.e. Istiff = Ibeam + Igusset).
The nonlinearity of the beam was simulated using the CPH (concentrated plastic hinge) approach;
hence a zero-length non-linear link element was used in the beam plastic hinge region. The non-linear
cyclic behaviour was simulated using the SAP2000 (CSI Berkeley, U. 2015) Wen plasticity model. The
non-linear spring degrees-of-freedom are: (i) axial deformation in translational direction 1 (i.e. U1)
with linear properties, (ii) shear deformation in translational directions 2 and 3 (i.e. U2 and U3
respectively) with linear properties, (iii) rotation in direction 3 (i.e. R3) with non-linear properties, and
rotational deformation restraint in directions 1 and 2 (i.e. R1 and R2). A high axial stiffness and shear
stiffness were provided. These stiffnesses were made high enough so the results were not sensitive to
the actual values, but low enough to avoid numerical instability problems.
The input properties required for the Wen plasticity spring are; (i) initial flexural stiffness, (ii) post-
yield stiffness ratio, (iii) yield moment, and (iv) exponent coefficient. The initial flexural stiffness of the
rotational spring (Kspring) was calculated using Equations 5 and 6, where Imod is the modified moment
of inertia of the beam, Ibeam is the beam moment of inertia about the major axis, and n is the
multiplication factor of 10 based on Ribeiro et al. (2015), and this represented the data.

6:E:Imod
Kspring ¼ n: (5)
L1

nþ1
Imod ¼ Ibeam : (6)
n
The initial flexural stiffness of the rotational spring was modified by a constant “n” to account for
the combined effect of the non-linear springs and the elastic beam-column element. The rotational
spring at the beam end was modelled as a rigid-plastic (by multiplying initial stiffness with “n”) so that
the numerical model does not pose any numerical instability issues (Ribeiro et al. 2015).
As reported by Ibarra and Krawinkler (2005), the overall hysteretic response of the beam is
a combination of the individual moment-rotation of the rotational spring and the elastic beam-
column element. In the non-linear time history analysis, the rotational spring dominates the overall
moment-rotation behaviour of the beam, and the beam-column element remains elastic. Since the
rotational spring and the elastic beam-column element are connected in series, the post-yielding to
elastic stiffness ratio (i.e. the strain hardening coefficient) was adjusted to obtain the strain hardening
coefficient of the rotational spring. The methodology suggested by Ribeiro et al. (2015). was adopted
here to obtain the strain hardening coefficient of the rotational spring. The post-yield stiffness ratio of
the rotational spring was calculated as:
α
αspring ¼ (7)
1 þ ½n:ð1 αÞ�

Where “α” is the nominal strain hardening ratio, which was considered equal to 3% (this value was
assumed based on the literature and the section analysis of the beam 310UB32), the post-yield stiffness
ratio of the rotational spring αspring was equal to 0.0028 (for n = 10). This value was used in the current
numerical simulation. Isotropic hardening was considered by increasing the predicted plastic
moment, Mp, to the likely maximum moment, Mm using the methodology suggested by Kawashima
et al. (1992). The flexural strength, Mm, is calculated as Zp(σu + σy)/2, where σy is the tension coupon
test yield stress, σu is the tension coupon test ultimate tensile stress, and Zp is the section plastic
modulus, which accounts for isotropic hardening. In the Wen plasticity model, the sharpness of the
hysteresis was influenced by the yielding exponent coefficient.
10 T. CHAUDHARI ET AL.

5.4.2. Shear Stud Idealisation

In the numerical model, the shear stud is represented using a non-linear link element available in
SAP2000 (CSI Berkeley, U. 2015). The force-displacement behaviour of the shear studs was obtained
from the methodology reported in the literature (EC4 2004, Hicks 2011; Lam and El-Lobody 2005). In
the current numerical simulation, the force–displacement relationship for the shear stud was calcu­
lated using the Johnson and Molenstra (1991) formulation,
�
P ¼ Prk 1 e βs α (8)

where “Prk” is the characteristic strength of the shear studs, which was calculated based on the
research study conducted by HERA (Hicks 2011), the values for constants “α” & “β” were selected as
0.989 and 1.535 mm−1 respectively as specified by Johnson and Molenstra (1991). The resulting shear
stud non-linear load-slip behaviour is shown in Fig. 7. The use of the shear stud spring element is
simple and gives good convergence (Wang and Tizani 2010). The shear spring idealisation is from the
beam mid-height to the mid-thickness of the topping concrete, as shown in Fig. 7b. Rigid constraints
assigned in the respective axes of the link element allow the shear stud force–slip relationship to
represent the slip between the top of the steel beam and the concrete decking. Shear spring character­
istics have been modified to represent the two shear studs placed beside each other. The various
nomenclature given in the Fig. 7 are, “tc” topping slab thickness, “hrc” height of steel deck rib.

5.4.3. Slab-Column Interaction Idealisation

The strut-and-tie formulation at the slab-column interaction zone considers that both force transfer
mechanisms (1 and 2) act in parallel as shown in Fig. 8. In the proposed model, they were represented by
an equivalent compression strut “keq” with the stiffness of a combined strut (keq= k1 + k2), where “k1” and “k2
” are the axial stiffness for Mechanism 1 and 2 respectively. Therefore, the equivalent strut is represented by
a single compression strut with an area equal to sum of those from the Mechanism 1 and 2.
This equivalent compression strut was assumed to act at the inclination of 45° and connected to the
column centre line with the non-linear compression-only contact spring shown in Fig. 9, representing
the slab-column interaction. The struts extend to Beff/4 in the slab longitudinal and transverse
directions, where Beff is slab effective width equal to L/4, where L is the distance between column
centres in the longitudinal direction according to NZS3404:Part1:1997(2007). The effective width of
the slab (Beff) represents the region of collection of slab force. However, the strut and tie model, Beff/2
are placed halfway along this collection region. The various nomenclatures given in the Fig. 9 are,
“Lstrut” the length of strut, “Astrut” the area of strut, “FT” the force in transverse tie, “FL” the force in
longitudinal tie, “Fc” the force in equivalent compression strut.
The non-linear contact spring force-displacement properties used the concrete compression stress-
strain envelope in Fig. 4. Here, the compression force, Fc, was obtained as the stress, fc, multiplied by
the strut area, Astrut. The corresponding displacement, ∆strut, was the corresponding concrete strain, εc,

Figure 7. Non-linear shear spring to model the behaviour of shear stud.

 JOURNAL OF EARTHQUAKE ENGINEERING 11

Figure 8. Equivalent strut comprising of mechanism 1 and 2 struts.

Figure 9. Idealised strut-and-tie model.

multiplied by the strut length, Lstrut. The contact spring force–displacement relationship is shown in
Fig. 10. Here, the peak of the force-displacement envelope of compression only non-linear contact
spring reparents bearing/crushing failure and was obtained as the maximum compression stress
(fc,max) multiplied by the strut area (Astrut).
The deck orientation effect has been considered while calculating the force-displacement properties of
the non-linear contact spring. In the transverse deck assembly, the column only bears against the topping
concrete, while for a longitudinal deck slab, the concrete in the decking trough may also contribute to the
bearing area. This aspect is considered in the current model while calculating the slab contact area.
In case of the full depth slab sub-assembly, the concrete in around the column in full depth portion
was confined using a reinforced cage. The effect of the concrete confinement is considered while
determining the non-linear contact spring properties using stress-strain properties of confined con­
crete based on the Mander, Priestley, and Park (1988). confined concrete model.
In between the composite slab section (zone-1) and the strut-and-tie (zone-2), the transition zone was
connected through the series of rigid axial elements, with the cross-sectional area, Astiff = Beff/2.tc, as shown
in Fig. 3. The slab in the composite section region was linear elastic flexural with an upper limit on tensile
12 T. CHAUDHARI ET AL.

Figure 10. Contact spring: force-displacement envelope.

strength (Ften) and it then immediately loses its strength after cracking, as shown in Fig. 4b. Here, Ften = f’ctAs
where f’ct is maximum tensile stress (i.e. 0.36√f’c, here f’c is in MPa) and “As” is the effective slab area (Befftc).
The steel reinforcing in this region was ignored for simplicity. The struts in the slab-column interaction zone
have compression-only properties, whereas the tie members carry tension only. Both the transverse and
longitudinal tie members were assumed linear elastic with the stiffness properties of the rebars.

5.5. Boundary Conditions and Loading Protocol

To replicate the boundary conditions of the tested frame subassembly, the beam ends were provided
with roller supports and the column bottom with pinned supports. Shear studs, and the corners of
strut-and-tie elements, were provided with out-of-plane restraints to avoid numerical instability as
depicted in Fig. 5a. The displacement control loading protocol specified in ACI T.1–01 (2001) (refer
Fig. 11) was applied to the column top as per the experimental tests.

6. Validation of Proposed Model

As part of the experimental research, beam-column subassembly tests with different slab configura­
tions were conducted at the University of Canterbury (Chaudhari 2015; Chaudhari et al. 2019).
Specimen details are summarised in Table 1 and shown in Figs. 12–16.

Figure 11. Displacement control loading regime.

 JOURNAL OF EARTHQUAKE ENGINEERING 13

Table 1. Details of experimental test configurations.

Active Force Transfer
Test Specimen Designation Deck Orientation Detail around the Column Mechanism
1 Bare Steel Frame - - -
(BSF)
2 Fully Isolated Slab Unit Transverse Deck All around isolation of slab from the None
(FI-SU) column
3 Shear Key Slab Unit Longitudinal Slab isolated on the column outer flange Mechanism-2
(SK-SU) Deck
4 Modified Shear Key Slab Longitudinal Slab isolated on the column outer flange Mechanism-2
Unit Deck
(MSK-SU)
5 Full Depth Slab Unit Transverse Deck Slab casted touching to the column on full Mechanism-1 and −2
(FD-SU) depth

Figure 12. Bare steel frame (BSF) test sub-assembly.

Figure 13. Fully isolated slab unit (FI-SU) test sub-assembly.

The comparison of the overall hysteresis behaviour of the frame sub-assemblies with different slab
configurations obtained using SAP2000 (CSI Berkeley, U. 2015) based numerical macro model with
the corresponding experimental test results are shown in Fig. 17. It can be observed that the numerical
model was able to simulate the cyclic behaviour of the frame sub-assembly with the different slab
configurations with reasonable accuracy. Also, the numerical model captured the initial lateral
stiffness. This is also summarised in Table 2. The deviation in the hysteresis loop of the BSF frame sub-
assembly is shown in Fig. 17a. This is due to the limitation of the non-linear spring. It fails to capture
the full Bauschinger effect and beam buckling effect as observed in the experimental test. However, the
14 T. CHAUDHARI ET AL.

Figure 14. Shear key slab unit (SK-SU) test sub-assembly.

Figure 15. Modified shear key slab unit (MSK-SU) test sub-assembly.

Figure 16. Full depth slab unit (FD-SU) test sub-assembly.

developed numerical model simulates the key envelope behaviour of the FI-SU (Fully Isolated Slab
Unit), SK-SU (Shear Key Slab Unit), MSK-SU (Modified Shear Key Slab Unit), and FD-SU (Full Depth
Slab Unit) frame assemblies (Chaudhari et al. 2019), as shown in Fig. 17b to 17e.
From above it can be seen that the strength deviation is about 7%. This is because of the
assumptions made in the model and the inability to accurately assess material properties for the
whole slab and beam elements. The stiffness deviation is between 8% and 12% for three slab assemblies
and 19% for full-depth slab assembly. The numerical model did not consider the struts at the beam
ends, or slop beside the pins in the loading system. Also, the non-linear nature of stress–strain
properties of concrete material as well as the concrete cracking, influences the stiffness estimation
 JOURNAL OF EARTHQUAKE ENGINEERING 15

Lateral Displacement (mm) Lateral Displacement (mm)

-125 -100 -75 -50 -25 0 25 50 75 100 125 -125 -100 -75 -50 -25 0 25 50 75 100 125
300 300
250 250
200 200
Lateral Force (kN)

Lateral Force (kN)

150 150
100 100
50 50
0 0
-50 -50
-100 -100
-150 -150
-200 EXP-BSF -200 EXP-FISU
-250 SAP-BSF -250 SAP-FISU
-300 -300
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
Lateral Drift (%) Lateral Drift (%)

a) Bare Steel Frame b) Fully Isolated Slab Unit

Lateral Displacement (mm) Lateral Displacement (mm)

-125 -100 -75 -50 -25 0 25 50 75 100 125 -125 -100 -75 -50 -25 0 25 50 75 100 125
350 350
300 300
250 250
Lateral Force (kN)

200
150
Lateral Force (kN) 200
150
100 100
50 50
0 0
-50 -50
-100 -100
-150 -150
-200 -200
-250 Exp-SKSU -250 EXP-MSKSU
-300 SAP-SKSU -300 SAP-MSKSU
-350 -350
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
Lateral Drift (%) Lateral Drift (%)

c) Shear Key Slab Unit d) Modified Shear Key Slab Unit

Lateral Displacement (mm)

-125 -100 -75 -50 -25 0 25 50 75 100 125
350
300
250
Lateral Force (kN)

200
150
100
50
0
-50
-100
-150
-200
-250 EXP-FDSU
-300 SAP-FDSU
-350
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
Lateral Drift (%)

e) Full Depth Slab Unit

Figure 17. Comparison of numerically simulated hysteresis plots with test results.

for the concrete elements. The stiffness of test specimens with slabs are greater than that of the
numerical analysis, this is because of the stiffness of experimental tests was computed in the initial
cycles with little cracking, while full cracking is assumed in the numerical model since the cracked slab
properties were considered. This difference is likely to be more significant in the case of the full depth
slab, where slab depth and concrete area is larger.
16 T. CHAUDHARI ET AL.

Table 2. Comparison of sub-assemblies initial stiffness and peak strength of numerical macro model with experimental results.
Test Specimen Initial Stiffness (kN/m) Peak Strength (kN)
Numerical Experimental Numerical Experimental Deviation
(N) (E) Deviation (N-E)/E (%) (N) (E) (N-E)/E (%)
Bare Steel Frame (BSF) 6,212 6,156 +1.0 214.1 206 +4
Fully Isolated Slab Unit (FI-SU) 7,879 8,747 −10 219.0 211.4 +4
Shear Key Slab Unit (SK-SU) 13,183 15,031 −12 277.6 263 +6
Modified Shear Key Slab Unit 13,283 14,506 −8 271.9 285.2 −5
(MSK-SU)
Full Depth Slab Unit (FD-SU) 13,298 16,402 −19 326.7 306.3 +7

7. Conclusions
A simple model is developed for steel beam-column joints with slabs using readily available computer
software, such that engineers can use in PBEE (performance-based earthquake engineering) time
history analysis assessments due to earthquake excitation. The following conclusions are made:

(1) A number of tests on steel beam-column moment frame subassemblies with composite slabs
indicate that the bare steel frame strength was increased up to 50% due to the presence of the
slab (full-depth slab configuration) and this effect is often not considered in the routine design.
(2) The macro-model can reasonably predict the frame sub-assembly lateral strength and initial
lateral stiffness. Specific considerations for the concrete bearing surface, shear stud character­
istic, slab confinement are made. It is shown that the approach represents the behaviour of
a number of subassemblies. However, there is some variation due to the model simplicity,
consideration of the boundary conditions and the concrete cracking effects.
(3) The model captures the overall beam-column slab subassembly experimental envelope beha­
viour under cyclic loading well for slabs detailed in five different ways. The configurations
involved: no composite slab; typical slab details; a highly confined slab; and two other inter­
mediate cases. Overall, the approach developed is good for use by practitioners.

Acknowledgements
The authors would like to acknowledge the MBIE Natural Hazards Research Platform for its support to conduct the
proposed research study as a part of the Composite Solution Research Project. Additional support was provided by
ComFlor New Zealand, John Jones Steel Research Funding, and Heavy Engineering Educational and Research
Foundation (HEERF). All opinions expressed remain those of the authors.

Disclosure statement
No potential conflict of interest was reported by the author(s).

ORCID
Gregory MacRae http://orcid.org/0000-0002-3011-5146
G. Charles Clifton http://orcid.org/0000-0003-0723-1699
Stephen J. Hicks http://orcid.org/0000-0002-6667-1531

References
ACI T1.1-01. 2001. Acceptance criteria for moment frames based on structural testing. Chicago, IL: American Concrete
Institute.
 JOURNAL OF EARTHQUAKE ENGINEERING 17

ANSI/AISC:341-10. 2010. Seismic provisions of structural steel buildings. Chicago, IL: American Institute of Steel
Construction.
Aslani, F., and R. Jowkarmeimandi. 2012. Stress–strain model for concrete under cyclic loading. Magazine of Concrete
Research 64 (8):673–85. doi:10.1680/macr.11.00120.
AS/NZS:2327. 2017. Composite structures – Composite steel-concrete construction in buildings. Wellington: Standards
New Zealand.
Braconi, A., A. Elamary, and W. Salvatore. 2010. Seismic behaviour of beam-to-column partial-strength joints for steel–
concrete composite frames. Journal of Constructional Steel Research 66 (12):1431–44. doi:10.1016/j.jcsr.2010.05.004.
Carr, A. 2004. Ruaumoko 2d–inelastic dynamic analysis. Christchurch: Department of Civil Engineering, University of
Canterbury.
Carreira, D. J., and K. H. Chu. 1985. Stress-strain relationship for plain concrete in compression. Journal of the American
Concrete Institute 82 (6):797–804.
Chaudhari, T. 2018. Seismic performance evaluation of steel frame building with different composite slab configurations.
Doctor of Philosophy in Civil Engineering Ph.D. Thesis, University of Canterbury, Christchurch, New Zealand.
Chaudhari, T. D., and G. A. MacRae. 2016. Selection of gap infill material for structural seismic applications. Journal of
the Structural Engineering Society of New Zealand Inc (SESOC) 29 (1):94–102.
Chaudhari, T., D. MacRae, G. Bull, S. Chase, G., Hicks, G.C. Clifton, and M. Hobbs 2015. Composite slab effects on beam-
column subassembly seismic performance. In STESSA 2015. Shanghai, China: China Architecture & Building Press.
Chaudhari, T., G. MacRae, D. Bull, C. Clifton, and S. Hicks. 2019. Experimental behaviour of steel beam-column
subassemblies with different slab configurations. Journal of Constructional Steel Research 162:105699. doi:10.1016/j.
jcsr.2019.105699.
Civjan, S. A. 1998. Investigation of retrofit techniques for seismic resistant steel moment connections. Doctor of
Philosophy, The University of Texas, Austin.
Civjan, S. A., M. D. Engelhardt, and J. L. Gross. 2001. Slab effects in SMRF retrofit connection tests. Journal of Structural
Engineering 127 (3):230–37. doi:10.1061/(ASCE)0733-9445(2001)127:3(230).
ComFlor80. 2014. Product guide ComFlor80. Accessed January 13, 2014. http://www.Comflor.co.nz/wp-content
/uploads/comflor/brochures/comflor80brochure.Pdf
CSI Berkeley, U. 2015. Analysis reference manual. SAP2000 V17.2.0.
Doneux, C. and H. Parung (1998). A study on composite beam-column subassemblages. Proceedings of the 11th ECEE
Conference, Paris.
EC4. 2004. Eurocode 4: Design of composite steel and concrete structures - Part 1-1: General rules and rules for buildings.
Brussels: European Committee for Standardization.
Elghazouli, A., J. M. Castro, and B. A. Izzuddin. 2008. Seismic performance of composite moment-resisting frames.
Engineering Structures 30 (7):1802–19. doi:10.1016/j.engstruct.2007.12.004.
EN1998-1. 2004. Eurocode 8: Design of structures for earthquake resistance - Part 1: General rules, seismic actions and
rules for buildings. Brussels: European Committee for Standardization
ENV1993-1-1. 1992. Eurocode 3: Design of steel structures – Part 1.1: General rules and rules for buildings. Brussels:
European Committee for Standardization.
FEMA451. 2006. NEHRP recommended provisions: Design example. Washington, DC: National Institute of Building Sciences.
Green, T. P., R. T. Leon, and G. A. Rassati. 2004. Bidirectional tests on partially restrained, composite beam-to-column
connections. Journal of Structural Engineering 130 (2):320–27. doi:10.1061/(ASCE)0733-9445(2004)130:2(320).
Hicks, S. 2011. Design resistances of 19 mm diameter headed stud connectors through-deck welded within the ribs of
Comflor 60 and Comflor 80 profiled steel decking. Manukau City: HERA (Received through private communication
from Steve Stickland).
Hobbs, M. 2014. Effects of slab-column interaction in steel moment resisting frames with steel-concrete composite floor
slabs. Master thesis, University of Canterbury, Christchurch, New Zealand.
Ibarra, L. and H. Krawinkler. 2005. Global collapse of frame structures under seismic excitations. John A. Blume
Earthquake Engineering Center Technical Report 152, Stanford University
Johnson, R. P. and N. Molenstra. 1991. Partial shear connection composite beams for buildings. Proceedings of Institute
of Civil Engineers, Part 2, 679–704.
Kattner, M., and M. Crisinel. 2000. Finite element modelling of semi-rigid composite joints. Computers & Structures
78 (1):341–53. doi:10.1016/S0045-7949(00)00064-X.
Kawashima, K., , G. A. MacRae, and K. Hasegawa. 1992. The strength and ductility of steel bridge piers based on loading
tests. Journal of Research, Japan 29:195.
Kim, J., J. Stanton, and G. MacRae. 2004. Effect of beam growth on reinforced concrete frames. Journal of Structural
Engineering 130 (9):1333–42. doi:10.1061/(ASCE)0733-9445(2004)130:9(1333).
Lam, D., and E. El-Lobody. 2005. Behavior of headed stud shear connectors in composite beam. Journal of Structural
Engineering-ASCE 131 (1):96–107. doi:10.1061/(ASCE)0733-9445(2005)131:1(96).
Lee, S.-J., and L.-W. Lu. 1989. Cyclic tests of full-scale composite joint subassemblages. Journal of Structural Engineering
115 (8):1977–98. doi:10.1061/(ASCE)0733-9445(1989)115:8(1977).
18 T. CHAUDHARI ET AL.

Leon, R. T., J. F. Hajjar, and M. A. Gustafson. 1998. Seismic response of composite moment-resisting connections. I:
Performance. Journal of Structural Engineering 124 (8):868–76. doi:10.1061/(ASCE)0733-9445(1998)124:8(868).
MacRae, G., and C. Clifton. 2013. Low damage design of steel structures. Christchurch: Steel Construction New Zealand.
MacRae, G. and G. Clifton. 2015. NZ research on steel structures in seismic areas. Proceedings of the 8th International
Conference on Behavior of Steel Structures in Seismic Areas, STESSA, Shanghai, China, 44–58.
MacRae, G. A., G. C. Clifton and N. Mago. 2007. Overstrength effects of slabs on demands on steel moment frames.
Pacific Structural Steel Conference 2007, Wairakai, New Zealand.
MacRae, G., M. Hobbs, D. Bull, T. Chaudhari, R. Leon, C. Clifton, and G. Chase. 2013. Slab effects on beam-column
subassemblies—beam strength and elongation issues. Queensland, Australia: Composite Construction VII
Conference VII:28–31.
Mago, N., and C. G. Clifton. 2008. Investigation of the slab participation in moment resisting steel frames (HERA report
R 4-140). Manukau City, New Zealand: New Zealand Heavy Engineering Research Association.
Mander, J. B., M. J. N. Priestley, and R. Park. 1988. Theoretical stress–strain model for confined concrete. Journal of
Structural Engineering 114 (8):1804–26. doi:10.1061/(ASCE)0733-9445(1988)114:8(1804).
NZS3101:1. 2006. Concrete structures standard: Part 1 - the design of concrete structures. Wellington: Standards New Zealand.
NZS3404:Part1:1997. 2007. Steel structures standard part 1-Incorporating amendment no 1 and no 2”, Standards New
Zealand Wellington.
Plumier, A., and C. Doneux. 2001. Seismic behaviour and design of composite steel concrete structures. In ECOEST2
and ICONS. Lisboa, Portugal: LNEC.
Rassati, G. A., R. T. Leon, and S. Noè. 2004. Component modeling of partially restrained composite joints under cyclic and
dynamic loading. Journal of Structural Engineering-ASCE 130 (2):343–51. doi:10.1061/(ASCE)0733-9445(2004)130:2(343).
Ribeiro, F. L. A., A. R. Barbosa, M. H. Scott, and L. C. Neves. 2015. Deterioration modeling of steel moment resisting
frames using finite-length plastic hinge force-based beam-column elements. Journal of Structural Engineering 141 (2).
doi:10.1061/(ASCE)ST.1943-541X.0001052.
Sheet, I. S., U. Gunasekaran, and G. A. MacRae. 2013. Modeling of steel beam-column connections under dynamic
excitation. Journal of Structural Engineering, Published by CSIR-Structural Engineering Research Centre, Chennai
40 (13):254–61.
Uang, C.-M., and C.-C. Fan. 2001. Cyclic stability criteria for steel moment connections with reduced beam section.
Journal of Structural Engineering 127 (9):1021–27. doi:10.1061/(ASCE)0733-9445(2001)127:9(1021).
Umarani, C., and G. MacRae. 2007. A new concept for consideration of slab effects on building seismic performance.
Journal of Structural Engineering, Structural Engineering Research Centre, Chennai, India 34-3:34.
Wang, Z. and W. Tizani. 2010. Modelling techniques of composite joints under cyclic loading. In Computing in Civil
and Building Engineering, Proceedings of the International Conference, Paper 254, 507. Nottingham, UK:
Nottingham University Press.
Webb, G., K. Kosinanonth, T. Chaudhari, S. Alizadeh and G. A. MacRae 2018. Column moment demands from
orthogonal beam twisting. Key Engineering Materials, 259–69. Trans Tech Publ.
Yamada, S., K. Kasai, Y. Shimada, K. Suita, M. Tada, and Y. Matsuoka 2009. Full scale shaking table collapse experiment
on 4-story steel moment frame: Part 1 outline of the experiment. Behaviour of Steel Structures in Seismic Areas:
STESSA 2009, 143–8. Philadelphia, Pennsylvania: CRC Press.