Академический Документы
Профессиональный Документы
Культура Документы
www.elsevier.com/locate/jcsr
Abstract
Although composite slabs are simple and economic construction elements, the verifica-
tions that are required for their design (structural safety, serviceability) are long and compli-
cated. Current design methods found in standards and guidelines rely on the results of costly
and time-consuming large-scale laboratory tests.
In this paper, a new design approach for the prediction of composite slab behaviour is
proposed. This new approach combines results from standard materials tests and small-scale
tests with a simple calculation model (referred to herein as the ‘‘New Simplified Method’’) to
obtain the moment–curvature relationship at the critical cross-section of a composite slab.
Unlike other recently proposed methods, the calculation method described herein does not
rely on computer driven numerical simulations. Results obtained using this new design
approach have been verified by comparison with large-scale tests.
# 2003 Elsevier Ltd. All rights reserved.
1. Introduction
Corresponding author. Tel.: +41-21-693-24-27/25; fax: +41-21-693-28-68.
E-mail address: michel.crisinel@epfl.ch (M. Crisinel).
0143-974X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0143-974X(03)00125-1
482 M. Crisinel, F. Marimon / Journal of Constructional Steel Research 60 (2004) 481–491
On the other hand, the verifications that are required for the design of these
structural elements (i.e. strength, serviceability) can be long and complicated. In
fact, they require the consideration of secondary effects that are not easy to model.
Examples of these effects include local buckling of the thin steel sheeting, partial
connection between the steel and concrete, punching shear due to concentrated
loads, dynamic effects, anisotropy, etc.
Normally, fabricators of steel sheeting provide engineers and builders with
design tables for commonly used spans and thicknesses in order to facilitate the
design of composite slabs. However, engineers who need to justify their calcula-
tions, or design slabs with non-standard dimensions generally will not have the
necessary information required to carry out the calculations on which these design
tables are based. This is because the information in the tables is determined using
current design methods that require experimental values. Similarly, a fabricator
wanting to develop a new sheeting profile currently does not have the means neces-
sary to predict the degree to which it will be able to act compositely with the cast-
in-place concrete.
Complications in the design of composite slabs, as mentioned above, have led
numerous researchers to develop new methods through the application of numeri-
cal solutions. This allows the designer to simulate the behaviour of composite slabs
using a numerical model, while decreasing (or eliminating) the number of large-
scale tests needed for design. In order to simulate the behaviour of the steel–con-
crete slab numerically, it is nevertheless necessary to know the behaviour of the
various slab components. This includes the material behaviour of the sheeting, con-
crete, reinforcing steel, etc. as well as the characteristic mechanisms acting at the
steel–concrete interface of the composite slab. Instead of using large-scale tests to
determine these mechanisms, small-scale tests have been developed that can provide
M. Crisinel, F. Marimon / Journal of Constructional Steel Research 60 (2004) 481–491 483
this information while, at the same time, providing a less costly and more flexible
alternative to large-scale testing.
2. Existing methods
Two design methods for the verification of composite slabs are described in
Eurocode 4 [1]—the shear-bond method [2] (also known as the m&k method [3])
and the partial connection method [4] (see Figs. 2 and 3). These methods are based
on a test program composed of at least six full-scale slab specimens (length 2–4 m,
width approximately 1 m, thickness 100–200 mm). Using these methods, the num-
ber of tests needed to determine the behaviour of the various existing commercial
products under service and ultimate loading becomes rather significant and testing
programs become very expensive.
Due to the semi-empirical nature of these two methods, neither model can be
said to result in a clear picture of the physical behaviour of the steel–concrete con-
nection. It would be very useful for practising structural engineers to develop a
design model and a set of associated design rules based on a more physical
interpretation of the connection behaviour. It is for this reason that the develop-
ment of a simple method is fundamental for the improvement of such structural
elements. Such a method should be based on a realistic theoretical model and a
simple and inexpensive test program, requiring a reduced number of small-scale
specimens.
Over the last decade, several attempts have been made to develop new design
methods for composite slabs based on the idea of using experimental values from
small-scale tests instead of the standard large-scale tests [5–7]. The objectives of
these developments have been threefold:
– to move away from the use of expensive and awkward large-scale tests,
– to propose a design model that resembles the models used for the design of
steel–concrete composite beams, and
– to take into account parameters that are ignored by the existing methods.
Fig. 4. Daniels pull-out test: specimen (a), and typical shear stress vs. slip result (b).
M. Crisinel, F. Marimon / Journal of Constructional Steel Research 60 (2004) 481–491 485
predict the behaviour of the slab and determine its flexural and longitudinal shear
resistance.
Recent work done by the Steel Structures Lab (ICOM) at the Swiss Federal
Institute of Technology in Lausanne (EPFL) has led to the development of a New
Simplified Method capable of predicting the behaviour of composite slabs [8]. The
proposed approach can be applied to slabs that exhibit both non-ductile and duc-
tile behaviour. It uses parameters describing the longitudinal shear connection
between the steel sheeting and concrete obtained from small-scale tests such as the
one developed at ICOM by Daniels [5] (Fig. 4). The effect of supplementary para-
meters such as friction and end anchorage can also be taken into account.
The physical model representing the composite slab in the New Simplified
Method can be likened to a composite beam. The steel sheeting is modelled as an
I-section with the same area and moment of inertia as the original sheeting section.
Similarly, the concrete is modelled as a rectangular section with a moment of iner-
tia and area identical to the original concrete section. The behaviour of the com-
posite slab is determined at the critical cross-section, that is, at the location of the
maximum sagging moment.
Material behaviour of the slab components is taken as elastic–perfectly plastic
for the steel (in tension and in compression) and concrete in compression, and as
elastic with brittle failure for the concrete in tension. Bernoulli’s assumption that
plane sections remain plane is presumed. Furthermore, the curvatures of the steel
and concrete elements are assumed to remain equal before and after slip has
occurred (i.e. no vertical separation between the concrete cover and the sheeting).
3.1. Steel–concrete connection
Fig. 5. Non-ductile (a) and ductile (b) s–s relationships obtained from small-scale tests and correspond-
ing models.
Phase I: Linear elastic behaviour, without concrete cracking and without slip
(total interaction between steel and concrete).
Phase II: Elastic or elasto-plastic behaviour, with concrete cracking and without
slip (total interaction).
Phase III: Non-linear elasto-plastic behaviour, with concrete cracking and with
slip (partial interaction).
In addition to these phases, it is also possible to define the three points that cor-
respond with the upper limit of each phase:
M. Crisinel, F. Marimon / Journal of Constructional Steel Research 60 (2004) 481–491 487
Fig. 7 shows the modelled cross-section, the strain distribution over the critical
cross-section and the resulting forces for the three phases (and corresponding
upper limits) described above.
Experimental, simulated (FEM model) and simplified moment vs. curvature (M–
/) relationships at the critical section for non-ductile and ductile composite slabs
are shown in Fig. 8. The FEM model was validated through comparison with
large-scale tests. These relationships are compared at the critical cross-sections of
simple span composite slabs loaded by two concentrated forces at the quarter
spans (Ls ¼ L=4, see Fig. 2). For experimental values obtained during the large-
scale testing of composite slabs, only the moments at first slip (Mtest,slip) and at
ultimate load (Mtest,ult) are shown since curvature was not measured during these
tests.
This comparison shows: (1) very good correlation between the simplified method
results and the numerical simulations; (2) good agreement between the calculated
moments and moments from the slab bending tests, both at the first slip and ulti-
mate load levels.
Fig. 8. Moment–curvature relationships for non-ductile (a) and ductile (b) composite slabs.
M. Crisinel, F. Marimon / Journal of Constructional Steel Research 60 (2004) 481–491 489
One could ask the question: of what importance is the development of a new
method for the design of composite slabs? The answer to this question is given here
by way of a discussion of the advantages of the proposed method both for the
practising structural engineer and for the profiled sheeting producer.
Composite beams
– Calculation of bending resistance: uses the stress distribution given by the curva-
ture of the composite cross-section.
– Calculation of vertical shear resistance: only the web of the steel section con-
tributes to shear resistance.
– Calculation of longitudinal shear resistance: the number of shear connectors
depends on the longitudinal shear forces at the steel–concrete interface; the
design shear resistance of the connectors is given by push-out tests on small-scale
specimens.
Composite slabs
– Calculation of bending resistance: uses the stress distribution given by the curvature
of the composite cross-section.
– Calculation of vertical shear resistance: only the concrete part of composite cross-
section contributes to shear resistance; the composite slab is treated as a rein-
forced concrete slab.
– Calculation of longitudinal shear resistance: currently requires full-scale testing of
composite slabs.
The producer is faced with the task of improving composite slab performances
within the constraints of a research and development budget. At present, in order
to provide load tables or charts for only one sheeting profile, he must conduct a
minimum of six full-scale slab tests. This number increases with the number of
parameters (protective coating of the profiled sheeting, end anchorage of the slab,
etc.) and can constitute a significant cost. If the profiled sheeting producer wants to
improve his product by changing the shape of the profile or modifying the emboss-
ments on the sheet walls, he must first develop and build a new profiling tool
before conducting new full-scale composite slab tests. This can be a very costly
process.
With the New Simplified Method, the producer can manually fabricate new
small-scale steel profile samples, cast them into pull-out specimen (see Fig. 4a) and
test them in a relatively inexpensive way. With this procedure, he will be able to
develop and put new products on the market in a much more economically feasible
manner. Load tables and charts can also be produced using the simplified method
without need of full-scale testing.
6. Conclusions
The New Simplified Method presented in this paper facilitates the calculation of
the load-carrying capacity of composite slabs by considering three phases of the
M–/ behaviour observed in composite critical cross-sections. It requires knowledge
of the geometric dimensions of the slab, the material properties (steel and concrete)
and the characteristic behaviour of the steel–concrete connection as determined
with tests on small-scale specimen.
Some technical aspects that require further analysis in order to improve our con-
fidence in the method and extend its range of applicability are as follows:
– The effect of the shear strain distribution along the length of the slab span,
– The effect of vertical separation between the profiled sheeting and the concrete
slab,
– The reduction of the mechanical connection strength due to elongation of the
sheeting in bending,
M. Crisinel, F. Marimon / Journal of Constructional Steel Research 60 (2004) 481–491 491
Acknowledgements
The authors would like to thank Technical Research and Consulting on Cement
and Concrete (TFB) of the Swiss Cement Industry and the Swiss Commission for
Technology and Innovation (CTI) for their financial support of the theoretical and
experimental work carried out for this project.
Thanks and recognition are also given to Pascal Guignard of Calatrava Valls,
Zurich (formerly of the Steel Structures Lab, ICOM), Ann Schumacher and Ahti
Lääne of ICOM, who were instrumental in the development of the New Simplified
Method. Thanks is also extended to Thomas Lang, of Swiss Federal Railways,
Bern (formerly of ICOM) and Philippe Edder of ICOM for their contributions to
the work presented herein.
References
[1] ENV 1994-1-1. European pre-standard Eurocode 4: design of composite steel and concrete struc-
tures: part 1-1 general rules and rules for buildings, CEN, Brussels; 1992.
[2] Schuster RM. Composite steel-deck-reinforced concrete systems failing in shear-bond. Preliminary
Report Ninth Congress IABSE Amsterdam. Zürich: IABSE; 1972, p. 185–91.
[3] Porter ML, Ekberg Jr. CE. Design recommendations for steel deck floor slabs. ASCE J Struct Div
1976;102(11):2121–36.
[4] Bode H, Sauerborn I. Zur Bemessung von Verbunddecken nach der Teilverbundtheorie. Stahlbau
1992;61(Heft 8):241–50.
[5] Daniels BJ. Comportement et capacité portante des dalles mixtes: modélisation mathématique et
étude expérimentale. Thèse no. 895, Ecole polytechnique fédérale de Lausanne; 1990.
[6] Patrick M. Shear connection performance of profiled steel sheeting in composite slabs. Doctoral
Thesis, School of Civil and Mining Engineering, The University of Sydney; 1994.
[7] Veljkovic M. Behaviour and resistance of composite slabs: experiments and finite element analysis.
Doctoral Thesis, University of Technology, Luleå, Sweden; 1996.
[8] Crisinel M, Schumacher A, Lääne A. Nouvelle méthode de calcul des dalles mixtes à tôles profilées,
Tracés no. 03, Lausanne, 2002, pp. 7–15.
[9] Lääne A, Edder Ph. Pull-out tests on steel–concrete composite slab small-scale specimens. Ecole
polytechnique fédérale de Lausanne, Report ICOM 451; 2002.