Comparison of Composite Floor Systems

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

© 2002 WIT Press, Ashurst Lodge, Southampton, SO40 7AA, UK. All rights reserved.

Web: www.witpress.com Email witpress@witpress.com


Paper from: High Performance Structures and Composites , CA Brebbia and WP de Wilde (Editors).
ISBN 1-85312-904-6

Comparison of composite floor systems


S. Silih and S. Kravanja
University of Maribor, Facultyof Civil Engineering, Slovenia

Abstract

The paper presents the economical comparison between different composite


floor system: composite I be a m and composite trusses. The comparison has
been made for simply supported b e a m for different spans and different loads.
Composite I beams and composite trusses were designed in accordance to
Eurocode 4 for the conditions of both the ultimate and the seviceability limit
state.
h order to carry out the comparison between the two different structural floor
systems, it was neccesany to obtain their optimal design solutions. The
optimization of the structures has been performed by t i e nonlinear programming
(NLP) approach. The economical objective function was defined for the
optimization and the comparison.

1 Introduction

This paper presents the economical comparison between different composite


floor system: composite I beams and composite trusses. The comparison has
been made for symply supported b e a m for different spans and differen loads.
The a m of the research is to find out the spans, at which each of the different
structures shows its advantages.
The up-to-date experience in engineering practice has shown that composite
trusses are economically viable for spans greater than 12 m. For spans between
12 and 15 m, the cost competitiveness of trusses is determined by floor to floor
height limitations, while composite trusses of spans above 18 m are generally the
most economic structural system. We wanted to verify these statements by using
also the structural optimization methods.
© 2002 WIT Press, Ashurst Lodge, Southampton, SO40 7AA, UK. All rights reserved.
Web: www.witpress.com Email witpress@witpress.com
Paper from: High Performance Structures and Composites , CA Brebbia and WP de Wilde (Editors).
596 High
ISBN Pe~ommnnce Strwtwesand ('otnposites
1-85312-904-6

Composite I beams and composite trusses were designed in accordance with


Eurocode 4 in order to satisfy the requirements of both the ultimate and the
serviceability limit state.
The optimization of both structural system was performed by the nonlinear
programming (NLP) approach. Optimization models COIvElOPT (COMposite
Beam OPTimzation) forthe cost optimization of composite I b e a m and
COMTOPT (COMposite Trusses OPTimization) for the cost optimization of
composite trusses were developed. Beside the optimal self-manufacturing costs
(material, weldmg, sheet-iron cutting, anti-corrosion resistant painting, fire
resistant painting, panelling and erection), the optimal concrete strength classes
and structural steel grades as well as all optimal dunensions have been obtained
for both composite I beams and composite trusses.
After obtaining the optimal design solutions for each combmation of span and
imposed load for both types of structures, it was possible to carry out the
economical comparison between the two structural systems.

2 Composite I beams
The composite I beam, dealt with in t h s research, are designedfroma
reinforced concrete slab of constant depth and from structural steel beams of
duosimetical welded I sections. These two members are connected together by
cylindrical shear studs, welded to the structural steel section and embedded in
concrete.
Composite b e a m were designed according to Eurocode 4. Whde the analysis
of reinforced concrete members additionally satisfies the requirements of
Eurocode 2, the analysis of structural steel was also performed accordmgly to
Eurocode 3.
The concrete slab is designed separately as a one way spanningslab of
constant depth, running continuously over the steel sections. Composite beams
have been checked for conditions at the ultimate limit and the serviceability l m t
states. When the ultimate l m t state was considered, the b e a m were checked
for:
l bending moment
l shear force
l shear buckling
l interaction between bendmg and shear
l shear between the web and the flange of the composite beam
The ultimate moment capacity was calculated by the plastic method.
Considered was only the case with the neutral axis lying withm the concrete slab,
whch is m general valid for the cases of simply supported composite beams with
duosymetical steel I sections subjected to positive bending moment.
h the case of the serviceability l m t state we calculated deflections by using
the elastic method. Creep and shrmkage of the concrete were also taken into
account. Both comhtions, deflection , S of the beam subjected to overall load
© 2002 WIT Press, Ashurst Lodge, Southampton, SO40 7AA, UK. All rights reserved.
Web: www.witpress.com Email witpress@witpress.com
Paper from: High Performance Structures and Composites , CA Brebbia and WP de Wilde (Editors).
ISBN 1-85312-904-6 lligh Pet$mnunce Stmctwes und Composites 597
and deflection S2 of the beam subjected to the variable imposed load, were
checked. Shear studs were designed by the plastic method.

3 Composite trusses

The composite truss girders are designed from a reinforced concrete slab ob
constant depth and from a Pratt truss girder (see Fig. 1). As in the case of
composite I beams, the two members are connected together bycylindncal
sherar studs, welded to the top chord of the truss and embedded in concrete.
The analysis of the composite trusses is c m e d out in acoordance with the
British standard BS 5950, whde the design constraints satisfy the requirements
of Eurocodes 4 and 3.
The composite trusses are checked at the ultimate limit state for moment and
shear capacity and at the serviceability limit state for vertical deflections. The
moment capacity of a steel truss system at the point of maximum moment is
determined by the tensile resistance ofthe bottom steel chord and the
compressive resistance ofthe concrete slab. The contribution of the top steel
chord is ignored because of concern about the amount of strain m the bottom
chord necessary before the h11 compressive action of the top chord is developed.

Figure 1: Composite truss system

The shear force gives rise to tensile/compressive axial forces m the bracing
members. The shear capacity is thus determined by the tensile and
compressivehckling resistance of the bracing members respectively. When
axial forces are calculated, all connections are assumed to be pinned.
The vertical deflections were calculated by the elastic method by considering
the effective second moment of area of the composite truss cross-section. The
effects of creep and shnnkage of concrete were also considered.
© 2002 WIT Press, Ashurst Lodge, Southampton, SO40 7AA, UK. All rights reserved.
Web: www.witpress.com Email witpress@witpress.com
Paper from: High Performance Structures and Composites , CA Brebbia and WP de Wilde (Editors).
598 lfigh Performance S ~ r z ~ c t w and
ISBN 1-85312-904-6 es (‘mposites
4 The optimization
The optimization models COMBOPT for composite I beams and COMTOPT
for composite trusses were modelled using the General Algebraic Modelling
System (GAMS). The optimizations were performed by the Nonlinear
Programming approach (NLP), where computer program MINOS (Reduced-
grahent method) was used.
The objective of the optimization was to minimize the mamfactunng costs of
the structure. The design variables of the optimization models are dimensions of
cross-sections (see Fig. 2 and 3 ) as well as concrete and structural steel material
characteristics. Considered were conrete strenghts fiom 25 to 50 h4Pa (C 25/30
to C 50/60) and stmctmal steels Fe 360, Fe 430 and Fe 510. Imput data for the
optimization are also the span and the imposed load as well as the number of
connections (internal spans) of composite trusses.

+ be be
4

L e
4

Figure 2: Vertical cross-section of composite I beam system

The cross-sections of the bottom chord and the bracing members of the trusses
are cyrcular hollow sections, while the top chord is a sqare hollow section, whch
enables the welding of the shear studs.
© 2002 WIT Press, Ashurst Lodge, Southampton, SO40 7AA, UK. All rights reserved.
Web: www.witpress.com Email witpress@witpress.com
Paper from: High Performance Structures and Composites , CA Brebbia and WP de Wilde (Editors).
ISBN 1-85312-904-6 599
High P e r J o r m n c e S t ~ ~ ~ m~ dwC oen zsp o s i r e s
l
bc bc
l ,

h t ~

Figure 3: Vertcal cross section of composite truss system

The material and labour costs, involved m the objective fimtions, are shown
m Table 1.

Table 1: Material and labour costs

Material costs for concrete C 25/30 85.00 EWm'


Material costs for reinforcing steel S 400 0.70 Euwi?
20 stud
Cylmdncal EUR
Pannehg costs (composite beams) 12.00 EWm2
I-beam truss
Material costs for structural steel Fe 360 0.3380.44 EUR/kg
Sheet-iron cutting costs 2.00 10.00 EWm'
Welding costs 3.00 5.00 EURm'
Anti-corrosion resistant painting 7.50 10.00 EWm2
Fire protection painting (F 30) 25.00 27.50 EWm2

The material costs of the cold-formed steel hollow sections for tuss system are
30 percent hgher than of the steel plates for the I beams. The cutting and
weldmg of the cold formed hollow sections are more comprehensive in
comparison with the cutting and welding of steel plates. There is also considered
a loss of painting material, when anti-corrosion and fire resistant paintings of
trusses are carried out.
© 2002 WIT Press, Ashurst Lodge, Southampton, SO40 7AA, UK. All rights reserved.
Web: www.witpress.com Email witpress@witpress.com
Paper from: High Performance Structures and Composites , CA Brebbia and WP de Wilde (Editors).
600
ISBN High Pet$ornzance Strzlctwes and C'onzpxites
1-85312-904-6

5 Comparison of results

The optimizations were performed for both types of structures for the spans
from 10 to 40 m. The structures were subjected to self weight and the variable
imposed load of 5 kN/m2.
The optimal results were obtained by t i e use of continuous nonlinear
programming approach. Topology of the structure (number of structural
elements) is fixed though each optimization. For each defined truss topology
(fiom 4 to 12internal partitions) we performed the optimization separately.
Figure 4 shows the obtained optimal self-manufacturing costs for different
topologies for the composite tn.ss system. Considered was span of 25 m. The
optimal topology was found at 8 internal partitions.

1300
1280

1180
1160
4 6 8 10 12
topology (number of internal partitions)

Figure 4: Obtained optimal costs, dependend on the topology

After the optimal designs for both composite I-beam and trusses were
obtained, we compared the self manufacturing costs of both types of composite
structures (see Fig. 5).
© 2002 WIT Press, Ashurst Lodge, Southampton, SO40 7AA, UK. All rights reserved.
Web: www.witpress.com Email witpress@witpress.com
Paper from: High Performance Structures and Composites , CA Brebbia and WP de Wilde (Editors).
ISBN 1-85312-904-6 lligll Petfomzance 3rzrctrw~s m d C’onzposites 60 1
4000
3500
3 3000
2 2500
2
v
2000
y 1500
B1000
500
0
10 15 20 25 30 40 35
truss
I-beam span m

Figure 5: Comparison of the self manufacturingcosts

The comparison shows, that the composite truss system represents m


economical more apropriate solution for all considered spans. The cost
difference grows with the span, from around 10 percent at the span of 10 m to
a l h o s t 40 percent at tie span of 40 m.
The manufacturing costs of walls, columns and foundations,which support the
floor structures, significantly depend on the mass of tie later structures. Figure 6
shows the comparison of the masses of the two different composite structures.
From Fig. 6 it is evident, that the composite I-beams become essentialy heavier
at the spans beyond 15 m.

10 15 20 25 30 35 40
I-beams truss
span (m)

Figure 6: Comparison of the masses


© 2002 WIT Press, Ashurst Lodge, Southampton, SO40 7AA, UK. All rights reserved.
Web: www.witpress.com Email witpress@witpress.com
Paper from: High Performance Structures and Composites , CA Brebbia and WP de Wilde (Editors).
602ISBNHigh Pet$ormance S t m c ~ ~ aensd Composites
1-85312-904-6

Another factor, that can considerably influence the costs of a building, is the
height of the floor structures. Figure 7 shows the comparison of the overall
heights for the two different composite structures for spans from 10 to 40 m. As
expected, the height of the composite truss system is corniderably larger. The
difference decreases slihgtly with the span, form 40 percent at the span of 10 m
to 35 percent at the span of 40 m.

450
400
350
300
250
200
G
150
100
50
0

I-beams - truss razpon (m)

Figure 7: Comparison of t i e overall heights

Conclusions
A comprehensive research work has been performed in order to compare t i e
different composite floor structures and to find out the benefits of each system.
The optimal design solutions for composite I b e a m and composite trusses were
found by the nonlinear programming approach. Beside the optimal self-
manufacturing costs (material, welding, sheet-iron cutting, anti-corrosion
resistant painting, fire resistant painting, panelling and erection), the optimal
concrete strength classes and structural steel grades as well as all optimal
dimensions have been obtained for both composite structures. Allthese
components, together with the dimensioning requirements, made the
optimization problems quite extensive and highly nonlinear. The comparison
was performed for simply supported system of spans between 10 and 40 m,
subjected to self weight and the variable imposed load of 5 kN/m2.
From the obtained results we conclude, that m general the composite truss
system represents a more economical solution than the composite I beam system.
When the mass and thus the costs of the supporting system are considered, we
conclude that the composite trusses are economically viable for spans greater
than 15m. When on tie other hand the height of the structure represents an
important factor, special care should be taken and the decision abot the structural
© 2002 WIT Press, Ashurst Lodge, Southampton, SO40 7AA, UK. All rights reserved.
Web: www.witpress.com Email witpress@witpress.com
Paper from: High Performance Structures and Composites , CA Brebbia and WP de Wilde (Editors).
ISBN 1-85312-904-6 Iiigigl? P e q h n n n c e .Stwctwes m d Composites 603
system depends on the nature of each individual building. The use of heigher
floor structures cause additional costs that should be included in the objective
fimction. We also conclude, that our result in general coincide with the
engineering experience, emphasized in the introduction of the paper.

References

Neal S., Johnson, R., DesignofComposite Trusses, The Steel


Construction Institute, Silwood Park, Ascont, 1992.
Eurocode2, Design of concrete structures, European Comtee for
standardization, 1992.
Eurocode 3, Design ofsteelstructures, European Comtee for
Standardization, 1992.
Eurocode 3, Part I-2 Structural jiredesign, European Comtee for
standadzation, 1999.
Eurocode 4, Design ofcompositestructures, European Comtee for
Standarduation, 1992.
Composite Construction I, European SteelDesignEducation
Programme, Group I O , Volume 13, British Steel, admmstered by The
Steel Construction Institute, 1994.
Composite Construction II, European Steel Design Education
Programme, Group I O , Volume 14, British Steel, admmstered by The
Steel Construction Institute, 1994.
O'Brien, E. J. and Dixon, A. S., Reinforced and Prestressed Concrete
Design - TheComplete Process, Longman Scientific 8~ Techmcal,
Essex, 1995.
Kravanja S. and h h S., The competitive spans of composite beams,
Proceedings of the Conference Eurosteel '99,eds. StudeniEka J., Wald
F. and MachaEek J., Czech Tehnical University in Prague: Prague,
pp. 623-626, 1999.

Brooke, A., Kendrick, D. and Meeraus, A., GAMS - A User's Guide,


Scientific Press, Redwood City, CA, 1988.

Murtagh, B.A. and Saunders, MA., MINOS User's Guide,Technical


ReportSOL83-20, System Optimization Laboratory, Department of
Operations Research, Stanford University, 1985.

You might also like