Engineering Structures 150 (2017) 711–718

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Ultimate behaviour and design of post-tensioned composite slabs

Gianluca Ranzi ⇑, Andrea Ostinelli
School of Civil Engineering, The University of Sydney, Sydney, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Post-tensioned composite slab systems consist of concrete slabs poured on thin-walled profiled steel
Received 18 February 2017 sheeting and prestressed soon after casting. This paper presents six experimental tests that investigate
Revised 8 June 2017 the influence of different parameters on the ultimate behaviour of post-tensioned composite slabs. For
Accepted 24 July 2017
this purpose, the samples were tested to failure in a simply-supported configuration. The parameters
Available online 4 August 2017
considered in this study included the slab thickness, the number of prestressing strands, the span length
and the continuity of the profiled sheeting. A design model is also presented for the evaluation of the ulti-
mate moment capacity by extending the applicability of procedures currently available for composite and
prestressed floor slabs. The proposed theoretical approach was validated against the measurements pre-
sented in the paper and available in the literature. The comparisons carried out between the experimen-
tal measurement and the calculated capacities showed close agreement.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction
Post-tensioned composite floors consist of concrete slabs
poured on thin-walled profiled sheeting and prestressed soon after
the casting. This structural solution is commonly used in concrete
construction where the post-tensioned composite slab spans
between reinforced or prestressed band beams, as illustrated in
Fig. 1. The potential advantages of this structural system based
on strength considerations were first pointed out, to the knowl-
edge of the authors, by Schravendeel et al. [1]. A decade later, fur-
ther research on post-tensioned composite slab systems was
carried out in Scandinavia placing particular focus on the charac-
terisation of the bond behaviour between the concrete and the
steel deck during post-tensioning and on the ultimate behaviour
of continuous slab arrangements. [2,3] In the last decade, an inno-
vative use of profiled sheeting combined with prestressing was
proposed in the UK where the tensioning of the tendons was
induced by the concrete self-weight. [4,5] For the post-tensioned
system investigated in this paper, the strands are not connected
to the sheeting and the prestressing takes place only after the hard-
ening of the concrete. Patrick and Lloyde [6] performed three tests
to compare the performance of a composite slab and of a post-
tensioned solid slab to the one of a sample combining the use of
the prestressing and the profiled sheeting. In this study, the pro-
filed sheeting was Fielders KF57 [7]. In recent years, six long-
term and ultimate tests were reported in [8,9] and related to sam-
ples cast on Stramit steel profiles [10].
⇑ Corresponding author.
E-mail address: gianluca.ranzi@sydney.edu.au (G. Ranzi).
In this context, the main contributions of the paper can be sub-
divided in: (i) the reporting of new experimental measurements to
be used as reference for the benchmarking of theoretical and
design models; and (ii) the development of a new theoretical
model suitable for routine design (currently included in the final
draft of the Australian composite code [11]) by extending the
applicability of procedures currently available for the composite
steel-concrete slab design, e.g. [12,13], and for the design of pre-
stressed solid slabs, e.g. [14–17]. In particular, the proposed exper-
imental programme, with samples cast on profiled sheeting,
evaluated the influence of different parameters, such as the slab
thickness, the number of prestressing strands, the span length
and the sheeting continuity over the supports, on the ultimate
post-tensioned composite response, as summarised in Table 1.
Comparisons carried out between the predictions obtained with
the proposed theoretical model and those measured during the
experiments presented in this paper and available in the literature
[9] showed good agreement.
2. Experimental programme
2.1. Test specimens
Six post-tensioned composite samples were prepared and
tested to failure in this experimental programme. The static config-
uration specified in the experiments was a simply-supported one.
The adopted steel sheeting was 1 mm thick and consisted of Stra-
mit Condeck HPÒ profile which possesses a cross-sectional area of
1620 mm2/m [10]. These specimens are referred to as S1-S6 in the
following. Their dimensions and number of strands are specified in

http://dx.doi.org/10.1016/j.engstruct.2017.07.075
0141-0296/Ó 2017 Elsevier Ltd. All rights reserved.
712 G. Ranzi, A. Ostinelli / Engineering Structures 150 (2017) 711–718

Fig. 1. Typical post-tensioned composite slab.

Table 1
Summary of test specimens.

Sample ID Thickness, Internal Span, Distance between line loads, No. of strands
Dc (mm) L (mm) L2 (mm)
S1 180 6000 600 3
S2 180 6000 3000 3
S3 180 6000 600 3
S4 180 6000 600 3
S5 225 7450 600 4
S6 225 7450 600 5

Table 1 considering the layout shown in Figs. 2a and b. The param-
eters varied in the preparation of the specimens included their
geometry, the sheeting continuity over the support, the number
of strands and the curing process. In particular, the overall length
of specimens S1-S4 was 7.2 m. The samples were 180 mm thick.
Their internal span between roller supports was equal to 6 m. Sam-
ples S5 and S6 had a larger thickness equal to 225 mm. Their over-
all length and internal span were 8.65 m and 7.45 m, respectively.
An overhang with a length of 600 mm was included at each end of
the slabs to minimise the prestressing anchorage influence over
the internal span length. Slab widths were measured before the
tests and were equal to 880 mm (for sample S1), 880 mm (for
S2), 890 mm (for S3), 895 mm (for S4), 900 mm (for S5) and 900
mm (for S6). Transverse N10 reinforcement (where N10 is the Aus-
tralian designation for a 10 mm diameter reinforcing bar with nor-
mal ductility) was specified at the top of the samples at a spacing
of 300 mm (this reinforcement is not shown in Fig. 2 for clarity). In
the preparation of the samples the profiled sheeting was termi-
nated before the supports by a distance of 100 mm (except S3)
as shown in Fig. 2c. This was carried out to better reflect the real
site conditions, in which case the sheeting is not able to develop
a reliable grip within the supporting band beams as depicted
schematically in Fig. 1b. In current construction practice, band
beams are cast on timber formwork. A continuous profiled sheet-
ing was adopted in the preparation of sample S3 to develop a fric-
tional resistance over the support and to determine whether there
could be ductility issues arising from this arrangement, particu-
larly significant in view of the brittle failure of the post-
tensioned composite slab reported in [6].
12.7 mm diameter strands were used for all prestressing oper-
ations. They were installed using flat metal duct. The strands were
tensioned to 85% of their capacity (i.e. equivalent to 156 kN/
strand). The number of strands specified varied between samples,
i.e. 3 strands for samples S1-S4, 4 strands for specimen S5 and 5
strands for S6 (Table 1). Over the overhang segments the prestress-
ing duct was located horizontally at mid-height of the slab, while
its profile varied parabolically between the roller supports (i.e. over
the internal span). The distances between the top of the slab and
the centreline of the strands were measured (at the mid-span) after
completion of the tests and after cutting the specimens in half. In
particular, these distances were equal to 123 mm, 137 mm, 130
mm, 146 mm, 175 mm and 168 mm for samples S1–S6, respec-
tively, while, for sample S2, the distance measured at the location
of the line load was about 120 mm.
2.2. Material properties
Average compressive cylinder strengths for the concrete were
measured at the time of the different slab tests and these varied
among the samples because the ultimate tests were carried out
 G. Ranzi, A. Ostinelli / Engineering Structures 150 (2017) 711–718
Fig. 2. Slab layout, load arrangement and support detail for the post-tensioned composite samples.
at different times. In particular, sample S6 was tested to failure
after one month from concrete casting and the average compres-
sive strength of the concrete was 41.5 MPa. Samples S1, S3 and
S5 were tested after about two months after casting with average
concrete compressive strength of 44.9 MPa, while the remaining
specimens S2 and S4 were loaded to failure after nearly four
months from pouring and had average concrete compressive
strength of 48.8 MPa and 47.6 MPa, respectively. The mix of the
concrete used for the casting of the specimens consisted of: cement
(323 kg/m3), 20 mm aggregate (755 kg/m3), 10 mm aggregate
(300 kg/m3), manufactured sand (270 kg/m3), fine sand (455 kg/
m3), fly ash (96 kg/m3), water reducer (1260 ml/m3) and water
(183 l/m3).
The yield and ultimate strengths for the sheeting were mea-
sured from tensile tests as 601 MPa and 629 MPa, respectively.
The characteristic minimum breaking load of the 12.7 mm diame-
ter strands was 184 kN.
2.3. Instrumentation layout and test set-up
The post-tensioned slabs were instrumented to monitor their
displacements at the quarter points. The relative movement
between the slab and the concrete was measured in the longitudi-
nal direction at the sheeting ends and in the transverse direction at
mid-span and at the sheeting ends. A view of the instrumentation
setup for the slip measurements at one sheeting end is provided in
Fig. 3. The load and the stroke induced by the actuator were
recorded as well. The testing layout is described in Fig. 2a and b
in which two line loads were applied in the central part of the span
at a distance L2 of 3000 mm for sample S2 and of 600 mm for the
remaining specimens, as specified in Table 1. Before the start of the
713
tests, the samples were cycled 1000 times up to a displaced shape
corresponding to the total deflection limit of span/250 as these can
affect the shear-bond behaviour between the concrete and the
sheeting, e.g. [11,13,18]. Once this was completed, the specimens
were loaded to failure under displacement control.
2.4. Slab post-tensioning
Prestressing of the slabs was carried out in two stages by post-
tensioning the strands from one end only (i.e. live end). In particu-
lar, the first stage of prestressing took place the day after the pour
when the strands were stressed to 25% of the jacking force. The
remaining 75% of the specified prestressing force was applied after
6 days from casting. The specimens (except S4) were wet cured for
7 days after casting. A water based acrylic curing compound (with
product identification of Concure A99) was applied after casting to
S4 to investigate its long-term response (not reported in this
study). After the completion of prestressing, the ducts were
grouted. The samples were then arranged in a simply-supported
configuration.
2.5. Test results
The ultimate loads recorded during the experiments are pre-
sented in Table 2. In particular, the values measured for the 180
mm thick samples were 100.0 kN (S1), 145.4 kN (S2), 113.1 kN
(S3) and 116.7 kN (S4), respectively (as expected, the presence of
the curing compound in samples S4 had no influence on the ulti-
mate response). The specimens prepared with the thicker 225
mm slab failed at 129.4 kN (S5 with four strands) and 138.3 kN
(S6 with five strands). The failure mode exhibited in the tests
714 G. Ranzi, A. Ostinelli / Engineering Structures 150 (2017) 711–718

steel sheeting (steel rib behind) concrete

metal rod embedded in concrete
and supporting LVDT through
hole inserted in the sheeting

LVDT
metal target
attached to sheeting
(a) diagram showing arrangement for longitudinal slip measurement

concrete

steel sheeting
(steel rib behind)

metal rod embedded in concrete

and supporting LVDT through
LVDT
hole inserted in the sheeting

(b) diagram showing arrangement for transverse slip measurement

(c) experimental arrangement where LVDTs for transverse and longitudinal slips measurements are

placed at same cross-section

Fig. 3. Setup for the slip measurements at one sheeting end.

Table 2
Summary of the comparisons between experimental and calculated ultimate moment capacities.

Sample ID Experimental ultimate load, Experimental ultimate moment, Calculated ultimate moment, Mu1/Mexp Calculated ultimate moment, Mu2/Mexp
Pexp (kN) Mexp (kN m)a Mu1 (kN m)b Mu2 (kN m)c
S1 100.0 152.1 140.2 0.92 144.1 0.95
S2 145.4 124.2 105.1 0.85 109.4 0.88
S3 113.1 170.0 159.8 0.94 163.5 0.96
S4 116.7 174.9 155.3 0.89 159.1 0.91
S5 129.4 250.9 239.2 0.95 244.6 0.97
S6 138.3 265.9 250.8 0.94 258.7 0.97
CK1 [9] 132.8 127.2 0.96 131.3 0.99
CK2 [9] 144.6 137.7 0.95 141.9 0.98
a
Experimental ultimate moment values include contribution due to self-weight of the sample and of the spreader beams.
b
Ultimate moment calculated using the simplified approach for the evaluation of the contribution of the prestressing strands (i.e. Eqs. (5)–(7)).
c
Ultimate moment calculated using the refined approach in the evaluation of the contribution of the prestressing strands (i.e. based on the stress-strain curve of the
prestressing strands).

subjected to the line loads 600 mm apart (i.e. L2 = 600 mm) was shear cracks. The ultimate moment capacities observed during
concrete crushing and this occurred in the mid-span region of the tests were equal to 152.1 kN m (S1), 124.2 kN m (S2), 170.0
the sample, while in the case of sample S2 (i.e. L2 = 3000 mm) kN m (S3), 174.9 kN m (S4), 250.9 kN m (S5) and 265.9 kN m
the concrete crushed near the loading points exhibiting flexural (S6), as shown in Table 2. These values include the contribution
 G. Ranzi, A. Ostinelli / Engineering Structures 150 (2017) 711–718 715

of the self-weight of the samples as well as the weight of the sprea-
der beams. The measured moment-deflection curves are shown in
Fig. 4, where vMid depicts the mid-span deflections while vR and vL
represent the deflections at L/4 from the right and left supports,
respectively. The plotted moments include the effects of the
applied external line loads and the self-weight contributions of
the specimen and spreader beams. The overall response of the
samples was ductile.
The variations of the longitudinal slips are reported in Fig. 5. In
particular, longitudinal slips remained negligible for most of the
tests during the loading procedure and only close to the peak
moments started to grow. This was not the case for sample S6 in
which case a longitudinal slip started to develop at the live end
when approaching about half the ultimate moment (Fig. 5f). Verti-
cal separation remained negligible for all specimens except for
sample S2 in which case separation occurred due to the vicinity
of the applied line load to the end of the profiled steel sheeting
(with vertical separations of about 1.9 mm and 1.2 mm at the live
and dead ends, respectively, when reaching the peak moment).
3. Theoretical model
A theoretical approach is presented in the following to deter-
mine the ultimate resistance of post-tensioned composite slabs.
The model builds on current design practice for composite slab
design, e.g. [13], and on the design procedure adopted for pre-
stressed floors, e.g. [14], and is currently included in the final draft
of the Australian composite code [11]. In the derivation of the
model, the following assumptions have been considered: (i) no
tensile stresses are resisted by the concrete; (ii) plane sections
remain plane and perpendicular to the member axis before and
after loading; (iii) a rectangular stress block is used to depict the
concrete stresses (here introduced, without any loss of generality,
based on Australian guidelines [11,15]); (iv) full and partial shear
connection is used to describe the interface behaviour between
sheeting and the slab (where the partial shear connection depicts
the condition in which the shear connection represents the critical
component, e.g. [19,20]); (v) the response of the prestressing is
described using a simplified procedure as well as relying on the

180 M (kNm) v Mid 180 M (kNm) v Mid

160 vR 160 vR
140 vL 140 vL
120 120
100 100
80 80
60 60
40 40
20 v mid-span (mm) 20 v mid-span (mm)
0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300
(a) sample S1 (b) sample S2

180 M (kNm) v Mid 180 M (kNm) v Mid

160 vR 160 vR
140 vL 140 vL
120 120
100 100
80 80
60 60
40 40
20 v mid-span (mm) 20 v mid-span (mm)
0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300
(c) sample S3 (d) sample S4

270 M (kNm) v Mid 270 M (kNm)

240 vR 240
210 vL 210
180 180
150 150
120 120 v Mid
90 90 vR
60 60 vL
30 v mid-span (mm) 30 v mid-span (mm)
0 0
0 100 200 300 400 500 0 100 200 300 400 500
(e) sample S5 (f) sample S6
Fig. 4. Experimental moment-deflections measured for all samples (moment values include contribution from the self-weight of the sample and of the spreader beams used
for loading).
716 G. Ranzi, A. Ostinelli / Engineering Structures 150 (2017) 711–718
Fig. 5. Longitudinal slip measurements (moment values include contribution from the self-weight of the sample and of the spreader beams).
stress-strain relationship of the prestressing strands; and (vi) the
components forming the structural member are assumed to be
ductile and possible failure due to limited ductility need to be ver-
ified separately, e.g. rupture of the steel sheeting as observed in the
composite post-tensioned test reported in Ref. [6].
With the proposed approach, the ultimate moment of the post-
tensioned composite section is determined as follows:
M u ¼ Nc zc  Npre zpre þ Mpr
where Nc and Npre describe the axial forces resisted by the concrete
component and by the prestressing strands, respectively, the lever
arms zc and zpre depict the distances of the concrete and prestress-
ing forces, respectively, from the line of action of the force resisted
by the profiled steel sheeting Np, and Mpr is the plastic resistance of
the profiled steel sheeting. The calculations of the terms contribut-
ing to the ultimate moment capacity Mu is outlined in the following
and the associated stress and strain diagrams are depicted in Fig. 6.
The concrete response at ultimate conditions has been
described, without any loss of generality, based on Australian
design recommendations [11,15], while material representations
of other design standards could be specified as well. In this man-
ner, the concrete strain ecu in the extreme fibre of the slab at the
ultimate moment condition is assumed to equal 0.003. The com-
pressive resistance of the concrete can then be calculated as:
[11,15]
0
Nc ¼ a2 f c cxpl b ð2Þ
which assumes a stress equal to a2fc0 (with f’c being the characteris-
ð1Þ
tic strength of the concrete and a2 being defined in Eq. (3a)) to be
resisted by the concrete over a height of cxpl, xpl defines the distance
between the neutral axis and the slab top fibre (reduced by c pro-
vided in Eq. (3b)), and b depicts the width of the slab. The parame-
ters c and a2 define the concrete stress block to be used in design
and is determined in the following based on Australian guidelines
[15]:
a2 ¼ 1:0  0:003f 0c within the limits of 0:67 6 a2 6 0:85
ð3aÞ
c ¼ 1:05  0:007f 0c within the limits of 0:67 6 c 6 0:85 ð3bÞ
 G. Ranzi, A. Ostinelli / Engineering Structures 150 (2017) 711–718
(a) typical post-tensioned cross-section
Fig. 6. Strain and stress distribution considered in the design model.
The axial contribution of the prestressing strands Npre is here
accounted for as follows:
Npre ¼ rpu Apre ð4Þ
where Apre depicts the area of the strands located at a distance dpre
from the top fibre of the cross-section to the centroid of the pre-
stressing steel, and rpu can be determined either using a simplified
expression (applicable when the minimum effective stress in the
strands is not less than 0.5fpb) available in Refs. [11,15] and
described below, or from the stress-strain relationship of the pre-
stressing strands following the procedure outlined in Appendix
for the evaluation of the strands’ strain. For the 7-wire ordinary
strands commonly specified for composite post-tensioned floors
and adopted in the tests reported in this paper the following prop-
erties apply: Apre = 98.6 mm2, fpb = 1870 MPa, and fypre = 1533 MPa.
With the simplified model, rpu is determined based on Australian
guidelines as [15]:


rpu ¼ f pb 1 
k1 k2
ð5Þ
c
with fpb being the characteristic minimum breaking strength of the
prestressing strand and parameters k1 and k2 being calculated from
[15]:
k1 ¼ 0:28 for f ypre =f pb P 0:9 ð6aÞ
k1 ¼ 0:4 for f ypre =f pb < 0:9 ð6bÞ
1  
k2 ¼ 0 Apre f pb þ Ape f yp þ ðAst  Asc Þf ys ð7Þ
bef dpre f c
where Ast and Asc define the tensile and compressive reinforcement
(i.e. steel reinforcing bars) with fys being its yield strength, bef rep-
resents the effective concrete width (for the specimens considered
in this study equal to the slab width b), and fypre and fyp describe
the yield strength of the prestressing strands and of the steel sheet-
ing, respectively.
Partial shear connection theory is used to consider the influence
of the profiled sheeting in the analysis. With this approach, the
force resisted by the sheeting Np is determined as follows:
Np ¼ minðNs ; NFSC Þ ð8Þ
in which Ns depicts the shear connection force relying on mechan-
ical interlock, and NFSC defines the axial force that the profiled steel
sheeting would resist in the case of full shear connection (i.e. when
the interface between the concrete and the sheeting is not the weak
component of the member). In particular, the value for Ns is calcu-
lated with respect to the interface shear strength su that depicts an
experimental representation of the shear-bond response exhibited
between the sheeting and the concrete. The value for su needs to
717
εcu α2 f c‘
Nc
xpl γ xpl
zc
Npre
Np Mpr zpre
(b) strain diagram (c) stress diagram
be identified experimentally for each profiled sheeting to be used
for post-tensioned floor applications based on the experimental
procedure specified in [11]. This testing procedure aims at account-
ing for the possible influence of the post-tensioning action on the
interface behaviour. The terms Ns and NFSC can be expressed as:
Ns ¼ su bLx ð9Þ
NFSC ¼ minðNyp ; NcðmaxÞ  NpreðmaxÞ Þ ð10Þ
where su is the interface shear strength between the sheeting and
the concrete obtained from experimental tests (e.g. following the
testing procedure presented in [11]), Lx represents the distance of
the cross-section under consideration to the nearest support, Nc(max)
and Npre(max) describe compressive and tensile capacities of the con-
crete and prestressing components, respectively, and Nyp depicts
the axial force resisted by the profiled sheeting at yield. These terms
can be calculated as:
0
NcðmaxÞ ¼ a2 f c cAc ; NpreðmaxÞ ¼ f pb Apre ; Nyp ¼ f yp Ape ð11a-cÞ
in which Ape and Ac are the areas of the profiled steel sheeting and of
the concrete, respectively. Obviously, the maximum compressive
force available in the post-tensioned solid slab (i.e. Nc(max) 
Npre(max) in Eq. (10)) depend on the prestressing force level and a
check needs to be carried out at the end of the calculations to
ensure that the adopted NFSC is representative of the actual level
of applied prestressing (as Npre(max) depicts an upper limit for the
prestressing force). For the profiled steel sheeting considered in this
study, su equals 210 kPa.
The flexural contribution of the profiled steel sheeting is then
determined using:


Np
Mpr ¼ 1:25Mpa 1  6 M pa ð12Þ
Nyp
where Mpa represents the plastic moment of the profiled sheeting.
4. Comparisons between experimental and calculated ultimate
moment capacities
The analytical model presented in the previous section is
applied in the following to calculate the ultimate moment resis-
tance of the specimens presented in this paper and those outlined
in [9]. The material and geometric properties adopted in the
numerical calculations are those obtained from material testing
and from measurements of the samples, therefore reflecting the
condition of the tested specimens.
Two approaches have been considered in the calculation of the
ultimate moment to account for the contribution of the prestress-
ing strands: (i) one approach based on the simplified procedure
outlined in Eqs. (5)–(7) whose calculated ultimate moments have
718 G. Ranzi, A. Ostinelli / Engineering Structures 150 (2017) 711–718
been referred to as Mu1 in Table 2 and (ii) a second approach based
on the stress-strain curve of the prestressing strand whose defor-
mation is calculated following the procedure outlined in Appendix
and whose ultimate moment predictions have been denoted as Mu2
in Table 2. Both approaches show good agreement with the exper-
imental measurements with the refined procedure (i.e. the one
based on the stress-strain relationship of the prestressing strands)
exhibiting a slightly higher accuracy. This can be observed consid-
ering the ratios of the estimated ultimate moments (i.e. Mu1 and
Mu2) over the measured capacities (i.e. Mexp) reported in Table 2,
where Mu1/Mexp and Mu2/Mexp remained in the range of 0.89–0.96
and 0.91–0.99, respectively, for all test samples except specimen
S2. In the case of sample S2, the specimen exhibited a flexural
shear failure and its ultimate moment predictions were
Mu1/Mexp = 0.85 and Mu2/Mexp = 0.88.
Based on the close matches between the experimental and cal-
culated ultimate moments the procedure based on the simplified
approach for the evaluation of the contribution of the prestressing
strands (i.e. reported as Mu1 in Table 2) is considered suitable for
routine design.
5. Conclusions
This paper presented new experimental data related to the
ultimate response of six specimens. These samples consisted of
post-tensioned slabs cast on profiled steel sheeting. The influence
of different parameters on the ultimate response of this form of
construction was investigated experimentally, including the slab
thickness (varied between 180 mm and 225 mm), number of pre-
stressing strands with diameter of 12.7 mm (varied between 3 and
5 strands), span length (with total lengths varied between 7.2 m
and 8.65 m) and continuity of the profiled steel sheeting. During
installation, the prestressing strands were post-tensioned to 85%
of their capacity. After the casting of the samples and once the
post-tensioning was completed, the specimens were arranged in
a statically simply-supported configuration till the time of their
ultimate tests. A new theoretical model was proposed and it was
based on a rigid-plastic approach that accounted for the effects
of partial shear connection. Its accuracy was validated against
the experimental measurements presented in this study and avail-
able in the literature showing good agreement between calculated
and measured values, therefore highlighting the suitability of the
proposed model for routine design.
Acknowledgements
The experimental work in this article was supported under
Australian Research Council’s Linkage Projects funding scheme
(project number LP0990190) and by Stramit Building Products
and Arup as Partner Organisations. The design model and numerical
predictions developed as part of this study were supported by the
Australian Research Council through its Future Fellowship scheme
(FT140100130). Support from the University of Sydney (Materials
& Structures Research Cluster) is gratefully acknowledged.
Appendix
The strain epu induced in the prestressing strands at ultimate
conditions is calculated as follows [14,21]:
epu ¼ ep;init þ ece þ ept
in which ep,init depicts the deformation due to the initial prestress-
ing force Pp,init, ece defines the decompression strain, and ept is the
additional strain to reach the ultimate condition. These strain com-
ponents are determined from:
 
Pp;init   1 RI RB
ep;init ¼ ; ece ¼ 1 ypre F f p;init ; F ¼
Apre Epre RA RI  R2B RB RA
ð13a-cÞ
" #
Apre Epre ep;init dpre  xpl
f p;init ¼ ; ept ¼ ecu ð13d; eÞ
ypre Apre Epre ep;init xpl
where Epre represents the elastic modulus of the prestressing ten-
dons, ypre depicts the level on the vertical axis of the prestressing
strands (assumed positive downwards), F accounts for the cross-
sectional properties, fp,init defines the initial prestressing, RA, RB
and RI are the rigidities associated with the cross-section (i.e. the
axial rigidity, the stiffness related to first moment of area and the
flexural rigidity, respectively). An iterative procedure is required
for the evaluation of the ultimate moment (Eq. (1)) that aims at
determining the value for xpl.
References
[1] Schravendeel PL, De Sitter WR, Klop GR. Prestressed post-tensioned composite
floor slabs. IABSE Conference; 1990.
[2] Koukkari H. Steel sheet in a post-tensioned composite slab. In: Proceedings of
Eurosteel 99, Prague, Czech Republic, 26–29 May 1999.
[3] Koukkari H. Anchorage zone in a steel-concrete composite slab with unbonded
tendons. In: Proceedings of the international symposium on connections
between steel and concrete, Stuttgart, Germany, 10–12 September 2001.
[4] Miller FR, Bailey CG. Design concept for a novel long span composite floor
system. Struct Eng 2007;85(21):38–47.
[5] Bailey CG, Currie PM, Miller FR. Development of a new long span composite
floor system. Struct Eng 2006;84(21):32–8.
[6] Patrick M, Lloyde M. Development of moment capacity tables for post-
tensioned composite slabs incorporating high-tensile steel decking, validated
by test data. In: Proceedings of the Australasian Structural Engineering
Conference (ASEC), Melbourne, Australia, 26–27 June 2008.
[7] http://www.fielders.com.au.
[8] Ranzi G, Al-Deen S, Ambrogi L, Uy B. Long-term behaviour of simply-supported
post-tensioned composite slabs. J Constr Steel Res 2013;88:172–80.
[9] Ranzi G, Al-Deen Hollingum G, Hone T, Gowripalan S, Uy B. An experimental
study on the ultimate behaviour of simply-supported post-tensioned
composite slabs. J Constr Steel Res 2013;89:293–306.
[10] Stramit. Stramit Condeck HP: composite slab system: product technical
manual, Stramit Building Products, Australia.
[11] Australia Standards. Draft for Public Comment Australia/New Zealand
Standard – composite structures – composite steel-concrete construction in
buildings DR AS/NZS2327-2017. Australia: Standards Australia; 2017.
[12] Ranzi G, Leoni G, Zandonini R. State of the art on the time-dependent
behaviour of composite steel-concrete structures. J Constr Steel Res
2013;80:252–63.
[13] EN 1994-1-1. EUROCODE 4: Design of composite steel and concrete structures
– Part 1–1: General rules and rules for buildings. European Committee for
Standardization; 2004.
[14] Gilbert RI, Mickleborough NC, Ranzi G. Design of prestressed concrete to
AS3600-2009. 2nd ed. London (UK): CRC Press; 2016.
[15] Australia Standards. Australian Standard for Concrete Structures AS3600-
2009. Australia: Standards Australia; 2009.
[16] The Concrete Society. Technical Report No. 43: Post-tensioned concrete floors:
design Handbook, 2nd ed., The Concrete Centre, UK; 2005.
[17] Bommer A, Aalami B. Design fundamentals of post-tensioned concrete slabs.
Post-Tensioning Institute; 1999.
[18] Cifuentes H, Medina F. Experimental study on shear bond behaviour of
composite slabs according to Eurocode 4. J Constr Steel Res 2013;82:99–110.
[19] Oehlers DJ, Bradford MA. Steel and concrete composite structural members:
fundamental behaviour. Oxford: Pergamon Press; 2015.
[20] Nethercot D. Composite construction. London: CRC Press; 2013.
[21] Gilbert RI, Ranzi G. Time-dependent behaviour of concrete
structures. London: Spon Press; 2011.
ð12Þ