©Civil-Comp Press, 2013

Proceedings of the Fourteenth International Conference on

Paper 174 Civil, Structural and Environmental Engineering Computing,
B.H.V. Topping and P. Iványi, (Editors),
Civil-Comp Press, Stirlingshire, Scotland

Tenso-Deformative Behaviour of Reinforced Precast

Concrete-Polystyrene Sandwich Panels
L. Pani, M. Valdes, C. Armas and G. Concu
Department of Civil, Environmental Engineering and Architectural
University of Cagliari, Italy

Abstract
In order to comprehend the actual seismic behaviour of structures realized with new
constructive technologies, a wide experimental programme has been carried out on
lightly reinforced precast concrete - polystyrene sandwich panels, composed of a
sheet of structural expanded polystyrene 160 mm thick inserted between two grids
of galvanized and electro welded steel wire, completed by spraying concrete to form
a composite structure. Experimental tests have been carried out on 16 panels set in
the vertical position, in-plane loaded by constant vertical loads and at the same time
subject to a cyclic horizontal concentred load. Results allowed the following
parameters to be determined: the first cracking horizontal load, the load-
displacement hysteretic curve, the collapse lateral displacement, the resisting shear
at the bottom, the resisting moment considering P-delta effect, and the equivalent
viscous damping coefficients.

Keywords: precast sandwich panels, lightly reinforced concrete, mechanical

properties, seismic behavior, numerical analysis.

1 Introduction
The need to reduce construction time and cost has led to introduce in the market
innovative construction technologies which consist in producing load-bearing panels
composed of reinforced concrete and lightening expanded polystyrene elements
acting as formwork or support for concrete casting. This construction technique
ensures both the advantages of prefabrication (ease of installation, reduced
construction time, construction flexibility and overall cost control) and the
monolithic behavior of the structure. Recently, studies have been carried out to
evaluate both static and seismic behavior of these innovative bearing elements [1, 2,
3]. In this context a construction technology has arisen, based on the possibility of
combining structural requirements and the need for heat insulation with the use of a

1
lightweight prefabricated sandwich panel composed of a sheet of expanded
polystyrene 160 mm thick inserted between two grids of galvanized and electro
welded steel wire completed, in situ, by spraying concrete to form a composite
structure of lightly reinforced concrete - polystyrene sandwich bearing squat walls.
The present paper illustrates the results of an experimental campaign carried out on
16 panels according to Guidelines [4], aimed at evaluating both seismic behavior
and mechanical performance [5, 6, 7, 8, 9].

2 Panels geometry
Each sandwich panel is composed of two lightly reinforced concrete sheets 40mm
thick and an inserted sheet of polystyrene 160 mm thick (Figure 1). The concrete
sheets are joined through proper connectors. Reinforced concrete curbs are present
in the top and the bottom of the panels. Figure 2 presents details of reinforced bars.
structural concrete
s1

polystyrene
s
s2

electrowelded steel wire mesh structural concrete

steel ties
Figure 1. Panel cross section

longitudinal
bars 4 Ø 12
stirrup Ø 8/400 mm U bar Ø 8/400 mm
210 210
170
200

670
670
2600

670
670

210
200

170

210
240

Figure 2. Panel cross section: details of reinforcement arrangements

2
Table 1 reports geometric characteristic of the panels.

Panel ID b (m) h(m) Thickness (mm) Typology

1 4,00 3,00 240 No openings
2 4,00 3,00 240 No openings
3 4,00 3,00 240 No openings
4 4,00 3,00 240 No openings
5 3,00 3,00 240 No openings
6 3,00 3,00 240 No openings
7 3,00 3,00 240 No openings
8 3,00 3,00 240 No openings
9 3,00 3,00 240 One window
10 3,00 3,00 240 One window
11 3,00 3,00 240 One window
12 3,00 3,00 240 One window
13 3,00 3,00 240 One door
14 3,00 3,00 240 One door
15 3,00 3,00 240 One door
16 3,00 3,00 240 One door
Table 1. Panels geometry and characteristics

3 Materials
The polystyrene is 16mm thick and has density ρ = 15 kg/m3. Concrete
reinforcement consists of galvanized and electro welded steel wire with yield point
fy = 650 N/mm2, tensile strength ft = 800 N/mm2 and ultimate percent elongation
A% = 4-8%. Concrete is cast in situ by the spritz beton technique. Table 2 reports
density and 28 days compressive strength of 6 cubic specimens 150 mm sized made
of the same concrete of the panels [10]. Eleven prismatic samples have been drown
from the panels for evaluating the actual concrete strength [11]. Table 3 reports the
dimensions, density and compressive strength of these samples.

Density
Specimen Density Rc Rc,Average
Average
ID
(kg/m3) (kg/m3) (N/mm2) (N/mm2)
1 1697,8 26,5
1785,2 27,8
2 1925,9 30,6
Density
3 1819,3 31,3 Rc,min
min
4 1834,1 1697,8 28,4 24,9
Density
5 1715,6 24,9 Rc,max
max
6 1718,5 1925,9 25,1 31,3
Table 2. Concrete characteristics measured on cubic samples 150 mm sized

3
 Average dimension Density
Sample Density Rc Rc,Average
(mm) Average
ID
a b h (kg/m3) (kg/m3) (N/mm2) (N/mm2)
1 40 31 43 1948,3 34,7
2 38 34 41 1900,8 1940,6 27,8 38,3
3 40 38 70 1932,3 36,3
Density
4 38 41 56 1900,3 24,9 Rc,min
min
5 40 31 43 1936,6 (kg/m3) 37,6 (N/mm2)
6 39 40 44 1982,9 42,1
1900,3 24,9
7 40 40 41 1976,2 52,7
Density
8 40 40 43 1949,1 38,8 Rc,max
max
9 40 41 43 1953,6 (kg/m3) 44,6 (N/mm2)
10 38 39 43 1955,0 47,8
1982,9 52,7
11 35 37 38 1912,0 33,8
Table 3. Concrete characteristics measured on samples drown from the panels

4 Experimental set-up
The panels have been pseudo-statically loaded by a monotonic vertical load and by
in-plane cyclic horizontal forces. Vertical loads have been applied by 500 kN
hydraulic jacks acting on a I-shaped steel beam placed on the top of the panel. The
jacks are able to run along the beam so as the same load is guarantee also when
panel displacements due to horizontal loads become significant. Three jacks have
been used for panels 1 - 4, 4 m long, and two jacks have been used for the other
panels, 3 m long. The horizontal load has been applied by a 3000 kN hydraulic
cylinder able to apply both positive and negative displacements. Constrains on the
base of the panel have been realized by restraining the bottom curb with lateral and
transversal bolted steel plates. Figure 3 reports the loading set-up.
Translations at the bottom of the panel have been measured through a mechanic
comparator, while horizontal displacements at the top of the panel have been
measured through a wire extensometer. Force and displacements real time
acquisition has been performed by a QCDAC unit.
Table 4 reports vertical loads applied on the panels. Loads of 75 kN/m and 12 kN/m
simulates the presence of two and one storey over the panel respectively.

Panel q Panel q Panel q Panel q

ID (kN/m) ID (kN/m) ID (kN/m) ID (kN/m)
1 75 5 75 9 75 13 75
2 75 6 75 10 75 14 75
3 12 7 12 11 12 15 12
4 12 8 12 12 12 16 12
Table 4: Vertical loads

4
The in-plane horizontal loads have been applied on the upper curb of the panels by
the hydraulic cylinder. For each imposed displacement three loops have been carried
out, till the collapse was reached.
The collapse condition was considered conventionally reached when cracks have
interested all the panel and horizontal cracks at the curbs level have arisen.
The loading test starts with the application of the vertical loads, kept constant for the
whole test. Then cyclic horizontal loads are applied, values increasing up to
collapse. The generalized tension-compression loop is carried out as follows:
1. A displacement ± ∆ is defined.
2. A compression force increasing from 0 to F(∆) is applied.
3. A tension force varying from F(∆) to F(-∆) is applied.
4. A compression force varying from F(-∆) to 0 is applied.
The loop is carried out three times.

Figure 3. Loading set-up for panels 1- 4

5 Experimental results
Figures 4-11 show the results of the cyclic loading. The abscissa addresses the
displacement at the top of the panel, while the ordinate addresses the applied
horizontal cyclic load, considered positive if it’s a tension load.
During the test no important translation has been measured at the bottom of the
panel, the effectiveness of the base constrain being so highlighted. Figure 12 reports
the horizontal displacements on both the top and the bottom of panels 7, 8 as a
function of horizontal cyclic loads on the top. Bottom displacements are less than
10% of top displacements. Similar results have been achieved for the other panels.

5
Figure 4. Panels 1and 2. In-plane horizontal cyclic loading

Figure 5. Panels 3and 4. In-plane horizontal cyclic loading

Figure 6. Panels 5 and 6. In-plane horizontal cyclic loading

6
 Figure 7. Panels 7 and 8. In-plane horizontal cyclic loading

Figure 8. Panels 9 and 10. In-plane horizontal cyclic loading

Figure 9. Panels 11 and 12. In-plane horizontal cyclic loading

7
 Figure 10. Panels 13 and 14. In-plane horizontal cyclic loading

Figure 11. Panels 15 and 16. In-plane horizontal cyclic loading

Figure 12. Panels 7 and 8. Top and bottom displacements vs horizontal load

8
Figure 13 shows for each panel the horizontal load producing the first crack visible
to the naked eye. For panels 10 and 13 the first crack has arisen in consequence of
vertical loads only.

Figure 13. First cracking horizontal load

Figure 14 shows for each panel the maximum horizontal displacement, considered
positive if deriving from a compression load.

Figure 14. Maximum horizontal displacements

Experimental resistant shear at the bottom of panels for both tension and
compression loads, reported in Figure 15, has been assumed as the maximum force
applied in tension and compression phases respectively.

9
 Figure 15. Experimental shear resistance

In-plane experimental resisting moment at the bottom of panels for both tension and
compression loads has been evaluated taking account of the P-Δ effect (Figure 16):

MR compr = Fmax compr · h + q · b ·∆max compr

MR tens = Fmax tens · h + q · b ·∆max tens

where h and b are panels height and base respectively and q is the distributed
compressive load.

Figure 16. Experimental bending resistance

10
6 Numerical analysis
A numerical analysis has been performed with the aim of obtaining the equivalent
viscous damping coefficient and the structural ductility of the panels.

6.1 Equivalent viscous damping coefficient

The experimental cyclic behavior of the panels was generally asymmetric, thus the
equivalent viscous damping coefficient has been evaluated for each loading cycle by
means of the model proposed by Rodrigues et al. [12]. This model calculates
autonomously the equivalent viscous damping coefficients for the compression and
tension phases.
For a perfectly symmetric hysteretic response and corresponding closed loop (as in
the case of pure harmonic loading), the hysteretic equivalent damping coefficient
ξhyst can be accurately given by the well known equation (1), where ED stands for the
dissipated energy within a given cycle, Aloop is the area of the corresponding closed
loop in the total restoring force-displacement diagram and ES0 is the elastic strain
energy associated with the maximum force Fmax and displacement Δmax reached in
the loop:

ξhyst = ED / 4 π ES0 = Aloop / 2π Fmax Δmax (1)

However, in the case of seismic loads or even for tests performed under
displacement controlled conditions, some asymmetries can be observed and the
loops may not be closed. Therefore, based on the work of Jacobsen [13] and
according to the procedure proposed by Varum [14], the equivalent hysteretic
damping can be evaluated for each half-cycle of the force-displacement curves as
described below:

1. Each half-cycle is identified, delimited by a pair of zero-force points.

2. For each force-displacement half-cycle, the maximum generalized force Fmax and
the maximum generalized displacement Δmax are evaluated, which allows
calculating the elastic strain energy ES0.
3. For each half-cycle, the dissipated energy ED is computed by performing the
integral of the force-displacement curve leading to the Ahalf-loop value.
4. The equivalent damping coefficient ξeq is computed with the following equation
2, for each half-cycle:

ξeq = Ahalf-loop / π Fmax Δmax (2)

Figures 17-24 report the average value of the equivalent viscous damping coefficient
obtained in the three loops as a function of the drift%= Δ/h ⋅ 100.

11
Figure 17. Equivalent viscous damping coefficient. Drift 0,1%

Figure 18. Equivalent viscous damping coefficient. Drift 0,2%

Figure 19. Equivalent viscous damping coefficient. Drift 0,3%

12
Figure 20. Equivalent viscous damping coefficient. Drift 0,4%

Figure 21. Equivalent viscous damping coefficient. Drift 0,5%

Figure 22. Equivalent viscous damping coefficient. Drift 0,6%

13
 Figure 23. Equivalent viscous damping coefficient. Drift 0,8%

Figure 24. Equivalent viscous damping coefficient. Drift 1,0%

6.2 Structural ductility

A numerical procedure for determining the structural ductility of the panels from
cyclic load-displacement diagrams has been applied, as describe in the following
and with reference to Figure 25:
1. The point (Δ1, F1) is identified, with F1= 0,6 Fmax and Δ1 obtained from the
intersection of the straight line y = F1 and the envelope curve.
2. The slope of the linear-elastic segment is determined as:

tg α = F1/Δ1 (3)

3. The straight line y = Fy is identified from the equivalence between the areas under
the envelope diagrams and the equivalent bilinear curve.

14
4. The yielding horizontal displacement Δy is obtained from the intersection of the
straight line y = Fy and the linear-elastic segment.
5. The ductility µ is defined as:

µ = Δmax / Δy (4)

where Δmax is the maximum horizontal displacement when collapse is reached.

Figure 26 shows both the envelope diagram and the equivalent bilinear curve for the
tested panels. Diagrams parameters and panels ductility are reported in Table 5.

Figure 25. Load-displacement diagram: envelope and equivalent bilinear curve

Panel Load (kN) Displacement (mm)

Ductility
ID F1 Fy Fmax Δ1 Δy Δmax
1 150 221 256 3,50 2,15 15,8 3,11
2 150 238 252 2,77 4,39 7,58 1,82
3 150 279 253 14,01 26,06 27 1,00
4 150 226 247 11,35 17,10 22,6 1,29
5 150 221 250 2,81 4,14 27 6,41
6 146 225 243 2,71 4,18 5,9 1,31
7 120 253 206 18,47 38,92 41,8 1,05
8 108 161 180 8,03 11,97 35,9 2,67
9 89 135 149 2,09 3,15 5,2 1,53
10 150 220 253 3,33 4,88 13 2,66
11 120 195 190 16,34 26,61 31,9 1,16
12 92 151 153 8,61 14,19 17,9 1,25
13 150 228 236 6,65 10,10 22,1 2,18
14 150 221 242 6,66 9,79 19,1 1,94
15 120 176 191 12,76 18,70 28,7 1,55
16 120 193 195 20,05 32,20 32,9 1,00
Table 5: Diagrams parameters and panels ductility

15
Figure 26. Load (kN) – displacement (mm) diagram: envelope and equivalent
bilinear curve for the tested panels

16
7 Conclusions
The results of pseudo-static tests carried out on 16 lightly reinforced precast
concrete - polystyrene sandwich panels loaded by both vertical loads and in-plane
cyclic horizontal loads are summarized in the following points:
- The loaded panels did not reach an experimental collapse nor did show
significant damage. After the test they were still able to bear vertical loads. The
collapse condition was considered conventionally reached when cracks have
interested all the panel and horizontal passing through cracks have arisen at the
curbs level.
- The first crack arose in consequence of an horizontal load of about 100kN for
panels without openings, while for the other panels the first cracking resistance
was lower. Two of the panel with openings experienced the first crack under
vertical loads only.
- The panels behavior under cyclic loading was generally asymmetric. Resisting
shear and in-plane resisting moment for the compression phase were generally
13% higher than those of the tension phase. On average the panels bore an
horizontal load of 200 kN and an in-plain bending moment of about 600 kNm.
- Panels experimental behavior showed some variability. This might depend on
the quality of concrete, that was casted by using the spritz beton technique in
summer when temperature was more than 30°C. Compressive tests carried out
on cubic samples made of the same concrete of the panels showed scattered
compressive strength values, ranging between 24,9 and 31,3 N/mm2. Also
compressive tests carried out on prismatic samples drown from the panels
showed scattered results, ranging between 24,9 e 52,7 N/mm2.
- The structural ductility, numerically extracted from the experimental results as
the ratio between collapsing maximum displacement and yielding displacement,
was lower than 2 for 60% of the panels. Collapsing and yielding displacements
were numerically obtained from the equivalent bilateral curve derived by the
envelope of the diagrams of cyclic loading.
- The equivalent viscous damping coefficient, which depends on the hysteretic
energy dissipation, was numerically evaluated from the experimental results by
applying a model taking account for the asymmetric behavior of the panels
during the cyclic loading. The variability of the equivalent viscous damping
coefficient with both drift% and load condition (pulling or pushing phase) was
pointed out.

References
[1] N. Ile, J.M. Reynouard, “Ligthly Reinforced Walls Subjected to Seismic
Excitations: Interpretation of CAMUS 2000-1 and 2000-2 Dynamic Tests”,
Journal of Earthquake Engineering, Vol. 12, 91-114, 2008.
[2] P.A. Hidalgo, C.A. Ledezma, R.M. Jordan, “Seismic Behaviour of Squat
Reinforced Concrete Shear Walls” , Earthquake Spectra, 18(2), 287-308,
2002.

17
[3] H.Y. Chai, J.D. Anderson, “Seismic response of perforated lightweight
aggregate concrete wall panels for low-rise modular classrooms”, Engineering
Structures, Vol. 27, 593-604, 2005.
[4] “Linee Guida per Sistemi Costruttivi a Pannelli Portanti basati sull’Impiego di
Blocchi Cassero e Calcestruzzo Debolmente Armato Gettato in Opera”,
Consiglio Superiore dei Lavori Pubblici – Servizio Tecnico Centrale, 2011.
[5] T. Paulay, M.J.N. Priestley, “Seismic Design of Reinforced concrete and
Masonry Buildings”, Wiley Interscience Press Publication, John Wiley & Sons
inc., 1992.
[6] R.E. Englekirk, “Seismic Design of Reinforced and Precast Concrete
Building”, Wiley Interscience Press Publication, John Wiley & Sons inc.,
2003.
[7] J.W. Wallace, J.P. Moehle, “Ductility and Detailing Requirements of Bearing
Wall Buildings”, Journal of Structural Engineering, ASCE 118(6), 1625-1644,
1992.
[8] S.L. Wood, “ Performance of Reinforced Concrete Building during the 1985
Chile Earthquake: Implication for the Design of Structural Walls”, Earthquake
Spectra, 7(4), 607-637, 1991.
[9] T.N. Salonikios, “Shear strength and deformation patterns of R/C walls with
aspect ratio 1.0 and 1.5 designed to Eurocode 8”, Engineering Structures 24,
39-49, 2002.
[10] UNI EN 206-1, “Calcestruzzo. Parte 1: Specificazione, prestazione,
produzione e conformità”, 2006.
[11] UNI EN 14481-1, “Prove su calcestruzzo proiettato parte 1: Campionamento
su calcestruzzo fresco e sul calcestruzzo indurito”, 2005.
[12] Rodrigues H., Varum H., Arede A., Costa A. “A comparative analysis of
energy dissipation and equivalent viscous damping of RC columns subjected
to uniaxial and biaxial loading” Engineering Structures 35, 149-164, 2012.
[13] Jacobsen L.S. “Steady Forced Vibrations as Influenced by Damping.” ASME
Transaction, 52(1): 169-181, 1930.
[14] Varum H. “Seismic assessment, strengthening and repair of existing
buildings” Ph.D. thesis. Aveiro: Civil Engineering Department, University of
Aveiro, 2003.

18