Sei sulla pagina 1di 12

FRC 2014 Joint ACI-fib International Workshop

Fibre Reinforced Concrete: from Design to Structural Applications

The flexural behaviour of SFRC flat slabs: the Limelette fullscale experiments for supporting design model codes
Benoit Parmentier 1, Petra Van Itterbeeck 1, Audrey Skowron 1
1

: Belgian Building Research Institute (BBRI), Limelette, Belgium.

Abstract
An experimental campaign on a full-scale SFRC flat slab with 70 kg/m, without any
traditional reinforcement, was performed in 2011. These experiments were based on
uniformly distributed loading as well as concentrated loadings. The ultimate flexural loadbearing capacity of the slab was checked according to the relationships given in the Model
Code 2010 recently published by fib (MC2010). The design was made on a full plastic
approach, using the characterization of the SFRC on small specimens. The tests confirmed the
ductile behaviour and the redundancy potential in SFRC flat slabs. The identification of the
worst yield lines pattern remains an important task for the designer, as it significantly
influences the results. The laboratory characterization showed discrepancies compared to
specimens directly sawn from the slab. The use of the in-situ properties in the design method,
nonetheless, seems to provide a good agreement with the results of the full-scale tests.

Keywords
Steel fibres, fibre reinforced concrete, flat slab, design, full-scale experiments.

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

Introduction

Nowadays, common applications of steel fibre reinforced concrete (SFRC) are, on the one
hand, applications in which the designer does not only rely on the reinforcement capacity of
the fibres (combination with traditional reinforcement) or, on the other hand, concrete
applications in which the consequences of deviations in performance produce minor effects in
terms of safety (e.g. slabs on ground). Recently, theoretical models to fully design structural
fibre reinforced concrete (FRC) elements with or without conventional reinforcement were
developed and translated into the new version of the fib Model Code 2010 [MC2010] (fib,
2013). Flat slabs are one application that could benefit from the use of fibres as main
reinforcement. Flat slabs without drops (thickened areas around the columns to resist
punching shear) are an interesting kind of construction that allow for a fast erection due to the
minimalist use of simple formworks and a rapid turn-around achieved by early striking
combined with flying systems (BCA, 2001). While the use of SFRC in this category of
applications with a high degree of redundancy seems well adapted and successfully achieved
in practice (Destre, 2004), the experience in calibrating the design rules for SFRC flat slabs
is rather limited (Gossla, 2005; Michels, 2009). This paper presents a first insight (mainly in
terms of loadbearing flexural capacity) of the full-scale experiments performed at BBRI in
2011. We compare the results with recent assumptions for fibre distribution, redistribution
effects and constitutive laws for the design of SFRC flat slabs developed within the MC2010.

Description of the experimental programme

A SFRC flat slab with a flat soffit has been erected on the BBRI site in Limelette
(Belgium) in October 2011. The slab consists of 9 bays (panels) of 6x6 m delimited by
sixteen circular concrete columns with a diameter of 0.3 m. The thickness of the slab is 0.2 m
and its full size is 18.3x18.3 m (see Figure 1a). It was poured from a pump moving from
spot to spot, without any vibration, and manually troweled afterwards with a bumpcutter. The
columns are reinforced with 6 longitudinal reinforcement bars of diameter 12 mm and circular
stirrups. They are, moreover, anchored in 1.5x1.5x0.35 m concrete footings. The
longitudinal rebars of the columns reach the mid-depth of the slab without any other
additional anchoring length. The centre-to-centre distance between the columns is 6 m in
both orthogonal directions. For each of the seven truck mixers delivering the concrete for the
slab, different samples were taken to evaluate the compressive strength (3 cubes by truck), the
consistency and to check the fibre dosage. In order to evaluate the dispersion of the concrete
properties and characterize the fibre concrete as input for theoretical models, a second batch
(2 m) with the same concrete composition was delivered in a second phase by the same
concrete supplier two days after having poured the slab.
The flexural behaviour of the slab is checked against its serviceability limit state (SLS) and
its post-cracking behaviour for ultimate limit states (ULS, see Figure 1b) and in particular its
ultimate bending capacity by applying a uniformly distributed load (UDL) and, later on, a
concentrated load in the centre of specific panels.

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

Steel frame

Loadcell

Jack

(a)

Figure 1:

2.1

(b)

(a) Slab geometry (dimensions in cm). Capital letters indicate (6x6m) panels while
the notation k# refers to the column under the slab and the dashed line shows
the path of the pump during the pouring sequence.
(b) Loading frame for the ULS tests.

Fibre reinforced concrete properties

As sole reinforcement of the slab, hooked end steel fibres with a length of 60 mm and a
diameter of 1.0 mm are used. The characteristics of the steel fibres are summarized in
Table 1. The concrete composition (see Table 2) was adapted in such a manner that the
SFRC, with 70 kg/m of fibres (volume fraction Vf = 0.89%), can be pumped. The coarse-tofine aggregate ratio is close to 50:50 in order to target a good workability with the fibres
(slump of 200 mm). This was confirmed on site by the slump measurements with the Abrams
cone. In the middle of the concrete discharging phase, the slump fluctuated between 200 mm
and 260 mm in function of the truck mixer.
Table 1:

Fibre
HE+ 1/60

Nominal properties of the fibres


Tensile
strength
[MPa]

Elasticity
modulus
[kPa]

Density

Length

Diameter

Aspect ratio

[kg/m]

[mm]

[mm]

[/]

1450

210

7850

60

1.0

60

The properties of the hardened concrete are presented in Table 3 for the two phases. For
the casting of the flat slab, the compressive resistance measured on 5 cubes of
150x150x150mm at 28 days was 43.7 MPa. The post-peak behaviour was characterized
according to EN 14651, using a three-point bending test on specimens taken from the last
truck mixer (phase 1) or specially delivered (phase 2). These specimens have a cross-section
of 150x150mm with a notch of 25 mm deep at mid-span (250 mm).
3

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

Table 2:

Concrete mix

Material

Quantity

Cement [kg/m] :

400

Portland CEM I 52,5R HES 200


CEM V/A (V-S) 32,5N LH HSR 200
Coarse aggregate [kg/m] :

840

Crushed limestone 4-8 400


Crushed limestone 6-14 440
Quartz sand 0/4 [kg/m]

850

Water [l/m]

215

High range water reducer [l/m]

1.6

Steel fibres HE+ 1/60 [kg/m]

70

Water/cement ratio (W/C)

0.54

The limit of proportionality fL, the peak flexural tensile strength fmax, the residual flexural
tensile strengths fR,1, fR,2 ,fR,3 and fR,4 for a crack mouth opening displacement (CMOD) of 0.5,
1.5, 2.5 and 3.5 mm, respectively, were measured for each series of specimens (see Table 3).
Some of these properties will be used to calculate the loadbearing capacity of the slab with the
help of the analytical model from the MC2010.
Table 3:

Mean properties of the hardened concrete at 28 days


(the coefficient of variation [%] is given in brackets)

Mean compressive resistance


fcm,cub [MPa]

Test standard
(# specimens)

Slab casting
Phase 1

Control casting
Phase 2

EN 12390-3
(5 specimens)

43.7 (2)

41.8 (1)

5.1 (6)

4.5 (11)

9.2 (19)

7.1 (22)

8.6 (19)

6.4 (25)

8.7 (18)

6.8 (22)

7.5 (20)

6.4 (20)

6.9 (19)

5.7 (18)

NA

33940 (2)

Limit of Proportionality fL [MPa]


Peak strength fmax [MPa]
Residual strength at 0.5 mm CMOD - fR1,m [MPa]
Residual strength at 1.5 mm CMOD - fR2,m [MPa]
Residual strength at 2.5 mm CMOD - fR3,m [MPa]

EN 14651
(9 specimens for
phase 1 and 11
specimens for
phase 2)

Residual strength at 3.5 mm CMOD - fR4,m [MPa]


Youngs modulus Ecm [MPa]

NBN B 15-203
(3 specimens)

The average curves of the bending tests exhibit residual strengths of more than 60% of the
peak up to a mid-span deflection = 3 mm that corresponds to a CMOD of 3.5 mm (see
Figure 2). All the specimens exhibited a strain-hardening behaviour in bending, with an
increased loadbearing capacity beyond crack initiation in the concrete (after the limit of
proportionality (LOP), fL). By comparing the specimens taken during the execution of the
slab and from the additional casting, we observe that the results differ in the mid-span
4

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

deflection range [0.05..2] mm with a larger increase of the load for the samples taken from the
casting sequence of the flat slab. This can be explained by the higher number of fibres
counted across the broken section of the specimens taken from the slab pouring phase.
Finally, the coefficient of variation of the residual strengths of the different series ranged from
18% up to 25%, which is quite common for this type of test characterised by a small cracked
section and no stress redistribution possibilities.
40
1st casting (flat slab)
2nd casting (control)

35

Load [kN]

30
25
20
15
10
5
0

0.5

2.5
2
1.5
Mid-span Deflection [mm]

3.5

(a)

Figure 2:

(b)

a) The control specimens from the first casting phase exhibit larger residual forces
after the peak in three-point bending up to 2.0 mm (average curve for each series)
b) Test setup used for the tests

The control of the fibre dosage in the fresh concrete is achieved by taking samples of +/5.5 litres from the pump in the middle of each concrete discharging phase. The average fibre
dosage measured from the different truck mixers (one sample taken from each of the 7 truck
mixers in the middle of the delivery) is 63.8 kg/m with a minimum of 52.2 kg/m and a
maximum of 69.3 kg/m. The observed individual values are lower than the target of 70
kg/m, but the samples represent every time only one sample of 5.5 litres on a total volume of
10 m of concrete (truck capacity) which is rather limited. Later on, additional tests on 11
cylinder cores (113 mm) drilled randomly from panels G-H-I showed an average dosage of
68.2 kg/m with a standard deviation of 7.4 kg/m. The analysis of the distribution of the
fibres along the depth (on the same samples) did not reveal any heterogeneity. Hence, we
assumed that fibres were distributed homogenously in (this part of) the slab. This should be
confirmed later by a detailed analysis on the full slab.

2.2

Uniformly distributed loads (SLS tests)

The slab is stripped 14 days after it was poured. At this stage, a first straight crack
appeared in the centre of the panels G-H-I along the E-W direction. The crack width was on
average around w = 0.3 mm. Another crack was also noticed in the centre of panels I-G to the
N-S direction. These cracks will be discussed later in 2.4.

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

The first phase of the project deals with the evaluation of serviceability limit states
(deflections, cracks) of the slab under uniformly distributed loads (UDL). In order to evaluate
the deflection of the SFRC in this case, two flexible polyester water tanks of 6x6 m are
installed on panels A and C (see Figure 3a). Thirty-six days after casting the slab, these
reservoirs are filled alternatively with water, at a rate of 8000 litres per hour. The loading
principle consists of the filling of the water reservoirs alternatively in steps of 1 kN/m (3.6
m), while the deflections of the slab are measured by means of linear variable displacement
transducers (LVDT) distributed over the bottom of the slab.
For the first three loading steps, the central deflection of the panels increases linearly. The
deflections are quite similar for panels A and C (see Figure 3b). After 3 kN/m a first visible
crack appears at the bottom of panel A and in a similar way later on in panel C. The crack
opening in both panels was on average 0.1 mm at this time. Because this value was small, it
was decided to continue the test up to 4 kN/m for each panel. At this load level, the
maximum crack opening reaches respectively on average 0.4 mm and 0.5 mm for panel A and
C and the central deflection is 9.4 mm and 7.2 mm. At this time it was decided to stop the test
and to keep this load level for one day. At the end of the next day, the panels were unloaded.
At this moment the central deflection had reached 15 mm for panel A and 13 mm for panel C
(increase of respectively 60% and 80%, see Figure 4) and the cracks at the bottom of the slab
opened up to on average 0.8-1.0 mm.
16

Unloading panel A

Central deflection [mm]

14
12

Unloading panel C
10
8

Loading stopped
(4 kN/m)

6
4

Panel A
Panel C

2
0

(a)
Figure 3:

2.3

200

400

600

800 1000 1200 1400 1600 1800


Time [hours]

(b)
(a) Uniformly distributed loads (UDL) on the slab. Note: dimensions in cm.
(b) Central deflection of panels A & C during loading/unloading with UDL.

Concentrated loads (ULS tests)

The second phase of the study focussed on the ultimate post-cracking behaviour of the
SFRC slab (ULS). To investigate this limit state, a concentrated load is applied to the centre
of different panels (centre, edge and corner panels) consecutively. The load is applied by
means of a jack with a capacity of 1000 kN and controlled by a hand pump. The jack was
installed in a reaction frame composed out of steel profiles HEA800 (as shown in Figure 1b).
6

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

The frame is anchored by using 2 dywidag bars (diameter 26 mm) drilled 30 cm deep in the
head of the columns through the thickness of the slab. This test setup does not allow the
investigating of potential punching shear problems due to the interaction between the reaction
frame and the slab at the connection with the columns. This failure mechanism was however
not considered a problem for the SFRC used in this structure (Gossla, 2005;Michels, 2012).
The load is applied through a 30x30x5 cm steel plate. The force, continuously measured
by a load cell, is applied in steps of 50 kN and kept constant during 10 minutes to achieve
deflection stabilisation. This procedure allowed us to make some manual measurements of
crack openings and to make sure that the effective stable peak load is identified. In order to
check this, the increase of panel deflection was continuously monitored and when the increase
in panel deflection drops below 10% of the deflection increase at the beginning of the loading
plateau the test is continued.
The ULS tests are initiated five days after the SLS tests, at a concrete age of 42 days. The
first of these tests relates to a concentrated load applied in the centre of panel E. This test
would give the maximum load because of the large stress redistribution capacity of the slab in
this configuration. As foreseen, the central panel exhibits the maximum capacity for this
loading scheme with a maximum load of 328 kN (Figure 4). This load was considered as the
peak load because it was not possible to go beyond this value, and the deflections became
unstable. This first ULS test also illustrated well the remarkable plastic behaviour potential of
the material at peak load, with more than 40 mm of deflection increase before considering that
the ultimate capacity is reached. Based on the change in slope on Figure 4a and compared
with the linear elastic FEM-based prediction, we noticed a major crack initiation at a load of
85 kN. We observed very similar deflections for the adjacent panels D and F while their
very low amplitude compared to those from the loaded panel E (see Figure 4a) demonstrate
the effect of the full plastic hinge developed in the top of the slab along the columns k2-k14
and k3-k15. We also observe the development of a larger crack in N-S direction at the bottom
centreline of panel E in comparison with the crack extending along the W-E direction (see
Figure 4b). At the end of this first test the maximal crack opening reaches on average 8.5 mm
at the bottom of panel E. The full description of the crack pattern is presented in 2.4.

350

350

300

300

250

250

Load [kN]

Load [kN]

Afterwards, the load was applied in the centre of panel H and finally in the centre of panel
I in order to test an edge panel and a corner panel of the slab, respectively. The peak loads
and corresponding deflections of these experiments are presented in Table 4.

200
150
Panel E (loaded)
Panel D
Panel F
Panel E (FEM lin elast.)

100
50
0

20

40
60
Central deflection [mm]

(a)

Figure 4:

80

200
150
100
Panel E : W-E crack
Panel E : N-S crack

50

100

4
6
8
Crack opening at mid-span [mm]

10

(b)

Deflection (a) and crack openings (b) versus load during the ULS 1 test

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

Table 4:

ULS tests on the flat slab Experimental results


ULS 1 - Pcentral panel (E) ULS 2 - Pedge panel (H) ULS 3 - Pcorner panel (I)

Test
Peak load [kN]

328

143

54

Deflection at peak load [mm]

27

51

19

2.4

Crack pattern

As mentioned before, the slab was already cracked in some parts due to its self-weight just
after the formwork was removed (see Figure 5a). Two long cracks appeared at the bottom of
the slab (w 0.3 mm) and at the top of the slab above edge and inner columns (w 0.06).
The particular cement type (CEM V/A) used as part of the concrete mix could be the reason
of the reduced flexural strength at early age but this was not confirmed so far. However,
because the largest cracks were far from the fields tested under UDL and the cracks at top of
the slab were limited in both length and crack opening, it was estimated that their influence on
the deflections behaviour was rather limited. For the ULS test, it is also our opinion that the
reduced time between the removal of the formwork and the ULS tests (28 days) did not
induce any major detrimental effect (corrosion, carbonation, etc.) on the flexural capacity of
the slab (in any way not exceeded) caused by the presence of these cracks.
At the end of the UDL tests, a long crack appeared along the W-E direction at mid-span of
panels A-B-C together with circular cracks around columns k1 and k4 (see Figure 5b). The
first ULS experiment in panel E was characterised by a large crack at mid-span over the
bottom side along the N-S direction (Figure 5c). This crack initiated in the centre of the panel
and proceeded to the free edges of the slab while the load increased up to the peak loading.
Another orthogonal crack appeared and was followed by radial cracks starting from the centre
of the panel. At peak load, the slab also presented radial cracks above the head of the internal
columns and a large circular one connecting them. This last crack was repeated with a larger
radius at some locations. The same cracking pattern was observed for the two other ULS tests
with the difference that the large circular crack flowing from column to column was not fully
completed.

(a)

Figure 5:

(b)

(c)

Cracking pattern induced by the self-weight (a), by the UDL tests (b) and by the
ULS 1 test (c). Note: dotted lines indicate cracks at the top of the slab.

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

Check of ultimate resisting moment of the slab

A first insight about the deflection and stress distribution in the slab was obtained with a
FEM-based commercial software. Then, the analytical model described in the MC2010 was
used to estimate the ultimate flexural capacity of the SFRC slab. In the following, we only
focus on the first ULS test to check the design relationships.

3.1

Calculation of the ultimate flexural capacity of the slab

The fib Model Code 2010 (MC2010), which was recently published, allows the designer to
calculate the ultimate capacity of any SFRC structure. In a particular case, we use the rigidplastic (RP) model approach to estimate the ultimate flexural capacity of the slab.
The RP model takes static equivalence into account as shown in Figure 6, i.e. fFtu results
from the assumption that the whole compression is concentrated in the top fibre of the section.

Figure 6:

Simplified rigid plastic model adopted to compute the ultimate tensile strength in
uniaxial tension fFtu by means of the residual nominal bending strength fR3

Hence, the equation for fFtu and wu=CMOD3 is obtained, from the rotational equilibrium at
ULS, when a stress block in tension along the section is taken into account:
f Ftu = f R 3 3

(1)

For slab elements without conventional reinforcement with prevalent bending actions, the
resisting moment mR can be estimated by considering a rigid plastic relationship:

mR ,lab =

f Ftu .t f R 3 .t
=
2
6

(2)

where t is the thickness of the slab. By introducing the average properties of the concrete
(worst results from phase 2, see Table 3) and the thickness of the slab into Eq. (2), we can
calculate the flexural resisting moment mR = mR,lab = 42.7 kNm/m. This is what we basically
can calculate from the laboratory characterization without any safety factor. But we need to
take two specific SFRC properties into account in the real structure compared to the
characterisation in the laboratory: the fibre orientation and the redistribution capacity of the
full-scale structure. For this project, a reduction coefficient K=0.9 is used to mainly express
the size effect on the post-peak performances (RILEM TC 162-TDF, 2003). Moreover,
redistribution factor KRd = 1.3 was used (Di Prisco et al., 2014).
9

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

3.2

Comparison with experimental results and discussion

We used the yield line theory to estimate the load producing the ultimate flexural strength
of the slab. The yield line theory is a kinematic plastic method which is an upper bound
solution for the design problem (Johansen, 1962). Therefore, it is crucial to identify the worst
case in terms of cracking pattern i.e. the situation providing the lowest ultimate loadbearing
capacity. For this project, we use the work method to compute the ultimate load. The method
is based on the assumption that at failure, the external energy expended by the moving load
(We) must be equal to the internal energy dissipated by rotations around the yield lines (Wi).
This method is particularly suited for the design of flat slabs (Kennedy and Goodchild, 2003).
We first need to identify the actual yield lines pattern which causes the ultimate failure
of the slab at the end of the ULS 1 test. Two patterns were assumed to be relevant regarding
to the cracking pattern observed after the test. The first one (see Figure 7a) starts with a long
crack surrounding the internal columns (ultimate negative moment m) associated with radial
cracks at the bottom of the slab (ultimate positive moment m). The second one (see Figure
7b) is more global with straight negative yield lines in the centre of the panels and positive
yield lines along the column axes crossing the full slab on its length (i.e. 3 panel lengths are
activated by the concentrated load and the self-weight). For the ULS tests, the loads included
in the analysis are the peak concentrated load Pu = 328 kN and the self-weight q = 5 kN/m.
For the circular yield line pattern, the radius of the crack is R=4,24 m. Moreover, if the fibres
are uniformly distributed along the depth of the slab, we can assume that m=m=mR. Based
on a slab geometry survey, we also reassess the thickness of the slab to be 210 mm. This
depth was used in further calculations. Other yield line patterns have also been considered:
(1) circular pattern with more radial cracks at bottom, (2) local fan pattern around the internal
column, etc. but were not in agreement with the real cracking pattern observed or generated
larger ultimate capacity.
Circular & radial yield lines pattern

Global folded plate failure mechanism

SEC. A-A
A

A
SEC. B-B
SEC. A-A

(a)

Figure 7:

10

(b)

Yield line patterns used for the calculation of the ultimate capacity of the slab

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

R2

We = Pu + 4 q
4
3

We = Pu + 2 q3L y x
2 2

(3) & (4)

2 R

n + m2 R
Wi = 4 m'
4

Wi = m'3L y n + m2pRp

(5) & (6)

If we assume = .R and n = p = ,

If we assume = .Lx/2 and n = p = ,

Pu = (8 + 2 )m R

qR 2
3

m R , FS = m R = 29.6 kNm/m

Pu = 24m R

3qL2
2

(7) & (8)

m R , FS = m R = 25.5 kNm/m

If we compare this value with mR,lab = 42.7 kNm/m (even without any redistribution
factor), we can conclude that there is a large discrepancy between the laboratory
characterization and the full-scale tests. In order to confirm this, prisms were sawn in the slab
from the same direction (N-S), specifically in panel I. Due to logistic problems, this was the
only panel where specimens could be sawn from. These prisms exhibited dimensions
equivalent to the requirements of the EN 14651 except the height which was 210 mm. Based
on the analysis of 9 small beams with a notch depth of 36 mm, we found an average residual
strength fR3,m = 3.5 MPa (COV = 21%) meaning a reduction of almost 50% compared to the
initial lab characterization. This result stresses the heterogeneous distribution of the fibres in
the slab or specific orientation of the fibres in the prisms poured from phase 2. However, if
this residual strength is used to compute the mR,lab and then multiplied by the orientation
factor (0.9) and a redistribution factor (1.3), we find mR,lab = 30.1 kNm/m which is very close
to what we found from the full-scale test analysis.

Summary and Conclusions

Full-scale tests on an SFRC flat slab (with 70 kg/m) without conventional reinforcement
were performed in order to analyse the potential of SFRC in this field of applications. The
study presented in this paper aimed at supporting the design relationships given in the fib
Model Code 2010 and in particular the calculation of the ultimate bending capacity from a
full plastic approach. Moreover, the tests allowed us to confirm the ductile behaviour and the
redundancy potential in flat slabs. The project was achieved without any conventional rebars
in order to only focus on the steel fibres contribution. In real situations, anti-progressive
collapse rebars would be used from columns to columns to provide a robust structure. The
shear, punching and other design situations were out of the scope of this analysis. The
failures observed on the full-scale experiment were clearly associated with bending.
At the serviceability limit states, the slab presented cracks up to 0.5 mm, for an alternate
uniformly distributed loading condition, up to 4 kN/m. Until now, the model of the MC2010
however does not provide an analytical solution to calculate the crack opening for elements
reinforced by fibres only. This appeared to be a problem only if the slab is cracked under
service conditions. While the short-term maximum deflection observed at q = 2 kN/m
reached 1.2 mm without major cracks, we can however argue that the slab would be cracked
11

FRC 2014 Joint ACI-fib International Workshop


Fibre Reinforced Concrete: from Design to Structural Applications

at long term because of the creep (assuming a final creep factor of around 2.5 generates a
long-term maximum deflection beyond the value of 2.0 mm which was the deflection
observed at cracking). It means that long-term deflections should be calculated according to a
nonlinear model unless the concrete quality is better than the one used within this study.
For the ultimate limit state of the central panel, the use of the yield line theory combining
with the observed crack pattern allows to calculate the resisting moment of the full-scale
application. This result was firstly not in agreement with the ultimate bending capacity
computed from the characterization bending tests achieved in the laboratory on small prisms.
Flexural resistances obtained from samples directly sawn from the slab revealed however
lower post-peak properties. The use of these characteristics within the design guidelines from
the MC2010, together with a redistribution factor of 1.3, seemed to give a much better
agreement with the full-scale results.
Finally, research is still ongoing on companion specimens (round slabs with a diameter of
1700 mm and small round slabs with a diameter of 800 mm) to evaluate if the casting process
of the slab had so much influence in comparison to the casting of the small prisms in their
moulds.

References

BCA (2001), Flat slabs for efficient concrete construction, British Cement Association, UK.
Destre X. (2004), Structural application of steel fibre as only reinforcing in free suspended
elevated slabs: conditions - design - examples. Proceedings of the 6th RILEM
Symposium on Fibre-Reinforced Concretes (FRC) - BEFIB 2004, di Prisco M., Felicetti
R., Plizzari G. (Eds), Varenna, Italy, pp. 1073-1082.
Di Prisco M., Martinelli P. and Dozio D. (2014), The structural redistribution coefficient
KRd : a numerical approach for its evaluation. Submitted manuscript for Cement &
Concrete Composites.
(2013), fib Model Code for Concrete Structures 2010, fdration internationale du Bton (fib),
Lausanne, Switzerland.
Gossla U. (2005), Development of SFRC free suspended elevated flat slabs, Internal report,
Aachen University of Applied Sciences, Germany.
Johansen K. W. (1962), Yield-line theory. Cement and Concrete Association, London, UK.
Kennedy G. and Goodchild C. (2003), Practical yield line design. British Cement
Association, United Kingdom.
Michels J. (2009), Bearing capacity of steel fibre reinforced concrete. Ph.D. thesis, University
of Luxembourg, Luxembourg.
Michels J., Waldmann D., Maas S. and Zrbes A. (2012), Steel fibers as only reinforcement
for flat slab construction - Experimental investigation and design. Construction and
Building Materials, Vol. 26, No. 1, pp. 145-155.
RILEM TC 162-TDF (2003), Test and design methods for steel fibre reinforced concrete Recommendations for the - design method. Materials and Structures, Vol. 36,
No. 262, pp. 560-567.

12

Potrebbero piacerti anche