See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/311102224

Strengthening of concrete flat slabs with an overlaid reinforced concrete layer

Conference Paper · November 2016

CITATION READS

1 134

3 authors:

Hugo Fernandes Valter Lucio

New University of Lisbon New University of Lisbon
5 PUBLICATIONS   4 CITATIONS    103 PUBLICATIONS   219 CITATIONS   

SEE PROFILE SEE PROFILE

António Pinho Ramos

New University of Lisbon
73 PUBLICATIONS   238 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Ancient masonry structures View project

Punching of flat slabs View project

All content following this page was uploaded by António Pinho Ramos on 29 November 2016.

The user has requested enhancement of the downloaded file.

 Strengthening of concrete flat slabs with an overlaid reinforced
concrete layer

H. Fernandes (1), V. Lúcio (2), and A. Ramos (2)

(1) - PhD Student, Civil Engineering Department, Faculty of Sciences and Technology, Universidade
NOVA de Lisboa, Caparica, Portugal.
(2) - CERIS/ICIST, Civil Engineering Department, Faculty of Sciences and Technology, Universidade
NOVA de Lisboa, Caparica, Portugal.

Abstract

Strengthening of concrete slabs with a new reinforced concrete layer applied on the tensile face has
proved effective for service and ultimate limit states. Although it requires the consideration for the
premature failure, correctly designed it stands as an efficient and economical method for structural
strengthening.

Different detailing solutions for the rebars crossing the interface between the two concrete layers
were tested and analysed in order to identify the main factors that influence the behaviour of the
composite section. The knowledge from this study was applied to the strengthening of flat slabs,
where large stress concentration occurs at the column-slab connection. The usual brittle failure of
these zones by punching further influence the behaviour of the interface between the two concrete
layers.

This paper presents the study performed on concrete flat slabs strengthened with a bonded concrete
overlay. The experimental research consisted on four specimens loaded monotonically and
concentrically, and the comparison of the results with current codes, namely the Eurocode 2 and the
Model Code 2010.

Keywords: Structural strengthening, Concrete overlay, Reinforced concrete, Flat slabs.

   

1 
 
1 Introduction
The broad use of flat slabs as a structural solution has lead over the years to the development of
strengthening techniques for the cases where structural strengthening is required. Punching failure,
although a local phenomenon, can cause global failure of the entire structure due to the progressive
collapse of the remaining structure. Strengthening techniques for flat slabs in the column regions can be
embedded in the structure, with minimal protruding of its components, such as post installed steel bolts
or drop panels, or applied bonded to the structural elements, like bonded fibre reinforced polymers or
concrete overlays. The strengthening technique with a bonded concrete overlay (BCO) can be applied
externally or embedded within the thickness of the existing slab (Fig. 1). Either way this technique is
applied, a good bond to the existing structure is a requirement for the effective strengthening of the
latter. Concrete to concrete interaction has been widely studied when concerning the strengthening or
retrofitting of reinforced concrete structures [1], [2].
The interface between the two layers has been the subject mostly studied when applying new concrete
to an existing structure ([3], [4]), since its behaviour can govern the performance of the strengthened
structure. In order to guarantee structural integrity upon failure, debonding of the overlaid concrete must
be controlled to avoid a brittle debonding and global failure. Several solutions exist ([5], [6]), usually
concerning reinforcement crossing the interface, in order to stitch the two layers and guarantee ductility
when debonding occurs. A rough or very rough surface is the desired finish for a concrete surface to be
cast against, since it will promote interlocking and further activate the stitching reinforcement between
the two layers.
The combination of the aforementioned solutions can result in performance gains and increase in
structural strength, along with structural predictability as to the ductility of its behaviour. The practicality
of the BCO technique in punching strengthening of flat slabs is its applicability in the top face of the
slab, with minimal or even without formwork, as illustrated in Fig. 1. It is an unobtrusive solution, since
it can be easily levelled with the new flooring due to the reduced thickness (Fig. 1, left). The existing
rebar and the irrespective cover may allow the new layer to be cast within the thickness of the original
slab, according to Fig. 1 (right).

Fig. 1 Overlay level with the floor (left) or embedded in the existing slab (right).
The resulting cross section effective depth from adding a new reinforced concrete layer decreases the
slab deformability, and increases the bending and punching strength. To guarantee the anchoring of the
added reinforcement, the perimeter of the new layer must be increased from the punching control
perimeter in order to accommodate for the required reinforcement anchorage length. The reinforcement
for the new layer should be anchored at the edges of the BCO area. This accounts for a stitching solution
of the BCO technique that further improves the performance of the strengthening system (Fig. 2, left).
Another solution for stitching the two layers is the placing of shear connectors in the interface, arranged
so that stress is dispersed through the interface (Fig. 2, right). This stitching solution is coherent with
the punching strengthening solution with post-installed transverse reinforcement. Specific detailing of
this reinforcement could combine the advantages of both the interface and punching strengthening.
Detailing of these elements should then be performed according to the punching strengthening solution,
radially or orthogonally centered with the column.

2 
 
 Fig. 2 Stitching solutions for connecting the BCO to the substrate.

2 Experimental research

2.1 Test specimens

The experimental programme for assessing the performance of the strengthening method presented in
this work consisted of four reduced scale reinforced concrete flat slab specimens, strengthened with an
overlaid reinforced concrete layer in the region of the slab-column connection, organized as follows:
SQ-REF - Reference specimen with only reinforced concrete overlay and no interface reinforcement;
SQ-STC - Shear connectors distributed across the interface, embedded 80 mm with cement grout in
the existing substratum;
SQ-ANC - Longitudinal reinforcement of the overlaid concrete anchored in the ends by embedding 80
mm in the existing substratum, with cement grout;
SQ-STANC - Combination of the aforementioned detailings of the interface.
Due to the lifting phenomenon observed in several works and its inclusion in design codes ([6], [7]),
detailing with reinforcement crossing the interface near the new layer ends directly addresses this
phenomenon, controlling the interface crack opening.
The reduced scale specimen’s substratum comprised the area dimensions of 2300x2300 mm2, with 150
mm in thickness, illustrated in Fig. 3 (left). Longitudinal reinforcement of this layer was carried out with
16 mm bars, spaced every 100 mm in each direction.

Fig. 3 Detailing for the reference (SQ-REF) specimen, substratum (left) and overlay (right).
The overlaid concrete was defined at four times the slab effective depth from the column face, in order
to comprise enough area for the punching shear crack to develop and anchoring of the longitudinal
reinforcement through bonding to surrounding concrete. A 60 mm overlaid reinforced concrete layer
was cast with area dimensions of 1700x1700 mm2, which totals 210 mm thickness of strengthened slab,
with both layers illustrated in Fig. 3 (right). For this layer, the longitudinal reinforcement consisted on
bundled 10 mm bars, in order to guarantee minimum clearance and concrete cover in a reduced
thickness, also spaced every 100 mm in each direction.
Surface preparation of the substratum was performed with an electric chipping hammer and steel moil
point. The resulting surface roughness was performed for each specimen with concern that the moil tip
should not go deeper than 10 mm at each stroke, thus protecting the longitudinal reinforcement of the
substratum and the minimum surrounding concrete cover.

3 
 
2.2 Test setup
The punching test setup is illustrated in Fig. 4. Test specimens were monotonically loaded by a 1000
kN hydraulic jack (ENERPAC RCH-1003), positioned below the slab at the geometric centre.

Fig. 4 General test setup, real life (left) and scheme (right).
Loading was performed in the upwards direction by a steel plate with 200x200 mm2 area, and 50 mm
thickness. Eight steel plates with 100x100 mm2 in area and 20 mm thickness guaranteed kinematic
boundary conditions at the zero moment line of the specimens. These elements were linked to spreader
beams by 15.2 mm high strength strands, and the whole support setup was fixed to the laboratory strong
floor by high strength 40 mm threadbars. All elements are illustrated in Fig. 5.

Fig. 5 General dimensions for the test setup, top view (left) and side view (right).
Applied load was measured at the supports by eight TML CNC-200KNA load cells. Vertical
displacements were measured by fourteen 100 mm TML CDP-100 displacement transducers, thirteen
on the top face and one on the bottom face, and four 50 mm TML CDP-50 displacement transducers at
the supports, illustrated in figures Fig. 4 and Fig. 5. Relative horizontal displacement between layers
was also measured with four 50 mm TML CDP-50 displacement transducers. Data acquisition was
performed with Spider 8 and Centipede data-loggers, both from HBM, all monitored by HBM Catman
V6.0 software. Loading until failure of the specimens was electronically controlled by a WALTER+BAI
PKNS19D hydraulic pump, at a speed of 0.25 kN/s.

2.3 Materials characterization

Material characterization was performed for different rebar sizes, grout for anchoring the rebars crossing
the interface, concrete compressive and tensile strength for both layers, and pull-off strength at the
interface between layers. Rebar testing was performed with a INSTRON 8874 tensile test machine,
according to EN 10002-1 [8] guidelines, with test results for yield and tensile strength in Table 1. The
4 
 
mechanical properties of the cement grout were characterized according to RILEM PC-5 [9] and PCM-
8 [10] guidelines for compressive and tensile strength, respectively. Bond strength was also assessed
with pullout tests for the irrespective rebar sizes. Material parameters for the cement grout were a
flexural tensile strength of 9.7 MPa, a compressive strength of 78.8 MPa, and bond strength of 16.2
MPa. The concrete characteristics were evaluated according to EN 12390-3 [11] for the compressive
strength, and EN 12390-6 [12] for the splitting tensile strength. The former was carried out on 150 mm
cubic specimens, and the latter on 150x300 mm cylinders. Pull-off strength was carried out with
150x150 mm2, 60 mm thick top layer bonded to the base during casting of the overlaid concrete. Pull-
off tensile strength results can be observed in Table 2.
Table 1 – Strength characteristics for the steel bars.

Ø6 mm Ø10 mm
fy [MPa] 541.0 530.6
fu [MPa] 692.7 627.5
fy – mean yield stress of steel.
fu – mean tensile strength of steel.

Table 2 – Concrete and bond strength characteristics.

Steel connectors +
Reference Steel connectors Anchored rebar
Anchored rebar
(SQ-REF) (SQ-STC) (SQ-ANC)
(SQ-STANC)
fc,cube [MPa] 32.8 26.4 34.8 25.6
Substrate
fc,t [MPa] 2.9 2.6 3.0 2.3
Bond σpull-off [MPa] 0.85 1.04 0.82 0.80
fc,cube [MPa] 36.9 34.3 37.2 39.3
Overlay
fc,t [MPa] 2.9 3.2 2.6 2.8
σpull-off – mean value for the pull-off tensile strength of the interface between layers.
fc,cube – mean value for the compressive strength for concrete cubic specimens.
fc,t – mean value for the tensile splitting strength for concrete cylinders.

2.4 Failure modes

The failure modes were identified, with punching failure of the SQ-STC and SQ-STANC specimens,
and debonding of the top layer for the SQ-REF and SQ-ANC specimens. The complete debonding of
the top layer, illustrated in Fig. 6, allows for the visual observation of the resulting failure surface.

Fig. 6 Failure (left) and interface aspect (right) for the SQ-REF specimen.
During testing of the reference specimen, radial cracks could be observed on the top face of the overlaid
concrete, which could indicate a significant amount of stress being transferred to the top layer. When no
more deformation could be supported by the interface, debonding occurred. The cracks never evolved
5 
 
from hairline cracks, and closed right after debonding. Analysing the resulting surface after debonding,
one can observe that the top layer debonded clean from the bottom layer, with the original surface after
preparation with almost no visible damage. Removing the overlaid concrete layer, the punching shear
cone that developed in the substratum can be observed, illustrated in Fig. 6 (right).
The failure mode was similar for the detailing with the longitudinal reinforcement anchored in the
substratum (SQ-ANC), as observed in Fig. 7. Although major cracks reached the top layer, this was due
to the anchoring of the longitudinal reinforcement at the overlay edges, which allowed for the
deformation of the two layers together after debonding.

Fig. 7 Failure and crack pattern for the SQ-ANC detailing.

Detailing with shear connectors allowed for the punching shear failure to reach the top face of the
overlaid concrete, as illustrated in Fig. 8. It should be noted that the cracking pattern due to punching
never reached the edges of this layer.

Fig. 8 Failure and crack pattern for the SQ-STC detailing.

The most prominent tangential crack can be observed at two times the substratum effective height (≈0.25
m), and visually the damage reached a distance that was 5 cm from the overlaid concrete layer edges.
For the specimen with all solutions combined the failure mode followed that of the latter, with punching
failure reaching the top face of the overlaid concrete layer, as observed in Fig. 9. Similar to those
specimens, main tangential cracking was also at 0.25 m from the column face.

6 
 
 Fig. 9 Crack pattern for the SQ-STANC detailing.

3 Discussion

3.1 Experimental debonding and failure loads

Loading until failure was monotonic for all specimens, with debonding of the overlaid concrete for the
reference and SQ-ANC specimens, and punching failure for the other specimens. Failure for the latter
specimens was characterized by a residual resisting capacity for larger displacements about two times
that of the reference specimen. The relationship between load and deflection at the centre of each
specimen can be observed in Fig. 10.
SQ-REF slab SQ-STC slab
600 600
Load [kN]
Load [kN]

400 400

200 200

0 0
0 10 20 30 40 50 0 10 20 30 40 50
Deflection [mm] Deflection [mm]

SQ-ANC slab SQ-STANC slab

600 600
Load [kN]

Load [kN]

400 400

200 200

0 0
0 10 20 30 40 50 0 10 20 30 40 50
Deflection [mm]   Deflection [mm]  
Fig. 10 Load deflection relationship at the centre of the test specimens.
The load-deflection relationships presented show a higher residual load for specimens with
reinforcement crossing the interface, when compared to the reference specimen. Analysing the failure
loads one can verify that they have the same magnitude. Considering the predicted loads provided by
the Model Code 2010 [7] for the substratum alone, the load increment due to strengthening with a
concrete overlay was greatest for the specimens with shear connectors, as observed in Fig. 11. Concrete
strength was considered for these calculations as 0.8 of the fc,cube listed in Table 2. A 65 % increment for
the SQ-STC detailing and 62 % increment for the SQ-STANC detailing were the largest values
7 
 
registered. This phenomenon attests the good performance for the detailing of the shear connectors
stitching the two layers.

600 559,8 kN 567,6 kN 549,8 kN

535,8 kN

fc=29.5 MPa

fc=29.7 MPa
fc=27.5 MPa
500

fc=31,5 MPa
+ 46 % + 65 % + 36 % + 62 %
400
Load [kN]

300

fc=27.9 MPa
fc=26.2 MPa

fc=21.1 MPa

fc=20.5 MPa
200

100

0
SQ-REF
1 SQ-STC
2 SQ-ANC
3 SQ-STANC
4
Overlay Substratum
Fig. 11 Experimental failure loads.

3.2 Comparison with design codes

The maximum loads attained were then compared with current design codes, namely the Eurocode 2
(EC2) [13] and the Model Code 2010 (MC2010) [7] shear provisions for slabs without shear
reinforcement. The former presents an empirical design equation for estimating the punching shear
strength of flat slabs, calculated according to:
. /
100 (1)

The main terms considered in this approach are concrete characteristic compressive strength in cylinder
specimens ( ), reinforcement ratio ( 0.02), and size effect ( 1 200/ 2.0).
To assess the experimental values from testing, the latter limiting values for and were not
considered, and the concrete strength considered was the average compressive strength on cylindrical
/
specimens. The minimum punching stress is quantified as 0.035 / .
The design shear force calculation provided in the MC2010 is based on a physical model (Critical Shear
Crack Theory – CSCT), and therefore differs from the latter design code. The punching shear strength
according to CSCT is calculated like:
/
⋅ ⋅ 2
⋅ ⋅

This model allows accounting for slab rotation ( ), which rules the shear crack opening ( ⋅ ), and
is calculated according to the level of approximation required for the analysis or design. Another
important factor accounted for is the shear crack roughness. This term rules the interlocking
phenomenon responsible for stress transfer in the shear crack, depending on the aggregate size ( ) and
reference aggregate size ( ).
When calculating the predicted punching capacity using equations (1) and (2), concrete strength
considered was that of the substratum (lower strength). The control perimeter considered for the loaded
area is one of the main differences between the two design codes. On the EC2 [13], the control perimeter
( ) is set at a distance of two times the effective depth of the slab, whilst on the MC2010 [7], the basic
control perimeter ( ) is set at a distance of half the slab’s effective depth. The effective depth for a
composite cross section such as the one presented in this work is not an objective measure, but can vary
between the effective depth of the substrate ( ) and overlay ( ) rebars. A quantitative measure ( )
can be calculated for this value, with the contributions of each rebar layer weighted on the irrespective
layer stiffness, as follows:
8 
 
 , ,
3
, ,

Where , and , are the substrate and overlay rebars cross sectional area. Accounting for the values
of each reinforcement cross-sectional area and irrespective effective depth, yields an equivalent effective
depth of approximately 0.138 m for all specimens. Reinforcement ratio also accounting for the
equivalent effective depth of the rebar layers and the total cross-sectional area yields a value of 0.026
for all specimens. The punching strength calculated according to each design code is presented in Table
3, where the Level of Approximation III was considered for the MC2010 approach. The strength of the
substrate layer alone is also presented to clarify the strength increase of the specimens.
Table 3 – Punching shear loads for the test specimens
Substrate layer Strengthened cross-section
EC2 MC2010 Test load EC2 EXP/EC2 MC2010 EXP/MC2010
[kN] [kN] [kN] [kN] [-] [kN] [-]
SQ-REF(*) 382,6 363,4 559,8 566,4 0,99 580,5 0,96
SQ-STC 343,1 337,9 567,6 526,7 1,08 520,6 1,09
SQ-ANC(*) 394,5 370,9 535,8 578,1 0,93 598,6 0,90
SQ-STANC 338,4 334,9 549,8 522,0 1,05 513,5 1,07
(*) debonding failure

Punching capacity provisions on both design codes yielded results that are more conservative for the
punching capacity when no debonding of the overlaid concrete occurred (SQ-STC and SQ-STANC).
The opposite could also be observed where the debonding phenomenon ruled the behaviour of the
specimens upon failure (SQ-REF and SQ-ANC). A slightly more conservative MC2010 [7] could be
observed over the EC2 [13] for the former specimens, and for the substrate without strengthening
predicted loads.

4 Conclusions

The test results presented allowed assessing the performance of flat slabs to punching loads,
strengthened with an overlaid concrete layer on the top face. Placing a new reinforced concrete layer on
top of an existing slab increased the punching strength of the test specimens, when compared with code
provisions for a slab without strengthening.
The MC 2010 [7] code provisions allowed for more conservative load predictions than compared to the
EC2 [13], both for the substrates and the full composite cross-section when no debonding occurred. The
latter cross-section comprised two reinforcement layers working together, which led to the estimation
of an equivalent effective depth that could represent the theoretical effective depth for a monolithic
cross-section. This value was weighted between the two rebar effective depths, yielding the tensile
strength of the cross-section below the interface. The latter statement implies the interface dependent
performance of the composite cross-section.
Concrete strength also played a major role in structural performance. The little variation in terms of
punching test loads attest how concrete strength affected the results, largely varying its compressive
strength through the tests. Considering both the MC 2010 [7] and EC2 [13] punching provisions, for
specimens with the higher concrete strength punching load was overestimated, as for the specimens with
the lower strength concrete punching load was underestimated.
More tests are required to attest the behaviour of flat slabs strengthened with an overlaid reinforced
concrete layer. Variations in the detailing of the interface can also affect the overall behaviour of the
strengthened structure. This should be assessed through varying the parameters such as cross-sectional
area, and placing this reinforcement nearer or further from the column face. Numerical validation
through non-linear analysis is also required to assess the stress distribution at the interface due to the
punching phenomenon.

9 
 
 5 References

[1] Randl N.; Design recommendations for interface shear transfer in fib Model Code 2010. Struct
Concr, 14:230–41, 2013.
[2] Santos PMD, Júlio ES.; Interface shear transfer on composite concrete members. ACI Struct J,
111:113–21, 2014.
[3] Wicke F, Randl M, Münger F, Wick M.; Design of Shear Transfer in Concrete-Co Principality
of Liechtenstein University of Innsb structures at Innsbruck Summary For the design of
concrete composite constructions, 2014.
[4] Walraven J, Frenay J, Pruijssers A.; Influence of concrete strength and load history on the
shear friction capacity of concrete members. PCI J, 32:66–84, 1987.
[5] Julio ES, Branco F, Silva VD.; Structural rehabilitation of columns with reinforced concrete
jacketing. Prog Struct Eng Mater, 5:29–37, 2003.
[6] Bissonnette B, Courard L, Fowler DW, Granju JL.; Bonded Cement Based Material Overlays
for the Repair , the Lining or the Strengthening of Slabs or Pavements: State-of-the-Art Report
of the RILEM Technical Committee 193-RLS, 2011.
[7] fib Model Code for Concrete Structures 2010. Ernst & Sohn, 434p., 2013.
[8] European Committee for Standardization; EN 10002: Metallic materials—tensile testing—part
1: method of test at ambient temperature, 2001.
[9] RILEM Committee TC113; PC-5: Method of test for compressive strength of polymer concrete
and mortar, 1995.
[10] RILEM Committee TC113; PCM-8: Method of test for flexural strength and deflection of
polymer-modified mortar, 1995.
[11] European Committee for Standardization; EN 12390-3:2001 - Testing hardened concrete Part 3
- Compressive strength of test specimens, 2001.
[12] European Committee for Standardization; EN 12390-6:2000 - Testing hardened concrete Part 6
- Tensile splitting strength of test specimens, 2000.
[13] European Committee for Standardization; EN 1992-1-1: Eurocode 2-Design of Concrete
Structures - Part I: General Rules and Rules for Buildings, Brussels, 2004.

10 
 

View publication stats