Engineering Structures 125 (2016) 8090

Contents lists available at ScienceDirect

Engineering Structures

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

Experimental study on the shear behaviour of precast concrete hollowcore slabs with concrete topping

http://dx.doi.org/10.1016/j.engstruct.2016.06.0050141-0296/ 2016 Elsevier Ltd. All rights reserved.

Corresponding author.E-mail addresses: eeznie@yahoo.com, iznisyahrizal@utm.my (I.S. Ibrahim).

I.S. Ibrahim a,, K.S. Elliott b, R. Abdullah c, A.B.H. Kueh c, N.N. Sarbini ca Forensic Engineering Centre, Institute for Smart Infrastructure and Innovative Construction (ISIIC), Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310 JohorBahru, Johor, Malaysiab School of Civil Engineering, University of Nottingham, University Park, Nottingham NG7 2RD, UKc Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310 Johor Bahru, Johor, Malaysia

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

Article history:Received 24 July 2014Revised 2 June 2016Accepted 6 June 2016

Keywords:Hollow core unitConcrete toppingComposite actionShear-flexure capacityInterface slip

In typical precast construction practice of floor slabs using precast concrete hollow core unit (HCU), in-situ concrete is cast on top of the HCU to obtain smooth and even floor finish. The surface of the HCU isseldom given proper treatment prior to casting the concrete topping. The texture and surface moisturecondition of the HCU just before receiving concrete topping may affect the overall strength of the slabswhen the concrete topping and the HCU act compositely during service. This paper presents the exper-imental study on shear-flexure capacity of composite slabs using HCU and concrete topping. Full scalethree point load test are carried out on 14 composite slab specimens with different surface roughnessand surface condition of the HCU before casting the concrete topping. The surface roughness consideredis smooth and rough, while the moisture conditions are dry, ponded and optimum wet. The effect of thelongitudinal joint between the HCU panels is also considered. The experimental results are also comparedwith predicted values using the available equation in Eurocode 2 and an equation published by a previousresearcher. The results of the experiment show that the HCU surface condition and longitudinal jointaffect the stiffness and shear-flexure strength of the slabs. The optimum HCU surface condition whichcan produce highest stiffness and shear strength is rough and wet conditions, while the longitudinal jointbetween HCU panels reduces the slab shear strength. The interfacial horizontal shear is not the factor thatgoverns the strength and behaviour of the slabs. The equation available in Eurocode 2 gives non-conservative prediction of the shear strength. In contrary, the equation published by the previousresearcher gives conservative prediction of the shear strength.

2016 Elsevier Ltd. All rights reserved.

1. Introduction

Precast concrete slab system for buildings that is widely avail-able in the market includes hollow core unit, double-tee, solidcomposite plank & beam, and composite plank. They offer speedyconstruction, reliable and reduce construction cost particularlyfor the construction of suspended floors. The system can be madewith variable lengths and is suitable for used in all types of build-ings. In most construction practices, cast in-situ concrete toppingsare added onto the concrete precast slab for the purpose of makingsmooth and even floor finish. Besides, the concrete topping canalso enhance the structural performance of the precast slab byproducing a composite structural system. Typically, the in-situ

concrete toppings are 40100 mm in thickness, and contain a smallamount of steel reinforcement, usually a prefabricated weldedmesh to control shrinkage. The concrete topping with the strengthranges from 25 to 40 N/mm2 are laid onto the aged precast slabunits. The most popular precast concrete slab system is the pre-stressed precast Hollow Core Units (HCUs). The HCU is manufac-tured using automated semi-dry extrusion where the finalproduct is high strength concrete.

Each year the UK industry constructs around 30 m of compos-ite hollow core floors slabs with no bona fide information abouttheir design, surface preparation and construction. Relative move-ment between the wet cast concrete topping and the HCU, theinjudicious placement of mesh reinforcement, and the presenceof construction joints may cause delamination, edge restraint, cur-vature and loss of serviceability (see Fig. 1). Ultimate failure modescould be brittle, especially on the precast prestressed floors thathave a high strength-stiffness ratio.

http://crossmark.crossref.org/dialog/?doi=10.1016/j.engstruct.2016.06.005&domain=pdfhttp://dx.doi.org/10.1016/j.engstruct.2016.06.005mailto:eeznie@yahoo.commailto:iznisyahrizal@utm.myhttp://dx.doi.org/10.1016/j.engstruct.2016.06.005http://www.sciencedirect.com/science/journal/01410296http://www.elsevier.com/locate/engstruct

(a) Overall problems

(b) Surface roughness related problems

Strand relaxation

Shrinkage

Debonding Interface

Surfaceroughness (see (b))

Creep

Tension

arrangementMesh

HCU

toppingConcrete

Load

shear stress

Concrete topping

HCU

Ponding waterDebris & dust

Roughness

Fig. 1. Problems related to concrete topping construction.

I.S. Ibrahim et al. / Engineering Structures 125 (2016) 8090 81

Typical work specification does not address the proper surfacepreparation of the HCU before casting of the concrete topping.Delays can also occur by not knowing when the conditions areright for laying the concrete topping. Consequently, the contractoroften neglects the surface preparation of the HCU during concretetopping construction. Improper surface preparations may causeproblems to the interaction between the concrete topping andthe HCUs. This may affect the overall structural behaviour andstrength when composite action is expected from both. Someattempts to quantify surface textures on its relation with the inter-facial shear in the composite floor structures are given in theFdration Internationale de la Prcontrainte (FIP) [1] document.

0.8

1.0

1.2

reng

th (N

/mm

2 )

BS 8110 (Smooth)

BS 8110 (Brushed)

BS 8110 (Washed)

FIP (Category I-II)

FIP (Category III-VI)

2. Related works

When concrete topping is cast onto the HCU without mechani-cal devices to strengthen the interfacial connection, the system inflexure may behave either partially of fully composite. Partial com-posite is obtained when the slippage occurs at the interface of thetwo concrete components in the horizontal direction as shown inFig. 2. The incomplete interfacial interaction may occur in the sys-tem where the two concrete components are cast at different timeand surface condition of the HCU is not sufficient to provide resis-tance to horizontal shear force. Horizontal shear transfer along theinterface between the HCU and the in-situ concrete topping is anessential requirement to ensure composite action of the twomembers.

Codes of Practice such as ACI 318 [2], BS 8110 [3] and Eurocode2 [4] specify that the interface shear strength between the concrete

Load kN/m

HCU

Concrete toppingInterfaceshear stress

Fig. 2. Horizontal shear stress along the interface of a composite member bent inflexure.

topping and the precast unit depends on two fundamental param-eters; (i) the surface roughness of the precast unit and (ii) thequantity of shear steel. In common construction practice usingHCU, there is usually no shear steel or mechanical device providedto connect between the concrete topping and the HCU surface. Insuch a case, the interface shear strength relies on the surfaceroughness of the HCU which provides natural friction betweenthe two materials. Fig. 3 shows comparison of interface shearstrength from different codes of practices, namely BS 8110, FIPand Eurocode 2. It can be seen from the figure that all codes con-sider the interface shear strengths to vary depending on the con-crete strength as well as on the different types of surfaceroughness of the precast units. The ACI 318 code specifies the min-imum interface shear strength value for 6.4 mm roughened surfaceas 1.79 N/mm2, a far higher value compared to other internationalcodes. Furthermore, ACI 318 does not consider the concrete com-pressive strength as a factor influencing the interface shearstrength.

Another significant difference between the codes is the cate-gorisation of the degrees of roughness. BS 8110 merely states thetype of instrument used to create the roughness, whereas Eurocode2 assigns measurable properties, i.e. 3 mm for rough surfaces and5 mm or greater for indented surfaces.

The FIP Guide to Good Practice [1] identifies ten categories ofthe type of surface which a precast unit may have, prior to receiv-ing the in-situ concrete. The categorise are based on the end pro-duction of the precast unit in terms of smooth and roughsurface despite of the difficulties in distinguishing the two cases.Within the FIP Commission itself there is a popular theory thatthe smooth but clean surfaces have better overall bond than rough-ened which is often dusty and dirty surfaces where localised bondfailures occur. FIP [1] recommends that contaminants should beremoved either by water flushing, compressed air or vacuumcleaning. Sweeping is not sufficient as it could not remove the dustcompletely especially on rough surface. Surface treatment is alsoneeded to control the moisture of the precast surface, because;

(a) If the surface of the precast member is very dry beforereceiving the concrete topping, it will absorb water fromthe in-situ concrete. As a result, the concrete near the inter-face that govern the interfacial shear capacity may havedegraded.

(b) If the surface is very wet, i.e. ponding, the water-cementratio at the interface will be very high, resulting in weakbond strength in the immediate strata.

When the surface pores are fully treated, it is said to be surface-dry and saturated (wet condition). If the precast surface unit was

0.0

0.2

0.4

0.6

25 30 35 40

Inte

rfac

e sh

ear

st

Concrete strength (N/mm2)

FIP (Category 2)

EC2 (Very smooth)

EC2 (Smooth)

EC2 (< 3mm rough)

EC2 (5mm rough)

ACI (Roughened)

Fig. 3. Comparison of interface shear strength with various codes of practice.

82 I.S. Ibrahim et al. / Engineering Structures 125 (2016) 8090

left to stand free in dry air before the casting of concrete topping,some of the water contained in the unit will evaporate and surfacewill be less saturated, i.e. air-dry condition. Prolonged drying in anoven or in a closed hot compound would reduce the moisture con-tent in the concrete until no moisture is left (bone-dry). This con-dition is not achievable for large scale production of precast unitsand therefore is not further discussed in this paper. For an extremecondition, surface moisture is left ponding, making it saturated andmoist (ponded condition). The various stages of surface conditionsare shown in Fig. 4.

The bond between HCU and the concrete topping is essentialand has to be checked in design, and ensured in construction. Scott[5], Ros et al. [6], Ueda and Stitmannaithum [7], Girhammar andPajari [8] carried out tests on HCU with concrete topping to studythe interaction behaviour up to the ultimate load capacity. The testparameters included surface roughness, load transmissionbetween a group of two slabs, pre-stressing force, tension rein-forcement ratio, shear span-to-effective depth ratio and concretetopping depth. In 2007, extensive work was carried out by Ajdukie-wicz et al. [9] with two different test setups; (i) short-term loadingsubjected to instantaneous bending tests until failure and (ii)6 months long-term loading followed by ultimate bending testsuntil failure. All the aforementioned researchers had come up withthe conclusion, that the ultimate load of HCU with concrete top-ping increased between 10% and 42% compared to the HCU alone.Further experimental test was then carried out by Adawi et al. [10]in 2015 to study the composite action between HCU and concretetopping. The experimental work includes pull-off, push-off and fullscale tests. They concluded that HCU with machine-cast surfacecan be considered to act compositely with the concrete toppingeven though in the current North American design codes statedthat it should not be considered. Baran [11] also found that theexistence of concrete topping resulted in improvements in thecracking moment and initial stiffness of HCUs.

Ueda and Stitmannaithum [7] compared the test results withthe predicted shear cracking strength based on the cracking pat-tern, i.e. web-shear cracking and flexure-shear cracking. Theweb-shear cracking strength is predicted using elastic analysis,which is the conventional method for ordinary prestressed con-crete member. As for the flexure-shear cracking strength, theycan be predicted using equations given in the ACI 318-83 [12]and the JSCE specification [13]. The ultimate shear strength, Vuwas further compared using the proposed equation by Niwa [14]and Okamura et al. [15]. Niwas formula for ordinary reinforcedconcrete beams without shear reinforcement where a/d 6 2.5 isgiven in Eq. (1) and Okamuras modified formula for ordinary rein-forced concrete slender beams without shear reinforcement wherea/d > 2.5 is given in Eq. (2):

Vu 0:244f 02=3ck 1ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi100qw

p 1 3:33w

d

11 a=d2

!bwd

if a=d 6 2:5 1

Absorbed water

(Dry)

Concrete topping

HCU

Bone-dry SAir-dry

HCU

Concrete topping

Fig. 4. Conditions at the interface w

Vu 0:20f 01=3ck 100qw13

1000d

14

0:75 1:4a=d

bwd if a=d > 2:5

2where fck is the cylinder compressive strength of the concrete top-ping, qw = As/bwd is the ratio of tensile reinforcement, As is the areaof tensile reinforcement, w is the distributed applied loading, bw isthe width of web, d is the effective depth of the tension reinforce-ment and a is the shear span. Both Eqs. (1) and (2) were obtainedempirically by best fit analysis after analysing the shear resistancemechanism carefully.

Ueda and Stitmannaithum [7] suggested that the equationsused to calculate the ultimate shear capacity of ordinary reinforcedconcrete beams without shear reinforcement are also applicable tothe HCU with concrete topping by using the strength of concretetopping and its width in the prediction. However, the applicationis limited to the condition that the size of hollow core is relativelysmall compared to that of an entire composite section.

In typical precast construction practice of floor slabs using pre-cast concrete HCU, the concrete topping thickness is usually in therange of 40100 mm, which is always lesser than the precast unitthickness (usually P150 mm). Because of the lesser thickness ofthe concrete topping, the failure mode that governs the slab beha-viour and strength is most unlikely the horizontal shear failure atthe interface between the precast unit and the concrete topping.This is due to the fact that the neutral axis of the system will occurin the precast unit rather than in the concrete topping. As a result,the interface position will be in the compression region underbending and hence, most unlikely will govern the failure mode.

Even though the interfacial shear is unlikely the controlling fac-tor in precast slabs with typical thickness, significant numbers ofresearches had reported on the shear interaction behaviourbetween the precast unit and the concrete topping. The findingsby Baran [11] concluded that the measured and computed inter-face shear strength to be significantly lower than the values spec-ified by the ACI [12] and AASHTO [16] specifications. Lesser workhad been discussed on the influence of other factors such as thesurface condition of the precast unit. Therefore, this research isconducted to study the effects of surface conditions of HCU namelythe surface roughness and moisture content on the behaviour andvertical shear strength of composite slabs made with HCU and in-situ concrete topping. The surface roughness considered is smoothand rough, while the moisture conditions are dry, ponded and opti-mum wet. Besides that, the effect of the longitudinal joint betweenHCU is also observed and the prediction of the shear strength usinga proposed modified BS 8110 and a published equation is alsodone.

3. HCU specimen

The HCU used in this study is obtained from a local manufac-turer. The nominal cross section dimensions are 1.2 m wide and

Free water

(Wet) (Ponded)Saturated and moist

HCU

Concrete toppingConcrete topping

HCU

urface-dry and saturated

ith different surface treatments.

I.S. Ibrahim et al. / Engineering Structures 125 (2016) 8090 83

150 mm deep with circular hollow cores of 100 mm diameter asshown in Fig. 5. The length is 4.3 m. The actual dimensions of thespecimens are also measured and given in Table 1. Other data sup-plied by the manufacturer are as follows: The units are pre-tensioned with nine number 7-wire helical strands of 9.3 mmdiameter, placed with 40 mm bottom cover. They were pre-tensioned to 70% of the ultimate strength, fpu = 1770 N/mm2 beforethe concrete was cast using dry mix extrusion techniques. Themaximum aggregate size was 10 mm for both the unit and con-crete toppings. The HCU was 4.3 m long, producing design service

1381381381381197

115.5

152

13813813846.5 137.5

Fig. 5. Original HCU sectio

Table 1Specimen name, test parameters, specimen dimensions and concrete toppings properties.

Specimen HCU depth,hs (mm)

Concrete toppingdepth, ht (mm)

Interface width,bv (mm)

Wb

Smooth Dry 1SL 153 76 1170 31SR4SL 153 81 1189 34SR

Ponded 2SL 154 76 1187 32SR5SL 152 81 1170 35SR

Wet 3SL 153 77 1190 33SR6SL 153 77 1200 36SR

Rough Dry 1RL 153 78 1176 31RR4RL 153 78 1190 34RR

Ponded 2RL 156 75 1180 32RR5RL 155 76 1184 35RR

Wet 3RL 152 80 1192 33RR6RL 155 76 1190 36RR

Control Rough 156 77 1160 3Control Smooth 154 75 1160 3

S = Smooth and R = Rough.L = Left end section and R = Right end section.

a Average of 3 tests.

Longitudinalconnections 37

152

520

Fig. 6. Cross section of the tested s

and ultimate bending moment of 45.22 kN m and 69.13 kN m,respectively, and an ultimate shear resistance of 96.98 kN (givenby the manufacturer). The average concrete compressive cubestrength at 28 days for 3 samples is 76.8 N/mm2.

Twelve numbers of HCUs are cut longitudinally along the corewithout affecting the strands to produce two units of 520 mmand 665 mm wide slabs. Both units are then place side by side asshown in Fig. 6. The purpose of constructing the slab specimenby joining the HCU specimens longitudinally is to observe theeffect of the longitudinal joint as found in the actual construction

115.5138138138

47.5137.5138138138

n (dimensions in mm).

eb width,w (mm)

Average corebreadth (mm)

Concrete cube compressivestrength, fcu0 (N/mm2)a

Concrete cylindersplitting tensilestrength, ft (N/mm2)a

38 99 29.3 4.41

63 98 34.7 4.18

69 97 30.2 4.63

54 97 31.7 4.33

82 97 32.5 3.76

89 96 33.0 3.96

78 96 47.0 4.46

87 95 32.8 4.07

76 97 47.5 4.89

82 96 32.7 4.12

82 96 48.0 4.43

88 96 34.5 3.23

70 96 30.1 4.5669 95

Nominal

665

pecimens (dimensions in mm).

84 I.S. Ibrahim et al. / Engineering Structures 125 (2016) 8090

practice. Two numbers of the slabs using uncut HCU as shown inFig. 5 are used as control specimens where comparison with thejointed specimens will be made.

4. HCU surface roughness and moisture condition

Two types of surface roughness including the controls for theHCU are prepared, that is smooth and roughened. The smooth sur-face is the original as-cast while the roughened surface is obtainedby scraping using steel brush. The surface roughness is then mea-sured along a sampling length of 200 mm using an instrumentdeveloped by Bensalem [17].

For each surface roughness, three different surface moisturecontents are prepared before casting the concrete topping. Theyare dry, optimum wet and ponded conditions. The dry surface con-dition is obtained by leaving the HCU in the ambient temperature.The optimum wet, which is the preferred condition as mentionedin the FIP document [1], is obtained by spraying the HCU surfacewith water and leaving the units to dry under ambient tempera-ture until the surface colour of the HCU turned to light dark greywith no access surface water found on it. The ponded conditionis obtained by pouring water until the excess water is left on thesurface with a depth of approximately 1.8 mm. Before getting thecorrect amount of water, a trial was carried out to produce theoptimum wet and ponded conditions on a 1 m 1 m surface area.The time required for the water to evaporate was recorded for thesurface to change to the light dark grey for the optimum wet con-dition. From the trial, it was found that optimum wet conditionrequired 9.5 l/m2 of water while the ponded condition used 45 l/m2 of water. The HCU surface moisture was also measured usinga moisture meter, where the readings were 1213% for the drycondition, 2628% for optimum wet condition and 4143% forponding condition. The results from the trial studies were thenused as a guide for the preparation of the required surface condi-tions of the HCU before casting the concrete topping.

Vertical slippotentiometer

200

75

50 RHSSpherical bearing

Load jackLoad cell

5. Concrete topping

The top surfaces of the HCU are cleaned thoroughly by vacuum-ing to remove any concrete laitance and debris. For rough surfacecondition, the surface of the HCU are scratched using steel brushbefore vacuuming. Prefabricated mesh of type A142 (R6-200 mm)with 25 mm cover was laid on the top surface of the HCU andlapped 300 mm in the mid-span region [1]. Then the concrete top-ping is cast using ready mix concrete in 5 batches. The maximumsize of concrete coarse aggregate is 20 mm. The concrete slumpis 50 mm and the mix conforms to BS 8500: Part 2 [18]. The con-crete compressive cube strength, fcu is designed to achieve 30 N/mm2 at 7 days. The concrete topping and the longitudinal jointare cast monolithically. The nominal depth of the concrete toppingis fixed at 75 mm throughout the 4 m effective length. At bothends, the HCU surface is sprayed for 150 mm to allow for the fixingof potentiometers for measuring slips. After concreting, all speci-mens are cured by covering with damp gunnysack at room temper-ature. Water is sprayed every day until the test day to keep theconcrete nearly saturated. The average strength of the concrete

Fig. 7. Cross section of the slab specimen.

topping is obtained from compression test of three 100 mm cubestogether with the cylinder splitting strength, fct on the slab testday. The slab specimen cross section is shown in Fig. 7.

6. Testing procedure

The specimens are subjected to static three-point bending on asimple span of 4 mwith roller supports at both ends (Fig. 8). Mean-while, the load is applied at 562 mm from the support which isequal to 2.5 times the total depth of the composite section. Theintention of the test is to obtain vertical shear failure. The locationof the applied load is chosen so that the shear capacity of the slabspecimens would be independent of the shear span-to-depth ratioa/d > 2.5 [19]. The vertical deflection, interface (horizontal) slip andvertical slip (or interface dilation) are measured using potentiome-ters. However, since there is no movement of vertical slip through-out the test, they will not be discussed in this paper. The verticaldeflection is measured at two positions below the line load, i.e.one at each slab segment. The interface slip is recorded by measur-ing the horizontal movement of concrete topping relative to theHCU using potentiometers that are attached to the HCU. Initially,the loading is applied with the increment of 10 kN. The readingis recorded 1 min after each load increment. After the first crackingis observed, the load increment is reduced to 5 kN until failure. Theslab is considered failed when it could no longer sustain the loadincrement. The load is removed before the slab undergoes total col-lapse. The final cracking is recorded after the removal of load andthe results of the test at first end are marked as left (L) end. Afterthe load is removed, it was observed that the slab is still in sturdycondition. Because of this condition, we are able to repeat the loadtest on the same slab by applying the load at the other end of theslab. The results of the tests at second end are marked a right (R)end. It is observed that the test results for left (L) and right (R)end loading for all specimens are almost equal.

7. Test results

The ultimate applied loads (including the self weight of the slaband test equipment), ultimate shear force, interface slip at failureand types of failure are shown in Table 2. It should be noted thatthe test at right end of the slab is conducted after the slab had par-tially damaged, but still in sturdy condition, due to the prior test atthe left end. Hence, it is prudent to compare the variation of theright end results from the left end ones. Comparison of the ulti-mate load between left and right end tests show that six slabshad sustained higher ultimate loads at the right end tests than atthe left ends ranging from 0.44% to 20.8%, while six other slabsshow lower ultimate loads at the right end tests than at the leftends ranging from 3.38% to 6.33% (see Table 2). Because the differ-

L = Left end and R = Right end

potentiometer 152Horizontal slip

40004300

Deflection potentiometer

562

SpecimenLH RH

Fig. 8. Shear test setup (dimensions in mm).

Table 2Test results.

Specimen Ultimate loadcapacity,Pult (kN)

% Difference of ultimateload at R endfrom ultimate load atL ends

Ultimateshearcapacity,Vult (kN)

Interface slip at ultimateload,ds,ult (mm)

Interface slipat failure,ds max (mm)

Failure types(WC =Web cracking& SF = Shear-flexuralfailure)

Smooth Dry 1SL 191.3 2.67 155.8 1.25 1.52 SF1SR 196.4 160.2 0.64 0.81 SF4SL 186.3 3.52 151.7 1.31 1.46 SF4SR 193.1 157.6 1.10 1.32 SF

Ponded 2SL 157.9 0.44 127.3 0.97 1.38 SF2SR 158.6 127.9 0.14 0.57 WC5SL 205.5 4.33 168.2 2.65 2.68 SF5SR 196.6 160.5 2.11 2.29 SF

Wet 3SL 209.9 3.38 172.1 0.27 0.41 SF3SR 202.8 166.0 1.20 1.45 WC6SL 219.5 6.33 179.9 2.95 3.49 WC-SF6SR 205.6 168.0 1.11 1.24 WC-SF

Rough Dry 1RL 187.1 11.92 152.2 0.38 0.50 WC1RR 209.4 171.3 0.45 0.89 WC4RL 203.5 14.40 167.0 0.20 0.62 SF4RR 232.8 192.1 0.04 0.50 SF

Ponded 2RL 222.6 20.80 183.2 0.33 0.49 WC2RR 268.9 223.0 0.07 0.07 SF5RL 265.7 7.00 219.9 1.73 1.86 WC-SF5RR 284.3 235.9 0.04 0.05 SF

Wet 3RL 317.9 5.00 264.9 0.06 0.06 WC-SF3RR 302.0 251.3 0.31 0.95 SF6RL 274.1 5.11 227.1 0.08 WC-SF6RR 260.1 215.1 0.02 0.02 WC-SF

Control Rough 328.6 273.1 0.01 0.02 SFControl Smooth 311.1 258.7 0.02 0.13 SF

S = Smooth and R = Rough.L = Left end section and R = Right end section.

I.S. Ibrahim et al. / Engineering Structures 125 (2016) 8090 85

ence is uniform with the same number of slabs having larger ulti-mate loads at either end, we can conclude that the data for bothends are valid and acceptable for further analysis.

Fifteen tests showed the slabs failed by combinations of verticalshear and flexure (SF) cracking. The shear crack occurred across theslab width which extended from underneath the point load anddiagonally spread towards the support as shown in Fig. 9. At thesame time, several numbers of flexural cracks formed near theloading in the longer span region. Six tests failed by combinationof shear-flexure and web-cracking (WC-SF), and five failed byweb cracking (WC) only. Web cracking as observed here is thecrack in the concrete mass either above or at the side of the hollowcores and spread along the longitudinal length of shear span(Fig. 10). Web cracking also occurred along the longitudinal joint.Crack along the interface between concrete topping and the HCUsurface is also observed (Fig. 9). In general, there is no particularcrack type that can be associated with any surface condition ofthe HCU. As such failure mode cannot be associated with any par-ticular surface roughness or surface moisture condition of theHCUs. The most dominant failure type is shear-flexure (SF).

Shear-flexure crack

Separation of concrete topping from HCU

Fig. 9. Shear-flexure crack and separation of concrete topping from HCU (WC-SF).

Graphs of shear force versus interface slip are shown in Fig. 11.For slabs on smooth HCU surface, the average slip at the ultimateload is around 0.1 mm and the slips increased at constant load tillfailure. The largest slip at failure is recorded between 0.41 mm and2.29 mm. For slabs on rough HCU surface, the slips at the ultimateload are mostly

Web crack

Web crack

Fig. 10. Web cracking (WC).

(a) Smooth surface

(b) Rough surface

0

50

100

150

200

250

300

0.5

Shea

r fo

rce

(kN

)

Interface slip (mm)

1SR-Dry 2SR-Pond

3SR-Wet 4SL-Dry

5SL-Pond 6SL-Wet

0

50

100

150

200

250

300

0.0 0.1 0.2 0.3 0.4

0.0 0.1 0.2 0.3 0.4 0.5

Shea

r fo

rce

(kN

)

Interface slip (mm)

1RL-Dry 2RR-Pond

3RR-Wet 4RR-Dry

5RR-Pond 6RR-Wet

Control

Control

Fig. 11. Shear force vs. interface slip relationship at the 665 mm wide unit up to0.5 mm.

(a) Smooth surface

(b) Rough surface

0

50

100

150

200

250

300

10

Shea

r fo

rce

(kN

)

Average deflection (mm)

1SR-Dry 2SR-Pond

3SL-Wet 4SL-Dry

5SL-Pond 6SL-Wet

0

50

100

150

200

250

300

0 2 4 6 8

0 2 4 6 8 10

Shea

r fo

rce

(kN

)

Average deflection (mm)

1RR-Dry 2RR-Pond

3RR-Wet 4RL-Dry

5RL-Pond 6RL-Wet

Predicted deflection

Predicted deflection

Control

Control

Fig. 12. Shear force vs. average deflection relationship up to 10 mm.

86 I.S. Ibrahim et al. / Engineering Structures 125 (2016) 8090

closely up to the cracking load, while those on smooth HCU surfacehave lower slope than the elastic prediction. This shows that thestiffness of the elastic section slab is higher for those built on roughHCU surface than those built on smooth HCU surface.

Beyond cracking the deflection increases and the slabs con-structed on rough HCU surface failed at larger deflection, i.e. moreductile, than the slabs constructed on smooth HCU surface. Thisbehaviour is true for HCU with all surface conditions.

9. Effect of the HCU surface roughness on the SLAB shearcapacity

The ultimate loads for rough HCU surface slabs are comparedwith the corresponding ultimate loads for smooth HCU surface asshown in Fig. 13. Comparison of the values indicates that, exceptfor the dry surface of left end data, the rough surface of the HCU

can increase the slab shear load capacity from 7% to 70%. It is alsofound that the rough HCU surface can reduce the interface slipbetween the concrete topping and the HCU. The interface slips atultimate loads are 0.1 mm for smooth HCU surface. The first crack occurred at lowerloads for slabs on smooth HCU surface.

10. Effect of moisture condition on the HCU surface

The ultimate loads for slabs where concrete topping cast on theHCU surface at different moisture condition are tabulated inTable 3. For the smooth HCU surface, the largest average ultimateload was recorded for the wet HCU surface condition, and loweston ponded HCU surface condition for left and right end tests. Thevalue differences with respect to the larger one are 5% and 15%.

0

50

100

150

200

250

300

350

0 50 100 150 200 250 300 350

Ulti

mat

e lo

ad fo

r ro

ugh

surf

ace

(kN

)

Ultimate load for smooth surface (kN)

Dry

Ponded

Wet

ControlDry condition from left end test

Fig. 13. Comparison of ultimate loads between rough and smooth surface for HCU.

Table 3Comparisons of ultimate loads of composite slabs on HCU surface with differentmoisture.

Ultimate load (kN)Smooth surface

Ultimate load (kN) Roughsurface

Wet Ponded Dry Wet Ponded Dry

Left end test 209.9 157.9 191.3 317.9 222.6 187.1219.5 205.5 186.3 274.1 268.9 203.5

Average 214.7 181.7 188.8 296.0 245.8 195.3Correction factor 1.00 0.85 0.88 1.00 0.83 0.66

Right end test 202.8 158.6 196.4 302.0 265.7 209.4205.6 196.6 193.1 260.1 284.3 232.8

Average 204.2 177.6 194.8 281.1 275.0 221.1Correction factor 1.00 0.87 0.95 1.00 0.98 0.79

I.S. Ibrahim et al. / Engineering Structures 125 (2016) 8090 87

On the contrary, for slabs built on rough HCU surface, the averageultimate loads are largest for the one cast on the wet HCU surfaceand lowest on the dry HCU surface for both left and right end tests.The value differences with respect to the larger one are 2% and 34%.From the findings, the study proposed correction factor given hereas cm when considering the surface moisture. The correction factoris determined by dividing the ultimate load of the dry and pondedwith the wet condition (by considering that the wet condition isthe perfect moisture condition on the HCU surface before placingthe in-situ concrete topping). The correction factor from the leftand right end tests are then averaged and proposed in this study.The correction factor is summarised as follows:

(a) Smooth surface: cm,dry = 0.91 and cm,ponded = 0.86.(b) Rough surface: cm,dry = 0.73 and cm,ponded = 0.90.

11. Analytical analysis on the shear capacity

The failure mechanism of the specimens (refer to Table 2, andFigs. 9 and 10) can be divided into three modes, that is (i) web-shear cracking failure in the HCU, Vwc (ii) shear failure at the inter-face between the HCU and concrete topping, Vi and (iii) shear-flexural failure of the composite slab, Vu.

The web-shear cracking strength can be predicted by elasticanalysis where the tension reinforcement ratio is neglected in thisprediction, thus;

Vwc vwcIcompbwScomp 3

where Icomp is the second moment area of the composite section,Scomp is the first moment of area above and about the centroidal axis

of the total composite section and bw is the effective web width. Theconcrete stress at web-shear cracking, vwc is given as [7];

vwc f ctffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1 f cc

f ct

s4

where fct is the splitting tensile strength of the HCU and fcc is thecompressive stress at level of highest shear stress from the effectiveprestressing force and external applied shear force.

The shear strength at the interface supposing that the web is ina cracked state can be determined from the following equation,where;

Vi v iIcompbiScomp 5

where bi is the interface width and vi is the shear stress at the inter-face given as [1]:

v i Vbtdcomp 6

where bt is the transverse interface width, dcomp is the effectivedepth to the tension reinforcement of the composite section andV is the applied shear force.

The ultimate shear capacity, Vu are compared with two equa-tions provided in Eurocode 2 [4] as discussed below and the Oka-muras equation, Eq. (2) as discussed in the introduction section.

Eurocode 2 gives an equation to estimate the shear resistancefor prestressed single span members without shear reinforcementof region cracked in bending as:

VRd;c 0:12k100q1f ck1=3 0:15rcpbwdcomp 7where fck is the concrete cylinder compressive strength, q1 Asbwd,k 1

ffiffiffiffiffiffi200d

qand rcp is the concrete compressive stress at the cen-

troidal axis due to axial loading and/or prestressing force.The expression above does not consider the hollow core and

therefore gives a constant value. It is applicable for beam withfixed cross section such as rectangular solid section. For sectionwith variable webs such as HCU, Eq. (7) highly underestimatesthe shear capacity. Therefore, a more suitable expression shouldbe used to take into account the hollow core. In region uncrackedin bending where the shear resistance should be limited by thetensile strength of the concrete, Eurocode 2 suggested the inclu-

sion of expression IcompbwScomp obtained using elementary beam theory

ignoring the normal stress gradient at the beam end, such that:

VRd;c IcompbwScompffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffif 2ct a rcp f ct

q8

where a is a factor taking into account of the fact that full prestress-ing is not yet developed at a distance x = d/2 from the inner edge ofthe support plate. The tensile strength of the HCU is evaluatedbased on Eurocode 2 [4] given as f ct 0:70f ctm, wheref ctm 2:12 ln 1 f cm10

h i> C60, fcm = fck + 8 (N/mm2), fck = 0.8fcu

and fcu = 76.8 N/mm2 which is the actual cube compressive strengthbased on an average of 3 tests. Hence, fct = 3.07 N/mm2.

Using the above values, the shear capacities are calculated usingEqs. (7) and (8). Comparison with Okamuras equation for ordinaryreinforced concrete slender beam without shear reinforcement, Eq.(2) is also done and the results are tabulated in Table 4, and plottedagainst the test results in Fig. 14.

Fig. 14 indicates that, most of the slabs capacities calculatedusing Eq. (7) is lower than the test value. Meanwhile, the calcu-lated shear capacities using Eq. (8) is higher than the test valueexcept the slabs on wet rough HCU surface where they are 28%lower than the test results. This shows that both Eqs. (7) and (8)

Table 4Values of VRd,c, Vu and test results.

Specimen Measured secondmoment of area,Icomp (107 mm4)

First moment areaabove and about thecentroidal axis,Scomp (105 mm3)

ScompIcomp

(mm1) Experimentalultimate shearcapacity, Vult (kN)

Calculated shearcapacity, Eq. (7),VRd,c (kN)

Calculated shearcapacity, Eq. (8),VRd,c (kN)

Okamuras shearcapacity, Eq. (2),Vu (kN) (See Note)

VRd,c from Eq. (7)/Vult VRd,c from Eq. (8)/Vult Vu/Vult

Smooth Dry 1SL 85.02 58.67 6.90 155.8 76.9 191.9 137.4 0.49 1.23 0.881SR 160.2 0.48 1.20 0.864SL 93.47 62.72 6.71 151.7 81.2 212.0 146.8 0.53 1.40 0.974SR 157.6 0.52 1.35 0.93

Ponded 2SL 92.56 59.67 6.45 127.3 82.3 224.3 139.9 0.65 1.76 1.102SR 127.9 0.64 1.75 1.095SL 93.94 60.30 6.42 168.2 79.7 216.1 140.9 0.47 1.28 0.845SR 160.5 0.50 1.35 0.88

Wet 3SL 83.68 60.44 7.22 172.1 84.5 207.3 143.8 0.49 1.20 0.843SR 166.0 0.51 1.25 0.876SL 93.73 60.64 6.47 179.9 85.8 235.9 142.5 0.48 1.31 0.796SR 168.0 0.51 1.40 0.85

Rough Dry 1RL 96.88 61.00 6.30 152.2 83.8 235.3 161.3 0.55 1.55 1.061RR 171.3 0.49 1.37 0.944RL 94.79 58.71 6.19 167.0 85.4 245.2 144.3 0.51 1.47 0.864RR 192.1 0.44 1.28 0.75

Ponded 2RL 100.20 61.54 6.14 183.2 83.5 240.0 162.3 0.46 1.31 0.892RR 223.0 0.37 1.08 0.735RL 95.43 58.51 6.13 219.9 84.5 244.2 143.5 0.38 1.11 0.655RR 235.9 0.36 1.04 0.61

Wet 3RL 99.89 62.00 6.21 264.9 84.6 241.5 163.9 0.32 0.91 0.623RR 251.3 0.34 0.96 0.656RL 96.35 59.54 6.18 227.1 85.6 246.4 146.7 0.38 1.08 0.656RR 215.1 0.40 1.15 0.68

Control Rough 93.33 52.17 5.59 273.8 82.5 259.6 137.9 0.30 0.95 0.50Control Smooth 95.35 52.83 5.54 258.7 82.3 251.6 139.9 0.32 0.97 0.54

S = Smooth and R = Rough.L = Left end section and R = Right end section.Note: Eq. (2) for a/d > 2.5 (562/182.35 = 3.08).

88I.S.Ibrahim

etal./Engineering

Structures125

(2016)80

90

0

50

100

150

200

250

300

0 50 100 150 200 250 300

Cal

cula

ted

ultim

ate

shea

r ca

paci

ty

(kN

)

Ultimate shear capacity from test (kN)

Okamura's Equation

EC2 Eq. (7) EC2 Eq. (8)

Fig. 14. Comparison of VRd,c and Vu with test results.

I.S. Ibrahim et al. / Engineering Structures 125 (2016) 8090 89

are non-conservative but nevertheless is able to predict variableshear capacities depending on the core opening size. On the otherhand, the Okumuras equation mostly predicts lower shear capac-ities than the test results in the range of 339% except for slabs onponded smooth HCU surface and the left end of slab on dry roughHCU surface where the test results are higher than the predictedvalues by 69%. The difference of values show that Okumurasequation is closer to the test results and is conservative comparedwith Eqs. (7) and (8) which mostly is either underestimates oroverestimates the shear capacities of the composite slabs. It shouldbe noted that Ueda and Stitmannaithum [7] had recommendedthat Okumuras equation should be limited to the condition wherethe core diameter is

90 I.S. Ibrahim et al. / Engineering Structures 125 (2016) 8090

[10] Adawi A, Youssef MA, Meshaly ME. Experimental investigation of thecomposite action between hollowcore slabs with machine-cast finish andconcrete topping. Eng Struct 2015;91:115.

[11] Baran E. Effects of cast-in-place concrete topping on flexural response ofprecast concrete hollow-core slabs. Eng Struct 2015;98:10917.

[12] ACI. Building code requirements for structural concrete. Detroit(MI): American Concrete Institute; 2005.

[13] JSCE. Standard specification for design and construction of concrete structures,Part 1. Tokyo: Japan Society of Civil Engineers (JSCE); 1986.

[14] Niwa J. Equation for shear strength of reinforced concrete deep beams basedon FEM analysis. Concr Libr Int 1984;4:28395.

[15] Okamura H et al. Reevaluation of the equation for shear strength of reinforcedconcrete beams without web reinforcement. Concr Libr Int 1987;9:6584.

[16] AASHTO. LFRD bridge design specifications. 5th ed. Washington(DC): American Association of State Highway and Transportation Officials;2010.

[17] Bensalem K. The structural integrity of precast concrete floor systems used ashorizontal diaphgrams. Nottingham: University of Nottingham; 2001.

[18] BSI. Concrete complementary British standard to BS EN 206-1 Part 2:specification for constituent materials and concrete: EN 206-1-2. London: British Standard Institute; 2006.

[19] CEN. Precast concrete products hollow core slabs forfloors. Brussels: European Committee for Standardization; 2002.

http://refhub.elsevier.com/S0141-0296(16)30270-X/h0050http://refhub.elsevier.com/S0141-0296(16)30270-X/h0050http://refhub.elsevier.com/S0141-0296(16)30270-X/h0050http://refhub.elsevier.com/S0141-0296(16)30270-X/h0055http://refhub.elsevier.com/S0141-0296(16)30270-X/h0055http://refhub.elsevier.com/S0141-0296(16)30270-X/h0060http://refhub.elsevier.com/S0141-0296(16)30270-X/h0060http://refhub.elsevier.com/S0141-0296(16)30270-X/h0065http://refhub.elsevier.com/S0141-0296(16)30270-X/h0065http://refhub.elsevier.com/S0141-0296(16)30270-X/h0070http://refhub.elsevier.com/S0141-0296(16)30270-X/h0070http://refhub.elsevier.com/S0141-0296(16)30270-X/h0075http://refhub.elsevier.com/S0141-0296(16)30270-X/h0075http://refhub.elsevier.com/S0141-0296(16)30270-X/h0080http://refhub.elsevier.com/S0141-0296(16)30270-X/h0080http://refhub.elsevier.com/S0141-0296(16)30270-X/h0080http://refhub.elsevier.com/S0141-0296(16)30270-X/h0085http://refhub.elsevier.com/S0141-0296(16)30270-X/h0085http://refhub.elsevier.com/S0141-0296(16)30270-X/h0090http://refhub.elsevier.com/S0141-0296(16)30270-X/h0090http://refhub.elsevier.com/S0141-0296(16)30270-X/h0090http://refhub.elsevier.com/S0141-0296(16)30270-X/h0095http://refhub.elsevier.com/S0141-0296(16)30270-X/h0095

Experimental study on the shear behaviour of precast concrete hollow core slabs with concrete topping1 Introduction2 Related works3 HCU specimen4 HCU surface roughness and moisture condition5 Concrete topping6 Testing procedure7 Test results8 Vertical shear force versus vertical deflection9 Effect of the HCU surface roughness on the SLAB shear capacity10 Effect of moisture condition on the HCU surface11 Analytical analysis on the shear capacity12 Summary and conclusionAcknowledgementReferences