Construction and Building

Construction and Building Materials 19 (2005) 595603

MATERIALS
www.elsevier.com/locate/conbuildmat

Repair and structural performance of initially cracked reinforced concrete slabs

Waleed A. Thanoon, M.S. Jaafar *, M. Razali A. Kadir, J. Noorzaei
Department of Civil Engineering, Universiti Putra Malaysia, UPM, Serdang, Selangor, 43400 Darul Ehsan, Malaysia Received 29 November 2002; received in revised form 10 September 2004; accepted 28 January 2005 Available online 21 March 2005

Abstract Crack is one of the most common defects observed in reinforced concrete slabs and beams. Major cracks in concrete structures may occur due to overloading, corrosion of reinforcement or dierential settlement of support. To restore the structural capacity of the distressed elements, retrotting and/or strengthening are needed. There are dierent techniques available for retrotting and strengthening of dierent reinforced concrete structural elements reported in the literature. This paper investigates the structural behaviour of cracked reinforced concrete one-way slab, which is repaired using dierent techniques. Five dierent techniques are used for the purpose of repair in the cracked concrete slab namely; cement grout, epoxy injection, ferrocement layer, carbon bre strip and section enlargement. The slabs were loaded to failure stage and the structural response of each slab specimens have been predicted in terms of deection, variation of strain in concrete and steel, collapse loads and the failure modes. The eciency of dierent repair and strengthening techniques and their eects on the structural behaviour of cracked one-way reinforced concrete slab had been analyzed. It was observed that the type of repair technique used will aect the load carrying capacity of the slab and will lead to a redistribution of the strains and hence stresses in both concrete and steel reinforcement. All repair techniques are found to be able to restore or enhance the structural capacity of cracked concrete slabs. 2005 Elsevier Ltd. All rights reserved.
Keywords: Repair and strengthening; Distressed concrete slab; Ferrocement; CFRP; Epoxy injection; Grouting; Section enlargement

1. Introduction In practice, situations arise where existing concrete structures or some of their components may, for a variety of reasons, be found to be inadequate and in need of repair and/or strengthening. The inadequacy may be due to mechanical damage, functional changes, overstress due to temperature changes, or corrosion of reinforcement. A common feature of a number of dierent causes of deterioration is that there is a reduction of the alkalinity of the concrete, which allows oxidation of the reinforcing steel to take place. This oxidation process leads to cracking of the concrete and possible spalling of the cover to the reinforcement.
*

Corresponding author. Tel.: +603 89466377; fax: +603 86567129. E-mail address: msj@eng.upm.edu.my (M.S. Jaafar).

Bridges are one of the concrete structures, which normally exhibit severe distress due to their exposure to harsh environment. Dierent repair techniques have been successfully developed to strengthen a given structure or part of it to restore its serviceability and strength. It is also prudent to consider durability aspect when repair or strengthening is carried out. With the advancement of new materials technology, which have superior mechanical properties and excellent resistance to electrochemical corrosion, many eective repairs and strengthening techniques have been developed. The nal selection of a suitable and most eective method generally depends on simplicity, speed of application, structural performance and total cost. Studies have shown that bre reinforced plates (FRP) increase the strength of exural members signicantly. Carbon bre reinforced polymer has a high strength to

0950-0618/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2005.01.011

596

W.A. Thanoon et al. / Construction and Building Materials 19 (2005) 595603

weight ratio, favourable fatigue behaviour and excellent resistance to electrochemical corrosion to make it practically suited for structural application [1]. A study conducted by Alfarabi et al. [2] showed that, although the FRP increase the failure loads, most of the beams strengthened by FRP started the failure at the curtailment zones of the plates. The epoxy used to laminate the plate at the sot of exural members only failed at loads much higher than the required level [3]. Similar study also found that the failure modes for repaired structures may change from ductile to brittle [4]. The probability of this change depends largely on the percentage of FRP being used, the location of FRP and the presence of shear reinforcement in the existing structures. Toong and Li [5] investigated the eect of using carbon bre reinforced polymers (CFRP) plates to strengthen one-way spanning slab to increase the exural capacity with particular emphasis on the cracking behaviour at working load. All the CFRP strengthened specimens exhibited large increase in load carrying capacity ranging from 60% to 140%. Ferrocement is a type of thin composite material made of cement mortar reinforced with uniformly distributed layers of continuous, relatively small diameter wire meshes. The use of ferrocement proper in repair was rst introduced by Romualdi [6] and Iron [7] in the early 1980s mainly as relining membranes for the repair of liquid retaining structures, such as pools, sewer lines, tunnels, etc. Investigation into the use of ferrocement as strengthening components for the repair and strengthening of reinforced concrete beams was reviewed by Paramasivam et al. [8]. In general, the damaged concrete and reinforcement (if also damaged) were removed and replaced with ferrocement, with or without any changes in overall dimensions of the beam. The beams were tested under static or cyclic loading [6] conditions. The strengthened beams were reported to exhibit improved cracking resistance, exural stiness and the ultimate loads compared to the original beams. These improvements; however, depend on the full composite action between the ferrocement layers. Al-Kubaisy and Zamin [9] presented the exural behaviour of reinforced concrete slabs strengthened with ferrocement tension zone cover. Twelve simply supported (500 mm2) reinforced concrete slabs were tested under exural load. The eect of the percentage of wire mesh reinforcement in the ferrocement layer, thickness of the ferrocement layer and the type of connection between the ferrocement layer and reinforced concrete slab on the ultimate exural load, rst crack load, crack width and spacing and loaddeection relationship were considered. Other technical methods used for repair of reinforced concrete structures are epoxy injection and cement grouting techniques. These techniques are widely used to

treat cracking problem in concrete. The procedure used is well established in the literature [1012]. This paper presents a study on the eects of dierent repair techniques on the structural response of one way reinforced concrete slab. The techniques include: (a) (b) (c) (d) (e) carbon bre reinforced polymers (CFRP) strip; cement grout, i.e., SikaGrout214; epoxy injection, i.e., Sikadur52; ferrocement cover and; section enlargement.

These techniques had been selected for their potential to either increase the structural capacity of members or to restore the original capacity of the sections. Furthermore, this study focuses on the serviceability, strength and ductility performance for each of the repair techniques to ascertain their potential application in cracked reinforced concrete slabs.

2. Experimental procedure and repair technique In order to investigate the eect of various repair techniques on the structural response of one way slab, a total of six full scale one-way reinforced concrete slabs having a dimension of 2.5 m long 1.0 m wide and 0.15 m thick are cast, cured and tested. The steel reinforcement consists of ve 10 mm diameter high-yield deformed bars with a characteristic strength of 460 N/ mm2. The 28-day cube compressive strength, fcu of the concrete used is 30 N/mm2. All specimens are tested under two-line load located in the middle third of the slab specimen as shown in Fig. 1. Initially, all slabs are loaded to 2/3 of their expected ultimate load capacity or after the development of cracks in the specimens (initial load ranges between 34 and 40 kN), except for the control slab, which is loaded until failure. Subsequently, the load has been released and the specimens

Fig. 1. Test set up.

W.A. Thanoon et al. / Construction and Building Materials 19 (2005) 595603

597

are removed from the testing frame so that repairs may be carried out. The resulting cracks in dierent concrete slab specimens were repaired using each of the techniques mentioned above. The specimens are then re-tested to failure after allowing a suitable curing period. The response of each specimen in terms of deection, stiness, cracking load, ultimate load, and failure pattern are analyzed. The repair techniques applied on cracked slabs are shown in Fig. 2.

2.1. Grout pouring technique using SikaGrout 214 (Slab S2) At the end of rst stage of loading (34 kN), two exural cracks of 0.88 mm width each were observed at the middle third of the slab. The resulted crack paths have been enlarged by 50 mm in width and 30 mm in depth to expose the main steel reinforcement. The exposed reinforcement and concrete surfaces were cleaned using steel brush, water jet and compressed air before grout is

Reinforced steel bar Fine crack

Mid-span of concrete slab

30 mm
Wiremesh

Ordinary reinforcement Skeletal steel

50mm

0.833 m

Before Repair Grout

Patch with cement mortar

(a) Grout pouring S2. (b) Epoxy injection S3.

Reinforced steel bar

0.15 m

Fine crack
1m 1.2 mm 50 mm

Epoxy (Sikadur30) Inlet Port Epoxy (Sikadur52)

(c) Ferrocement layer S4.
2m 2.5 m Carbon Fibre Reinforced Polymer (CFRP)

(d) CFRP strip S5.

Applied Load

0.15 m 0.05 m 0.15 m 2.5 m

5Y 10 4R 10 + 5 R 8

Epoxy

132 mm 50 mm 15 mm 1m

5 Y 10 4R 10 + 5 R 8

(e) Section enlargement S6. Fig. 2. Five dierent repair techniques applied on cracked reinforced slabs.

598

W.A. Thanoon et al. / Construction and Building Materials 19 (2005) 595603

Fig. 3. Cracks repaired using grout pouring technique.

ll the crack is Sikadur52, which is low viscosity epoxy, free owing and fast curing injection resin based on 2component solvent free epoxy resin. It has a density of 1.1 kg/L, tensile strength of 25 N/mm2 (7 days) and a compressive strength of 40 MPa at 20 C within 24 h. The viscosity of Sikadur52 is equal to 290 and 130 cps at a temperature of 20 and 30 C, respectively. The crack width limits for Sikadur52 is between 0.2 and 5 mm. Sikadur52 was injected from one end of the crack until the material exudes from the next nipple. The injection process was repeated until the whole crack is lled with the epoxy material. 2.3. Ferrocement layer (slab S4) The crack pattern observed at the end of initial load stage (38 kN) is shown in Fig. 5(a). The maximum crack width found at this load was equal to 0.75 mm. In this technique, a 30 mm depth concrete from the bottom of the slab was removed using a concrete chisel and hammer. This concrete layer has been removed only from the middle third portion of the slab (with dimensions of 850 mm 850 mm) as shown in Fig. 5(b). Two layers of 12.5 mm2 opening galvanized welded wire mesh of 1.25 mm diameter and a layer of skeletal steel (5R6) are xed with the original reinforcement of the slab after the concrete surface was roughened as shown in Fig. 5b. Cement mortar (cement to sand ratio is 1:2 with w/c ratio equal to 0.5) is applied and cured for 28 days. 2.4. Carbon bre reinforced polymers strip (slab S5) The crack pattern observed at the end of initial load stage (34 kN) is shown in Fig. 6(a). The maximum crack width found at this load was equal to 0.6 mm. A 50 mm wide and 2 m long CFRP strip having 1.2 mm thickness has been externally bonded to the tension face of the reinforced concrete slab using Sikadur30 epoxy adhesive (bonding agent). The carbon bre strip has been kept at

poured into the enlarged extents, see Fig. 3. The grout used is SikaGrout214 of a density 2.2 kg/L. It is a non-shrink premixed high strength cementitious grout. The mixing ratio used is 25 kg of grout to 4.4 kg of water. The specimen has been tested after allowing 7 days for curing. 2.2. Epoxy injection technique (slab S3) Similar to slab specimens S2, two exural cracks of 0.65 mm width each were observed at the middle third of the slab under a total load of 37.4 kN. In this repair method, injection nipples are installed along the crack path at 200 mm centers as shown in Fig. 4. Sikadur30 was used to x the injection nipples in position as well as to seal the surface of the cracks. The epoxy used to

Fig. 4. Cracks repaired using epoxy injection technique.

(a) Crack pattern before repair.

(b) Wire mesh used for repair.

Fig. 5. Cracks repaired using ferrocement cover.

W.A. Thanoon et al. / Construction and Building Materials 19 (2005) 595603

599

(a) Crack pattern before repair.

(b) CFRP laminate in position.

Fig. 6. Strengthening the slab with CFRP laminate.

the central part of the slab. The choice of CFRP was based on theoretical analysis using strain compatibility method and assuming both steel and CFRP yielded at the same time with the concrete compressive strain reaches 0.0035. The analytical calculation shows that the expected failure load of the strengthened slab with CFRP is equal to 106 kN (double the capacity of the slab). The concrete surface where the CFRP strip will be located was roughened and cleaned using compressed air and water jet. Sikadur30 epoxy adhesive was next applied on the roughened concrete surface (23 mm thick). The CFRP was then xed on the Sikadur30 adhesive layer as shown in Fig. 6(b). The tensile strength of the carbon bre strip is 2800 N/mm2, its modulus of elasticity is 165,000 N/mm2 and the density is 1.5 kg/L. The main characteristic of the Sikadur30 epoxy is presented in Table 1. 2.5. Section enlargement (slab S6) The crack pattern observed at the end of initial load stage (34 kN) is shown in Fig. 7(a). The maximum crack width found at this load was equal to 0.65 mm. In this repair technique, the bottom of the cracked slab is reinforced with additional 50 mm thick concrete layer rein-

Table 1 Characteristic of Sikadur30 epoxy Characteristics Sag ow Compressive strength Tensile strength Shear strength E-modulus (static) Shrinkage Glass transition point Guide values 35 mm at 35 C 75100 N/mm2 2030 N/mm2 1520 N/mm2 800016,000 N/mm2 0.040.08% 5070 C

forced with additional steel reinforcement. The strengthened slab was designed to fail at ultimate load of 104 kN. The bottom surface of the slab was roughened and a number of holes have been drilled to a depth equal to the eective depth of the slab. After cleaning the dust, 48 pieces of R10 steel bars (act as shear connectors) of 155 mm long were inserted in the holes and xed in position using Sikadur30 epoxy adhesive as shown in Fig. 7(b). Additional exural steel reinforcement 5R10 and 5R8 was next xed to the shear connectors. Similar concrete mix is used to cast the additional 50 mm concrete layer and kept for 28 days for curing before retesting the specimen.

(a) Crack pattern before repair.

(b) Roughened surface and steel provided.

Fig. 7. Strengthening by enlarging the slab section.

600

W.A. Thanoon et al. / Construction and Building Materials 19 (2005) 595603

160 140 120
(+ 1 29.22 % ) 13 0.93 kN

3. Results and discussions 3.1. Cracking and ultimate loads The initial cracking loads for dierent reinforced concrete slab specimens are shown in Fig. 8 along with the control slab (S1). All the repaired concrete slabs exhibit higher cracking load compared to the control slab, except S5. The repaired specimens using grout pouring, epoxy injection and section enlargement techniques show 35% increase in the cracking loads compared to the control slab. While the use of ferrocement layer increase the cracking load by 17.8%. The strengthening of the slab by using the CFRP at its sot, without repairing the initial cracks, improves the crack width only. In this specimen, new cracks are developed at slightly lower load compared to the original slab. Fig. 9 shows the ultimate failure loads for all the slab specimens. It could be observed that all the repair techniques used in this study are capable of restoring the ultimate capacity of the defected slab except specimen S3 where the cracks have been treated by epoxy injection. However, the reduction in strength is only 8.6% compared with the control slab. The ultimate capacity of slabs S5 and S6, which are repaired by CFRP and section enlargement, respectively, show 77.4% and 130% higher ultimate load capacities compared to the control slab. The increase in the ultimate strength for slab specimen S5 is in agreement with the result reported by Toong and Li [5] even though the ratio of CFRP strip area to the overall cross-section area used in this study is very small (0.04). Moreover, there is no increase in the ultimate capacity of the specimen repaired by ferrocement cover compared to the control slab which is matching well with the conclusion reported by Al-Kubaisy and Zamin [9]. Both crack and ultimate loads for all slabs also indicate that the repaired structures had a high degree of integrity. The main concern that engineers normally have are related to the ability of the repair material to integrate and act compositely with the parent materials
60 50

(+ 77.39 % ) 101 .4 1 kN

Load (kN)

100 80 60 40 20 0 S1 S2 S3 S4 S5 S6
(+ 17 .2 3 % ) 66.96 k N (- 8.61 % ) 52 .2 k N

C o n tro l S lab 5 7.12 kN

(0 % ) 57.12 k N

Slab
Fig. 9. Ultimate loads for slabs 16.

did not appear to be a problem for any of the repair techniques being investigated in this study. 3.2. Slab deections Fig. 10 shows loaddeection for each of the slabs. These deections are recorded at the mid-span. The slab specimens show almost similar stiness except specimen S6, where the deection has decreased due to the stiness of the extra concrete layer of concrete that has been added. The stiness has increased more signicantly compared to other specimens. Deection patterns for the control slab S1 and specimens S2, S3 and S4 showed that all of them had similar initial stiness. However, after two-third of the ultimate load, these specimens exhibit dierent level of ductility pattern. The maximum deections observed in S3 (epoxy injection) and S4 (ferrocement layer) are 15% lower compared to the control slab, while pouring the crack with grout show 20% increase in maximum deection compared to the original slab.
140 18.83 mm 120 100 38.65 mm

Load (kN)

40 30 20 10 0 S1
Control Slab 27.60 kN

(+ 17.83 %) 32.52 kN (- 9.42 %) 25.00 kN

(+ 34.06 %) 37.00 kN

Load (kN)

(+ 35.65 %) 37.44 kN

(+ 35.65 %) 37.44 kN

80 60 40 20 0 15.83 mm 22.90 mm 19.00 mm 16.31 mm

Control Slab - S1 Grout Pouring - S2 Epoxy Injection - S3 Ferrocement - S4 CFRP Strip - S5 Enlargement - S6

s2

s3

s4

s5

s6

10

20

30

40

50

Slab
Fig. 8. Cracking loads for slabs 16.

Deflection (mm)
Fig. 10. Loaddeection curves for all slabs.

W.A. Thanoon et al. / Construction and Building Materials 19 (2005) 595603

601

On the other hand, specimen S5 (CFRP) and specimen S6 (Section enlargement), show dierent variation in loaddeection curves compared to all other specimens. The specimen repaired by section enlargement exhibits higher stiness and the loaddeection curve shows much stier behaviour. The maximum deection observed is almost the same as observed in the control slab but occurred at more than twice the ultimate load. Moreover, non-ductile variation in the loaddeection curve could be observed in this specimen, which change the ductile behaviour observed in the control slab. The specimen reinforced with the carbon bre strip shows no change in the initial stiness compared to the control slab. With the increase of loading, the stiness decreases at a higher rate compared to specimen S1, S2, S3 and S4. Hence, the CFRP strip has a signicant eect on the stiness in the advance stage of loading. 3.3. Strain distribution Fig. 11 shows the variation of the concrete compressive strain at the mid-span at a distance of 25 mm below the top bre of the reinforced concrete slab specimens versus the applied load. This location however, is very close to the N.A of the slab section and hence the values of the strains are very small. It was recorded during the test that the N.A. is shifting up as the applied load increases and its depth during dierent loading stages changed from 40 to 30 mm (approximately) in specimens S1, S2, S3 and S4. For specimens strengthened with CFRP and reinforced concrete layer (S5 and S6), the depth of N.A changed approximately from 70 to 45 mm near failure load. Fig. 12 represents the concrete strain in dierent slab specimens when the applied load is equal to 50 kN (near by failure load of the control slab). All the slab specimens exhibit lower strain values compared to the control slab. The decrease in the concrete strain in specimens S2, S3 and S4 has been found ranging between 30% and

450

400

360
350

Concrete Strain (x 10 )

-6

300

250

230 185 195

200

150

115
100

60
50

0 S1 S2 S3 S4 S5 S6

Slabs
Fig. 12. Concrete compressive strain at 50 kN load level.

50%, while in S5 and S6 specimens, the reduction in compressive strain of concrete has been recorded at 65% and 85%, respectively. For the specimen reinforced with additional reinforced concrete layer, it was observed that the strain in the additional reinforced concrete layer is not compatible with strain in the original slab although initially both layers act as a composite section. This is due to providing insucient number of shear connectors which leads to the occurrence of horizontal longitudinal cracks between the two layers. 3.4. Failure modes and mechanism Fig. 13 shows the crack pattern of all the strengthened slab specimens at failure. The tests were stopped when excessive deection and/or excessive wide cracks were observed although the specimens did not completely collapse. It is clear from the crack patterns in the slabs S1, S2, S3 and S4 almost similar modes of failure have been observed. Cracks started at the tension sides and increased in width and length with the applied loads. In the control slab the neutral axis location is shifted upwards until the concrete strain reaches its ultimate value. At this stage, the steel reinforcement is yielded which quickly led to compressive crushing of concrete. This failure mechanism is a typical ductile failure observed in under-reinforced concrete sections. However, in the repaired specimens, the ductility is not clearly observed as in the control slab. The failure mechanism in Slab S5, which has been strengthened by CFRP strip, is dierent from other slab specimens since CFRP laminates is additional reinforcement. The failure is characterized by shearing of the concrete interface with the CFRP strip (relative

140 120 100 177

314
Control Slab - S1 Grout Pouring - S2 Epoxy Injection - S3 Ferrocement - S4 CFRP Strip - S5 Enlargement - S6

Load (kN)

80 60 40 20 0 0 100 200 226 257 206

539

Strain (x 10-6)

300

400

500

600

Fig. 11. Variation of concrete compressive strain for all slabs.

602

W.A. Thanoon et al. / Construction and Building Materials 19 (2005) 595603

Fig. 13. Crack patterns and modes of failures for strengthened slabs.

slippage) associated with less warning compared to other slab specimens. The failure was sudden and occurred immediately after the peeling of the CFRP strips. This is due to insucient anchorage length of the CFRP strip. The strain measured before failure in the CFRP laminate is 60% of its yielding strain which comply with peeling failure (and not rupture failure) observed in this specimen. The number of cracks observed at failure is more but the cracks width are smaller compared to other specimens. For S6 slab specimens, in addition to the transverse cracks, horizontal crack had been formed at the boundaries between the original slab and the additional concrete layer due to shearing.

(ii) The repairs using grout, epoxy injection and ferrocement layers showed behaviours similar to that of the control slab in terms of strength and ductility performance. In other words, these repair techniques can safely adopt the normal reinforced concrete design for concrete slabs. (iii) CFRP and section enlargement repair techniques for the cracked slabs showed superior structural performance in terms of strength. The ductility performance for these slabs, however, is less than that of the control slabs. (iv) It could also be concluded that all repairs techniques used are eective to at least restore the structural performance of cracked reinforced concrete slabs.

4. Conclusion Acknowledgements Based on this study, the following conclusions could be drawn: (i) The repaired structures had similar or higher cracking and ultimate loads compared to the control slab. The authors thank Sika Kimia Sdn. Bhd., Malaysia for providing the CFRP strip and other Sika products. Moreover, the authors acknowledge Mr. Wong Chee Wai and Mr. Tan Khong Yee for their contribution in the experimental work.

W.A. Thanoon et al. / Construction and Building Materials 19 (2005) 595603

603

References
[1] Clarke JK, Waldron P. The reinforcement of concrete structures with advanced composites. Struct Eng 1996;74(17/3):1996. [2] Alfarabi S, Alsulaimani GJ, Basunbul IA, Bamch MH, Ghaeb BN. Strengthening of initially loaded reinforced concrete beams with using FRP plates. ACI Struct J 1994 (March):1608. [3] El-Mihilmy MT, Tedesco JW. Analysis of reinforced beams strengthened with FRP laminates. J Struct Eng 2000:68491. [4] Arduni M, Tommaso AD, Nanni A. Brittle failure in FRP plate and sheet bonded beams. ACI Struct J 1997;94:36370. [5] Toong K Chan, Li Xiaoan. Improving crack behaviour of oneway slabs with carbon bre plates. In: Proceedings of the 4th Asia-Pacic structural engineering and construction conference, September 1315, 2000, Kuala Lumpur, vol. 2; 2000. p. 35158.

[6] Romualdi JP, Lim CTE, Ong KCG. Strengthening of RC beams with ferrocement laminates. Cement Concrete Compos 1998;20:5365. [7] Irons M. Ferrocement for infrastructure rehabilitation. Concrete Int Design Constr 1987;9:248. [8] Paramasivam P. Laminated ferrocement for better repair. Concrete Int: Design Constr 1987;9:348. [9] Al-Kubaisy MA, Jumaat Mohd Zamin. Flexural behaviour of reinforced concrete slabs with ferrocement tension zone cover. Constr Build Mater 2000;14:24552. [10] Emmons PH. Concrete repair and maintenance. R.S. Means INC; 1994. [11] Allen RTL, Edwards SC. The repair of concrete structures. Blackie; 1987. [12] Raina VK. Concrete bridge practice-construction, maintenance and rehabilitation. 2nd ed. New Delhi: Tata McGraw Hill; 1993.