Abstract: The current research addresses Engineered Cementitious Composites (ECC) as a new alternative for retrofitting damaged concrete beams. Twenty-one plain concrete beams with pre-defined artificial cracks were prepared and repaired using different combinations of ECC alone or together with Carbon Fiber Reinforced Polymers (CFRP). The study showed that replacement of the inferior layer from the bottom of the deteriorated beams with a thin layer of ECC could be able to restore the beam to a condition better than its original state. Moreover, the repair with ECC was found effective in enhancing the member ductility as well. It was also shown that pasting CFRP directly over ECC substrate resulted in shear failure rather than the undesirable interfacial debonding mode of failure that typically occurs in case of concrete substrates.
INTRODUCTION
Deterioration of concrete structures can appear in the form of extra deformations and cracking of the structure elements at the most stressed portions. Repair of concrete structures requires sufficient experience to the various ranges of the repairing materials with their different physical and chemical properties as well as the optimum method for their application. Compatibility with the original construction material, the ease of use in a wide variety of situations, simplicity, speed of application and the total cost are all crucial in selecting the appropriate repairing material. Many researches have studied the repair and retrofitting of concrete members containing cracks. Waleed et al. (2005) used cement grout, epoxy injection, ferrocement layer, carbon fibre strip and section enlargement to repair cracked concrete slabs.
The introduction of a high performance fiber reinforced cementitious composites called Engineered Cementitious Composites (ECC) by Li (2002), nearly two decades ago, attracted many researchers to utilize this material for repairing and strengthening purposes. ECC has strain hardening behavior when exposing to tensile forces. It also exhibits a strain level two orders of magnitude higher than that of concrete (Maalej and Leong, 2005a, b). The many cracks with very small crack width are considered one of its superior characteristics. The fibers inserted in ECC help in bridging the applied forces through the generated cracks. Cost wise, ECC is very expensive to be used on a big scale and should only be applied at particular places of the damaged structures. Li et al. (1995) and Lim and Li (1997) overlaid the concrete substrate by layers of ECC, It was shown that the delamination and spalling modes did not occur; micro size cracks were found instead. Kamada and Li (2000) found that smooth surface preparation of the substrate concrete achieved good structure performance than the corresponding cases with rough surface. Zhang et al. (2006) conducted theoretical and experimental studies to check the suitability of using ECC as a strengthening material for beams, the results showed enhancement in the flexure behavior of the beams with increasing ECC thickness. On the other hand, Maalej and Leong (2005a) conducted a primary study to strengthen reinforced concrete beams, containing ECC layer located at its base, with CFRP. However, the study was preliminary and required further investigations.
In the current research, plain concrete beams were prepared and tested destructively. Each specimen contained artificial crack with a depth equal to 50% of the beam thickness. ECC with different thicknesses was used as a repair material alone or simultaneously with Carbon Fiber Reinforced Polymers (CFRP). Different curing ages were deliberately selected for both the substrate concrete and the repairing material, ECC. It was thought that this combination might be close to reality. Moreover, the effect of ECC stiffness to the original beam stiffness was also tested. The results showed enhancement in the section capacity as well as the ductility of the repaired beams.
MATERIALS AND METHODS
Experiment Procedure
Twenty-one scaled down plain concrete beams with dimensions (depthxwidthxlength)
equal to (100x100x400 mm) were prepared. Small size beams were selected
to minimize the cost of the used ECC. The specimens contained artificial
cracks, 50 mm depth and 2 mm width, located at their mid-spans. The cracked
beams were designed so that their load carrying capacity should not exceed
20% of the sound specimens. The above condition was confirmed by conducting
many side experimental tests. The cracks were conducted by placing separators
in the moulds before casting the concrete. The combined concrete/ECC beams
were prepared through two stages; first, the concrete was cast leaving
a free space in the moulds as shown in Fig. 1, second,
after 28 days of water curing to the concrete, ECC with different thicknesses
was added to fill the remaining free spaces in the moulds. ECC was placed
locally around the crack, or with thicknesses of 10, 30 and 50 mm from
beam bottom. For more details about the notation of specimens as well
as the number of replications for each case, refer to Table
1. After casting ECC, the final repaired specimens were again kept
for curing in water for another 28 days.
The specimens were mainly divided into seven groups. Group 1A, represented the original state of the beams before deterioration and was kept as a target condition to all the deteriorated specimens.
| Fig. 1: | Specimens before repair |
| Table 1: | Description of the examined specimens |
| *A: Denotes to specimens without CFRP, **B: Denotes to specimens with CFRP pasted at the beam soffit | |
| Table 2: | Mix proportions of concrete |
| Table 3: | Mix Proportions of ECC per unit volume |
| *Diluted with 25 times of water mass | |
Group 2A, represented the deteriorated severe cracked case with no repair. The current research aimed to retrofit the beams from condition similar to group 2A to reach its original state (i.e., group 1A). The other five groups represented beams with different repair combinations.
To distinguish the groups, either a letter A or B was added to the group number. The letter A means that the conducted repair method was by using ECC only. The letter B denotes groups where CFRP sheets were applied simultaneously with ECC.
The specimens were subjected to four points loading bending test with a bending span of 300 mm. Load deflection curves were measured at the mid spans of beams. In the current study, the repair length of ECC was 280 mm, i.e., 20 mm lesser than the bending span. The difference in age of both concrete and ECC might also coincide with real conditions. It was thought that these configurations could help in studying the spalling of ECC from concrete substrate, if any and could end-up with a better understanding of the flexure performance of the repaired beams.
Materials
The mix proportion for both the concrete and the ECC are shown in Table
2 and 3, respectively. The concrete mix was designed as
per the American Concrete Institute, ACI 211.1-32. The used ECC was in the form
of ready mixed powder. ECC incorporated high modulus polyethylene fibers. The
fiber length is 12.7 mm and diameter 38 micron, tensile strength of 1690 MPa
and an elastic modulus of 40,600 MPa. An expansive agent and an alcohol-type
shrinkage reducing agent were added. A bio-saccharide type viscous agent was
also applied to provide compatibility between fluidity and fiber dispersibility.
In addition, a polycarboxylic-acid-based superplasticizer was simultaneously
used. The ECC mixture was brought from Kajima Technical Research Institute.
More details about the used ECC were reported by Kanda et
al. (2006).
| Table 4: | Physical properties of CFRP sheets* |
| *As supplied by manufacturer | |
| Table 5: | Physical properties of the bonding epoxy* |
| *As supplied by manufacturer | |
The measured compressive and flexural strengths of concrete after 28 days were 35 and 4 MPa, respectively. The corresponding values for ECC were 27 and 9 MPa, respectively. At the test day, a little deviation was noted on the concrete and the new values at 56 days were 36 and 4.5 MPa, respectively.
In case of applying CFRP, the required surfaces were polished and smoothened; primer coat followed by a two compound epoxy risen were applied. The detailed application procedure for the unidirectional CFRP was noted by Anwar et al. (2007). CFRP was applied at the beam soffit leaving a 5 mm spacing between the cut-off and the support from both sides. Table 4 and 5 show the physical properties of the used CFRP and epoxy, respectively.
RESULTS AND DISCUSSION
Beam Strength
The load-mid-span deflection curves for the conventionally used methods
are shown in Fig. 2. The current repairing methods are shown
in Fig. 3 and 4. Noticed enhancement in
the section capacity was observed. Specimen repaired only with traditional epoxy
injection (C100-cr-epoxy) showed a 130% enhancement in the section capacity
over the cracked specimen (C100-cr). It, however, failed to reach the section
capacity of the beam original state (C100). It was also noted that the traditionally
used repair technique (C100-cr/CFRP) exhibited a section capacity 133% higher
than the target beam (C100). In the current research, the repair with ECC alone
was found more effective. Both the loading capacity and the ductility of the
repaired beams were enhanced. For (C100-El), ECC was only applied locally, around
100 mm at the beam base and extended to the full crack depth. The enhancement
in the section capacity was 220% more than the damaged state (C100-cr). Although
this combination was still not able to reach the initial original state of the
beam; it was found better than the traditional case (C100-cr/epoxy). The low
loading capacity might be attributed to the relatively small contact area between
ECC and the base concrete at a zone of extremely high normal stresses. The application
of ECC layers, with different thicknesses with a longer distance among the beam
soffit, improved the flexure performance a lot. It was observed that ECC acted
as a bridge to convey the internal forces between both sides of the crack. Retrofitting
of the damaged beams with ECC showed that the original integrity was not only
restored but enhanced as well. The section capacity for (C90E10), (C70E30) and
(C50E50) was found 12, 42 and 75%, respectively, more than the target specimen
(C100) and 505, 640 and 790% higher than the damaged specimen (C100-cr). Generally,
the increase in ECC thickness led to the increase in the section load carrying
capacity.
| Fig. 2: | Load-deflection curves for the traditional repairing methods |
| Fig. 3: | Load-deflection curves for repair using ECC Only |
| Fig. 4: | Load-deflections curves for repair using combinations between ECC and CFRP |
| Table 6: | Flexure test results for all the examined specimens |
The same combinations were again repeated after applying CFRP sheets at the beam soffit. The results showed that the mitigation in the section capacity for (C100El/CFRP), (C90E10/CFRP), (C70E30/CFRP) and (C50E50/CFRP) over the target beam (C100) were 127, 153, 115 and 119%, respectively. The previous values did not show substantial variation and a little bit lesser than that of (C100-cr/CFRP) except for the case where ECC was of minimum thickness. The application of CFRP led to an increase of the absorbed energy by the beam until failure. This energy was in the form of high loading capacity accomplished with small deflection. That was the reason for having small variation in the load carrying capacity for all the specimens incorporated CFRP. Table 6 shows the flexure test results for all the examined specimens. Moreover, it was noted that 30-50% reduction in the effective shear resisting cross-section for both (C70E30/CFRP) and (C50E50/CFRP), respectively, was behind the observed pure shear failure and the reduction in the total load carrying capacity.
For instance, it can be suggested that the use of ECC with CFRP did not influence the section loading capacity compared with cases with no ECC. ECC was found significant to act as a transient material between concrete of low strain capacity and CFRP of relatively high stain capacity. ECC was also able to change the mode of failure of the beams.
Modes of Failure
In general, the mode of failure changes by changing the repair method. Figure
5 shows the modes of failure for the examined specimens. It was observed
that both the control (C100-cr) and target (C100) specimens exhibited pure flexure
mode of failure. For the commonly used repair method, (C100-cr-epoxy), flexure
failure had also occurred; the crack started exactly from the same place of
the pre-existing fault after peeling of the epoxy from one of the crack sides.
For (C100-El), flexure crack with spalling of the local ECC was noticed. This
revealed that the contact between the concrete layer and the ECC was not capable
of withstanding the induced tensile stresses. For (C90E10) and (C70E30), flexure
mode of failure started exactly at the crack position and continued through
the whole ECC thickness. It was noted that the contact length was sufficient
and spalling of the ECC layer was not observed. The effect of the repairing
material stiffness was significantly noticed when ECC thickness reached 50 mm.
For (C50E50), the failure of one of the specimens was observed early after debonding
and spalling of the ECC layer, nearly at a position just under the applied load
point. The relatively high rigidity of the ECC layer prevented the ECC from
following the same deformation of the concrete beam, thus spalling occurred.
The second specimen failed almost by the same mode but the separation was only
limited to one side surface of the ECC layer, the crack started at one side
of ECC and followed vertically causing separation of ECC from concrete; the
crack was then preceded directly to the support.
| Fig. 5: | Typical modes of failure for the examined specimens |
It was shown that in the repairing process, the thickness along with the length of the applied ECC should be precisely selected and more investigations are still required.
After applying CFRP sheets at the beam soffit, the mode of failure was changed. It was however noted by many researchers that applying CFRP sheets directly to concrete surface could end up with undesirable interfacial debonding mode of failure. The concentration of the interfacial shear stresses near the vicinity of the crack was the main cause of the delamination of the CFRP sheet (Maalej and Leong, 2005b). For (C100-cr/CFRP), flexural crack induced interfacial debonding mode of failure was observed, the debonding location started just underneath the load position, this fact was also noted by Aprile et al. (2001). For (C100-El/CFRP), flexure shear crack followed by interfacial debonding of CFRP was observed.
In contrast with concrete, for all the other specimens where CFRP was applied over the ECC layer with different thicknesses, the common interfacial debonding of CFRP was changed to shear failure. Flexure shear crack was observed in case of (C90E10/CFRP) while (C70E30/CFRP) and (C50E50/CFRP) underwent a pure shear mode of failure and the crack was found extending from the support directly to the nearest loading point. The shear crack split the concrete near the support leaving a small portion attached to the ECC layer. This emphasizes the great influence of ECC on delaying the interfacial debonding failure or even avoiding its occurrence. It was thought that if the beams were well confined in shear, the previous trends might have little changes. Moreover, it might be expected that the interfacial debonding would have happened but after achieving higher loading value. The reasons for the strong bonding between ECC and CFRP might be attributed to the ability of ECC to develop very high strain comparing to concrete; along with the small size cracks that were created in case of ECC. The previous reasons might lead to minimize the induced shear stress in the adhesive layer and decelerated the debonding failure. Moreover, the presence of large amount of fibers on the prepared surface and just underneath this surface created also strong bonding between CFRP and the ECC substrate; it effectively helped in resisting both the induced shear and normal stresses generated along the intermediate adhesive.
Finally, it can be concluded that the dissipation of the consumed energy during loading for system repaired with CFRP might follow either, (i) developing of a localized crack at the, most stressed, bottom surface followed by the debonding failure mode when pasting the CFRP directly over concrete surface or (ii) developing of another crack failure mode - here it was shear crack - rather that the interfacial debonding failure mode when pasting CFRP directly on ECC.
Deformability
Ductility is the ability of the repaired beams to carry deformation
under peak load without full collapse. Figure 3 shows
also the enhancement in the member ductility with the increase of ECC
thickness. The maximum deflection was also noted and shown in Table
6. It was clear that the undergone deflection was directly proportional
to the ECC thickness. The application of CFRP showed only minor changes
in the maximum deflection achieved by the specimens. The decrease in the
maximum deflection of (C70E30/CFRP) and (C50E50/CFRP) was due to the occurrence
of shear failure that prevented the beams from reaching their ultimate
deflection limits. It was however suggested to carry out similar combinations
again after confining the high shear stress zones so that the comparison
would be more realistic.
CONCLUSION
In the current research, twenty-one plain concrete specimens were prepared to examine the effectiveness of using ECC as a repair material for cracked concrete beams. Different combinations were done with and without CFRP and the following were concluded:
ECC had the ability to bridge the internal forces between the crack sides due to its high ductility and perfect contact with the substrate concrete without the use of any adhesives.
Local placing of ECC at the crack position had no great significance in restoring the integrity of the damaged structures.
As the thickness of ECC increased, both the section capacity and the ductility of the damaged beam were enhanced.
The repair incorporating CFRP sheets showed almost similar flexure strength achievements. However, when ECC was used as a substrate to the CFRP, the mode of failure was changed from the undesirable interfacial debonding to shear failure.
The minimum the ECC thickness followed by CFRP application was found optimum in enhancing both section capacity and ductility with low probability of interfacial debonding.
The change in the curing ages between the ECC and the concrete beam did not show any weakness in their contact.