Analysis of reinforced concrete beams strengthened in flexure with composite laminates: Discussion

Discussions 835 The authors are to be commended for attempting to pro- vide a refined, yet simple, flexural analysis tool for rein - forced concrete beams strengthened with composite laminates. Although there is much merit in the paper, there are also some aspects of the work that need careful examina- tion before anyone begins to apply it to any real structure. The purpose of this discussion is to point out the danger that is inherent in analyses that (i) assume perfect bond between the FRP laminate and the concrete and (ii) assume the shear and normal stresses at the interface of the laminate and the substrate to be governed by classical beam theory. The writer is confident that the authors know about these dan- gers, but have forgotten to mention them in the paper. Both experiments (Tumialan et al. 1999) and theory (e.g., Rabinovitch and Frostig 1999) have shown that the shear and normal stresses at the interface of the laminate-epoxy and epoxy-concrete have a complex distribution which can- not be found by using the simple beam theory. In fact, beam theory erroneously gives the normal stresses perpendicular to the interface to be zero, but more refined analyses based on theory of elasticity indicate significant tensile stresses de - veloping at the interface. The beam theory also leads one to believe that the interfacial longitudinal shear stresses can be calculated using the elementary shear stress equations given in strength of material books. In reality these stresses have a much more complex distribution. Furthermore, when a con- crete beam cracks, the stresses at the interface in the vicinity of the crack greatly deviate from the distribution predicted by simple beam theory or by any other theory that ignores the presence of cracks. The experimentally observed delamination of either the substrate concrete or the FRP laminate is caused by the combined effect of these high in- terfacial stresses. Given that premature delamination will cause a strength- ened reinforced concrete beam to fail at a load lower than predicted by the program described in the paper, the results of the program could lead to unsafe design. This is the dan- ger that users of the program must be aware of. Figure 6 of the paper illustrates this point where the predicted and the actual load differ significantly. Note that the measures taken to prevent delamination, namely, the use of the U-shaped composite anchors, cannot be modelled by the program as described. The program is unable to quantify the effect of the end anchors and thus unless it can be shown by some other method that delamination cannot occur, it may overes- timate the actual capacity of a member with or without an- chors.