Discussion of "An experimental investigation of steel-free deck slabs"

We are pleased to note that the authors have described the ultimate load tests on the seven full-scale models of steelfree deck slabs meticulously and have also clearly presented all the relevant results. However, several of their speculative conclusions are not supported by their test data. As noted in the following discussion, our experimental research has confirmed that a combination of steel straps for confinement and a mesh of glass fibre reinforced polymer (GFRP) bars of minimal cross-sectional area for crack control leads to the most economical and durable bridge deck. With the help of two doctorate and one masters student, we have conducted further research in the cracking and fatigue aspects of steel-free deck slabs. Some of the findings of this research have already been published (Mufti et al. 2001; Banthia et al. 2002; Mufti et al. 2002; Limaye et al. 2002; Memon and Mufti 2002), but most were published after the authors had submitted their paper to the Canadian Journal of Civil Engineering on 3 April 2002. It is not surprising that the authors were not aware of this later research, in the light of which in particular we would like to note the following. Banthia et al. (2002) have reported on the development of a prestressed concrete strap, which is partially embedded in the steel-free deck slab. The premise behind this development, confirmed by the experimental study, is that concrete provides substantial restraint to the deck slab until it cracks. Transverse prestressing of the deck slab, whether internal or external, raises the threshold at which the concrete cracks. Transverse prestressing can help to increase the failure load only when it is high enough to keep the concrete from cracking at higher loads. The authors’ conclusion with regard to the effect of limited transverse prestress on the failure load of the deck slab is not surprising. The authors have noted in two places that a longitudinal crack in a steel-free deck slab is “dangerous and marks the end of the serviceability limit state because it breaks the continuity of the slab.” (page 835). On page 839, they contend “that this crack would always tend to continue propagating towards the edges and reach the upper surface of the slab, causing full-depth cracking that breaks the continuity of the slab.” In the following text, we would like to dispel the authors’ fears that longitudinal cracks are dangerous and render the slab unserviceable. We are aware that the longitudinal cracks at the bottom of steel-free deck slabs eventually develop into full-depth cracks. During static and single pulsating loads, the compressive stress regime around the load keeps the bottom surface cracks from reaching the top (e.g., see Mufti et al. 2001). The first part of the work of the Ph.D. student,V. Limaye, was to determine the number of passes of a given wheel load that would cause the bottom longitudinal crack to reach the top surface. It was found that a steel-free deck slab, having nearly the same dimensions as the seven slabs described by the authors, developed a bottom surface longitudinal crack after only 10 cycles of a pulsating load peaking at 391 kN. Thereafter, the pulsating load was applied alternately at two locations. After 1700 passes of 391 kN load, the crack extended to the top surface of the slab (Limaye et al. 2002). In one of the panels of the model slab, a 75 mm high insert, placed midway between the girders, broke the transverse continuity of the slab in its bottom portion. A fulldepth crack directly above the insert, appearing because of a fatigue test on an adjacent panel, offered an opportunity to allay the concerns about the full-depth crack breaking the continuity of the top surface of the slab. It is noted that besides the authors, many other engineers have expressed similar concerns to us. As shown in Fig. 1, a pulsating load, peaking at 391 kN, was placed on one side of the crack. Even after 1700 cycles of this very high load, no discontinuity was observed on the top surface of the slab. As shown by Mufti et al. (2002), 6115 passes of the 391 kN wheel induce the same fatigue damage in a deck slab as the damage by the maximum number of wheels, being 372 million ranging from 10 to 156 kN, during the 75 year life of the bridge. It is noted that the test had to be stopped because of the demands on the use of lab-