Coupled Experimental and Computational Investigation of the Interplay between Discrete and Continuous Reinforcement in Ultrahigh Performance Concrete Beams. I: Experimental Testing

Ultrahigh performance concrete (UHPC) is a special class of cementitious composites with superior strength and durability characteristics. Typically, UHPC contains very fine aggregate (size less than 1 mm) and short fiber reinforcement (less than 15 mm long). This sets it apart from the family of regular fiber-reinforced concrete (FRC) that has aggregate and fiber of similar lengths since the aggregate is typically coarse (10-20 mm) and that makes fiber effects on heterogeneity and crack bridging in UHPC more dominant compared to FRC. Additionally, since fiber content is relatively high (around 2% by volume), it is expected that a significant interplay exists between fiber content and traditional continuous reinforcement (rebars). Such interplay could affect the modes of failure, load-carrying capacity, and ductility of reinforced-UHPC (R-UHPC) members. Given such differences, prediction of failure mechanisms for R-UHPC members via extrapolation of design guides available for normal concrete (NC) and FRC could be debatable. Therefore, this two-part study discusses a coupled experimental and computational investigation analyzing the effect of fiber-rebar interplay on the modes of failure, strength, and ductility of R-UHPC prismatic members. The experiments included an initial campaign of testing 12 R-UHPC beams with different fiber contents and reinforcement ratios. Beams were tested with an eccentric three-point bending setup along with material level companion tests on specimens cast from same batches. Interestingly, all beams with fiber failed in bending without noticeable shear failures. The data from this initial campaign were used to calibrate and validate a computational modeling framework. This validated model was further used to predict the needed dimensions to achieve shear failure. Given these dimensions, a second experimental campaign was performed, and all tested beams failed in shear as predicted. These experimental programs are reported in the present Part I. The computational counterpart is described in the companion paper as Part II.