Blast retrofit of shear-deficient high-strength concrete beams with ultra-high performance concrete

Abstract

This research has investigated the ability of ultra-high performance concrete (UHPC) to improve the blast performance of high-strength concrete (HSC) beams. As part of the study 4 shear-deficient HSC beams built without stirrups and with dimensions of 125 mm × 250 mm × 2440 mm were retrofitted with a UHPC jacket and tested under simulated blast loads using a shock-tube. The retrofit involved replacing the existing cover with a cast in-situ UHPC jacket (with thickness of 20–40 mm). A companion set of beams was also tested under quasi-static four-point bending. The results from the blast and static tests are compared to a control set of HSC beams built without and with stirrups to examine the ability of UHPC to increase shear and flexural resistance, respectively. The effect of steel ratio (ρ = 1.6% or 2.4%), an important parameter which affects the behavior of HSC and UHPC beams, was also investigated in beams with 15M and 20M bars. Under static loading, the results show that the UHPC jacket was able to prevent shear failure, and improve flexural performance by increasing strength by 40–125%, stiffness by 80–100% and overall toughness by 56–144%, when compared to the companion HSC beams. Under blast loading, the UHPC increased shear capacity, prevented shear failure and improved blast performance by reducing displacements, increasing overall blast capacity and improving damage tolerance. For example, in the 20M set, the UHPC-retrofitted beam showed reductions of 26–63% in maximum displacements, and 88–100% in residual displacements under varying blast intensities. The steel ratio was found to play an important role on the failure mode of the retrofitted beams under static loading, with bar fracture observed in the beam with lower steel ratio of 1.6%. As part of the numerical study, the blast response of the UHPC retrofitted beams was predicted using 3D finite element (FE) modeling. The FE models were able to predict the failure modes and maximum displacements of the control and UHPC retrofitted beams, with an average numerical-to-experimental displacement ratio of 1.03 and an average error of 8%.