Experimental investigation and multiscale modeling of ultra-high-performance concrete panels subject to blast loading

Tailored cementitious materials, such as Ultra-High-Performance Concrete (UHPC), may significantly improve the blast resistance of structural panels. To understand and quantify the performance of UHPC panels subject to blast loading, four 1626- by 864- by 51-mm UHPC panels without steel rebar reinforcement were subjected to reflected impulse loads between 0.77 and 2.05 MPa-ms. The UHPC material was composed of a commercially available UHPC premix, high-range water reducing agent, 2% volume fraction of straight, smooth 14-mm-long by 0.185-mm-diameter fibers, and water. Experimental results determined that the UHPC panel fractured at a reflected impulse between 0.97 and 1.47 MPa-ms. These results were used to validate a multiscale model which accounts for structure and phenomena at two length scales: a multiple fiber length scale and a structural length scale. Within the multiscale model, a hand-shaking scheme conveys the energy barrier threshold and dissipated energy density from the model at the multiple fiber length scale to the model at the structural length scale. Together, the models at the two length scales account for energy dissipation through granular flow of the matrix, frictional pullout of the fibers, and friction between the interfaces. The simulated displacement and fracture patterns generated by the multiscale model are compared to experimental observations. This work is significant for three reasons: (1) new experimental data provide an upper and lower bound to the blast resistance of UHPC panels, (2) the multiscale model simulates the experimental results using readily available material properties and information regarding mesostructure attributes at two different length scales, and (3) by incorporating information from multiple length scales, the multiscale model can facilitate the design of UHPC materials to resist blast loading in ways not accessible using single length scale models. (C) 2014 Elsevier Ltd. All rights reserved.