Ji Woon Park, John E. Bolander, Yun Mook Lim
Quantifying the rate dependency of the fiber-matrix interface is crucial to evaluating the dynamic behavior of fiber-reinforced cementitious composites (FRCCs). Numerous studies have directly conducted high-speed pullout tests on single or multiple fibers. However, current test methods are limited with respect to quantifying the effects of random fiber dispersion, irregular crack surfaces, and the resulting variations in the inclination and embedment length of fibers within FRCC materials. To address these needs, this study uses a novel inverse analysis procedure to accurately capture the dynamic pullout behavior of randomly distributed fibers in composite materials. This has been achieved using a recently developed multi-scale lattice model combined with an optimization algorithm. The numerical model is capable of simulating the mesoscale response of FRCCs while accounting for the rate-dependent fiber-matrix interface at the microscale. The inverse determination of rate-dependent behavior of the reinforcing fibers in strain hardening cementitious composites is consistent with experimental findings. The results provide new insights into the effects of multiple fiber bundling and inclination within the cementitious composite specimens. In particular, the analysis showed that multiple randomly oriented fibers caused more matrix spalling than single-fiber pullout tests, in agreement with experimental observations. Additionally, fiber pullout behavior was governed more by the propagation speed of the dominant crack than by the overall specimen displacement rate. • Rate-dependent models simulated the observed experimental behavior of FRCC under varying loading rates, reflecting material heterogeneity. • The frictional phase of fiber pullout primarily influences the post-crack tensile strength and material toughness. • The dynamic fiber pullout behavior within the numerical composite model was more dominantly influenced by the progression speed of the most severe crack rather than by the overall displacement rate. • The analysis revealed that the multiple randomly distributed fibers within the composite materials had greater matrix spalling than the single-fiber pullout tests. • When analyzing deformed steel fibers, fibers with less surface deformation were evaluated to have higher DIF values, regardless of the matrix strength.