*** 9712291 Geubelle An experimental and analytical research project will investigate the fiber debond and pull-out processes in fiber reinforced polymeric and ceramic matrix composites subjected to dynamic loading conditions. While extensive work has been performed regarding these two failure mechanisms under quasi- static loading conditions, and while all preliminary investigations on that topic indicate a strong dependence on the loading rate, very little is currently known on how these processes take place under dynamic loading conditions. The joint experimental and analytical research project aims at shedding some light on the dynamic debonding and pull-out processes and constitutes a mandatory preliminary step for the design of impact-tolerant composite structures. The experimental part of the project consists of dynamic push-out and pull-out tests performed on various model composites with the aid of a specially adapted Split Hopkinson Bar apparatus. In addition to macroscopic load/displacement curves obtained from the elastodynamic response of the bars, in situ optical observations of the debonding and pull-out processes will be performed using a photoelastic set-up and a high speed camera, providing independent measurements of the debonding length and speed. In situ thermographical measurements on the model composite and post mortem microscopic analyses of the fracture surfaces will also be performed to capture the possible occurrence of localized frictional heating which has been shown in some preliminary investigations to drastically affect the pull-out process.*** The analytical part of the project has two basic objectives: firstly, detailed finite element simulations of the experimental set-up will be performed to characterize with precision the dynamic loading of the composite specimen during the tests. Secondly, two distinct and complementary numerical schemes (a special volumetric/cohesive dynamic finite elemen t scheme and a spectral form of the boundary integral method) will be used to investigate the adequacy of existing cohesive and friction models to capture the observed fiber debond and pull-out processes.***
