Multiscale Simulation of Dynamic Fracture in Polycrystalline SiC-Si3N4 Using a Molecularly Motivated Cohesive Finite Element Method

Abstract

An advanced nanocomposite microstructure such as that of polycrystalline Silicon Carbide (SiC)-Silicon Nitride (Si3N4) nanocomposites contains multiple lengthscales with grain boundary (GB) thickness of the order of 50 nm, SiC particle sizes of the order of 200300 nm and Si3N4 grain sizes of the order of 0.8 to 1.5 μm. Recent developments in failure analyses of such materials focus on continuum calculations with an account of the corresponding atomistic deformation mechanisms. In the presented research one such analysis approach is applied to analyze continuum level deformation in polycrystalline SiCSi3N4 nanocomposites. The continuum bilinear cohesive law is motivated from the atomistic SiC-Si3N4 interfacial separation analyses. The cohesive finite element method (CFEM) based analyses of dynamic fracture at a loading rate of 2 m/sec in bi-modal SiC-Si3N4 nanocomposites with an explicit account of the multiple length scales associated with GBs, second phase (SiC particles), and the primary phase (Si3N4 matrix) are performed. For CFEM analyses bimodal polycrystalline SiC-Si3N4 nanocomposite structures are generated with grain sizes of Si3N4 in the range of 0.8 to 1.5 μm and SiC particle size varying between 200 nm and 300 nm. The volume fraction of the SiC phase is fixed at 30%. In order to analyze the effect of GBs each sample of SiC-Si3N4 nanocomposite has two corresponding meshes: one with finite element (FE) mesh resolving GBs and the other with FE mesh neglecting GBs. Since, a given unique set of phase morphology defining parameters (such as location of SiC particles, SiC or GB distribution etc.) corresponds to a multiple sets of morphologies, three different random sets of morphologies are used to characterize the material behavior corresponding to one unique set of phase morphology parameters. Analyses clearly show that the GBs have a strong effect on dynamic fracture in the nanocomposites.