FRACTURE STRENGTH SIMULATION OF SiC MICROTENSILE SPECIMENS - ACCOUNTING FOR STOCHASTIC VARIABLES

This paper summarizes work performed to predict the room-temperature strength of SiC micro-scale tensile specimens with introduced stress concentration using NASA Glenn Research Center's Ceramics Analysis and Reliability Evaluation of Structures Life prediction program (CARES/Life). A polycrystalline and a single crystal SiC material were tested. CARES/Life tended to over predict strength of the polycrystalline material but had relatively better success predicting the strength response for the single crystal material. Large columnar grains and rough side-walls for the polycrystalline material likely reduced the applicability of continuum analysis and therefore the accuracy of the CARES/Life analysis. The single crystal material did not have (or had less of) these complications. Significant specimen-to-specimen dimensional variation existed in the specimens. To simulate the additional effect of the dimensional variability on the probabilistic strength response for the single crystal specimens the ANSYS Probabilistic Design System (PDS) was used with CARES/Life. The ANSYS/PDS - CARES/Life simulation was not rigorous. Isotropic material and fracture behavior was assumed in the analysis. However the feasibility of using these programs to account for multiple stochastic variables on the strength response of a structure was demonstrated.