Estimation of high temperature creep parameters for structural ceramics

Power-law high temperature creep parameters of structural ceramic materials are commonly deduced from load-point displacement data obtained form bend tests on the assumption that tensile and compressive behaviours obey the same constitutive law. However, it is now well recognized that this premise may not always be valid because of microcracking and cavitation. Meanwhile, former methods of creep parameters estimation are not suitable, due to a demand for many CPU time. Therefore, a new optimal estimation method is proposed for the structural ceramic creep parameters. Its governing equations are derived from non-coincidence of the neutral axis of a beam under bending with the centroids of the cross-sections, and from the creep response in terms of both curvature rate and load-point displacement rate as functions of the applied moment and creep parameters. Numerical solution of the governing equations is completed with golden section method. In order to obtain optimal estimation of the creep parament, a static optimal calculation program is used which turns the constrained minimum problem into an unconstrained minimum problem by penalty function. Then the problem is solved with variable matrix method that includes both DFP method and BFGS method. An example is given for a set of load point displacement data of sintered nitride silicon beam specimens crept at 900°C and 1200°C from three-point bend tests. The results show that the conventional method overestimates or underestimates the creep rates in compression or tension, and the more accurate analysis presented here is needed.