A Finite Element Study of Four-Point Bending Creep Tests

Abstract

Due to the cost and difficulty of conducting direct tensile and compression creep testing on engineering ceramics, four-point bending creep test methods are often used as an alternative. Stress distribution in the bending specimen is nonlinear, so a proper interpretation method is needed to get creep properties from data produced by four-point bending creep tests. The method of Hollenberg et al. and the method of Chuang are among the methods to predict the creep parameters from bending creep test data. However, bending creep test methods are often doubted for quantitative creep analysis with reasons like uncertainties from contact point shift or frictional effects in four-point bending creep tests. Finite element simulations of the four-point bending creep tests were performed to evaluate the limitations and abilities of four-point bending creep tests and the methods to predict creep parameters from bend test data. Material model for asymmetric creep behavior (different creep rate in tension and in compression) of ceramics material were developed by modifying the existing symmetric creep model and implemented in the inhouse non-linear finite element code. Explicit finite element method (dynamic relaxation) was successfully used to consider both, the frictional effects between loading rollers and specimen, and asymmetric creep properties of ceramics material. The developed asymmetric creep material model was verified by the simulation of C-ring compression creep test and comparison with published experimental data. It was found that when friction between loading rollers and specimen was not applied in the simulations, both Hollenberg’s and Chuang’s methods well predicted creep parameters from bend creep simulation data. But, when friction was high as in normal bend creep tests, the pre-exponent (A) was highly underestimated. Prediction of stress exponent (n) was not affected much by friction. Bend test set-up with rolling-pins in ASTM C 1211 was recommended to reduce the frictional effects in bend creep tests and a proof simulation was performed. The simulation showed that the test set-up in ASTM C 1211 effectively removed the frictional effect of the frequently used creep test set-up and the effect of bending moment increase due to the rolling of loading pin was minor.