Estimation of Creep Voids Using a Progressive Damage Model and Neural Networks

Abstract

In order to develop nondestructive techniques for the quantitative estimation of creep damage, a series of crept copper samples were prepared and their ultrasonic velocities were measured. Velocities measured in three directions with respect to the loading axis decreased nonlinearly and their anisotropy increased as a function of creep-induced porosity. A progressive damage model was described to explain the observed void-velocity relationship, including the anisotropy. The model study showed that the creep voids evolved from sphere toward flat oblate spheroid with its minor axis symmetrically distributed with respect to the stress direction. This model allowed us to determine the average aspect ratio of voids for a given porosity content. The back-propagation neural network (BPNN) was applied for estimating the porosity content. The measured velocities were used to train the BPNNs, and its performance was tested on another set of creep samples containing 0–0.7% porosity. When the void aspect ratio was used as input parameter in addition to the velocity data, the neural network algorithm provided a much better estimation of the void content.