Data for: A multi-parameter ultrasonic evaluation of mean grain size using optimization

Published: 27 May 2019| Version 1 | DOI: 10.17632/v487vwmd7r.1
Xi Chen, Guanhua Wu, Zhenggan Zhou, chen hao


The data obtained from ultrasonic and quantitative metallography measurements on GH4169 samples, is used to built a an optimized multi-parameter ultrasonic evaluation model for grain diameter. The performance of the new model described in this study was benchmarked against three single parameter models (i.e. velocity, attenuation, and backscattering) using two test samples. The abstract are list follow. An evaluation model based on multiple ultrasonic parameters was developed to control mean absolute error between an actual measured value and a model calculation value. Superalloy GH4169 was used to validate the presented model. A multi-dimensional ultrasonic parameter set was built for each sample by ultrasonic and nonlinear ultrasonic test and parameter calculation. The feature scale of different parameters was restricted to the same range using normalization. The dimensionality of the parameter set was descended by a correlation metric with grain size. A new one-dimensional ultrasonic characteristic parameter was comprised of the select normalized set of ultrasonic parameters under application of a mapping function in the form of a quadratic polynomial with undetermined coefficients. The relationship between the new ultrasonic multiple eigenvalues mapping parameter and grain size was expressed using a linear fitting function. An optimization problem aimed at minimizing the mean absolute error between the mapping and fitting parameters was solved using evolutionary algorithms to obtain coefficients for the mapping and fitting functions. These calculations were used to generate the full multi-parameter ultrasonic evaluation model, which was benchmarked against three single parameter evaluation models (i.e. velocity, attenuation coefficient, and backscattering). The mean grain size evaluation value of GH4169 using this new multi-parameter model is sound and precise.



Materials Science, Materials Testing, Materials Characterization