Error estimation for computed polycrystalline texture characteristics by varying measurement parameters in electron microscopy methods

Abstract

A two-dimensional mathematical model is proposed for a polycrystalline specimen and an electron microscopy experiment with varying measurement parameters, such as the scanning step and the threshold disorientation angle. Experimental results are used to compare specimen texture characteristics and measurements: the grain size distribution, average grain size, variance; disorientation angle distribution, average disorientation angle, variance; and estimates of the orientation distribution function in three-dimensional form in a one-parameter representation. All these distributions are tested by applying a chi-square homogeneity test. The most important aspects of the experiment are formulated as propositions.