Mechanical blending and microscopic image processing for evaluating homogeneity in nano- and micron-sized powder mixtures

Abstract

This paper presents the results of mechanically blending AlSi₁₀Mg micropowder with Co₃O₄ nanoparticles, resulting in the formation of a nanoparticle coating on the surface of the coarse particles. An automated computer program was developed using Python's OpenCV library to process EDX elemental mapping images and evaluate the homogeneity of a mixture of nano- and micron-sized powders through surface analysis of sample images. The proposed methodology employs Otsu's thresholding technique to segment images into regions of interest and calculates quantitative data on the areas occupied by micron- and nanoscale components, which are key indicators for assessing powder mixture homogeneity. The developed technique is cost-effective and demonstrated its efficacy in measuring the homogeneity distribution of nanoparticles within a micropowder using a set of microscopy images. The only requirement for the method's applicability is a difference in the elemental composition of the powders. In this work, the homogeneity of micron-sized AlSi₁₀Mg powder mixed with varying concentrations of cobalt oxide nanoparticles was analyzed using the processing of elemental mapping images. The results showed that 8 h of blending in a planetary ball mill with grinding balls was sufficient to achieve a homogeneous powder mixture.