Minimum mesh quality for reliable morphology characterization of 3D soil particles

Abstract

Remarkable advancements in computer-based optical imaging technology have greatly enhanced particle shape quantification and characterization with high efficiency. However, the varying sensitivity of different particle shape descriptors to mesh quality is often overlooked, which can affect the reliability and accuracy of particle shape characterization. In response to this issue, this study systematically investigates the impact of three-dimensional (3D) particle mesh quality on seven commonly used shape descriptors, showing that lower mesh quality can lead to greater deviations in 3D particle shape characterization. These shape descriptors are then categorized according to their scale and sensitivity to mesh quality. From this analysis, minimum mesh quality criteria are established to ensure reliable particle shape characterization. The applicability of the criteria to particles with diverse forms is evaluated, and further validation is conducted using four scanned morphologically featured particles. These findings provide essential technical guidance for determining acceptable levels of 3D particle mesh model simplification in geotechnical engineering applications.