An Enhanced Automated Particle Angularity Measurement Method

Abstract

Abstract Angularity is an important aspect of particle morphology, directly correlated to the mechanical response and strength of granular materials. Many researchers have proposed multiple direct and indirect methods such as visual charts, analytical formulas, and advanced image-based techniques for qualitative and quantitative assessment of particle angularity, but no single method has been fully successful in distinguishing all types of particles and obtaining an accurate quantitative value of angularity. Nevertheless, because of the significant importance of the parameter, researchers have been using a few generally accepted angularity quantification methods. This paper evaluates the conceptual merits and demerits of four such methods, namely, two analog methods called “roundness index” and “degree of angularity” coined by Wadell and Lees, respectively, as well as two image-based methods called “angularity using outline slope” and “gradient angularity index” proposed by Rao, Tutumluer, and Kim and Chandan et al., respectively. The latter two methods were developed for use in conjunction with the University of Illinois Aggregate Image Analyzer and the Aggregate Imaging System, respectively. A new approach for angularity quantification is proposed herein based on insights from a rigorous analysis and critical assessment of these methods. The proposed method is compared against the existing methods based on Ward’s linkage method of clustering and is shown to be superior at distinguishing between various classes of particles. The proposed approach is demonstrated to be a better measure of the sharpness of the meso-level geometrical features of importance along a particle’s outline that are the controlling parameters contributing to particle angularity, which consequently correlates to the kinematic and mechanical response of granular materials.