Of fiery sparks & glittering spots: Melting-resolidification and spherical particle formation in abrasion

Abstract

The curious occurrence of perfectly spherical particles when a steel substrate is slid against a hard abrasive was first observed and documented by Robert Hooke in the 17$^{th}$ century. Similar particles have subsequently been observed in numerous other abrasion-type processes, ranging from grinding of steels to sliding rock faults. The prevalent hypothesis, originally proposed by Hooke, is that these particles are formed due to high local temperatures between the abrasive and the substrate, resulting in melting, droplet ejection and subsequent resolidification -- the melting-resolidification hypothesis. In this work, we revisit this phenomenon using \emph{in situ} analysis of a model steel-abrasive contact geometry, complemented by analytical calculations. It is found that the temperature within the contact zone, for typical contact conditions used, is far from the melting point and that spherical particles do not form in the absence of oxygen. We thereby propose a modification of the melting-resolidification hypothesis, involving an intermediate exothermic oxidation stage, and provide quantitative evidence for each step of the process. Our results have implications for a wide class of abrasive systems that involve the formation and utilization of spherical metallic particles