Answers are grounded in this paper and cite their sources.
In bio-lubricant research, worn surfaces are usually interpreted qualitatively from scanning electron microscopy (SEM). This study tested how much tribological information objective, brightness-invariant texture descriptors can recover from an archive of such micrographs. Fifty-two ×1000 fields of brass, aluminum and mild steel worn against stainless steel at 40 N under fully synthetic oil and a 50/50 blend with waste cooking oil (WCO) at 22, 40 and 80 °C, plus four unworn references, were described by ten orientation, correlation-length, spectral and feature descriptors. Each condition is one specimen, so fields are repeated measurements; the analysis is descriptive at field level and inferential only at specimen level (13 specimens). Aluminum surfaces were less directional than brass and mild steel (orientation order 0.71 against 0.96 and 0.91), but a similar difference existed before wear, and at specimen level it was not significant (p = 0.09). Mild steel at 80 °C was strongly grooved (anisotropy 21 against 1.7 at 40 °C), with a fifteen-fold rise in measured roughness. Replacing half the synthetic oil with WCO left no detectable signature, and none of the 30 descriptor–response correlations were significant. Operator-selected single SEM fields document gross morphology but cannot rank lubricants or infer wear rate; an imaging protocol is proposed.
No takes yet. Share an insight, caveat, or question.
Alajmi et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: