Lubricant quality control: A chemometric approach to assess wear engine in heavy machines

Abstract

Used lubricants assessment could be suitable to reveal wear mechanisms undergoing in a machine. In the proposed work an effective, reliable and cost-efficient technique combining inductively coupled plasma optical emission spectrometry (ICP-OES) and chemometrics was developed to perform wear analysis for industrial application. Thus, multielemental determination of a large amount of used lubricants from diesel heavy machines (truck, bulldozer, loaders and backhoe) was carried out by ICP-OES. Multivariate analysis was done by the non supervised method principal component analysis (PCA) and supervised method sequential discriminant analysis (SDA). The recognition ability for the groups was highly satisfactory. For PCA, the main three principal components explained 99.98% of total variance for samples lubricants. SDA allowed, in training and prediction sets, a correct classification of 97.12% and 95.75% respectively for the lubricant analyzed samples, obtaining three groups classified according to the use. The outcomes demonstrate that lubricant wear effect could be assessed quickly and accurately, having great potential to become a very useful tool in the machine monitoring industry.