A Novel Signal Processing Method for Friction and Sliding Wear

Abstract This current study proposed a new computationally efficient and comparatively accurate algorithm for calculating both static and dynamic coefficients of friction from high-frequency data. Its scope embraced an application in a real-time friction-based system, such as active braking safety systems in automobile industries. The signal sources were from a heavy-duty reciprocating dry sliding wear test platform, focused on experimental data related to friction induced by stick-slip phenomena. The test specimen was a polytetrafluoroethylene (PTFE)-coated basalt/vinyl ester composite material, tested at a large scale. The algorithm was primarily aimed to provide scalability for processing significantly large tribological data in a real-time. Besides a computational efficiency, the proposed method adopted to evaluate both static and dynamic coefficients of friction using the statistical approach exhibited a greater accuracy and reliability when compared with the extant models. The result showed that the proposed method reduced the computation time of processing and reduced the variation of the absolute values of both static and dynamic frictions. However, the variation of dynamic friction was later increased at a particular threshold, based on the test duration.