Uses of empirical mode decomposition and multi-entropy techniques to establish the correlations among vibrations, friction coefficients and component wear of ball-bearing-like specimens

Ball-bearing-like specimens with two groove factors (GFs) are prepared to investigate surface roughness (Ra) effects on tribological parameters and vibrations. The self-adaptive decomposition (SAD) and the intrinsic mode function (IMF) are used to evaluate the multi-scale entropy (MSE). Define entropy tolerance (ET) as the difference in maximum and minimum MSE values. ET variations for the grove-like vibrations created at various IMFs are obtained. Average friction coefficients ((mu) over bar) and wear losses (WL) are evaluated. ((mu) over bar)(outer) << ((mu) over bar)(inner) << ((mu) over bar)(spacer) are the controlling factors for average vibration (AV). Increasing the ball-spacer gap (GD) and lowering GF can reduce AV. Smaller AV, AET, and (WL)(spacer) results are obtained from relatively smaller ((mu) over bar)(spacer). The developments of IMF and MSF provide a novel and efficient way to identify bearing's contact components governing the vibrations and tribological parameters of the tribo tests, and further establish the correlations for the result of AV, AET, (mu) over bar, and (WL)(spacer). (C) 2019 Elsevier Ltd. All rights reserved.