Transient Stray Flux Analysis Via MUSIC Methods for the Detection of Uniform Gearbox Teeth Wear

Gearboxes are very important transmission components widely used for industrial productions in aerospace, automotive and heavy industry due to their high transmission efficiency and strong load-bearing capacity. During their service life, gearboxes are subjected to constant stresses, which causes an imminent and gradually wear on the tooth gear. Most current methodologies for gearbox fault diagnosis are based on vibration and current signals, which relies on the feature of specific fault harmonics. However, their respective time-frequency and power spectrum representations are blurred due to noisy signals, compromising the final diagnosis. This paper presents a novel methodology for uniform tooth wear in gearboxes with two main contributions: the proposal is based on stray flux signals, and their transient analysis with a high-resolution MUSIC method. This solution mitigates the effects of processing noisy signals and evidences only larger frequency components. The obtained results are promising and demonstrates the effectiveness of the proposal to become an excellent alternative to classical approaches.