Spalling in rolling element bearing is a common localized defect generated during the operation of bearing due to rolling fatigue. Size estimation of such localized defect can be helpful in determining severity of the fault. This further can be used as an input for predicting the remaining useful life of the bearing. The popular approach for estimating the size (in terms of width) of spall is to trace the entry and exit events of rolling element while interacting with the fault. The time estimated between entry and exit of rolling element from pit like spall can be converted to the geometric estimation is the fault size from vibration signature. The present approach demonstrates the use of Singular Spectrum Analysis (SSA) to accomplish the task. The vibration signals generated from the interaction of rolling element with the localized fault is a hybrid signal consisting of low frequency stepped response generated while the rolling element enters the fault and is superimposed on to the high frequency impact generated during the reentry of the rolling element in to the raceway from the spall. The signal information is enhanced via preprocessing the signal with Total Variance Regularization (TVR) filtration. The informative signal, which is extracted from the row temporal signal via SSA, aids in the accurate identification of entry and exit events. The proposed method integrating TVR with SSA for faut size estimation is validated using simulated signals and experimental signals from independent resources. The result shows strong agreement with the accuracy level of size estimation.

No abstract

No abstract

No abstract

No abstract