Method for predicting destruction of a bearing utilizing a

Abstract

A method for predicting destruction of a bearing preliminarily finds frequency range of acoustic emission signals indicating fatigue of bearings. In order to find the fatigue-related frequency range, frequency components of electrical noise, mechanical noise, acoustic emissions (AE) due to slip of a bearing, and AE due to plastic-deformation of the bearing are respectively obtained under corresponding conditions, using a test equipment. Then, under a certain condition, combined AE signals of electrical noise, mechanical noise, slip-related AE, plastic-deformation-related AE and rolling-fatigue-related AE are generated. By subtracting the frequency components of electrical noise, mechanical noise, slip-related AE, plastic-deformation-related AE from frequency components of the combined AE signals, the fatigue-related frequency range is obtained. This frequency range is used to predict destruction of a bearing. AE signals in the preliminarily found frequency range are extracted from all acoustic emission signals detected. The extracted AE signals are converted into digital signals and compared with a predetermined reference value. When the digital signals exceed the predetermined reference value, a preliminary indication of inner destruction of the bearing is discriminated.