Detection and localization of isolated wear bearing faults of rotating machinery by spectral analysis

Abstract

Rotating machines play a strategic role in a manufacturing process, and it is the case of a cement mill. These machines are composed of fragile bodies (bearings and gears, etc.) subjected to mechanical stress and harsh industrial environment.Sources of failure are many; the wear and the spalling of bearing, the wear and fracture of a tooth of a gear, the misalignment of the axes, etc. To avoid unexpected downtime and costly, those bodies must be monitored continuously. There are several ways to do that; vibration analysis, lubricant analysis, infrared thermography and acoustic analysis. Vibration analysis (frequency) is most often used by operating a signal from an accelerometer. Whenever a fault participated in a contact, it changed the vibration characteristics of the system. By analyzing the acceleration measurements from an accelerometer, we can identify and quantify these changes in order to establish a relationship between the measures and the type of defect, which will be used as signatures for fault diagnosis system. In a system of power transmission (gear unit of a cement mill) the frequency of occurrences of faults can be known from the geometry and kinematics of the system.This article deals with the detection of bearing defects isolated in a gear unit of a cement mill by spectral analysis of vibration; which based on a systematic analysis spectral and envelope spectrum of vibration signal for the presence of images of all vibrational defects may affect the body in question.