Detection and diagnosis of electrical motor-pump faults using computer based frequency analysis technique

In developed countries, a considerable amount of improvement in production rate and saving has been achieved by means of maintenance techniques using vibration measurement and tracking. In recent years predictive maintenance (PdM) as an application of vibration engineering has been of interest in our country and many companies have started to examine the technology. The aim of this study is to show that the technology which has been used by the industrialized companies can be used in our country and to show the benefits of the technology for the industry. For that reason, some pumping station of Istanbul Water and Sewerage Administration (İSKİ) are selected as case studies and the PdM method was applied. A historical view to the fault detection by using vibration analysis is given in Chapter I. and studies performed till now are evaluated. In Chapter 2, vibration theory is worked out and basic vibration types and vibration terminology are explained. Relationship between displacement, velocity and acceleration is explained. In Chapter 3, Classical maintenance method and PdM methods are explained and compared with each other, and advantages / disadvantages are given. In Chapter 4, FFT methods which is used to convert time domain vibration data in to frequency domain are explained. Sampling techniques used during vibration measurement and signal aliasing, leakage, windowing, averaging and overlopping are explained. In Chapter 5, types of the transducers used in vibration measurement, transducer selection and determination of the measurement points and direction on the motopomp (motor and pump) are shown. In Chapter 6, detection of bearing faults, unbalance, mechanical looseness, eccentricity, cavitations, resirculation, turbulence, problems of stator and rotor using PdM are given. This study is compared with those of Crawford and Eshleman. Economical merits of this study are explained through an example. The working condition to utilize PdM method satisfactorily, motopomp test scheduling and tracking of vibration trends are explained in Chapter 7. In Chapter 8, diagnostic application of each fault detected by means of the vibration analysis techniques given in Chapter 6 are performed, the vibration spectrum showing these faults are analyzed and the motopomp trend curves are derived. In conclusion, advantages of PdM for a company, types of faults which can be detected by using vibration analysis and the parameters used in PdM methods are explained. Keywords : Predictive maintenance, vibration analysis, vibration measurement, Diagnosis, Fault diagnosis, pumps