This thesis investigates the application of various types of piezoelectric materials including piezo crystalline and piezo polymers in condition monitoring of electric machines. The condition monitoring is a procedure for detection of defects and failures in mechanical and electrical parts of an electric machine which can prevent severe damages to the structure machine by early detection of the failures. Piezoelectric materials are among smart materials which are sensitive to mechanical strain and they can generate electric due to the mechanical vibration. Thus, despite passive strain and accelerometer sensors, a piezoelectric vibration sensor does not need an extra power source as a bias for sensing vibrations. This feature enables using piezoelectric materials as a self-power sensor in remote and wireless sensing applications. Currently, the induction machines are horse power of various industries and the condition monitoring of these machines, due to vast amount of their applications in recent decades, is among demanding research topics. Condition monitoring of an electric machine is conventionally established based on processing of the signals of the vibration sensors in frequency domain. Owing to the recent advances in microelectronics in development of high-speed and inexpensive processors, discrete signal processing (DSP) algorithms can now be readily implemented within the DSP boards. The condition monitoring algorithms are often realized based on digital processing of captured signals from accelerometers and vibration sensors. Thus, the sensitivity of sensors to vibration and the signal-to-noise ratio of sensors are significantly important in fast and accurate detection of failure in electric machines. In this thesis, firstly the suggested processing algorithms for condition monitoring are reviewed and the most suitable methods which can be used for piezoelectric sensors are selected. Then, various piezoelectric sensors for this application are practically investigated. Since the investigation of the piezoelectric sensors needs the model and characterization of the sensors, a vibration test bed is developed that mainly includes a computer-controlled vibration plate with electromagnetic vibration exciter. The vibration test setup is equipped with an accelerometer sensor which is used for characterizing various piezoelectric sensors. Then, after characterizing the piezoelectric sensors, the application of these sensors are investigated using an experimental test setup for condition monitoring of the electrical induction machines. In these tests, the performance of polymer and crystalline types of piezoelectric sensors in detection of bearing defections in electric machine are investigated. The test results show that piezo polymer sensors due to flexibility and ease of installation as well as high signal to noise ratio are among promising sensors for condition monitoring of electric induction machines. Key words Condition monitoring, induction machines, piezoelectric sensors, sensor characterization, test-bed shaker.
