Crack detection of Alstom diesel locomotive liners using wavelet and short time Fourier transform

Due to the important and large contribution of rail transport and the high costs of machinery and equipment used in this section، the use of modern maintenance and repair methods is an inevitable necessity. The present study has been done experimentally in cooperation with “Bafgh locomotive repair factories complex” and has been defined due to its specific need. In this study the vibration analysis capabilities in order to find the cracked liner of Alstom diesel locomotives، operating in Iran’s railroads، have been discussed. The goal is that by mounting accelerometer sensors on the outer wall of locomotive engine and processing and analysis of collected signals، the faulty liner could be identified. Due to the non-stationary nature of collected signals، Short Time Fourier Transform and Wavelet Transform have been adopted for processing of these signals. The results show that both of these transforms are efficient in distinguishing the faulty and normal liners and can be an appropriate alternative to the current traditional methods. Among the benefits of replacement of this method with current methods، repair time reduction، running costs reduction، engine down time period in repairs section reduction، efficiency increment and timely and reliable detection can be named. In case of lack of timely diagnosis and completion of crack، irreparable damages will come. Sometimes the damage is so that the engine is completely outdated. This was the main motivation of the present study.