Iranian Researchers Develop Technique for Gear Fault Detection In Machinery

Hassan Alavi, PhD student at Amirkabir University of Technology’s Faculty of Mechanical Engineering and executive director of the "simultaneous detection of defects in helical gears based on vibration and sound signal processing" project, said that early detection of faults in machinery matters a lot as it reduces maintenance and repair costs considerably.

“Several faults usually appear simultaneously in machines. Many researches have been done to detect only one defect, but there are few researches in the field of simultaneous defects detection,” Alavi said, according to the public relations department of Amirkabir University of Technology.

“In this research, new methods for identifying and classifying faults at the same time were proposed and they were evaluated from both theoretical and experimental perspectives. These proposed methods perform better as compared to previous methods,” he continued.

“In this project, the use of sound signal from the machine was studied in fault detection. since the measurement and recording of signal is simple and accessible, sound-based fault detection makes the fault detection process easier," the researcher said. 

“In this research, theoretical simulations and, in parallel, extensive experimental tests were used in the acoustic research laboratory of the Mechanical Engineering Faculty of Amirkabir University of Technology to find out the effect of defects on the vibration and sound signal produced by the gearbox,” Alavi added.

According to him, in the end, creative techniques to separate the symptoms of simultaneous defects and new methods in the field of artificial intelligence, which are known as deep learning, have been used to classify the type and severity of defects.

 “Due to the widespread use of gearboxes in various machines, the results of this research can be used in the maintenance and repair of industrial machinery, wind turbines, some flying objects and machines such as helicopters and UAVs, as well as cars and even some house appliances such as washing machines,” he went on to note.

"With our proposed methods, detecting very slight defects and simultaneously detecting faults in more complicated machines that happen in other parts of the machine other than the gearbox becomes possible," he went on to say. 

According to the researcher, the methods they have proposed are completely new and have no domestic or foreign equivalents.

He continued that improving the accuracy of faults detection and simplifying the detection of defects through sound signal and vibration is one of the advantages of their method.

“Creating or improving machinery condition monitoring and troubleshooting systems is another feature of this project,” Alavi concluded.

4155/i