A Novel Approach of Gearbox Fault Diagnosis by Using Time Synchronous Averaging and J48 Algorithm

Abstract

Gearbox fault diagnosis is a major research area in mechanical engineering at the current generation. Due to the important role in industries, gearbox makes a huge impact on mechanical engineering research. A lot of faults occur in the gearbox such as tooth crack, wear, plastic flow, and spelling. These types of gearbox faults can lead to major production disaster which is the cause of huge financial losses and decrease the efficacy in the production unit. Thus, it is very important to detect any fault at an initial stage to stay away from an unexpected breakdown of the gearbox. Therefore, this paper presents a unique gearbox fault diagnosis method based on time synchronous averaging (TSA) and J48 algorithm. An electromechanical system was used for experimental work. Two types of faulty gearset were introduced along with a healthy gearset for the study of the gearbox vibration. A computer-aided DAQ was used for acquiring the vibration signal. The experimental work was carried out by three different loads and speeds conditions with duration of 30 s. The results of this work are very promising and make a clear idea of some future research work.