Title: High strength aluminium alloy fatigue damage alert of high speed train gearbox shell using acoustic emission instrument

Authors: Yibo Ai; Tao Lv; Weidong Zhang

Addresses: National Center for Materials Service Safety, University of Science and Technology Beijing, Beijing 100083, China ' National Center for Materials Service Safety, University of Science and Technology Beijing, Beijing 100083, China ' National Center for Materials Service Safety, University of Science and Technology Beijing, Beijing 100083, China

Abstract: As a key component of high speed train, the gearbox shell must be running safely. The main damage form of high speed train gearbox shell is fatigue, and to effectively predict the working state and give out safety alert is of great significance of operation safety. In this study, the acoustic emission instrument has been used for real-time and non-destruction monitoring fatigue damage progress of high strength aluminium alloy which is the material of high speed train gearbox shell. By comparing with the fatigue damage progress, the feature parameter and its threshold of acoustic emission (AE) signal for classifying the states has been defined. The consistence of the feature is discussed by Hurst index method. A particle swarm optimisation-least square support vector machines (PSO-LSSVM) prediction model has been designed to predict the feature of next step, and the safety alert is given by comparing with the threshold of the feature. In this study, the prediction result is about 600s to 1600s earlier than the critical time, and by comparing acceleration test and real condition, it can give enough time for the train to stop and evacuate passengers.

Keywords: high strength aluminium alloy; material fatigue damage; acoustic emission; safety alert; PSO method; SVM method; high speed train gearbox; manufacturing system.

DOI: 10.1504/IJMMS.2018.091176

International Journal of Mechatronics and Manufacturing Systems, 2018 Vol.11 No.1, pp.36 - 52

Received: 11 Jul 2017
Accepted: 28 Nov 2017

Published online: 11 Apr 2018 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article