Non-invasive dynamic condition assessment techniques for railway pantographs

Xin, Tingyu (2019). Non-invasive dynamic condition assessment techniques for railway pantographs. University of Birmingham. Ph.D.

Preview

Xin2019PhD.pdf
Text - Accepted Version
Available under License All rights reserved.
Download (19MB) | Preview

Abstract

The railway industry desires to improve the dependability and longevity of railway pantographs by providing more effective maintenance. The problem addressed in this thesis is the development of an effective condition-based fault detection and diagnosis procedure capable of supporting improved on–condition maintenance actions.
A laboratory-based pantograph test rig established during the course of the project at the University of Birmingham has been enhanced with additional sensors and used to develop and carry out dynamic tests that provide indicators that support practical pantograph fault detection and diagnosis. A 3D multibody simulation of a Pendolino pantograph has also been developed.
Three distinct dynamic tests have been identified as useful for fault detection and diagnosis: (i) a hysteresis test; (ii) a frequency-response test; and (iii) a novel changing-gradient test. These tests were carried out on a new Pendolino pantograph, a used pantograph about to go for an overhaul, the new pantograph with individual parts replaced by old components, and on the new pantograph with various changes made to, for example, the greasing or chain tightness. Through a comparison of absolute measurements and features acquired from the three dynamic tests, it was possible to extract features associated with different failure modes.
Finally, with a focus on the practical constraints of depot operations, a condition-based pantograph fault detection and diagnosis routine is proposed that draws on decision tree analysis. This novel testing procedure integrates the three dynamic tests and is able to identify and locate common failure modes on pantographs. The approach is considered to be appropriate for an application using an adapted version of the test rig in a depot setting.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):