Intelligent condition monitoring of high-power laser systems

Howlett, R.J., Dawe, G. and Walters, Simon (2007) Intelligent condition monitoring of high-power laser systems Journal of Systems Science, 33 (2). pp. 51-59. ISSN 0137-1223

Full text not available from this repository.


Powerlase Limited designs and manufactures diode-pumped solid state Neodymium Yttrium Aluminium Garnet (Nd:YAG) lasers for industrial applications in the materials processing and microelectronics marketplaces. Powerlase has developed new advanced technology which enables the combination of high intensity short pulse outputs with high energies and high repetition rates to yield several hundred watts of good beam quality. The Company's products find application in aerospace, motor vehicles, microelectronics, PCB production, ablative lithography, and many other areas. These lasers are used in manufacturing processes where the avoidance of unplanned downtime and the ability to maintain high beam stability and consistent power output is important. This paper is a case study describing a project to investigate, design and implement a condition monitoring system utilising intelligent techniques. The aim of the project was the detection of the degradation in the system, as an indicator of future problems, but before it became severe enough to adversely affect the manufacturing process. The development of a strategy is described for the analysis of the system to deduce the physical system variables to be monitored. A brief account is given of a rule-based methodology for analysing the system variables to predict the early onset of failure. A neural network technique is described that detects reduction in power output through the automated analysis of the drive current-power characteristics of the light amplification units. The power level monitor created permits measurement of the output of the laser using an inexpensive power sensor and without the necessity for an expensive high-accuracy power meter.

Item Type: Journal article
Uncontrolled Keywords: condition monitoring; diagnostics; neural networks; lasers; manufacturing; fault finding
Subjects: H000 Engineering > H300 Mechanical Engineering > H330 Automotive Engineering
H000 Engineering > H300 Mechanical Engineering
Faculties: Faculty of Science and Engineering > School of Computing, Engineering and Mathematics > Engineering and Product Design Research > Automotive Engineering
Depositing User: editor environment
Date Deposited: 14 Mar 2008
Last Modified: 29 Apr 2015 09:07

Actions (login required)

View Item View Item


Downloads per month over past year