Electronic system modelling of UT pulser-receiver and the electron beam welding power source

Abstract

Continuous improvements to industrial equipment used in essential industrial applications are a key for the commercial success to the equipment manufacturers. Industrial applications always demand optimum performance and reliability and almost all equipment used in industrial applications is complex and are very expensive to replace. Often modifications to hardware and retrofitting additional hardware are encouraged by most equipment manufacturers and operators. The complexity of these systems however, makes assessment of modifications and design change difficult. This research implemented system modelling techniques to overcome this issue, by developing virtual test platforms of two distinctive industrial systems for enhancement assessment. The two distinctive systems were the electronic equipment called pulser-receiver used in ultrasonic non-destructive testing of safety critical oil & gas pipelines and a high voltage power supply used in high energy electron beam welding. Optimisation with emphasis on portability of the pulser-receiver and rapid weld recovery after a flashover fault condition in the electron beam welding application required assessment before design changes were made to hardware. SPICE based simulators LTSpice and PSpice were used to model and simulate the pulser-receiver and the welding power supply respectively. All the models were evaluated appropriately against theoretical data and published datasheets. However, validation of low level component models developed in the research against measurement data at a component level suffered due to system complexity and resource constraints. Close mapping of simulation results to measurement data at a system level were obtained. The research helped build up a wealth of knowledge in the development of circuit simulation models that can be analysed in the time domain with no non-convergent issues. Simulation settings were relaxed without compromising accuracy of model performance.