THE DESIGN AND IMPLEMENTATION OF A SYSTEM FOR MACHINE MONITORING

This thesis describes the research, design, manufacturing and testing of a novel multi-sensor machine monitoring system: this system integrated with existing programmable logic controller (PLC) systems and wirelessly transmitted condition data. The research began by exploring machine monitoring strategies and the individual sensor topic areas. Methods for interfacing with PLCs and microcontroller-based wireless transmission systems were also explored. Potential electrical and physical interferences in the system were identified alongside methods to minimise them. The designs were then simulated, prototyped and revised before final manufactured designs were practically tested on factory machines. The Clutch and Brake (C&B) sensor measured the wear of the pads in 0.08cm windows with a minimum detectable increment of 0.02cm. The C&B sensors total range was 26cm which meant the sensor could measure to a resolution of 0.08% of its total range with a sensitivity of 0.27V/cm. The wear conditions were correctly reported to the PLC which activated the programmed corrective actions. The Vibration Analysis (VA) sensor measured vibration in different motors and highlighted mechanical looseness in a motor which had recently undergone maintenance. The bandwidth of the VA sensor was 1kHz with an amplitude of acceleration within the ±16g range. The resolution of the acceleration measurement was 3.9mg which was 0.02% of the established 20g measuring range. The conditions of the motors were determined by programming which was precise to 3dp. The Non-Intrusive Pipe Pressure (NIPP) sensor monitored the pressure inside a pipe in 10bar brackets which equated to 18mV: the entire sensor range covered 0.18V which was only 3.6% of the available microcontroller range. This allowed a clear press pressure cycle to be graphically displayed from the NIPP sensor data. A leak was simulated in the PLC program and the pumps were promptly shutdown as intended. All the sensors wirelessly transmitted their conditions to a central graphical device which generated an accompanying graphical user interface. Practical methods utilised to minimise the effect of interferences in the system such as decoupling capacitors, data averaging and filtering ensured that the conditions were reported accurately in the factory environment; achieving the aims of the project. The developed multi-sensor system improved the condition monitoring strategies of the factory by reducing the downtime of equipment and enabling better utilisation of maintenance hours. The automated monitoring aspect of the system – coupled with the integration with the PLC system – made it viable for wide implementation across the factory. Overall, this system identified conditions accurately, automatically and succinctly which allowed it to succeed in a working factory environment.

Grantor dc:publisher

Oxford Brookes University

Year dc:date

2020

Author dc:creator

• Noble, Ben

Contributors dc:contributor

• Barker, Steve
• Hayatleh, Khaled
• Ben-Esmael, Mohamed

Statement dc:rights

• All rights reserved

Language dc:language

en

DOI dc:identifier

https://doi.org/10.24384/mjt8-t610

OAI identifier oai:identifier

tle:e22e9c32-4e6a-4fe1-96a1-8e72f961c7d2:d6bd9758-527a-46cd-bfe2-c433766e8fca:1