THERMAL ENERGY HARVESTING IN WIRELESS SENSOR NODES USED FOR CONDITION MONITORING

Presently, wireless sensor notes (WSN) are widely investigated and used in condition monitoring on industrial process monitoring and control, based on their inherent advantages of lower maintenance cost, easy installation and the ability to be installed in places not reached easily. However, current WSN based monitoring system still need dedicated power line or regular charging / replacing the batteries, which not only makes it difficult to deploy it in the fields but also degrades the operational reliability.

This PhD research focuses on an investigation into energy harvesting approaches for powering WSN so as to develop a cost-effective, easy installation and reliable wireless measurement system for monitoring mission critical machinery such as multistage gearbox.

Among various emerging energy harvest approaches such as vibrations, inductions, solar panels, thermal energy harvesting is deemed in this thesis to be the most promising one as almost all machines have frictional losses which manifest in terms of temperature changes and more convenient for integration as the heat sources can be close to wireless nodes. In the meantime, temperature based monitoring is adopted as its changes can be more sensitive to early health conditions of a machine when its tribological behaviour is starting to be degraded. Moreover, it has much less data output and more suitable for WSN application compared the mainstream vibration based monitoring techniques.

Based on these two fundamental hypothesis, the research has been carried out according to two main milestones: the development of a thermoelectric harvesting (TEH) module and the evaluation of temperature based monitoring performances based on an industrial gearbox system. The first one involves the designing, fabricating and optimising the thermal EH module along with a WSN based temperature node and the second investigates the analysis methods to detect the temperature changes due to various faults associated with tribological mechanisms in the gearbox.

In completing the first milestone, it has successfully developed a TEH module using cost-effective thermoelectric generator (TEG) devices and temperature gradient enhancement modules (heat sinks). Especially, the parameters such as their sizes and integration boundary conditions have been configured optimally by a proposed procedure based on the fine element (FE) analysis and the heat generation characteristics of machines to be monitored.

The developed TEG analytic models and, FE models along with simulation study show that three different specifications of heat sinks with a Peltier TEG module are able to produce power that are consistently about 85% of the experimental values from offline tests, showing the good accuracy in predicting power output based on different applications and thus the reliability of the models proposed. And further investigation shows that a Peltier TEG module based that the thermal energy harvesting system produces is nearly 10 mW electricity from the monitored gearbox. This power is demonstrated sufficient to drive the WSN temperature node fabricated with low power consumption BLE microcontroller CC2650 sensor tags for monitoring continuously the temperature changes of the gearbox.

Moreover, it has developed model based monitoring using multiple temperature measurements. The monitoring system allows two common faults oil shortages and mechanical misalignment to be detected and diagnosed, which demonstrates the specified performance of the self-power wireless temperature system for the purpose of condition monitoring.