On-machine evaluation of micro-EDM process signature in radio frequency (RF) domain: A step towards cost-effective data collection in a multiphysical process

In the growing age of digitization, there is a need for cost-effective data collection techniques to enable manufacturing data collection of production processes in small and medium scale enterprises (SMEs). This goes further where electrophysical processes, such as micro electrical discharge machining (micro-EDM) are employed in high-end micro-components such as fuel injection nozzles and precision components for automotive and aerospace sectors. There have been several efforts to get process information using electrical and acoustic monitoring, AI-based and multiphysics models, and fundamental physics experiments. All aforementioned techniques have contributed to improving understanding of the discharge behaviour in micro-EDM processes. However, they are all expensive and invasive in nature, thereby limiting their application for SMEs in enabling digitalisation of process-chain involving micro-EDM process.

To address this gap, this paper investigates micro-EDM process signature in radio frequency (RF) domain and hence RF-based strategy is proposed as a low-cost and non-invasive process monitoring technique. Research is conducted to investigate the radiation sources and potential influencing factors during the EDM process. First, the RF mechanism in EDM is examined by analysing the monitored RF signals across different discharge stages of various pulse waveforms. From this analysis, the corresponding equivalent RF radiation models are established. Second, through an experimental design, the impact of different machining parameters, dielectric fluids, discharge states, monitoring equipment, and workpiece materials on RF signals are explored in both the time and frequency domains. The results indicate that only parameters related to breakdown voltage and gap current affect RF intensity, however, machining parameters show to be not significant on the radiation spectrum. The influence of dielectrics and workpiece materials on radiation behaviour is primarily associated with their corresponding equivalent resistance. Notable differences in both discharge pulses and antenna devices in monitoring RF signals are observed. Additionally, the monitorability of RF strategy is also evaluated under two application scenarios. By investigating the RF signals in EDM, it is expected to provide a non-invasive, interference-free, and low-cost method for discharge state monitoring.