Abstract Electrochemical discharge machining (ECDM) is an emerging non-traditional manufacturing process suitable for machining of advanced materials such as hardened metal alloys, ceramics, metal composites, and glass. Its ability to machine both conductive and non-conductive materials, irrespective of their hardness, provides a unique advantage over other machining processes. However, the process remains underutilized in modern industrial setups owing to the unpredictability and stochastic nature of gas film formation and material removal mechanisms in ECDM. Existing studies mostly focus on correlating the input parameters to the post-process machining outputs, however an in-situ measure to quantify gas film stability is still lacking. Stable gas film is required to ensure uniform machining performance and stable discharges. To study the gas film dynamics, ECDM experiments were performed, and high-speed camera images were captured on a custom setup, with ultrasonic vibrations induced at the base of the electrolyte bath. Features associated with gas film dynamics, including thickness variations, gas accumulation within the film, geometric spread, and bubble density, were extracted from image data. Using these features, a novel stability index is proposed in this paper to quantify gas film variability across machining conditions. This index is developed by tracking temporal evolution of image features using a multivariate exponentially weighted moving average (MEWMA). Through image analysis, reduction in gas film thickness was observed under ultrasonic vibrations as the electrolyte concentration increased. Based on Pearson’s correlation analysis, the stability index shows statistically significant correlations with machining metrics like, overcut ( r = − 0.712, p = 0.009), circularity ( r = 0.691, p = 0.013), and maximum heat-affected zone thickness ( r = − 0.832, p = 0.001), supporting its applicability as an indicator of process behavior. The proposed stability index can serve as a process indicator to be used for selecting appropriate input parameters to maintain stability in ECDM.
Gore et al. (Thu,) studied this question.