Stochastic modeling of EDM with a view to off-line optimization

The working principle of electrical discharge machining (EDM) is discussed and its importance to the manufacturing industry is emphasized. Current research thrusts are outlined, and the need for the proposed methodology, Data Dependent Systems (DDS), to deal with the random and complex nature of EDM, is cited. A pilot experiment carried out to collect surface profiles for characterization by DDS models is described and the main experiment designed to verify erosion rate and surface roughness formulas is examined. The derivation of expressions for melt depth and radius of a crater are given in terms of machining parameters and material properties. Attempts to identify the distinguishing features of arcing and sparking and their use in developing a strategy for online computer contol of EDM are explored. Possible adaptive control strategies based on these features are proposed.