A DEA approach for optimization of multiple responses in Electrical Discharge Machining of AISI D2 steel

Present research proposes an optimization methodology for the selection of best process parameters in multi-response situation. Experiments have been conducted on a die-sinking electric discharge machine under different conditions of process parameters. A response surface methodology (RSM) is adopted to establish effect of various process parameters such as discharge current (I p), pulse on time (Ton), duty factor (tau) and flushing pressure (Fp) on four important responses like material removal rate (MRR), tool wear rate (TWR), surface roughness (Ra) and circularity (r(1)/r(2)) of machined component. Since the natures of responses are contradicting in nature, it is difficult to find a single combination of machining parameters that provides the best performance satisfying all responses simultaneously. In order to achieve best machining condition, an equivalent single response capable of representing all individual responses is needed. The work includes data envelopment analysis (DEA) to obtain relative efficiency for each experimental run treating as decision making unit (DMIJ). Each DMIJ is evaluated usin LINGO software to obtain relative efficiency. The relative efficiency is ranked in ascending order and average ranked value (ARV) is calculated to find the optimal solution. Finally, the optimal setting capable of improving all the responses simultaneously is found to be Ip=7 amp, Ton= 200 mu s, tau = 90%, and Fp = 0.4 kg/cm(2) With this best combination of factorial level, the experimental values of responses are obtained as MRR=13.9600 mm(3)/min, TWR=0.0201 mm(3)/min, Ra=4.9300 mu m and circularity=0.840 1.