Abstract
Tungsten carbide cobalt composite (WC-Co) is frequently employed for tool and die making due to its superior mechanical properties; however, its high hardness and composite nature makes it a difficult-to machine material with conventional processes. This difficulty is further augmented when intricate and complex profiles are to be machined at high level of precision. Geometrical accuracy of these profiles has paramount importance as WC-Co tools have to generate the replica of their shape on workpiece. Wire electric discharge machining (WEDM) is capable of processing very hard materials, yet it has its own challenges related to geometrical accuracy and surface quality of machined parts which need to be investigated. In the present research work, WEDM has been employed for machining a very small-sized rectangular and tapered cavity in WC-Co. Six WEDM process parameters, namely open voltage (O-V), pulse duration (P-ONT), pulse interval (P-OFFT), servo voltage (S-V), wire tension (W-T), and flushing pressure (D-P), have been taken as input variables. The effect of WEDM parameters on geometrical errors at top and bottom surfaces (GE(TS) and GE(BS)) of machined cavity and surface roughness (R-a) has been assessed by analysis of variance and main effect plots. Experimental results reveal that value of GE(TS) is higher as compared to corresponding value of GE(BS) due to taper cutting. O-V is the most influential parameter for GE(TS), GE(BS), and R-a with a contribution of 48.4%, 20.4%, and 31.5%, respectively. GE(TS), GE(BS), and R-a have shown a decreasing trend with increase in D-P. GE(TS) decreases with increase in P-ONT. Both GE(BS) and Ra tend to decrease with increase in S-V. SEM images indicate that deposition of recast layer inside machined cavity and fractured layer due to cracking are potential sources of higher geometrical error. Single parametric optimization using S/N ratio method has resulted in improvement of 27%, 75%, and 16% for GE(TS), GE(BS), and R-a, respectively. Moreover, application of gray relational analysis (GRA) technique leads to simultaneous optimization of all three responses with a reduction of 36.4%, 31.3%, and 12% in GE(TS), GE(BS), and R-a, respectively.