Keywords

1 Introduction

The hypothesis of the electrical discharge machining (EDM) was established by Lazarenko [1]. Wire electrical discharge machining (WEDM), a variant form of EDM, turn out to be one of the most useful non-conventional machining process in the modern era of machining [2]. WEDM becomes the best alternative to produce miniaturized scale products with the high accuracy in dimensional precision as well as high degree of surface. By adopting this thermo-electrical approach, almost any conductive material can be machined regardless of their hardness and mechanical properties [3]. Combination of pulse generator, wire electrode and de-ionized medium are the main feature behind a simple WEDM. Electrical spark is generated by pulse generator between workpiece and wire electrode when immersed into de-ionized medium. Such repeated tiny spark causes electro-erosion to erode material from workpiece which also increases the machine zone temperature near about 10,000 °C [4]. Higher and lower flushing arrangement flushes away these eroded materials from the machining zone. These constant sparking, erosion and flushing makes uneven, hard and brittle surface [5].

However, complex mechanism, ample of conflicting process parameters, instability in the manufacturing process makes it awfully difficult to set ideal process parameter setting to get most preferred responses. Here comes the necessity of multi-criteria decision making (MCDM) approach. This operation research technique exceptionally appraises various contradictory criteria to ease in decision making. As a potential tool, MCDM approach widely used to analyze complex real problems to get possible collection of the best choices which will be cost effective for all the manufacturing processes [6]. Though there are substantial MCDM approach are available for selection of ideal setting [7,8,9,10,11,12,13], multi-objective optimization on the basis of ratio analysis (MOORA) is detected to be simple and computationally easy.

Madic et al. [14] showed application potential of MOORA to solve problem related to non-conventional machining methods. Comparison with technique for order preference by similarity to ideal solution (TOPSIS) indicated perfect correlation to solve this kind of problems. Gadakh et al. [15] applied MOORA to solve different multiple objective optimization difficulty in welding which proved MOORA as a potential, flexible technique. Patel et al. [16] successfully applied MOORA method in connection with AHP to select optimum value of WEDM process parameters when machining EN31 alloys steel using brass wire. Achebo et al. [17] successfully applied MOORA coupled with standard deviation (SDV) to find out the ideal welding process parameters.

Inconel 718, a nickel based super alloy, used in many fields like spacecraft, gas turbines, nuclear reactors, rocket motors, etc. where the material can withstand at very high temperature [18]. High toughness, strength and work-hardening characteristic holds a great challenge to cut this difficult to machine material by conventional machining techniques [19]. Thus, the use of WEDM is very much prominent to machine this type of super alloys. Due to the plenty option of process parameters it is very much necessary to select optimum condition in WEDM. In this work, to regulate WEDM process parameters namely pulse-on time (TON), pulse-off time (TOFF), pulsed current (I) and servo voltage (SV) MOORA was used as an application potential of MCDM approach.

2 Materials and Methods

2.1 Materials, Experimental Setup and Data Collection

Following Taguchi’s L9 orthogonal array, 9 experiments was accomplished on 4-axis AGIECUT CNC WEDM (maker: AGIE). Four important input variables namely pulse-on time (TON), pulse-off time (TOFF), pulsed current (I) and servo voltage (SV) were considered to detect significant machinability features like cutting rate (CR), arithmetic mean roughness (Ra) and machining time (MT). De-ionized water and brass wire having diameter 0.25 mm was utilized as dielectric medium wire electrode and respectively. 5 mm length was cut by respective input setting from 5 mm thickness plate of inconel 718. Chemical composition of inconel 718 is shown in Table 1. CR was taken from WEDM machine monitor during respective setting machining and average value from five measurements was booked for final CR value. Using stopwatch MT was taken for every experiment. After machining, Ra values of the machined surface were measured utilizing Taylor Hobson 3D profilometer and average value from five measurements were taken as final Ra value. All the experiment was accomplished in a single loading to diminish loading-unloading time.

Table 1. Chemical composition of inconel 718 [20].
Full size table

2.2 Methodology

In this study, a MCDM model MOORA has been practiced to optimize different WEDM process parameter for inconel 718.

2.2.1 Multi-objective Optimization on the Basis of Ratio Analysis (MOORA)

To simultaneously optimize different conflicting attributes MOORA, which was first introduced by Brauers [21], is used subject to some definite limitations [22]. The associated steps were involved:

1ststep: After defining the objective, decision matrix was formed to represent different performance features with respect to diverse variables.

$$ Q = \left[ {\begin{array}{*{20}c} {q_{11} } & {q_{12} } & { \ldots .} & { \ldots .} & {q_{1n} } \\ {q_{21} } & {q_{22} } & { \ldots .} & { \ldots .} & {q_{2n} } \\ { \ldots .} & { \ldots .} & { \ldots .} & { \ldots .} & { \ldots .} \\ { \ldots .} & { \ldots .} & { \ldots .} & { \ldots .} & { \ldots .} \\ {q_{m1} } & {q_{m2} } & { \ldots .} & { \ldots .} & {q_{mn} } \\ \end{array} } \right] $$
(1)

where, qmn = Performance measure of the mth alternative on nth response.

m = Number of variables.

n = Number of performance features.

2ndstep: Decision matrix was normalized to turn it dimensionless quantity to compare all components. Normalization was done following Eq. 2.

$$ q_{mn}^{*} = \frac{{q_{mn} }}{{\sqrt {\sum\limits_{i = 1}^{r} {q_{mn}^{2} } } }} $$
(2)

where, \( q_{ij}^{*} \) display the normalized value mth alternative on nth response \( (0 < q_{ij}^{*} < 1) \).

3rdstep: In the next step, normalized values were added together for beneficial condition and subtracted for non-beneficial condition to find out overall assessment of the performance measures.

$$ y_{i} = \sum\limits_{n = 1}^{n} {q_{mn}^{*} - \sum\limits_{n = g + 1}^{n} {q_{mn}^{*} } } $$
(3)

Usually few responses are more dominant than others. Corresponding weights were multiplied with the specific objective to give it more preference [23]. Overall assessment value (Yi) calculated as follows:

$$ y_{i} = \sum\limits_{n = 1}^{n} {w_{n} q_{mn}^{*} - \sum\limits_{n = g + 1}^{n} {w_{n} q_{mn}^{*} } } $$
(4)

where,\( w_{n} \) known as the weight of n-th response.

4thstep: In the last step, overall assessment values were arranged in descending order. The highest value of Yi signifies the best optimized setting while lowest value of Yi signifies the least preferred setting.

3 Results and Discussion

The responses measured in the current exploration are cutting rate (CR), arithmetic mean roughness (Ra) and machining time (MT) while varying TON, TOFF, I and SV. Among these responses, Ra and MT are non-beneficial condition whereas CR is beneficial condition. Using Eq. 2, normalized values for each response were calculated. Relative weights for each responses were given as CR = 0.3; Ra = 0.3 and MT = 0.4. Overall assessment value for each experimental setting was calculated following Eq. 4 and tabulated in Table 2. After that, respective overall assessment value prepared in descending order to find out the best setting. It was calculated that, experiment no. 7 has the highest Yi value. So, according to MOORA the optimized process parameter setting like TON = 120 µs., TOFF = 46 µs., I = 230 A. and SV = 20 V.

Table 2. Calculation using MOORA.
Full size table

4 Conclusions

The workpiece inconel 718 was machined using WEDM practice and the outcomes were optimized using MCDM approach MOORA. Following major conclusions might be drawn:

  • MCDM approach MOORA found fairly easier supportive strategy which involves less mathematical calculations.

  • However applying MCDM approach MOORA gave optimized setting as TON = 120 µs., TOFF = 46 µs., I = 230 A. and SV = 20 V which yield the preferred results.

  • The capacity to illuminate process fluctuation makes MOORA more useful for those conditions where numerous responses are need to be improved all along.