This art icle was downloaded by: [ Anna Universit y] On: 03 June 2015, At : 02: 35 Publisher: Taylor & Francis I nform a Lt d Regist ered in England and Wales Regist ered Num ber: 1072954 Regist ered office: Mort im er House, 37- 41 Mort im er St reet , London W1T 3JH, UK Materials and Manufacturing Processes Publicat ion det ails, including inst ruct ions f or aut hors and subscript ion inf ormat ion: ht t p: / / www. t andf online. com/ loi/ lmmp20 Multiresponse Optimization of Abrasive Water Jet Cutting Process Parameters Using TOPSIS Approach a N. Yuvaraj & M. Pradeep Kumar a a Depart ment of Mechanical Engineering, CEG, Anna Universit y, Chennai, Tamil Nadu, India Accept ed aut hor version post ed online: 16 Dec 2014. Published online: 16 Dec 2014. Click for updates To cite this article: N. Yuvaraj & M. Pradeep Kumar (2015) Mult iresponse Opt imizat ion of Abrasive Wat er Jet Cut t ing Process Paramet ers Using TOPSIS Approach, Mat erials and Manuf act uring Processes, 30: 7, 882-889, DOI: 10. 1080/ 10426914. 2014. 994763 To link to this article: ht t p: / / dx. doi. org/ 10. 1080/ 10426914. 2014. 994763 PLEASE SCROLL DOWN FOR ARTI CLE Taylor & Francis m akes every effort t o ensure t he accuracy of all t he inform at ion ( t he “ Cont ent ” ) cont ained in t he publicat ions on our plat form . However, Taylor & Francis, our agent s, and our licensors m ake no represent at ions or warrant ies what soever as t o t he accuracy, com plet eness, or suit abilit y for any purpose of t he Cont ent . Any opinions and views expressed in t his publicat ion are t he opinions and views of t he aut hors, and are not t he views of or endorsed by Taylor & Francis. The accuracy of t he Cont ent should not be relied upon and should be independent ly verified wit h prim ary sources of inform at ion. Taylor and Francis shall not be liable for any losses, act ions, claim s, proceedings, dem ands, cost s, expenses, dam ages, and ot her liabilit ies what soever or howsoever caused arising direct ly or indirect ly in connect ion wit h, in relat ion t o or arising out of t he use of t he Cont ent . This art icle m ay be used for research, t eaching, and privat e st udy purposes. Any subst ant ial or syst em at ic reproduct ion, redist ribut ion, reselling, loan, sub- licensing, syst em at ic supply, or dist ribut ion in any form t o anyone is expressly forbidden. Term s & Condit ions of access and use can be found at ht t p: / / www.t andfonline.com / page/ t erm s- and- condit ions Materials and Manufacturing Processes, 30: 882–889, 2015 Copyright # Taylor & Francis Group, LLC ISSN: 1042-6914 print=1532-2475 online DOI: 10.1080/10426914.2014.994763 Multiresponse Optimization of Abrasive Water Jet Cutting Process Parameters Using TOPSIS Approach N. Yuvaraj and M. Pradeep Kumar Downloaded by [Anna University] at 02:35 03 June 2015 Department of Mechanical Engineering, CEG, Anna University, Chennai, Tamil Nadu, India This paper describes how optimization studies were carried out on an abrasive water jet (AWJ) cutting process with multiresponse characteristics based on Multi Criteria Decision Making Methodology (MCDM) using the Technique for Order Preference by Similarity Ideal Solution (TOPSIS) approach. The process parameters water jet pressure, traverse rate, abrasive flow rate, and standoff distance are optimized with multiresponse characteristics, including the depth of penetration (DOP), cutting rate (CR), surface roughness (Ra), taper cut ratio (TCR), and top kerf width (TKW). The optimized results obtained from this approach indicate that higher DOP and CR and lower Ra, TCR, and TKW were achieved with combinations of the AWJ cutting process parameters, such as water jet pressure of 300 MPa, traverse rate of 120 mm/min, abrasive flow rate of 360 g/min, and standoff distance of 1 mm. The experimental results indicate that the multiresponse characteristics of the AA5083-H32 unit used during the AWJ cutting process can be enhanced through the TOPSIS method. Analysis of variance was carried out to determine the significant factors for the AWJ cutting process. Keywords Aluminium alloy; AWJ cutting; Multiresponse; Optimization; TOPSIS. taper cut ratio (TCR), top kerf width (TKW), and surface roughness (Ra) by the combined effect of the multi-criteria decision-making methodology (MCDM) and analysis of variance (ANOVA). The chemical composition of AA5083-H32 is 94.24% Al, 4.73% Mg, 0.65% Mn, 0.054% Si, 0.060% Cr, 0.055% Ti, 0.005% Cu, and 0.15% Fe; this aluminium alloy was used as the workpiece material and its main properties are shown in Table 1. It is a work-hardened alloy that is stabilized at low-temperature heat to quarter hard in the 5xxx series and offers high strength, hardness, excellent weldability, enhanced corrosion resistance, etc.; it is widely used in the manufacture of various components with marine, automotive, aerospace, and railway applications. [5]. It is enriched with magnesium which enhances its strength through work hardening or strain hardening. Due to its poor machinability, it undergoes partial tempering which slightly reduces its strength but improves ductility [6]. However, machining is difficult by conventional and other nontraditional machining processes because of high tool wear, heat-affected zone, material distortion, high cutting force, and reduction in fatigue life. Because this alloy cannot operate above 65 C and it performs better under low temperatures [7, 8], researchers have stated their preference for the AWJ cutting process to machine this type of aluminium alloy. Few researchers have attempted to investigating abrasive contamination in AA5083-H32 and therefore no specific machinability data are available on these alloys in regard to the AWJ cutting process. This study represents the first attempt to investigate multiresponse optimization of AWJ cutting process parameters in AA5083-H32 alloy using the Technique for Order Preference by similarity INTRODUCTION Abrasive water jet (AWJ) cutting is a successful modern machining technique which is widely applied in the cutting of hard or complex-to-cut materials, such as titanium, stainless steel, and aluminium alloys and it has also been used in the aircraft and automobile Industries. In AWJ cutting, the material can be cut through by erosive action due to abrasive particles mixing with high-pressure water, which acts as a cutting tool, and material removal takes places through the cutting and deformation wear mode [1, 2]. The AWJ cutting process offers a wide range of benefits, such as reduced heat-affected zone, minimal cutting forces, and no contact tool that might break during machining [3]. The cutting parameters of AWJ can be characterized as hydraulic, cutting, mixing, abrasive, etc. [2]. In contrast, use of AWJ makes it difficult to analyze multiresponse parameters, such as depth of penetration, cutting rate, taper cut, kerf width, and roughness [1, 4]. In order to obtain a smooth surface finish, reduction of taper, and higher depth of penetration, the selection of process parameters is of paramount importance for manufacturing industries. These industries require a better cut quality of the product under optimum conditions, which must be met by decision-making methodologies. This study examines product cut quality based on depth of penetration (DOP), cutting rate (CR), Received September 4, 2014; Accepted November 30, 2014 Address correspondence to N. Yuvaraj, Department of Mechanical Engineering, CEG, Anna University, Chennai, Tamil Nadu 600025, India; E-mail: yuvaceg09@gmail.com or yuvabitt09@gmail.com Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/lmmp. 882 MULTIRESPONSE OPTIMIZATION OF ABRASIVE WATER JET CUTTING PROCESS PARAMETERS TABLE 1.—Main properties of aluminium alloy AA5083-H32. Property Downloaded by [Anna University] at 02:35 03 June 2015 Hardness (Brinell scale) Density Ultimate tensile strength Yield strength Modulus of elasticity Shear strength Thermal conductivity Melting point Value 87 HB 2.66 g=cm3 320 MPa 250 MPa 70.3 GPa 195 MPa 117 W=m-k 638 C Ideal Solution (TOPSIS) approach with Simos’ weighting criteria method. In earlier studies on optimization, the Taguchi method was a tool frequently used for improving quality and productivity at low cost. It is used to optimize a single response characteristic. However, it requires further research efforts to investigate multiresponse characteristics [9]. For AWJ cutting, DOP and CR are ‘‘the larger the better’’ performance characteristics, while Ra, TCR, and TKW are ‘‘the smaller the better’’ performance characteristics. Hence, multiresponse optimization is much more difficult than the optimization of a single response characteristic [10]. The selection of optimal process parameters for multiresponse characteristics in the AWJ machining process involves a problematic criterion which necessitates the establishment of adequate MCDM. In MCDM involving engineering problems there are several decision-making methods available, such as grey relational analysis, TOPSIS, Elimination and Et choice Translating reality I, II, III, VIKOR, and Analytic Hierarchy Process. Several researchers used these techniques in the selection of the nontraditional machining process [11], green manufacturing [12], and selection of warehouse location [13]. However, it was found that few researchers had applied these techniques for the selection of manufacturing process and very little is known about the selection of the process parameters. Furthermore, there are no reports in the literature on the optimization of AWJ cutting process parameters with MCDM. Therefore, in this work we aimed to investigate AWJ cutting parameters by varying the level of the water jet pressure (P), traverse rate (TR), abrasive flow rate (AFR), and standoff distance (SOD) and, while machining the AA5083-H32 aluminium alloy, the output parameters DOP, CR, TCR, TKW, and Ra were considered together as a multi-objective problem. From the literature, it is also observed that researchers have developed MCDMs, such as Grey, Electre I, II, TOPSIS, FUZZY TOPSIS, and ELECTRE for solving multi-objective problems. Among these decision-making techniques, TOPSIS was found to be the most successful in the selection of parameters [10] and machining processes [11, 14, 15]. Behzadian et al. [16] conducted review studies on the TOPSIS method associated with multiple fields, such as water resource management, human resource management, and supply chain management. They found only a few research studies on the selection of optimal 883 parameters for machining processes. Lan [17] optimized turning process parameters by combining the TOPSIS and Taguchi methods. Their results indicated that this combination produced a satisfactory performance for computer numerical control (CNC) turning operations. Gauri et al. [18] proposed the selection of optimal machining parameters for an ultrasonic process using the TOPSIS approach. It was observed that for primary component analysis, TOPSIS performed better than grey relational analysis. Durai Prabhakaran et al. [19] used the TOPSIS technique for the selection of a composite product system, observing that this technique met the challenges of composite industry through customization and reliability of these techniques. Rao [20] investigated the machinability of work material by their proposed model of the TOPSIS–AHP method. They observed that the combined effect of TOPSIS and AHP selected the best tool–work combinations for machining operations. This method uses the concept of an ideal solution which is nearest to the choice of attributes [21]. In general, obtaining the ideal solution is a difficult task in real-time problems, so that this method adopts two different kinds of ideal solution, based on the best or worst performance of the alternatives associated with multiple criteria. The ranking-based approach [22] is used to select the best and worst alternatives from positive (as near as possible to the ideal solution) and negative (as distant as possible from the ideal solution) ideal solutions. The advantages of the TOPSIS system include an easily understandable, simple computation technique which has ease of implementation [11]. Maity and Chakraborty [23] studied the selection of abrasive material for grinding wheels using a fuzzy TOPSIS method. They found this method capable of make the best decision based on selection criteria and their interrelationships. Recently, many computational techniques have evolved for optimizing machining process parameters such as Response Surface Methodology [24], Genetic Algorithm [24], Particle Swarm Optimization (PSO) [25–28] and Neural Networks [26]. Some research related to optimization of machining processes parameters using evolutionary algorithms such as PSO and neural networks is briefly presented below. Vázquez et al. [25] used the PSO technique for selection of micromachining parameters in the fabrication of medical devices. They observed that accuracy and surface roughness were improved through optimal process parameters using this evolutionary algorithm. Ciurana et al. [26] used neural network modeling and PSO in pulsed laser micromachining for hardened AISI H13 steel. They investigated the neural network model for prediction of experimental work, and multi-objective process parameters were carried out by PSO. It was observed that the neural network and PSO are suitable for setting optimal process parameters. Ulutan and Özel [27] predicted cutting forces and residual stress values on nickel alloy during turning operations. They conducted physics-based simulations along with PSO for predicting values which help minimize tensile residual stress and improve compressive residual stress on the turned surface through the selection of optimal process parameters. 884 N. YUVARAJ AND M. PRADEEP KUMAR Downloaded by [Anna University] at 02:35 03 June 2015 FIGURE 1.—Decision model of the AWJ cutting process parameter selection. Pawar et al. [28] carried out PSO for optimization of grinding parameters for maximization of production rate and surface finish. However, it was observed that these optimization techniques require a coding system (algorithm) to resolve MCDM problems, and were time consuming. Among these methods, TOPSIS was found to be more efficient in solving MCDM problems because it offers a simple computational technique, less computational time, and values are close to the ideal solution. So, researchers have a keen interest in using the TOPSIS method to optimize the selection of process parameters and determine the optimal solution. The TOPSIS method requires precise input data in solving multi-criteria problems to assign weights to the criteria, which emphasizes the relative importance of multi-criteria (output responses) in real-time problems [29, 30]. To use the weighting criterion, Simos’ procedure was carried out. In this study, each response was designated by a gaming card with output responses arranged from the least to the most important. According to the importance of the output responses, the decision maker classifies the parameter responses from the least to the most important (e.g., DOP, CR, TCR, TKW, Ra). The details and steps of Simos’ procedure are given in Ozcan et al. [13] with reference to Figueira and Roy [31]. This procedure is a key tool for analyzing real-time problems and is adopted by decision makers for various reasons, such as robustness, and it has to produce results more rapidly than the other weighting computational techniques. Ozcan et al. also carried out Simos’ procedure through data collection, computation of normalized weights, and minimization of errors through rounding off the normalized values. Figure 1 represents the decision model of the AWJ cutting process parameter selection problem. In this figure, each output response is related to four different process parameters of the AWJ cutting, and these responses are taken to optimize the cutting parameters using the TOPSIS method. MATERIALS AND METHODS The experimental setup used for this study is shown in Fig. 2. The experiments were conducted using a water jet injection-type AWJ machining center (S 3015, Germany; FIGURE 2.—Experimental setup. maximum pressure: 620 MPa, traverse speed: 10 m=min, cutting head movement in x- and y-axes: 3 m  1.5 m, water discharge: 3.2 l=min. L27 orthogonal array was used to conduct the experimental run, and each process parameter was allocated to a separate column. All experiments were conducted on AA5083-H32 (120 mm (length)  50 mm (width)  64.065 mm (thickness)), with the jet traversing a predetermined length of the workpiece. For each set of process parameters the jet produced a different penetration depth into the target material. Penetration depth was determined through measuring the length of cut (L) on the workpiece surface, as shown in Fig. 3. The list of AWJ process parameters and their levels are shown in Table 2. The experimental design of the AWJ cutting process parameters is given in Table 3. The other AWJ cutting process parameters were maintained at a constant level (impact angle: 90 , garnet mesh size: 80, tungsten carbide nozzle diameter: 1.1 mm, sapphire orifice diameter: 0.35 mm, and jet FIGURE 3.—Lengths of cut (L) on the workpiece surface (front view). 885 MULTIRESPONSE OPTIMIZATION OF ABRASIVE WATER JET CUTTING PROCESS PARAMETERS and the computational steps of this procedure are shown in Table 4. By using this procedure, the weighting of the criteria is calculated as follows. TABLE 2.—Input process parameters and their levels. Level Symbol A B C D Factor Level 1 Level 2 Level 3 Water jet pressure (P) (MPa) Traverse rate (TR) (mm=min) Abrasive mass flow rate (AFR) (g=min) Standoff distance (SOD) (mm) 250 120 240 1 275 150 300 2 300 180 360 3 traverse length: 60 mm. The output responses considered in this study were DOP, CR, TCR, Ra, and TKW. These response parameters were used to evaluate the performance of the AWJ cutting process by varying the levels of input process parameters. DOP (mm) was measured by a dial gauge of accuracy 0.05 mm. CR (mm2=s) was measured using the following equation [2]: Downloaded by [Anna University] at 02:35 03 June 2015 CR ¼ DOP  TR ð1Þ TKW (mm) was measured by micro-hardness testing microscope of accuracy 0.01 mm. TCR was determined by the ratio of top cut width (bT) to bottom cut width (bB). A Surftronic-3þ Hobson was used to measure workpiece roughness (mm), with a traverse length of 4 mm and cut-off length of 0.8 mm. In this paper, the key subjective input weights of the output responses were determined by Simos’ procedure . Definite set of criteria ¼ [DOP, TR, TCR, TKW, and Ra] . Criteria (output responses) were arranged by decision makers according to their importance, from least to most important (e.g., TKW, TCR, Ra, CR, and DOP). In this study, TCR and Ra are considered as equal weights, as are DOP and CR. Nevertheless, because the decision makers needed to develop the importance of the DOP and CR criteria they put white cards between two successive criteria in the criteria set. In this procedure, white cards play an important role by assigning the weight of criteria from least to most important, the cards representing the relative importance of two successive criteria. This step gives more weightage or importance to the criteria selected. The optimization steps of TOPSIS were calculated using the following procedure [29]. Step 1. The TOPSIS method is a best-ranking method which selects alternatives that eliminate the units of all criteria, and it takes a normalized value. Table 3 also shows the normalized performance matrix (rij) obtained using the following equation: TABLE 3.—Design of experiment and normalized values of output response parameters. Normalized matrix value (rij) Ex. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 A B C D DOP CR TKW TCR Ra 250 250 250 250 250 250 250 250 250 275 275 275 275 275 275 275 275 275 300 300 300 300 300 300 300 300 300 120 120 120 150 150 150 180 180 180 120 120 120 150 150 150 180 180 180 120 120 120 150 150 150 180 180 180 240 300 360 240 300 360 240 300 360 240 300 360 240 300 360 240 300 360 240 300 360 240 300 360 240 300 360 1 2 3 2 3 1 3 1 2 1 2 3 2 3 1 3 1 2 1 2 3 2 3 1 3 1 2 0.182 0.188 0.214 0.136 0.162 0.162 0.115 0.136 0.152 0.193 0.214 0.225 0.157 0.183 0.214 0.157 0.173 0.188 0.240 0.251 0.266 0.204 0.193 0.250 0.178 0.182 0.188 0.150 0.154 0.175 0.139 0.166 0.166 0.142 0.167 0.187 0.159 0.176 0.184 0.161 0.188 0.219 0.194 0.213 0.231 0.197 0.206 0.218 0.209 0.198 0.257 0.219 0.224 0.231 0.184 0.187 0.218 0.190 0.194 0.139 0.184 0.124 0.159 0.193 0.208 0.227 0.194 0.210 0.157 0.187 0.155 0.213 0.200 0.204 0.234 0.215 0.208 0.179 0.202 0.181 0.204 0.152 0.195 0.154 0.165 0.143 0.181 0.231 0.145 0.383 0.144 0.189 0.148 0.155 0.136 0.185 0.171 0.139 0.363 0.137 0.180 0.114 0.139 0.120 0.159 0.202 0.135 0.313 0.200 0.197 0.179 0.214 0.205 0.220 0.252 0.193 0.189 0.191 0.175 0.157 0.205 0.177 0.187 0.236 0.216 0.182 0.174 0.169 0.146 0.189 0.166 0.174 0.209 0.188 0.168 886 N. YUVARAJ AND M. PRADEEP KUMAR TABLE 4.—Computational steps of Simos’ weighting procedure. Subset criterion TKW TCR, Ra White cards DOP, CR Total Number of criteria Number of position Non-normalized weighted matrix Total 1 2 1 2 6 1 2,3 (4) 5,6 17 1=17 100 ¼ 5.88  6 2.5=17 100 ¼ 14.70  15 6 30 5.5=17 100 ¼ 32.35  32 64 100 Xij ffi rij ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Pm 2 X i¼1 ij Downloaded by [Anna University] at 02:35 03 June 2015 TABLE 5.—Closeness coefficient values and ranking of alternatives. Ex. no. i ¼ 1; 2; . . . . . . 27; j ¼ 1; 2; . . . . . . 5 ð2Þ where i ¼ number of alternatives (experimental runs) j ¼ number of criteria (output responses) xij ¼ normalized value of ith experimental run associated with jth output response. Step 2. The weighted normalized matrix (Wij) is obtained by the product of the normalized value and weighted values: Wij ¼ wj  rij i ¼ 1; 2; . . . . . . 27; j ¼ 1; 2; . . . . . . 5 ð3Þ 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Dþ i D i Ci Ranking 0.04456 0.04376 0.03220 0.05751 0.04536 0.04661 0.06504 0.05113 0.05904 0.03992 0.03337 0.02771 0.04745 0.03497 0.02393 0.04321 0.03471 0.04624 0.02194 0.02041 0.01378 0.02638 0.03025 0.00990 0.03474 0.02957 0.04022 0.041401 0.037499 0.04899 0.033531 0.040201 0.035201 0.022553 0.038606 0.02202 0.044816 0.045699 0.053411 0.037817 0.046851 0.051175 0.038535 0.04745 0.039044 0.058293 0.058259 0.069447 0.052116 0.051744 0.067331 0.042704 0.051591 0.040844 0.48160 0.46149 0.60342 0.36829 0.46987 0.43027 0.25746 0.43023 0.27166 0.52888 0.57800 0.65838 0.44351 0.57260 0.68135 0.47139 0.57750 0.45782 0.72653 0.74051 0.83444 0.66394 0.63108 0.87180 0.55141 0.63564 0.50387 17 20 10 25 19 23 2 24 26 15 11 7 22 13 5 18 12 21 4 3 2 6 9 1 14 8 16 rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2ffi X27   Wij  Sj ¼ i¼1 Step 3. Every response that is an ideal alternative to the best (Sþ) and worst alternative performance (S) was identified. If the jth criteria have the optimal performance: D i Sþ ¼ Step 5. For each alternative, closeness coefficient (Ci) values were determined using the following equation:    maxðSij Þj j 2 J  or minðSij Þj j 2 J 0 i ¼ 1; 2;......27 ð4Þ where Sþ represents a positive ideal solution [0.085, 0.082, 0.00742, 0.01703, 0.022] Similarly the S values were determined if the jth criteria have the worst performance [0.03687, 0.04456, 0.0140, 0.0570, 0.03782] where S represents a negative ideal solution. Step 4. In this step, the performance of the criteria was measured as the best alternative distance (Dþ ij ) from the Sþ values and the worst alternative distance (D ij )  from the S values. The values of Dþ and D were ij ij determined using the equations below (5, 6). Table 5 shows the performance of each alternative under the best and worst conditions. Dþ i rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2ffi X27  þ ¼ Wij  Sj i¼1 ð5Þ ð6Þ where i ¼ 1,2,3. . .,27 Ci ¼ D i þ D i þ Di i ¼ 1; 2; . . . . . . : 27; 0  Ci  1 ð7Þ The best alternative was chosen according to preference ranking ordered by the Ci value, which is very close to the ideal solution. RESULTS AND DISCUSSION Table 5 also indicates the closeness coefficient values of each experimental run using L27 orthogonal array. Among the values of the 27 closeness coefficients, experiment no. 24 showed the best multiresponse characteristics because it represents the maximum closeness coefficient; the values recorded in each experiment are shown in Fig. 4. A higher closeness coefficient means that the corresponding experiment is closer to the ideal value; experiment no. 24 yielded optimal process parameters for the multi-criteria output responses of the AWJ cutting process. 887 Downloaded by [Anna University] at 02:35 03 June 2015 MULTIRESPONSE OPTIMIZATION OF ABRASIVE WATER JET CUTTING PROCESS PARAMETERS FIGURE 4.—Effect of process parameters in each experiment. FIGURE 5.—Mean response values of AWJ process parameters. Table 6 shows the closeness coefficient values of the process parameters at the factor level. The values were determined by the closeness coefficient value of each level of the cutting parameter in the L27 orthogonal array, and their mean. These represent the optimal level of process parameter by factor level. From the mean responses shown in Fig. 5, it is observed that the optimal parameter setting is maintained at P 300 MPa, TR 120 mm=min, AFR 360 mm=min, and SOD 1 mm, thus distinguishing between maximum and minimum values of closeness coefficient. P is the most effective process parameter and its maximum–minimum value is higher than that of the other input parameters. Higher P values result in maximum DOP and CR and minimum TKW, TCR, and Ra. ANOVA was conducted to determine the significant level of each input process parameter affecting the multi-performance characteristics of AWJ. Results show that the F-ratio indicates the significant process parameters of the AWJ cutting process, determining whether the statistical test is significant at the confidence level selected. Significance level is based on the critical value: if it exceeds the critical value the test is significant, otherwise it is not [32]. In the present study, ANOVA was carried out at a confidence level of 95% and significance level of 5%. Table 7 shows the results of ANOVA for the effects of input and interaction parameters on multiresponse parameters. It was found that P and TR are the most significant and significant factors, respectively, in the assessment of DOP, CR, TKW, TCR, and Ra, because higher pressure resulted in increased DOP and CR and decreased TCR, TKW and Ra, regardless of the material used. Higher pressure results in the production of high kinetic energy by the abrasive particles impacting on the workpiece surface, thereby enhancing cutting of the material with better surface quality, high jet penetration, and minimization of kerf width and taper cut. At lower TR there is a possibility of increasing the number of abrasive particles, which impacts on the workpiece per unit time and hence an improved surface finish, penetration depth, CR, kerf width, and taper cut are obtained [2, 33]. The other input process parameters and interaction parameters were found to be less significant due to failure of the statistical test. Table 7 also indicates the influence of input process parameters on AWJ cutting. The most influential parameter was P (54.03%), followed by TR (20.93%), SOD (7.46%), AFR (6.86%), P  TR (2.68%), P  AFR (1.40%), and P  SOD (0.66%). The influence of each process parameter on multiresponse characteristics is shown in Fig. 6. From Table 7 it is observed that P plays a crucial role in affecting multiresponse characteristics while SOD and AFR contribute to the cutting quality of the machined surface along with P and TR. Lower SOD produced a narrow abrasive water jet diameter resulting in reduced kerf width [33] while increased AFR, along with higher P and lower TR, resulted in enhanced quality of the machined surface due to the additional particles created which serve to smoothen the workpiece surface [2]. The comparative test results for initial and optimal selection of AWJ process parameters (predicted and TABLE 7.—Results of ANOVA for multiresponse parameters. Source of process parameters TABLE 6.—Mean response for closeness coefficient. Average closeness coefficient Symbol A B C D Process parameter Water jet pressure Traverse rate Abrasive mass flow rate Standoff distance Level 1 Level 2 Level 3 Maximum.– Minimum 0.4194 0.5522 0.6844 0.6237 0.5703 0.4619 0.4992 0.5663 0.5903 0.2650 0.1618 0.0911 0.5960 0.4988 0.5611 0.0972 P TR AFR SOD P  TR P  AFR P  SOD Error Total   DOF SS MS F P value % contribution 2 2 2 2 4 4 4 6 26 0.3160 0.1224 0.0401 0.0436 0.0157 0.0082 0.0038 0.0349 0.5849 0.1580 0.0612 0.0201 0.0218 0.0039 0.0020 0.0009 0.0058 27.15 10.51 3.45 3.75 0.68 0.35 0.17 0.001 0.011 0.101 0.088 0.633 0.834 0.948 54.03 20.93 6.86 7.46 2.68 1.40 0.66 5.98 100 Most significant parameter. Significant parameter. 888 N. YUVARAJ AND M. PRADEEP KUMAR FIGURE 6.—Process parameter influence, by percentage. TABLE 8.—Initial and predicted test results. the maximum closeness coefficient value for multiresponse optimization of AWJ cutting parameters: water jet pressure (P): 300 MPa, traverse rate (TR): 150 mm=min, abrasive flow rate (AFR): 360 mm=min, and standoff distance (SOD): 1 mm. The optimum levels of input process parameter combinations were identified according to the response of closeness coefficient values: 300 MPa (P), 120 mm=min (TR), 360 mm=min (AFR), and 1 mm (SOD). ANOVA was conducted to investigate the significant parameters for the multi-performance characteristics of AWJ. From the above analysis, P was found to be the most significant factor, followed by TR, SOD, and AFR. It was observed that the proposed combination of TOPSIS and ANOVA was more effective in solving AWJ multiresponse problems than previously used methods. Output response Downloaded by [Anna University] at 02:35 03 June 2015 Condition Initial setting parameter Predicted Experimental Setting level DOP (mm) Ci CR TKW Ra (mm2=s) (mm) TCR (mm) value A3B3C2D1 35.017 105.05 1.172 1.34 3.52 0.6356 Optimal process parameters A3B1C3D1 A3B1C3D1 48.17 96.34 1.23 0.8387 1.15 2.96 0.8617 Improvement in TOPSIS grade: 0.2261. experimental conditions, respectively) are shown in Table 8. Once the optimal level of machining parameters was determined, confirmation tests were carried out to validate enhancement of the multiresponse characteristic of AWJ. Using the optimal level of AWJ cutting parameters, predicted response value (cpredicted) can be estimated from the following equation [9]: Xn ðc  cm Þ ð8Þ Upredicted ¼ Um þ j¼1 o where Um is the overall mean multiresponse value, Uo is the mean multiresponse value at the optimum level of factors, and n is the number of input process parameters. These test results indicate that the overall Ci value of the optimal parameter condition (A3B1C3D1) is higher than that of the initial setting parameter condition (A3B3C2D1), and also that the predicted response value is close to the experimental value. CONCLUSION In the present work, use of the TOPSIS method with orthogonal array was carried out to optimize process parameters in the abrasive water jet (AWJ) cutting process of AA5083-H32 alloy for multiresponse characteristics. Simos’ weighting criteria method was used to rank the output response parameters. An optimal combination of AWJ cutting parameters and their levels was identified in regard to achieving better depth of penetration (DOP), cutting rate (CR), taper cut ratio (TCR), and top kerf width (TKW) and lower surface roughness (Ra). Experiment no. 24 yielded ACKNOWLEDGMENTS The authors acknowledge the experimental facilities provided by the Director, Water Jet Germany Ltd, Chennai. FUNDING The authors are grateful to the Council of Scientific and Industrial Research (CSIR) under the Senior Research Fellowship (SRF), Government of India, New Delhi, for their financial support to this research work. REFERENCES 1. Hashish, M. A Model for abrasive water jet machining. Transactions of ASME Journal of Engineering Materials and Technology 1989, III, 154–162. 2. Momber, A.; Kovacevic, R. Principles of Abrasive Water Jet Machining; Springer-Verlag: London, UK, 1998. 3. Folkes, J. Waterjet–an innovative tool for manufacturing. Journal of Materials Processing Technology 2009, 209 (20), 6181–6189. 4. Hashish, M. Optimization factors in abrasive-waterjet machining. Transactions of ASME Journal of Engineering for Industry 1991, 113, 29–37. 5. Kechagias, J.D.; Ziogas, C.K.; Pappas, M.K.; Ntziatzias, I.E. Parameter optimization during finish end milling of Al alloy 5083 using Robust design. In Proceedings of the World Congress on Engineering, London, UK, July 6–8, 2011. 6. Totten, G.E.; Scott Mackenzie, D. Handbook of Aluminium Physical Metallurgy and Processes; Marcel Dekker Incorporation: New York, NY, 2003. 7. Selvakumar, G.; Sarkar, S.; Mitra, S. An experimental analysis of single pass cutting of aluminium 5083 alloy in different corner angles through WEDM. International Journal of Machining and Machinability of Materials 2013, 13 (2–3), 262–275. 8. Stournaras, A.; Stavropoulos, P.; Salonitis, K.; Chryssolouris, G. An investigation of quality in CO2 laser cutting of aluminum. CIRP Journal of Manufacturing Science and Technology 2009, 2 (1), 61–69. 9. Lin, J.L.; Lin, C.L. The use of the orthogonal array with grey relational analysis to optimize the electrical discharge MULTIRESPONSE OPTIMIZATION OF ABRASIVE WATER JET CUTTING PROCESS PARAMETERS 10. 11. 12. 13. Downloaded by [Anna University] at 02:35 03 June 2015 14. 15. 16. 17. 18. 19. 20. 21. machining process with multiple performance characteristics. International Journal of Machine Tools and Manufacture 2002, 42, 237–244. Singh, A.; Datta, S.; Mahapatra, S.S. Application of TOPSIS in the Taguchi method for optimal machining parameter selection. Journal of Manufacturing Science and Production 2011, 11, 49–60. Temuçin, T.; Tozan, H.; Vayvay, O.; Harnicarova, M.; Valicek, J. A fuzzy based decision model for nontraditional machining process selection. The International Journal of Advanced Manufacturing Technology 2014, 70, 2275–2282. Sivapirakasam, S.P.; Mathew, J.; Surianarayanan, M. Multiattribute decision making for green electrical discharge machining. Expert Systems with Applications 2011, 38, 8370–8374. Ozcan, T.; Celebi, N.; Esnaf, S. Comparative analysis of multi-criteria decision making methodologies and implementation of a warehouse location selection problem. Expert Systems with Applications 2011, 38, 9773–9779. Chau, O.L.; Parkan, C. Selection of a manufacturing process with multiple attributes—A case study. Journal of Engineering Technology Management 1995, 12, 219–237. Chakladar, N.D.; Chakraborty, S. A combined TOPSISAHP-method-based approach for non-traditional machining processes selection. Journal of Engineering Manufacture 2008, 222, 1613–1623. Behzadian, M.; Otaghsara, S.K.; Yazdani, M.; Ignatius, J. A state-of the-art survey of TOPSIS applications. Expert Systems with Applications 2012, 39, 13051–13069. Lan, T.S. Taguchi optimization of multiobjective CNC machining using TOPSIS. Information Technology Journal 2009, 8 (6), 917–922. Gauri, S.K.; Chakravorty, R.; Chakraborty, S. Optimization of correlated multiple responses of ultrasonic machining (USM) process. The International Journal of Advanced Manufacturing Technology 2011, 53 (9–12), 1115–1127. Durai Prabhakaran, R.T.; Babu, B.J.C.; Agrawal, V.P. Optimum selection of a composite product system using MADM approach. Materials and Manufacturing Processes 2006, 21, 883–891. Rao, V.R. Machinability evaluation of work materials using a combined multiple attribute decision-making method. The International Journal of Advanced Manufacturing Technology 2006, 28, 221–227. Chen, S.J.; Hwang, C.L. Fuzzy Multiple Attribute Decision Making Methods and Applications; Springer-Verlag: New York, NY, 1992. 889 22. Hwang, C.L.; Yoon, K. Multiple Attribute Decision Making Methods and Applications; Springer-Verlag: Berlin, Germany, 1981. 23. Maity, S.R.; Chakraborty, S. Grinding wheel abrasive material selection using fuzzy TOPSIS method. Materials and Manufacturing Processes 2013, 28, 408–417. 24. Kilickap, E.; Huseyinoglu, M. Selection of optimum drilling parameters on Burr height using response surface methodology and genetic algorithm in drilling of AISI 304 stainless steel. Materials and Manufacturing Processes 2010, 25, 1068–1076. 25. Vázquez, E.; Ciurana, J.; Rodrı́guez, C.A.; Thepsonthi, T.; Özel, T. Swarm intelligent selection and optimization of machining system parameters for micro-channel fabrication in medical devices. Materials and Manufacturing Processes 2011, 26, 403–414. 26. Ciurana, J.; Arias, G.; Özel, T. Neural network modeling and Particle Swarm Optimization (PSO) of process parameters in pulsed laser micromachining of hardened AISI H13 steel. Materials and Manufacturing Processes 2009, 24, 358–368. 27. Ulutan, D.; Özel, T. Multiobjective optimization of experimental and simulated residual stresses in turning of nickel-alloy IN100. Materials and Manufacturing Processes 2013, 28, 835–841. 28. Pawar, P.J.; Rao, R.V.; Davim, J.P. Multiobjective optimization of grinding process parameters using particle Swarm optimization algorithm. Materials and Manufacturing Processes 2010, 25, 424–431. 29. Olson, D.L. Comparison of weights in TOPSIS models. Mathematical and Computer Modelling 2004, 40, 721–727. 30. Opricovic, S.; Tzeng, G.H. Compromise solution by MCDM methods-A comparative analysis of VIKOR and TOPSIS. European Journal of Operational Research 2004, 156, 445–455. 31. Figueira, J.; Roy, B. Determining the weights of criteria in the ELECTRE type methods with a revised Simos’ procedure. European Journal of Operational Research 2002, 139, 317–326. 32. Azmir, M.A.; Ahsan, A.K.; Rahmah, A. Investigation on abrasive waterjet machining of Kevlar reinforced phenolic composite using Taguchi approach. In Proceedings of the International Conference on Mechanical Engineering, Dhaka, Bangladesh, December 29–31, 2007. 33. Karakurt, I.; Aydiner, G.A.K. An experimental study on the depth of cut of granite in abrasive waterjet cutting. Materials and Manufacturing Processes 2012, 27, 538–544.