Application of analytic hierarchy process (AHP) to find proper combination of process parameters of GMAW for getting good quality of welding joint for a particular set-up IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 252 Vol. 5 Issue 2 2013 ISSN 1936-6744 SELECTION OF APPROPRIATE PROCESS PARAMETERS FOR GAS METAL ARC WELDING OF MEDIUM CARBON STEEL SPECIMENS Kazi Sabiruddin Mechanical Engineering Department Kalyani Government Engineering College Kalyani- 741 235, West Bengal, India E-mail: kazi4sabir123@gmail.com Subhajit Bhattacharya Mechanical Engineering Department Kalyani Government Engineering College Kalyani- 741 235, West Bengal, India E-mail: sbhattacharya63@gmail.com Santanu Das Mechanical Engineering Department Kalyani Government Engineering College Kalyani- 741 235, West Bengal, India E-mail: sdas.me@gmail.com ABSTRACT Gas Metal Arc Welding (GMAW) is a semi-automated process used widely for accurate welding in the fabrication industry. The selection of appropriate process parameters of GMAW is essential to obtain the desired weld quality. In the past, much work has been done investigating a variety of workpiece and electrode material combinations. In the present work, an Analytical Hierarchy Process (AHP) based parametric optimization is tried in gas metal arc welding of C45 medium carbon steel specimens using carbon dioxide as the gas shield. The experiments were performed by varying three process parameters, weld speed, weld voltage and weld current. The AHP facilitates the selection of suitable process parameters to obtain a sound weld. In the present experimental domain, optimal conditions are evaluated to be at a weld voltage of 30 V, weld current of 160 A with a weld speed of 475.75 mm/min. Keywords: AHP, welding, GMAW, MAG, parametric optimization 1. Introduction The Analytic Hierarchy Process (AHP) is a simple, widely used decision making tool that can effectively solve a variety of complex multi-criteria problems hierarchically (Saaty, 1977; Saaty, 1980; Vargas et al., 1990). The AHP was and is being employed to solve several managerial, manufacturing and production related decision making problems. This has also been utilized for the optimum selection of process parameters in different mailto:sdas.me@gmail.com Rob Typewritten Text http://dx.doi.org/10.13033/ijahp.v5i2.184 IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 253 Vol. 5 Issue 2 2013 ISSN 1936-6744 welding processes. In a recent work, Saaty (2009) discussed a method of taking a judgment in decision-making process in contrast to that used in usual science experiments. Gas metal arc welding (GMAW) utilizes an arc maintained between the workpiece and an automatically fed wire electrode. Argon, helium or a mixture of the two is usually used for welding different metals, and primarily nonferrous metals. When welding steel, some oxygen or carbon dioxide is usually added to improve the arc stability and reduce weld spatter. Less costly CO2 can be used alone when welding steel, provided that a deoxidizing electrode wire is employed in this process of GMAW known as MAG (Metal Active Gas) welding (Khanna, 1995; Nadkarni, 1996). The GMAW process can be easily mechanized to guarantee high productivity while maintaining good quality. However, to achieve good results, process variables of GMAW need be selected appropriately. Nadkarni (1996) reported the relationship of mechanical properties of a welded joint with the degree of compositions of base material, and the effect of main process parameters of welding on the quality of the weld. Detailed investigation of the effect of the chemistry of base material on the softening of HAZ was made by Mohandas et al. (1999). Hardness and microstructure were compared with the variation of the chemistry of the parent metal and the welding process to gain an understanding of the influence of the alloy chemistry, and the effect of different welding processes on the same low alloy steel. In another work, Zumelzu et al. (1999) observed the effect of post-weld heat-treatment and external cooling on the GMAW product. They investigated the quality of the joining of 316L stainless steel specimens under varying conditions through the analysis of microstructure observation. Kim and Basu (1998) employed mathematical models of the GMAW process to select welding process parameters for obtaining the required weld-bead geometry. All these works reported some success in the respective electrode-workpiece material combinations under variations of the welding processes. In other works, Modensi et al. (1999) evaluated the influence of small differences in wire characteristics on operational conditions of CO2 gas shielded GMAW. Data were evaluated using factorial analysis and graphical techniques to assess the effect of different wire characteristics on the weldment. The results showed that differences in wire diameter produced varying quality of a welded joint. An abductive polynomial network model of the GMAW process was established by Simpson and Hughes (2006). This network model enabled establishing the relationship between GMAW process parameters such as wire diameter, gas flow rate, welding speed, arc current and welding voltage on the weld bead penetration. The estimated value of weld bead penetration derived from network training was compared with the measured value. Jones et al. (1992) observed a relationship between power input to the arc in GMAW, metal transfer process and base plate heating. Optimized parameters evolved within the respective experimental domain in other investigations by Sabiruddin and Das (2005) and Jaubari et al. (2007) involving MAG welding of different steels under varying conditions. Jaubari et al. (2007) recommended a gas mixture of argon, CO2 and oxygen for the GMAW process to obtain substantial cost savings with a good control of spatter. In order to discover appropriate process parameters for desired weld quality, a number of works were also done employing the Analytical Hierarchy Process (AHP) for optimal IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 254 Vol. 5 Issue 2 2013 ISSN 1936-6744 selection of process variables. Ravisankar et al. (2006) and Sabiruddin et al. (2009) tried to obtain quality butt joints of aluminium alloys and steels respectively through the selection of a suitable welding process and corresponding process parameters applying the AHP. The selection of appropriate process parameters was also successfully carried out by Lai et al. (2009) by applying the AHP for resistance spot welding, and choosing typical edge preparation for obtaining sound welding was tried by Liu et al. (2011) using the AHP that could have long fatigue life. In all these works involving the AHP, appropriate process conditions could be achieved to apply in practice. Because the optimal process parameters are vital to the quality of the weldment, in this work, a number of experiments have been conducted to determine these parameters. Using the AHP, different parameters of CO2 gas shielded GMAW process were varied in order to find out an appropriate combination of process parameters. 2. Details of experiments In this experimental work, gas metal arc welding of medium carbon steel flats is carried out on an ESAB India Ltd. made GMAW set up with an AUTOK 400 model. An indigenously developed system is used to move the welding torch along a straight path along the gap between the two steel flats to weld with a set speed to have weld deposition under a carbon dioxide gas shield. Although there are many factors that influence a weld, we have chosen three main factors that determine heat input to the weld to investigate in this work. These factors are welding current, welding voltage and welding speed. Heat input (Q) is quite important in welding, and during the GMAW process, heat input is calculated by: Q = 0.8 V I / S when V is weld voltage, I is weld current, and S is weld speed. Based on the trial tests, a welding current of 140 A, 150 A and 160 A, a welding voltage of 25 V and 30 V, and a welding speed of 370.5 mm/min and 475.75 mm/min are chosen for the present experimental work on joining C45 medium carbon steel specimens as detailed in Table 1. Twelve experiments are carried out, and the parameters corresponding to each experiment are shown in Table 1. Without any preheating, specimens (size: 120 mm x 50 mm x 5 mm) are joined by a double-butt joint (in which both sides of the joint are welded) with a root gap of 1.5 mm. The weldment is brought to room temperature by air cooling. The joint is made in a horizontal position with the torch angle of 75° with the horizontal, using a low carbon steel wire electrode of 1.2 mm diameter. The weldments are visually inspected and tested through dye penetration. The presence of a visible crack, a blow hole, and the extent of spatter and uniformity of weld metal deposition is discovered through visual inspection. At some experimental conditions, bubbles of molten metal are scattered around the weld resulting in less penetration and reducing the aesthetic look of the weldment. Penetration of weld metal is the depth of penetration of the weld metal going into the gap between two specimens being joined, and is observed through polishing a cut section of the weldment along its cross section. A IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 255 Vol. 5 Issue 2 2013 ISSN 1936-6744 bend test is done on a universal testing machine (Fine Spavy Associates & Engineers Pvt. Ltd., Miraj, India, model- TUN 200: 97/333) that observes the bending strength of the weldment. In this test, the butt welded specimen is placed on two supports, and a downward load is placed onto it at its middle and around the weld region. The bend test is continued up to a bend angle of 45 0 , or when any crack is formed in the weldment, and the corresponding bending load is noted. These observed results are utilized to design the AHP model that will be used to discover the appropriate process parameters. Table 1 Experimental conditions (alternatives) for welding medium carbon steel flats 3. Discussion of experimental results The experimental results that were obtained are given in Table 2. Observation of weld quality, such as spatter, blow holes, penetration at the joint, uniformity of weld, and presence of surface cracks are shown in tabular form against each experiment. Bending load obtained through the bend test, is also included in Table 2. At a low travel speed of 370.5 mm/min with an weld voltage of 25 V and weld current of 140 A (experiment 1), large spatter and blow holes are found with less penetration; transverse and longitudinal surface cracks are also observed indicating quite poor weld quality. The bending load for this case is moderate. When weld current is set at 150 A in experiment 2, less spatter and thin welds are noticed. Although good penetration is achieved, a few blow holes and apparent toe cracks are found. However, the weld joint appears to be good as it sustains a high bending load of 15.8 kN up to a bend angle of 45 o without fracturing. At a weld current of 160 A (experiment 3), spatter and blow holes are present less, and good penetration is observed. The presence of transverse and longitudinal cracks limits the bending load when the weld gets fractured. When weld voltage is increased to 30 V in experiments 4-6, increase in weld current also causes an increase in heat input from 0.54 kJ/mm to 0.62 kJ/mm. This results in deep penetration of the weld metal inside the joint. However, this high heat input to the weld Sl. No. (Alternatives) Weld speed (mm/min) Weld voltage (V) Weld current (A) Heat input (kJ/mm) A1 370.5 25 140 0.45 A2 370.5 25 150 0.49 A3 370.5 25 160 0.52 A4 370.5 30 140 0.54 A5 370.5 30 150 0.58 A6 370.5 30 160 0.62 A7 475.75 25 140 0.35 A8 475.75 25 150 0.38 A9 475.75 25 160 0.40 A10 475.75 30 140 0.42 A11 475.75 30 150 0.45 A12 475.75 30 160 0.48 IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 256 Vol. 5 Issue 2 2013 ISSN 1936-6744 caused non-uniform contraction upon cooling leading to different types of cracks detected in the weld (see Table 2). This gives low bending load of the weld. Table 2 Experimental observation of the weldment In experiments 7-9 with a welding speed of 475.75 mm/min, heat input is less (0.35 kJ/mm to 0.4 kJ/mm) leading to less penetration and a poor weld joint. This resulted in a considerably less bending load and indicated the presence of different types of cracks. On the other hand, at the weld speed of 475.75 mm/min, an acceptable quality of weld was observed corresponding to a weld voltage of 30 V at all the weld currents selected (experiments 10-12) having good bending strength. No crack, blow hole or spatter was Sl. No. Weld Speed (mm/ min) Weld Voltage (V) Weld Current (A) Spatter Pene tration Blow Hole Uniformity of Weld Deposition Observed crack Bend ing Load (kN) A1 370.5 25 140 Large Less Large Poor weld deposition Transverse, longitudinal under-bead crack 8.8 A2 370.5 25 150 Less Good Less Thin weld deposition Toe crack 15.8 A3 370.5 25 160 Less Good Less Thin weld deposition Transverse, longitudinal crack 9.2 A4 370.5 30 140 No Very good Very less Good, continuous deposition Longitudinal crack 7.8 A5 370.5 30 150 No Very good Very less Good, continuous deposition At HAZ 9 A6 370.5 30 160 No Good No Good, continuous deposition Transverse crack 10.2 A7 475.75 25 140 Some Less Medium Disconti nuous deposition Toe crack 7 A8 475.75 25 150 Little Less Medium Not a smooth deposition Transverse, longitudinal, under-bead crack 7.8 A9 475.75 25 160 No Medium No Good deposition Transverse, longitudinal, root crack 6.5 A10 475.75 30 140 No Good No Good deposition No crack 13 A11 475.75 30 150 No Good No Good deposition No crack 13.4 A12 475.75 30 160 No Good No Good deposition No crack 16 IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 257 Vol. 5 Issue 2 2013 ISSN 1936-6744 detected in the weld portion or in the HAZ (see Table 2). In these cases of 475.75 mm/min weld speed and weld voltage of 30 V, increase in heat input from 0.42 to 0.48 kJ/mm resulted in an increased of bending load. 4. Optimal selection of process parameters The Analytic Hierarchy Process (AHP) was introduced by Saaty (1977). The hierarchy structure used in this work is shown in Figure 1. The goal or objective of the decision- making process is placed at the top level of the hierarchy. The goal or objective of the AHP in this work is the selection of optimum process parameter combination. The criteria and decision alternatives come in the subsequent descending levels. Six criteria, as detailed in Table 3, are considered in order to determine the best alternative out of a total of 12 alternatives listed in Table 1. Each alternative corresponds to a typical parametric combination for welding test corresponding to a typical experimental run. In this way, at two levels of weld speed and two levels of weld voltage, weld current is varied at three levels, and hence, twelve (2x2x3 = 12) experimental runs are performed. From these runs, the set of process parameters giving the best quality weld will be selected. The pair wise comparison matrices are formed by comparing an element with the elements of the next higher level. This determines the local priority weights. A typical pair wise comparison matrix (A) is shown in Equation (1). Here, aij (for i, j = 1,2,3,…….n) is the strength of preference of the alternative Ai over Aj corresponding to the criterion, C, aji = 1 / aij and aii = 1 for all values of i and j. A = (Equation 1) The numerical values of aij are taken from the ratio scale (Table 4). When all the elements of the matrix are selected, consistency of the entries of the matrix needs be checked. A comparison matrix is said to be consistent if, aij ajk = aik for all values of i, j and k (Equation 2) For a consistent matrix, aij = wi /wj for all values of i and j (Equation 3) where, w is the priority weight. C A1 A2 … An A1 A2 A3 . . An a11 a12 … a1n a21 a22 … a2n a31 a32 … a3n . . . . . . an1 an2 … ann IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 258 Vol. 5 Issue 2 2013 ISSN 1936-6744 Usually, matrix A is rarely consistent, that is, aij ≠ aik akj for some elements of the matrix. Then, a priority weight can be evaluated solving Equation 4; Aw = λm w (Equation 4) where, w = (w1,w2,w3,..) T , λm  n, and λm is the largest eigen value of the matrix A. On the other hand, for a consistent matrix, Equation 4 becomes; Aw = nw (Equation 5) For an inconsistent matrix, the degree of inconsistency is measured by consistency index (CI). CI = (λm - n) / (n - 1) (Equation 6) A random index (RI) is computed through evaluating the consistency index of a matrix with the elements randomly generated from the range of ratio scale (1/9, 1/8, 1/7, …1, …, 7, 8, 9). The consistency ratio (CR = CI/RI) is then calculated, and a consistency ratio of up to 10% is considered acceptable. Table 3 Criteria selected for judging a sound weld Local weights, wi of a comparison matrix for a criterion or an alternative are next determined by solving the Equation 7. n wi = Σ (aij wi )/λm, i=1,2,3,…….,n (Equation 7) j=1 if Pj (j =1, 2, 3, …, m) are the priority weights of n alternatives with respect to the jth criterion, and if qij are the priority weights of the criteria, then global weights (ri) of alternatives are calculated as m ri = Σ (Pj qij ), i =1,2,3, ………,n (Equation 8) j=1 Criterion No. Criterion C1 No spatter C2 Good penetration C3 No blow hole C4 Good weld deposition C5 No surface crack C6 Good bending load IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 259 Vol. 5 Issue 2 2013 ISSN 1936-6744 The largest global weight thus obtained is the optimum one, and a corresponding alternative is recommended as the optimum solution after Saaty (1977), Saaty (1980), Vargas (1990) and Das et al. (2003). ------ Goal ------ Criteria - - - - - -- ------ Alternatives Figure 1. Hierarchy structure of the AHP used Table 4 Ratio scale of comparison matrix The pair wise comparison matrix for criteria is constructed to solve the present problem of the selection of optimum parametric combination, and is given in Table 5. This table shows preferences for selection of a criterion compared with the other criterion to judge a quality weld. Good penetration and good bending load are more highly preferred than spatter, as a high bending load indicates good load sustaining capability of the weld. Good weld penetration facilitates this. These weights of preferences have been introduced based on the experiences from different welding tests. Good, uniform weld deposition has slightly less preference compared to penetration and bending load, as it has less influence Preferential Judgment Rating Extremely preferred 9 Very strongly to extremely preferred 8 Very strongly preferred 7 Strongly to very strongly preferred 6 Strongly preferred 5 Moderately to strongly preferred 4 Moderately preferred 3 Equally to moderately preferred 2 Equally preferred 1 A11: Expt. 11 A2: Expt. 2 A1: Expt. 1 A12: Expt. 12 C5: No Crack C4: Uniformity in Weld Deposition C3: No Blow Hole C2: Good Penetration C1: No Spatter C6: Bending Load Selection of optimum parameter combination in GMAW IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 260 Vol. 5 Issue 2 2013 ISSN 1936-6744 than the other factors when determining a good weld. The presence of cracks and blow holes is next in order of preference. Even if there is an apparent presence of a crack and blow hole in a weld, if there is high bending strength along with deep penetration, the weld may still be usable. If a crack cannot propagate, and it is arrested summarily, it may not cause any failure in the component. A blow hole presents some discontinuity; however, if bending strength of the weld is good in spite of presence of a blow hole, the weld may be acceptable. With these considerations, priority weights of the criteria matrix are chosen, and the consistency ratio (CR) of the matrix comes out to be less than 10% which signifies consistency of the chosen values. In this work, no commercial software was used for calculations. The local weight is not calculated raising the powers to the pairwise comparison matrix; however, it is calculated using a computer programme written by the authors in C++ language in the following manner: i) First, the elements of each column are normalized by dividing each element of a column by the arithmetic sum of elements of that column. ii) Local weight of a row is then calculated by arithmetic mean of the normalized row elements. Table 5 The criteria matrix Optimum Quality Weld C1 C2 C3 C4 C5 C6 Local Weight C1 1 1/7 1/3 1/6 ¼ 1/8 0.0317 C2 7 1 3 1 4 ½ 0.2295 C3 3 1/3 1 1/3 ½ 1/5 0.0716 C4 6 1 3 1 3 ½ 0.2107 C5 4 ¼ 2 1/3 1 1/4 0.0958 C6 8 2 5 2 4 1 0.3606 Principal eigen value, λmax = 6.1521, CR = 0.004469 For each criterion (C), preferences of the alternatives (A) are tabulated in Table 6 through Table 11. Table 6 shows the relative priorities within any two alternatives (experiments) considering the occurrence of no spatter (criteria, C1). As A1 alternative (experiment 1) has large spatter, and alternatives A4, A5, A6, A10, A11 and A12 show no spatter, compared to A1 alternative, these six alternatives are assigned a ‘very strong preference’ (a preferential strength of 7). On the other hand, presence of low spatter in A2 and A3 compared to that of A1 alternative, results in assigning the preferential rating of 4 (that is, moderately to strongly preferred). Table 7 shows the pair-wise comparison matrix for alternatives with respect to criterion, C2 which is good penetration. Compared to less penetration observed in A1 alternative, A7 and A8 alternatives have similar less penetration, and hence, are assigned a value of equal preference (that is 1). Similarly, alternative A6 shows good penetration and has astrength of preference of 5 which signifies ‘strongly preferred’ compared with alternative A1. IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 261 Vol. 5 Issue 2 2013 ISSN 1936-6744 Table 6 Pair-wise comparison matrix for alternatives for criterion 1 (no spatter) C1 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 Local weight A1 1 1/4 ¼ 1/7 1/7 1/7 1/5 1/6 1/6 1/7 1/7 1/7 0.0133 A2 4 1 1 1/5 1/5 1/5 ½ 1/3 1/3 1/5 1/5 1/5 0.0266 A3 4 1 1 1/5 1/5 1/5 ½ 1/3 1/3 1/5 1/5 1/5 0.0266 A4 7 5 5 1 1 1 4 2 2 1 1 1 0.1263 A5 7 5 5 1 1 1 4 2 2 1 1 1 0.1263 A6 7 5 5 1 1 1 4 2 2 1 1 1 0.1263 A7 5 2 2 ¼ 1/4 ¼ 1 ½ ½ ¼ 1/4 1/4 0.0387 A8 6 3 3 ½ 1/2 ½ 2 1 1 2 2 2 0.103 A9 6 3 3 ½ 1/2 ½ 2 1 1 2 2 2 0.103 A10 7 5 5 1 1 1 4 ½ ½ 1 1 1 0.103 A11 7 5 5 1 1 1 4 ½ ½ 1 1 1 0.103 A12 7 5 5 1 1 1 4 ½ ½ 1 1 1 0.103 Principal eigen value, λmax = 12.689, CR = 0.0039 Table 7 Pair-wise comparison matrix for alternatives for criterion 2 (good penetration) C2 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 Local weight A1 1 1/3 1/3 1/4 1/4 1/5 1 1 1/3 1/3 1/3 ¼ 0.0263 A2 3 1 1 1/3 1/3 1/4 3 3 1 1 1 1/3 0.0592 A3 3 1 1 1/3 1/3 1/4 3 3 1 1 1 1/3 0.0592 A4 4 3 3 1 1 1/3 4 4 3 3 3 1 0.1329 A5 4 3 3 1 1 1/3 4 4 3 3 3 1 0.1329 A6 5 4 4 3 3 1 5 5 4 4 4 3 0.2354 A7 1 1/3 1/3 1/4 1/4 1/5 1 1 1/3 1/3 1/3 ¼ 0.0263 A8 1 1/3 1/3 1/4 1/4 1/5 1 1 1/3 1/3 1/3 ¼ 0.0263 A9 3 1 1 1/3 1/3 1/4 3 3 1 1 1 ½ 0.0608 A10 3 1 1 1/3 1/3 1/4 3 3 1 1 1 ½ 0.0608 A11 3 1 1 1/3 1/3 ¼ 3 3 1 1 1 ½ 0.0608 A12 4 3 3 1 1 1/3 4 4 2 2 2 1 0.1190 Principal eigen value, λmax = 12.3832, CR = 0.0019 Table 8 illustrates the pair-wise comparison matrix for alternatives with respect to the criterion C3, that is, lack of a blow hole. Compared to presence of large blow holes in experiment 1 (alternative A1), occurrence of less or no blow hole is assigned a priority ratio of moderately to strongly preferred (3 to 5). Similarly, Table 9, 10 and 11 are constructed for pair-wise comparison matrix of alternatives with respect to criteria C4 (uniformity of weld deposition), C5 (no crack) and C6 (bending load). IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 262 Vol. 5 Issue 2 2013 ISSN 1936-6744 Table 8 Pair-wise comparison matrix for alternatives for criterion 3 (no blow hole) C3 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 Local weight A1 1 1/3 1/3 1/4 1/4 1/5 1 1 1/3 1/3 1/3 ¼ 0.0188 A2 3 1 1 1/3 1/3 1/4 3 3 1 1 1 1/3 0.0188 A3 3 1 1 1/3 1/3 1/4 3 3 1 1 1 1/3 0.0371 A4 4 3 3 1 1 1/3 4 4 3 3 3 1 0.0685 A5 4 3 3 1 1 1/3 4 4 3 3 3 1 0.0685 A6 5 4 4 3 3 1 5 5 4 4 4 3 0.1429 A7 1 1/3 1/3 1/4 1/4 1/5 1 1 1/3 1/3 1/3 ¼ 0.0371 A8 1 1/3 1/3 1/4 1/4 1/5 1 1 1/3 1/3 1/3 ¼ 0.0371 A9 3 1 1 1/3 1/3 1/4 3 3 1 1 1 ½ 0.1429 A10 3 1 1 1/3 1/3 1/4 3 3 1 1 1 ½ 0.1429 A11 3 1 1 1/3 1/3 1/4 3 3 1 1 1 ½ 0.1429 A12 4 3 3 1 1 1/3 4 4 2 2 2 1 0.1429 Principal eigen value, λmax = 12.3832, CR = 0.0017 Table 9 Pair-wise comparison matrix for alternatives for criterion 4 (good weld deposition) C4 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 Local weight A1 1 1 1 1/3 1/3 1/4 1 1 1/2 1/6 1/5 1/5 0.0285 A2 1 1 1 1/3 1/3 1/4 1 1 1/2 1/6 1/5 1/5 0.0285 A3 1 1 1 1/3 1/3 1/4 1 1 1/2 1/6 1/5 1/5 0.0285 A4 3 3 3 1 1 1/3 3 3 2 1/5 1/3 1/3 0.0708 A5 3 3 3 1 1 1/3 3 3 2 1/5 1/3 1/3 0.0708 A6 4 4 4 3 3 1 4 4 3 1/3 1/2 1/2 0.1185 A7 1 1 1 1/3 1/3 1/4 1 1 1/2 1/6 1/5 1/5 0.0285 A8 1 1 1 1/3 1/3 1/4 1 1 1/2 1/6 1/5 1/5 0.0285 A9 2 2 2 1/2 1/2 1/3 2 2 1 1/4 1/3 1/3 0.0506 A10 6 6 6 5 5 3 6 6 4 1 2 2 0.2354 A11 5 5 5 3 3 2 5 5 3 1/2 1 1 0.1557 A12 5 5 5 3 3 2 5 5 3 1/2 1 1 0.1557 Principal eigen value, λmax = 12.2903, CR = 0.0015 IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 263 Vol. 5 Issue 2 2013 ISSN 1936-6744 Table 10 Pair-wise comparison matrix for alternatives for criterion 5 (no surface crack) C5 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 Local weight A1 1 1/5 1/3 1/5 1/5 1/5 1/5 1/2 1/2 1/7 1/7 1/7 0.0148 A2 5 1 3 1 1 1 1 4 4 1/3 1/3 1/3 0.0866 A3 3 1/3 1 1/3 1/3 1/3 1/3 2 2 1/4 1/4 1/4 0.0410 A4 5 1 3 1 1 1 1 4 4 1/3 1/3 1/3 0.0866 A5 5 1 3 1 1 1 1 4 4 1/3 1/3 1/3 0.0866 A6 5 1 3 1 1 1 1 4 4 1/3 1/3 1/3 0.0866 A7 5 1 3 1 1 1 1 4 4 1/3 1/3 1/3 0.0866 A8 2 1/4 ½ 1/4 1/4 1/4 ¼ 1 1 1/5 1/5 1/5 0.0250 A9 2 1/4 ½ 1/4 1/4 1/4 ¼ 1 1 1/5 1/5 1/5 0.0250 A10 7 3 4 3 3 3 3 5 5 1 1 1 0.1536 A11 7 3 4 3 3 3 3 5 5 1 1 1 0.1536 A12 7 3 4 3 3 3 3 5 5 1 1 1 0.1536 Principal eigen value, λmax = 12.404, CR = 0.0019 Table 11 Pair-wise comparison matrix for alternatives for criterion 6 (good bending load) C6 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 Local weight A1 1 1/5 1 2 1 1 2 1 2 1/4 1/4 1/5 0.444 A2 5 1 5 6 5 5 6 5 6 2 2 1 0.2074 A3 1 1/5 1 2 1 1 2 1 2 1/4 1/4 1/5 0.4532 A4 ½ 1/6 1/2 1 1 ½ 1 1/2 1 1/5 1/5 1/6 0.2758 A5 1 1/5 1 1 1 ½ 1 1/2 1 1/5 1/5 1/6 0.3167 A6 1 1/5 1 2 2 1 2 1 2 1/4 1/4 1/5 0.4709 A7 ½ 1/6 1/2 1 1 ½ 1 1/2 1 1/5 1/5 1/6 0.2758 A8 1 1/5 1 2 2 1 2 1 2 1/4 1/4 1/5 0.4709 A9 ½ 1/6 1/2 1 1 ½ 1 1/2 1 1/5 1/5 1/6 0.0307 A10 4 1/2 4 5 5 4 5 4 5 1 1 1/2 0.1384 A11 4 1/2 4 5 5 4 5 4 5 1 1 1/2 0.1464 A12 5 1 5 6 6 5 6 5 6 2 2 1 0.2065 Principal eigen value, λmax = 12.1907, CR = 0.0010 Combining the pair-wise comparison matrix for criteria and that for alternatives, a global matrix is found as shown in Table 12. In the present work, a computer programme is developed by the authors using C++ computer programme that takes the input of the element data of all matrices, and computes the consistency ratio and global weight following the same steps as that of manual calculations. IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 264 Vol. 5 Issue 2 2013 ISSN 1936-6744 Table 12 Global weights for alternatives Alternatives Weld Speed (mm/min) Weld Voltage (V) Weld Current (A) Global Weight A1 370.5 25 140 0.0314 A2 370.5 25 150 0.103 A3 370.5 25 160 0.0425 A4 370.5 30 140 0.0715 A5 370.5 30 150 0.073 A6 370.5 30 160 0.1174 A7 475.75 25 140 0.0331 A8 475.75 25 150 0.0372 A9 475.75 25 160 0.0505 A10 475.75 30 140 0.1459 A11 475.75 30 150 0.1292 A12 475.75 30 160 0.1655 5. Discussion of AHP results Many process parameters influence GMAW or MAG performance, and three main parameters are selected for the present investigation. The results, as detailed in Table 2, show that the relationship among parameters chosen is not simple enough to draw a clear conclusion. Therefore, the AHP is used in this work to discover the appropriate combination of process variables to obtain sound welding. Experimental observations made in GMAW show that at a welding voltage of 30 V, 140-160 A welding current and 475.75 mm/min speed condition, a good quality weld is obtained. At a lower weld voltage, weldments begin to exhibit a number of weld defects. The AHP is used to find out the optimized process conditions by choosing suitable weights in the criteria matrix and the alternative matrices, and finally combining these weights to find the global matrix as shown in Table 12. The expertise of the authors is utilized to choose the pair-wise comparison ratio, and these are comparable with some other published articles by Sabiruddin et al. (2009) and Muralidharan et al. (1999). If global weights against each alternative are arranged in descending order, the same appears to be: A12 > A10 > A11 > A6 > A2 > A5 > A4 > A9 > A3 > A8 > A7 > A1. Therefore, the AHP indicates that the A12 alternative be chosen for GMAW for joining C45 medium carbon steel specimens. This corresponds to a setting of a weld voltage of 30 V, welding current of 160 A and 475.75 mm/min speed of the welding torch. Although, a weld voltage of 30 V with 140-160 A weld current at 475.75 mm/min speed condition (experiments A10 and A11) have been found experimentally to be somewhat good for having a sound weld, the AHP refines the experimental results further to give the optimum welding process parameters within the domain of conditions considered in this work. Conditions for experiments 6 and 2 may also be considered since they show global weights slightly less than that obtained from experiments 12, 10 and 11. IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 265 Vol. 5 Issue 2 2013 ISSN 1936-6744 6. Conclusion The following conclusions may be drawn from the present investigation on joining C45 medium carbon steel specimens using gas metal arc welding employing 100% carbon di- oxide as the shielding gas, and to find out the optimal set of process parameters utilizing the AHP. Three process parameters, weld speed, weld voltage and weld current were varied to evaluate the best combination of process parameters corresponding to an experimental run within the domain of the present work. As these process parameters have varying influence on weld quality, the AHP was employed to discover the experimental run(s) giving the desired quality of weld. The AHP analysis considered six criteria for joining medium carbon steel specimens optimally, and a weld voltage of 30 V, weld current of 160 A, and welding speed of 475.75 mm/min were chosen for the selected electrode and workpiece. This result corresponded to the maximum global weight of the A12 alternative. This is also agreeable with the experimental results. At this condition, heat input is supposed to be quite favourable to facilitate good weld penetration, high bending strength, and lack of the presence of a crack, spatter and blow hole. Therefore, this condition may be recommended for implementation to obtain sound welding. Hence, while gas metal arc welding of medium carbon steel workpieces is used, the AHP helps managerial decision-making so that the management may prepare the process sheet specifying the evaluated optimized process parameters to set in order to have a defect- free, good welded joint. IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 266 Vol. 5 Issue 2 2013 ISSN 1936-6744 REFERENCES Das, S., Islam, R. & Chattopadhyay, A.B. (1997). A simple approach for on-line tool wear monitoring using the analytical hierarchy process. Proceedings of the IMechE, Journal of Engineering Manufacture, 211, 19-27. Jatua, R., Sarkar, A., Nandi, S. & Das, S. (2003). On the effects of various parameters for gas metal arc welding of steels. Proceedings of Conference on Advances and Recent Trends in Manufacturing, Kalyani, India, 143-150. Jaubari, M.K., Prasad, N.V., Rao, A.T., Premchander, V. & Nagaraju, J.S. (2007). Innovative cost effective concept in GMAW process to improve/ enhance weld aesthetics, quality and fatigue life. Indian Welding Journal, 40(1), 40-45. Jones, L.A., Eager, T.W. & Lang, J.H. (1992). Metal transfer control in gas metal arc welding. (Unpublished dissertation). Massachusetts Institute of Technology, USA. Khanna, O.P. (1995). Welding technology. New Delhi: Khanna Publishers. Kim, I.S. & Basu, A. (1998). A mathematical model of heat transfer and fluid flow in the gas metal arc welding process. Journal of Materials Processing Technology, 77, 17-24. Lai, X., Ji, C., Luo, X. & Deng, L. (2009). Application of AHP method of orthogonal trial to selection of parameters in resistance spot welding. Electric Welding Machine, 49, 7-8. DOI: cnki: ISSN: 1001-2303.0.2006-09-003. Liu, X. & Gong, S.L. (2011). Evaluation on the effect of weld shape on fatigue performance by analytic hierarchy process, Advanced Materials Research, 146/147, 1839-1842. Modensi, P.J. & DeAvelar, R.C. (1999). The influence of small variations of wire characteristics on gas metal arc welding process stability. Journal of Materials Processing Technology, 86, 226 – 232. Mohandas, T., Reddy, G.M. & Kumar, B.S. (1999). Heat-affected zone softening in high strength low-alloy steels. Journal of Materials Processing Technology, 88, 264–294. Muralidharan, C., Anantharaman, N., Balasubramanian, V. & Deshmukh, S.G. (1999). Selection of a welding process using analytic hierarchy process. Journal of the Institution of Engineers (India), 80, 51–54. Nadkarni, S.V. (1996). Modern arc welding technology. New Dehli: Oxford and Indian Book House. Ravisankar, V., Balasubramanian, V. & Muralidharan, C. (2006). Selection of welding process to fabricate butt joints of high strength aluminium alloys using analytic hierarchic process., Materials & Design, 27, 373-380. http://www.scientific.net/AMR.146-147.1839 http://www.scientific.net/AMR.146-147.1839 IJAHP Article: Sabiruddin, Bhattacharya, Das/ Selection of appropriate process parameters for gas metal arc welding of medium carbon steel specimen International Journal of the Analytic Hierarchy Process 267 Vol. 5 Issue 2 2013 ISSN 1936-6744 Saaty, T.L. (1977). A scaling method for priorities in hierarchical structures. Journal of Mathematical Psychology,15(3), 234-281. Saaty T.L. (1980). Analytic Hierarchy Process. New York: McGraw-Hill. Saaty, T.L. (2009). An essay on how judgment and measurement are different in science and in decision making. International Journal of the Analytic Hierarchy Process, 1(1), 61-62. Sabiruddin, K. & Das, S. (2005). Characterisation of gas metal arc welding of different steels. Proceedings of All India Workshop on Advances in Welding Science and Technology, Shibpur, Howrah, India. Sabiruddin, K., Das, S. & Bhattacharya, A. (2009). Application of the analytic hierarchic process for optimization of process parameters in GMAW. Indian Welding Journal, 42(1), 38-46. Simpson, W. & Hughes, P.W. (2006). Industrial application of welding signatures for real time quality assurance and process improvement, (Unpublished dissertation). School of Electrical and Information Engineering, University of Sydney, Australia. Vargas, L.G. (1990). An overview of the analytic hierarchy process and its applications. European Journal of Operations Research, 48, 72–80. Zumelzu, E., Sepulvede, J. & Ibarra, M. (1999). Influence of microstructure on the mechanical behaviour of welded 316L SS. Journal of Materials Processing Technology, 94(1), 36–40.