Al-Khwarizmi Engineering Journal Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80 (2014) Determination of Welding Velocity and Arc Energy for Fusion MAG Welding Joint Salah Sabeeh Abed-Alkareem Department of Machines and Agricultural Equipment / University of Baghdad Email: dr.salah2007@yahoo.com (Received 22 April 2014; accepted 27 September 2014) Abstract This paper is an experimental work to determinate the effect of welding velocity and formed arc energy for CO2- MAG fusion weld pool. The input parameters (arc voltage, wire feed speed and gas flow rate) were investigated to find their effects on the weld joint efficiency. Design of experiment with response surface methodology technique was used to build empirical mathematical models for welding velocity and arc energy in term of the input welding parameters. The predicted quadratic models were statistically checked for adequacy purpose by ANOVA analysis. Additionally, numerical optimization was conducted to obtain the optimum values for welding velocity and arc energy. A good agreement was found between experimental and predicted results. Keywords: Welding Velocity, Arc Energy, Mathematical Modeling, Numerical Optimization, Joint Efficiency. 1. Introduction Carbon Dioxide, as inert gas is normally considered effective in MAG welding process, since the heat of the arc breaks down the CO2 into carbon monoxide and free oxygen. The oxygen will combine with elements transferring across the arc to form the oxides which are released from the weld pool in the form of slag and scale. Although CO2 is an active gas and produces an oxidizing effect, efficient joint welds are achieved almost free of porosity and weld defects. CO2 is widely used for welding mostly for low carbon steel due to its common availability, quality weld efficiency and low cost [1]. CO2 -MAG is an electric arc welding process which joins metals by heating them with an arc established between the electrode and the work. Industrially, CO2-MAG welding is one of the most important processes for welding sheet metal engineering applications, such as automobiles, structures and marine parts. CO2-MAG welding is versatile, gives very little loss of alloying elements and can be operated as semi as well as fully automated. Many commercial metals can be welded by the CO2- MAG process, including carbon steels, stainless steels, Aluminum, copper and nickel alloys [2]. The recommended arc energy (rate of heat input) results in good mechanical properties in the heat affected zone. The rate of heat input supplied by the welding process affects the efficiency of the welded joint. This is described by the arc energy that can be calculated taking in account arc voltage, welding current and speed of welding [3]. Arc energy is one of the most important process parameters in controlling weld response. It can be referred to as an electrical energy supplied by the welding arc to the weldment. In practice, however, arc energy can approximately (i.e., if the arc efficiency is not taken into consideration) be characterized as the ratio of the arc power supplied to the electrode to the arc travel speed [4]. The welding speed has a major influence on the arc energy formed during CO2-MAG welding, since a high welding speed will provide a lower energy input per unit of length of a welded joint mailto:dr.salah2007@yahoo.com Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 65 [5]. And, this will result in insufficient melting of the base metal. Many researchers have been previously carried out by using CO2-MAG welding processes considering mainly the effect of process parameters on the structure and mechanical properties, but there is little works have focused on studying the influence of these parameters on the welding velocity and arc energy using the Design of Experiment (DOE) and Response Surface Methaolodgy (RSM) technique for modeling and optimization purposes for CO2- MAG welding [6-13]. Therefore, the aim of this paper is to determine experimentally the effect of input parameters of this welding process on the welding velocity and arc energy. 2. Experimental Procedure 2.1. Used Material and Samples Preparation Low carbon steel material type AISI 1010 in form of plate with 5 mm thickness in the hot rolled condition was used in this work to prepare samples for welding test. Chemical analysis for this material was carried out, and the results are given in Table 1. Also, the mechanical properties of this steel were obtained by tensile test according to ASTM-E8 standard, and the resulted data are listed in Table 2, presenting the average of three readings for three tested samples .Samples were then prepared with dimensions of 50 mm× 25 mm×5 mm to be welded in a closed Butt weld joint design by CO2-MAG process. Table 1 and Table 2 indicate that the used material matches to the standard base metal [14]. 2.2. Used Welding Parameters The effective selected input factors of CO2- MAG welding in this work were welding speed, arc voltage and wire feed speed in two levels, as shown in Table 3. These parameters were used based on the ability of welding machine and experimental skill of the welder operator. Table 1, Chemical composition of AISI 1010 the Steel Plate With Standard Type (wt%). Table 2, Mechanical Properties for Steel. Elongation (%) Tensile strength (MPa) Yield strength (MPa) Material 42 391 262 Steel Sample Material C Si Mn P S Cr Mo Used Material 0.13 0.015 0.450 0.003 0.003 0.001 0.002 Standard Steel AISI 1010 [12] 0.08 – 0.13 0.1 max 0.3 - 0.6 0.04 max 0.05 max -- -- Material Ni Al Co Cu V Fe Used Material 0.043 0.036 0.007 0.001 0.001 Bal. Standard Steel AISI 1010 [12] -- -- -- -- -- Bal. Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 66 Table 3, Levels of Input Parameters Used with Respective Coding. Input parameter Unit Low Level - 1 High Level + 1 -alpha +alpha Voltage volt 19 21 18 22 Wire feeding speed cm/min 125 175 100 200 Gas flow rate L/min 8 12 5 14 2.3. Welding Test CO2-MAG welding tests were conducted for twenty samples using the welding factors mentioned above and depending on the design matrix established by Design of Experiment Software, as given in Table 4. These tests were achieved randomly to prevent any systematic error. The wire filler type ‘AWS ER70S-6’1.2 mm diameter in form of rod was used for welding samples. Table 4, Design Matrix for Input Factors and Experimental Values of Output (Responses). Std. No. Run No. Type of point Voltage (volt) Wire feed speed (cm/min) Gas flow rate (L/min) Welding velocity (mm/min) Arc Energy (Joul/mm) 1 12 Factorial 19 125 8 64.66 920 2 7 Factorial 21 125 8 73.14 950 3 8 Factorial 19 175 8 90 1640 4 1 Factorial 21 175 8 105 1400 5 14 Factorial 19 125 12 69.72 925 6 4 Factorial 21 125 12 85.68 900 7 16 Factorial 19 175 12 103.44 1380 8 18 Factorial 21 175 12 114 1325 9 9 Axial 18 150 10 66.66 1600 10 15 Axial 22 150 10 85 1530 11 6 Axial 20 100 10 65 230 12 2 Axial 20 200 10 125 1180 13 19 Axial 20 150 6 95.32 1050 14 10 Axial 20 150 14 115 850 15 3 Center 20 150 10 70 1750 16 11 Center 20 150 10 75 1700 17 17 Center 20 150 10 74.5 1690 18 5 Center 20 150 10 71 1715 19 13 Center 20 150 10 75.5 1800 20 20 Center 20 150 10 73.8 1640 2.4. Determination of Welding Velocity Speed of welding is defined as the rate of travel of the electrode along the seam or the rate of travel of the work under the electrode along the seam. Therefore, the welding speed during welding each sample was calculated using the following formula [15]: S = d t …(1) Where, S = welding Speed (mm/min). d = Travel of electrode (mm). t = Arc time (min). In the present, the travel of electrode (d) was first measured using a digital venire (with accuracy of 0.01mm). Then, the arc time (t) was measured for the period between the start and end of achieving the welding pass using a stop watch. According to equation (1), the welding velocity (S) was calculated. Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 67 2.5. Determination of Arc Energy Arc energy (heat input rate) is a relative measure of the energy transferred per unit length of weld. It is typically calculated as the ratio of the power (i.e., voltage, current) to the velocity of the heat source (i.e., the arc) as follows [16]. Q = V∗I∗ 60 S … (2) Where, Q = Arc energy (J/mm), V = arc voltage (volts) and I = welding current (ampere). For calculating the arc energy (Q), the reading of both arc voltage (V) and welding current (I) were taken from the welding machine indicators during the welding process. Then, using the calculated welding velocity (S) as mentioned in section (2.4), the arc energy (Q) was obtained by equation (2).The results of measurements and calculations for these responses are also listed in Table 4. 3. Results and Discussion 3.1. Mathematical Model of Welding Velocity For the welding velocity parameter, the analysis of variance (ANOVA) was established by DOE software as shown Table 5. illustrating that the input parameters individually as well as the quadratic terms of voltage, wire feeding speed and gas flow rate are all statistically significant and have the greatest influence on the welding velocity response according to their P-values (< 0.05). The lack of fit test indicates a good model, since it is insignificant with P-value greater than 0.05. So, this analysis indicates that this model is significant at 95% confidence. In addition, this model showed a good agreement between the predicted and actual values for welding velocity, as shown in Fig.1. Therefore, the final predicted equation for the welding velocity in terms of the coded input factors is: Welding velocity = +74.28 + 5.42 * A + 14.95 * B + 4.96 * C + 5.35 * B2 + 7.89 * C2 … (3) And , the final equation in terms of actual factors: Welding velocity = + 241.40000 + 5.41750 * Voltage - 1.97110 * Wire feeding speed - 36.98125 * Gas flow rate + 8.56400E-003 * Wire feeding speed2 + 1.97313 * Gas flow rate2 … (4) Table 5, Analysis of Variance (ANOVA) for Response Surface Reduced Quadratic Model (Welding Velocity). Source Sum of squares df Mean square F value p-value Prob > F Model 6508.89 5 1301.78 243.24 < 0.0001 significant A-Voltage 469.59 1 469.59 87.74 < 0.0001 B-Wire feeding speed 3577.24 1 3577.24 668.42 < 0.0001 C-Gas flow rate 394.02 1 394.02 73.62 < 0.0001 B² 754.99 1 754.99 141.07 < 0.0001 C² 1641.57 1 1641.57 306.73 < 0.0001 Residual 74.93 14 5.35 Lack of Fit 49.33 9 5.48 1.07 0.4970 not significant Purr Error 25.60 5 5.12 Core Total 6583.81 19 Std. Dev. = 2.31 R-Squared = 0.9886 Mean = 84.87 Adj R-Squared = 0.9846 C.V. % = 2.73 Pred R-Squared = 0.9792 PRESS = 137.26 Adeq Precision = 50.037 Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 68 Fig. 1. Predicted Versus Actual Welding Velocity Data. The statistical properties of this model were diagnosed, and it was found that the residuals that falling on a straight line implying errors are normally distributed, as shown in Fig.2. Additionally, the residuals versus predicted actual for welding velocity data revealed no obvious pattern or unusual structure implying models are accurate as shown in Fig.3. Fig. 2. Normal Probability Plot of Residuals for Welding Velocity Data. Fig. 3. Residuals versus Predicted Welding Velocity Data. Design-Expert® Software Welding velocity Color points by value of Welding velocity: 125 64.66 Actual P r e d ic t e d Predicted vs. Actual 60.00 70.00 80.00 90.00 100.00 110.00 120.00 130.00 60.00 70.00 80.00 90.00 100.00 110.00 120.00 130.00 Design-Expert® Software Welding velocity Color points by value of Welding velocity: 125 64.66 Internally Studentized Residuals N o r m a l % P r o b a b il it y Normal Plot of Residuals -2.00 -1.00 0.00 1.00 2.00 1 5 10 20 30 50 70 80 90 95 99 Design-Expert® Software Welding velocity Color points by value of Welding velocity: 125 64.66 Predicted I n t e r n a ll y S t u d e n t iz e d R e s id u a ls Residuals vs. Predicted -3.00 -2.00 -1.00 0.00 1.00 2.00 3.00 60.00 70.00 80.00 90.00 100.00 110.00 120.00 130.00 Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 69 The perturbation of the predicted welding velocity response resulted by varying only one parameter at a time from the center point of the investigated region is shown in Fig.4. It can be seen that increasing all the three input parameters generally increase the welding velocity, since these input parameters increased the fusion effect of the weld joint, which necessitates increasing the welding velocity to keep the stability of the welding process. Fig. 4. Perturbation of Welding Velocity on Wire Feeding Speed and Gas Flow Rate. Due to no statistical problems found, the response surface plots were generated in terms of 2D surface plot as shown in Figs.5-7 depicting the welding velocity as a function of voltage and wire feeding speed at various gas flow rates 8, 10 and 12 L/min, respectively. These figures indicate that both voltage and wire feeding speed have greater influence on increasing the welding velocity than the gas flow rate which has a slight effect. This is possibly due to increase of molten material accumulated in the weld joint caused by higher voltage and wire feeding speed. Also, this is more likely ascribed to the increased chemical reaction of CO2 with the accumulated molten material in the weld joint. Fig. 5. Contour Graph of Welding Velocity as A function of Voltage and Wire Feeding Speed Gas Flow Rate 8 L/min. Design-Expert® Software Factor Coding: Actual Welding velocity Actual Factors A: Voltage = 20 B: Wire feeding speed = 150 C: Gas flow rate = 10 Perturbation Deviation from Reference Point (Coded Units) W e ld in g v e lo c it y -1.000 -0.500 0.000 0.500 1.000 60 70 80 90 100 A A B B C C Design-Expert® Software Factor Coding: Actual Welding velocity Design Points 125 64.66 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 10 19 20 20 21 21 125 135 145 155 165 175 Welding velocity A: Voltage B : W ir e f e e d in g s p e e d 60 65 70 75 80 85 90 95 6 Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 70 Fig. 6. Contour Graph of Welding Velocity as function of Voltage and Wire Feeding Speed Gas Flow Rate 10 L/min. Fig. 7. Contour Graph of Welding Velocity as A function of Voltage and Wire Feeding Speed Gas Flow Rate 12 L/min. Figures 8-10 show 3D surface plots for welding velocity as a function of voltage and wire feeding speed at different gas flow rates, showing similar behavior as mentioned above. Eventually, these observations are confirmed by the cube plot for the welding velocity, as shown in Fig. 11. Fig. 8. 3D Graph of Welding Velocity as A function of Voltage and Wire Feeding Speed at Gas Flow Rate 8 L/min. Design-Expert® Software Factor Coding: Actual Welding velocity Design Points 125 64.66 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 8 19 20 20 21 21 125 135 145 155 165 175 Welding velocity A: Voltage B : W ir e f e e d in g s p e e d 65 70 75 80 85 90 95 100 Design-Expert® Software Factor Coding: Actual Welding velocity Design Points 125 64.66 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 12 19 20 20 21 21 125 135 145 155 165 175 Welding velocity A: Voltage B : W ir e f e e d in g s p e e d 75 80 85 90 95 100 105 110 Design-Expert® Software Factor Coding: Actual Welding velocity 125 64.66 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 8 125 135 145 155 165 175 19 20 20 21 21 40 60 80 100 120 140 W e ld in g v e lo c it y A: Voltage B: Wire feeding speed Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 71 Fig. 9. 3D Graph of Welding Velocity as A function of Voltage and Wire Feeding Speed at Gas Flow Rate 10 L/min. Fig. 10. 3D Graph of Welding Velocity As a function of Voltage and Wire Feeding Speed at Gas Flow Rate 12 L/min. Fig. 11. Cube Shape of Welding Velocity. Design-Expert® Software Factor Coding: Actual Welding velocity Design points above predicted value Design points below predicted value 125 64.66 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 10 125 135 145 155 165 175 19 20 20 21 21 40 60 80 100 120 140 W e ld in g v e lo c it y A: Voltage B: Wire feeding speed Design-Expert® Software Factor Coding: Actual Welding velocity Design points above predicted value Design points below predicted value 125 64.66 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 12 125 135 145 155 165 175 19 20 20 21 21 40 60 80 100 120 140 W e ld in g v e lo c it y A: Voltage B: Wire feeding speed Design-Expert® Software Factor Coding: Actual Welding velocity X1 = A: Voltage X2 = B: Wire feeding speed X3 = C: Gas flow rate Cube Welding velocity A: Voltage B : W ir e f e e d in g s p e e d C: Gas flow rate A-: 19 A+: 21 B-: 125 B+: 175 C-: 8 C+: 12 62.0713 72.1688 92.1488 102.246 73.0788 83.1763 103.156 113.254 6 Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 72 3.2. Mathematical Model of Arc Energy Similarly, the analysis of variance (ANOVA) for RSM reduced quadratic model was determined for the arc energy as given in Table 6. The results in this table show that the wire feeding speed (B) and gas flow rate (C) are statistically significant, since their P-values were very small (< 0.5).While the voltage (A) has no influence on the weld joint. Moreover, this table also indicate that the lack of fit was insignificant (P-value > 0.05), indicating that this model is adequate and significant at 95% confidence. So, the final predicted equation for the arc energy in terms of the coded input factors is: Arc energy = + 1706.02 - 26.88 * A + 246.87 * B - 48.75 * C - 42.61 * A2 - 257.61 * B2 -196.36 * C2 … (5) And , the final equation in terms of actual factors is: Arc energy = - 30222.61364 +1677.67045 * Voltage +133.52955 * Wire feeding speed + 957.44318 * Gas flow rate - 42.61364 * Voltage2 - 0.41218 * Wire feeding speed2 - 49.09091 * Gas flow rate2 … (6) Table 6, Analysis of Variance (ANOVA) for Response Surface Reduced Quadratic Model (Arc energy). Source Sum of squares df Mean square F value p-value Prob > F Model 3.257E+006 6 5.428E+005 114.66 < 0.0001 significant A-Voltage 11556.25 1 11556.25 2.44 0.1422 B-Wire feeding speed 9.752E+005 1 9.752E+005 206.00 < 0.0001 C-Gas flow rate 38025.00 1 38025.00 8.03 0.0141 A2 45657.47 1 45657.47 9.65 0.0084 B² 1.669E+006 1 1.669E+006 352.49 < 0.0001 Residual 61538.07 13 4733.70 Lack of Fit 46617.23 8 5827.15 1.95 0.2390 not significant Purr Error 14920.83 5 2984.17 Core Total 3.318E+006 19 Std. Dev. = 68.80 R-Squared = 0.9815 Mean = 1308.75 Adj R-Squared = 0.9729 C.V. % = 5.26 Pred R-Squared = 0.9529 PRESS = 1.563E+005 Adeq Precision = 37.446 The adequacy of this model was checked to examine the predicted model. Two types of model diagnostics, the normal probability plot and residuals versus the actual values plot, were used for verification, as shown in Figs. 12 and 13 for arc energy respectively. It can be observed from these plots that there was no violation of the normality assumption, since they normal probability plot followed a straight line pattern, the residual was normally distributed, and as long as the residuals versus the predicted values show no unusual pattern and no outliers. Also, this model shows a good agreement between the predicted and actual values for arc energy, as depicted in Fig. 14. Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 73 Fig. 12. Normal Probability Plot of Residuals for Arc Energy Data. Fig. 13. Residuals versus Predicted Arc Energy Data. Fig. 14. Predicted Versus Actual Arc Energy Data. The perturbation plot of the predicted responses caused by changing only one factor at a time from the center point of the experimental region is shown in Fig. 15. This figure indicates that, individually, the wire feeding speed has greater effect than the gas flow rate on arc energy, Design-Expert® Software Arc energy Color points by value of Arc energy: 1800 230 Internally Studentized Residuals N o r m a l % P r o b a b il it y Normal Plot of Residuals -2.00 -1.00 0.00 1.00 2.00 1 5 10 20 30 50 70 80 90 95 99 Design-Expert® Software Arc energy Color points by value of Arc energy: 1800 230 Predicted I n t e r n a ll y S t u d e n t iz e d R e s id u a ls Residuals vs. Predicted -3.00 -2.00 -1.00 0.00 1.00 2.00 3.00 0.00 500.00 1000.00 1500.00 2000.00 Design-Expert® Software Arc energy Color points by value of Arc energy: 1800 230 Actual P r e d ic t e d Predicted vs. Actual 0.00 500.00 1000.00 1500.00 2000.00 0.00 500.00 1000.00 1500.00 2000.00 Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 74 while the voltage is not influential. This is more probably because of the increasing wire feeding speed resulted in an increase in the welding velocity, leading to more accumulation of molten material due to more thermal effect and less chemical affinity of the CO2 gas with the weld joint material. Fig. 15. Perturbation of Arc Energy on Wire Feeding Speed and Gas Flow Rate. Because of no statistical problem with the model, Fig. 16 shows the 2D contour plot for the arc energy as a function of voltage and wire feeding speed at gas flow rate of 10 L/min. Whereas, Figs. 17-19 depict 3D surface plots for the arc energy at gas flow rate of 8, 10 and 12 L/min, respectively. It can be noted from these figures that increasing both wire feeding speed and gas flow rate increases the arc energy due to the increase of quantity of the molten material that resulted by the increasing of the welding velocity and thermal input. Finally, these observations are confirmed by the cube plot for arc energy, as shown in Fig.20. Fig. 16. Contour Graph of Arc Energy as A Function of Voltage and Wire Feeding Speed at Gas Flow Rate 8 L/min. Design-Expert® Software Factor Coding: Actual Arc energy Actual Factors A: Voltage = 20 B: Wire feeding speed = 150 C: Gas flow rate = 10 Perturbation Deviation from Reference Point (Coded Units) A r c e n e r g y -1.000 -0.500 0.000 0.500 1.000 1200 1300 1400 1500 1600 1700 1800 A A B B C C Design-Expert® Software Factor Coding: Actual Arc energy Design Points 1800 230 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 10 19 20 20 21 21 125 135 145 155 165 175 Arc energy A: Voltage B : W ir e f e e d in g s p e e d 1200 1200 1300 1400 1500 1600 1700 1725 1650 1750 1765 6 Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 75 Fig. 17. 3D Graph of Arc Energy as A function of Voltage and Wire Feeding Speed at Speed at Gas Flow Rate 10 L/min. Fig. 18. 3D Graph of Arc Energy as A function of Voltage and Wire Feeding Speed at Gas Flow Rate 10 L/min. Fig. 19. 3D Graph of Arc Energy as A function of Voltage and Wire Feeding Speed at Gas Flow Rate 12 L/min. Design-Expert® Software Factor Coding: Actual Arc energy 1800 230 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 8 125 135 145 155 165 175 19 20 20 21 21 0 500 1000 1500 2000 A r c e n e r g y A: Voltage B: Wire feeding speed Design-Expert® Software Factor Coding: Actual Arc energy Design points above predicted value Design points below predicted value 1800 230 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 10 125 135 145 155 165 175 19 20 20 21 21 0 500 1000 1500 2000 A r c e n e r g y A: Voltage B: Wire feeding speed Design-Expert® Software Factor Coding: Actual Arc energy Design points above predicted value Design points below predicted value 1800 230 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 12 125 135 145 155 165 175 19 20 20 21 21 0 500 1000 1500 2000 A r c e n e r g y A: Voltage B: Wire feeding speed Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 76 Fig. 20. Cube Shape of Arc Energy. 3.3. Numerical Optimization The numerical optimization is provided by the Design of Experiment software to find out the optimum combinations of parameters in order to fulfill the requirements as desired. Therefore, this software was used for optimizing the welding velocity and arc energy; based on the data from the predicted models as a function of three factors: arc voltage, wire feeding speed and gas flow rate. Table 7 lists the constrains of each variable for numerical optimization of the welding velocity and arc energy. According to this table, one possible run fulfilled the specified constrains to obtain the optimum values for welding velocity, arc energy and desirability, as listed in Table 8. It can be noted that this run gave a desirability of 0.849 with the optimum values of the Voltage (20 volt), Wire feeding speed (153 cm/min), Gas flow rate (10 L/min), Welding velocity (77.071 mm/min) and arc energy (1722 Joule/mm). Figures 21-23 show 3D surface plots for desirability, optimum value of welding velocity and optimum value of arc energy, respectively as a function of voltage and wire feeding speed at 10 L/min gas flow rate. Table 7, Constrains Used for the Numerical Optimization. Goal Lower Limit Upper Limit Lower Weight Upper Weight Importance A:Voltage is in range 19 21 1 1 3 B:Wire feeding speed is in range 125 175 1 1 3 C:Gas flow rate is in range 8 12 1 1 3 Welding velocity minimize 64.66 125 1 1 3 Arc energy maximize 230 1800 1 1 3 Table 8, Optimum Solutions of the Desirability. Design-Expert® Software Factor Coding: Actual Arc energy X1 = A: Voltage X2 = B: Wire feeding speed X3 = C: Gas flow rate Cube Arc energy A: Voltage B : W ir e f e e d in g s p e e d C: Gas flow rate A-: 19 A+: 21 B-: 125 B+: 175 C-: 8 C+: 12 1038 941 1532 1434 984 887 1478 1381 6 Number Voltage (volt) Wire feeding speed (cm/min) Gas flow rate (L/min) Welding velocity (mm/min) Arc energy (Joul/mm) Desirability 1 20 153 10 77.071 1722 0.849 Selected Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 77 Fig. 21. 3D Graph for Desirability as A function Of Voltage and Wire Feeding Speed at Gas Flow Rate 10 L/min. Fig. 22. The Optimum Value for Welding velocity at 10 L/min Gas Flow Rate. Fig. 23.The Optimum Value for Arc Energy At 10L/min Gas Flow Rate. Design-Expert® Software Factor Coding: Actual Desirability 1.000 0.000 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 10 125 135 145 155 165 175 19 20 20 21 21 0.400 0.500 0.600 0.700 0.800 0.900 D e s i r a b i l i t y A: Voltage B: Wire feeding speed 0.8490.849 Design-Expert® Software Factor Coding: Actual Welding velocity 125 64.66 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 10 125 135 145 155 165 175 19 20 20 21 21 50 60 70 80 90 100 W e l d i n g v e l o c i t y A: Voltage B: Wire feeding speed 77.040777.0407 Design-Expert® Software Factor Coding: Actual Arc energy 1800 230 X1 = A: Voltage X2 = B: Wire feeding speed Actual Factor C: Gas flow rate = 10 125 135 145 155 165 175 19 20 20 21 21 1100 1200 1300 1400 1500 1600 1700 1800 A r c e n e r g y A: Voltage B: Wire feeding speed 17221722 Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 78 3.4. Weld Joint Efficiency Calculation: In order to obtain the efficiency of the weld joint obtained by CO2-MAG welding of low carbon steel AISI 1010, three tensile samples were first welded with the optimum welding condition given in Table – and then tensile tested to determine the ultimate tensile strength of the weld joint. The average tensile strength was found to be 285 MPa. Therefore, the efficiency of the weld joint was calculated to be 73% according to the joint efficiency definition which is the ratio of the tensile strength of the weld joint to the tensile strength of the base metal (Table 2). This result indicates the importance of using CO2-MAG welding process and its effectiveness and suitability for welding steel AISI 1010 from strength point of view. 4. Conclusions 1. Regarding the welding velocity, a quadratic model was obtained by DOE with RSM technique for the optimum welding velocity response in terms of input welding parameters. His model indicated that the arc voltage, wire feeding speed and gas flow rate largely effective on welding velocity. 2. Concerning the arc energy, a quadratic model was obtained for the optimum arc energy response in terms of input welding parameters. This model shows that the wire feeding speed has greater impact than gas flow rate on arc energy, while the arc voltage was found not affected. 3. By numerical optimization, the optimum values of the voltage, wire feeding speed, gas flow rate, welding velocity, arc energy and desirability are (20Volt),(153cm/min),(10L/min), (77.071mm/min) (1722 Joule/mm ) and ( 0.849), respectively. 4. DOE with RSM was found a useful tool for predicting the responses in MAG-CO2 welding technique for any given input parameters. 5. References [1] Katokihiko, Ikeda Rinsei and Yasuda Koichi," Development of Ultra-low Spatter CO2 Gas-shielded Arc Welding Process", JFE GIHO, No. 16, p. 50–53, June 2007. [2] Parth D. Patel and Sachin P Patel,"Prediction of weld Strength of Metal Active Gas (MAG) Welding Using Artificial Neural Network", International Journal of Engineering Research and Applications (IJERA), ISSN: 2248-9622, Vol. 1, Issue 1, pp.036-044, 1-1- 2014. [3] OP khanna, "A text book of welding technology", Dhanpat Rai Publications Ltd., pp.351, 2006. [4] Neeraj Kumar Sharma and Kunal Sharma, "To Study The Effect of Process Variable on Properties of 2.25Cr-1Mo Steel In GMAW", International Journal of Engineering Research & Technology (IJERT), Vol. 2, Issue 5, ISSN: 2278-0181, May - 2013. [5] Miloš JOVANOVIC , Janez GRUM and Miro URAN 1," Influence of lack-of-fusion defects on load capacity of MAG welded joints", 17th World Conference on Nondestructive Testing, Shanghai- China , 25-28 Oct 2008 . [6] V. Gunaraj and N. Murugan, " Application of response surface methodology for predicting weld bead quality in SAW of pipes", Journal of Mater Processing Technology, Vol. 88, 266-275, 1999. [7] A. I. Khuri and J. Cornell," Response Surfaces Design and Analysis", 2nd ed, Marcel Dekker, New York, (1996). [8] Gunaraj and N. Murugan , " Prediction and Optimization of Weld Bead Volume for the Submerged Arc Process — Part 1", Welding Research Supplement, 286-s - 294-s , October2000. [9] Vinod Kumar, " Modeling of Weld Bead Geometry and Shape Relationships in Submerged Arc Welding using Developed Fluxes". Jordan Journal of Mechanical and Industrial Engineering, Volume 5, Number 5, ISSN 1995-6665 Pages 461 – 470, Oct. 2011. [10] S. Thiru Chitrambalam, Tan Wee Ming, Imran Syakir Mohammad and Shafizal bin Mat , "A Study on Relationship between Process Variables and Weld Penetration for Gas Metal Arc Welding (GMAW)", International Conference and Exhibition on Sustainable Energy and Advanced Materials , Solo-Indonesia,p. 237, October 3-4, 2011. [11] Nischal Chhabra , Nirmal S. Kalsi and Dilbag Singh, " Effect of Shielding Gases on Micro Hardness of FE 410 (AISI 1024) Steel Welded Joint in GMAW Process" , International Journal on Emerging Technologies 5(1), ISSN No. (Online): 2249- 3255, 8-13(2014). Salah Sabeeh Abed-Alkareem Al-Khwarizmi Engineering Journal, Vol. 10, No. 4, P.P. 64- 80(2014) 79 [12] S. Thiru chitrambalam, Chew Lai Huat, Phang Boo Onn, S.Hemavathi, Imran Syakir Mohammad and Shafizal bin Mat. "An Investigation on Relationship between Process Control Parameters and Weld Penetration for Robotic CO2 Arc Welding Using Factorial Design Approach",The Journal of Mechanical Engineering and Technology, Volume 4, Pages 1-16.Publisher University Technical Malaysia Melaka, 2012. [13] Edwin Raja Dhas J and Jenkins Hexley Dhas S," A Review on Optimization of Welding Process". Procedia Engineering 38 (2012) 544 – 554 Elsevier. [14] http://www.efunda.com/materials/alloys/carb on_steels/show_carbon.cfm?ID=AISI_1010 &prop=all&Page_Title=AISI%201010. [15] Jatinder Gill and Jagdev Singh," Efect Of Welding Speed and Heat INPut RateOn Stress Concentration Factor Of Butt Welded Joint Of IS 2062 E 250 A STEEL", International Journal of Advanced Engineering Research and Studies, E- ISSN2249–8974 IJAERS/Vol. I/ Issue III/April-June, 2012/98-100. [16] Ajay N.Boob and Prof.G. K.Gattani, " Study on Effect of Manual Metal Arc Welding Process Parameters on Width of Heat Affected Zone (Haz) For Ms 1005 Steel", International Journal of Modern Engineering Research (IJMER), Vol. 3, Issue. 3, pp-1493- 1500, ISSN: 2249-6645, May, June. 2013. http://www.efunda.com/materials/alloys/carbon_steels/show_carbon.cfm?ID=AISI_1010&prop=all&Page_Title=AISI%201010 http://www.efunda.com/materials/alloys/carbon_steels/show_carbon.cfm?ID=AISI_1010&prop=all&Page_Title=AISI%201010 http://www.efunda.com/materials/alloys/carbon_steels/show_carbon.cfm?ID=AISI_1010&prop=all&Page_Title=AISI%201010 (2014)64- 80 ، صفحة4، العذد10دجلة الخىارزهي الهنذسية الوجلصالح صبيخ عبذ الكرين م 80 (MAG)أيجاد سرعة اللحام وطاقة القىس لىصلة لحام القىس الوعذني االنصهاري صالح صبيخ عبذ الكرين جايعت بغذاد/ انضساعتكهٍت /لسى انًكائٍ واَالث انضساعٍت dr.salah2007@yahoo.com :االنكخشوًَ انبشٌذ الخالصة سبىٌ فً حضًٍ هزا انبحث دساست عًهٍت نغشض اٌجاد حأثٍشسشعت انهحاو وطالت انمىط انخً حخكىٌ نطشٌمت نحاو انمىط انًعذًَ بغاص ثاًَ أوكسٍذ انكاي وانخً حى دساسخها ( ص سشعت حغزٌت سهك انهحاو ويعذل جشٌاٌ انغا, فىنخٍت انمىط)أٌ انعىايم انذاخهت راث األهًٍت نهزِ انطشٌمت هً . بشكت انهحاو االَصهاسي حى اسخعًال حمٍُت حصًٍى انخجاسب يع طشٌمت االسخجابت انسطحٍت نبُاء يىدٌالث سٌاضٍت حخص سشعت انهحاو . ألٌجاد حأثٍشاحها عهى كفاءة انىصهت انًهحىيت ٌعٍت انخً حى انحصىل عهٍها وانخً حى انخُبأ بها لذ دلمج انًىدٌالث انشٌاضٍت انخشب" . وطالت انمىط نىصهت انهحاو بذالنت عىايم انهحاو انذاخهت وانًزكىسة آَفا أيثهٍت انخحهٍم انعذدي انخً حى انحصىل يٍ خالنها انى أدق انمٍى نسشعت فضال عٍ. وحممج أغشاضها بًىثىلٍت بعذ ححهٍهها بطشٌمت ححهٍم انخباٌٍ " أحصائٍا .ة يع انُخائج انخً حى انخُبأ بها يٍ جشاء هزا انبحث وجذث بأَها راث حطابك جٍذ حى يماسَت انُخائج انًسخحصم.انهحاو و طالت انمىط نىصهت انهحاو mailto:dr.salah2007@yahoo.comالبريد mailto:dr.salah2007@yahoo.comالبريد