Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 267 TAGUCHI METHOD TO OPTIMIZE MACHINING CONDITIONS AFFECTING CEMENTED CARBIDE TOOL'S WEAR LIFE IN TURNING AISI 1020 STEEL Kawthar Y. AL-Dulaimi 1 , Dr. Haydar A. H. Al-Ethari 2 , BogdanWarcholinski 3 1(Msc. Student at Dept. of Metallurgical Eng., College of Material's Eng University of Babylon-Hilla-IRAQ) 2(Prof. at Dept. of Metallurgical Eng., College of Material's Eng University of Babylon-Hilla-IRAQ) 3(Prof. at Koszalin University of Technology- Poland) Received 27December 2015 Accepted 21 January 2016 ABSTRACT: The present work studied the performance of a carbide tool with a chemical composition of (65% W, 14 % Ti, 9 % Co and 12 % C). Turning tests were conducted on a workpiece of mild steel (AISI 1020) using four spindle speeds (80, 315, 500 and 800 rpm), two feed rates (0.2 and 0.5 mm/rev) and two depth of cut (0.5 and 0.7 mm). Taguchi method is a statistical approach to optimize the process parameters and improve the quality of components that are manufactured. The objective of this study is using Taguchi method to optimize the machining conditions of a turning operation such as spindle speed; feed rate and depth of cut. Orthogonal array, signal-to-noise ratio, and analysis of variance were employed using Mtb14 software to study the performance characteristics on turning operation represented by the tool life. Accordingly, a suitable mixed orthogonal array L16 (3×4) was selected. The tool life was measured basing on a maximum flank wear width of 0.3 mm.Optimum parameter values were obtained and confirmation experiments were carried out. The analysis results showed that the parameter design of the Taguchi method provides a simple, systematic, and efficient methodology for optimizing the process parameters. Only 6.4 % error was recorded. The regression analysis was applied using Datafit ver.9 software. The results of the analysis showed that the non-linear quadratic polynomial appears to be more suitable to represent the relation of the spindle speed, feed rate and depth of cut with the tool's wear life. . Key words: Tool life, Machining parameters, Taguchi method, Flank wear, Optimization, Modeling. طريقة تاكوجي ألمثلية طرق التشغيل المؤثرة على عمر البلى للعدد الكاربيدية المسمنتة في (AISI 1020خراطة الفوالذ) الخالصة: ( . W, 14 %Ti, 9 % Co and 12 % C %65) ترك ل ك م لي ي رمنلم العمل الاليلي رل أد ء ال العل ك الييأة رل ,500 ,315 ,80)ةيسلمعمي اأةعل سلرو أا ل (AISI 1020)اخمبيأات الخراط ق اجررت علل شغللةل شل ذلةط ش للي Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 268 طررق تليكةجي ( شلم .( and 0.7 0.5شلم/ أك , ق مم لعمق الق ع (and 0.5 0.2) أك/ ق ق , ق م م لمع الملذر ( 800 عمل تاس ةع المركبيت المصنع . ه ف هذه ال أاس هة تةض ح األجلرال المسلمعم هي شنهج إحصي ي لماس شا ات ال . شثلل السللرع ال أا لل , شعلل الملذرلل عمللق الق للع ذللي اسللمخ اي طررقلل تلليكةجي لماسلل يللر ف المغللل ذللي عمل لل الخراطلل ل أاسل شم لتات اط اك ذلي Mtb 14ممعيش ك , سب األشيأك ال النةضيل تال المبلير قل تلم اسلمخ اشهي ةمسلمعمي المصفةذ ال قل تلم اخم يأهلي. عملر العل ك قل تلم L16 (3×4). ذقيً لذلك , شصفةذ شمعيش ك شخمل ل شنيسلب مل الخراط شممثل ةعمر اط اكع . الظر ف المثل ق تم الاصلة عل هلي اخمبليأات الماقلق قل تلم (mm 0.3)لةصي اطقص ق يسه اعممي ا عل عرض البل الخ تنف ذهي . تال النمي ج المجررب ايهرت ءن تصم م المعلم ة ررقل تليكةجي د ت طررقل ةسل , شنهج ل ذعيلل لجعل يلر ف اضيذمه ةيسمعمي ةر ليشج اتيذلت النسلخ الميسلع . ملي ج هلذا . تال المراجع ق تم سب خ م ق سجلت % 6.4العمل شثيل . ذقط المال ايهرت ان الممع الا ش ال أج الثي ل ي لر الخ لي هلة اكثلر شل مل لممث ل العلقل ةل السلرع ال أا ل , شعل . . الملذرللللللللللللللللللللل عملللللللللللللللللللللق الق لللللللللللللللللللللع شلللللللللللللللللللللع عملللللللللللللللللللللر األ اك 1. INTRODUCTION: The tool is an important part required in material removal process in order to produce the components. It not only provides product but also maintains the geometricaltolerances, dimensional accuracies and surface finish of the outcome. The manufacturing industry is striving to improve the productivity, quality and longer tool life. Achievement of longer tool life depends on different factors such as cutting parameters, tool wear, hardness of the work piece and machine tool materials [1,2]. Many improvements can be done to obtain longer tool life duration like changing the tool material, tool geometry, tool edge honing, optimizing machining conditions, using coolants and applying thin layer coating. Tool wear influences cutting power, machining quality, tool life and machining cost. Whentool wear reaches a certain value, variation in cutting force and cutting temperature will causesurface integrity deteriorated and dimension error greater than tolerance. When machining steel with uncoated carbide tools, different tool wear mechanisms occur, such as: abrasion, adhesion, oxidation and even some diffusion, which act simultaneously and mainly depending on the temperature [3,4]. Bala Murugan Gopalsamy et al. in 2009 [5] applied Taguchi method and analysis of variance ANOVA for machining parameters optimization during machining hardened steel. ANOVA, signal-to-noise ratio, and L18 array had been used to study the performance characteristics of the machining conditions ( feed rate, cutting speed and depth of cut ) with consideration of tool life and surface roughness. S.R. Das et al. in 2014 [6] experimentally investigated the effects of the machining conditions on the behaviour characteristics (surface roughness, cutting force and flank wear) by using multi layer coated carbide insert in hard turning of AISI 4340 steel. Taguchi standard has been used to develop the regression models for machining response. Results showed that, the most significant factor on flank wear is the cutting speed and that the feed rate has statistical significance on the surface roughness. The objective of the present work is studying the optimization process of the machining conditions that affecting the tool wear. Basing on the analyses of multiple regression method, mathematical predictive model was designed and validated to select an optimum combination of the studied parameters.Taguchi design is successfully used for optimizing the cutting conditions that affecting tool life. To achieve the analysis of the present work, Datafit ver9 and Mtb14 softwares had been employed. 2. EXPERIMENTAL PROCEDURE : 2.1. MATERIALS USED : 2.1.1. CARBIDE TOOL (TIP): Carbide tips type P10 with a chemical composition of (65% W, 14 % Ti, 9 % Co and 12 % C) were used in this investigation. The tip has a tool angle of 60 o and a nose radius of 1.6 mm. Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 269 2..1.2. WORKPIECE MATERIAL : The machining experiments were performed using cylindrical work-piece of AISI 1020 steel with a chemical composition demonstrated in Table (1). The tests were carried out for a length of (210 mm) and (40 mm) diameter. This kind of steel is the most common form of steel because it’s price is somewhat low while it offers material characteristic that are tolerable for many applications. 2.2. DESCRIPTION OF TURNING PROCESS: In this work, external longitudinal turning operation was used for machining experiments. The 210 mm length and 40 mm diameter low carbon steel bar was divided into working regions of 50mm length as shown in Fig.(1). For all machining experiments, dry cutting condition was used. Four spindle speed (80, 315, 500, and 800 rpm) for each of which two feed rates 0.2 and 0.5 mm/rev and two values for the depth of cut (0.5 and 0.7 mm) were used. The inserts used were rigidly mounted on a tool holder type (TAK Holder). 2.3. TESTS ( MATERIAL TESTING) OR FLANK WEAR : A maximum width of 300 µm for the flank wear was considered as a criterion for the tool life. The width of the flank wear was measured every one minute of machining. Optical microscope type (Electron Eyepiece, model YJEYE01, resolution of 1280(H)*1024(V), China) integrated with CCD camera was used to capture the image of worn tools and to measure the flank wear width. 3. RESULTS AND DISCUSSION: Table (2) demonstrate the results of the machining experiments according to the used machining program. Figure (2) represent examples for the method used in estimating the tool life. These figures shows the typical behavior of the flank wear width (VB) with the machining time. An accelerated increase of VB can be noticed with the machining process at the first stage of cutting, then a constant rate of this increase at the second stage, after which a very rapid increase at the third stage that will continue till premature failure of the tool. The figures also indicate that a higher cutting speed causes a higher value of VB. This is due to the increase in cutting temperature accompanying with the increase in cutting speed and this causes increase in adhesion wear on tool cutting edge, also the increase in temperature may soften a very thin surface layer of the tool cutting edge. In addition to that a higher cutting speed means a higher repeated contact between the machined surface and the flank surface which increases the scratching action of machined material (i.e. the abrasive wear). Figure (3) shows examples of the wear occurred at the flank surface of the tool during. Both fundamental and empirical approaches can be used to establish models or equations for quantitatively predicting the cutting performance such as tool life and surface roughness. In the case of the empirical approach, the machining characteristic values which experimentally measured (tool life) are related to the cutting conditions by regression analysis. Several forms of mathematical models, as linear, exponential, power function, and non- linear quadratic polynomial were used to exam the approach. .4. DEVELOPMENT OF MATHEMATICAL PREDICTIVE MODEL: Mathematical predictive models for the tool life had been developed in forms shown in Table (3). Such models were constructed basing on the statistical data of the carried out machining experiments. Data fit ver.9 software was used for constructing these models basing on multiple Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 270 regression analyses. The value of the coefficient of multiple determination (R 2 ) for the non-linear quadratic polynomial is highest rank among all forms of other models. Hence, this model is selected to predict value of tool life. So the final mathematical predictive model developed will be: T=2.3931+5.5524X1+13.4310X2+13.3856X3+8.6831X1 2 +10.9889X2 2 +6.6217X3 2 +7.2496X1X2 +10.3714X2X3 +4.0852X1X3+8.7428X1X2X3 (1) Where: T = Tool life (min.); X1= Spindle speed (rpm); X2= Feed rate (mm/rev.), and X3= depth of cut (mm). Fig.(4) shows the matching between the experimental values of the tool life and their predicted values due to the designed models. The figure indicates their close matching. 5.TAGUCHI APPROACH : 5.1. Selection of control factors and noise factors: Cutting speed, feed rate, and depth of cut have a significant influence on tool life in the turning operations, so they were considered as the controllable factors in this study, while tool life as response factor as shown in Table (4). Design of experiments via Taguchi method and L16 (3×4) mixed orthogonal array is utilized for the parametric design [7,8,9]. To obtain optimal cutting conditions, the larger-the-better is used as S/N ratio of quality characteristic. This expressed as [10]: Where: η = signal to noise ratio (S/N), yi = observed value of the response, n = number of observation in a trial. Table (5) shows the experimental results and the corresponding S/N for tool life. Table (6) shows the response values for signal- to-noise ratios and the response values formeans for tool life. Figure (5) shows the results of Taguchi design analysis (main effect plot for Means and main effect plots for S/N ratio) for tool life. Also, Taguchi orthogonal array is demonstrated in Table (5). Figure (5) indicate that the optimal machining conditions representing the conditions at which the tool life will be maximum are: spindle speed, Vc =315 rpm, feed rate, f= 0.4 mm/rev, and depth of cut, t= 0.6 mm. 5.2. Analysis of Variance ANOVA: Determining which machining parameter significantly affect the quality characteristic (tool life) is the main purpose of ANOVA. This can be done by separating the total mean S/N ratio from the total variability of the S/N ratios which is determined by sum of square deviations. First, the total mean S/N ratio ηm from the total sum of squared deviations SST can be calculated as : i= 1, 2, … n (2) Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 271 ∑ ( ) (3) Where ηi is the mean S/N ratio for the i th experiment and n is the number of experiment in the orthogonal array. Table (7) shows the ANOVA (with 95% confidence level) and F-test values with the P- value which reflects effectiveness of the individual studied parameters on the tool life. The table indicates that the feed rate is the most significant parameter for maximum tool life. 5.3. Predicting optimum performance : The optimum tool life can be predicted by using the designed model (eq. 2) based on the determined optimum cutting conditions (i.e. at Vc = 315 rpm, f = 0.4 mm/rev., and t = 0.6 mm) at which the tool life equals to 7.8 min. 5.4. Confirmation of The Optimum Results: The purpose of the confirmation experiment in this study was to validate the optimum cutting conditions. The predicted tool life was calculated basing on the optimum machining conditions resulted by Taguchi analysis. Also basing on these machining conditions, experimental values for tool life were measured. Table (8) shows the comparison of the actual tool life with the predicted tool life using optimal cutting parameters for the used inserts with an error percentage of 6.4 % for tool life. 6. CONCLUSIONS: In this study, the Taguchi method was used to predict the most turning conditions of the AISI 1020 steel affecting the cemented carbide tool life, according to the results obtained, the following points can be concluded: 1-The Taguchi method is a robust orthogonal array design method, suitable for analysis of the tool machining parameters have an effect on the tool life. Among these parameters feed rate has the largest effect on the tool life. 2- Multiple regression analysis can be applied to develop a mathematical model for tool life prediction. The models developed are found to be reliable. An assessment of turning of AISI 1020 steel can be achieved by the designed empirical model for selecting the appropriate machining conditions for a required carbide tool life. 3- An optimum machining combination of AISI 1020 steel for maximum tool life is spindle speed of 315 rpm, feed rate of 0.4 mm/rev., and depth of cut of 0.6 mm. 7. REFERENCES: [1] M. Narasimha, D. Tewodros, R. Rejikumar, Improving Wear Resistance of Cutting Tool By Coating, IOSR Journal Of Engineering (IOSRJEN) , Vol. 04, Issue 05 (May. 2014), PP (06-14). [2] B. Fnides, S. Boutabba, M. Fnides, H. Aouici, M. A. Yallese, Cutting Tools Flank Wear and Productivity Investigation In Straight Turning of X38crmov5-1 (50 HRc), International Journal of Applied Engineering and Technology, 2013, Vol. 3 (1) January-March, PP (1-10). Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 272 [3] Bin Li, A Review Of Tool Wear Estimation Using Theoretical Analysis and Numerical Simulation Technologies, Int. Journal of Refractory Metals and Hard Materials 35, 2012, PP (143– 151). [4] Abdul Kareem Jaleel, Kareem Abdulla Hadi, Coated Carbide Cutting Tools Performance In High Speed Machining Processes, The Iraqi Journal For Mechanical and Materials Engineering, Vol.12, No.1, 2012. [5] Bala Murugan Gopalsamy, BiswanathMondal and Sukamal Ghosh, Taguchi Method and ANOVA: An Approach For Process Parameter Optimization of Hard Machining While Machining Hardened Steel, Journal of Scientific and Industrial Research Vol. 68, August 2009, PP(686-695). [6] S. R. Das1, R.P. Nayak, D. Dhupal, A. Kumar, Surface Roughness, Machining Force and Flank Wear in Turning of Hardened AISI 4340 Steel With Coated Carbide Insert: Cutting Parameters Effects, International Journal of Automotive Engineering Vol. 4, Number 3, Sept 2014. [7] Rajesh Nayak, Raviraj Shetty, Sawan Shetty, Investigation of Cutting Force In Elastomer Machining Using Taguchi’s Design of Experiments”, International Journal of Advanced Technology & Engineering Research (IJATER), ,Volume 4, Issue 4, July 2014. [8] Roy Ranjit K, Design of Experiments Using The Taguchi Approach : 16 Steps To Product and Process Improvement, John Wiley & Sons, Inc. (US),2001. [9] Haydar A. H. Al-Ethari, Kadhim Finteel Alsultani, Nesreen Dakhel F., Optimization of Chemical Machining Conditions of Cold Worked Stainless Steel 420 Using Taguchi Method, International Journal Of Mechanical Engineering and Technology (IJMET), Volume 5, Issue 3, March (2014), PP (57-65). [10] Genichi Taguchi, Subir Chowdhury, Yuin Wu,Taguchi’s Quality Engineering Handbook, 2005 John Wiley & Sons, Inc., Hoboken, New Jersey. Table (1): Chemical composition of the workpiece material (wt %). C Si Mn P S Cr Mo Ni Al Co Cu Fe 0.15 6 0.004 8 0.38 6 0.018 1 0.020 0 0.024 8 0.005 1 0.017 6 0.007 7 0.015 7 0.01 39 99.3 (Balance) Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 273 Table (2) : Results of machining experiment for the used tool. Table (3) : Predictive models for the tool life. No.No. Vc (rpm) f (mm/rev) t (mm) T (min) S/N ratio 1 1 1 1 3.6 11.1261 2 1 2 2 4 12.0412 3 1 3 3 3.8 17.9525 4 1 4 4 3.8 11.5957 5 2 1 2 3 9.5424 6 2 2 1 3.9 11.8213 7 2 3 4 2.7 8.6273 8 2 4 3 2.9 11.1261 9 3 1 3 2.4 12.0412 10 3 2 4 2.8 15.563 11 3 3 1 2.2 19.0849 12 3 4 2 2.2 13.9794 13 4 1 4 1.9 13.0643 14 4 2 3 2.1 13.9794 15 4 3 2 1.4 13.2552 16 4 4 1 1.5 18.0618 Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 274 Table (4) :Parameters and Levels for Experimental Design. Parameter Symbol Unit Level 1 Level 2 Level 3 Level 4 Spindle speed X1 rpm 80 315 500 800 Feed rate X2 mm/rev 0.2 0.3 0.4 0.5 Depth of cut X3 mm 0.5 0.6 0.7 0.8 Table (5): Experimental results and the corresponding S/N for tool life. Table (6) : Response values for signal- to-noise ratios and the response values for means. Response values for Signal to Noise Ratios (Larger is better) Response Values for Means Level Vc F t Level Vc F t 1 13.90 11.57 11.72 1 5.325 4.038 4.506 2 14.65 15.49 15.63 2 5.831 6.300 6.288 3 14.01 16.39 12.98 3 5.613 6.900 5.138 4 12.58 11.69 14.81 4 4.919 4.450 5.756 Delta 2.07 4.82 3.92 Delta 0.912 2.863 1.781 Rank 3 1 2 Rank 3 1 2 Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 275 Table (7) : One way ANOVA results. T versus Vc T versus f T versus t Source DF SS MS F P Source DF SS MS F P Source DF SS MS F P Vc 3 7.46 2.49 0.46 0.709 f 3 93.07 31.02 7.88 0.00 t 3 28.49 9.50 1.89 0.140 Error 12 321.89 5.36 Error 12 236.28 3.94 Error 12 300.86 5.01 Total 15 329.35 Total 15 329.35 Total 15 329.35 *DF – Degree of freedom, SS : Sum of Squares, Ms : Mean Square, P: probability. Table (8) : Confirmation test results for the used insert. Figure (1) : Workpiece regions. Performance Predicted Experimental Error% Tool life (min) 7.8 7.3 6.4 Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 276 Figure(2): Estimation of the tool life. 0 100 200 300 400 500 0 1 2 3 4 5 w id th o f fl a n k w e a r ,V B (m ic ro ) cutting time ,(min). Feed Rate =0.2 mm/rev DOC = 0.5 mm 80 rpm 315 rpm 500 rpm 800 rpm 0 100 200 300 400 500 0 1 2 3 4 5 w id th o f fl a n k w e a r, V B (m ic ro ) Cutting Time,(min) Feed Rate =0.5 mm/rev DOC = 0.5 mm 800 rpm 500 rpm 315 rpm 80 rpm 0 100 200 300 400 500 0 1 2 3 4 5 w id th o f fl a n k w e a r ,V B (m ic ro ) cutting time ,(min). Feed Rate =0.2 mm/rev DOC = 0.7 mm 800 rpm 500 rpm 315 rpm 80 rpm VB max 0 100 200 300 400 500 0 1 2 3 4 5 w id th o f F la n k w e a r, V B (m ic ro ) Cutting Time ,(min) Feed Rate = 0.5 mm/rev DOC = 0.7 mm 80 rpm 315 rpm 500 rpm 800 rpm Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 277 Figure (3) : The flank wear of the tools during the machining process. Spindle speed= 80 rpm Spindle speed= 315 rpm Spindle speed= 315 rpm Spindle speed= 315 rpm Spindle speed= 315 rpm Spindle speed= 500 rpm Spindle speed= 500 rpm Spindle speed= 500 rpm Spindle speed= 500 rpm Spindle speed= 800 rpm Spindle speed= 800 rpm Spindle speed= 800 rpm Spindle speed= 800 rpm Spindle speed= 80 rpm Spindle speed= 80 rpm Spindle speed= 80 rpm Al-Qadisiyah Journal For Engineering Sciences, Vol. 9……No. 2 ….2016 278 Fig.(4): Scatter plot of the experimental and predicted values of tool life. Fig.(5): Results of Taguchi design analysis: (a) main effect plot for Means; (b)main effect plots for S/N ratio for tool life. 0 1 2 3 4 5 6 0 1 2 3 4 5 6 P re d ic te d t o o l li fe ( m in ) Measured tool life (min) Tool Material : P 10 Workpiece : AISI 1020 Cutting Parameters : Vc=(80, 315, 500 ,800 rpm), Feed rate (0.2 and 0.5 mm/rev) and depth of cut (0.5 and 0.7 mm).