Published on

IJERA (International journal of Engineering Research and Applications) is International online, ... peer reviewed journal. For more detail or submit your article, please visit

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide


  1. 1. Upinder Kumar Yadav, Deepak Narang, Pankaj Sharma Attri / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue4, July-August 2012, pp.2060-2065 Experimental Investigation And Optimization Of Machining Parameters For Surface Roughness In CNC Turning By Taguchi Method Upinder Kumar Yadav*, Deepak Narang**, Pankaj Sharma Attri*** *(Department of Mechanical Engg., M.M. University, Mullana (Haryana) ** (Department of Mechanical Engg., P.C.E.T., Lalru (Punjab) ***(Department of Mechanical Engg., M.M. University, Mullana (Haryana)Abstract Medium Carbon Steel AISI 1045 has a conditions include speed, feed and depth of cut andwide variety of applications in vehicle component also tool variables include tool material, nose radius,parts & machine building industry. Surface rake angle, cutting edge geometry, tool vibration, toolroughness is the main quality function in high overhang, tool point angle etc. and work piecespeed turning of medium carbon steel in dry variable include hardness of material and mechanicalconditions. In this study, the effect and properties. It is very difficult to take all theoptimization of machining parameters (cutting parameters that control the surface roughness for aspeed, feed rate and depth of cut) on surface particular manufacturing process. In a turningroughness is investigated. An L’27 orthogonal operation, it is very difficult to select the cuttingarray, analysis of variance (ANOVA) and the parameters to achieve the high surface finish. Thissignal-to-noise (S/N) ratio are used in this study. study would help the operator to select the cuttingThree levels of machining parameters are used parameters.and experiments are done on STALLION-100 HS The work material used for the present studyCNC lathe. The optimum value of the surface is AISI 1045 medium carbon steel of composition {roughness (Ra) comes out to be 0.89. It is also Carbon (0.43-0.50)%, Silicon (0.2-0.3)%,concluded that feed rate is the most significant Magnesium (0.60-0.90)%, Phosphorus (0.05)%factor affecting surface roughness followed by Sulphur (0.05)% }. Its tensile strength is (620- 850)depth of cut. Cutting speed is the least significant Mpa. This Carbon steel is suitable for shafts andfactor affecting surface roughness. Optimum machinery parts. It is mostly used in Automobileresults are finally verified with the help of parts, in gears and machine building industry.confirmation experiments. This paper is about experimentally investigating and optimizing the machiningKeywords: ANOVA, CNC Turning, parameters for surface roughness in cnc turning byOptimization, Surface Roughness, Taguchi taguchi method. Taguchi’s orthogonal arrays areMethod. highly fractional designs, used to estimate main effects using few experimental runs only. These1. Introduction designs are not only applicable for two level factorial Quality plays a major role in today’s experiments, but also can investigate main effectsmanufacturing market. From Customer’s viewpoint when factors have more than two levels. Designs arequality is very important because the quality of also available to investigate main effects for someproduct affects the degree of satisfaction of the mixed level experiments where the factors includedconsumer during usage of the product. It also do not have the same number of levels. For example,improves the goodwill of the company. High speed a four-level full factorial design with five factorsturning is a machining operation which is done on requires 1024 runs while the Taguchi orthogonalcnc lathe. The quality of the surface plays a very array reduces the required number of runs to 16 only.important role in the performance of dry turning David et al. (2006) described an approach forbecause a good quality turned surface surely predicting Surface roughness in a high speed end-improves fatigue strength, corrosion resistance and milling process and used artificial neural networkscreep life. Surface roughness also affects on some (ANN) and statistical tools to develop differentfunctional attributes of parts, such as, contact causing surface roughness predictors.surface friction, wearing, light reflection, ability ofdistributing and also holding a lubricant, load bearing Srikanth and Kamala (2008) proposed a realcapacity, coating and resisting fatigue. As we know coded genetic algorithm (RCGA) for findingin actual machining, there are many factors which optimum cutting parameters and explained variousaffect the surface roughness i.e. cutting conditions, issues of RCGA and its advantages over the existingtool variables and work piece variables. Cutting approach of binary coded genetic algorithm (BCGA). 2060 | P a g e
  2. 2. Upinder Kumar Yadav, Deepak Narang, Pankaj Sharma Attri / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue4, July-August 2012, pp.2060-2065According to Roy, R. K. (2001)., the very intention 3. Design of experimentof Taguchi Parameter Design is for maximizing the In this study, three machining parametersperformance of a naturally variable production were selected as control factors, and each parameterprocess by modifying the controlled factors. was designed to have three levels, denoted 1, 2, and 3 (Table 1). The experimental design was according to Experiments were designed using Taguchi an L’27 array based on Taguchi method, while usingmethod so that effect of all the parameters could be the Taguchi orthogonal array would markedly reducestudied with minimum possible number of the number of experiments. A set of experimentsexperiments. Using Taguchi method, Appropriate designed using the Taguchi method was conducted toOrthogonal Array [7, 8] has been chosen and investigate the relation between the processexperiments have been performed as per the set of parameters and response factor. Minitab 16 softwareexperiments designed in the orthogonal array. Signal is used to optimization and graphical analysis ofto Noise ratios are also calculated for analyzing the obtained dataeffect of parameters more accurately. Results of the Table 1 Turning parameters and levelsexperimentation were analyzed analytically and also Symbol Turning Level Level Levelgraphically using ANOVA. ANOVA used to parameters 1 2 3determine the percentage contribution of all factors V Cutting 175 220 264upon each response individually. speed(m/min) F Feed 0.1 0.2 0.32. Taguchi method rate(rev/min) Traditional experimental design methods are D Depth of 0.5 1.0 1.5very complicated and difficult to use. Additionally, cut(mm)these methods require a large number of experimentswhen the number of process parameters increases 4. Experimental details[16, 17, 18]. In order to minimize the number of tests Medium Carbon Steel (AISI 1045) of Ø: 28required, Taguchi experimental design method, a mm, length: 17 mm were used for the turningpowerful tool for designing high-quality system, was experiments in the present study. The chemicaldeveloped by Taguchi. This method uses a design of composition, mechanical and physical properties oforthogonal arrays to study the entire parameter space AISI 1045 can be seen in Tables 2 and 3,with small number of experiments only. Taguchi respectively. The turning tests were carried out torecommends analyzing the mean response for each determine the surface roughness under variousrun in the inner array, and he also suggests to analyze turning parameters. STALLION-100 HS CNC lathevariation using an appropriately chosen signal-to- used for experimental investigations.noise ratio (S/N). There are 3 Signal-to-Noise ratiosof common interest for optimization of static Table 2 Chemical composition of Medium Carbonproblems: Steel (AISI 1045) Element Percentage (I) SMALLER-THE-BETTER: 𝐒 ∑𝐘𝐢 𝟐 = −𝟏𝟎 𝐥𝐨𝐠 Carbon (0.43-0.50)% 𝐍 𝐧 Silicon (0.2-0.3)% (II) LARGER-THE-BETTER: 𝐒 𝟏 = −𝟏𝟎 𝐥𝐨𝐠⁡ ∑ )/𝐧) (( Magnesium (0.60-0.90)% 𝐍 𝐲𝐢² (III) NOMINAL-THE-BEST: Phosphorus (0.05)% max. 𝑺 𝒚𝟐 𝑵 = 𝟏𝟎 𝐥𝐨𝐠⁡ 𝒔 𝟐 ) ( Sulphur (0.05)% max. Here Yi is the ith observed value of theresponse, n is the no. of observations in a trial, y is Table 3 Mechanical and Physical properties ofthe average of observed values (responses) and s is Medium Carbon Steel (AISI 1045)the variance. Regardless of category of the Property Valueperformance characteristics, the higher S/N ratio Yield strength (Mpa) (300-450)corresponds to a better performance. Therefore, the Tensile strength (Mpa) 640optimal level of the process parameters is the level Hardness (HB) 170-210with the highest S/N value. The statistical analysis of Density(g/cm³) 7.85the data is performed by analysis of variance Poisson’s Ratio 0.27-0.30(ANOVA) [9, 10] to study the contribution of the Elongation 14% min.factor and interactions and to explore the effects of Modulus of elasticity 205each process on the observed value. (Mpa) 2061 | P a g e
  3. 3. Upinder Kumar Yadav, Deepak Narang, Pankaj Sharma Attri / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue4, July-August 2012, pp.2060-20655. Results and discussion Table 5 ANOVA Table for Surface roughness5.1 Experiment results and Taguchi analysis In high speed turning operation, surface Var DF SS MS F P Contrroughness is an important criterion. The purpose of iabl ibutiothe analysis of variance (ANOVA) is to investigate e nwhich design parameter significantly affects the Facsurface roughness. Based on the ANOVA, the torsrelative importance of the machining parameters with CS 2 0.496 0.248 28.8 0.00 2.71respect to surface roughness was investigated to 3 0 **determine the optimum combination of the F 2 17.42 8.713 1011 0.00 95.23machining parameters. 6 .42 0 *** A series of turning tests are conducted to D 2 0.177 0.088 10.2 0.00 0.96*assess the effect of turning parameters on surface 8 6roughness in turning of AISI 1045. Experimental CS 4 0.048 0.012 1.41 0.31 0.26results of the surface roughness for turning of AISI *F 31045 with different turning parameters are shown in CS 4 0.036 0.009 1.05 0.44 0.19Table 4. Table 4 also gives S/N ratio for surface *D 0roughness. The S/N ratio for each experiment of L’27 F * 4 0.043 0.010 1.26 0.36 0.23was calculated using smaller the better approach. The D 0objective of using the S/N ratio as a performance E 8 0.068 0.008 0.37measurement is developing products and process T 26 18.29insensitive to noise factor. 7Table 4 EXPERIMENTAL RESULT ANDCORRESPONDING S/N RATIO Where CS-Cutting Speed,Ex. CS ( F ( mm/ D ( SR, S/N F-Feed Rate,No. m/min) rev) mm) Ra(μm) Ratio D-Depth Of Cut,1 175 0.1 0.5 1.46 -3.28 SR-Surface Roughness,2 175 0.1 1.0 1.24 -1.86 E-Error,3 175 0.1 1.5 1.25 -1.93 T-Total,4 175 0.2 0.5 1.88 -5.48 DF-Degree of Freedom,5 175 0.2 1.0 2.20 -6.84 SS-Sum of Squares,6 175 0.2 1.5 1.91 -5.62 MS-Mean of Squares,7 175 0.3 0.5 3.35 -10.50 F-a statistical parameter,8 175 0.3 1.0 3.48 -10.83 P-Percentage9 175 0.3 1.5 3.15 -9.9610 220 0.1 0.5 1.17 -1.36 Here *** & ** represents most significant and11 220 0.1 1.0 1.09 -0.74 significant parameters and * as less significant.12 220 0.1 1.5 0.88 1.1113 220 0.2 0.5 1.93 -5.71 S/N Ratio of Surface Roughness is calculated by : 𝑆 ∑𝑌𝑖 214 220 0.2 1.0 1.83 -5.24 𝑁 = −10 log 𝑛15 220 0.2 1.5 1.76 -4.9116 220 0.3 0.5 3.14 -9.9317 220 0.3 1.0 3.00 -9.54 Here yi is the ith observed value of the response , n is18 220 0.3 1.5 3.01 -9.57 the no. of observations in a trial.19 264 0.1 0.5 1.23 -1.7920 264 0.1 1.0 1.02 -0.1721 264 0.1 1.5 0.95 0.4422 264 0.2 0.5 1.80 -5.1023 264 0.2 1.0 1.74 -4.8124 264 0.2 1.5 1.63 -4.2425 264 0.3 0.5 3.00 -9.5426 264 0.3 1.0 2.95 -9.3927 264 0.3 1.5 2.71 -8.65 2062 | P a g e
  4. 4. Upinder Kumar Yadav, Deepak Narang, Pankaj Sharma Attri / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue4, July-August 2012, pp.2060-2065 Residual Plots for SN ratios Normal Probability Plot Versus Fits Interaction Plot for SN ratios Data Means 99 0.50 0.1 0.2 0.3 90 0.25 0 cutting Residual Per cent speed 50 0.00 -5 175 cutting speed 220 -0.25 10 264 -10 -0.50 0 1 feed -0.50 -0.25 0.00 0.25 0.50 -10.0 -7.5 -5.0 -2.5 0.0 rate Residual Fitted Value -5 0.1 feed rate 0.2 Histogram Versus Order 0.3 -10 6.0 0.50 0 depth of cut 4.5 0.25 0.5Fr equency -5 Residual depth of cut 1.0 0.00 1.5 3.0 -10 -0.25 1.5 175 220 264 0.5 1.0 1.5 -0.50 Signal-to-noise: Smaller is better 0.0 -0.6 -0.4 -0.2 0.0 0.2 0.4 2 4 6 8 10 12 14 16 18 20 22 24 26 Residual Observation Order Fig 3: Interaction Plot for S/N Ratios of Surface RoughnessFig 1: Residual analysis of Surface Roughness of firstdegree polynomial equation From the ANOVA table it is clear that maximum contribution factor is feed rate having percentage contribution up to 95.23%. After that second main contribution is of cutting speed. Hence the individual Main Effects Plot for Means ranking of these three parameters on the average Data Means value of means of Surface roughness:- cutting speed feed rate 3.0 Table 6 Mean values of process parameters for material removal rate 2.5 Level Cutting speed Feed rate Depth of 2.0 (CS) (F) cut (D) 1 2.213 1.143 2.107 Mean of Means 1.5 2 1.979 1.853 2.061 1.0 175 220 264 0.1 0.2 0.3 3 1.892 3.088 1.917 depth of cut Rank 2 1 3 3.0 2.5 5.2 Predictive Equation and Verification The predicted value of Surface roughness at 2.0 the optimum levels [13, 14] is calculated by using the relation given as: 1.5 𝒐 1.0 ň = 𝒏𝒎 + (𝒏𝒊𝒎 − 𝒏𝒎) 0.5 1.0 1.5 𝒊=𝟏 Where ň is the predicted value of the surfaceFig 2: Effect of turning parameters on Material roughness after optimization, nm is the total meanremoval rate value of surface roughness for every parameter, nim 2063 | P a g e
  5. 5. Upinder Kumar Yadav, Deepak Narang, Pankaj Sharma Attri / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue4, July-August 2012, pp.2060-2065is the mean surface roughness at optimum level of 5.3.1 Surface Roughnesseach parameter and o is the number of main It has been found that feed rate is the mostmachining parameters that affects the response significant factor for surface roughness and itsparameter. contribution is 95.23% . The best results for Surface roughness would be achieved when AISI 1045By applying this relation, the predicted value of medium carbon steel work piece is machined atsurface roughness at optimum conditions is cutting speed of 264 m/min, depth of cut of 1.5 mm,obtained:- feed rate of 0.1 mm/rev. With 95% confidenceň Ra = 2.028 + [(1.892-2.028) + (1.143-2.028) + interval, feed rate affects the surface roughness most(1.917-2.028)] significantly.ň Ra = 0.89 μm 6. Conclusion The effectiveness of this parameter The current study was done to study theoptimization is verified experimentally. This requires effect of machining parameters on the surfacethe confirmation run at the predicted optimum roughness. The following conclusions are drawnconditions. The experiment is conducted at the from the study:predicted optimum conditions and the average of the 1. The Surface roughness is mainly affected by feedresponse is comes out to be 0.93 μm. The error in the rate and cutting speed. With the increase in feed ratepredicted and experimental value is only 4.4%, so the surface roughness also increases & as the cuttinggood agreement between the experimental and speed decreases the surface roughness increases.predicted value of response is obtained. Because the 2. From ANOVA analysis, parameters makingpercentage error is less than 5%, it confirms that the significant effect on surface roughness are feed rateresults have excellent reproducibility. The results and cutting that using the optimal parameter setting 3. The parameters taken in the experiments are(V3F1D3) the lower surface roughness is achieved. optimized to obtain the minimum surface roughnessTable 7 shown below shows that optimal values of possible. The optimum setting of cutting parameterssurface roughness lie between the optimal range. for high quality turned parts is as :- i) Cutting speed i.e. 264 m/min.Table 7 Optimal values of machining and response ii) Feed rate i.e. 0.1 mm/rev.parameters iii) Depth of cut should be 1.5 mm.CP OV OL POV EOV ORCS 264 V3- 0.89 0.93 0.89 F1- < Ra >F 0.1 D3 0.93D 1.5Where,CP-Cutting ParametersOV-Optimal Values of ParametersOL-Optimum Levels of ParametersPOV-Predicted Optimum valueEOV-Experimental Optimum ValueOR-Optimum Range of Surface Roughness5.3 Results The effect of three machining parameters i.e.Cutting speed, feed rate and depth of cut and theirinteractions are evaluated using ANOVA and withthe help of MINITAB 16 statistical software. Thepurpose of the ANOVA in this study is to identify theimportant turning parameters in prediction of Surfaceroughness. Some important results comes fromANOVA and plots are given below: 2064 | P a g e
  6. 6. Upinder Kumar Yadav, Deepak Narang, Pankaj Sharma Attri / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue4, July-August 2012, pp.2060-2065References Scientific and Industrial Research, vol. 68, [1] David V, Rubén M, Menéndez C, Rodríguez (2009), pp. 686-695. J, Alique R . Neural networks and statistical [13] Cayda U., “Machinability evaluation in hard based models for surface roughness turning of AISI 4340 steel with different prediction. International Association Of cutting tools using statistical techniques ”, Science and Technology for Development, Proceedings of the Institution of Mechanical Proceedings of the 25th IASTED Engineers, Part B: Journal of Engineering international conference on Modeling, Manufacture, (2010). indentification and control, (2006) pp. 326- [14] Thamizhmanii S., Saparudin S. and Hasan S, 331. “Analysis of Surface Roughness by Using [2] Srikanth T, Kamala V . A Real Coded Taguchi Method”, Achievements in Genetic Algorithm for Optimization of Materials and Manufacturing Engineering, Cutting Parameters in Turning. IJCSNS Int. (2007), Volume 20, Issue 1-2, pp. 503-505. J. Comput. Sci. Netw. Secur,, [16] Lan T.-S., Lo C. Y., Wang M.-Y. and Yen (2008),8(6):189-193. A-Y, “Multi Quality Prediction Model of [3] Roy, R. K. (2001). Design of experiments CNC Turning Using Back Propagation using the Taguchi approach: 16 steps to Network”, Information. Proceeding of product and process improvement. New American society of mechanical engineers, York: Wiley. (2008). [4] Arbizu IP, Pérez CJL (2003). Surface [17] Palanikumar K. and Karunamoorthy L., roughness prediction by factorial design of “Optimization of Machining Parameters in experiments in turning processes. J. Mater. Turning GFRP Composites Using a Carbide Process. Technol., 143-144: 390-396. (K10) Tool Based on the Taguchi Method [5] Tsao CC, Hocheng H (2008). Evaluation of with Fuzzy Logics”, metals and materials the thrust force and surface roughness in International, Vol. 12, No. 6 (2006), pp. drilling composite material using Taguchi 483-491. analysis and neural network. J. Mater. [18] Dhavamani C. and Alwarsamy T., “review Process. Technol., 203(1-3): 342-348. on optimization of machining”, international [6] Zhang JZ, Chen JC, Kirby ED . Surface journal of academic research, (2011), Vol. 3. roughness optimization in an end-milling No. 3. May, II Part. operation using the Taguchi design method. [19] Yang, W. H., & Tarng, Y. S. (1998). Design J. Mater. Process. Technol., (2007), 184(1- optimization of cutting parameters for 3): 233-239. turning operations based on the Taguchi [7] Munteanu, “Application of taguchi method method. Journal of Materials Processing for cutting force in turning high density Technology, 84, 122-129. polyethylene”, academic journal of [20] Lin, C. L. (2004). Use of the Taguchi manufacturing engineering, vol. 8, (2010), method and grey relational analysis to issue 3. optimize turning operations with multiple [8] Kolahan Farhad and Manoochehri Mohsen, performance characteristics. Materials and “Parameters and Tool Geometry Manufacturing Processes, 19(2), 209-220. Specifications in Turning Operation of AISI 1045 Steel”, World Academy of Science, Engineering and Technology (2011). [9] Selvaraj D. Philip and Chandarmohan P., “optimization of surface roughness of AISI 304 austenitic stainless steel in dry turning operation using Taguchi design method”, Journal of Engineering Science and Technology Vol. 5, No. 3 (2010) 293 - 301. [10] Kamarudin K. and Rahim E. A., “Optimizing Surface Roughness and Flank Wear on Hard Turning Process Using Taguchi Parameter Design”, Proceedings of the World Congress on Engineering July 2 - 4, (2007), London, U.K. [11] Gopalsamy B.M., Mondal B. and Ghosh S., “Taguchi method and ANOVA: An approach for process parameters optimization of hard machining while machining hardened steel”, Journal of 2065 | P a g e