3 (1), 2009, 140 - 148
©BEYKENT UNIVERSITY
PREDICTION AND OPTIMIZATION OF
SURFACE ROUGHNESS IN BALL BURNISHING
PROCESS USING RESPONSE SURFACE
METHODOLOGY
Aysun SAĞBAŞ, Funda KAHRAMAN
Mersin University, Tarsus Technical Education Faculty, Department of Mechanical Education, 33400 Tarsus-Mersin/TURKEY
e-mail: asagbas@mersin.edu.tr e-mail: fkahraman@mersin.edu.tr
Received: 09.12.2008, Revised: 11.02.2009 Accepted: 13.02.2009
ABSTRACT
In this study, a regression model was developed to predict surface roughness in burnishing process of 7178 aluminum alloy by using Response Surface Methodology (RSM). In the development of predictive models, burnishing force, number of passes, feed rate and burnishing speed were considered as model variables, surface roughness were considered as a response variable. The statistical analysis showed that, burnishing force and number of passes have the most significant effect on surface roughness. After building the regression model, a numerical optimization technique using desirability functions was carried out to minimize the surface roughness. The lowest value of surface roughness (0.469 //m) is obtained at a burnishing force of 10 N, a feed rate of 0.30 mm/rev and a burnishing speed of 80 m/min with two passes of the burnishing tool.
Keywords: Burnishing process, Surface roughness, Optimization, RSM
ÖZET
Bu çalışmada, galetaj işlemi uygulanmış 7178 aluminyum alaşımının yüzey pürüzlülüğünü tahmin etmek için Tepki Yüzeyi Tasarımı yöntemi kullanılarak bir regresyon modeli geliştirilmiştir. Geliştirilen tahmin modelinde, kuvvet, paso sayısı, ilerleme miktarı ve hız model değişkeni olarak, yüzey pürüzlülüğü ise çıktı değişkeni olarak seçilmiştir. İstatistiksel analizler sonucu yüzey pürüzlülüğünü etkileyen en önemli parametrelerin kuvvet ve paso sayısı olduğunu belirlenmiştir. Regresyon modeli kurulduktan sonra, yüzey pürüzlülüğünü minimize etmek için, istenilirlik fonksiyonları kullanılarak bir sayısal optimizasyon tekniği uygulanmıştır. En düşük yüzey pürüzlülüğü
Anahtar Kelimeler: Galetaj işlemi, Yüzey pürüzlülüğü, Optimizasyon, Tepki
Yüzeyi Metodu
1. INTRODUCTION
In today's manufacturing industry, special attention is given to dimensional accuracy and surface finish. Thus, measuring and characterizing the surface roughness can be considered as the predictor of the machining performance. Burnishing is a process that leads to an accurate change in the surface roughness of the workpiece by a minor amount of plastic deformation. In burnishing, the metal on the surface of the workpiece is redistributed without material loss. Besides producing a good surface finish, the burnishing process has additional advantages such as securing increased hardness, corrosion resistance and fatique life as result of the produced compressive residual stress
[1-3].
A literature survey shows that, work on burnishing has been conducted by many researchers. Esme et al. [4] developed an artificial neural network model for the prediction surface roughness of AA 7075 aluminum alloy in ball burnishing process.. See mikery et al. [5] focused on the surface roughness, microhardness, surface integrity and fatique life aspects of AISI 1045 work material using full factorial design of experiments. Hassan et al. [6] investigated the effect of the measure parameters (burnishing force and number of passes) on the surface roughness using RSM. They established a mathematical model to correlate burnishing force and number of passes with surface finish. El-Axir et al. [7] studied on the surface finishing of 2014 aluminum alloy by ball burnishing process. The experiments were designed on the basis of RSM with Central Composite Design (CCD). Khabeery and El-Axir [8] examined the use of the roller burnishing process to improve surface integrity for 6061 aluminum alloy using a vertical milling machine. To explore the optimum combination of burnishing parameters, the experiments were designed based on RSM with CCD. Loh et al. [9] investigated the effects of various parameters on the surface roughness of aluminum alloy. They discussed optimum burnishing parameters and conditions. El-Axir et al. [10] studied the use of the internall ball burnishing process to improve surface characteristics for 2014 aluminum alloy using CNC lathe. They developed response models using RSM.
This paper examines the use of the ball burnishing process to minimize the surface roughness for 7178 aluminum alloy. The experiments designed on the basis of RSM with CCD. The optimum process conditions were determined by using desirability functions.
Response Surface Methodology
2. EXPERIMENTAL DETAILS
In this study, 7178 aluminum alloy was used as workpiece material. The workpieces were received in the form of bars diameter of 30 mm. These bars were cut to 100 mm in length. Stainless steel ball (HRc 62) was used for
burnishing process. A ball diameter of 18 mm was implemented. The experiments were performed on an industrial type CNC lathe. No coolant was used during burnishing. A schematic presentation of the burnishing process is shown in Figure 1.
•^1'lTllUt
Figure 1. Experimental set up of the burnishing process
The empirical model using RSM with CCD was developed to investigate the influences of four burnishing parameters, including burnishing force, number of passes, feed rate and burnishing speed on the performance characteristics of surface roughness. The range of each parameter was coded in five levels. The burnishing conditions are shown in Table 1. The optimum burnishing parameter combinations were obtained by using a numerical optimization technique.
Table 1. Factors and levels for CCD
Factors/Levels Lowest Lower Center Higher Highest
Burnishing force (N) 10 90 170 250 330
Number of passes 1 2 3 4 5
Feed rate (mm/rev) 0.10 0.30 0.45 0.60 0.80
Burnishing speed (m/min) 18 54 72 90 126
3. RESULTS AND DISCUSSION
The arrangement and the results of the 30 experiments carried in this work based on the rotatable CCD are shown in Table 2. Regression analysis indicates quadratic model
Ra = 0.57 + 0.28x1 + 0.14x2 + 0.10x12 + 0.15x22 + 0.022x32 + 0.032x42 + 0.049x1x2 (1)
where; xx : burnishing force; x2: number of passes;
burnishing speed
Table 2. CCD design matrix and experimental results
x3: feed rate; x4: Exp. No Burnishing force (N) Number of passes Feed rate (mm/rev) Speed (m/min) Ra (/am) 1 90 2 0.30 54 0.55 2 250 2 0.30 54 1.09 3 90 4 0.30 54 0.69 4 250 4 0.30 54 1.39 5 90 2 0.60 54 0.53 6 250 2 0.60 54 1.00 7 90 4 0.60 54 0.69 8 250 4 0.60 54 1.32 9 90 2 0.30 90 0.48 10 250 2 0.30 90 0.93 11 90 4 0.30 90 0.65 12 250 4 0.30 90 1.42 13 90 2 0.60 90 0.52 14 250 2 0.60 90 1.02 15 90 4 0.60 90 0.68 16 250 4 0.60 90 1.32 17 10 3 0.45 72 0.44 18 330 3 0.45 72 1.48 19 170 1 0.45 72 0.80 20 170 5 0.45 72 1.45 21 170 3 0.10 72 0.65 22 170 3 0.80 72 0.61 23 170 3 0.45 18 0.60 24 170 3 0.45 126 0.69 25 170 3 0.45 72 0.57 26 170 3 0.45 72 0.57 27 170 3 0.45 72 0.51 28 170 3 0.45 72 0.58 29 170 3 0.45 72 0.56
Response Surface Methodology
30 170 3 045 72 0.61 An ANOVA table is commonly used to summarize the tests performed. It was statistically studied the relative effect of each burnishing parameters on the surface roughness by using ANOVA. The ANOVA table for response surface quadratic model for the surface roughness is given in Table 3.
Table 3. The ANOVA table for the quadratic model
Source Sum of Square Degrees of Freedom Mean Square F-value p- value Model 3.21 14 0.23 89.10 < 0.0001 X1 1.92 1 1.92 743.99 < 0.0001* X
2
0.46 1 0.46 180.55 < 0.0001* X3
1.667E-003 1 1.667E-003 0.65 0.4336 X4
1.067E-003 1 1.067E-003 0.41 0.5295x2
0.30 1 0.30 116.07 < 0.0001* X2 0.58 1 0.58 225.94 < 0.0001* X3
0.013 1 0.013 5.10 0.0393* X42 0.028 1 0.028 10.82 0.0050*X!
X2 0.038 1 0.038 14.77 0.0016*X!
X3 3.025E-003 1 3.025E-003 1.18 0.2955 X1 X4 2.500E-005 1 9.711E-003 0.62 0.9228 X X r-o Co 1.600E-003 1 2.500E-005 0.97 0.4428 X2 X4 2.500E-003 1 1.600E-003 1.40 0.3400X3X4 3.600E-003 1 2.500E-003 1 0.2554
Residual 0.039 15 2.574E-003
Total 3.25 29
* significant factors at 5% significance level
The F statistic is a statistic for a test concerning the differences among means. The F statistic is typically used for comparing two population variance. In statistical hypothesis testing, the p-value is the probability of obtaining a result at least as extreme as the one that was actually observed, given that the null
value showed that, the quadratic model fits well to the experimental results. Higher F value indicates that the variation of the process parameter makes a big changes on the surface roughness. As seen in Table 3, burnishing force and number of passes is the most significant parameters, feed rate and burnishing speed is the least significant parameters and so they are absent in regression model. All the squared terms and among interaction term burnishing force and number of passes appears to be highly significant. The relationship of the burnishing force and the number of tool passes on the surface roughness is shown in Figure 2. It can be realized that the combination between low number of passes and low burnishing force results in a considerable reduction in burnished surface roughness.
•2J00 -2JQ0
Figure 2. Three dimensional plot of surface roughness
We consider a measure of the model's overall performance referred to as the coefficient of determination and denoted by R2. In the model, R2 is obtained
equal to 0.98. The R2 value is high, close to 1, which is desirable. The
comparison of experimental surface roughness results with the model surface roughness predictions is presented in Figure 3. As seen in Figure 3, the fits between actual surface roughness and predicted surface roughness values are very good.
Response Surface Methodology
A c t u a l
Figure 3. RSM predictions against the experimental results
The comparisons of experimental results with the RSM predictions have been depicted in terms of percentage absolute error for validation set of experiments. In the prediction of surface roughness values the average absolute errors for RSM is found to be as 3.5 %. After building the regression model, a numerical optimization technique using desirability functions can be used to optimize the responses. The objective of optimization is to find the best settings that minimize a particular response. [11]. Table 4 shows four alternative solutions of the optimization approach used to determine the optimum processing conditions.
Solutions Burnishing Number force (N) of passes Feed rate (mm/rev) Burnishing speed (m/min) Ra (
/m
) Desirability 1 10 2 0.33 81 Q.471 Q.92Q 2 10 2 Q.3Q 8Q Q.469 Q.92Q 3 10 2 Q.22 95 Q.481 Q.91Q 4 10 2 0.37 102 0.511 0.901The achieved maximum desirability of 0.92 means that it is possible to meet surface roughness value. The lowest value of surface roughness obtained is Ra (0.469 /Lim) at a burnishing force of 10 N, a feed rate of 0.30 mm/rev and a burnishing speed of 80 m/min with two passes of the burnishing tool.
5. CONCLUSION
In this experimental study, RSM-based approach has been suggested and successfully implemented. It was used to develop an empirical model with burnishing force, number of passes, feed rate and burnishing speed on the surface roughness of the 7178 aluminum alloy. The predictive power of this model was tested with supplementary experimental surface roughness data and a good fit was observed. The average absolute error between experimental and predicted values was calculated as 3.5 % sufficiently low to confirm the high predictive power of model. It was observed that significance of the main variables is burnishing force and number of passes. RSM approach can help manufacturers to determine the appropriate burnishing conditions, in order to achieve specific surface roughness.
Response Surface Methodology REFERENCES
[1] Luo, H., Wang, L., and Zhang, C.; Study on the aluminum alloy burnishing processing and the existence of the outstripping phenomenon. Journal of Materials Processing Technology 116 (2001), 88-90.
[2] Hassan, A.M., and Al-Bsharat, A.S.; Improvement in some properties of non-ferrous metals by the application of ball burnishing process. Journal of Materials Processing Technology 59 (1996), 250-256.
[3] Luca, L., and Ventzel-Marinescu, I.; Effects of working parameters on surface finish in ball-burnishing of hardened steels. Precision Eng. 29 (2005), 253-256.
[4] Esme, U., Sagbas, A., Kahraman, F., Kulekci, M.K.; Use of artificial neural networks in ball burnishing process for the prediction of surface roughness of AA 7075 aluminum alloy. Materiali in Tehnologije 42 (2008), 215-219.
[5] Seemikery, C.Y., Brahmankar, P.K., and Mahagaonkar, S.B.; Investigation on surface integrity of AISI 1045 using LPB tool. Tribology International 41 (2008), 724-734.
[6] Hassan, A.M., Al-Jalil, H.F., and Ebied, A.A.; Burnishing force and number of passes for the optimum surface finish of brass components. Journal of Materials Processing Technology 83 (1998), 176-179.
[7] El-Axir, M.H., Othman, O.M., and Abodiena, A.M.; Study on the inner surface finishing of aluminum alloy 2014 by ball burnishing process. Journal of Materials Processing Technology 202 (2008), 435-442.
[8] El-Khabeery, M.M, and El-Axir, M.H.; Experimental techniques for studying the effects of milling roller burnishing parameters on surface integrity. International Journal of Machine Tools & Manufacturing 41 (2001), 1705-1719.
[9] Loh, N.H., Tam, S.C., Miyazawa, S.; A study of the effects of ball burnishing parameters on surface roughness using factorial design. Journal of Mechanical Working Technology 18 (1989), 53-61.
[10] El-Axir, M.H., Othman, O.M., Abodiena, A.M.; Improvements in out-of-roundness and microhardness of inner surfaces by internal ball burnishing. Journal of Materials Processing Technology 196 (2008), 120-128.
[11] Myers, R.H. and Montgomery, D.C.; Response surface methodology: process and product optimization using designed experiments, John Wiley & Sons Inc., NY (2002).
