为了正常的体验网站,请在浏览器设置里面开启Javascript功能!

灰色关联

2014-03-10 10页 pdf 537KB 31阅读

用户头像

is_915724

暂无简介

举报
灰色关联 Procedia Engineering 64 ( 2013 ) 868 – 877 Available online at www.sciencedirect.com 1877-7058 © 2013 The Authors. Published by Elsevier Ltd. Selection and peer-review under responsibility of the organizing and review committee of IConDM 2013 doi: 10.1016/j...
灰色关联
Procedia Engineering 64 ( 2013 ) 868 – 877 Available online at www.sciencedirect.com 1877-7058 © 2013 The Authors. Published by Elsevier Ltd. Selection and peer-review under responsibility of the organizing and review committee of IConDM 2013 doi: 10.1016/j.proeng.2013.09.163 ScienceDirect International Conference On DESIGN AND MANUFACTURING, IConDM 2013 Analysis of Process Parameters in Wire EDM with Stainless Steel using Single Objective Taguchi Method and Multi Objective Grey Relational Grade M. Durairaja, D. Sudharsunb,*, N. Swamynathanb aSr. Assistant Professor, Department of Mechanical Engineering, Tagore Engineering College, Chennai -127 bStudent, Department of Mechanical Engineering, Tagore Engineering College, Chennai - 127 Abstract With the increasing demands of high surface finish and machining of complex shape geometries, conventional machining process are now being replaced by non-traditional machining processes. Wire EDM is one of the non-traditional machining processes. Surface roughness and kerf width are of crucial importance in the field of machining processes. This paper summarizes the Grey relational theory and Taguchi optimization technique, in order to optimize the cutting parameters in Wire EDM for SS304. The objective of optimization is to attain the minimum kerf width and the best surface quality simultaneously and separately. In this present study stainless steel 304 is used as a work piece, brass wire of 0.25mm diameter used as a tool and di 16, orthogonal array has been used. The input parameters selected for optimization are gap voltage, wire feed, pulse on time, and pulse off time. Dielectric fluid pressure, wire speed, wire tension, resistance and cutting length are taken as fixed parameters. For each experiment surface roughness and kerf width was determined by using contact type surf coder and video measuring system respectively. By using multi objective optimization technique grey relational theory, the optimal value is obtained for surface roughness and kerf width and by using Taguchi optimization technique, optimized value is obtained separately. Additionally, the analysis of variance (ANOVA) is too useful to identify the most important factor. © 2013 The Authors. Published by Elsevier Ltd. Selection and peer-review under responsibility of the organizing and review committee of IConDM 2013. Keywords -16 orthogonal array; Surface Roughness; Kerf Width; Taguchi Optimization Method; Grey Relation Analysis. * D. Sudharsun. Tel.: +91-962-941-8418. E-mail address: sunsudharsun@gmail.com © 2013 The Authors. Published by Elsevier Ltd. Selection and peer-review under responsibility of the organizing and review committee of IConDM 2013 869 M. Durairaj et al. / Procedia Engineering 64 ( 2013 ) 868 – 877 1. Introduction WEDM is based on Electrical Discharge Machining Process, which is also called electro-erosion machining process. When the gap voltage is sufficiently large (i.e. reaches the breakdown voltage of dielectric fluid), high power spark is produced, which increase the temperature about 10,000 degrees Celsius. By this way the metal is removed from the work piece. Stainless Steel (S304) is used as a work piece. Stainless Steel 304 is a nickel and chromium based alloy, which is widely used in valves, refrigeration equipment, evaporators, cryogenic vessels due to their exceptional corrosion resistant, high ductility, non-magnetic and it retains solid phase up to 1400 degree Celsius. The chemical composition of the work material is shown in table 1. Table 1. Chemical Composition of Stainless Steel Chemical Composition Wt% Carbon (C) Manganese (Mn) Silicon (Si) Sulphur (S) phosphorous (P) Nickel (Ni) Chromium (Cr) Work Material 0.08 1.86 0.41 0.016 0.026 8.42 18.26 Range Up to 0.08 Up to 2.00 Up to 0.75 Up to 0.030 Up to 0.045 8.00 - 10.50 18.00 - 20.00 Brass wires are alloys of copper and zinc possesses reasonable conductivity with high tensile strength when compared to the copper wires. In this experiment, brass wire having 65% of copper and 35% of zinc is selected as a tool due to its properties, availability and low cost. The gap between the wire and work piece usually ranges from 0.025 to 0.075 mm and is constantly maintained by a computer controlled positioning system [1]. The selection of optimum machining parameters in WEDM is an important step [2, 3]. Taguchi Optimization technique is single parameter optimization based on the signal to noise ratio. Grey relational analysis is applied to optimize the parameters having multi-responses through grey relational grade. Furthermore, a statistical analysis of variance (ANOVA) is performed to see which process parameters are statistically significant [4]. Nomenclature GV gap voltage (V) WF wire feed (mm/min) TON pulse on time (μs) TOFF pulse off time (μs) Ra Surface Roughness Kf Kerf Width %P Percentage Contribution Units V Volts mm Millimetre min Minute μs Micro Seconds μm Micro meter Abbreviation WEDM wire cut electrical discharge machining DOF degrees of freedom LB lower the better SS stainless steel 870 M. Durairaj et al. / Procedia Engineering 64 ( 2013 ) 868 – 877 2. Design of experiments Taguchi Technique is applied to plan the experiments [5]. Orthogonal arrays were introduced in the 1940s and have been widely used in designing experiments [6]. It is used to reduce the number of experiments needed to be performed than the full factorial experiment. Based on the machine tool, cutting tool and work piece capability, the process parameters and the level for the process parameters were selected and listed in Table 2. Table 2. Machining parameters and their levels Taguchi proposed to acquire the characteristic data by using orthogonal arrays, and to analyze the performance measure from the data to decide the optimal process parameters. The designed combination of input parameters and its corresponding surface roughness and kerf width is shown in table 3 respectively. Table 3. Experimental Results 3. Optimization of process parameters is the key step in the Taguchi method to achieve high quality without increasing cost [7]. However, originally Taguchi method was designed to optimize single performance S. No Process Parameter Unit Level 1 Level 2 Level 3 Level 4 1 Gap Voltage V 40 45 50 55 2 Wire Feed mm/min 2 4 6 8 3 Pulse ON time (Ton) μs 4 6 8 10 4 Pulse OFF time (Toff) μs 4 6 8 10 Exp. No. Process Parameters Surface Roughness [μm] Kerf Width [mm] Gap Voltage [V] Wire Feed [mm/min] Pulse on Time [μs] Pulse off Time [μs] 1 40 2 0.315 4 2.36 0.315 2 45 4 0.308 6 2.28 0.308 3 50 6 0.308 8 2.17 0.308 4 55 8 0.297 10 2.51 0.297 5 50 4 0.299 4 2.56 0.299 6 55 2 0.300 6 2.85 0.300 7 40 8 0.296 8 2.48 0.296 8 45 6 0.289 10 2.34 0.289 9 55 6 0.293 4 2.02 0.293 10 50 8 0.301 6 2.22 0.301 11 45 2 0.294 8 2.05 0.294 12 40 4 0.298 10 2.39 0.298 13 45 8 0.305 4 2.32 0.305 14 40 6 0.308 6 2.32 0.308 15 55 4 0.297 8 2.21 0.297 16 50 2 0.311 10 2.49 0.311 871 M. Durairaj et al. / Procedia Engineering 64 ( 2013 ) 868 – 877 characteristics [8]. According to Taguchi method, the S/N ratio is the ratio of Signal to Noise where signal represents the desirable value and noise represents the undesirable value. The response Ra and Kf reported in Table 3, which is used to calculate the Signal to Noise Ratio (S/N) using the equation (1). The experimental results are now transformed into a signal-to-noise (S/N) ratio [9]. Since surface roughness and kerf width is desired to be at minimum, so Lower the Better characteristic is used for S/N ratio calculation. The optimal setting would be the one which could achieve lowest S/N ratio [10]. The S/N Ratio for the experiments conducted is shown in Table 4. 2 1 110 log( ) r iLB i S yN r (1) where S/NLB is the Signal to noise ratio (Lower the better), yi - output characteristic(Ra) and r no of trials . Table 4. Corresponding S/N ratios for Surface Roughness and Kerf Width Exp No. S/N Ratio for surface roughness S/N Ratio for kerf width 1 -7.4582 10.0338 2 -7.1587 10.2290 3 -6.7292 10.2290 4 -7.9935 10.5449 5 -8.1648 10.4866 6 -9.0969 10.4576 7 -7.8890 10.5742 8 -7.3843 10.7820 9 -6.1070 10.6626 10 -6.9271 10.4287 11 -6.2351 10.6331 12 -7.5680 10.5157 13 -7.3098 10.3140 14 -7.3098 10.2290 15 -6.8878 10.5449 16 -7.9240 10.1448 The mean value of S/N ratio for surface roughness and kerf width is tabulated for four levels are tabulated as shown in Table 5 and Table 6 respectively. Table 5. S/N Ratio Mean for Surface Roughness S. No. Process Parameters S/N Ratio Mean Level 1 Level 2 Level 3 Level 4 1 Gap Voltage -7.5563 -7.0220 -7.4363 -7.5213 2 Wire Feed -7.6786 -7.4448 -6.8826 -7.5299 3 Ton -7.3349 -8.1338 -6.7093 -7.3579 4 Toff -7.2600 -7.6231 -6.9353 -7.7175 From Fig.1, it can be seen that S/N Ratio decreases up to a short period then increases correspondingly to gap voltage. The S/N Ratio decreases up to a certain limit then increases correspondingly to the wire feed. When pulse on time and pulse off time increases, the S/N ratio will be deflected with increasing and decreasing. 872 M. Durairaj et al. / Procedia Engineering 64 ( 2013 ) 868 – 877 Fig.1. Effects of Process Parameters on Mean S/N Ratio for Surface Roughness Table 6. S/N Ratio Mean for Kerf Width S. No. Process Parameters S/N Ratio Mean Level 1 Level 2 Level 3 Level 4 1 Gap Voltage 10.3382 10.4895 10.3223 10.5525 2 Wire Feed 10.3173 10.4441 10.4757 10.4655 3 Ton 10.2592 10.5751 10.5600 10.3082 4 Toff 10.3743 10.3361 10.4953 10.4969 Fig.2. Effects of Process Parameters on Mean S/N Ratio for Kerf Width From Fig.2, it can be seen that S/N Ratio increases up to a short period then decreases gradually when the pulse off time and wire feed increases. With respect to the increase in pulse on time, the S/N Ratio decreases up to a short period and then increases gradually. The S/N ratio will be deflected with increasing and decreasing when the Gap Voltage increases. -8.3 -8.1 -7.9 -7.7 -7.5 -7.3 -7.1 -6.9 -6.7 -6.5 1 2 3 4 S/ N R at io M ea n No. of Levels Gap Voltage [V] Wire Feed [mm/min] Pulse on time [μs] Pulse off time [μs] 10.2 10.25 10.3 10.35 10.4 10.45 10.5 10.55 10.6 10.65 1 2 3 4 S/ N R at io M ea n No. of Levels Gap Voltage [V] Wire Feed [mm/min] Pulse on time [μs] Pulse off time [μs] 873 M. Durairaj et al. / Procedia Engineering 64 ( 2013 ) 868 – 877 4. Grey Relational Analysis Grey theory has been widely used in engineering analysis, and it reveals the potential to solve the setting of optimal machining parameters associated with a process with multiple output parameters [11]. The steps to be carried out are 1. Grey Relational Normalization 2. Grey Relational Gathering 3. Grey Relational Coefficient 4. Grey Relational Grade Step 1: Normalize the measured values of surface roughness and kerf width ranging from zero to one. This process is known as Grey relational normalization. Step 2: From the Grey relational normalization values, the grey relational gathering value can be determined using the required characteristics. Since both surface roughness and kerf width cannot be optimized for minimum value, lower the better and nominal the better characteristics are used to get the minimum surface roughness and nominal Kerf width respectively. Step 3: The Grey relational co-efficient is calculated to represent the relationship between the desired and actual data. Step 4: The average value of the grey relational co-efficient value of surface roughness and kerf width is known as overall Grey relational grade. Now, the multiple objective optimization problems have been transformed into a single equivalent objective function optimization problem using this approach. The four above mentioned values are shown in the table 4. From the grey relational grade values obtained, the means of the grey relational grades at different levels of process parameters were calculated [12]. Table 7. Grey Relational Gathering, Co-efficient and Grades Values Exp. No Grey Relational Normalized Values Grey Relational Gathering values Grey Relational Coefficient values Grey Relation Grade Values Surface Roughness Kerf Width Surface Roughness Kerf Width Surface Roughness Kerf Width 1 0.8281 1.0000 0.5903 1.0000 0.5496 1.0000 0.7748 2 0.8000 0.9778 0.6868 0.7309 0.6149 0.6501 0.6325 3 0.7614 0.9778 0.8194 0.7309 0.7346 0.6501 0.6924 4 0.8807 0.9429 0.4097 0.3079 0.4586 0.4194 0.4390 5 0.8982 0.9492 0.3496 0.3842 0.4346 0.4481 0.4414 6 1.0000 0.9524 0.0000 0.4230 0.3333 0.4643 0.3988 7 0.8702 0.9397 0.4457 0.2691 0.4742 0.4062 0.4402 8 0.8211 0.9175 0.6144 0.0000 0.5646 0.3333 0.4490 9 0.7088 0.9302 1.0000 0.1539 1.0000 0.3714 0.6857 10 0.7789 0.9556 0.7593 0.4618 0.6750 0.4816 0.5783 11 0.7193 0.9333 0.9639 0.1915 0.9327 0.3821 0.6574 12 0.8386 0.9460 0.5543 0.3455 0.5287 0.4331 0.4809 874 M. Durairaj et al. / Procedia Engineering 64 ( 2013 ) 868 – 877 13 0.8140 0.9683 0.6387 0.6158 0.5805 0.5655 0.5730 14 0.8140 0.9778 0.6387 0.7309 0.5805 0.6501 0.6153 15 0.7754 0.9429 0.7713 0.3079 0.6862 0.4194 0.5528 16 0.8773 0.9873 0.4337 0.8461 0.4920 0.7646 0.6283 Since the experiment is done by Taguchi L-16 orthogonal array, the separation of the effect of each machining parameter on the grey relational grade at different levels is tabulated as shown in table 8. Table 8. Grey Relational Grade value for corresponding levels S. No Process Parameter Grey Relational Grade Mean Level 1 Level 2 Level 3 Level 4 1 Gap Voltage 0.5778 0.5780 0.5851 0.5191 2 Wire Feed 0.6148 0.5269 0.6106 0.5076 3 Pulse-on Time 0.6347 0.4324 0.6006 0.5924 4 Pulse-off Time 0.6187 0.5562 0.5857 0.4993 From Fig.3, it can be seen that S/N Ratio slightly increases up to a certain limit then decreases when gap voltage increases. The S/N ratio will be deflected with increasing and decreasing when the wire feed, the Pulse on time and the Pulse off time increases. Fig.3. Effects of Process Parameters on Mean Grey Relational Grade 5. Analysis of Variance ANOVA is a statistically based, objective decision-making tool for detecting any differences in the average performance of groups of items tested [13]. The purpose of the Analysis of Variance (ANOVA) is to investigate which machining parameter significantly affects the performance characteristic [14]. 5.1. S/N Ratio for Surface roughness The results obtained by using analysis of variance to find out the percentage contribution of each input factor on the surface roughness are shown in table 5. It is seen that the pulse on time [μs] has the major influence on the surface roughness [15]. 0.4 0.45 0.5 0.55 0.6 0.65 1 2 3 4 G re y R el at io na l G ra de M ea n No. of Levels Gap Voltage [V] Wire Feed [mm/min] Pulse on time [μs] Pulse off time [μs] 875 M. Durairaj et al. / Procedia Engineering 64 ( 2013 ) 868 – 877 Table 9. ANOVA table for Surface roughness Factors Sum of Square DOF Mean Squares %P (Percentage Contribution) Gap Voltage 0.7293 3 0.2431 9.3451 Wire Feed 1.4527 3 0.4842 18.6146 Pulse ON time 4.0817 3 1.3606 52.3020 Pulse OFF time 1.5404 3 0.5135 19.7383 5.2. S/N Ratio for Kerf width The results obtained by using analysis of variance to find out the percentage contribution of each input factor on the kerf width are shown in Table 5. It is seen that the pulse on time [μs] has the major influence on the kerf width. Table 10. ANOVA table for Kerf width 5.3. Grey Relation Grade The results obtained by using analysis of variance to find out the percentage contribution of each input factor on both surface roughness and kerf width are shown in Table 5. It is seen that the Pulse ON Time has the major influence on the Surface Roughness and Kerf Width. Table 11. ANOVA table for Surface roughness and Kerf width 6. Results & Discussions The results obtained from the Taguchi Optimization technique to get the minimum Surface Roughness and minimum Kerf Width are shown in table 12. The grey relational analysis result is also shown in the table 13 to get the minimum surface roughness and nominal Kerf Width. Factors Sum of Square DOF Mean Squares %P (Percentage Contribution) Gap Voltage 0.1539 3 0.0513 24.4791 Wire Feed 0.0647 3 0.0216 10.2911 Pulse ON time 0.3277 3 0.1092 52.1234 Pulse OFF time 0.0824 3 0.0275 13.1064 Factors Sum of Square DOF Mean Squares %P (Percentage Contribution) Gap Voltage 0.0114 3 0.0038 6.4334 Wire Feed 0.0371 3 0.0124 20.9368 Pulse ON time 0.0978 3 0.0326 55.1919 Pulse OFF time 0.0309 3 0.0103 17.4379 876 M. Durairaj et al. / Procedia Engineering 64 ( 2013 ) 868 – 877 Table 12. Optimum conditions using Taguchi Optimization method S. No. Process Parameter Units Surface Roughness Kerf Width Best Level Value Best Level Value 1 Gap Voltage V 1 40 3 50 2 Wire Feed mm/min 1 2 1 2 3 Pulse on time μs 2 6 1 4 4 Pulse off time μs 4 10 2 6 Table 13. Optimum Conditions using Grey Relational Analysis S. No. Process Parameter Units Best Level Value 1 Gap Voltage V 3 50 2 Wire Feed mm/min 1 2 3 Pulse on time μs 1 4 4 Pulse off time μs 1 4 Fig.4. Range of occurrence of surface roughness Fig.5. Range of occurrence of kerf width 877 M. Durairaj et al. / Procedia Engineering 64 ( 2013 ) 868 – 877 From Fig. 4, it can seen that for a particular value of input parameter the corresponding range of occurrence of surface roughness can be determined and vice versa. From Fig. 5, it can be seen that for a particular value of input parameter the corresponding range of occurrence of kerf width can be determined and vice versa. 7. Conclusion Experimental investigation on wire electrical discharge machining of Stainless Steel (SS304) has been done using brass wire of 0.25mm. The following conclusions are made. ization method, the optimized input parameter combinations to get the minimum surface roughness are 40V gap voltage, 2mm/min wire feed, 6 μs pulse on time, 10 μs pulse off time and similarly optimized conditions to get the minimum kerf width are 50V gap voltage, 2mm/min Wire Feed, 4 μs pulse on time, 6 μs pulse off time. Based on the Grey relational analysis, the optimized input parameter combinations to get both the minimum surface roughness and the nominal kerf width are 50V gap voltage, 2mm/min wire feed, 4 μs pulse on time and 4 μs pulse off time. The Analysis of Variance resulted that the pulse on time has major influence on the surface roughness (μm) and kerf width (mm) in both the Taguchi optimization method and Grey relational analysis. The objectives such as surface roughness and kerf width are optimized using a single objective taguchi method and multi objective grey relational analysis and the same has been validated with the experimental results. References [1] Sorabh, Manoj Kumar, Neeraj Nirmal, 2013, A Literature review on Optimization of Machining Parameter in Wire EDM, International Journal of Latest Research in Science and Technology vol 2(1), p.492 [2] V. Muthu Kumar, A. Suresh Babu, R. Venkatasamy, M. Raajenthiren, 2010, Optimization of the WEDM Parameters on Machining Incoloy 800 Super Alloy with Multiple Quality Characteristics, International Journal of Science and Technology Vol 2(6), p.1538 [3] S.R. Nipanikar, 2012, Pameter Optimization of
/
本文档为【灰色关联】,请使用软件OFFICE或WPS软件打开。作品中的文字与图均可以修改和编辑, 图片更改请在作品中右键图片并更换,文字修改请直接点击文字进行修改,也可以新增和删除文档中的内容。
[版权声明] 本站所有资料为用户分享产生,若发现您的权利被侵害,请联系客服邮件isharekefu@iask.cn,我们尽快处理。 本作品所展示的图片、画像、字体、音乐的版权可能需版权方额外授权,请谨慎使用。 网站提供的党政主题相关内容(国旗、国徽、党徽..)目的在于配合国家政策宣传,仅限个人学习分享使用,禁止用于任何广告和商用目的。

历史搜索

    清空历史搜索