CHAPTER 3 EXPERIMENTAL PROCEDURES
4.3 Experimental results of RSW process
With combination of this Taguchi method and a neural network, the optimal welding conditions for Max. Load with RSW process were electrode tip size at φ3 mm, welding current at 7800 A, electrode force at 3.0 kN and welding time at 15 cycles. Table 4-4 shows the experimental results obtained with above optimal welding parameters. Table 4-5 shows the experimental results with the conditions of production operation currently (Aφ4mmB7800AC1.8kND8cycles).
Comparison of Table 3-24 with Table 4-5 shows that the increase in average Max. Load from the initial conditions to the initial optimal parameters (apply Taguchi method only) is 0.309 kN. Comparison of Table 4-4 with Table 4-5 shows that the increase in average Max. Load from the initial conditions to the real optimal parameters (apply Taguchi method and neural network) is 0.566 kN.
The surface condition of specimens for different parameters is shown in Fig.4-2.
In summary, the quality of RSW process for high strength steel sheet can be efficiently improved with the proposed approach.
Table 4-4 Results of the proposed approach in RSW process
Max. Load Trial no.
N1 specimens N2 specimens
Average (kN)
21 4.310 4.169 4.112 3.746
22 4.153 3.973 4.522 3.876
4.108
Table 4-5 Results of the initial conditions in RSW process
Max. Load Trial no.
N1 specimens N2 specimens
Average (kN)
23 3.329 3.518 3.605 3.344
24 3.673 3.575 3.626 3.669
3.542
Fig.4-2 Surface conditions of specimens for validation (a) Initial conditions
(b) Apply Taguchi method only
(c) Apply Taguchi method with proper regulation (d) Apply proposed approach
CHAPTER 5
CONCLUSION
This dissertation presents an integrated approach of the combination of Taguchi method and neural networks to optimize the process conditions of GTA welding, laser-micro weld and RSW process. Based on the results obtained from this research, the following conclusions can be drawn from this dissertation.
1. In GTA welding process, the improvement of average depth-to-width ratio from initial optimal parameters (apply Taguchi method) to the optimal parameters (apply proposed approach) is about 11.96%. The largest depth-to-width ratio of the initial optimal parameters by Taguchi method is 0.712. The largest depth-to-width ratio of the optimal parameters by proposed approach is 0.806.
2. The ANOVA result indicates that, the electrode angle, welding current, and travel speed are the significant parameters in affecting the depth-to-width ratio of weld pool geometry in GTA welding process.
3. In Nd:YAG laser micro-weld process, The improvement of the defective rate from initial conditions to the initial optimal parameters (apply Taguchi method) is 3.30%; from initial conditions to the optimal parameters (apply proposed approach) is 6.67 %.
4. The simulating results indicate that, the specimens (AA3003 aluminum alloy) with 50% cleanliness contributed most to the Nd:YAG laser micro-weld process. In order to improve the welding quality efficiently, the cleaning treatment to the safety vent and cathode lead of lithium-ion secondary batteries must be corrected.
5. In RSW process, the improvement of the average tensile-shear strength from initial conditions to the initial optimal parameters (apply Taguchi method) is about 8.72%. The improvement from initial conditions to the optimal parameters (apply proposed approach) is about 15.98%.
6. The ANOVA result indicates that, the size of electrode tip and welding current were the significant parameters in affecting the tensile-shear strength in RSW process for high strength steel sheet.
7. Compare with the results of ANOVA, there are 26.88% of error contribution in GTA welding, 7.75% of error contribution in Nd:YAG laser micro-weld process, and 19.72% of error contribution in RSW process. It shows that the experimental error of Nd:YAG laser micro-weld process is least and GTA welding is largest.
8. From the results of confirmation test in these welding processes, the conformity of reproducibility for the experimental results has been confirmed 9. The proposed approach is relatively effective and ease for engineers to apply
to a range of other processes. The LMBP algorithm neural network is easy-and-quick to explore a nonlinear multivariate relationship between parameters and responses. It was proved successfully and effectively in this study.
10. In addition, applying the proposed approach allows engineers to directly use neural network software to optimize the parameters without any theoretical knowledge of neural computing.
REFERENCE
1. H.B. Cary, Modern Welding Technology, Prentice-Hall, New Jersey, 1994.
2. G. Taguchi, E.A. Elsayed and T.C. Hsiang, Quality Engineering in Production Systems, McGraw-Hill, New York, 1989.
3. C.T. Su, C.C. Chiu, and H.H. Chang, “ Parameter design optimization via neural network and genetic algorithm”, Int. J. Ind. Eng., 7(3), 224-231, 2000.
4. I.S. Kim, Y.J. Jeong, and C.W. Lee, “ Prediction of welding parameters for pipeline welding using an intelligent system”, Int. J. Adv. Manuf. Technol., 22, 713-719, 2003.
5. H.K.D.H. Bhadeshia, D.J.C. Mackay, and L.E. Svensson, Mater. Sci. Technol., 11, 1046-1051, 1995.
6. John F.C. Khaw, B.S. Lim, and Lennie E.N. Lim, “ Optimal design of neural networks using Taguchi method”, Neurocomputing, 7, 225-245, 1995.
7. P.J. Ross, Taguchi Techniques for Quality Engineering, McGraw-Hill, New York, 1988.
8. G. Taguchi, Taguchi methods: Design of Experiments, American Supplier Institute, Inc, Michigan, 1993.
9. R.K. Roy, A primer on the Taguchi Method, Van Norstrand Reinhold, New York, 1990.
10. M.S. Phadke, Quality Engineering using Robust Design, Prentice-Hall, New Jersey, 1989.
11. S.C. Juang and Y.S. Tarng, “ Process parameter selection for optimizing the pool geometry in the tungsten inert gas welding of stainless steel”, J. Mater.
Process. Technol., 122, 33-37, 2002.
12. Wei Li, S. Jack Hu, and Shao-Wei Cheng, “ Robust design and analysis for manufacturing processes with parameter interdependency”, Manuf. Syst.,
21(2), 93-100, 2002.
13. Y.S. Tarng, S.C. Juang and C.H. Chang, “ The use of grey-based Taguchi methods to determine submerged arc welding process parameters in hardfacing”, J. Mater. Process. Technol., 128, 1-6, 2002.
14. C.T. Su, and T.L. Chiang, “Optimizing the IC wire bonding process using a neural networks/genetic algorithms approach”, J. Intell. Manuf., 14, 229-238, 2003.
15. D.W. Coit, B.T. Jacson, and A.E. Smith, “Static neural network process model: considerations and cases studies”, Int. J. Prod. Res., 36(11), 2953-2967, 1998.
16. K. Funahashi, “ On the approximate realization of continuous mapping by neural network”, Neural Netw., 2, 183-192, 1989.
17. M.T. Hagan, H. Demuth and M. Beale, Neural Network Design, PWS Publishing co., Boston, 1996.
18. M.T. Hagan and M.B. Menhaj, “ Training feedforward networks with the Marquardt algorithm”, IEEE Trans, Neural Netw. , 5(6), 989-993, 1994.
19. F. A. Dias, A. Antunes, J. Vieira and A. Mota, “ A sliding window solution for the on-line implementation of the Levenberg-Marquardt algorithm”, Engineering Applications of Artificial Intelligence, 19, 1-7, 2006.
20. K. Kumar and M.A. Alsaleh, “ A comparative study for the estimation of parameters in nonlinear models”, Applied Mathematics and Computation 77, 179-183, 1996.
21. S. Haykin, Neural Networks-A Comprehensive Foundation, Macmillan College Publishing co., New York, 1994.
22. H. Demuth and M. Beale, Neural Network Toolbox-For use with MATLAB, The Math Works, Inc., Boston, 1998.
23. I.S. Kim, I.S. Son, C.E. Park, I.J. Kim and H.H. Kim, “ An intevestigation into an intelligent system for predicting bead geometry in GMA welding process”, J. Mater. Process. Technol., 159, 113-118, 2005.
24. C.S. Wu, T. Polte and D. Rehfeldt, “ Gas metal arc welding process monitoring and quality evaluation using neural networks”, Sci. Technol. Weld Joi. , 5(5), 324-328, 2000.
25. D.S. Nagesh and G.L. Datta, “ Prediction of weld bead geometry and penetration in shielded metal-arc welding using artificial neural networks”, J.
Mater. Process. Technol., 123, 303-312, 2002.
26. G.E. Ridings, R. C. Thomson and G. Thewlis, “ Prediction of multiwire submerged arc weld bead shape using neural network modelling”, Sci.
Technol. Weld Joi., 7(5), 265-279, 2002.
27. J.Y. Jeng, T.F. Mau and S.M. Leu, “ Prediction of laser butt joint welding parameters using back propagation and learning vector quantization networks”, J. Mater. Process. Technol., 99, 207-218, 2000.
28. J.I. Lee and K.W. Um. Tarng, “ A prediction of welding process parameters by prediction of back-bead geometry”, J. Mater. Process. Technol., 108, 106-113, 2000.
29. J.M. Vitek, S.A. David, M.W. Richey, J. Biffin, N.Blundell and C.J. Page,
“ Weld pool shape prediction in plasma augmented laser welded steel”, Sci.
Technol. Weld Joi., 6(5), 305-314, 2001.
30. Y.S. Tarng, H.L. Tsai and S.S. Yeh, “ Modeling, optimization and classification of weld quality in tungsten inert gas welding”, Int. J. Machi.
Tools & Manuf., 39, 1427-1438, 1999.
31. Y.L. Han, “ Artificial neural network technology as a method to evaluate the fatigue life of weldments with welding defects”, Int. J. Pres. Ves. & Piping,
63, 205-209, 1995.
32. H. Rowlands, M.S. Packianather and E. Oztemel, “ Using artificial neural networks for experimental design in off-line quality”, Journal of Systems Engineering, 6, 46-59, 1996.
33. C.C. Chiu, C.T. Su, G.H. Yang, J.S. Huang, S. C. Chen and N.T. Cheng,
“Selection of optimal parameters in gas-assisted injection moulding using a neural network model and the Taguchi method”, Int. J. Quality Sci., 2, 106-120, 1997.
34. J. Norrish, Advanced Welding Processes, IOP Publishing Ltd., New York, 1992.
35. S.C. Juang and Y.S. Tarng, “ Process parameter selection for optimizing the pool geometry in the tungsten inert gas welding of stainless steel”, J. Mater.
Process. Technol., 122, 33-37, 2002.
36. Paulo J. Modenesi, Eustaquio R. Apolinario and Iaci M. Pereira, J. Mater.
Process. Technol., 99, 260-265, 2000.
37. Walter W. Duley, Laser Welding, John Wiley & Sons, Inc., New York, 1998.
38. Christopher Dawes, Laser Welding, McGraw-Hill Inc., New York, 1992.
39. L.K. Pan, C.C. Wang, Y.C. Hsiao and K.C. Ho, “ Optimization of Nd:YAG laser welding onto magnesium alloy via Taguchi method”, Opt. Laser Technol., 37, 33-42, 2004.
40. J.M. Jr. Sawhill and J.C. Baker, “Spot weldability of high-strength sheet steel”, Weld. J., 59(1), 19-s to 30-s, 1980.
41. Z. Han, J.E. Indacochea, C.H. Chen and S. Bhat, “ Weld nugget development and integrity in resistance spot welding of high-strength cold-rolled sheet steels”, Weld. J., 1993, 72(5), 209-s to 216-s.
42. W.F. Savage, E.F. Nippes and F.A. Wassell, “Dynamic contact resistance of
series spot welds”, Weld. J., 1978, 57(2), 43-s to 50-s.
43. Yu Thomas, Study on the Pulse Nd:YAG laser welding for packaging process parameters, National Tsing Hua University (Taiwan), Thesis, 2002.
AUTHOR’S PUBLICATION LIST
A. Referred Papers
1. H. L. Lin and C. P. Chou., Optimisation of the GTA welding process using Taguchi method and a neural network, Science and Technology of Welding and Joining, 2006, 11(1), 120-126.
2. H. L. Lin and C. P. Chou., Modeling and optimization of Nd:YAG laser micro-weld process using Taguchi Method and a neural network, Accepted to International Journal of Advanced Manufacturing Technology, 2006.
3. H. L. Lin, Ting Chou and C. P. Chou., Optimization of resistance spot welding process using Taguchi Method and a neural network, Accepted to Experimental Techniques, 2006.
4. H. L. Lin, Ting Chou and C. P. Chou., “Modeling and optimization of the resistance spot welding process in automobile industry,” Submitted to Proceedings of the Institution of Mechanical Engineers Part D Journal of Automobile Engineering ,Under review, 2006.
5. 林玄良、周長彬，Nd:YAG 雷射銲接穩健性製程設計之研究，銲接與切割，
15(4)，PP.56-61，2005。
6. 林玄良、周定、周長彬，建立汽車鋼板接合參數預測模型之研究，銲接 與切割，16(1)，PP.56-61，2006。
B. Conference Papers
1. H. L. Lin and C. P. Chou, Modeling and Optimization of resistance spot welding process in automobile industry, The First South-East Asia - International Institute of Welding Congress, Bangkok, Thailand, 21-22 November 2006.
2. H. L. Lin, C. P. Chou, and M. S. Chen, Selection of the GTAW-flux process parameters on type 310 stainless steel via a Taguchi-Neural approach, Abstract of paper has been accepted to The International Welding/Joining Conference, Seoul, Korea, 10-12 May 2007.
3. 林玄良、蔡勝禮、周長彬，雷射微接合製程參數最佳化之實務研究，中
C. Others
1. 林玄良，探討田口方法在自動化銲接之應用，銲接與切割，11(2)，
pp.35-38，2001。
2. 林玄良，應用田口方法改善工程中的品質問題，機械技術雜誌，195，
pp.146-150，2001。
3. 林玄良，遺傳演算法在銲接工程之應用，銲接與切割，11(6)，pp.14-18，
2001。
4. 林玄良，淺談汽車用鋁合金及其最新鑄造技術，機械技術雜誌，200，
pp.137-139，2001。
5. 林玄良，雷射銲接在汽車工業的最新應用，機械技術雜誌， 206，
pp.147-152，2002。
6. 林玄良，淺談鋁合金半固態成形技術，機械技術雜誌，210，pp.124-127，
2002。
7. 林玄良，淺談摩擦攪拌銲接技術及其最新發展，機械技術雜誌，210，
pp.148-152，2002。
8. 林玄良，銲接製程之可靠度分析，銲接與切割，12(4)，PP.31-40， 2002。
9. 林玄良，電漿電弧銲接在微接合加工技術之應用，機械技術雜誌，216，
pp.113-117，2003。
10. 林玄良，銲接對壓力容器的影響及問題對策，工業安全衛生月刊，164，
pp.32-40，2003。
11. 周長彬、林玄良、黃處明、潘聖富，高熵合金銲接性能研究結案報告，
工業技術研究院分包學術機構研究計畫，新竹，2003.12。