雷射銲接鋁合金之特性探討及預測模式之建立

An Investigation of the Characteristics and a Predictive Model Establishment in Laser Welding Aluminum Alloy

黃耀賢(Yao-Hsien Huang)

指導教授 ： 簡文通

國立屏東科技大學/工學院/機械工程系所/碩士(2010年)

https://doi.org/10.6346/NPUST.2010.00063

摘要

論文摘要內容：
本研究之目的為探討Nd:YAG雷射對6061-T6鋁合金之銲接特性，並建立銲件抗拉強度之預測模式。選用之雷射加工參數為尖峰功率、脈衝寬度、脈衝頻率、聚焦位置、銲接速度和脈衝重複率等。利用田口法之直交表規劃實驗配置共25組用以執行實驗，並量測銲件之抗拉強度和銲道截面積。實驗結果的分析及應用可分為兩部份，第一部份為探討6個雷射加工參數分別對銲件抗拉強度及銲道截面積之影響性，並用來求取鋁合金對雷射能量吸收率的關係。然後藉由調整雷射加工參數以提高吸收率，增進雷射銲接鋁合金材料的加工效率。第二部份為架構一雷射銲接鋁合金工件之抗拉強度預測模式，使用的原理為倒傳遞類神經網路，上述之實驗結果則被用做訓練及回想範例。實驗結果顯示最佳參數組合經過確認實驗之後得到抗拉強度為157.66MPa，與25組中最佳的第21組相比增加了約7.84%；同時也可得到抗拉強度與銲道截面積接近線性的正比關係。當改變影響性較大的銲接速度、尖峰功率與脈衝寬度等三個雷射參數，可達到雷射能量吸收提升的效果。抗拉強度的改善方面，當改變銲接速度到0.1mm/sec時，與最佳參數組合相比提高了6.48%；當將尖峰功率改變至2300W時，與最佳參數組合相比提高了2.59 %；當脈衝寬度提高至12.5ms時，與最佳參數組合相比提高了4.23%。最後，經由驗證實驗得知利用倒傳遞類神經網路所架構之鋁合金抗拉強度預測模式的平均誤差為6.53%，顯示此模式具有良好的預測能力。本研究之成果將有助於克服鋁合金材料對雷射能量高反射率之問題，對相關研究或產業應用上有實質上的貢獻。

關鍵字

雷射銲接 ； 鋁合金 ； 田口法 ； 倒傳遞類神經網路 ； 預測模式

並列摘要

The Contents of Abstract in this Thesis :
The purpose of this study is to investigate the welding characteristics of 6061-T6 aluminum alloy by Nd:YAG laser, and construct a predictive model for tensile strength of the laser welded workpiece. The laser welding parameters chosen in this study are peak power, impulse width, impulse frequency, focusing position, welding speed and gas pressure. An orthogonal array of Taguchi method is used to arrange experimental tests. After each test the tensile strength and the welded cross section area are measured for analyzing and applying in two parts. In the first part, the influence on the tensile streugth and the cross section area is investigated by the 6 laser welding parameters, respectively. The results are further used to find the relationship between the absorption rate for aluminum alloy and the provided laser energy. By adjusting the amount of laser welding parameters, the improvement of the absorption rate can be achieved. Therefore, the increasing productivity for laser welding aluminum alloy can be achieved. In the second part to construct a predictive model for the tensile strength in laser welding aluminum alloy has been conducted. A back-propagation artificial neural network theory is used in this program as the training and recalling patterns. The experimental result showed that after conducting a confirmation experiment an optimum parameter combination has been found to get the tensile strength is 157.66MPa. When compared with the best results of the 25 tests, an increase of approximately 7.84% has been found. It is also found that the relationship between the tensile strength and the weld bead cross-sectional area is expressed in a linear propagation. The effect of laser energy absorption is improved when the more important affecting parameters for welding speed, peak power and pulse width has been adjusted, respectively. By comparing with the results obtained with the optimum parameter combination, an improvement for tensile strength in 6.48% is achieved when the welding speed is 0.1 mm/sec; in 2.59% is achieved when the peak power is 2300W; and in 4.23% is achieved when the pulse width is 12.5ms. Finally, it showed that a mean error 6.53% is found for the constructing predictive model when compared with the verifying experiments. It is can be concluded the predictive model has a good predicting ability for the tensile strength in laser welding aluminum alloy. The achievement in this study should provide a direction to overcome the high reflection problem for aluminum alloy in a laser fabrication process. Hence the contribution in this study can be used for related research field or industrial application.

並列關鍵字

Laser Welding ； Aluminum Alloy ； Taguchi Method ； Back-propagation Artificial Neural Network ； Predictive Model

參考文獻

[13] 徐上祐，2008，雷射切割鎂6061鋁合金薄板之製程參數最佳化分析，碩士論文，國立屏東科技大學，機械工程系，屏東。

[19] 程昭榮，2008，Nd：YAG雷射銲接鎂合金最佳化參數及預測模式之探討，碩士論文，屏東科技大學，機械工程系，屏東。

[36] Hirose A., Todaka H., Yamaoka H., Kurosawa N. and Kobayashi K. F., Metallurgical and Materials Transactions, Vol.30A, (1999), pp.2115-2120.

[38] Tarng Y. S., and Yang W. H., 1998, “Design optimization of cutting parameters for turning operations based on Taguchi method,” Journal of Materials Processing Technology, Vol. 84, pp. 122-129.

[39] Lin T. R., 2002, “Experimental design and performance analysis of TiN-coated carbide tool in face milling stainless steel,” Journal of Materials Processing Technology , Vol. 127, pp. 1-7.