Nd:YAG雷射銲接鎂合金最佳化參數及預測模式之探討

Investigation on the Optimum Parameters and Predictive Model on Nd:YAG Laser Welding Magnesium Alloy AZ31B

程昭榮(Jhao-Rong Cheng)

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(2007年)

https://doi.org/10.6346/NPUST.2007.00021

摘要

本研究之目的在探討鎂合金AZ31B雷射銲接處之特性並建立其抗拉強度預測模式。首先利用田口法直交表規劃實驗參數的配置，以銲接後之抗拉強度及銲道凹陷量為銲接品質兩個目標函數。藉由田口法對此二目標函數分別進行銲接參數最佳化分析，並探討雷射加工參數對目標函數的影響趨勢。相關實驗數據並用做為架構以倒傳遞類神經為理論基礎之抗拉強度預測模式。整體實驗規劃分為三部分，第一部分經由前置實驗結果選定影響雷射銲接品質較為重要的六個參數，分別為尖峰功率、脈衝頻率、脈衝寬度、銲接速度、聚焦位置和氣體壓力，依照田口法直交表L2556配置雷射銲接參數進行實驗。第二部份則是搭配倒傳遞類神經網路以第一部份的25組實驗數據，提供倒傳遞類神經網路架構之預測模式所需的練習範例及回想範例，再利用田口法之平均數分析找出最佳化網路參數組合。第三部分則是配置9組不包含回想範例及訓練範例之雷射銲接參數進行實驗以驗證預測模式之準確性。結果顯示在個別單一品質目標之最佳銲接參數組合中，分別得到抗拉強度為原材料的71.3%，及銲道凹陷量為0.21mm等為最佳銲接品質。而影響銲接品質較重要的參數則分別為脈衝頻率，及銲接速度。此外，經由驗證實驗得知利用倒傳遞類神經網路所架構之抗拉強度預測模式的平均誤差為6.89%，顯示此模式具有良好的預測能力。本研究的過程及結果對於雷射銲接鎂合金AZ31B之加工品質及結果預測可提供實質上的協助及參考。

關鍵字

雷射銲接 ； 鎂合金AZ31B ； 最佳化參數 ； 預測模式

並列摘要

The purpose of this study is to investigate the characteristics for the welded portion of magnesium alloy AZ31B in a laser welding process and construct a predictive model for predicting its tensile strength. Firstly, the experiments arranged by an orthogonal array of Taguchi method then after welding conducted the tensile strength and the weld under-cut were selected for two single quality objectives. The analytic methods of Taguchi method were used to find the optimum welding parameters combination for these two objectives separately, and the influence of each parameter on the objective was also discussed simultaneously. The experimental results can be further used for constructing a welding tensile strength predictive model that is based on a back-propagation neutral network algorithm. The overall experimental procedure is divided into three parts. In the first part, according to the preliminary tests there are six laser welding parameters that showed stronger influence on welding quality have been selected; namely, peak power, pulse frequency, pulse width, welding speed, focus position, and gas pressure. The experiments were conducted based on the arrangement of laser welding parameters by an L2556 orthogonal array of Taguchi method. In the second part, the former experimental results were used for training patterns and recalling patterns of the predictive model which was constructed base on a back-propagation neutral network. The optimum network parameters were attained by the average mean analysis of Taguchi method. In the third part, there are nine sets of verifying experiments without including any experiment of the training patterns or the recalling patterns have been conducted to validate the accuracy of this predictive model. It has shown that a 71.3% for the tensile strength of the original material, and a 0.21mm for the weld under-cut were obtained separately form each single quality objective base optimal welding parameters. The most important affecting factor on the welded tensile strength is impulse frequency, and on the weld under-cut is welding speed. Moreover, the results of the verifying experiments showed that a mean error as 6.89% was found when the predictive values were compared. It indicates the predictive model developed base on a back-propagation neural network has good predicting ability for the laser welded tensile strength. The processes and results in this study provide substantial assistance and reference for efficiently laser welding magnesium alloy AZ31B.

並列關鍵字

Laser Welding ； Magnesium Alloy AZ31B ； Optimum Parameters ； Predictive Model

參考文獻

[23]楊俊彬，2005，「3C產業用鎂合金外殼鍛件之開發」，鍛造，第2卷，第14期，第35-45頁。

[36]Chen C. B., Lin C. T., Chang C. W., and Ho C. P., 2000, “Grey Relation for Solving Multi-Quality characteristics Problems of Taguchi Methods,” Journal of Technology, Vol. 15, No.1, pp. 25-33.

[37]Tsao Chung-Chen, 2002, “Prediction flank wear of different coated drills for JIS SUS 304 stainless steel using neural network,” Journal of Materials Processing Technology, Vol. 123, pp. 354-360.

[38]Darwish S. M., 2002, “Characteristics of weld-bonded commercial aluminum sheets,” International Journal of Adhesion and Adhesives, Vol. 23, pp. 169-176.

[39]Davim J. P., 2003, “Study of drilling metal-matrix composites based on the Taguchi techniques,” Journal of Materials Processing Technology, Vol 132, pp. 250-254.