本研究主要目的是探討槍鑽深加工最佳化孔壁品質的加工參數及刀具幾何參數,最終目標即是獲得較佳且分佈均勻的深孔孔壁表面粗度。 本研究首先利用田口式直交表進行規劃本研究所需的實驗配置,實驗配置將考慮包括主軸轉速、進給率、以及外切削角、內切削角、鑽頂位置、外傾角與內傾角等刀具幾何作為實驗因子。利用實驗規劃配置的參數組合實際進行鑽孔加工實驗,並分段量測鑽孔孔壁表面粗糙度值。再採用統計學觀念將每個深孔的表面粗糙度分佈值以平均值和標準差來表示,並以誘導式類神經網路系統建立孔壁表面粗糙度平均值和標準差兩者與加工參數和刀具幾何間的關係模式,這些模式經實驗結果證實具有相當可用性。最後利用模擬退火演算法進行參數最佳化,以期達到整體深孔孔壁表面粗糙度最佳化。經實驗結果證實,最佳化參數組合應用在槍鑽深孔加工時,可以獲得較佳且均勻的孔壁表面品質。
In this research the main goal is to discuss the optimum machining parameters and tool geometry for the hole wall quality in gun drilling and the final purpose is to obtain the better and distribute-uniformly roughness of the hole wall. The Taguchi’s orthogonal arrays is first used to plan the experiments in this research; the considered factors included the spindle speed, feed rate and the out point cutting angle, inner point cutting angle, position of the point, outer rake angle and inner rake angle of the tool geometry. The experiments will be done by gun-drill machine and the roughness in every interval of hole wall will be measured. The distributing value of surface roughness in every deep hole is expressed as the average and the standard deviation from the statistical viewpoint and the Abductive Networks is used to establish the models of averages and the standard deviations between the machining parameters and the tool geometry. At least the simulated annealing algorithm is used to optimize the machining parameters and the tool geometry so that the optimization of surface roughness of the hole wall in the whole deep hole can be achieved. The experimental results show that the combination of the optimum parameters for gun drilling can obtain the better and more uniform surface quality of the hole wall.