%0 Journal Article
%T 共轨管微小孔磨粒流抛光实验研究与表面粗糙度预测<br>Experimental Study on Abrasive Flow Polishing of Common-rail Tube Micro-holes and Surface Roughness Prediction
%A 蔡智杰
%A 刘薇娜
%A 高彬彬
%A 任成祖
%J 机械科学与技术
%D 2017
%X 针对共轨管微小孔电火花加工表面的光整加工问题，采用磨粒流抛光工艺，并通过正交实验探索了加工参数及其交互作用对孔道表面粗糙度的影响规律；基于二阶响应曲面模型和幂函数模型分别建立了表面粗糙度预测多元非线性回归模型。研究结果表明：加工参数对抛光效果的影响显著，而交互作用对其影响较小；抛光压强、磨料浓度及加工时间对孔道表面粗糙度的影响均为负效应，磨粒粒径大于148 μm时对表面粗糙度的影响为正效应，粒径小于该临界值时表现为对抛光效率的正效应影响；在最优参数组合条件下，孔道表面粗糙度值（Ra）由初始的1.31 μm降至0.20 μm；二阶多项式回归模型相对于幂函数回归模型有更高的预测精度，相关系数高达0.990，预测误差在9.54%以内。<br>Aiming to the surface finishing of common-rail tube micro-holes manufactured by electrical discharge machining, abrasive flow polishing was adopted. The effects of the processing parameters and their interaction on the surface roughness have been studied via orthogonal experiments. The prediction model for surface roughness via multiple nonlinear regression equation was established based on the second order response surface model or power function model. The results showed that the processing parameters had a significant influence on the polishing effect, but the interaction had less impact on it. The surface roughness of the tunnels has a negative correlation with the pressure, concentration and processing time, but the positive correlation with the grain size when it exceeds a critical value of 148 μm. At grain size below critical value, it shows a positive correlation with polishing efficiency. The combinatorial optimization is obtained, and under this condition, the surface roughness value decreased from 1.31 μm to 0.2 μm. The prediction accuracy via second-order polynomial regression equation was higher than that via power function regression equation. The correlation coefficient was up to 0.990, and the prediction errors was less than 9.54%
%K 共轨管
%K 磨粒流抛光
%K 表面粗糙度
%K 回归分析
%K 共轨管
%K 磨粒流抛光
%K 表面粗糙度
%K 回归分析&lt
%K br&gt
%K common-rail tube
%K abrasive flow polishing
%K surface roughness
%K regression analysis
%K common-rail tube
%K abrasive flow polishing
%K surface roughness
%K regression analysis
%U http://journals.nwpu.edu.cn/jxkxyjs/CN/abstract/abstract6866.shtml