基于径向基函数神经网络的CFRP切削力预测
Prediction of cutting force in CFRP based on radial basis function neural network

碳纤维增强树脂基复合材料(CFRP)加工中基体相极易因切削力过大而破坏,并迅速扩展至加工表面以下而形成损伤。为了准确预测其切削力并加以控制,基于实验切削力数据建立了人工神经网络切削力模型,预测了不同纤维角度、切削深度和刀具角度下加工CFRP的切削力变化规律,并完成了不同刀具角度及切削参数下典型纤维角度CFRP单向板的直角切削实验,对预测模型进行验证,其预测精度可达85%以上。结合成屑过程在线显微观测结果可知:纤维角度是影响CFRP切削力的主要因素, 0°~135°范围内,切屑形成方式为切断型和开裂后弯断型;切削力随纤维角度增大呈先减小后增大的趋势, 135°时最大,随切削深度增加,切削力总体呈增大趋势。 In machining of carbon fiber reinforced polymer (CFRP), the matrix phase is easy to fail due to the excessive cutting force, which extends to the underneath of machining surface and forms damage rapidly. For the precise prediction and control of cutting force, we established an artificial neural network cutting force model based on the experimental cutting force data to predict the cutting force change rule in machining CFRP under different fiber orientations, cutting depths and tool angles. Orthogonal cutting experiments on CFRP unidirectional laminate of typical fiber orientation with different tool angles and cutting parameters were conducted to verify the predicting model, and the predicting precision is up to 85%. Combined with the online microscopic prediction results of chip forming process, it is concluded that the fiber orientation is the primary factor affecting the cutting forces of CFRP, as varies from 0° to 135°, chip formation ways include the crushing-dominated failure type and bending failure type; the cutting force firstly decreases and then increases as the fiber orientation angle increases, and is maximum at 135°; the cutting force increases generally as cutting depth increases. 国家"973"计划(2014CB046503);国家自然科学基金创新研究群体(51321004);教育部新世纪优秀人才支持计划(NCET-13-0081)