超声振动辅助磨削牙科氧化锆陶瓷切削力预测模型研究  

Prediction model of cutting force in ultrasonic vibration assisted grinding of zirconia ceramics

作　　者：肖行志[1] 郑侃[1] 廖文和[1] 

机构地区：[1]南京理工大学机械工程学院,南京210094

出　　处：《振动与冲击》2015年第12期140-145,共6页Journal of Vibration and Shock

基　　金：国家自然科学基金项目(51305206);中央高校基本科研业务费专项资金(2012XQTR001)

摘　　要：通过单因素实验设计,开展了牙科氧化锆陶瓷的超声振动辅助磨削实验,分别建立了超声振动辅助磨削加工切削力指数预测模型、BP神经网络预测模型以及理论预测模型。通过验证实验对比分析了三种模型的预测精度,并阐明了误差产生的原因。结果表明,基于BP神经网络的切削力预测模型相对于指数和理论模型具有较高的预测精度,其平均相对误差仅为9.60%,理论模型因未能考虑材料的塑性流动去除,导致预测精度较低。Ultrasonic vibration assisted grinding （UVAG）experiments on dental zirconia ceramics were conducted by using single-factor method.An index prediction model,a BP neural networks prediction model,as well as a theoretical prediction model of cutting force in UVAG were proposed.A verification experiment was carried out to study the prediction accuracy of these three models,and the causes of errors were revealed.The results indicate that the BP neural networks prediction model has higher precision compared with the two others,and the relative error is only 9.60%.The prediction accuracy of theoretical model is poor due to the neglect of plastic flow removal during UVAG.

关 键 词：超声振动辅助磨削 牙科氧化锆陶瓷 切削力预测 指数模型 BP神经网络 理论模型 

分 类 号：TH-16[机械工程]