摘要
针对曲轴滚压强化工艺中滚压工艺参数难以确定的问题,利用Deform-3D有限元软件对康明斯双缸空压机曲轴进行滚压数值模拟,并运用BP神经网络对滚压工艺参数和残余应力进行训练,得到二者的映射关系。然后利用该BP神经网络预测曲轴在特定工况下的滚压工艺参数,为实际生产过程中滚压工艺参数的合理选取提供参考。
It is difficult to determine the crankshaft rolling process parameters in rolling strengthening technology.The paper simulates the operation status of the crankshaft of Cummins two-cylinder compressor based on Deform-3D software,uses the BP neural network to train the rolling process parameters and residual stresses,and obtains the relationship rolling process parameters and residual stresses. It determines the rolling process parameters under certain conditions and BP neural network,provides the theoretical guidance for the selection of rolling process parameters in practical production.
作者
阮景奎
倪帆
周学良
RUAN Jingkui NI Fan ZHOU Xueliang(School of Mechanical Engineering, Hubei Institute of Automobile Industry, Hubei Shiyan, 442002, Chin)
出处
《机械设计与制造工程》
2017年第2期55-59,共5页
Machine Design and Manufacturing Engineering
基金
湖北省自然科学基金资助项目(2014CFB633)
关键词
滚压强化
数值模拟
残余应力
BP神经网络
rolling strengthening
numerical simulation
residual stress
BP neural network
