慢走丝线切割机床丝架的热误差建模

摘要

慢走丝线切割机床上下丝架的误差受环境温度影响变化较大,且其相对位置的改变直接造成电极丝的位姿发生变化,从而使加工精度降低. 针对某大型慢走丝线切割机床,用旋量理论建立其理论热误差模型,采用有限元的方法计算在环境温度改变时上下丝架在x、y、z方向上的变形量,并用MATLAB处理得出电极丝位姿在笛卡儿坐标系中的变化情况. 构建实验平台实测不同环境温度下线切割机床上下丝架的热变形,基于径向基函数( radical basis function,RBF)神经网络的M-RAN算法建立热误差模型. 结果表明:该误差模型残差小于1μm,可以良好地反映出热误差分布及变化的规律.%The wire is the main processing unit of the low speed wire-cut electrical discharge machine ( WEDM-LS) . The wire rack is a cantilever structure. Its support member—the up and down wire racks is a cantilever structure. The error of the up and down wire racks is significantly affected by ambient temperature, and the change of the up and down wire racks' relative position makes the wire's position and orientation change directly, which reduce the machining precision. In this paper, the screw theory was used to establish the WEDM-LS' theoretical thermal error model. The finite element method was used to calculate the deformation of the upper and lower wire racks in x, y, and z directions when the ambient temperature changes, and drawn the change of the wire's position and orientation with MATLAB in the Cartesian coordinate system. An experimental platform was built to measure the wire rack's thermal deformation under different ambient temperatures, and establish a thermal error model based on the M-RAN RBF neural network algorithm, which provides a basis for compensating for the numerical control system. Results show that the model residual error is less than 1 μm, and it can well reflect the distribution and the variation of the thermal error.