机床基础大件支撑预变形设计方法研究
Research on Design Methodology for Support Pre-deformation of Large Machine Tools Body

针对精密机床装配过程中导轨基面静变形对几何精度的影响，提出一种预先消除该静变形的方法，即对整机进行静力学分析，获得装配后导轨基面的静变形，通过调整支撑方式使该变形在装配前呈现，便于将此变形提前消除。以一精密卧式加工中心为例，详细阐述了该方法，该方法的关键是通过优化支撑方式获得导轨基面最终静变形。以支撑点的定位尺寸为设计变量，采用正交试验的设计方法选取样本点构建BP神经网络模型，利用遗传算法对支撑位置进行优化，使导轨基面呈现目标变形。结果表明，采用优化后的支撑方式，导轨基面变形和目标变形形状相同，且最大变形分别为7.982 8 μm和8.2 μm，两者之间的误差为0.21 μm，取得了较好的优化效果。
Aiming at the geometric error induced by guide surface deformation in precision machine tools assembly process, an idea of deformation pre-elimination is proposed. The final guide surface deformation after assembly can be obtained by the static analysis of the machine tools. Through the reproduction of the deformation in support state, it can be eliminated in advance. A precision horizontal machine center was taken as an example to elaborate this method, the key point of this method is to obtain the final deformation of guide surface in support state. Taking the position dimensions as the variables, a neural network was constructed with the samples which were chosen by the orthogonal experiment, and genetic algorithm was used to get the best locations of the supports, so the surfaces of guides can present the target deformation. The results show that using the optimized support, the static deformation of the guide surface and the target deformation present the same shape and the maximum deformation, which are 7.982 8 μm and 8.2 μm, respectively. The error between them is only 0.21 μm, which denotes that the optimization has good effect