【作者】 高岩；

【导师】 周明；

【作者基本信息】 哈尔滨工业大学 ， 机械制造及其自动化， 2009， 硕士

【摘要】 KDP晶体是一种优质的非线性材料,广泛应用于激光和非线性光学领域,尤其在惯性约束聚变固体激光驱动器、强激光武器及核武器等关键设备所需的众多光学元器件中,因此对KDP晶体的加工表面质量要求很高。但由于KDP晶体本身的机械与物理特性,使其在传统加工中表现出强烈的各向异性,难以达到要求的表面质量。目前对晶体加工表面质量的研究多数限于表面质量的单一参数的控制,因此多参数综合评价晶体加工表面质量的研究工作具有重要的意义。同时随着对KDP晶体材料需求的不断增加,研究在一定表面质量范围内晶体加工的高效性也对生产具有现实指导意义。为研究晶体加工表现出的各向异性,分析了KDP晶体本身的结构和物理机械特性,应用弹性力学的理论计算出了其各个晶面的弹性模量和剪切模量;为确定最佳切削晶向,进行晶体的超精密车削试验,通过分析试验数据,综合评价粗糙度、残余应力,最终确定切削加工的最佳晶向为40°晶向。确定了SPDT(单点金刚石切削)切削试验的变量参数及相应水平值,设计了正交试验方案,采用原子力显微镜测量表面粗糙度、白光干涉仪测量亚表面损伤深度、低载荷原位纳米压痕力学系统测量残余应力,获得了刀具参数和工艺参数对粗糙度、亚表面损伤深度和残余应力的影响规律,确定了各参数对表面质量影响的主次关系。以单点金刚石切削试验数据为训练样本及预测样本,设计了神经网络拓扑结构,建立了人工神经网络预测模型,通过对预测结果进行误差分析,得出最大相对误差不超过5%,证明该神经网络的预测精度较高,从而实现了对不同加工参数形成的表面质量的预测,降低了生产成本;建立了表面质量的目标函数,确定约束条件,选择优化方法,对加工参数进行了优化,得出在粗糙度不超过10nm、亚表面损伤不超过100nm、残余应力不超过150 Mpa的表面质量条件下,最佳的切削参数为主轴转速n=386r/min、进给量f=43μm/s、背吃刀量ap=9μm和刀具前角γ0=-36.5°,从而实现高表面质量、高效的加工。更多还原

【Abstract】 KDP crystal is a kind of high-quality non-linear material, which is widely applied in the laser and nonlinear optical fields, especially in optical components of inertial confinement fusion solid laser driver, laser weapon and Nuclear weapons. so it demands high machined surface quality of KDP crystal. But because of its mechanical and physical characteristics, it showed strongly anisotropic in the traditional processing that it is difficult to meet the requirements of the surface quality.In order to research anisotropic of crystal processing, this paper has analyzed the structure and physical and mechanical properties of KDP crystal, applied the theory of elastic mechanics to calculate the Young’s modulus and shear modulus of each crystal face. To determine the optimal cutting crystallographic orientation, this paper has executed ultra-precision turning experiment. Through the analysis of experiment data, the roughness and residual stress has been evaluated by synthesis and determined that the best cutting crystallographic orientation is 40°.The parameter of the SPDT (single point diamond turnning) and the corresponding level have been determined. Orthogonal experiment scheme has been designed. This paper has been used atomic force microscope to measure surface roughness, white interferometer to measure and low load in situ nano creasing mechanical system to measure residual stress. the influence law of tool parameters and process parameters on the surface roughness, subsurface damage depth and residual stress has been obtained. Moreover this paper has determined the primary and secondary relation of influence on the surface quality to various parameters.With single-point diamond turnning test data for the training sample and predicting sample, this designed neural network topology structure, established the artificial neural network model for prediction. Through the error analysis of predicted results, obtained that the maximum relative error is less than 5%. It proves that the prediction precision of this neural network is enough. so we can achieve the high precision of different machining parameters for surface quality and reduce the production costs, This paper has established the objective function of the surface quality, choosed optimization method and determined the constraint conditions. It has achieved optimization of processing parameters, obtained the best cutting parameters which are spindle speed n= 386r/min, feed rate f=43μm/s, cutting depth ap=9μm and tool orthogonal rakeγ0=-36.5 degrees under surface quality constraint conditions that are roughness is no more than 10nm, subsurface damage does not exceed 100nm and residual stress is no more than 50Mpa. Thus, this paper achieved high surface quality and high efficient processing.更多还原