中文摘要：

用白光干涉检测（WLI）法检测微结构表面形貌时,其光滑结构面和边角数据很容易丢失,因此本文提出了自适应方向WLI检测法.该方法分别沿着每个微斜面法向进行自适应方向检测,评价分析其面形、特征轮廓和特征点的加工精度.首先,采用# 3000金刚石砂轮微细尖端在Si表面加工出高度为50 μm、宽度为56 μm且光滑的微锥塔结构.然后,利用四次检测点云对微磨削表面进行拼接与重建.最后,分析面形误差、特征轮廓误差和特征点误差.实验显示：自适应方向WLI检测可以重构出完整的微锥塔结构表面,微磨削的面形误差为5.3 μm,表明该微磨削技术可以确保微锥塔结构光滑Si表面的加工精度.但是,对微锥塔结构表面特征轮廓误差及特征点误差的评价表明,特征轮廓误差高达7.7μm,而特征点误差约为15 μm,约为面形误差的3倍.分析认为这些误差是由微金刚石砂轮V形尖端磨钝及微磨粒钝化造成的.

英文摘要：

When a White Light Interferometry （WLI） is used to measure the micro-structured surface topography,the measured points on smooth micro-structured surfaces and edges are easy to lose.Therefore,an adaptive-orientation WLI was proposed.The method measured each steep surface along its near normal-orientation for the evaluation of its surface shape,feature profile and micro-machining accuracy.First,a # 3000 diamond wheel V-tip was employed to fabricate a micro-pyramid-structure Si surface with a depth of 50 μm and a width of 56 μm.Then,four adaptive-orientation measured point clouds were used to splice and reconstruct whole micro ground surface.Finally,the microground form error,feature profile error and the feature point error were investigated.Experiments show that the adaptive-orientation measurement can reconstruct integrated micro-pyramid-structured surface and the form error is 5.3 μm,which means that the micro-grinding may assure the machining accuracy of micro-pyramid-structured Si surface.In contrast,the evaluation for the feature profiles and feature points of a micro-pyramid-structure indicates that the feature profile error reaches 7.7μm,and the feature point error is about 15 μm,which is about 3 times of its form error.It suggests that these errors come from the passivations of diamond wheel V-tip and micro grains.