探讨数字图像相关方法测试结果的误差分布规律对提高该方法的测试精度具有重要意义。本文从理论上分析了基于双线性位移模式数字图像相关方法的子区变形场测试误差分布规律,结果表明,当试件的真实变形可由双线性位移模式描述时,变形场最大测试误差通常出现在子区的边界或节点处。因此,若试件发生均匀拉伸等常应变变形,可利用相关系数选取一个最优子区,认为测得最优子区中心应变为试件真实应变。零变形实验验证了该测试方法的可靠性。最后对手机导光板试件在单轴拉伸载荷下的数字图像进行分析,并利用本文方法测试了其弹性常数。
Error distribution investigation of measured results in digital image correlation method (DIC)is important to improve measurement accuracy. Measurement error distribution of deformation field in a sub-region by DIC was theoretically analyzed based on bilinear displacement mode. Results show that when actual deformation of a specimen can be described by bilinear displacement approximation, the maximum deformation measurement error usually appears at the boundary or the node of sub-region. Therefore, if specimen is subjected to constant strain deformation, such as in uniform tencile, an optimal sub-region can be selected by correlation coefficients, and measured strain at the center of optimal sub-region can be considerd as the true strain of specimen. Zero deformation experiment was used to verify the reliability of this measurement scheme. At last, digital images of a light guide plate in mobile phone subjected to uniaxial tensile load were analyzed, and its elastic constants was measured by above proposed method.