基于数字图像关联技术的沥青应变计

Measurement Accuracy Comparison of InSitu Asphalt Pavement Strain Sensors Based on Digital Image Correlation Technique

摘要:

提出一种基于DIC(digital image correlation)技术的沥青混合料表面平均应变测量方法，以之作为基准比对不同种类沥青应变计的测量准确度.首先基于DIC技术实现试件表面亚像素级位移场识别，进而研发基于二次曲线拟合的沥青混合料平均应变基准值算法.校准该算法后比对两种常用应变计的测量准确度.研究结果表明，有不低于95%的置信度认定DIC法测量应变误差小于12×10-6，通过比较应变计测量结果与DIC法平均应变基准值之间的差异，能够比对不同应变计测量准确度优劣.此外研究结果也证实，荷载模式对测量准确度具有影响，移除塑性应变或建立校准方程校准应变计测量结果均能够提高应变计测量准确度.

Abstract:

A method to measure strain distribution on the asphalt mixture surface was developed based upon DIC (digital image correlation) in order to compare the measurement accuracy of insitu asphalt pavement strain sensors. First, the displacement was distinguished at subpixel level using the DIC technique. Then an algorithm to calculate average tensile strain was proposed to eliminate the adverse effect caused by asphalt mixture inhomogeneous deformation in strain measurement. After testing the reliability of the algorithm, the measurement accuracy of two common insitu asphalt pavement strain sensors were calibrated. It was found that the error of the DIC method was less than 12×10-6 at the 90% confidence level, which met the requirement of accuracy comparison. Besides, it was proved that the measurement accuracy could be reflected by the error between the measurement result of the sensors and the method developed in this paper. It was also proved that load type had an influence upon sensor’s accuracy. Utilization of calibration function or eliminatation of the cumulative plastic strain all contributed to the improvement of accuracy.