PIV技术在非饱和冻土冻胀模型试验中的实现与灰度相关性分析

Application of PIV in model tests on frozen unsaturated soils and grayscale correlation analysis

粒子图像测速法（PIV）已经在流体力学和岩土工程中得到广泛应用,并取得了良好的试验结果。但是,由于黏性土缺乏表面纹理,PIV技术尚未在冻结黏性土颗粒迁移分析中得到成功应用。针对黏性土冻结过程中的土颗粒迁移问题,采用自制的PIV冻土模型试验箱,以非饱和粉质黏土作为试验土样,结合Canon EOS 1300D相机和GeoPIV软件,提出了PIV技术在非饱和冻土中的实现方法。结果表明：（1）试验过程中光线的变化会极大的影响试验结果,故需要在摄影棚中构建恒定光场;（2）为便于PIV分析,试验前需对冻土进行表面纹理构建,粒径0.3 mm石英砂是纹理构建的最佳示踪粒子;（3）非饱和粉质黏土在冻结过程中的冰水相变会导致图像灰度变化,根据统计结果,将种子区域灰度相关性及整体灰度相关性分别调整为0.85和0.69。

PIV is a velocity-measuring technique commonly used in the field of experimental fluid mechanics and geotechnical tests, but few previous reports are found on the use of this technique in frozen silty clay（especially unsaturated condition） due to the lack of surface texture. The self-made model test apparatus for frozen soil, Canon Eos camera and GeoPIV software, are used to analyze the movement of soil particles. The experiment results demonstrate that：（1） The constant light area is needed to be built in studio which significantly affects the PIV results.（2） Due to the insufficient surface texture of clayey soils, PIV cannot be used. Therefore, to allow the analysis, some materials are used to make the artificial texture of silty clay and the silica particles with a diameter 0.3 mm are the best tracking particles.（3） Change of ice-water phase during freezing will cause the grayscale variation of silty clay images inevitably. The minimum acceptable correlation coefficient tolerance for seed is 0.85 and the minimum correlation coefficient tolerance is 0.69 according to the statistical results.