摘要
通过引入粒子计数法,对基于二进制图像分析算法(BICC)的粒子图像速度场仪(PIV)的粒子图像跟踪技术进行了改进.数值模拟了3个典型的两相流稀相微粒流场,即等线速度旋转流场、旋涡流场和两个相向运动的射流汇合流场,并把改进后的微粒跟踪技术用于模拟流场中粒子对的识别研究.结果表明:改进后的方法能很好地识别与跟踪粒子对,计算速度快,准确率高达97%,适用于两相流稀相微粒速度场的测量;随着粒子对数目的增大,相关系数降低;随着粒子半径的减小,相关系数分布渐趋陡峭;最佳粒子对数目为8～12,最佳粒子图像半径为3～8像素.
In the image analysis technique of particle image velocimetry (PIV) based on the binary image crosscorrelation (BICC) algorithm, the correlation coefficient expression is simplified using the count method. Three dilutephase particle velocity fields in the fluidparticle twophase flow are numerically simulated, i.e. rotary flow with equal line speed, vortex flow, and two opposite mixing jet flow. These simulated flow fields are then adopted to verify the improved algorithm. It is shown that the new algorithm has better ability in tracing the particle couples. The computing speed is higher and the probability of error is less that 3%. The correlation coefficient rises with increasing particle couple number, and the coefficient distribution becomes sharper with decreasing particle radius. The optimal particle couple number is 8～12 and the perfect radius of the particle image is 3～8 pixels under the above condition.
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2003年第7期734-737,共4页
Journal of Xi'an Jiaotong University
关键词
粒子图像速度场仪
粒子对识别
两相流
particle image velocimetry
particle couple identification
two-phase flow
