On the accuracy of the random walk simulation of pollutant dispersion
In fluid, mass transport phenomena exist widely and have thus attracted researches for decades.Numerical methods are often used to study them.The Eulerian and the Lagrangian approaches are the two options in numerical simulations.Compared with the Eulerian approach, the Lagrangian approach has advantages in stability and simplicity, but suffers from the heavy computational cost.The random walk method we used in this paper belongs to the Lagrangian approach.To control the computational cost without losing accuracy, we introduce a linear model that helps predict the accuracy of the random walk method in several computational configurations.We conducted parametric studies and showed that the linear model introduced can explain satisfactorily the computational error, which is measured by the mean-squared-error (MSE) in all the cases we tested in this paper.Extra tests were carried out to validate the performance of the linear model established.Furthermore, we find a constant dimensionless coefficient in the linear model which links the MSE and the number of particles in different flow conditions.
Random walk method pollutant transport dispersion linear regression error analysis
Xuefei Wu Dongfang Liang
Department of Engineering, University of Cambridge Cambridge, UK MOE Key Laboratory of Hydrodynamics, School of Naval Architecture, Ocean and Civil Engineering,Shang
国际会议
The Second Conference of Global Chinese Scholars on Hydrodynamics(第二届全球华人水动学学术会议)
江苏 无锡
英文
972-978
2016-11-11（万方平台首次上网日期，不代表论文的发表时间）