一种基于STL文件的LBM前处理高效算法

An efficient algorithm for pre-processing of lattice Boltzmann method based on STL files

LBM是一种计算流体数值方法,计算过程中需要确定流场格点属于流体点还是固体点,通过前处理获得格点类型信息。其中一种方法是通过STL文件格式描述的物体模型判断格点在物体内外信息来确定格点类型,从而实现STL几何信息到LBM计算模型信息的重构。为了能够快速重构计算模型,本文提出了一种快速生成计算模型算法。该算法根据面三角形找到模型的边界点集,然后由边界点判定出模型内部点和外部点从而完成对模型的重建。算法从面三角形出发,大量减少了计算量,节约了时间,提高了效率。基于本文算法和直接法对不同复杂程度的圆球、NACA0012翼型、CHN-T1飞机标模开展了前处理建模比较,结果表明,随着几何复杂度增加,直接法耗时急剧增加,而本文算法一直保持低耗时(例如CHN-T1模型前处理网格生成,直接法采用120核并行运算耗时11 h,而本文算法采用单核仅耗时20 s),极大地提高了LBM针对复杂几何外形的前处理效率。使用基于本文算法前处理获得的网格开展流场计算,验证了该前处理方法的适用性。

Lattice Boltzmann method(LBM)is a method for computational fluid dynamic simulations.An important pre-processing procedure of LBM is to obtain the information of the types of the lattice so that the solid and fluid areas can be identified.One approach for this is to judge the lattice type based on the solid object model described in the STL file format,followed by a mapping from the STL geometry to the LBM computational model.In order to reconstruct the computational model more efficiently,the present study proposed a fast algorithm for the generation of the computational model.The algorithm finds the point set of the model boundary based on the surface triangles,and then determines the internal and external points of the model according to the boundary points.As the searching starts from the surface triangle,the algorithm significantly reduces the computation cost,saves the computational time and improves the efficiency.The model reconstruction efficiency is compared between the new algorithm and the direct method for a sphere,the NACA0012 airfoil and the CHN-T1 aircraft with different complexity,and the results show that the computational time of the direct method increases rapidly with the increase of geometric complexity,whereas the new algorithm always keeps low computational cost.For example,the direct method takes 11 hours on 120 cores to generate the pre-processing grid for the CHN-T1 model,while the new algorithm takes only 20 seconds on a single core.The new algorithm greatly improves the pre-processing efficiency of LBM for complex geometries.Numerical simulations are performed using the generated grid,and the simulation results further demonstrate the applicability of the new algorithm.