喷雾仿真中的网格依赖性

基于离散油滴模型,比较了原始喷雾模型与修正喷雾模型的网格依赖性.计算结果表明:原始喷雾模型由于采用O'Rourke碰撞模型以及最近节点法耦合算法,使得预测结果出现明显的网格依赖性.相反,修正喷雾模型采用独立的拉格朗日碰撞单元降低碰撞过程的网格依赖性,联合使用拉格朗日插值算法以及气相射流模型降低气液耦合过程对于网格的依赖性.因此修正喷雾模型能获得较好的低网格依赖性从而保证粗大网格下的计算精度. 

Investigation of the grid dependency from the spray simulation

Based on the discrete droplet model, the grid dependencies of the original spray model and revised spray model were compared. The results show that the apparent grid dependency exists in the simulated results with original spray model due to the application of O'Rourke collision model and momentum coupling algorithm with nearest node. However, the separate Lagrangian collision mesh is utilized to reduce the mesh dependency from the collision prediction, and the Lagrangian interpolation coupling algorithm and the gas-jet model are combined to cut down the mesh dependency from the momentum coupling calculation. Therefore, the revised spray model successfully achieves significant lower grid dependency so that the simulated accuracy of coarse grid is improved.