An efficient separated method for fluid flows on collocated grid, named EPPL(Error Production and Propagation Limited), has been proposed involving robustness and convergence. The main features of the method include changing the solution procedure according to the matching degree of velocity and pressure dynamically, optimizing the under-relaxation strategy according to the characteristics of momentum interpolation, and enhancing the robustness of pressure equation. In this paper, the basic ideas of the three improvements and the overall solution procedure of the EPPL algorithm are introduced in details. Theoretical support for each improvement is provided by three benchmark compressible/incompressible flow tests, demonstrating the superior performance of the EPPL algorithm. Tests show that the EPPL algorithm achieves significant improvements in both robustness and convergence. In addition, the EPPL algorithm is simple and easy to implement, and the ideas in it are very suitable for porting to existing codes. Finally, the real performance of EPPL for complex case is demonstrated by simulating the interaction of jet and supersonic cross-flow in the background of fuel injection of scramjet.

提出了一种有效的分离方法，用于在并置网格上进行流体流动，该方法称为鲁棒性和收敛性，称为EPPL（错误产生和传播有限公司）。该方法的主要特点包括根据速度和压力的匹配程度动态改变求解程序，根据动量插值的特点优化欠松弛策略，增强压力方程的鲁棒性。本文详细介绍了这三种改进的基本思想以及EPPL算法的整体求解过程。三种基准可压缩/不可压缩流量测试为每种改进提供了理论支持，证明了EPPL算法的优越性能。测试表明，EPPL算法在鲁棒性和收敛性上均取得了显着改善。此外，EPPL算法简单易行，并且其思想非常适合移植到现有代码。最后，通过在超燃冲压燃料喷射的背景下模拟射流和超音速横流的相互作用，证明了EPPL在复杂情况下的真实性能。