arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1410.0010astro-ph.IM
收藏

ENZO的自适应粒子网格重力求解器

An adaptive particle-mesh gravity solver for ENZO

Jean-Claude Passy, Greg L. Bryan

摘要:

我们描述并实现了一个自适应粒子网格算法来解决泊松方程的网格为基础的流体动力学代码与嵌套网格。该算法的实施和广泛的测试内的天体物理代码恩佐对多重网格求解器默认可用。我们发现,虽然这两种算法表现出类似的精度光滑的质量分布,自适应粒子网格算法是更准确的点群众的情况下,一般是少噪音。我们还表明,两体问题可以准确地解决了嵌套网格的配置。此外,我们还讨论了子循环的影响,并证明了以相同的时间步长发展所有级别会产生更高的精度。

英文摘要:

We describe and implement an adaptive particle-mesh algorithm to solve the Poisson equation for grid-based hydrodynamics codes with nested grids. The algorithm is implemented and extensively tested within the astrophysical code Enzo against the multigrid solver available by default. We find that while both algorithms show similar accuracy for smooth mass distributions, the adaptive particle-mesh algorithm is more accurate for the case of point masses, and is generally less noisy. We also demonstrate that the two-body problem can be solved accurately in a configuration with nested grids. In addition, we discuss the effect of subcycling, and demonstrate that evolving all the levels with the same timestep yields even greater precision.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑