基于能量密度分布的辐射源粒子空间抽样方法研究∗

摘要

利用隐式蒙特卡罗方法模拟热辐射光子在物质中的输运过程时，物质辐射源粒子是需要细致处理的物理量。传统的物质辐射源粒子抽样方法是体平均抽样方法，对于大多数问题，这样处理不会带来大的偏差。但是对于一些辐射吸收截面大、单一网格内温差显著的问题，体平均抽样方法的计算结果偏差较大。分析了产生偏差原因，提出一种基于辐射能量密度分布的辐射源粒子空间位置抽样方法，并推导了相应的抽样公式以解决此类问题。数值实验表明，新方法计算结果明显优于原方法且与解析结果基本一致。%When using the implicit Monte Carlo method to simulate the thermal radiative photon transfer in materials, the radiative source particles emitted by the material should be treated carefully. In traditional sampling method for the location of radiative source particles can be regarded as uniform in a single cell, and this is reasonable for most of thermal radiative transfer problems. But it may result in unacceptable error in the problem which has a large absorption cross-section and remarkable temperature difference in a single cell. The reason for this error is discussed in this paper. A new method based on radiation energy density distribution is proposed and the sampling formula is derived. Numerical test indicates that the new method will give results much better than the old method and in agreement with the analytical results.