Abstract

This study focuses on the calculation of the energy deposition in the Transformational Challenge Reactor by two major Monte Carlo codes: Serpent and MCNP. The first software computation relies on Kinetic Energy Released per unit Mass (KERMA) factors while the second one relies on Q-values. The results from these two independent computation methodologies are in very good agreement; however, Serpent runs much faster than MCNP (for the same computational model) and allows for a detailed energy deposition distribution from a 1-mm-side square mesh with a relative statistical error between 0.5% and 1%. This detailed energy deposition is suitable for multiphysics analyses aimed at design optimizations. In order to calculate the energy deposition, Serpent needs enhanced ACE files (distributed by the software developers). Unlike other Monte Carlo software that uses inputs based on Python or Java languages, the Serpent input syntax is very similar to that of MCNP; a Python script can convert a MCNP input to a Serpent input in seconds. For simulations not requiring the calculation of the energy deposition, Serpent can also read nuclear data from MCNP ACE files, which eventually improves the comparison of the results of the two codes.

摘要

本研究的重点是通过两种主要的蒙特卡罗代码计算转化挑战反应堆中的能量沉积：Serpent 和 MCNP。第一个软件计算依赖于每单位质量释放的动能 (KERMA) 因子，而第二个依赖于 Q 值。这两种独立计算方法的结果非常一致；然而，Serpent 的运行速度比 MCNP（对于相同的计算模型）快得多，并且允许从 1 毫米边的方形网格获得详细的能量沉积分布，相对统计误差在 0.5% 到 1% 之间。这种详细的能量沉积适用于旨在优化设计的多物理场分析。为了计算能量沉积，Serpent 需要增强的 ACE 文件（由软件开发人员分发）。与其他使用基于 Python 或 Java 语言输入的 Monte Carlo 软件不同，Serpent 输入语法与 MCNP 的语法非常相似；Python 脚本可以在几秒钟内将 MCNP 输入转换为 Serpent 输入。对于不需要计算能量沉积的模拟，Serpent 还可以从 MCNP ACE 文件中读取核数据，最终改善了两种代码结果的比较。