论文部分内容阅读
内存不足是蒙特卡罗方法大规模输运模拟的关键问题。对于反应堆燃耗分析,需在输运过程中统计大量反应截面数据,计算机内存限制了燃耗计算规模。本文基于反应堆蒙特卡罗程序(RMC),利用数据分解方法对计数器数据并行存储,并与点燃耗并行耦合,实现计数器数据分解和燃耗数据分解的综合并行方法。对全堆基准题进行数值测试,结果表明综合并行方法可明显降低计算内存,验证了数据分解对蒙特卡罗大规模燃耗分析的有效性。
Insufficient memory is a key issue for the Monte Carlo approach to large-scale transport simulations. For the analysis of reactor burnup, a large amount of reaction cross-section data needs to be counted during transport, and the computer memory limits the calculation scale of fuel burnup. Based on the reactor Monte Carlo (RMC) method, a data decomposition method is used to store the counter data in parallel and couple it with the ignition loss in parallel to achieve a comprehensive parallel method of counter data decomposition and fuel consumption data decomposition. The numerical test of the full stack of benchmark questions shows that the comprehensive parallel method can significantly reduce the computational memory and validate the effectiveness of the data decomposition on Monte Carlo large-scale fuel consumption analysis.