RMC程序敏感性分析功能的并行策略与验证

Parallel Strategies and Verification of Sensitivity Analysis Function of RMC Code

针对堆用蒙特卡罗程序（RMC）中的有效增殖系数（keff）敏感性分析功能,为提高计算效率并降低内存占用,在RMC中实现了分别基于下一代裂变中子数以及中子产生率的伴随注量率估计方法的2种并行策略,并使用无限均匀介质多群算例和连续能量算例对该功能进行验证。结果表明,2种并行策略所求得的敏感性系数均与解析解、蒙特卡罗粒子输运程序（MCNP6）的计算结果吻合良好,计算速度为MCNP6的3倍左右,统计不同核素的总截面敏感性系数的品质因子为MCNP6的4~8倍左右。

In order to improve the computational efficiency and reduce the memory usage of the function in the Reactor Monte Carlo code RMC to compute the effective multiplication factor （keff） sensitivity coefficients with regard to nuclear data, two parallel strategies which are based on two ways to estimate the adjoint flux, namely, the next generation neutron number estimator and the neutron production rate estimator, are implemented in RMC. A multi-group infinite-medium test and a continuous-energy test are used to verify the new strategies. Results show that the sensitivity coefficients computed by the two parallel strategies agree well with the analytic solutions and those computed by MCNP6. Furthermore, the new strategies run 3 times as fast as MCNP6 and the figure of merits of the new strategies to compute the sensitivity coefficients to total cross section of different isotopes are 4 to 8 times as high as MCNP6.