Numerical simulation study of LARCH software based on union energy grid method
-
摘要: LARCH是中广核研发的三维蒙特卡罗软件，兼顾了反应堆辐射屏蔽设计和反应堆核设计校算两大基本需求。介绍了在LARCH软件采用统一能量网格，该方法可以替代传统的二分查找方法和对数查找方法，减少粒子能量查找的次数和单次查找时间，在此基础上研发的优化delta-tracking算法，可以提高约25%的蒙卡软件堆芯临界计算效率。初步数值结果表明，与传统的蒙特卡罗软件相比，LARCH 1.0软件能够更高效地模拟反应堆问题。Abstract:
Background With the continuous development of nuclear power technology, reactor design has put forward higher requirements for the accuracy, efficiency and multi-functionality of nuclear computing software. The current mainstream Monte Carlo software has deficiencies in the balance between reactor radiation shielding design and nuclear design calibration, which restricts the critical simulation efficiency of the reactor core. Therefore, CNPRI has specifically developed the 3D Monte Carlo software LARCH 1.0 to meet the actual needs of nuclear power engineering design.Purpose To optimize the particle energy search mechanism in Monte Carlo simulation and address the pain point of low efficiency in traditional search methods; Thirdly, based on the optimized search method, the delta-tracking algorithm is further improved to enhance the efficiency of core critical calculation and provide efficient and accurate calculation support for reactor design.Method During the development of the LARCH software, the core technological innovation lies in the adoption of a unified energy grid design to replace the traditional binary search and logarithmic search methods. Through the standardization and unification of the energy grid, the number of searches in the particle energy matching process is reduced, and the time consumption of a single search is shortened. Based on the technology of unified energy grid, further develop and optimize the delta-tracking algorithm to achieve the improvement of computing efficiency; By designing a targeted numerical verification scheme, the LARCH 1.0 software and the traditional Monte Carlo software were compared and tested in the reactor problem simulation.Results The optimized technical solution has achieved remarkable results. The search method based on the unified energy grid has significantly reduced the time cost of particle energy search compared with the traditional method. Based on this, the optimized delta-tracking algorithm has increased the critical computing efficiency of the Monka software core by approximately 25%.Conclusions The unified energy grid method and the optimized delta-tracking algorithm adopted by the LARCH 1.0 3D Monte Carlo software provide an effective technical path for the efficiency improvement of the Monte Carlo software and significantly enhance the critical computing efficiency of the reactor core. The application potential of this software indicates that it can provide more efficient and reliable numerical simulation tools for reactor design. Subsequently, more extensive engineering verification and functional iterations will be further carried out.-
Key words:
- Monte-Carlo /
- Larch 1.0 /
- speed-up /
- union energy grid /
- computational efficiency
-
表 1 统一能量网格简化前后对比
Table 1. Comparison of Union Energy Grid
information energy bin/number memory usage/MB calculate time/s keff traditional union energy grid 668268 1927 4056 1.18192 simplified union energy grid 142342 820 2853 1.18199 Note: The verification problem is at the pin-cell level. Subsequently, the improvement in calculation efficiency was also verified at the assembly and core levels, with an overall 25% improvement in the simulation efficiency of criticality problems. 
下载: 导出CSV
表 2 查找方法对比
Table 2. Comparison of Search Methods
grid information binary search method energy box search method grid search time/s 901 154 Note: The verification problem is at the pin-cell level, with an 83% improvement in search efficiency.
下载: 导出CSV
表 3 采用不同输运方法的对比
Table 3. Comparison of Two Transportation Methods
result traditional surface tracking method hybrid transport method keff 1.18184 1.18169 cell1-volume flux(cm3/s) 1.926 1.925 cell2-volume flux(cm3/s) 0.08369 0.08373 cell3-volume flux(cm3/s) 0.6214 0.6204 cell4-volume flux(cm3/s) 3.571 3.568 calculate time/s 859 409 Note: The verification problem is at the pin-cell level, with a 52% improvement in calculation efficiency.
下载: 导出CSV
表 4 VERA基准题1信息
Table 4. VERA benchmark question1 detail imformation
id fuel temperature (K) moderator density (g/cc) moderator temperature (K) 1B 600K 0.661 600 1C 900K 0.661 600 1D 1200K 0.661 600 1E 600K 0.743 600
下载: 导出CSV
表 5 VERA基准题2信息
Table 5. VERA Benchmark Question2
id fuel temperature (K) moderator density (g/cc) moderator temperature (K) 2B 600 0.661 600 2C 900 0.661 600 2D 1200 0.661 600 2E 600 0.743 600 2F 600 0.743 600 2G 600 0.743 600 2H 600 0.743 600 2I 600 0.743 600 2J 600 0.743 600 2K 600 0.743 600 2L 600 0.743 600 2M 600 0.743 600 2N 600 0.743 600 2O 600 0.743 600 2P 600 0.743 600
下载: 导出CSV
表 6 VERA基准题验证结果
Table 6. VERA Benchmark Question Verification Result
id keff/pcm deviation/pcm id keff/pcm deviation/pcm LARCH 1.0 KENO-VI LARCH 1.0 KENO-VI 1B 1.18187 1.18215 5 2H 0.78855 0.78822 33 1C 1.17220 1.17172 48 2I 1.17955 1.17992 −37 1D 1.16331 1.16260 71 2J 0.97520 0.97519 1 1E 0.77106 0.77169 −63 2K 1.02005 1.02006 −1 2B 1.18302 1.18336 −34 2L 1.01868 1.01892 −24 2C 1.17315 1.17375 −60 2M 0.93808 0.93880 −72 2D 1.16635 1.16559 76 2N 1.04795 1.04773 22 2E 1.06970 1.06963 7 2O 0.92788 0.92741 47 2F 0.97664 0.97602 62 2P 0.78855 0.78822 33 2G 0.84813 0.84769 44
下载: 导出CSV
表 8 统一能量网格优化对比
Table 8. Optimization comparison table of union energy grid
problem binary search keff binary search time tb/min union grid keff union grid time tu/min speed up ratio (1−tu/tb) VERA 1B 1.18187 41.95 1.18199 30.10 28.2% VERA 2M 0.93808 22.03 0.93815 16.41 25.5% core 0.99687 7465.56 0.99686 5638.81 24.5%
下载: 导出CSV
-
[1] Briesmeister J F. MCNP: a general Monte Carlo N-particle transport code[R]. LA-13709-M, 2000. [2] Kulesza J A, Adams T R, Armstrong J C, et al. MCNP® code version 6.3. 1 theory & user manual[R]. LA-UR-24-24602, 2024. [3] Wang Kan, Li Zeguang, She Ding, et al. RMC-A Monte Carlo code for reactor core analysis[J]. Annals of Nuclear Energy, 2015, 82: 121-129. doi: 10.1016/j.anucene.2014.08.048 [4] 丘意书, 佘顶, 范潇, 等. 堆用蒙特卡罗程序RMC的全堆计算研究[J]. 核动力工程, 2013, 34(s1): 1-4,23 doi: 10.3969/j.issn.0258-0926.2013.z1.001Qiu Yishu, She Ding, Fan Xiao, et al. Analysis of full-core calculation of RMC[J]. Nuclear Power Engineering, 2013, 34(s1): 1-4,23 doi: 10.3969/j.issn.0258-0926.2013.z1.001 [5] Romano P K, Forget B. The OpenMC Monte Carlo particle transport code[J]. Annals of Nuclear Energy, 2013, 51: 274-281. doi: 10.1016/j.anucene.2012.06.040 [6] Romano P K, Horelik N E, Herman B R, et al. OpenMC: a state-of-the-art Monte Carlo code for research and development[J]. Annals of Nuclear Energy, 2015, 82: 90-97. doi: 10.1016/j.anucene.2014.07.048 [7] Leppänen J. Two practical methods for unionized energy grid construction in continuous-energy Monte Carlo neutron transport calculation[J]. Annals of Nuclear Energy, 2009, 36(7): 878-885. doi: 10.1016/j.anucene.2009.03.019 [8] Leppänen J. Development of a new Monte Carlo reactor physics code[R]. VTT Publications 640, 2007. [9] Liu Yuxuan, She Ding, Wang Kan, et al. Optimization treatment of point-wise nuclear data in Monte Carlo criticality and burnup calculations[J]. Annals of Nuclear Energy, 2011, 38(7): 1489-1495. doi: 10.1016/j.anucene.2011.03.016 [10] She Ding, Liu Yuxuan, Wang Kan, et al. Development of burnup methods and capabilities in Monte Carlo code RMC[J]. Annals of Nuclear Energy, 2013, 51: 289-294. doi: 10.1016/j.anucene.2012.07.033 [11] Leppänen J. Performance of Woodcock delta-tracking in lattice physics applications using the Serpent Monte Carlo reactor physics burnup calculation code[J]. Annals of Nuclear Energy, 2010, 37(5): 715-722. doi: 10.1016/j.anucene.2010.01.011 [12] Morgan L W G, Kotlyar D. Weighted-delta-tracking for Monte Carlo particle transport[J]. Annals of Nuclear Energy, 2015, 85: 1184-1188. doi: 10.1016/j.anucene.2015.07.038 [13] Kowalski M A, Cosgrove P M. Acceleration of surface tracking in Monte Carlo transport via distance caching[J]. Annals of Nuclear Energy, 2021, 152: 108002. doi: 10.1016/j.anucene.2020.108002 [14] Liu Minyun, Ma Yugao, Guo Xiaoyu, et al. An improved tracking method for particle transport Monte Carlo simulations[J]. Journal of Computational Physics, 2021, 437: 110330. doi: 10.1016/j.jcp.2021.110330 [15] Godfrey A. VERA core physics benchmark progression problem specifications[R]. Revision 4, CASL Technical Report: CASL-U-2012-0131-004, 2014. -
点击查看大图
图(5) / 表(8)
计量
- 文章访问数: 134
- HTML全文浏览量: 101
- PDF下载量: 3
- 被引次数: 0
