Massively Parallel Method of Characteristics Neutron Transport Calculation with Anisotropic Scattering Treatment on GPUs

Author(s):  
Namjae Choi ◽  
Junsu Kang ◽  
Han-gyu Joo
Author(s):  
Liang Liang ◽  
Hongchun Wu ◽  
Liangzhi Cao ◽  
Youqi Zheng

The method of characteristics (MOC) has been widely used in lattice code for its high precision and easy complement. However, the long characteristics method needs large quantity of PC memory when dealing with large scale problems. The modularity MOC method could significantly reduce the PC memory when calculating the problem which contains lots of repeatedly geometries, like the fuel assembly in the reactor. In this method, only typical geometric cells are selected to trace the rays, and then the geometry information of these cells is stored. So, the modularity MOC method is feasible to perform well in the calculation with large scale. When tracing the rays, the technique of mesh ray generating and the corresponding azimuthal quadrature set are both applied. The techniques make sure that each ray has the reflected ray in the boundary so it is convenient to describe the boundary condition. The optimal polar angle and the Guass quadrature set are selected as the polar quadrature set. Furthermore, the coarse mesh finite difference (CMFD) is employed to accelerate the calculation. A pin cell is chosen as the coarse mesh. The CMFD solution provides the MOC with much faster converged fission and scattering source distributions. The LOTUS code is developed and the numerical results show that the code is precise for engineering application and the CMFD acceleration is effective.


2011 ◽  
Vol 38 (2-3) ◽  
pp. 447-454 ◽  
Author(s):  
Zhouyu Liu ◽  
Hongchun Wu ◽  
Liangzhi Cao ◽  
Qichang Chen ◽  
Yunzhao Li

2019 ◽  
Vol 131 ◽  
pp. 257-272 ◽  
Author(s):  
Peitao Song ◽  
Zhijian Zhang ◽  
Liang Liang ◽  
Qian Zhang ◽  
Qiang Zhao

2014 ◽  
Vol 38 (5) ◽  
pp. 058201 ◽  
Author(s):  
Zhen-Ping Chen ◽  
Hua-Qing Zheng ◽  
Guang-Yao Sun ◽  
Jing Song ◽  
Li-Juan Hao ◽  
...  

2019 ◽  
Vol 34 (1) ◽  
pp. 13-28
Author(s):  
Donghoon Kim ◽  
Lucas Albright ◽  
Brittney Saenz ◽  
Tatjana Jevremovic

A novel polar angular quadrature set called the Kim-Jevremovic polar angular quadrature set is derived for the method of characteristics. It is based on neutron anisotropic scattering cross-sections in the Evaluated Nuclear Data File. This new set is implemented within the state-of-the-art neutron transport code AGENT and tested in comparison to MCNP6 as well as to other known quadrature sets for the UO2 unit cells, the well-known C5G7 benchmark, unreflected cylinders of uranyl-fluoride solutions in heavy water, and the University of Utah 100 kWth TRIGA MARK-I reactor core. These comparisons show that the newly proposed polar angular quadrature set provides better agreements than other quadrature sets for the lower order of anisotropic scattering expansions. This paper presents a complete derivation of the Kim-Jevremovic polar angular quadrature set and the analysis for the mentioned bench-mark examples.


Sign in / Sign up

Export Citation Format

Share Document