Massively Parallel, Three-Dimensional Transport Solutions for the k-Eigenvalue Problem

Integrated Radiation Transport and Nuclear Fuel Performance for Assembly-Level Simulations

“…Although the Denovo run time went up significantly, the increase is largely explained by a significant increase in the number of iterations required (from 15 to 27), possibly due to inadequate selection of subspace dimensions for the larger problem. It does not indicate a lack of scalability on the part of Denovo, because it has been shown to scale well on similar problems in previous studies [5]. The time required to complete the AMPFuel portion of the calculation actually decreased from the coarse mesh to the fine mesh.…”

“…It has the capability to solve the two-dimensional (2D) and three-dimensional (3D) deterministic multigroup discrete ordinates equations. A detailed description of the parallel algorithms and solvers can be found in References [5] and [6].…”

“…The accuracy of Denovo has been verified for nuclear reactor simulation through the use of standard numerical benchmarks for radiation transport [16]. Additional parallelism in the form of an energy decomposition is also available, extending the parallel scalability to hundreds of thousands of processors [5,6]. Numerous spatial discretizations are available, as well as a variety of angular quadrature sets and boundary condition options.…”

Piecewise Diffusion Synthetic Acceleration scheme for neutron transport simulations in optically thick systems

“…Even though it ignores the time variable, the 3D stationary BTE is still set in a 6-dimensional phase space (3 for space, 2 for travel direction and 1 for energy). Its discretization at the scale of the full reactor core therefore quickly produces very large problems of size in the order of 10 10 to 10 12 degrees of freedom, whose solution has remained mainly out of reach before the early 2010s [4,5,6], when large enough supercomputers became available, along with numerical methods able to efficiently harness them. Devising and implementing parallel methods able to efficiently solve the transport equation for such large problems is in itself no easy task, the major difficulty lying in the fact that the hyperbolic nature of the transport equations implies dependencies between cells.…”

Iterative stability analysis of spatial domain decomposition based on block Jacobi algorithm for the diamond-difference scheme

“…To solve this type of numerical problems one relies on massively parallel computations using high-performance computers. Parallel methodologies for transport problems are based on various transport sweeping algorithms and domain decomposition methods [3][4][5][6][7][8][9][10]. Let us consider that the spatial domain D of the transport problem (1) is divided into a number of subdomains D l (D = l D l ).…”