ANS is committed to advancing, fostering, and promoting the development and application of nuclear sciences and technologies to benefit society.
Explore the many uses for nuclear science and its impact on energy, the environment, healthcare, food, and more.
Explore membership for yourself or for your organization.
Conference Spotlight
2026 Nuclear Energy Conference & Expo (NECX)
August 24–27, 2026
Dallas, TX|Hilton Anatole
Latest Magazine Issues
Jun 2026
Jan 2026
2026
Latest Journal Issues
Nuclear Science and Engineering
July 2026
Nuclear Technology
Fusion Science and Technology
May 2026
Latest News
Breaking ground on a new approach to construction
The drive to Kairos Power’s reactor demonstration site in Oak Ridge, Tenn., is not only scenic—it’s historic. Nearly 85 years ago, roughly 30,000 construction workers transformed orchards and farmland into a key Manhattan Project site. Depending on your route, you may pass by one of the three gatehouses that were once military checkpoints controlling access to Atomic Energy Commission production facilities.
Hussein Khalil
Nuclear Science and Engineering | Volume 90 | Number 3 | July 1985 | Pages 263-280
Technical Paper | doi.org/10.13182/NSE85-A17768
Articles are hosted by Taylor and Francis Online.
A diffusion theory method is developed for synthetic acceleration of nodal Sn calculations in multidimensional Cartesian geometries. The diffusion model is derived from the spatially continuous diffusion equation by applying spatial approximations that are P1 expansions of the corresponding approximations made in solving the transport equation. The equations of the diffusion model are formulated in a way that permits application of existing and highly efficient nodal diffusion theory techniques to their numerical solution. Test calculations for several benchmark problems in X-Y geometry are presented to illustrate the efficiency and stability of the acceleration method when applied to a “constant-linear” nodal transport approximation. The method is shown to yield point-wise flux convergence of 10-4 in fewer than ten synthetic iterations for all problems considered and to require substantially less computational effort than unaccelerated solutions.