Performance of ANL hexagonal-geometry nodal diffusion methods
- Argonne National Lab., IL (United States)
The DlF3D nodal scheme was the first widely used hexagonal-geometry extension of the transverse integration nodal approach. This scheme is extremely effective for its primary application (fast reactors), its accuracy typically being similar to that of finite difference (FD) solutions employing 24 triangular mesh per hexagon and a factor of {approximately} 5 finer axial mesh. For applications requiring a finer FD mesh, e.g., some thermal systems, the accuracy of DIF3D nodal is thus likely to be insufficient. This accuracy limitation stems from approximations made not only for within-node one-dimensional (1-D) (transverse-integrated) flux shapes but also for the corner-point fluxes that appear in the hexagonal-geometry relations between node-surface currents and 1-D fluxes. Thus unlike Cartesian geometry, use of the actual 1-D fluxes does not ensure exact flux-current coupling relations. This theoretical drawback of the transverse integration approach in hexagonal geometry can be avoided by employing multidimensional expansions for the within-node flux, as is done in the VARIANT variational nodal diffusion and transport capabilities recently implemented in DIF3D. Here we compare the performance of the DIM nodal and VARIANT options in the solution of fast and thermal reactor benchmark problems with homogenized assemblies and demonstrate the superior accuracy achievable with the VARIANT scheme.
- OSTI ID:
- 89274
- Report Number(s):
- CONF-941102--
- Journal Information:
- Transactions of the American Nuclear Society, Journal Name: Transactions of the American Nuclear Society Vol. 71; ISSN 0003-018X; ISSN TANSAO
- Country of Publication:
- United States
- Language:
- English
Similar Records
Accuracy of nodal predictions of HWR core physics parameters
Polynomial expansion nodal transport method in hexagonal geometry