Negative Flux and Precursor Concentrations

[nhsnow2]

Hello, I am working with a 24-group solution for the MSFR based on the MOLTRES examples, documentation, and benchmark work to put together a 2D MSFR model, but currently my flux and precursor concentrations show negative lower bounds. This sometimes results in negative values for the power_density/fission_rate auxiliary variables. The negative lower bounds are orders of magnitude lower than the positive maxima and normally occur only in low-flux regions (near zero).

Is this an expected behavior or a result of numerical error?

Things I have checked/tried:

• Adjusting linear and non-linear convergence criteria
• Adjusting normalization factors
• Ensured cross sections are positive
• Used various single-temperature cross section data

Thank you for your help.

Example:

Input file, geometry, and cross section data for replication:

moltres_question.zip

[Jeremyb8707bigbrain]

Maybe I am missing something but the legend on the right depicts the lowest bounds as 4.3e9 and 1e0. I dont see any negative numbers here just that the numbers are going lower near lower flux zones.

[nhsnow2]

Hello, the bottom figure is log-scaled to help identify where the negative flux is occurring. Negative numbers are shown as '1.0' on this scale. The top figure is in linear scale and shows the negative lower bound.

[Jeremyb8707bigbrain]

I apparently just didnt see the negative sign infront of the -4.9, my apologies. I hope this gets fixed for you!

[smpark7]

You need a finer mesh along the boron carbide interface to "accurately" capture the strong neutron flux gradients (as accurate as neutron diffusion can be). You can refer to page 25 of my masters thesis (https://www.ideals.illinois.edu/items/116168) for the mesh I used for my six group MSFR model.

[nhsnow2]

Thank you!