Problem with reflective surface

Hello everyone, I have a question regarding reflective surfaces.
I am attempting to perform a depletion calculation for a single fuel assembly, so I designed a hexagonal prism-shaped fuel assembly cell, with the outermost hexagonal boundary and the top and bottom surfaces set as reflective surfaces.
I built an identical model in both OpenMC and MCNP for comparison purposes and found that the resulting keff values differ significantly:
• OpenMC: 0.97712 ± 0.00021
• MCNP: 0.98608 ± 0.00018

I would like to ask: in addition to specifying boundary_type=‘reflective’ in OpenMC, is there anything else I should pay special attention to when defining reflective boundaries?
Thanks!

The difference is almost certainly due to the depletion calculation, not reflective surfaces. You can verify this by running a calculation with identical fresh materials and the same cross-section library and comparing it, and then doing the same with an infinite homogeneous medium.

1 Like

Thank you for your reply!

My calculation does not involve depletion—only eigenvalue calculations.
I re-ran the problem using a simpler geometry:
A spherical fuel region with a radius of 20 cm, surrounded by a reflector region extending to a radius of 25 cm.
Material temperatures were set identically to those in MCNP, and the cross-sections should all be from ENDF/B-VII. (However, I’m not entirely familiar with specifying cross-sections in OpenMC, so I’m not sure if that’s causing the discrepancy.)
Still, the calculated results differ significantly:
MCNP: 1.53379 ± 0.00020
OpenMC: 1.53695 ± 0.00023

I’m still quite new to OpenMC, and I’m concerned I might have missed something in my input settings. Could someone please take a quick look to see if there’s anything configured incorrectly? I’d sincerely appreciate any guidance or suggestions you might offer.
Thank you very much for your kind help!

The OpenMC code is attached, and here is the MCNP input:
c Cell card
1 1 -6.2088444 -1 imp:n=1
2 2 -10.53 1 -2 imp:n=1
3 0 2 imp:n=0

c Surface card
1 so 20
*2 so 25

The material compositions are identical, with the fuel set at .73c and the reflector (m2) at .71c.

so.ipynb (24.2 KB)

Looks like you’re using NNDC cross sections for your OpenMC run. Refer to