Effect of sub-pin feedback in coupled MCNP6/CTF

There has been a trend towards coupled Monte Carlo (MC) neutronic and Thermal Hydraulic (TH) subchannel codes to be used to create reference solutions. Typically the pin average feedback assumption is used to decrease the computational requirements. The pin average feedback assumption is where the fuel region in the neutronics code is modeled as a single region. In this paper the number of sub-pin fuel regions are varied in a coupled MC neutronics and TH subchannel code to get a measure of their effect on a coupled solution as well as to verify the pin average feedback assumption. The pin average feedback assumption was found to give errors of over 100 pcm and over 15% in the fuel centerline temperature. For most of the cases, decreasing the number of sub-pin regions in MCNP6 gave conservative results, while decreasing the number of sub-pin regions in CTF gave non conservative results, with an exception of the eigenvalue. The effect on the eigenvalue was reversed. (authors)