I’m just computing the K-eff at different temperatures, then linearly interpolating to find the k-eff at a specific temperature.
I’ve been told that the cross section should be fairly linear in the elementwise sense, but I’m concerned there might be some ‘phase shift’ (to borrow a term) in the resonance region.
Depending on what temperature difference you’re interpolating between, linear interpolation may or may not be accurate. I would suggest looking over an excellent paper by Tim Trumbull who looked at temperature interpolation in considerable detail. It is published in Nuclear Technology, volume 156, pages 75-86. It is also available in a free version at:
http://www.osti.gov/energycitations/purl.cover.jsp?purl=/875458-xJORlP