I am modeling a domain, part of it is rotating (a ball), and part of it is stationary (a chamber). Should I use multiple rotating frame or dynamic mesh? Thanks

this is pretty interesting case, with MRF, for spheres i ahve observed that fluent calculates negative lift coefficient(i ahev done more than 10 different cases with it). its better to use dynamic or moving mesh with it. at least the sign of the lift is okey.

Really? They yield different results? But the latter seems more difficult

yes, but we are not able to find the solution to this problem, fluent support also is answerless, fianlly we are only using sliding mesh approach which is easy to use (if does not give u negatve volumes or handedness prob)