Pat Flannery writes:

To totally eliminate that effect at one g, you need something around the
diameter of the station in 2001.

Rather than a centrifuge, why not just spin the whole craft? That eliminates
bearings and vibrations (to some degree) and helps spin-stabilize the craft.
Also the entire mass of the spacecraft can offset small shifts in mass inside
the craft, eliminating the need for a high rim mass centrifuge to achieve the
same effect. Also it's not clear you need to spin to 1g when something smaller
might suffice and thereby avoid the coriolis effects on the inner ear at a
higher spin rate.

This of course puts some design constraints on the spacecraft and requires a
fair degree of axial symmetry. There are always trade-offs....

?

Dave