Hmmm, I don't think that expansion of the ring and crucible probably make much difference to the slipping problem as the crucible will settle to wherever it needs to maintain a slip fit . Yup I do know the fit can be adjusted.
For me the question still stands: where should it properly be located when making one? Is it half way up, two thirds, etc? I wish Ironman still visited here, as I wouldn't doubt that he has a useful opinion on it.
And also I, like you, Andrew, have made shank rings out of round bar and about 2/3 up which felt right as a mere guess. But I experienced the problem and ruined a pour. I do wonder if round bar contributes to the slip out problem. Would, for instance, square section bar be preferable? Or even flat stock? Would it be an advantage if that were shaped to a cone segment? Much to ask about this particular topic!
I suppose a shank could be tested in advance over a padded surface to protect the crucible. That might have been smarter of me than finding it out during pours. Also to mimic the real action maybe a few largish ball bearings poured in the process to mimic the molten iron emptying out and shifting the balance.
I should confess that I eventually gave up on ring shanks altogether and just poured directly with the tongs I made. But I'd still like to know the solution to the problem.
Edit: Okay, I'm going to throw out a new (for me ) guess, that actually the ring should be located nearer the bottom end of the crucible than its balance point when on its side. Theory, mind you, not necessarily reality.....