The fast Sn cited first by Impaler is very interesting. They derive an expansion velocity from line-width, then compare that with the radius obtained at different times assuming a black-body radiator. There results are, by my calculation, fairly consistent with no expansion and way off for conventional BB.
