Hi KK
Yes, welds are small castings! Welding problems and casting problems are very similar to each other.
Regarding your question concerning cooling rates, I believe that the answer can be too complex to really answer it here. Typically, the faster the cooling rate through the solidification range, the finer the dendrite (and therefore grain) structure will be. In addition, if the cooling rate is high enough, not only will the dendrites start to form (nucleate) from the base metal, but will also start to nucleate within the melt itself. This will only happen if the moulten metal is "supercooled", because it takes a great deal of "driving force" for this to happen. Under these circumstances, the segregation mechanism will probably have less time to take place, but will none the less take place. It may reduce the grain boundary layer, but will probably increase the shrinkage stresses. As such, I am not sure that you will have any great benefit either way. This is only speculation on my part, as I have not seen studies looking at the effect of solidification rates on hot cracking. It may be an interesting exercise to try.
Regards
Niekie Jooste
Just had a look at my AWS welding metallurgy book, and they recon that with a decrease in grain size, hot cracking decreases because the amount of surface area of the grains increase. This allows the contamination rich material to be spread over a larger area, reducing the concentration of the contaminant. Based on this I would think that increasing the cooling rate would reduce the hot cracking tendancy of an alloy such as 304 or 316.
Regards
Niekie Jooste