More info please. Preheat, amps, etc. You know the drill.

Hello jhopwood, as others have mentioned you may want to spell out your specific application in more detail.
     From a general standpoint, when I weld on cast iron, whether it be to steel, a grade of stainless steel, or simply cast iron to cast iron, I opt for a nickel-based filler more along the lines of Ni-Rod. I have often chipped the flux off of Ni-Rod and used the core wire after removal of the flux and thorough cleaning with scotchbrite so as to leave a clean metal core wire to be used as a filler. There are also specific GTAW fillers for these purposes, yet my experiences in this area have been with a Eutectic filler metal and it was very pricey.
     It sounds like you are fabricating parts where you can't really run your welds in a manner that is conducive to allow for the rigidity differences of the two metals. To better explain I'll term it like this: cast iron is a very rigid material and not very forgiving in it's ability to withstand stresses introduced by welding, the stainless on the other hand generally will withstand welding stresses very readily. You would think that this would make this joining process very simple as the stainless should "theoretically" be able to absorb and minimize the stresses that develop during the welding. I believe the problems that you are probably running into are centerline cracking of the beads or possibly cracking between the weld bead and the cast iron side of the weld bead. Differences in co-efficients of expansion between the two materials is likely partly the problem and another issue is likely due to the amount of dilution of the weld material because of using the GTAW process. It is also likely that the weld metal is shrinking at a different rate between the stainless side and the cast side causing bead issues while the weld metal is in it's plastic state. An example might be considering what happens to a soldered joint when it is stressed or moved before the lead has had a chance to adequately set.
     Whenever I weld cast iron I use minimal heat, welding short small passes, and peening well after each application of weld material. Depending on the design and application of the weld joint I try to distribute the welding stresses equally around or through the weld joint in a skip fashion or side-to-side type application whichever applies. A preheat of the cast iron portion of the weldment might aid in preventing or minimizing some of the cracking. If you run with a preheat of some sort you will likely run the risk of hardening of the completed part, so to help with this I might suggest using some sort of slow-cooling method which might include heat blankets or possibly lime. Since most of my repairs can't utilize pre-heat ovens and such I try to take another tack and keep all of the welding at a temperature where I can lay my hand on the part between each small weld application and include a generous dose of peening before going on to the next one.
     If you are familiar with the search function of the forum you might also try searching under the heading of "welding cast iron" or similar topics. There are numerous threads that have dealt with similar issues over the years. Good luck and best regards, Allan

Hello again jhopwood, thought about this a bit more and have a few additional observations that could impact your success. Grade of cast, type of cast, quality of cast. Many cast iron items can have consistency issues with regard to quality of casting, sand pits, or lack of them, carbon/graphite dispersion in the casting and amount present. Have these cast parts seen any cyclic heat cycles? If they have and if the cycles have included extreme temperature ranges you can end up with collection and migration of undesirable components in the cast that will adversely affect weldability.
     In some instances you may be better off to consider an SMAW process to reduce weld metal dilution, possibly even going to GMAW with a nickel wire might be a possible answer. Yet another possibility would be to go with a silicon-bronze filler with the GTAW process to help with your crack issues provided that the deposit will meet the mechanical strength requirements/chemical/environmental constraints of the intended application. The silicon-bronze doesn't actually melt the base materials so it may provide the answer to your crack prevention while still meeting other requirements. Keep us informed and please include any other details as you can. Best regards, Allan