The D2 tool steel I mentioned in the thread Nosetackle referenced is 1.5% carbon [1.5% C, 0.5%Mn, 0.3% Si, 12.0% Cr, 0.8% Mo, 0.9% Vanadium] It was welded many times, usually sucessfully with 9018 stick. There was pre and post weld heat treatment. As I mentioned in the other post, this does not provide a weld as strong as the parent metal, but it was always a catostropic screwup that broke these parts in the first place. I am not suggesting 9018 for Felipe's situation, but depending on what is expected of this weld, there may well be some way to do it. What made the repairs I described possible is that We could do the post weld heat treat properly, and not end up with untempered martinsite in the HAZ.