Hotfoot
05-21-2008, 11:22 PM
I came across this on the web, and it sounds excellent. It was submitted to another forum (on brazing) by a Andy Pugh: I pull it to this forum FYI.
I read your article on mig-repairs to cast iron with some interest, and having been through a similar train of thought, but coming to a different conclusion, I thought I would share it with you.
The basic problem with welding cast iron is that it is, as everyone knows, brittle. It will crack if subjected to tensile forces. As your weld contracts it will tend to crack the iron due to differential thermal contraction. The traditional way to avoid this is to avoid the "differential" bit and to pre-heat the whole thing to as near the melting point of the filler rod used as practicable. This is the approach generally used with cast-iron filler rod.
There is another issue with cast iron, of a chemo-metallurgical mature. What makes it cast iron rather than steel is the very high carbon content. This tends to diffuse into your filler rod, causing unfortunate metallurgical transformations on cooling. An attempt to use mild steel to weld cast iron will leave you with a fast-quenched, high-carbon steel weld. You will almost certainly end up with Martensite, brittle and unmachinably hard.
The phase-transformation in which austenite turns to martensite (or pearlite, ferrite, or bainite, depending on carbon content and cooling rate) is at the heart of the problem. This occurs at about 720 degrees. It is also accompanied by a change in volume, which may cause more differential stress cracking issues. Both the austenite-ferrite (hot-cold low carbon) and austenite-martensite (hot-cold high carbon) transformations lead to an increase in volume, so ought to reduce the residual stresses slightly, and under conditions of slow-cooling this is probably true. However this sudden expansion on cooling of the weld pool will tend to crack the already solidified and transformed weld behind it.
None of this metallurgy happens in Nickel, so nickel rods side-step the problem, as does brazing. I am not sure what sort of bond nickel rods make with iron, but brazing material needs to "wet" the surface to stick properly in just the same way as solder. This might be why your thin-sheet mig-brazing worked well, and your cast iron brazing less so, perhaps the greater section thickness resulted in the material not getting hot enough for the braze to take.
There is, however, another common mig-welding filler rod which does not go through the austenite phase transformation, and that is stainless steel. As long as the stainless is austenitic (easily checked with a magnet, austenite is non-magnetic) then the phase change on cooling won't happen, and so you can afford a great deal of dissolved carbon without fear of creating a brittle, unmachinable, weld.
Armed with this theoretical knowledge I attacked the broken guide-arm of a mechanical hacksaw I had been given. Using stainless wire in short bursts (to avoid too much heat input creating thermal stresses elsewhere) I successfully repaired the arm, and it has shown no signs of breaking again yet, after 10 years of irregular use.
I read your article on mig-repairs to cast iron with some interest, and having been through a similar train of thought, but coming to a different conclusion, I thought I would share it with you.
The basic problem with welding cast iron is that it is, as everyone knows, brittle. It will crack if subjected to tensile forces. As your weld contracts it will tend to crack the iron due to differential thermal contraction. The traditional way to avoid this is to avoid the "differential" bit and to pre-heat the whole thing to as near the melting point of the filler rod used as practicable. This is the approach generally used with cast-iron filler rod.
There is another issue with cast iron, of a chemo-metallurgical mature. What makes it cast iron rather than steel is the very high carbon content. This tends to diffuse into your filler rod, causing unfortunate metallurgical transformations on cooling. An attempt to use mild steel to weld cast iron will leave you with a fast-quenched, high-carbon steel weld. You will almost certainly end up with Martensite, brittle and unmachinably hard.
The phase-transformation in which austenite turns to martensite (or pearlite, ferrite, or bainite, depending on carbon content and cooling rate) is at the heart of the problem. This occurs at about 720 degrees. It is also accompanied by a change in volume, which may cause more differential stress cracking issues. Both the austenite-ferrite (hot-cold low carbon) and austenite-martensite (hot-cold high carbon) transformations lead to an increase in volume, so ought to reduce the residual stresses slightly, and under conditions of slow-cooling this is probably true. However this sudden expansion on cooling of the weld pool will tend to crack the already solidified and transformed weld behind it.
None of this metallurgy happens in Nickel, so nickel rods side-step the problem, as does brazing. I am not sure what sort of bond nickel rods make with iron, but brazing material needs to "wet" the surface to stick properly in just the same way as solder. This might be why your thin-sheet mig-brazing worked well, and your cast iron brazing less so, perhaps the greater section thickness resulted in the material not getting hot enough for the braze to take.
There is, however, another common mig-welding filler rod which does not go through the austenite phase transformation, and that is stainless steel. As long as the stainless is austenitic (easily checked with a magnet, austenite is non-magnetic) then the phase change on cooling won't happen, and so you can afford a great deal of dissolved carbon without fear of creating a brittle, unmachinable, weld.
Armed with this theoretical knowledge I attacked the broken guide-arm of a mechanical hacksaw I had been given. Using stainless wire in short bursts (to avoid too much heat input creating thermal stresses elsewhere) I successfully repaired the arm, and it has shown no signs of breaking again yet, after 10 years of irregular use.