aluminum weld cracking

admin

I am  more than a novice but not an expert by any means. I have not run into this before so I have to ask. After Tig welding a crack in a bell housing, which I have done 50 times with no issues, it cracked while cooling. I groung it back down, prepped it and tried again. Same thing. About 2 minutes after the weld. CRACK!  I am doing it the same way I have always done it. Same settings, wire, gas, amperage. Any help/suggestion arew appreciated.
Reply:You need to find a way to slow the cooling down.  Preheat an area about 3" surrounding the weld to 300-350F before you weld, and again after the weld is done.  Then let cool slowly.  Either wrap in welding blankets, bury in sand, ect.Usually this is more of a problem when working on thick aluminum which I doubt the bellhousing is, but I imagine that the casting is so that is dissipates the heat very quickly.Have we all gone mad?
Reply:Did it crack along the edge of the bead, center or whole bead pop back off?Also any spyder cracks?Matt
Reply:Sand is not an option as the bellhousings are not usually removeable. But I have tried a heat lamp to keep it warm longer. I just don't get what's diferent on this particular trans.Thanks for the info
Reply:Crack is dead center of the stack of dimes.
Reply:Originally Posted by gitardedSand is not an option as the bellhousings are not usually removeable. But I have tried a heat lamp to keep it warm longer. I just don't get what's diferent on this particular trans.Thanks for the info
Reply:I'd suggest 5356 for filler vs. 4043. I find a lot less cracking while cooling on the beads done with 5356. If removal isn't a good option, nor is sand, get a putty for heat sink; heat it in an oven to 300-400 degrees. You can actually slap the putty right on the welds, and mold it into place. It will cool the material much slower, minimizing the chances of cracking.And then, after so much work...... you have it in your hand, and you look over to your side...... and the runner has run off. Leaving you holding the prize, wondering when the runner will return.
Reply:Rojo,IMO it would be iffy to use 5356 to weld a weldable aluminum bellhousing, because of the 150F service temperature limit with the filler.  And engines and trans assemblies and engine bays and such are usually running at greater than 150F.And about preheat?  In general, you do NOT-NOT-NOT preheat aluminum to greater than 200F.  Heating a piece of aluminum up to 400F is in the temperature zone for aging of heat-treatable aluminum alloys.  Which a bell housing might or might not be, depending on exactly which alloy the casting was made with.  A 3xx.x casting alloy has silicon and magnesium as principal alloying elements, is heat-treatable, and from an alloying perspective is similar to the 6xxx wrought alloys.  Which are prone to hot cracking.  Usually, most of the alloys in those families are able to be weldedwith 4043 filler but adequate weld puddle dilution with the filler alloy is needed to keep the weld and parent material from hot cracking.Per Dave/Blackbird, remove the housing and prep the part well and then weld it (it -is- a weldable alloy? ) making sure to have adequate weld puddle dilution with the 4043 filler as you weld (V or U groove the joint as opposed to a butt joint to make room for more filler).  No more than 200F preheat if at all, and keep interpass temps below 250F to minimize weakening the parent metal.  The best laid schemes ... Gang oft agley ...
Reply:Well, I am using the 4043 filler, and the transmission is out of the car, and I did make a pass on both the inside and the outside with only the outside cracking. HOWEVER you said something about the stress on the belhousing and my dumb A$# forgot I had to clamp the broken piece in in order to weld it. It was broken on the curve of the motor mount and had to be pulled back in shape to weld. THAT stress is most likely the cause of the crack. Sorry to bother you guys and thanks for the info.
Reply:Originally Posted by gitardedWell, I am using the 4043 filler, and the transmission is out of the car, and I did make a pass on both the inside and the outside with only the outside cracking. HOWEVER you said something about the stress on the belhousing and my dumb A$# forgot I had to clamp the broken piece in in order to weld it. It was broken on the curve of the motor mount and had to be pulled back in shape to weld. THAT stress is most likely the cause of the crack. Sorry to bother you guys and thanks for the info.
Reply:Originally Posted by MoonRiseadequate weld puddle dilution with the filler alloy is needed to keep the weld and parent material from hot cracking.
Reply:Originally Posted by MoonRiseRojo,IMO it would be iffy to use 5356 to weld a weldable aluminum bellhousing, because of the 150F service temperature limit with the filler.  And engines and trans assemblies and engine bays and such are usually running at greater than 150F.And about preheat?  In general, you do NOT-NOT-NOT preheat aluminum to greater than 200F.  Heating a piece of aluminum up to 400F is in the temperature zone for aging of heat-treatable aluminum alloys.  Which a bell housing might or might not be, depending on exactly which alloy the casting was made with.  A 3xx.x casting alloy has silicon and magnesium as principal alloying elements, is heat-treatable, and from an alloying perspective is similar to the 6xxx wrought alloys.  Which are prone to hot cracking.  Usually, most of the alloys in those families are able to be weldedwith 4043 filler but adequate weld puddle dilution with the filler alloy is needed to keep the weld and parent material from hot cracking.Per Dave/Blackbird, remove the housing and prep the part well and then weld it (it -is- a weldable alloy? ) making sure to have adequate weld puddle dilution with the 4043 filler as you weld (V or U groove the joint as opposed to a butt joint to make room for more filler).  No more than 200F preheat if at all, and keep interpass temps below 250F to minimize weakening the parent metal.
Reply:Rojo,T'was just trying to clarify a bit about "preheat".See, you say 'preheat' to a steel guy, and he or she might run the item up to 200F or 350F or 500F for some alloy stuff.Say 'preheat' before welding to a cast iron item, and you could be talking 800-1200F.Aluminum is a bit more, ummm, 'delicate'.        Heat aluminum up to 1200F and you have a melted blob of aluminum.Interesting tip/hint about heating up the clay/putty.  Sounds like it's working by added some preheated thermal mass to the objects being worked on.And jakeru, piling a weld bead higher in an attempt to get more puddle dilution to prevent hot cracking of aluminum may work for that aspect but will -totally- kill the fatigue life/strength of the aluminum if there is a dynamic load condition (varying loads or vibration or such) and not just a static load condition (a non-moving weight or load that doesn't vary or vibrate at all).  Putting a "good" weld bead on metal reduces the fatigue limit stress by about 25-30%.Putting a poorly profiled but still 'good' weld bead in a 'critical' direction can reduce the fatigue limit stress to 1/4 or 1/5 of the original metal properties.Putting a poorly profiled AND 'bad' weld (porosity, slag inclusions, incomplete penetration/fusion, undercut, starts/stops, craters, etc, etc) can reduce the fatigue limit stress even lower than that.And aluminum doesn't even have an indefinite fatigue limit to begin with, unlike steel.Numbers?Let's say a piece of wrought/billet steel had an original fatigue limit stress of 200 (units left out, but these are "real" numbers).The "good" weld with the "good" bead profile in a non-critical direction could have the steel reduced to a fatigue limit stress of about 140-150.A 'bad' bead profile (humped up) in a critical direction could end up with the fatigue limit stress reduced to 45.Put in some undercut, incomplete penetration/fusion (one-sdied weld perhaps), slag inclusions AND a humped-up bead profile and you might be down to a fatigue limit stress of 20.  The best laid schemes ... Gang oft agley ...