Cracks when tacking aluminium?

[FASTdan]

Sometimes (probably 20% of the time) my tacks will crack across their full length - can anyone advise why this occurs? The frustrating thing is once its done it its actually quite hard to re-do without the same thing happening.

As I understand cracking is usually down to not enough filler being applied? It wouldnt matter if I could just go over it with another run but as I say it seems to repeatedly do it once its happened once!

My actual weld runs are fine (on the odd occasion there might be a small crack at the start of the run but as I'm welding tubes I ensure I cover this as I finish the run.

The materials being welded are 45mm OD tube with a 3mm wall thickness fillet welded to a 10mm thick flange plate.

 

[robwizard]

Could it be the type of rod you are using? As in not the correct filler for the material? I don't know by the way this is merely a guess..

 

[hotrodder]

Most of the time the cracks are a result of there not being enough metal to resist the shrinkage stress... ally's not very strong, especially when hot, and has a high coefficient of expansion. This is also why craters at the end of an ally weld are far more prone to solidification cracking than steels. More filler wire is one solution BUT only if 'burnt' in properly i.e. i tack ally autogenously (no filler) a lot of the time, even the flavours that are very prone to solidification cracking without problems but talking thnner sheet/tube. Something 10mm thick needs some heat throwing at it, especially for the first few tacks- as it warms up it gets easier to tack/weld. If the inital tacks are on the cold side adding more wire/slightly longer tacks might not be enough on their own

4xxx series filler wire is more forgiving in that Si lowers the melting temp i.e. weld freezes over a wider temp range and this extended 'mushy' range helps it cope with shrinkage stresses better than other flavours. The downside is something like 5356 is both stronger and less brittle than 4043, with butts the weak spot is (or should be) the annealed HAZ, with fillets the weld itself can play a bigger role. Having said that 4043 is still usually the 'go to' choice for 6xxx series ally (virtually all tube/extrusions) and more than good enough for an inlet manifold

 

[FASTdan]

Most of the time the cracks are a result of there not being enough metal to resist the shrinkage stress... ally's not very strong, especially when hot, and has a high coefficient of expansion. This is also why craters at the end of an ally weld are far more prone to solidification cracking than steels. More filler wire is one solution BUT only if 'burnt' in properly i.e. i tack ally autogenously (no filler) a lot of the time, even the flavours that are very prone to solidification cracking without problems but talking thnner sheet/tube. Something 10mm thick needs some heat throwing at it, especially for the first few tacks- as it warms up it gets easier to tack/weld. If the inital tacks are on the cold side adding more wire/slightly longer tacks might not be enough on their own

4xxx series filler wire is more forgiving in that Si lowers the melting temp i.e. weld freezes over a wider temp range and this extended 'mushy' range helps it cope with shrinkage stresses better than other flavours. The downside is something like 5356 is both stronger and less brittle than 4043, with butts the weak spot is (or should be) the annealed HAZ, with fillets the weld itself can play a bigger role. Having said that 4043 is still usually the 'go to' choice for 6xxx series ally (virtually all tube/extrusions) and more than good enough for an inlet manifold

Thanks for the comprehensive reply. I am using 5356 rods so that perhaps explains why I am more prone to cracking. As you say longer tacks seem to help - I just worry about the plate distortion the longer I spend tacking (I usually tack prior to bolting down purely as it makes the workpiece easier to manouvre.

I've also noticed as you state that heat helps - in the winter I preheat the plates on my woodburner (as its lit, I may aswell) but if thats not lit then they dont get preheated. Lately of course its not been lit and I'm seeing more cracked tacks.

I've just run out of rods so maybe I should try 4043....

 

[Will_Davis]

If it is a thermal stress sort of problem then why not try annealing the tacked areas with a propane (or soft oxy-A) torch?

I was told by me grandad that the annealing point of ally could be checked with a bar of soap: rub the soap over the area in a criss-cross pattern ( joint or bend line ) then heat gently until the soap lines turn black - result soft annealed ally!

 

[scott0031]

If it is a thermal stress sort of problem then why not try annealing the tacked areas with a propane (or soft oxy-A) torch?

I was told by me grandad that the annealing point of ally could be checked with a bar of soap: rub the soap over the area in a criss-cross pattern ( joint or bend line ) then heat gently until the soap lines turn black - result soft annealed ally!

It doesn't apply in this case. You will get localized annealing of the weld area anyway due to the heat input form the torch, assuming the aluminium in question is heat treatable. Its the rapid expansion of a small area in a cold lump of material, The heat is then drawn away rapidly by the cold material this creates two force working against each other. If the whole part was a even temp then cooled at an even rate the cracking would not occur. This is only tacks so no big deal really.

Hotrodder is totally spot on as ushaual as to be expected with someone of his experience.


Kind regards,


Scott