TIG cracking on roller forming machine

[dssobral97]

Hello


I work at an exhaust system producing factory for various types of vehicles, in Europe.
We've been having problems with a TIG welding roller forming machine at my work, the photos bellow will show you.
Basically, we have this machine that uses TIG welding on a metal plate that is rolled previously (in the same machine).
The plate, now rolled into a can, goes inside a welding cage, where it's guided by two wheels that press both ends together and moves forward (with the assistance of a mechanical pusher) as the weld is made. The welding torch is stationary, the height is adjustable through a HMI.
Then, the can goes through an expanding device to reach the desired diameter of the rolled plate.


I cannot go into details on the characteristics and model of the machine, as it is one of a kind and confidential, so there isn't anything about it out there.


The problem is (as you can see in the pictures): The weld cracks at the end (sometimes at the beginning too), after it goes through the expanding device.


We are able to adjust the desired diameter, we could put it lower and solve the problem, but then the product isn't compatible with the next procedures, as the diameter isn't enough to produce a good part.


We've tried multiple things: Adjusting welding parameters (lower and higher amperage, lower and higher voltage, distance at which the welding starts and stops, we've switched the TIG torch, we've checked if there's any gas leak and if the gas system is working properly) but nothing seems to work.


We thought it was a problem in the rolling area, as it seems like the plate's ends don't meet each other close enough, but it isn't, as we've adjusted the rolling section to put them as close as possible together, and they only tear apart after the expanding process.


We can't prolong the weld much more, as it creates holes in the end of it, as a result of lack of material I guess.


If you have any idea what might cause this I appreciate the help
I will leave some information that might be useful:


- The welding rod is 3,2mm in diameter and the angle accepted by the machine is 30º
- The welding machine is a Fronius Kuhlgerat FK 4000-R FC
- The usual voltage we use to weld is 10V, it's the only way we can get acceptable penetration
- The amperage is always around 190A-200A
- The protective gas we use is Argon


If you need any other information that I might be able to provide, please ask.

[snoeproe]

Looks like aluminized coated metal?

[dssobral97]

I forgot the material sorry, it's stainless steel, according to the supplier

[thegary]

Have you tried to back step the last 10 to 15 mm of weld by reversing the feed so the end of the weld is not at the very end of the tube. I would also still add some filler while doing it. Also are you keeping a shielding gas cover on the weld as it cools after the arc goes out?

Can you show us the inside of the pipe at the weld point? Is the machine set up to back purge the weld since it is Stainless or is there a backing behind the welded seem ?

[walker]

What filler metal is being used?

[thegary]

It seems to me that your amperage is very high. I presume that is because of the speed at which the part is welded. At the end of the tube there is no where left for the heat to go and is overheating the the end of the weld. Is there any way to reduce the amperage for the last few meters of the pipe? The other thought is to make the pipe longer and sacrifice the fist and last few meters of pipe. This would involve another step in the process though.

[arcflash]

Try bringing the weld out past the end of the tube and peen it.

[snoeproe]

If that’s stainless, it looks to have a heavy oxide coating on it. I would start simple and remove all that oxide coating from the metal in the weld area. See how this helps.

[dssobral97]

Have you tried to back step the last 10 to 15 mm of weld by reversing the feed so the end of the weld is not at the very end of the tube. I would also still add some filler while doing it. Also are you keeping a shielding gas cover on the weld as it cools after the arc goes out?

Can you show us the inside of the pipe at the weld point? Is the machine set up to back purge the weld since it is Stainless or is there a backing behind the welded seem ?

We've tried to keep the weld away from the end, but then there's a gap too big at the end of the tube where there's no weld. The acceptable amount is 4mm gap, the expanding device makes it way past that.
There's really no way to add filler, as it uses a non-consumable tungsten electrode
We are using Argon as protective gas, it stays at the end for 4 seconds before ejecting the can from the welding cage. We've tried to extend this period, but the result is the same.
There is a back purge system incorporated during the weld. There's a piece, a welding arm we call it, inside the welding cage, that stays under the 2 ends of the plate being welded, where there's multiple exit holes for the gas to be expeled. There's also the gas coming from the torch, so it's well protected.
I will try to give you a picture from the inside, there's plenty penetration, even excessive sometimes, except at the end of the weld. By the way, the end of the weld, in these pictures, is the one shown in the last picture. The others are all the beginning.

[dssobral97]

What filler metal is being used?

No filler metal. Non-consumable tungsten electrode

[dssobral97]

It seems to me that your amperage is very high. I presume that is because of the speed at which the part is welded. At the end of the tube there is no where left for the heat to go and is overheating the the end of the weld. Is there any way to reduce the amperage for the last few meters of the pipe? The other thought is to make the pipe longer and sacrifice the fist and last few meters of pipe. This would involve another step in the process though.

There's no way we can change the size of the pipe, as it would involve changing entire projects with huge losses to the company
We've tried to reduce and increase amperage and voltage, no results

[snoeproe]

So this is an autogenius weld? (No filler added)
I suspect this is your main problem. Especially with all the oxide that hasn’t been removed from the weld area. You need to use some filler rod.
Another issue, if your not using internal shielding of some sort, you will have sugaring on the inside of the tube which is a defect that happens on stainless tube/pipe when it’s tig welded without an internal gas purge or shielding of some sort.
It doesn’t sound like this weld procedure has been verified as solid by your employer.
Hate to say it but Your employer is setting you up for failure with his one.

[Woznme]

I see two issues you have to deal with here. Firstly you have to get the bead to flow right to the end without leaving the fish eye. Secondly stainless steel work hardens during the expansion process, the heat affected zone is likely to be failure point.
Introducing pulse with increased peak current may help with both of these issues.
Worth a try, you may even be able to increase travel speed.

[mrmikey]

So this is an autogenius weld? (No filler added)
I suspect this is your main problem.

My thoughts also, lack of reinforcement. Any chance you could add a run off tab, it may help but adding filler would be my first fix if you can do that as well as ramping the heat down near the end........Mike

[Woznme]

Another thing you can try is trimming a few mm off the end prior to expansion to ensure a true end without imperfections to propagate cracks.

[walker]

I would think a post expansion weld to fix the crack would be fairly easy to do.

[villageblacksmith]

There's no way we can change the size of the pipe, as it would involve changing entire projects with huge losses to the company
We've tried to reduce and increase amperage and voltage, no results

It looks to me like there will be heavy losses to the company at some point anyway unless something changes. I don't know how automated the welding process is, but one thing to try if the equipment is capable of it is to start the weld at one end like I assume is being done, but only weld to the middle of the pipe instead of letting the weld progress to the other end. Then, travel to the other end of the pipe, start the weld, and travel towards the welded portion, finishing the cycle by overlapping the previous weld by maybe 1/2 inch.

[ronsii]

It looks to me like there will be heavy losses to the company at some point anyway unless something changes. I don't know how automated the welding process is, but one thing to try if the equipment is capable of it is to start the weld at one end like I assume is being done, but only weld to the middle of the pipe instead of letting the weld progress to the other end. Then, travel to the other end of the pipe, start the weld, and travel towards the welded portion, finishing the cycle by overlapping the previous weld by maybe 1/2 inch.

From his first description of the machine process I would guess this is not possible... sounds like the machine rolls the stock and welds in one operation... so would be a one way gig.

[Welder Dave]

Could manually back welding the end (1/2" or so) so the automatic weld ended on it be an option? Other option is to cut the the crater end off after welding. Would be quick with a metal chop saw or bandsaw. Might be faster than playing around with other welding options.

[thegary]

Take this for what its worth but I think I have an idea to try. I will say I have no experience with automated welding but I have welded for a long time. Have you tried to slow down the welding process by reducing both the feed speed and the amperage. I know this will slow down production and is not going to make management happy but I think it has merit . I would suggest cutting your amperage in half and adjust the welding speed down until you get the acceptable penetration needed. This will probably involve voltage adjustment also and will take some trial and error to get good results. To solve the cold start and the cracking at the beginning of the weld maybe you can program a pause in the feed rate for the start of the weld.

[Xsbank]

How about preheating the pipe with some sort of torch before welding and expanding to allow more uniform cooling and less stress? The steel is being bent before welding, maybe it needs a sort of annealing process before welding? Likely only necessary on the ends, you don't get cracking anywhere else?

Some stainless is very susceptible to work-hardening. Maybe if you can keep the ends together, the actual use as an exhaust system will provide enough heat for the system to "normalize?"

[shovelon]

i am going to blame bad material. There should not be a problem if the right alloy is chosen for autogenous welds.

[MinnesotaDave]

I would be looking to manually tig weld the start/finish either before or after the automated process.

Prior would be my first choice, it would still look like an automated job and a little bit of filler in the beginning and end could fix the problem.

[shovelon]

I would be looking to manually tig weld the start/finish either before or after the automated process.

Prior would be my first choice, it would still look like an automated job and a little bit of filler in the beginning and end could fix the problem.

With a little filler to check the cracking tendency.

[MinnesotaDave]

With a little filler to check the cracking tendency.

Exactly my thoughts as well.