Avoid buckling of plate when welding circular frame?

[jellywerker]

I'm working on a project with a number of steel discs that need to be banded with a rolled frame, and then seam welded on the exterior so they can be ground down for a cosmetic finish. All materials are 1/8in mild steel. The band is flush with the disc on one side. Think of a round tray.

I am finding that despite taking efforts to limit heat input, by the time I've seam welded the entire circle, the contraction from the welds cooling is enough to force the base plate to deform/oil can - about 1/8in over 20in.

I am stitch welding the seam, about 2in at a time and rotating to the opposite side. 70-80 amps DC tig, minimal filler. Edges are prepped with only a tiny bevel, as it's purely cosmetic and I don't want to put too much metal in.

Any thoughts on mitigating this? Would peening out the weld be helping? I'd be happy to torch it back hot and hammer around.

 

[Rob F.]

Try no bevel and no filler, just fusion weld it. In my experience filler seems to make more shrinkage. You may play around with joint design so the weld still has strength after blending.

 

[memphisjed]

Mig warps less than tig. Weld your ring then weld your cover which will be smaller than id of the ring, stitch weld with each set of stitch passes becoming bigger (tacks first pass, 1/2 in second, 1 inch third....)
The stresses will pull the plate tight like a drum.

Or gas weld it with a torch.

 

[IceCzar]

I'd be thinking if there was a thick spacer\plate\jig I could clamp it to, to keep it flat then normalize the whole thing with heat, cool and extract it.

 

[farmersamm]

Peening mostly gets rid of stress as the weld is cooling..........used for highly restrained joints to prevent cracking.

Preheating can be helpful. It lowers the temperature gradient, and will limit distortion somewhat.

But, by in large, once you strike an arc, you're dealing with distortion.....even if you clamp all pieces(they'll simply spring back once the clamps are removed)

Ultimately, I'm thinking you'll have to do some heat shrinking to get rid of the distortion.

 

[jellywerker]

Thanks for all the helpful feedback!

It's a cosmetic piece that will get plated, so mig is a no go due to potential porosity (also I don't own a mig welder), and I can't get away with just fusion welding as the slight dips and valleys where the puddle sinks into the seam can't be ground out without rounding the edges too much.

Jed, I think that's a great idea and will do the next one like that. My process until now has been to wedge the ring tight to the plate, tacking as I go around, then final fit the ends, tack, flip and seam weld.

Big round plate to normalize on makes sense, might be able to give that a shot on the current finished pieces.

Thankfully the finished product can accept some dishing, but I feel it reflects poorly on me as a fabricator to send out a finished product with this much distortion.

 

[farmersamm]

Are we talking normalizing, or stress relieving? Both require different temperatures I believe.

Stress relieving involves heating just below the critical temp. While normalizing requires heating slightly above critical temp.

Get above that critical temp, and you're looking at changing mechanical properties. Dunno if it would be a huge thing with mild steel, but just wondered. I do know that there are strict limits on heat input when straightening structural steel. 1200F is the upper limit.

 

[plastikdreams]

How bout some pictures to help us see what you are talking about.

 

[IceCzar]

experiment
there is probably an answer to be had with all the variables nailed down, the alloy and the right partial differential equations
I just go at it. The other advantage of the jig is it will wick away some of the heat.

 

[jellywerker]

Attached below. Nothing fancy, just a big disc with a border. This one is 2in but same story.

 

[plastikdreams]

Your heat affected zone is huge...use more current and run much faster.

 

[jellywerker]

I'll give that a shot, thanks. Self taught on TIG, so sometimes I don't know what I don't know.

 

[DDoug]

Your heat affected zone is huge...use more current and run much faster.

Yup, newbies want to go slow, turn down the heat.
Crank it up, hit it fast, and get off it.
MIG would be much better here, but since you won't get one, might be best to sub it out.

 

[farmersamm]

I think you set yourself a high bar on this one. I really don't see how it can be done without some distortion.

Is it a stand-alone piece, or will it be attached to something else? If it's gonna be welded to another something or other, it might be held relatively flat when welded.

There's just too damn much weld on it. If it has to be welded like that, it's just the price you pay I guess. Doggone shame.

 

[john.k]

A lot more preheat of the whole thing.

 

[metalmagpie]

What if there were a round cake of ice lying centered on your disc? Or some similar way to wick the heat out of the disc?

 

[BT Fabrication]

Attached below. Nothing fancy, just a big disc with a border. This one is 2in but same story.

View attachment 384723

Id second that your heat input is massive, how big/ what tig welder are you using?
and with tig, the part always gets much hotter then with mig, especially if moving too slow and cooking it. 1/8" you should be able to light it up and just fly through it.
the weld profile is fat and wide, which is causing all the pulling and distortion. Only other way to keep that thing flat is to clamp it, or best bet. weld a bar across it and tack it in the middle, break off when done and blend back smooth. Im guessing the middle is popping up like a huge mountain, as huge plates of that thin of steel will always do that just due to size, as when you weld, you add heat in a localized zone, it expands on the weld side, and you continue to weld it. The middle is most likely dished like a lake then, the weld stops and cools, and shrinks on the weld side pulling it and pulling the top closer to the edge and making a dome.
After it happens you can only live with it, or heat the one side to 4-500F and heat warp it opposite and closer to flat. 1/8" isnt too bad depending what its for.
you also might get a bit better if you can have small tacks all over the inside preventing it from pulling.

 

[Ries]

I always overlap the plate, and have the ring below, so the weld is on the side, not the top, cosmetic face. Sanding something like this is inevitable, and its easier to get a square corner that way. (although rounded corners are my choice- they look nicer, and hold the final finish better, as sharp corners are places where paint, powder, or plating always fail first)
And I would always do tiny tacks on the inside first. I would probably cut some backing rod the same height as the ring, and clamp the sheet down to the backing rod and the table, so its clamped in place while being welded, then let it cool while clamped. Clamping something flat has minimized heat distortion for me on dozens of projects- it doesnt eliminate it completely, but it helps.
This looks to me like something that was designed by either an artist or an architect, and they always want the impossible, cheap.

 

[DanielG]

If you can't get it flat with just welding technique, there is a cool flame straightening technique for tightening up thin flat plates. See 7:44 at

 

[jellywerker]

I appreciate all the continued advice.

It's a standalone design object piece, seen from all sides. There's an insert that goes in the frame, so some distortion is acceptable. Last time I made one was with the same processes, but I had access to an Acorn table. I don't think I realized how much that big block of cast iron was probably helping me out.

DDoug, I'd be happy to try it with MIG in the future, just don't have the money to buy one right now for how often I'd use it. Sub out isn't really an option, not for this order at least, maybe in the future.

Sanding/grinding is expected, I just meant I'd have to bevel the edge down a noticeable amount to fully smooth out a fusion weld, whereas with filler there's material to keep that corner crisp and square after grinding.

The customer will be happy with them, I'm just not. Sounds like the first thing I should be doing is welding the next one hotter and faster.

Ries - you nailed it.

BT - it's an Everlast 210, would happily be putting out more amps, I just didn't know to do so. Thanks for all the other tips as well.

Quite a few days when I feel like I should have had more experience before hanging out my shingle, but you do what you can...