What's new
What's new

preventing weld deflection on a bolt flange

GMBM

Aluminum
Joined
Dec 20, 2006
Location
NY
Trying to solve a problem when welding a bolt flange to a box the weld will pull the flange so there is a gap at the edges when assembled
The thickness of the plate ranges from ½ to 1.375 depending on the assembly and the flange will range from 1.25 to 4.24 past the end of the box

On the ones that get gussets there will be less of a problem but the deflection is still there unfortunately we can not put gussets on all of them

We use spray transfer weld with solid wire, I know this is hotter but there is no post cleanup work

Most get 3 passes .045 wire, 30-32 volt, 300-375 inch per minute, depending on the box, we let the weld cool between passes

What I would like to know if there is any change in the weld prosses that might help to prevent this

Really not possible to rest the box in a water bath do to the size

I hope I have given enough information to explain the problem to hopefully get some possible solutions

Thank you
Greg
 
Set-up the workpiece out of "Square" inversely proportionate to the deflection amount of current technique.
Fine tune it by tack weld prior to the beads. can clamp, though less consistent.
Note, sequence of tack etc. makes a huge difference.
 
John K's suggestion of preheat will reduce thermal gradients/thermal stresses, and should reduce warp. Letting it cool between passes would seem to me to increase it. Preheat or letting it get very hot from multiple passes, while it is bolted to a strongback of some sort, that holds it flat 'til it is cold should keep it flat, as long as the ductility is enough that it will not crack.
There is a very long, very contentious thread here some where about vibratory stress relief. It can be done very cheap, andi ti has worked to reduce warping for me.
 
Preheat let it cool to maximum temperature for you steel between passed. Strongback tacked on, gussets... everything said above. You live with flanges pulling. That is what grout and jack nuts are for. If bolting to another plate later, they still ride with the bow.
That is insanely hot for .045 wire. Switch over to flux core at that size weld. Unless you are using dual feeder you are on the very edge of 3 pass weld. More Small passes with no weave are stronger and warp less.
 
The 2 things that come to mind is interpass temps are too hot and the fact of using a spray weld vs a dual shield. The dual shield will probably give you better results for the weld, but will have to to some post cleanup. This cleanup should be a lot less then you think if done properly.
 
Thank You for the responses

It is bolted to a plate thicker than the one being welded

I might be misunderstanding but is it better to let the weld cool or keep the weld hot for the 3 passes
 
I can't see your part but if you are able to offset the flange out past the edge of what it welds to you can do an internal open corner weld first, which will pull it one way, then your real fillet which will pull it back straight. I've never done it on that thick of metal though. Usually limited to 1/4". I have also done a cross bend in a square 1/4" flange to prebend it opposite distortion, and we have a few production parts that get a 3 degree prebend so they come out flat after welding.
 
Thank You for the responses

It is bolted to a plate thicker than the one being welded

I might be misunderstanding but is it better to let the weld cool or keep the weld hot for the 3 passes
welding that hot gives more distortion then not being as crazy.
im ususally about 26-27V and about 320ipm on 0.035 wire or about i think 275IPM with 0.045 and maybe 0.5 to 1V higher at most.
 
Yes post machining is really the way to go but the size of the assemblies make that not a possibility for us

Fortunately we do all the assembly in house so we can make sure everything fits correctly

We are looking into changing the way we handle the welding so was trying to see if there was any welding prosses that would improve on the problem
 
Yes post machining is really the way to go but the size of the assemblies make that not a possibility for us

Fortunately we do all the assembly in house so we can make sure everything fits correctly

We are looking into changing the way we handle the welding so was trying to see if there was any welding prosses that would improve on the problem
Besides already stated way to hot, three passes on 1” is not enough- more smaller stringers warp less than big fat globs of weld. Big is relative- we run .06 and 3/32 wire on the heavier plates. Still, more small stringers than single pass weaves.
 
I also see this problem all the time at facilities that do sanitary tube. They fusion weld so hit that it warps through ferrule and they wonder why they need a bug wrench to do what you should be able to do with your fingers.

I have seen it so bad that on a 2" ferrules with od of 2.3" the difference can be 1/16". That is almost thicker then the gasket.
 
Need a different recipe for the different sections. Tack weld the corners first then tacks in between. 1/2" is doable with .045 & 300amps but you need strongbacks clamped to the flange. The thicker section parts you'd run a similar torch travel speed with the wire upped to 375. Really getting into sub-arc territory here.
Good luck, Matt
 
I agree with all of the comments about changing your weld to a multipass but here's another thought-

Can you change the design of your strongback to preload the flange? Something like a spacer separate the 2 so you may pull a little harder in the 'flat' direction?

Side story- I bought a skid from a mfgr once, it was the reboiler for a dehydration unit. 8' wide, 40' long 8" i beam frame, heavy xmbrs, and a 1/4" seal welded deck set just below the top flange. That fucker looked like a dorito at first inspection. It was unfortunate, but we made them scrap it. On unit #2 they opted to not let the welder see how long he could run a single bead.
 








 
Back
Top