Panel Distortion

[atomant48]

Hi all

After some thoughts on a little project. I am replacing a large part of a car bonnet and have been quite successful using thin metal technique in getting a good join without burn through or warping. The problem i am having though is a "valley" is forming along the long edge in the photo as it cools. So instead of grinding back to a flat join, i am left with a slightly depressed valley along the weld.

It is not structural, and has plenty of rigiidity, so any ideas to "flatten it out"? I am wondering if it is possible to shrink the metal perhaps to pull it up?. The other option is a smear of filler along the join (fortunately it has depressed rather than risen so quite a straight foward fill).

I estimate (using a steel rule) it is about 2-3mm at the deepest part.

Through trial and error, I think some of the issue is caused by not leaving enough of a gap in the join, and the 2 edges seem to push against eachother as they heat/cool.

So, any ideas? Thanks in advance. (Note i havent yet welded the RH edge in the photo!)

 

[ChrisJohnsSculpture]

When doing something like this a bit of distortion is inevitable, the two main reasons being firstly that the weld itself will shrink as it cools and that the heat which goes into the surrounding metal will allow any residual stresses from the original manufacturing process to warp it somewhat as it locally softens as it gets hotter.

This effect is really something which is only ever managed rather than entirely eliminated.

Welding the reverse side alternately as you go along can help (although this is not always desirable for other reasons) as this tends to equalise the forces exerted by the shrinkage of the weld itself, at least to a degree.

If anything the gap may be a bit on the large side...the bigger the gap the more heat you need to put into the join.

The solution will probably be a combination of shrinking as close to shape as possible and then a bit of filler to get it spot on.

 

[hotrodder]

If there's no 'oil canning' (see this thread... http://www.mig-welding.co.uk/forum/showthread.php?t=22538&highlight=shrink) going on then it's sounding like the weld needs stretching not shrinking*.

As said, welds shrink. Using a traditional coachwork approach (with oxy or these days maybe TIG instead) the welds would be planished smooth instead of ground down. This stretches them while general heat distortion is dealt with at the same time. Sometimes referred to as 'hammer welding' although in most cases/circles hammerwelding refers to the practice of hotworking a short bead immediately after welding it i.e. while it's still 'kin hot

You get less distortion with MIG (especially using the thin metal technique) which along with it being a one handed process has made it more popular but... welds still shrink. MIG welds are typically too big and have too much reinforcement for hammering to have any effects, well positive ones anyway, at least until the weld has been ground out. MIG welds also tend to be harder which can sometimes result in strange behaviour when trying to planish them as they can be harder than the rest of the panel



* that's kinda a guess in that without 'reading' the panel it's impossible to be sure. It could be that the low spots just need bumping up and that purposely stretching the weld might result in oil canning

 

[the snooper]

may have been easier to joggle it and plug weld then skim with filler

 

[atomant48]

Thanks guys. I did consider joddling the edges but was concerned about causing distortion as the long join is curved upwards and I thought it might flatten it out.

I was quite constrained about where i made the join due to the bracing underneath. In hindsight (as always !) I would have cut out the bracing and then welded it back in once finished.

It was a fantastic way of learning the technique for thin metal though. I may try a skim of filler and see how one edge comes out.

Anyone have thoughts on maximum thickness for filler? Build it up in layers?

 

[steveo3002]

in my experiance no need to build it up in layers , ive seen it used pretty thick and never seen any cracks or problems over sound metal ...only time it fails is over rot

just use a good brand

 

[atomant48]

Thanks I'm feeling a bit more confident then. I've got some of the dinitrol alusoft I think it is called. Another skill to learn and will see how it goes. All takes time.

 

[hotrodder]

Anyone have thoughts on maximum thickness for filler?

Personally i'm not happy if something i did needs more than the thinnest of thin skims, virtually all of which gets sanded off. On the other hand i've seen some beautiful looking sleds/kustoms and VW campers that turned out to have been pretty much carved out of plod- 3/4" or so in places Especially if they've recently been imported from the states

 

[langy]

Thanks guys. I did consider joddling the edges but was concerned about causing distortion as the long join is curved upwards and I thought it might flatten it out.

I was quite constrained about where i made the join due to the bracing underneath. In hindsight (as always !) I would have cut out the bracing and then welded it back in once finished.

It was a fantastic way of learning the technique for thin metal though. I may try a skim of filler and see how one edge comes out.

Anyone have thoughts on maximum thickness for filler? Build it up in layers?

As thin as you can get away with, Modern fillers don't crack like the old stuff did, just put it on in thin coats, Upol Fantastic is my Fave, Don't use cheap filler as it will sink and come back to haunt you.

 

[atomant48]

Yep I agree- which is why I was interested in ways of pulling the join back up. I take it by your "hotrodder" name you're a bit of a whizz at this stuff? I love watching the "celebrity" bike and hotrod tv guys working magic. I actually contacted a local speed shop to see if they would take it on from scratch but unfortunately they wouldn't.

I have also looked into having a fibreglass bonnet made, so if I can get a decent 'cosmetic' finish it may be ok to use as the mould.

 

[hotrodder]

Dunno about whizz but most of my work is with ally that won't be painted and often has to be polished to appeal to the magpie in everyone. Filler isn't an option

My previous post was half tongue in cheek. The 2 - 3mm mentioned is nothing but with a bit of reading up on panel beating and some practice on scrap you should be able to improve on that if you want to/time isn't an issue etc. What you might find is distortion getting worse and/or oil canning at the corners- i'm assuming you haven't finished welding it yet or is just the pic?

 

[atomant48]

Thanks. No time isn't an issue, it's actually just a distraction from cutting out rot and previous bodged repairs during the rest of the "resto". Nice to do something different than lying under a bunch of rust with a grinder!

I actually had managed to tack it in quite a few places without any problems and it was when adding to the tacks that I noticed the "sinking". I guess i got too confident with the number of pulses at each location!

 

[atomant48]

And sorry, as you said, no I haven't finished the front edge, but the back edged went fine. This was really close to the "folded edge" at the rear of the bonnet (next to the windscreen) so had plenty of strength to resist warping i guess.

 

[hotrodder]

Yeah, when i'm making something i place welds where there's some shape as it reduces distortion and it's easier to correct what does occur. Not a lot of choice when working with a low crown panel (doors, most bonnets and many roof skins) but weld placement can still help... if i was letting in a blister/bulge/scoop into a bonnet i'd typically turn the edges of the hole up to meet the scoop so that the weld is surrounded by some geometry to stiffen things up