Vehicle roof patch help

[chevroletfreak]

Hey all, I've got a 69 Nova that had a vinyl top. I am replacing it and of course there is major pitting along with some pinholes. Unfortunately, no one makes a replacement roof for this body.

Most of the pinholes are in high contour areas around the front and rear edges of the roof panel. I am able to hand form the patches close to the correct profile. The problem I am having is that when I TIG weld the patch in, the sheetmetal shrinks and causes the patch to flatten out and lose its contour.

I have tried both stitch and spot welding with the TIG, but the same thing happens.

The best method I have come up with is over bending the patch and then welding 1 entire side of the patch at a time (instead of alternating to keep cool). Then I will go to the opposite side after it has cooled and do the same thing. This makes it more of an artform trying to play the heat shrinkage game.

Another thought I have considered is creating a flange to mount the patch, but I am concerned that the thicker areas of the roof where the flange is will be evident when the roof naturally flexes in the sun. It is possible that the new vinyl texture will hide that. Don't know, just thoughts.

Any ideas are appreciated. I would prefer the science method over guestimating method.

 

[CNGsaves]

You've had enough posts . . . surely you've seen this.

Here you go:

 

[chevroletfreak]

LOL. You're right.
Really, it is just a cutout hole and some rusty pits, but I'll get some tonight when I get home.

 

[creativecars]

How big and how many patches are you talking about? It is better to have larger/fewer if that is possible. Do you have access to a mig? Preferably with .023 wire and argon/co2 mix.

 

[rsanter]

That is because you are getting shrink. The base metal is cold the welding filler goes on hot. When the filler cools it shrinks and pulls on the metal around it.
You need to stretch the metal a little with a hammer and dolly and it will gomright back to the shape you want it to be.
Do the hammering on the weld so flatten and stretch it before grinding and it will release some of the pull it has on the base metal.

Another option is to take a torch and anneal the weld out and that will release some of the welding stress

Bob

 

[Vegaman_Dan]

MIG and small wire at the lowest setting will help with heat distortion. I'm fighting that myself currently and there is a bit of a learning curve to it to be sure.

If you go with a lapped joint, the joint may be visible under paint as the metal will be double thickness there and expand at a different rate in the sunlight. As this would be covered by the vinyl, then I doubt you'll ever see any sign of this. I think in your case, I'd leave the flat patch in there, skim coat with filler, call it good. The vinyl will hide all the work you're putting into this.

 

[MP&C]

Check out post # 19 on this thread

http://www.garagejournal.com/forum/showthread.php?t=86935


While not identical to what you describe, the patches for a roof at a window opening will normally include the pinch weld flange of the window opening and extend out far enough into the roof to fix all of the associated rust, and/or place the weld at a location where you have access inside the roof to planish the inner weld. Typically the spot welds / plug welds on the flanges do not require planishing, but the **** weld where your patch is installed into the roof panel will.


This one shows a larger roof patch, but it attaches to a drip rail instead of the a window opening.


http://www.jalopyjournal.com/forum/showthread.php?t=337728


The question as to your methods is are you adding filler rod or performing a fusion weld with your TIG? If adding filler, the speed of your process (or lack thereof) may be causing the HAZ to become larger, affecting more metal, resulting in more shrinking. With a fusion weld process, you would need absolute tight joints, but can perform the weld faster with higher heat, resulting in a smaller HAZ for less effects of shrinking in the panel...


...and just as a clarification, I'll disagree with Dan's one statement above. I never base the control of distortion on using the lowest setting. The heat setting should be based on the material thickness and providing a full penetration weld to that thickness. Any distortion that results can be resolved with planishing to stretch the weld and HAZ where the heat from the weld process has caused shrinking. Trying to control it by using minimal heat will normally result in poor penetration of the weld, and risks failure of the joint.

 

[91bronc300]

Can you just smear RTV over the pinholes?

 

[MP&C]

Can you just smear RTV over the pinholes?

The owner can choose to use whatever method he likes, but while it is apart, it is much cheaper to fix correctly now than have the problem re-appear in two years and need to repeat this process again. Since there are no pictures I am basing my thoughts on a worst case scenario, but the mention of PINHOLES in the original post means replacement metal to me. Seldom is the case that the few pinholes that are showing are the only problems brewing inside. Open it up, fix it right, fix it once.

 

[Vegaman_Dan]

...and just as a clarification, I'll disagree with Dan's one statement above. I never base the control of distortion on using the lowest setting. The heat setting should be based on the material thickness and providing a full penetration weld to that thickness. Any distortion that results can be resolved with planishing to stretch the weld and HAZ where the heat from the weld process has caused shrinking. Trying to control it by using minimal heat will normally result in poor penetration of the weld, and risks failure of the joint.

Agreed. I would always use the lowest heat setting needed to do the job correctly. Putting more heat into the joint than necessary is also risking a poor weld joint while causing distortion to the patch. Getting advice from veterans is very helpful indeed, so I would definitely listen to comments like yours.

It's all about balance. Use what works best for the situation as needed.

 

[CARS]

Can you just smear RTV over the pinholes?

Dad??? When did you join the forum?? (you butcher )

 

[chevroletfreak]

Pics...finally!

It is really hard to get a pic of the pitting. One is of the entire roof. The other shows the largest area I am attempting to replace. There are others though.

There has been some excellent advise. Just a couple of notes. I am trying to keep this as original as possible, but maintain its driver quality. So, I would like to repair the roof for usability by stopping water intrusion and continued rusting. On the flip side, I really don't want to replace the roof, but will if I have to.

Another note, as you can see where I cutout I am unable to do any hammer and dolly work because there is a channel under the patch

 

[chevroletfreak]

In the interest of pictures, here is a bonus. This is one of the car right after I bought it. It is numbers matching drivetrain (350, 3-spd on the floor) with 81k miles.

 

[chevroletfreak]

The question as to your methods is are you adding filler rod or performing a fusion weld with your TIG? If adding filler, the speed of your process (or lack thereof) may be causing the HAZ to become larger, affecting more metal, resulting in more shrinking. With a fusion weld process, you would need absolute tight joints, but can perform the weld faster with higher heat, resulting in a smaller HAZ for less effects of shrinking in the panel...

I actually tried both methods. My joints aren't perfect, but are close. I am bending by hand and then using a flap disk to tune the edges.

Do you recomend a series of tacks like MIG or just go for it with a fusion weld and add filler if it opens up?

...and just as a clarification, I'll disagree with Dan's one statement above. I never base the control of distortion on using the lowest setting. The heat setting should be based on the material thickness and providing a full penetration weld to that thickness. Any distortion that results can be resolved with planishing to stretch the weld and HAZ where the heat from the weld process has caused shrinking. Trying to control it by using minimal heat will normally result in poor penetration of the weld, and risks failure of the joint.

I am using a Miller Diversion 165 set to 20g and using the thumbwheel to fine tune the arc. I am by no means a pro...just a hobbiest. With that said, Inormally I would expect to do some dolly work. In this instance I am unable to due to a channel directly under the patch.

I finally got some lighting in my garage and hope to work on the patches this weekend. Any other advice is appreciated.

 

[hemifalcon]

I'd recommend getting your hands on a donor hardtop--or a sedan and using patch sections as necessary.. I'm not a TIG owner, and do bodywork with my Miller 140 without issue.. But--getting rid of all those pits is going to be difficult. Selecting a good quality epoxy filler for the pitted areas and smoothing everything out would be sufficient..
The other alternative is swapping the entire roof from a good hardtop donor..

 

[MP&C]

The factory panel should have been 19 ga, which is hard to find in some areas of the country. I primarily use 18 ga and for that my welder is set at the 16 ga setting. Unless you are using 20 ga, which IMO is too light, a 20 ga welder setting will likely give you a cold joint on the Mig if using single tacks. If you look at my shop projects thread linked in my signature, the last couple pages show some "dot" welding using the Mig. If you would rather use the Tig, I would recommend either a fusion weld, hotter and faster speed, or a continuous weld while adding filler. If you attempt dot welds as you do with the Mig using the Tig, you will have considerable HAZ as you heat and add filler. I believe in a heads up Mig vs Tig in dot welding, the Mig will have less HAZ as the filler is there instantly, unlike a manual feed.


As an alternative to fabbing individual pieces, the purchase of a full replacement skin or fabbing a partial may be another option. In the following link is a windshield panel we made for a chevy p/u, not exactly the same panel, but yours may be made using a similar process in a larger panel for less welds...

http://www.garagejournal.com/forum/showthread.php?t=149532

 

[ng8264723]

The easiset thing would be a new roof. I would reattach using panel adhesive. I have replaced a roof via welding and via panel adhesive. Guess which i will never do again?

 

[chevroletfreak]

Apparently I was not subscribed to my own thread. Thanks for all the input.

I did not try thicker metal with the TIG, but did attempt to fuse the panels together. The metal was just to thin and it varied in thickness due to the pitting. So, I could not find a happy medium and kept punching through.

I located and purchased a donor vehicle with a good roof (not available aftermarket).

I also borrowed a Hobart 210 MIG from a buddy. However, I will have to buy a bottle and wire to convert it to GMAW. Is this necessary or would flux core be acceptable.

I will start on the donor skin first by melting out the lead seams with a torch. I will use a combination of grinder, cutoff disks and spot weld bit to remove it. The keeper vehicle should be a little easier since the skin is bad anyhow. I will have to Spot weld around the window seams and am considering 3M panel adhesive to attach the panel to the drip rails above the doors.

Is 3M 8115 or 8116 better?

 

[ng8264723]

DO NOT WELD IT ON!!!!!!!!!!!!
Use the panel adhesive. First tape off the seemed area then epoxy primer the underside. That way there will be no uncovered metal. Get a million clamps. Wood clamps work well. Use the panel adhesive and clamp evry couple inches. the next day your are done

panel prep.

Donor car
melt out lead with torch (propane is lwerr heat and works well for this) use a wire brush to move the lead.
drill each spot weld with a 1/8 inch bit
then drill using a 3/8 bit.
use strong flat scraper to break through remaiing spots
gently hammer and dolly edges.

actual car
drill with spot weld cutter
use the non hole saw type
grind clean
i would use an air hammer and cut off the roof first leaving approx an inch or two wear it is spot welded
that way you can wiggle the pieces back and forth

after all clean you are ready for the adhesive. use wax paper so it doesn't stick to the clamps. don't worry about the holes fill them with adesive

 

[MP&C]

There's really nothing wrong with welding it on, the factory had pretty good luck with it... I used a Miller 220v spot welder on a 55 wagon roof repair, it hasn't fallen off yet. I think either would be a viable option, but I'd prefer the factory method myself.

The spot welder in action:

This looks darn near a factory finish!

The spot welds completed...

 

[chevroletfreak]

I see the plus of both sides.

Weld - 3M 8115 and 8116 both show that quarter panels recommend welding at the rear of the panel and any seams. I expect the panel to require a bit of stretching and pulling to get to the correct shape. I suspect welding will hold the panel in place better reducing the stress at the connection points.

Panel adhesive - 8115 seems to be the standard, but 8116 is GM approved. Is either actually better? Using strictly adhesive, I could much easily remove the panel if I screw up

I am heading out now to remove the donor roof.

For future reference what is the recommended procedure for connecting the previously leaded seam at the sail panel and windshield post?

 

[chevroletfreak]

Well, my buddy kept bugging me so all I got done last night was melting the lead out. That was pretty stright forward and may have taken half an hour. Just heat it with a torch and it'll drip out leaving nice clean metal below. I was supprised at how much lead was in there. I estimate it was more than an inch wide by 3/8" deep.

That leaves me with 100+ spot welds to drill out tonight. Then on to my keeper Nova.

I'll try to get some pics up tonight and order some 3M 8116.

Any recomendations for filling in the seam where the lead was?

 

[chevroletfreak]

I got 4 hours into the removal of the donor vehicle roof last night. That got me almost half way to drilling out all the spot welds. I was running out of time so I ended up just cutting it off at the pillars to bring it home for completion. Nothing else exciting at this point.