AAE
Learner
Like Santa?
Oil Canning Roof
- Thread starter Machspeed
- Start date
You could not have said it any better!
The OP has had "oil canning" because his roof sheet metal is now longer and wider than it was before blasting. It is no longer taut in the same "footprint" and has yielded in tension and become thinner somewhere. What yields and elongates in length also contracts and thickens due to conservation of mass somewhere. And it's not always evident where it might be on first thought. If something stretches in one or two dimensions then in the third it must contract and thicken somewhere--otherwise it separates. Carbon steel welds don't "shrink" except when they go from liquid phase to solid phase delta iron crystal structure at about 2700F and then from delta to gamma phase at around 1700F or so. They expand and grow in volume as they pass on cooling at about 1330F. The volume change at room temperature will be an increase in thickness whether filler metal is added or not. The thermal energy absorbed in heating something to about 1100F is simply to add in energy to lengthen something in three dimensions is just reversible on cooling back to room temperature because no crystal phase change has occurred. But go past red heat and it's going to be something else from a phase change. Find someone with a ultrasonic thickness (UT) meter and you will find where you are thin and thick. They are cheap now and mine is accurate to about 0.002" in 2"--- at least at room temperature
The picture of a 2" boiler tube bent in a hair pin on a 4 1/2" centerline is a damn tight bend and bend "cold" at room temperature. 40% of the wall thinned on the outside and that surface lengthened dramatically. The tube's wall measures 0.132" unbent and 0.080" thinned on the outside. It breaks in tension at about 48% wall reduction consistently. The inside surface grew in thickness and yielded in compression--essentially the crystals slid and piled up on one another--it got dramatically shorter. People focus too much on the stretch-out without realizing the thickening. You don't get one without the other. I sectioned that bend longitudinally to "open it up" to measure. That sample was about 100,000th one out perhaps 28 years ago in a production run. That same machine is well over 1.3 million bends on hairpins and 45 degree bends. It still bends the same if the carbon and manganese content of the tube is low.
The same fundamentals apply to the OP's problem. Since your roof panel is pinned from movement at the edges--that's likely where it has thickened since metal crystals slipping and moving had no more room to flow from blasting except to go the edges. Those are your fixed points. Nothing can move here at the drip line to increase in length from the blasting in width and the same for length with the windshield and back window framework . You need to get metal back to where it is thin from where it is thick. You can do that with "gathering with a hammer and dolly" or by what I described above. Shrinking discs are bad news since you can get to a phase change to thicken just from frictional heat alone from the discs on the very, very surface crystals of the sheet metal but not the through thickness. Then you don't have an even pull. Whole thickness of the sheet metal or nothing if you want to stay put out in the hot summer sun outside or a -30F day in the winter after working it.
That is one of the best posts I've ever seen...
I had my car dipped. After watching countless car shows where they immediately send everything to the blaster, I thought that was a mistake. I've been watching these guys youtube channel and it has what I consider a lot of good info. It shows even careful blasting distorts the steel.
With the shrinking disk I bought many years ago, it will heat the sheet metal panel red hot all the way through, so not sure about the phase change only at the surface. Also, the technique is a heat/quench process. I only ever had mild results as I think there is a large learning curve and I wouldn't "practice" on a good roof.
Machspeed
Member
This roof dilemma has really set me back. In fact, I'm now scared to jack with it. Right now, it looks perfect just the freaking oil canning. I had a couple of guys on another forum tell me to leave it and stick sound deadener to the underside of the roof, something I intended to do anyway. Seems kinda halfazz but far better than me warping the crap out of it and using a ton of filler......arghhh! I just don't know!
JimKueneman
Mopar Nut
easy for me to say because I have been using the donut dolly and have a feel for it but I would use it and just a few love taps and see if it starts to get better or worse... if worse stop. I have found it is hard to really screw it up with that method... the movement and subtle and you can stop if it looks like it is going the wrong way.
MJM
Promoted Users
1. Hot glue some paint sticks to the underside of the roof.
2. Spray two coats of epoxy.
3. Lightly sand epoxy to reveal any high spots.
4. Remove paint sticks as necessary to use donut techic..
5. Replace paint sticks if needed to stiffen metal for skim coating.
6. Skimcoat roof with filler.
7. Sand roof.
8 epoxy prime roof again
9. Lightly sand roof again to check for highs and lows.
2. Spray two coats of epoxy.
3. Lightly sand epoxy to reveal any high spots.
4. Remove paint sticks as necessary to use donut techic..
5. Replace paint sticks if needed to stiffen metal for skim coating.
6. Skimcoat roof with filler.
7. Sand roof.
8 epoxy prime roof again
9. Lightly sand roof again to check for highs and lows.
MP&C
Member
The support for your roof skin is similar to the support for an elliptical roof on a building structure. Once a dent occurs, it moves forces around and interrupts that support. You'll see tight areas and loose areas and, on a good day with the roof in front of you, can find it very much a challenge to work out the area to address and those to leave the hell alone. My suggestion in order to LOCATE the area that needs addressing.....
Typically you'll cycle the oil can in and out in order to see the outer perimeter of the affected area. You've already done that and identified the perimeter, whether you know it or not... But lets continue for identifying the areas not so easy to pick out... Once the perimeter of the oil can is located, mark the outer perimeter using tape, sharpie, etc. Now that the outer perimeter of the oil can is located, using your thumb on one hand, apply slight thumb pressure along the outer perimeter and while holding this pressure, use the other hand to cycle the oil can in (and back out if it doesn't do it by itself) Then move the thumb pressure around about an inch on the perimeter, hold the same pressure, and cycle again. What you are looking for is a location along this perimeter that will LOCK the oil can from cycling. If you locate one, continue around the entire perimeter, repeating this process as there may be more than one "sweet spot" that needs work. This (these) are the area in need of work to resolve the issue and not compound things by guesswork. But before starting down that rabbit hole, read on.....
In looking at your initial video about two thirds of the way in, it appears in (picture) that your perimeter cycles just shy of that body filler area (as one would expect) I don't feel you can address anything oil can related without starting here first. A dent is a STRETCH. Aggressive media blasting is a STRETCH. Any area welded is a SHRINK. Whatever caused your issue, the opposite will need to occur to fix the situation. I would start by removing the filler and bumping/shrinking that dented area as needed, then see how the oil can has responded. A dent has introduced a "lock" and caused an interruption of forces across the entirety of the panel, as can be seen in your video. As a side effect, the dent may also pull from other adjacent areas, possibly making them loose. In the interest of showing why you don't fuck around with guessing on an oil can, IMO the majority of your issue was caused by that dent and NOTHING should be touched until the dent is fixed and the results of said repair are assessed. Anything else done before addressing that dent just risks changing forces in the panel unnecessarily.
Last edited:
Machspeed
Member
I could not have asked for a more detailed explanation and how to than this!!! This should be put up somewhere for reference. Thank you Robert (MP&C), this is not the first time you've bailed me out!
Gentlemen, thank you for your input! You all are truly a blessing to me. I will report back on how this transpired.
Machspeed
Member
So, I've not messed with the roof yet other than to remove the filler that I put on it, sand it, and apply a couple coats of epoxy. Tonight, I plan to block sand it again in order to find the highs and lows.
Since my original post, I've continued to research, ask questions and weigh my options. Can't tell you how many stories I've read where nightmares were created trying to fix such a problem, even by supposed skilled individuals. Spoke with a local guy who worked in the business and taught in a Vo-Tech for many years. I sent him my video as shown above. He's also a big Mustang guy and knows these cars in and out. He weighed out three options for me:
Something noted and not sure it matters, but I was in the shop last night and pushed down on that roof and it seem more taught, or rather difficult to push down on. Still cans, just not as easy. The shop is cold, in the 40's, as I've not been out there working and the heat is off. This makes me think that the roof is indeed stretched through the media blasting and therefore needs shrinking. Thoughts???
Since my original post, I've continued to research, ask questions and weigh my options. Can't tell you how many stories I've read where nightmares were created trying to fix such a problem, even by supposed skilled individuals. Spoke with a local guy who worked in the business and taught in a Vo-Tech for many years. I sent him my video as shown above. He's also a big Mustang guy and knows these cars in and out. He weighed out three options for me:
- Attempt to fix and hope for the best. Not knowing exactly what caused this makes the fix scary.
- Put a brace in the roof, either welded in or glued in with panel bond.
- Cut the top off and put another on.
Something noted and not sure it matters, but I was in the shop last night and pushed down on that roof and it seem more taught, or rather difficult to push down on. Still cans, just not as easy. The shop is cold, in the 40's, as I've not been out there working and the heat is off. This makes me think that the roof is indeed stretched through the media blasting and therefore needs shrinking. Thoughts???
1A Rock
Promoted Users
Does the roof have a specific spot where it cans. A tiny bit of manipulation there may be in order, the most active place may very well be the most stretched, Adding anything extra always seems cheesy to me. Take your time and try to correct the issue is better than treating the symptoms. JMO
MP&C
Member
The trouble with panel bond is that it will not allow movement. So when your roof wants to expand in the hot sun, anything glued to the bottom side will want to keep it from doing so, and you’ll likely see a pucker in the roof as the rest of the unglued portions do still expand. So I hope he wasn’t gluing the roof skin
Your roof panel contracted in length both across the roof and over its fore to aft dimension with a drop in temperature in your shop. Steel contracts or expands 8 millionth of an inch per inch of length for every 1 degree F temperature change. Thickness contracts too--but its dimension relative to length dimensions is insignificant. That's why your looseness feels less when colder. But by the same token--take your roof out as oil canned in the hot OK summer sun where you live and let it go to 125F or so and your problem will be even more--horrendously more.
(2.) isn't going to solve that problem, it will break the panel bond or depress the brace or shift the growth to can or distort another part of the sheet when exposed to the sun.
(2.) isn't going to solve that problem, it will break the panel bond or depress the brace or shift the growth to can or distort another part of the sheet when exposed to the sun.
Machspeed
Member
I get that, but what I was trying to relate to was a possible fix, as I'm not 100 percent certain what caused it. I believe the media blasting was the cause, but not certain. The cold made the metal shrink and stiffened the roof, so I'm thinking shrinking in the right places would remove or decrease the canning??? The media blasting, according to MP&C stretched the roof, makes sense that it needs shrinking. There are a few small dents in the roof that I plan to address first, then will go from there......slowly.
I see the logic mentioned here in the brace and will abort.
Thanks for the input!
I think that's a good plan, work out those small dents and the one that had filler in it. Careful not to aggressively hammer on dolly much, the donut dolly with a light touch mentioned earlier may come in handy and checking your progress often. I'm eager to watch you get this done, there's no heavy work and it'll be good bragging rights to the guys that get it.
Damn I just repeated what Robert said above...
Damn I just repeated what Robert said above...
Machspeed
Member
Blocked the roof last night with 180 and exposed the highs and lows. The areas circled are the areas of concern.
Thanks!!!
- At the very front of the roof are several shallow dings and dents. I had previously filled this with filler for a smooth surface. I'm not sure how I would address this area, as there is a brace right behind it all and I can't get a dolly back there. I do have a stud gun that I could use to pull up the lows. Thoughts?
- The area circled in the top of the picture (passenger side roof) is the biggest ding in the roof. I took a picture of the inside of the roof to show it's obvious downward protrusion. See photo below. I will probably address this dent first and see how it affects the oil canning.
- The area at the back of the roof is a large very shallow low. As seen in my initial video, you can see that I had filler in that area too. That also has been removed. Using Robert's (MP&C) instructions on localizing the issue, I found this area to be the most probable cause. When I pushed up from the inside in that area, I found the oil canning to pretty much diminish. Looking for suggestions on addressing this area? My thought was to address this area from the underside of the roof with a shrinking disc???
Thanks!!!
MP&C
Member
The dent at the top (passenger area) will pull other adjacent metal along for the ride as it went inwards. This is the first area to address, and then see if the larger area at the back has changed. Any disruption to the roof’s support system (the dent) can affect the support in other adjacent areas, especially lower crown areas.