Pitfalls of flanged weld seams

[MP&C]

Ok, you guys have heard me talk incessantly about the downside of using flanged seams in sheet metal repair, discussing the ghost lines that can and will occur with these type of repairs. Today I got a short video to show the effect. This car was neat as a pin, with extensive rust repair performed, in basically the same locations as the 55 we're working on. The only flaw I could find in the paint was a line in the finish where the rear tailgate repair patch was seamed. I asked the owner if he had used flanged seams in the repair, and he said yes. He did offer for me to take pictures that others may learn from it..

Butt welds people!!!!!!!


https://www.youtube.com/watch?v=OGhFEfVqxb0


.

[NickP]

dang it................................................ ................................................sa d

[WagonCrazy]

Good eye Robert.

[chevynut]

Wow that's a shame. I agree that you shouldn't ever use flanged seams and butt welds are the only way to go. However, you said this, and I disagree:

".......shows how the differing expansion rates of one layer vs. two layers eventually results in a ghost line exactly where the repair took place."

The expansion of metal doesn't depend on the thickness, assuming it's the same temperature throughout. Since steel is a good conductor of heat, I doubt the temperature varies measurably from the top sheet to the bottom sheet. I'm betting the line is a result of the different expansion rate of the filler material versus the steel. In fact, it may be due to shrinkage of the filler. Does the line show regardless of temperature? If so, it's probably due to shrinkage of the filler in the seam. I'll bet a leaded seam wouldn't show that line.

[chevynut]

Also, I wanted to comment on something that we discussed years ago on the other site. A guy had a trunk that was sandblasted and there were dips at every location where you could access the sheetmetal from the bottom. I mentioned that it was due to the sand stretching just one side of the metal, and guys said that couldn't happen because the metal was too thin. Of course it can happen. I think Robert proved that was the cause with some experimentation.

I actually was a mild victim of this. I blasted my Nomad along edges and jambs with 80 mesh silica sand after stripping the paint. I blasted the roof brace and barely hit the roof sheetmetal with the sand. Same with the hood....I blasted the rear brace from the bottom and hit the upper sheetmetal a little, even though I was trying to be careful not to do so. After priming and blocking, I noticed I had a dip along each brace. On the roof it only took another coat of primer to fill, but I actually had to do some bodywork to get the hood bent back up. It was down 1/32" or so....I didn't measure. I should have put a sheetmetal shield between the braces and outer sheetmetal before blasting. This can happen on fenders too....probably did to me.

A couple of weeks ago I had a rusty sheetmetal cover panel that was oilcanning. I put it into my glass bead blaster and blasted it clean, then I kept blasting the edges of the cover until it wouldn't oilcan any more. It was clear to me that the metal was stretching from the glass beads hitting it. I put a flat piece of sheetmetal in the blaster and blasted just one side.....it caused it to curve. When I blasted the other side it straightened out. And that was with glass beads that are almost like powder....imagine what sand would do.

So this shows that you CAN stretch just one side of a piece of sheetmetal and cause it to warp. Also, if you know what you're doing and have access you could probably straighten it back out by blasting the other side. If I'd known that at the time, I probably would have just blasted a line on my roof and top of the hood.

[MP&C]

Wow that's a shame. I agree that you shouldn't ever use flanged seams and butt welds are the only way to go. However, you said this, and I disagree:

".......shows how the differing expansion rates of one layer vs. two layers eventually results in a ghost line exactly where the repair took place."

The expansion of metal doesn't depend on the thickness, assuming it's the same temperature throughout. Since steel is a good conductor of heat, I doubt the temperature varies measurably from the top sheet to the bottom sheet. I'm betting the line is a result of the different expansion rate of the filler material versus the steel. In fact, it may be due to shrinkage of the filler. Does the line show regardless of temperature? If so, it's probably due to shrinkage of the filler in the seam. I'll bet a leaded seam wouldn't show that line.

Therein lies the problem, it doesn't remain the same temperature throughout.. To better illustrate, let's use an extreme example. Take a 1' square of 18 ga sheet metal and a 1' square of 1/2" plate out of your air conditioned shop and place them in direct sunlight. Do they both increase in temperature equally, or does the thin piece warm up more quickly? Does it not also cool off more quickly when the sunlight has gone for the day?

In the same fashion, it will take longer to heat up two layers of sheet metal over one single thickness. Differing heat rates, differing expansion rates, and ghost lines that show exactly where the flange seam is located. It does not happen overnight, it may take a year or two, but given sufficient heating and cooling cycles, the ghost lines will appear.

[Rick_L]

Much if not all of the reason for a ghost line is that the lap joint is stiffer than the surrounding metal. It doesn't matter if the joint heats up at a different rate than the surrounding metal, it's just that the surrounding metal moves more than the stiff lap joint. Same for cold. It doesn't matter what rate it heats or cools.

We can argue all day over the details, but it happens and that can't be denied.

The other disadvantage of a lap joint is that because the joint is significantly stiffer, you can't straighten it after welding, It's even difficult to beat a high spot down.

[chevynut]

Therein lies the problem, it doesn't remain the same temperature throughout.. To better illustrate, let's use an extreme example. Take a 1' square of 18 ga sheet metal and a 1' square of 1/2" plate out of your air conditioned shop and place them in direct sunlight. Do they both increase in temperature equally, or does the thin piece warm up more quickly? Does it not also cool off more quickly when the sunlight has gone for the day?

What you're saying is true. Thicker material has more thermal capacity, but it equalizes at some temperature pretty quickly. And it has the same expansion rate. If the seam only showed up during temperature transitions, the difference in thickness might make sense. But I'll bet you can see it in the middle of the day when the car has been in the sun and the temperature is stabilized. That tells me thermal capacity has nothing to do with it.


In the same fashion, it will take longer to heat up two layers of sheet metal over one single thickness. Differing heat rates, differing expansion rates, and ghost lines that show exactly where the flange seam is located. It does not happen overnight, it may take a year or two, but given sufficient heating and cooling cycles, the ghost lines will appear.

If it takes years to happen that proves that it's not related to metal thickness, as that never changes. What does change is the filler shrinks. If it was due to metal thickness you'd see it immediately.

I also don't think the line has anything to do with the stiffness of the metal. When metal expands, nothing is going to stop it. It expands at the same rate whether it's hardened or annealed, and a small area like that line isn't going to bow differently just because it's stiffer due to thickness. You'd see it in a larger area of the panel. Same argument if it was thickness that was causing it.....you'd see it over the entire doubled up area.

I maintain it's the filler that you're seeing. There's a narrow channel of filler that's applied on the lap joint and the thermal expansion of filler is at least 6 times that of the steel (polyester is 10X). That's a huge difference. If the lap seam was fully welded and ground, or filled with lead that has a thermal expansion closer to steel (but still 2x) and no shrinkage over time, I don't think you'd see that line.

Is it time for an experiment? Make a few panels with lap seams and fill with one with lead and another with polyester filler. Or take a panel and roll a narrow bead in it and fill it with filler. Then paint the panels and put them in the sun.

[chevynut]

Out of curiosity and since I can't do anything but sit around with my messed up leg :( I decided to do some research on this topic. There are literally hundreds of posts about ghost lines on dozens of boards. Guys are seeing ghosting over all sorts of welds, even butt welds ground smooth on both sides. Some guys have even metalworked the joint to perfection, used no filler, and still see the ghost lines. Nobody seems to know what the true root cause is but there are a lot of theories, some that make some sense and some that don't.

Guys have speculated about it being due to the hardness of the MIG weld. Hardness should not affect thermal expansion. Some say it's due to filler shrinkage, or difference in thermal expansion rates, but that doesn't explain it all when no filler is used.

Some guys have this problem and some don't. Seems like guys who use a lot of bondo or other filler don't see it as much. That's probably because they're covering it up but it's still happening under the filler.

So I got to thinking about what else could cause it. It looks like the thermal expansion of cold-rolled steel may be slightly anisotropic, meaning it's different in different directions. It does happen in stainless steel and some other metals:

https://www.google.com/search?q=expansion+of+steel+directional+rolled+col d&ie=utf-8&oe=utf-8#q=anisotropic+thermal+expansion+of+cold+rolled+s teel+


We all know that any rolled sheetmetal, steel or aluminum, has a "grain" direction as the crystal structure is altered from rolling. Aluminum should usually be bent across the grain because it can crack if bent with the grain in some situations. I don't think anyone has really studied the thermal expansion in sheet steel. So is it possible that when a patch is installed with the "grain" of the sheet non-parallel with the patched panel they expand at slightly different rates? That would clearly show a line at the weld in a highly polished surface. But it would show up immediately, not after years of temperature cycling.

In the case of the tailgate Robert posted, it would be interesting to tear it apart and study it to see how it was bent, how it was welded, how it was filled, and if the grain of the patch is in the same direction as the grain of the rest of the tailgate. There has to be an answer somewhere, but it seems elusive. Now I'm worried about my welds. :eek:

[shine]

in the early 80's i owned a van shop. we painted 10 per day . at the end of the year all that was left was 2 tone chevy duallys . there was a 3/4 inch tape strip . we removed it, da sanded with 80 grit primed and painted . outside the 3/4 stripe would still be visible. also seen this with many cars that had factory tape stripes. especially the firebird . those eagles would ghost back every time. this was lacquer then urethane clear.

[55 Rescue Dog]

I've always wondered if that is why gas welding was a preferred method that has used by some builders since the early days. The welding temp is much lower, and the weld more ductile.

[NickP]

in the early 80's i owned a van shop. we painted 10 per day . at the end of the year all that was left was 2 tone chevy duallys . there was a 3/4 inch tape strip . we removed it, da sanded with 80 grit primed and painted . outside the 3/4 stripe would still be visible. also seen this with many cars that had factory tape stripes. especially the firebird . those eagles would ghost back every time. this was lacquer then urethane clear.

Shine! Welcome

[BamaNomad]

I need to do some research on this, but off the top of my head, I'm not sure I agree with CN's statement below...??

"Guys have speculated about it being due to the hardness of the MIG weld. Hardness should not affect thermal expansion."

?? I think checking the thermal expansion coefficients of various hardness steels would supply an answer, but I don't have a manual handy...?

[Rick_L]

From what I know, if it's carbon steel of any hardness (carbon content could vary), thermal expansion will be the same for all.

Stainless steels can have different thermal expansion rates, but they will only vary in a fairly small range, unless it's a specialty alloy with small thermal expansion rate (and that may not even be "stainless steel" per se).

Aluminum will have greater expansion rate, and the various alloys will vary, but again not by much.

Again, I don't think thermal expansion rate is what's happening when you have a "ghost line", it's primarily joint stiffness, and maybe the amount of metal in a given location - I.e., a lap joint has roughly twice the metal as the surrounding panel. Also a lap joint is 4 times stiffer than a panel, or a butt welded panel.

Remember too, that an adhesive bonded lap joint will also exhibit "ghost lines". So it's not just welded joints either.

[LEE T]

Shine is right about the old pin stripes showing through. In the 1970s and 80s a lot of the big trucks had multi colors with pin stripes between the colors, and when a driver changed companies and got a new paint job, those stripes would show no matter how well it was stripped and prepped before paint.

[chasracer]

I worked in a body shop doing mostly tear downs and some final assembly work. On the odd car I was taught some metal patching. The owner had started working in the business back in the 40's after the war so he was up in age. The one thing I remember he said about metal from this work lesson is that unless you can use a patch from the same era of car as the one you are working on, you can never get it perfect - the metals are different and that's just the way it is. But with a big smile, he also said that means repeat business too!

[chevynut]

The one thing I remember he said about metal from this work lesson is that unless you can use a patch from the same era of car as the one you are working on, you can never get it perfect - the metals are different and that's just the way it is. But with a big smile, he also said that means repeat business too!

I don't think it's necessary to get a patch from the same era car but there could be slight differences in metal composition that could change the thermal expansion rate slightly. I don't know how much of a difference in thermal expansion rate would start to show problems in normal use. I would never use stainless steel to patch mild steel panels but I'm sure it's been tried.

[chasracer]

I don't think it's necessary to get a patch from the same era car but there could be slight differences in metal composition that could change the thermal expansion rate slightly. I don't know how much of a difference in thermal expansion rate would start to show problems in normal use. I would never use stainless steel to patch mild steel panels but I'm sure it's been tried.

I believe what he was referring to is that over time, the metal used for automotive panels has changed dramatically. We're taking a 1950's metal that is totally different from the 4x12 sheet of patch metal and combining the two together. Each metal is going to have it's own properties. Then we add a high amount of heat, plus another filler material to combine them. Today's high strength steels are both thinner and different from most of the metal in our cars. I'm not sure we can expect a perfect application and the lap seam that started this discussion, in and of itself probably leads more to the ghost line than anything else.

[Rick_L]

Low carbon steel was preferred then for body panels, as it is now for patch panels. If the steel composition had changed, so would the alloy designation. I don't think mild steel has changed a bit, let alone "dramatically".

Also, as we all know from the current discussion on steel and aluminum tariffs, most of the mild steel these days is imported. And don't think for a minute that steel from 3rd world countries is "advanced". It's made there the way it was made in the US 75 years ago. That's why they are 3rd world and we aren't.

[chevynut]

I believe what he was referring to is that over time, the metal used for automotive panels has changed dramatically. We're taking a 1950's metal that is totally different from the 4x12 sheet of patch metal and combining the two together. Each metal is going to have it's own properties. Then we add a high amount of heat, plus another filler material to combine them. Today's high strength steels are both thinner and different from most of the metal in our cars. I'm not sure we can expect a perfect application and the lap seam that started this discussion, in and of itself probably leads more to the ghost line than anything else.

Yes I agree that if you take a patch piece from a car from the 50's it's going to be different composition than metal taken from a car built today. Today's cars are made with thinner, stronger steel. I wouldn't think of using metal from a modern car for a simple patch panel.

However, that doesn't mean you can't get metal of the same composition that they used in the 50's, which was probably a mild (low carbon) steel. Plain mild steel sheet you can get today is usually something like 1008 to 1020 steel (.08% to .20% carbon) and it's likely similar to what they used back then. I'm not sure but I doubt the thermal expansion of mild steels differs much.

[BamaNomad]

Yes I agree that if you take a patch piece from a car from the 50's it's going to be different composition than metal taken from a car built today. Today's cars are made with thinner, stronger steel. I wouldn't think of using metal from a modern car for a simple patch panel.

However, that doesn't mean you can't get metal of the same composition that they used in the 50's, which was probably a mild (low carbon) steel. Plain mild steel sheet you can get today is usually something like 1008 to 1020 steel (.08% to .20% carbon) and it's likely similar to what they used back then. I'm not sure but I doubt the thermal expansion of mild steels differs much.


CN? Do you have some information to back up this statement?? " Today's cars are made with thinner, stronger steel. "

I don't think there's any doubt of today's sheet metal (used in cars) being THINNER, but 'stronger'?? That differs from my understanding. I think today's car bodies likely use 'bends and contours' to make the panel strong enough, in spite of it being thinner (and likely less strong)...' If today's steel used in auto bodies is stronger, I'd really like to see some detailed information so I can modify my opinion...

[Rick_L]

I think that all the "stronger" steel is not the steel panels you see - but the structure they are spot welded to. Which doesn't really compare to any kind of 50s/60s maybe even 70s cars.

Also think about that structure. It needs stiffness as well as strength. Stiffness doesn't come from special alloys. It comes from shape and thickness.

One thing that IS different about the modern bodies is that the steel has galvanized coating that's far more corrosion resistant than mild steel.

[chevynut]

CN? Do you have some information to back up this statement?? " Today's cars are made with thinner, stronger steel. "

I don't think there's any doubt of today's sheet metal (used in cars) being THINNER, but 'stronger'?? That differs from my understanding. I think today's car bodies likely use 'bends and contours' to make the panel strong enough, in spite of it being thinner (and likely less strong)...' If today's steel used in auto bodies is stronger, I'd really like to see some detailed information so I can modify my opinion...

http://www.worldautosteel.org/steel-basics/automotive-steel-definitions/

http://www.worldautosteel.org/steelyourworld/steel-your-strength/ (http://www.worldautosteel.org/steel-basics/automotive-steel-definitions/)

[BamaNomad]

Those articles were not clear (to me) as to whether they were referencing properties of the 'raw steel' used to make the panels, OR strength characteristics achieved AFTER forming (and then various treatments of the formed part) ??

[chevynut]

Unless it's a critical structural car part, heat-treating after forming isn't usually done with steel parts. The properties are engineered into the sheet steel as it comes out of the mill. These modern steels are engineered to have specific properties, unlike the steels of the 50's which were lower quality and lower strength plain carbon steels. This allows different steels to be used for different purposes depending on the strength and formability requirements. You can take a low-strength mild steel and make a stiff structure by design. A high strength steel can be made thinner yet still have the same resistance to deformation as a mild steel due to it's higher yield strength.

http://www.worldautosteel.org/steel-basics/

This type of steel is used for exterior body panels:

http://www.worldautosteel.org/steel-basics/steel-types/dual-phase-dp-steels/

[chasracer]

I wish I had the links to it but a few years back I read a report that was talking about the higher strength of automotive bodies today. And while you would always have to question some of the results, the information basically said that if you (and I don't mean this disrespectfully) remove deaths from no seat belts, windshields, air bags, non-collapsible columns and etc that the body of the car today protects us much better than earlier vehicles that basically just folded up on us given the same type of impact crash.

A bad comparison is that of a high school friend driving a late '57 2 door that did not survive a residential intersection crash that hit him in the passenger door. That car was literally folded in half and he never had a chance. My daughter on the other hand had a similar accident where she turned in front of a SUV with a small sedan on a busy 4 lane highway with a speed limit of 55 mph. Car was crushed up to the console, she walked away. How much force did the metals in those cars absorb and dissipate? I will never know but she now has her own family and my grandson is just 18 months from his driver's permit.

[Rick_L]

Strength and stiffness are two different things.

You can move a noodle anywhere you want and it won't break. In a collision there are two basic things happening. One is whether the structure is stiff enough to keep from collapsing BEFORE it actually breaks. This is a function of shape and design rather than metal strength. The second is when the damage is severe enough to actually break, in which case the original stiffness is changed because not all the metal is still connected.

[LEE T]

Its impressive to see it in slow motion
https://www.youtube.com/watch?v=fPF4fBGNK0U

[BamaNomad]

Today's fenders and hoods are designed with small 'indentations or dimples' every foot or so in the reinforcing or structural backup panels behind, which are a 'weak point', so that when it is hit with sufficient force, those parts 'fold up' sorta like an accordion... basically bending the hood or fender in half or quarters, etc.. which takes out a lot of the energy of the collision before actually getting into the firewall and passenger compartment.

[LEE T]

I have always used original 1950s metal for replacement panels and patches, but I'm running low on 1/4 panels now so I bought a golden star reproduction quarter for a 2 dr HT, and a sherman USA made quarter. The golden star has crisper lines, a much better wheel opening, and of course it has the flanges. They are both the same thickness as the original, but both are softer than the original. I can actually use my hand as a dolly for the back side and hammer on the front to work out the dents.

The golden star is about 1/4" short length wise. Although the HT dip is better on the golden star, neither one is good.

I'm shrinking the panel in areas to make it look better and noticed something very strange. Some areas on the golden star don't seem to want to be shrunk with a stud gun, although the shrinking disc works good. Instead of the target area shrinking, it may be an ajacent area.

Like I said, they are the same thickness and certainly not HSS, but the metal is different than what was used in the 50s. I understand that a lot of cars have to use these reproductions, but its surprising to me to see repairable panels that some people cut off and replace with the reproductions available today.

I also bought a repro front fender for an unbelieveable price because it has a few dents, and it is also soft and very pliable.

I remember when the only thing available was a flat panel with a formed wheel opening, so these panals are way better than that.

[chevynut]

They are both the same thickness as the original, but both are softer than the original. I can actually use my hand as a dolly for the back side and hammer on the front to work out the dents.

I'm shrinking the panel in areas to make it look better and noticed something very strange. Some areas on the golden star don't seem to want to be shrunk with a stud gun, although the shrinking disc works good. Instead of the target area shrinking, it may be an ajacent area.

I also bought a repro front fender for an unbelieveable price because it has a few dents, and it is also soft and very pliable.

What do you think makes the older steel harder than the new steel? I have a hard time believing it's the composition so it must be work-hardening from cold rolling or forming. Cold rolled steel is hardened by the rolling process. Perhaps they're now annealing the steel after cold-rolling or using hot-rolled and pickled steel so it forms easier.

[Rick_L]

Lee T, it seems to me that a softer steel is a good thing for the replacement quarter, given that it's not formed correctly, and you have to work it to get it right. You want metal that responds to the work you put in it.

If you follow the metalshaping community, they seem to prefer "deep draw quality" steel, which I think has a bit more aluminum in it ("AK" is the designation I think). Maybe MP&C can comment on that, since he's an expert metalshaper.

Also, looking at Cnut's links on steel composition, looks like the high strength stuff has a narrower range between yield strength and ultimate strength. If you're going to hammer/dolly, form, shrink, etc., seems like you'd want a wide range between yield strength and ultimate strength. That means you can shape it without cracking it, etc.

[LEE T]

I don't have any idea what the difference is, but the old metal is way easier to work. This stuff doesn't seem to have much memory either.
I don't do much metal shaping because I almost only work on tri-5s, and I have a stash of cars and parts to pick from. I would much rather repair an original panel than try to make one---but to each his own.

[Rick_L]

Lee T, a question on perspective. When you say "repair an original panel" are you talking about fixing a dent or crease by working the metal that's there, or are you talking about repairing rust damage by welding in a rust free panel cut from an original car?

Most of us don't have access to original rust free panels for the places that typically need that sort of work.

[LEE T]

I understand that a lot of cars have to use these reproductions, but its surprising to me to see repairable panels that some people cut off and replace with the reproductions available today.

I remember when the only thing available was a flat panel with a formed wheel opening, so these panals are way better than that.




You can buy good sheet metal today for patches, so I don't understand why people are throwing away good repairable panels. BTW, I have used a section an old rust free rocker panel to repair the bottom of a quarter panel.




I don't do much metal shaping because I almost only work on tri-5s, and I have a stash of cars and parts to pick from. I would much rather repair an original panel than try to make one---but to each his own.

Some metal shapers say just make a new panel, so I was giving my reason for not getting into metal shaping myself. Looking back, the last remark could have been left out.

[MP&C]

Cold rolled, drawing quality, deep draw. The drawing quality steels should have less carbon in the composition, are softer to better accommodate the movement needed in stampings and even hand manipulated forming.