Welds and bolts Vs Gluing panels on, science project.

Well, we have talked about this a number of times, can "gluing" panels on a car possibly be as strong as welding it on. Well, my 10 year olds science project may have the answer!

He cut 12 pieces of 20 gauge steel. We then welded them together in pairs. We butt welded with the MIG, "plug welded" with the MIG, spot welded with a $20,000 ProSpot STRSW and drilled 1/4" holes and bolted two as well. We then bonded with Winzer Universal Panel bonding adhesive another piece of metal to each of the pairs of welded and bolted sets.

We put clamps on them and pulled them apart using the Chief frame rack at work (10 year old boys love destroying things, don't they?).

Here are the results, the glue won every single time! Pretty impressive I would say.

Brian

 

That's awsome, Winsor always had good products. Used to use alot of them back in the 80's.:beers2: Bill

 

We are going to do it again, this time with more welds and less bond area. He made the bond area a little bigger than it would be on your average pinch weld. And I want to put completely weld the butt weld this time.

Brian

 

Neat exercise! Do you have any way to measure the force? That would be an interesting way to see what the exact differences are and if you are approaching the design strength!

- Bill

 

to measure force he needs a Tinius Olsen machine like this one. :dollar:


however how he's doing it works great as a comparison test with the ability to see what bonding agent is stronger.

If i were still in school for engineering this would make a fun project for a few of my classes though.

 

Interesting. Well done!

As an aside, when my Dad was an Advanced Manufacturing Engineer for GM, they made a run of Fiero space frames out of aluminum and adhesively bonded together.

To the best of my knowledge it was the first frame (and the only one since) to pass the durability tests without incident and without any revisions.

:Smarty:

 

I wonder if it is because the glue spreads the load over a larger surface area (rather than the welds, which are quite concentrated by comparison).

K

 

I would have lost money in that bet:beer

 

Looks like there is a lot bigger glued area than welded area. I would say to make it fair you would have to overlap the steel like you did when you glued it and then weld both seams, not a butt weld. That would be more apples to apples.
No doubt the glue is great stuff though.

 

The glue will win out everytime, more area of coverage. Plus the tensile strength of the glue is actually higher than the metal... Most cargo trailer are glued together, Try sikaflex 555 amazing stuff, Oh be sure it is were you want it, you're not moving it afterwards, They do lots of low end body work this way now.. Good luck,
Scott

 

Real world considerations:

- How well does the glue survive vibration over 20 years?
- How well does the glue maintain its properties over varying temperatures (-20 to +120) and humidity?
- How well does the glue hold up to UV degradation?
- Can the glue resist road chemicals like tar and salt?

Always interesting to learn more!

-Bob C.

 

That's what I was thinking...contact area. To make the test more meaningful, make the glued contact area equal to the weld contact area.

 

"Low end"? How about a BMW 5 series factory recommendation to GLUE the frame rail using NO bolts or welds, ONLY adhesive!

No, it is FAR from "low end" it is becoming the norm. There are many cars like the Saturn that have a bonded on roof panel, floors like the new Caravan "stowage" area that is bonded in, Mercedes has bonded in trunk floors, and the list goes on an on.

Adhesives are here to stay.

No, that would not be "apples to apples" that would be "apples to oranges" being the bonding is just not used that way. That is like saying you can't plug in the welder because the adhesive applicator wasn't plugged in.

We do plan on doing another test with a "peel" as well as more welds and less bonding. But it is only for Mythbusting make it fail be damned.

The car you are driving is held together by thousands of spot welds, about an inch or so apart, so the lack of welds on the test pieces represents these welding practices that cars are made from.

First, from Fighter and Commerial planes to multi hundred foot tall tv and radio antennas are bonded together these days, I think they have proved themselves. UV for instance, it is painted on most every car so that is a moot point.

Second, what makes welds any better at all at these things? Welds are the FIRST place that fails when it comes to corrosion. First off, the metal is compromised right at the weld do to the heat. Secondly the area surrounding the weld is poorly protected from the elements because the paint can't get in there at the edge of the weld between the panels, even if dipped. Even on a nice California car that is basically "rust free" you will see rust froming at the spot welds! Split the panel off the car and you will see rust in between the panels where it was "pinch welded" together! I will pull apart one year old cars and see rust in between the panels that were welded together. You just DON'T see that on panels that are glued. There will be ZERO rust where the glue is!

Temps, that I don't know. But with a little research I'll be we would find out that it is tested from well under 0 to up into the hundreds, I'll just bet.

I am not saying it is going to or should replace welding. What I am saying is that this stuff is amazing and we shouldn't be so afraid of it.

Brian

 

You may as well just set the part up on edge then with the glue. thats all the contact patch you have with a weld. A failure on a good weld joint will not be at the weld but the material near the weld. Just like glues, different welds have different strengths in tension, compression, or shear.

btw. i'm also debating on gluing my rear quarters on when that time comes to re do it no warping and just having to put some body glaze over the new seam. and on the back side some seam sealer can do wonders and you can get to nearly all of it.

-nate

 

I can't argue with the surface area comments. However, even if a glued joint is stronger in this manner, it doesn't mean it's a stronger bond per square-inch. That would be apples to apples.
The point of the test is that the glue method of body panel repair appears stronger than welding.

On the other hand, welding body panels seems to be sufficient. Both are apparently good enough.
So the strength of a glue bond isn't as important as cost, ease of use, and possible rust resistance.

I'd be interested in whether the glue reacts in a corrosive manner and/or weakens over time (like the 30 and 40 years our cars have lasted) relative to a weld-repair.

Make duplicates of your last test pieces and have him save them for a high school or college science project... to see how they all hold up over time. I'd not expect a huge difference by high school - but it would look really good for a grade to see he had a follow-up study planned.
If you can test the force, even if the results are the same in the future, perhaps the force to tear them apart might change.

 

about 1/4" of overlap is common practice for gluing body repairs. It is not common practice to lap the parts over one another and then weld both sides. that would be a heck of a lot of work.

with the glue ya need one of these. and little warping when done.

 

I know that you wouldn't do it that way. I just brought that up because of the large amount of overlap and glue.
No way a butt weld is going to compare to that.

 

The comparison is not actually welding vs. gluing.

The useful experiment actually compares the welding *process* to the gluing *process* (which is a process that uses greater surface area).

If the purpose is to determine which would be better for manufacturing, you need to do the testing like the processes would be in the real world. Everyone knows welding is stronger than gluing when you measure tensile strength in PSI.

-Bob C.

 

Very good point Bob! Yeah, I didn't think of it that way, but that is exactly what we are talking here.

On the overlap, he overlaped these about an inch. In your typical pinchweld on a quarter panel replacement or something it is closer to 3/4" or less. I agree, he went a little over, that is why we plan on doing it again with more weld and less adhesive along with a "peel" test!

Brian