Roof beams supports not attached to roof

I've got the ceiling panels out of my van to install a backup camera and a dash camera. I was about to start re-assembling things when I noticed that a couple of the roof support beams running from left to right across the van are not attached to the roof. I looks like the normal assembly process is with a black adhesive. I also have a different colored adhesive - looks like honey - in a few spots . That adhesive has also failed.

The gap between the beam and the roof is around 1/4 to 1/2 inch (around 6 to 12 mm). It's only the last two beams that have come apart (not the one right above the rear doors, but the next to forward from there. All of the others are securely attached.

My gut reaction is to leave things alone. I can't push the beam up to meet the roof, and the roof really wants to be where it is. It's a passenger van, manufactured in March 2013. It could have been rather cool in Germany that time of year, and it's now summer with warmer temperatures - and the van is painted black. I'm guessing it could be something as simple as the thermal expansion of the metals causing them to shift around a bit more than the adhesives could handle.

A couple of photos are below. Any comments are welcome.

Re: Roof beams not attached to roof

I've seen that a lot when I've had headliners out on the sprinters. If you wanted to fix it the best bet would be to clean all of that out and use SEM panel bond I assure you it'll never come apart again as long as the prep is good. As far as getting the braces to the roof you could use some floor jacks and 4x4's to push it up, could even make a little table to bring the floor jacks up off the ground some to get them closer to the roof, we do it all the time when bracing a vehicle for a subwoofer install where we know the roof will be a weak point, usually go to the junkyard and salvage some more roof braces out of the same model vehicle. As far as actual issues I've never seen it cause any. I know on newer GM SUVs they were having noise complaints when the roof braces separated, sounded like a buffeting/throbbing noise from the long roof moving up and down rapidly.

Re: Roof beams not attached to roof

98Firebird said:

I've seen that a lot when I've had headliners out on the sprinters. If you wanted to fix it the best bet would be to clean all of that out and use SEM panel bond I assure you it'll never come apart again as long as the prep is good. As far as getting the braces to the roof you could use some floor jacks and 4x4's to push it up, could even make a little table to bring the floor jacks up off the ground some to get them closer to the roof, we do it all the time when bracing a vehicle for a subwoofer install where we know the roof will be a weak point, usually go to the junkyard and salvage some more roof braces out of the same model vehicle. As far as actual issues I've never seen it cause any. I know on newer GM SUVs they were having noise complaints when the roof braces separated, sounded like a buffeting/throbbing noise from the long roof moving up and down rapidly.

Click to expand...

No.

Do not 'lift' the braces. The braces are designed to be stiff and not bent under normal applications.

Place weight on the roof panel. The braces are 'beams', the roof panel is a 'skin'.



.

Re: Roof beams not attached to roof

OrioN said:

No.

Do not 'lift' the braces. The braces are designed to be stiff and not bent under normal applications.

Place weight on the roof panel. The braces are 'beams', the roof panel is a 'skin'.



.

Click to expand...

All I can say is I've never had any issues with that method. It doesn't actually permanently bend the roof brace if done correctly, it will pull back down with the roof skin attached after releasing the jacks, bringing the skin to the brace is a good idea though if he can get enough weight up there and also not damage the paint in the process.

Re: Roof beams not attached to roof

98Firebird said:

All I can say is I've never had any issues with that method. It doesn't actually permanently bend the roof brace if done correctly, it will pull back down with the roof skin attached after releasing the jacks, bringing the skin to the brace is a good idea though if he can get enough weight up there and also not damage the paint in the process.

Click to expand...

Two bags of sand and some blankets....

Re: Roof beams not attached to roof

OrioN said:

Two bags of sand and some blankets....

Click to expand...

Sounds like a good plan

Re: Roof beams not attached to roof

An enquiring feeble mind wants to know how long you have owned the van before becoming aware of what is obviously a subpar design and manufacturing process? And, during this period, did you notice any related bothersome issues?

I hear Vic's fatherly voice sayin, "If it ain't broke, don't fix it."

Meanwhile, slap them headliners back in and forget about it! Then go drink a beer.

Re: Roof beams not attached to roof

Thanks for the discussion. Very thought provoking.

Upon further inspection, it looks like the worst offending brace is slightly twisted. The gap on one side (the one I photographed) is considerably larger than the gap on the other side (which is almost non-existent). A twisted brace implies that a really proper fix would be to cut the brace out and install a new one. And that is getting much too far afield of my current skills and time constraints.

I think I'm going to install a shim with adhesive on both sides to maintain the current gap and reduce the chances of annoying rattles. If the roof skin shrinks slightly in cooler weather, it will just hold the shim a bit tighter.

Re: Roof beams not attached to roof

A prudent course of action. You could use a bit of expanding foam used for window installations. Dunno, just a suggestion.

Re: Roof beams not attached to roof

Bobnoxious said:

An enquiring feeble mind wants to know how long you have owned the van before becoming aware of what is obviously a subpar design and manufacturing process? And, during this period, did you notice any related bothersome issues?

Click to expand...

I have owned the van for about 20 months. I first became aware of this issue about 6 hours ago. The headliner has been out for two weeks, but I never really took a good look at any areas outside my current work area until now. The honey-colored adhesive caught my eye this morning, and I started looking closer.

I have not noted any issues, and would not expect to notice any unless I found myself in an unplanned test of the van's structural integrity. I would not put it down to bad design, as the design looks fine. If anything is sub-par, it would be the assembly process.

As an aside, it looks like the roof and braces are primed before installation. The adhesive between the two is clearly installed after the primer, but before the final coat of paint. The alternative assembly process would be to [spot] weld the two parts together. But that would need to be done with the steel bare. Then it would be hard to properly coat the welded parts with primer and paint, leading to a source of rust. So I have no complaints about the design. When you think about it, is amazing how rust-resistant most modern cars really are.

Re: Roof beams not attached to roof

ptheland said:

I have owned the van for about 20 months. I first became aware of this issue about 6 hours ago. The headliner has been out for two weeks, but I never really took a good look at any areas outside my current work area until now. The honey-colored adhesive caught my eye this morning, and I started looking closer.

I have not noted any issues, and would not expect to notice any unless I found myself in an unplanned test of the van's structural integrity. I would not put it down to bad design, as the design looks fine. If anything is sub-par, it would be the assembly process.

As an aside, it looks like the roof and braces are primed before installation. The adhesive between the two is clearly installed after the primer, but before the final coat of paint. The alternative assembly process would be to [spot] weld the two parts together. But that would need to be done with the steel bare. Then it would be hard to properly coat the welded parts with primer and paint, leading to a source of rust. So I have no complaints about the design. When you think about it, is amazing how rust-resistant most modern cars really are.

Click to expand...

The beam will twist readily along the length when it is not attached to the roof panel. What you are seeing as a twist in it is from either one side now not having enough adhesion as compared to the other side or the warpage of the roof panel is greater on one side.

Before you cutout the beam, clean all of the adhesive from one end to next and re-glue.

Re: Roof beams not attached to roof

OrioN said:

The beam will twist readily along the length when it is not attached to the roof panel. What you are seeing as a twist in it is from either one side now not having enough adhesion as compared to the other side or the warpage of the roof panel is greater on one side.

Click to expand...

Both sides of the beam are completely detached from the roof skin. So there is no adhesion anywhere along the length of the beam. But your thought made me look at things a bit differently.

I got a ladder out and took a look at things from the outside. The biggest gaps are on the beam that coincides with the seam in the roof skin. (I also noted a small rust bubble, which I will definitely want to address!) Looking at things closer, the skin under the air conditioner is a bit flatter than the skin aft of the AC unit. From the inside, that makes it look like the beam is twisted. Now I suspect the beam was installed too low. It should be supporting the roof skin, but its too low to accomplish that.

I still think I'll put a couple of shims in to provide some support. Perhaps I'll use enough to push the roof skin up ever so slightly on the forward side of the beam, and then let the shims just touch the skin on the aft side.

Before you cutout the beam ...

Click to expand...

I've decided cutting out the beam is crazy. That ain't happening.

I'm going to stick to finding a way for the skin and beam to stick together to provide the structural support that was originally intended.

Re: Roof beams not attached to roof

ptheland said:

Both sides of the beam are completely detached from the roof skin. So there is no adhesion anywhere along the length of the beam. But your thought made me look at things a bit differently.

I got a ladder out and took a look at things from the outside. The biggest gaps are on the beam that coincides with the seam in the roof skin. (I also noted a small rust bubble, which I will definitely want to address!) Looking at things closer, the skin under the air conditioner is a bit flatter than the skin aft of the AC unit. From the inside, that makes it look like the beam is twisted. Now I suspect the beam was installed too low. It should be supporting the roof skin, but its too low to accomplish that.

I still think I'll put a couple of shims in to provide some support. Perhaps I'll use enough to push the roof skin up ever so slightly on the forward side of the beam, and then let the shims just touch the skin on the aft side.



I've decided cutting out the beam is crazy. That ain't happening.

I'm going to stick to finding a way for the skin and beam to stick together to provide the structural support that was originally intended.

Click to expand...

Only use the shims while the adhesive is setting/curing. Once cured, remove the shims and replace with more adhesive. The adhesive you need to use is one that flexes as thermal and torsional forces will always be happening and you don't want to have the adhesive crack or separate again.

Re: Roof beams not attached to roof

So I ended up putting in some shims so that the roof skin was supported by the roof brace. Here's the quick and dirty method.

Front of the van is to the left. That gap was smaller, so I pushed the shims in a bit further. Then on the aft side, I pressed the shim in just until it filled the gap. That should raise the skin slightly on the forward side of the brace, which would bring it closer to in line with the skin rear of the brace.

I ended up skipping any adhesive. There was plenty of the original adhesive left on both the roof and the brace. That should hold the shims in place.

Rock wool insulation over that, then the headliner. So it's all tidy and hidden.

Re: Roof beams not attached to roof

ptheland said:

...

I ended up skipping any adhesive. There was plenty of the original adhesive left on both the roof and the brace. That should hold the shims in place.

Rock wool insulation over that, then the headliner. So it's all tidy and hidden.

Click to expand...

That looks good to me.

Personally I would have used a dense adhesive between the shims. From there the shims could either be left in place, or removed with additional adhesive filling the void. Similar to what OrioN suggested.

The reason. If there is reason to be up on the roof it may be good to have a more continuous support for the roof skin.

The shims will work for general support and reduce flexing so I wouldn't disassemble to change anything.

vic

FWIW, I concur with OnioN [sic] and AP.

In support of my decision, I’ll point out that I’m trading support along the entire edge of the roof strut for support over the surface area of the whole shim. I don’t know if that’s better or worse support than the original design.

Fortunately, I have little need to be on my roof. So I’m not going to sweat it. I do agree that a more continuous support is a good goal. If I had more shims on hand, I would have used them.

I noticed the same thing on my van (2015 short tall). I fixed it my applying a bit of force to separate the skin from the beam and squirting poly urethane construction adhesive into the gap. It noticeably reduced road noise after I fixed it. I think the skin was vibrating at speed and now it doesn't.

Frito (an 04) suffered the same crap design/manufacture issue at several locations. I noticed it when I was insulating the roof. Nearly a half inch gap at one location. I stuffed the gaps with high density polyethylene foam and forgot about it.

Until now.

Just a little follow up. After a bunch of other work, I washed the van yesterday. While on the ladder to wash the roof, I could see that the curve in the roof skin was more consistent across the roof support that I shimmed up. And a light press on the roof showed it to be well supported. The roof skin did not move.

Unfortunately, the one I chose NOT to shim now has a noticeable difference in the curvature. It also moves just from the weight of my long handled washing brush. I'm going to pretend I didn't notice it. However, if I ever need to get into that area again, I'll add some shims at that roof support as well.