Built Up Curved Plywood Beam 1

[Stenbrook]

So I have an existing LG Stud wall that is curved around a stair case. The architect wants to cut and remove the lower portion of the wall to open up the second floor. What we have come up with is to attach several plies of plywood to the existing curved stud wall just above where they are going to cut the wall to act as a built up beam and support the beam on a post every so often.

So essentially we have 8 plies of 7/16" thick plywood curved along the stud wall and supported by light gauge posts every 15ish feet. What I need to know, is what kind of adhesive should be specified to create this built up plywood beam. Or maybe wood glue would be better for this application over adhesive? I'm not too terribly worried about the shear strength of the adhesive as the load that is being carried by the beam is only 60 plf. I am more worried about the tension forces being applied on the adhesive at the ends of the beam due to the curvature of the wall and the fact that the plywood is going to want to straighten back out.

Has anyone ever encountered something like this and can give me some help here?

 

[Triangled]

"Has anyone ever encountered something like this ...?" Can't say that I have, and having some trouble visualizing the issue to see where tension occurs.

"and the fact that the plywood is going to want to straighten back out....." seems to me a shear inducing phenomena, perhaps like the multiple laminations in a curved glulam beam for which adhesives have long been used...phenal resorcinal are the words that spring to mind, but they have changed over the years, but if that seems like a solution you're interested in, check with AITC or modern glulam literature as it should be common.

 

[Stenbrook]

This might be a classic example of me overthinking something. I know that shear flow will play into, but for some reason I was still thinking there would be some tension at the ends.

 

[jdgengineer]

I personally would be hesitant to try and do this built-up beam out of plywood that is curved. I'm sure it would work, but seems like you are on your own off the reservation. I'd use a curve steel beam in a heartbeat over the plywood beam.

 

[MotorCity]

I doubt you would have 100% contact between all of the plies as you try to bend plywood that is almost 1/2" thick (especially just using adhesive)

 

[Stenbrook]

We are going to be using wood screws in addition to the adhesive.

 

[Tmoose]

Pictures would sure help.

In the meantime -

http://www.schmeling.com/reference/pdf_files/plywood_bending_radius.pdf

 

[SwinnyGG]

Yes, you'll have some tension in the adhesive due to springback forces in the plywood. If you're already using screws, just put some in the ends.

With that said, still sounds a bit sketchy to me. Steel would be better in almost every way except for cost, and I bet you'd make up the cost difference through less labor- bending, gluing, screwing, clamping all that plywood is going to cost you or your contractor a lot of time.

 

[Canuck65]

I have fabricated beams using multiple plies of plywood before. I do take a somewhat conservative approach and as a result have created some beams with phenomenal capacities. I have done curved stair stringers before and the end result was very stiff and client was quite satisfied. All the plies were glued and screwed together.

You should search and download the following documents from the APA website "Design and Fabrication of All-Plywood Beams" and "Design and Fabrication of Glued Plywood-Lumber Beams" plus perhaps the "Plywood Design Specification" for some background information. Admittedly, these do not directly address curved beams, however, some careful thought and analysis should be sufficient to justify their performance.

 

[Brad805]

Your radius must be quite large to use 7/16" ply? I have done bent lamination's in my vac press, but the radii usually require 1/8" laminations. Field adhesive options will be limited. I would look at West Epoxy 610 or PL Premium. Both are structural adhesives and develop a reasonable bond without ridiculous high clamping pressure. I prefer West, but it is more difficult to find and tubes are about triple the cost of PL Premium. A sketch would be helpful.

 

[Stenbrook]

My radius is a little over 15 ft. We are talking about holding up a few feet of wall. The client wants a quick and easy solution that doesn't involve steel. To use a steel beam, there would be fabrication time in addition to the cost of the steel which would be unnecessary when the contractor already has the plywood on site and can go out and get an adhesive at Home Depot. If designed properly, I don't believe that this would be a "sketchy" design at all. If this is considered sketchy, then maybe all glulam beams are as well.

I was looking at the PL Premium adhesive, but I wasn't sure if there was a specification out there that set some requirements for adhesives in built up beams. Hence, why I started this post. I tried uploading a picture, but for some reason it froze up on me and won't do it, but its really not complicated. They are going to take a sheet of plywood and curve it to match the existing circular wall and attach it with screws to the metal studs. At that point they can adhere the next sheet to the first sheet with the adhesive and then screw them together, and so on and so on until they reach the required number of plies.

 

[Canuck65]

The adhesive for field glued floors in Canada needs to conform to the CAN-CGSB-71.26-M88 standard.

From the APA, adhesives conforming to State of California Specification AC 25-4 are suggested for laminating plywood webs, flanges, splice plates and stiffeners. They also reference ASTM Specifications D4689 or D2559.

The design guide also steers one towards combining staples and glue.

 

[ornerynorsk]

It might just be that I'm a bit old fashioned, but I cringe when I think about adhesives out of a caulking cartridge for structural applications.

"The client wants a quick and easy solution". Does the client want a sound, long-term solution, or are quick and easy the only requirements?

Resorcinal glue is worth a serious look.

It is better to have enough ideas for some of them to be wrong, than to be always right by having no ideas at all.

 

[Stenbrook]

Would you also be opposed to FRP strengthening of structures that uses epoxy to bond FRP plates to the existing structure? There have been many many many tests that show that it is a sound solution. How is this application any different? I think with anything, the soundness of the solution will be determined by the care of the contractor during installation. You could completely over design the beam, but if the contractor is lazy or half asses it, then there is nothing you can do as an engineer.

 

[structSU10]

how is the wall attached at the top? is there any way to hang it from whatever is above it? is it just a partition wall or is it structural?

maybe you need to restrain the new bottom of the wall with some kickers, but if it can be hung you could just add a bottom plate and some kickers and be done.

 

[Stenbrook]

We originally looked into hanging the structure from above, but the current wall does not go all the way up to the roof deck. The ceiling is also largely inaccessible. Installing the plywood beam would be easier than splicing studs on and hanging from the roof and installing the necessary bracing that would be needed as well.

 

[mike20793]

Sounds like you need a custom glulam to me. That would be no problem for them to manufacture and back up with sound structural design values.

 

[Brad805]

I doubt you will find an appropriate spec that will fit this case. Those codes will all indicate installation methods that will be difficult to mimic in the field. This will come down to your engineering judgement.

Resorcinol might not be a great adhesive as it requires quite a large clamp pressure to achieve the bond strengths tabulated. Clamping pressures in the realm of 100 - 200psi is not an uncommon spec for many of those adhesives. This is the glue of choice for glulam companies so you can google glulam clamping to see what they use for clamps. It would be very difficult to achieve that kind of pressure in the field and using this in a vertical application would be complicated due to its viscosity. It is also quite caustic to work with. I have used this as well, and I have reactions using it carefully with a mask. The danger label on the container made me pause. The west epoxy suggested (

http://www.westsystem.com/ss/new-six10-epoxy-adhesive/)

will develop the same bond stress as resorcinol and only needs about 10psi for clamping. It is easy to find at boat repair shops as they use it commonly to re-construct structural components that have rotted out. In my case a tube was $22 vs $8 for PL Premium. The tubes also have less volume due to the mixing nozzle. It has a long open time (45-60min) and if you are concerned about the application you can specify to use a V notch trowel. A 1/8" trowel would likely work well in this case.

If you start spec'ing FRP you are heading down a path that I doubt will be much of a savings over a curved glulam or steel and it will be a colossal nuisance for you. I suppose you could specify a licensed contractor for that work, but I suspect your client will get annoyed at the cost for what I suspect many will see as a minor issue.

If your wall is somewhere around 6" thick and your radius is 15'-0" and we assume a span of something like 4', it looks to me (see pic) like we are talking about 1.5" of curve. I am reading the sequence is to install the curved elements first, and cut out the opening after. If they were prepared to install some temporary bracing I would think you have room for a straight beam with an enlarged top flange in that space.

Given your light loads I suggest a design that relies upon screws as the primary mechanism with the adhesive being a safety factor. This does not seem very complicated and using your judgement I am sure you will devise a reasonable solution. I would still lean towards PL Premium.

 

[capeengineerma]

I think your plywood built up beam will work fine for the application here seeing that the loads are minimal. PL or the newer poly glues should be fine, especially when screwed as well. The only issue could be the thickness of the plywood panels. Might be less bonding stress if the panels were thinner, like 1/4" or less.