Which glue?

[Cecil Arnold]

I searched the threads before I posted, and while there may be an answer to my question I didn't find it. I am constructing some lightweight shelves and want to make a foam sandwitch structure using poplar frames with 1/8" BB skins and either some 1/2" or 3/4" foam siding cut to fit inside the frame. I need to know what glue to use to insure that the BB adhears to the foam, since almost any decent glue will work for the wood to wood joint. I think I read somewhere that TB or Elmers will work but wanted to ask the experts to be sure. Thanks.

[Jamie Buxton]

PVA (that is, Elmers or Titebond) doesn't bond well to plastic. Polyurethane (Gorilla glue) does stick to plastic, but is has the nasty habit of expanding as it cures, so it may push your laminates apart.

You might consider using a wood spacer grid instead of the foam. The advantage of the wood grid is that good ol' PVA will bond to it. You can make the grid from the same eighth-inch plywood you're using for skins. It can probably weigh almost as little as the foam. Rip strips as wide as you want the grid's thickness. Each strip will run the full width of the grid, or the full length. At each cross-point in the grid, cut slots halfway through each strip so they can fit past each other. The slots aren't joinery; they just are clearance. I cut them with a standard cabinet-saw blade. It makes a kerf which is .125" wide, so the "eighth-inch" plywood will slip right through. I gang-cut the slots. That is, I stack all the strips together and with one pass of the table saw make slots in all the strips. I strap the strips together with that saran-wrap strapping material, and cut right through it.

[Vaughn McMillan]

I've not used foam siding, but I'd think contact cement would work as long as it doesn't dissolve the foam.

- Vaughn

[Lee DeRaud]

Why do you need/want to bond the foam? It's not going to add any structural strength (IMHO).

(just trying to get in my quota of stupid questions for today...)

[Jamie Buxton]

Why do you need/want to bond the foam? It's not going to add any structural strength (IMHO).

(just trying to get in my quota of stupid questions for today...)


Lee, I think the bonded foam would add a lot of strength. What he's building is a torsion box. He's using foam for the spacer between the two skins. If he doesn't bond the spacer to the skins, the stiffness of the assembly is just the two thin skins. If he does bond the skins, the stiffness of the assembly is the full thickness of the assembly.

[Doug Shepard]

Cecil
Some other type may work fine, but they actually make adhesive for gluing foam. It's in a caulking gun tube normally near the regular construction adhesives in the stores. I've used it for gluing 1" styro insulation to AC plywood and it works fine. I'm not too sure but I think there are issues with some of the other types of construction adhesive disolving the foam.

[Lee DeRaud]

Lee, I think the bonded foam would add a lot of strength. What he's building is a torsion box. He's using foam for the spacer between the two skins. If he doesn't bond the spacer to the skins, the stiffness of the assembly is just the two thin skins. If he does bond the skins, the stiffness of the assembly is the full thickness of the assembly.I can see the foam adding strength in compression, but not in shear. It's going to help keep the skins flat, but if the outer "box" is going to be subjected to torsional loads that the skins can't handle, the foam will fail anyway.

If he really needs a torsion box, your wood spacer idea is a whole lot better idea: the "half-laps" where the spacers cross add a lot of shear strength.

[Jamie Buxton]

I can see the foam adding strength in compression, but not in shear. It's going to help keep the skins flat, but if the outer "box" is going to be subjected to torsional loads that the skins can't handle, the foam will fail anyway.

If he really needs a torsion box, your wood spacer idea is a whole lot better idea: the "half-laps" where the spacers cross add a lot of shear strength.

You can build a perfectly sturdy torsion box with quasi-half-laps in which the strips don't touch each other at all. The spacer elements don't need to provide shear strength. Mostly what they do is keep the skins the same distance apart as they are when the torsion box is unstressed. If the skin spacing is constant, any attempt at bending the box must stretch the skin on the outside of the bend, and/or compress the skin on the inside of the bend. That's where the strength of the assembly comes from: the skins.

[Lee DeRaud]

You can build a perfectly sturdy torsion box with quasi-half-laps in which the strips don't touch each other at all. The spacer elements don't need to provide shear strength. Mostly what they do is keep the skins the same distance apart as they are when the torsion box is unstressed. If the skin spacing is constant, any attempt at bending the box must stretch the skin on the outside of the bend, and/or compress the skin on the inside of the bend. That's where the strength of the assembly comes from: the skins.Agreed. Is the foam strong enough in tension to do that well?

[Charlie Plesums]

Agreed. Is the foam strong enough in tension to do that well?
The foam is never in tension. The skin on the outside of the bend is in tension, and will try to move towards the inside of the bend. The skin on the inside of the bend is in compression, and will try to move to the outside of the bend. The filler in the tension box, whether a wood grid, foam, or the corrugated cardboard in hollow core passage doors is in compression, working to keep the skins apart. By being apart, the leverage dramatically increases the bending stiffness of the torsion box.

Sort of like an I-beam. The top and bottom of the beam provide the strength, but the middle of the I beam only has to keep the top and bottom apart, thus can be much thinner than the overall width of the beam (top and bottom).

[Jamie Buxton]

Agreed. Is the foam strong enough in tension to do that well?

Good question, and I don't know the answer. My guess probably like yours --- if you load the foam-filled torsion box to the point that it crumples, it'll be the foam which is the failure point. However, my guess is also that if you're working well below the failure load, a foam-filled torsion box might not behave very differently from one filled with a wood rectangular grid (or a cardboard hex-grid for that matter).

[Lee DeRaud]

The foam is never in tension. The skin on the outside of the bend is in tension, and will try to move towards the inside of the bend. The skin on the inside of the bend is in compression, and will try to move to the outside of the bend. The filler in the tension box, whether a wood grid, foam, or the corrugated cardboard in hollow core passage doors is in compression, working to keep the skins apart.As Jamie says, the foam is there to keep the two skins parallel. If one of the skins is trying to bow out, the foam is in tension perpendicular to its (glued) interface with the skin. Conversely if one of the skins is trying to bow in, the foam will be in compression along that same axis.

But if the foam really is always in compression, you don't need any glue to hold it to the skin, which is how this discussion got started.

[Cecil Arnold]

Actually gluing the skins to the foam will add to the strenght. The idea of using foam came from some aircraft designs, one of which uses honeycoamb foam/cardboard with a fiberglass skin (military fighter) and another that is foam cored FG sandwitch (Windecker Eagle in case anyone has heard of it) as well as some of the designs of Burt Rutan the designer who did the Voyager round the world flight and just (with financing from Paul Allen of Microsoft fame) copped the prize as the first civilian to launch into space. Unless I can find an accptable glue solution--Doug, I'll look for the calk-- I will have to experiment with with epoxy and/or just build a reguliar torsion box as Jamie suggested. Thanks all for the input.