Best material for bent lamination

Other Versions
Searching for the most suitable wood product to use with bent lamination. July 22, 2003

I have been contracted to build a series of really wild desks. The desktops have a combination of curves with a radius anywhere between 24 and 48 inches. The drawer faces for the desk need to fallow the same curve as the desktop. I need to make a bent lamination for these faces that are 17" wide, and range from 12" to 9" tall. The finish spec is high density plastic laminate. What would be the best wood to use on this bending project? Any help on this subject would be great!

Forum Responses
(Furniture Making Forum)
If you are going to cover the wood with plastic laminate, I suggest using curve ply since it will be hidden anyway. You can always edge band it to hide the core. Unless your customer specifically asked for all wood, this is the route I would choose.

From the original questioner:
What do you mean by "curve ply"? Are you talking about compwood? I do intend on banding the edges, so the material I use does not need to be of finish quality. It just needs to be sturdy enough to hold laminate well.

Check with Outwater Plastics. They have a selection you could probably use. Pliable MDF, Flexply, ect.

From the original questioner:
Thank you. Found what I was looking for. Does anyone know if there is any spring back in the flex ply after removing the clamps? I have done bent laminations a few times before and always had a hard time getting the curve right. This product would make my shop so much more productive if there is no spring back.

From contributor R:
Flex ply will not work to hold a curve. It has to be attached to a frame built with the desired curve. 1/8 luan "mahogany" is excellent material for curves that stand alone without framework

A bending ply (as opposed to a normal ply used in a bent laminating situation) should have little to no spring back. The 48" radius should be no problem, as it is a fairly gentle curve. The 24" radius, however, will be the one that will want to spring back on you. I'd check out 1/8" or 1/16" (if you could find it) for this one. I believe you'll be able to go a little thicker on the larger radius curve. As a general rule of thumb, whatever thickness ply you're using should lay down easily over the form with light hand pressure. The harder it is for the laminations to conform to the jig, the greater the chance of spring back. To be safe, it's always optimal to do a test run, see what happens, and make adjustments if necessary.

Choosing the right adhesive will also help your chances of eliminating spring back. An adhesive that sets up really hard and stiff will better help the lay-up to hold shape than say a PVA (white or yellow) wood glue, which remains flexible longer through the curing process, and in their cured state will maintain a greater degree of flexibility... A good choice would be Unibond, epoxy (like West System), or urea formaldehyde glue. I really like Unibond in this situation best (or the West, depending on your clamping condition) because it is a two part adhesive with a base powder and chemical liquid hardener, which sets up through an exothermic chemical reaction (as does the epoxy).

With either of these, you are not adding water (moisture) to the work piece, as you might typically when using a urea formaldehyde glue which is basically a powder that you add water to. I have found it better to keep water out of the mix when laminating curves, especially if you are going into a vacuum bag, where the moisture has no place to go for the duration of the glue up, which depending on the adhesive and production schedule, could be anywhere from maybe 4-6 hours on the short side, to overnight or 24 hours on the long side. It might be better over a form, mechanically clamped, out in the air if you were using urea formaldehyde. The water would definitely have a better chance to evaporate sooner.

Exercise caution when mixing, too. The ratios of powder to hardener need to be right on to achieve optimal consistency - too thick and it's difficult to spread; too thin and it's a wet soupy mess which won't set up hard. Wear a respirator, too. All of these adhesives (save the PVA's) are nasty bad for you.

Hopefully the normal contact cement route will work for the laminate - test that, too. If the laminate pulls up at the edges, maybe there is some way to use one of the other adhesives I've described. Not positive on this - just throwing out possible alternatives. I've never used anything other than contact cement for laminate, and would choose regular over waterbased for this application.

From contributor B:
We very often will use either 1/8" bending poplar or 3/8" bending luan (wacky wood) for bent laminations where the grain direction is not an issue. Just last week we did a 5 1/4" radius x 6" deep half circle with 18" legs out of 1 layer of 3/8" bending luan with a single layer face of 1/8" bending poplar. The small jamb held its radius just fine when it came off the form. In fact, over the weekend with the heat on in the shop it actually tightened up a bit so that the straight legs were sprung in a bit.

From the original questioner:
Thank you for all the help! Last night I put three 6" wide 1/8 inch luan strips on a 24 inch diameter circle jig in a fashion such as contributor B described above, glued (Unibond) and clamped overnight. In the morning, like a child on Christmas, ran to the shop and ripped off the clamps. I found that my piece has sprung back about 2 inches on each end. I for some reason never thought about using luan as a lamination. The only drawback is the time-consuming effort to calculate the spring back. I have seen many formulas for spring back and I'm not interested in that route. I called the local wood distributors and could not find anything thinner than 1/8" luan, or the above mentioned wacky wood or bending poplar. So my search continues for the magic "no spring" bending plywood. I would like to work on more curved pieces in my shop, but can't afford to stay in business on the trial and error method of bending.

From contributor B:
Both bending plywood products are nearly universally available. Try another panel distributor. If you can't find one try your local lumber yard and see if they can get from their distributor for you.

From contributor R:
Are you guys sure? I tried laminating curved doors with rubberply "wackywood." All I got was a thicker panel of wackywood that relaxed to the point it almost straightened out. It was way too flexible to stay at the curve I wanted. I only use it to wrap tight radius on cabinetry.

From the original questioner:
What do you use now for bent laminations? 1/8" luan? How do you account for spring back? I did more experimenting and feel that six plys of 1/8" luan would be needed to achieve the desired stiffness needed to support plastic laminate. Is this contrary to anyone else's findings? Or is this the standard for free standing/no framework bent laminations finished with plastic laminate?

From contributor R:
6 layers is just right. I always cross cut my material across the 48" direction instead of the length of sheet. I make up my clamping template using a router and a 1.25" straight cutter. This makes my inner and outer clamping pattern. The extra 1/2" I fill with extra 1/8 ply as clamping culls. The trick is to use more culls on one side than the other to decrease the radius to allow for spring back. You can use a 1.75" bit if you need even more. Use a hard setting glue. I use yellow glue. Don't be in a hurry to pull it from the jig. The material contributor B uses must be different from what I've used because the rubberply I used will not work.

There's another little trick I was taught by a old timer - always make your doors first on curved cabinets and always install the cabinet with the doors on. The reasons are: you can always modify your cabinet slightly to match the doors - no one will know and it's a lot easier modifying the box than the curved door. And if you install the cabinet without the doors, you might not get the doors to work if they're even slightly out.

From the original questioner:
The first desk turned out wonderful. Almost finished, just need to find some half round T moulding (or T-cap, whatever it's called). Customer was very happy! The last tip that contributor R posted saved much time, money and aggravation. Even though the curves on the desk were slightly off from the spec drawings (due to spring back, even though I tried to compensate for it on the pattern) the customer would never know unless he brakes out a micrometer and protractor. I was going to create the top first, but it's a good thing I didn't - it wouldn't have worked out.

The comments below were added after this Forum discussion was archived as a Knowledge Base article (add your comment).

Comment from contributor A:

I was at a veneer layup shop not long ago and they were throwing away "cut offs" that were huge stacks of raw veneer, about 12" wide x 24" long. It seems to me that finding such a source in your town can't be too hard, and you might also find a glue you can spray rather than roll. A little wax paper and the right stream of raw materials and you should crank out true curves without buying very much.

Comment from contributor C:
I just finished a couch project with 16 curves of 25-50" radius. One had springback, the rest were fine. The difference was I changed glue. John Ersing (Veneer Systems) replied and recommended their glue and it held like a rock. The glue made all the difference.

Comment from contributor D:
What you need to use for bent wood lamination is 1/8" bending poplar plywood that comes in either long or short grain. These sheets come in 4' x 8'. If your radii is tight, you can use 1/16" Wisa Italian bending plywood. The more seams you have, the less chance for springback. Having an odd number of seams will help a more solid lamination and smaller springback. You should be able to order these materials from any decent lumberyard/hardwood store.

Comment from contributor L:
I really appriciate all the info that I found on this site. My project is a cabinet door with a 13 inch radius. After studying what I found here, I built a jig and took a 1/8 inch birch door skin and glued 7 layers of it together with TiteBond II, expecting to get a couple inches of springback. What I got was absolutely no change at all when I took it off the jig. Test completed, I am now trying the door on my jig revised to expect no springback.

Comment from contributor M:
Don't use yellow glue (Titebond) for bent lamination. Use plastic resin glue or epoxy. Yellow glue is slightly flexible and will creep over time, leading to spring back.