Dear All,
I have a bookcase that I am making for a client, the architect has specced 1 1/2″ shelves, so I glued two layers of 3/4″ birch ply together. I used yellow glue, a multitude of screws, and let them set up on a flat surface for a few days. Three out of four sets have warped badly. I have put weight on the warped pieces to no avail. I am not sure what to do now. It would seem silly to take the same approach, although, this is the first time that I have had this problem. I am considering using 1″ birch & 1/2″ birch, but I am open to suggestions.
Thank you for your opinions.
Best,
John
Replies
The flatness of plywood products has deteriorated immensely in the 40 years I have been using it. Your best bet is to make torsion boxes with thin plywood faces and veneered or solid wood exposed edges. This construction will yield light weight shelves also. The length of the shelves is important also: too long and they will bend or break. An expensive and heavy alternative is solid wood. Good luck!
Cadiddlehopper
Dear cadid,
I am not clear as to what a torsion box is, , but I suppose that I could look it up. I have considered going back to solid wood shelves a number of times as the labor involved in all these glue ups is a bit much. If I have a 12" jointer I would do it. I have even considered Grizzly's new combo just for this purpose. Perhaps this will push me over the top on that. Right now, on these longer shelves, it rip the plywood, glue it up, bondo the nail / screw holes, add the edging, bondo that joint and then sand the whole mess down. Not very fun. I have never jointed 12" wide maple, but I am willing to bet that there are a lot less steps, less labor, and no bondo! What's a Cadiddle?Best,John
I would have to look up the definition of cadiddle.Somewhere among my Taunton pubs is an article about torsion boxes written by Ian Kirby, as I remember. I looked in an index to no avail, but I would think that you could find it right here on this web site. I suppose that you noticed that more than one poster recommended the torsion box approach.Cadid
I am not clear as to what a torsion box is
John -
It's a rainy Sunday afternoon and I'm semi-bored so I decided that your statement was a great excuse to play with my new CAD program.
Here - for all the world to see - is a torsion box. I drew it with 1/2" ply top and bottom and 1/2" thick grid material. You can actually use whatever thicknesses you feel would work best for you.
Before anyone dumps on me too hard for the geeky looking wood, please accept my apology in advance. I'm getting pretty good with the drawing part, but my rendering still needs a little work - lol.
Dear Dave,
Well that sure clarifies things... WOW! Thanks! Is it safe to assume that the final connection between face & frame is glue & brads?Best,John
I just use glue. When I make them, I'll make a "stack" beginning with a piece of plywood, beads of glue for the grid, then the grid pieces, then more glue on them, then another piece of plywood. I sometimes use small pieces of tape to keep the grid pieces in place while I lay on the top sheet of plywood.
I'll make a framework to keep things from sliding around during clamping and I make the shelves a little oversized so I can trim to final dimensions after the whole thing has dried.
I used the same technique to make my workbench/TS outfeed table. No bottom piece of plywood, just a 2" x 6" grid assembled with screws and joist hangers and a piece of ACX ply screwed to the grid. My benches take a real beating so I scrape the top and apply a couple of coats of poly every year. I figure that this top will be up for replacement in 2010.
Don -
I tried that flipping over business.....................once! Maybe my technique needed work, but it turned into a serious train wreck. - lol
Edited 2/11/2007 8:46 pm by Dave45
Your drawing is more than adequate. Thanks for taking the time to make it.In its simpliest form, the middle frame is made of single pieces running the longest dimension and shorter pieces butt-jointed between them. Some articles advocated simply holding the middle frame pieces together with staples - no real fasteners or adhesive - just the minimum to keep the middle frame together during assembly. The closer the spacing of this framework the stronger the result. Spread adhesive on the upper surfaces of the middle frame and lay on the top plywood surface. Turn the assembly over, spread adhesive on the other side of the middle frame, lay on the other plywood surface and clamp the sandwich until the glue cures. Curved cauls are one way to clamp, I usually simply laid a number of sheets of additional 3/4" plywood on top overnight.An alternative middle would be resin coated homeycomb paper, available in small quantities at http://www.vacupress.com and perhaps some other sites. Prepare an outer frame of hardwood the same thickness or perhaps 1/32" less and an inch or so wide. Lay down the bottom plywood surface and apply adhesive. Place the outer frame and lay in the honeycomb paper. Apply adhesive to the underside of the other plywood surface, turn over and place on the sandwich. Clamp as above.Two articles on torsion boxes are:
“The Torsion Box” by Ian Kirby. Fine Woodworking #32 Jan/Feb 1982, pp 96-102; and
“Tackling Large Tabletops” by Kim Carleton Graves. Fine Woodworking #151 Sep/Oct 2001, pp 76-81
There is surprisingly little published material on torsion boxes, other than the above (to my knowledge).
There's also a very thorough chapter on torsion boxes in Ken Horner's "More Woodworkers' Essential Facts, Formulas, and Short-Cuts". Lots of tables, rules of thumb, and practical advice, plus useful formulas if you're not math-averse.
Horner's book is the most detailed info on torsion boxes available by far. The torsion box is probably one of the misunderstood things to construct. Even David Marks doesn't do it justice. He took Kirby's simple elegant approach and made it more complicated. Kirby used staples to hold the frame work together before gluing. I've seen more bad info on torsion box construction than good on the web.
I think your drawing looks sharp -- particularly the shading -- very realistic. Good job.
Michael
I would bet that your warping is caused by the moisture you introduce with the glue. It's "trapped" between your pieces of plywood and causes the plys to expand as it migrates thru the plywood. If you're using an "economy" grade of plywood (i.e. 5 plys), the problem can get worse.
Depending on the length of your shelves, I would either build torsion boxes, or use 3/4" ply with a 1-1/2" "band" rabbeted across the shelf front. It will look like a thick shelf without the added weight.
Either way, use a cabinet grade plywood (i.e. more plys). It's much more stable.
Dear Dave,
You said "I would bet that your warping is caused by the moisture you introduce with the glue. It's "trapped" between your pieces of plywood and causes the plys to expand as it migrates thru the plywood."You could be absolutely right, but wouldn't the expansion offset each other as each sheet tried to expand away from the other?I tend to agree that the grade of plywood is to blame. (It certainly can't be me!!). The width of these shelves (39") prevents using a single layer of 3/4 as it will deflect.Thanks,John
The expansion wouldn't necessarily be uniform throughout both pieces of plywood. Plywood cores aren't exactly top quality material and may (or may not) respond the same way to moisture, temperature,................etc. You also mentioned using a lot of screws and they may even contribute to the problem. Screws create their own localized stresses and who knows how they will affect the glued up "sandwich".
Depending on the load they're intended to carry, 39" shelves can be made from a single piece of 3/4" ply - if you use a good front edge to provide some strength. In a few situations, I've even used a shelf support pin in the back to help.
If you aren't comfortable with that idea, go with the torsion boxes. I would make them from 1/2" Baltic Birch ply with a grid of 1/2" birch between the pieces of ply. They're easy to make and they're seriously stout!! - lol
Hi,
I am considering a torsion box for my assembly/outfeed table. One of the questions that I have is;
Do you need a perfectly flat surface to assemble the torsion box on? Does the box just depend on how true the spars of wood are that make up the frame? Also, would you go with poplar or oak for the spars?Thanks,John
Dad -
Actually, that was three questions. - lol
"Perfectly flat" can be a little subjective, but a flat surface is better than laying it on the ground. Your workbench top should be fine - or you can lay some plywood on sawhorses.
The grid basically gives you the thickness you need without the weight of a solid piece of "filler". It's the glue that does the work. If your pieces are slightly warped, I would lay them out so the warps want to correct each other. If it's clamped flat, they should stay put when the clamps come off.
I would use poplar over oak just because of the cost difference. I suppose it could be argued to use the same species for the grid as the species of the plywood to minimize differential expansion, but it's the plywood core material that expands and contracts.
My workbench/outfeed table grid was made with doug fir 2" x 6" that I ripped to about 5" wide.
Edited 2/12/2007 11:05 am by Dave45
Build torsion boxes or else use MDF.
David Ring
http://www.touchwood.co.il/?id=1&lang=e
Would a non-water product react the same way?
How about epoxy type adhesive. Ron
I'm not gonna pretend that I understand all the factors involved here. But experience has taught me that gluing up ply sheets like this is asking for problems. Don't search for the right adhesive to do the wrong thing. If you want a shelf of 1½" thickness, you'll get both a lighter and stronger result using 2 sheets of ½" ply with ½" of air between them.Or as I said before, doubling MDF seems to have no ill effects.David Ring
http://www.touchwood.co.il/?id=1&lang=e
"Or as I said before, doubling MDF seems to have no ill effects."
Don't mean to be contrary BUT :)
1 1/2" of MDF will sag under it's own weight...
I would make a torsion box and face it with 1/4" baltic birch.
Cheers, DonDon Kondra - Furniture Designer/Maker
You're not contrary, in fact I agree about the MDF. I was just making the point that one can glue up multiples of MDF with no apparent warpage, as opposed to ply.David Ring
http://www.touchwood.co.il/?id=1&lang=e
We've glued up many 4x8 stacks of plywood in my shop with little or no warpage. In fact, we've not measured for warp because the glue-up simply lies flat. These 4x8's were subsequently cut up into stair treads..... still no warpage. Gluing up plywood in this manner yields a very stiff section that is, in my experience, stable in all aspects. We normally use Unibond 800 as the glue, but have had equal success with Weldwood (powder mixed with water), and PVA (Titebond).I'm guessing that the original poster here started with warped sheets of plywood. He may have very likely stacked them so that the warps intermeshed..... so that they reinforced each other rather than cancelling themselves out. Likely also is the supporting bench may not have been flat to start with. That could introduce a warpage similiar to intentionally laying up a curved panel over a form.I find it difficult to believe that the little bit of moisture contained in a PVA glue can overcome the inherrant strength gained by laminating two sheets of commercial plywood together. My own workbench is three sheets of 3/4" ply glued together in my vac press. It is as flat a surface as my layup table is (a torsion box, 4" thick), and that is very flat indeed.What has long been a mystery to me is why is some commercial plywood warped at all? Construction grade cdx is often like a potato chip ("crisps" to some). Even the more expensive hardwood veneered ply often has a significant "wow" to it. The laying up of alternating bands of ply is supposed to neutralize the stresses. Maybe the factories don't let the glue fully set before they leave the assembly line. Or maybe it's the tree spirits getting back at society for overharvesting the forests.
Hi sapwood ,
Some of the possible reasons why plywood is warped when we get it are as follows . If the unit or plywood gets stacked sloppy with say one sticker in the middle , most common in lumberyards the panels will warp . Also when the ply leaves the press at the mill it is warm and probably not 100% cured and vulnerable if not stacked flat .I have been shipped panels that are still warm several times .One of the other main reasons for the warp is if the core is not stable to begin with , or not dried properly or even of a species that goes to heck in a hand basket , like the ply we have seen recently from China .
dusty
My experince has been terrible with expensive plywood, it all warps. It is stacked flat outside when I buy it, and then, I bring it inside, and as soon as it is warm, warp. This is $90 ply. Talking to other users, and they have the same experince the same thing. We think it is poorly made ply.
Pedro
$90 Ply ?????
that sure is some expensive plywood, are we talking about ebony veneer here or what???
We are talking about birch 3/4' plywood, good two sides. Its just in the middle that it isn't any good! Its part of the joys of living in the north. They say its the freight!!
The clue may be in your thread. You say you "...let them set up on a flat surface for a few days." Does that mean you put them directly onto another surface? If so, air can not get equally to both sides and as the moisture content changes, the opposite surfaces will develope different moisture content. This will cause unequal expansion/contraction between the two surfaces and warp is the result.
Sometimes, you can salvage the panels by stickering them off a flat surface so air can get equally to both sides. You can try weighting them down on top but, again, be sure that air can freely get to both sides. Never put a panel flat down on a surface. Always either stand them directly on edge of sticker them up so air can ciculate.
On another point, the shop I was involved with probably made a couple of thousand laminted panels using plywood, particleboard and MDF. We always used solvent based contact cement to glue layer the panels. Water based adhesives take a lot longer to dry and can actual be prevented from drying propertly. In addition, it's impossible to get enough clamping pressure to get a good bond. Contact cement will bond immediately and can be worked within a couple of hours.
Dear Howie,
I had them stacked upon each other & stickered. Contact cement............Hmmmmmmmmm. I never thought of that! I like it. I'm not sure how that will work out with my wood stove heating the place, but hey.... what is the worst that could happen? If I blow the shop up, I'll never know about it, of course, with my luck, I'll probably end up only getting hurt. I'll have to think about that. I may be able to do that on site. That's a pretty good idea. Thank you very much!How do you apply the initial pressure? Walk on it, clamps? what would you suggest?Best, John
I just use a J-roller.Howie.........
Jmartinsky
I hope you said that tongue in cheek, or you intend to use a water-based contact cement.
When you open a gallon can of solvent based cement and start to work, the air starts to fill with the fumes of the volatile chemicals in the cement (like toluene). When the fumes hit the flame, the air ignites and you are surrounded by fire, the air burning around you. The fire follows the fumes to the gallon can and that starts to burn. By now, most of the fumes have burnt out, but the gallon is raging. If you are lucky, no clothing caught fire and no one was hurt yet, but you still need to put out the fire with the ABC type fire extinguisher that you probably do not have handy if you are prepared to use solvent based contact cement in the presence of an open flame or spark.
If you walk away, and still have a standing shop, you will have a black soot over everything in your shop, including the walls, ceiling and floor (and all your equipment). Work safely and intelligently. JL
Dear JL,
"When you open a gallon can of solvent based cement and start to work, the air starts to fill with the fumes of the volatile chemicals in the cement (like toluene). When the fumes hit the flame, the air ignites and you are surrounded by fire, the air burning around you. The fire follows the fumes to the gallon can and that starts to burn. By now, most of the fumes have burnt out, but the gallon is raging. If you are lucky, no clothing caught fire and no one was hurt yet, but you still need to put out the fire with the ABC type fire extinguisher that you probably do not have handy if you are prepared to use solvent based contact cement in the presence of an open flame or spark. "Yes, but I bet it will kill all of the spiders! I was kidding, of course, but thanks, you never know........John
THat is my guess.Pardon my spelling,
Mike
Make sure that your next project is beyond your skill and requires tools you don't have. You won't regret it.
Other posters have suggested that uneven water migration through the plywood may be causing the warpage.
If you have enough time, you might try laying your shelves concave side down on an unsealed concrete floor and weighting the middle to flatten them. The down side will pick up moisture gradually from the concrete and swell a little bit, tending to even out the warping. You can wipe the surface first with a damp rag to accelerate the process, but don't soak the surface.
This technique has worked well for me on cupped solid wood, even on oak. If you have time, it might be worth a shot.
A torsion box works on the principal of a box girder, giving a great deal of rigidity and strength with a lot less weight than solid material (like hollow core doors). You would use thinner birch plywood for top and bottom of your shelves, say 3/8", glue solid 1x2 poplar or alder around all 4 edges, then face the edges for appearance.
When laminating plywood I've had the best results using solvent based contact cement or plastic resin glue, Unibond 800 is one brand I've used. In the winter I use solvent based contact cement because my shop will get below 60 degrees when I air out the vapors from the Plastic resin glue or the contact cement. Plastic resin glue doesn't do well in lower temps.
I do carefully select the plywood sheets I'm using and I plan on some loss. For example I glued up 8 sheets of 1/4 inch cherry ply, I could get sheets of 1/2 inch that were good enough on both sides. I was making up frame and panels that would be seen from both sides. This gave me 4 panels when if all had been perfect I only needed 3 to get the door panels.
Again carefully select the plywood, carefully consider the glue ups so you don't match warps, and be sure to cover all surfaces with glue.
not sure what caused the problem, but you could remove one piece of plywood (i'd suggest taking it off with the jointer in a single 3/4" deep pass) and try again...
sorry, couldn't pass up an opportunity to rib you a little....
never had any problems with laminated layers of plywood remaining flat...are you positive that the "flat" surface you used is really flat? if it had any bow in it, your glued up panel would probably reflect that....
if you are trying to salvage what you already have, what about using 3/4" x 1-1/2" solid wood on the edges? is there enough flex in your shelf to straighten it out in that manner? if so make sure you put them crown up (almost like it is pretensioned to resist the shelf load)
it might be best to just start over, you already have several good suggestions on alternate ways to construct the shelf...
Dear C,
If I thought that I could salvage it, your suggestion might be viable, although I would have to to reduce the current cut of my jointer to do it in one pass...........I think that I need to start over. I like the contact cement idea, Although I have to think about a way to apply the pressure. I may try 1" ply with 1/2" to build it up. Thanks for the thoughts.John (Evil)
I've seen people use bags of sand or water to apply evenly distributed pressure...
i've used cauls, but it is a pain in the ####...a vaccum bag is the way to go....
Dear C,
The vaccum bag is the way to go.... if I had one. the other method that I keep kicking around is the solid setup. That Grizzly 12" jointer looks attractive for the job as well. For the moment, I need to complete this, so experiments are out. Best,John
Any reason nobody suggested a solid glulam or timberstrand rim board? 1 1/4" Rimboard is common place in houses and you could laminate a 1/4" birch or resaw birch stock veneer for the surface. The rimboard is far straigther than that wonderful plywood. I purchased some lower grade birch that I found out comes from China. Boy was it fun to try to build square projects with that useless crap.
Did you happen to buy that plywood at HD? I noticed the other day a stack of birch ply - made in China - that was really waffly. If the cores are not dried evenly the panels will won't stay flat. You mentioned that this is the first time you have had this problem, so if your technique is the same its reasonable to suspect the materials. I've glued up lots of plywood and never had the problem you describe. Wondered about the fact that you screwed the panels together, perhaps setting up stresses while the glue dried. But if you have done this before with no ill results then we're back to the materials as the culprit.
This forum post is now archived. Commenting has been disabled