5

I'm building some 7' by 3' bookcases with mix of plywood and solid oak for the sides and back (mostly plywood), but full solid oak shelving for strength. I'm worried that the ~9" depth to the shelves will allow up to about 1/8" of movement. If I used the online formula correctly. Since the sides and shelves are constructed so differently, and will thus move differently, I'm assuming I can't just secure the fixed shelves to the sides and back.

My primary question is: am I right about the movement, and if so, is it sufficient to place the fixed shelves in a dado and only glue the front couple inches? (I want the front of the fixed shelves to look solidly attached throughout the year.) Secondarily, can I "hide" the movement at the back, for instance by running a 1/4" dado and only insert the first 1/8" inch during construction (it's winter, so it's probably near my low humidity point)? I didn't plan quite far enough ahead to select my boards for grain direction, so I'm assuming the worst.

Here's a rough look at my design, for reference; green is plywood, and the rest is solid. The topmost and bottom visible shelf, as well as one not visible near the very bottom are the fixed shelves I'm asking about attaching. (The rest of the shelves will be adjustable, so I just need to allow room for their largest size, and accept some gap.)

Bookcase Design

Michael Urman
  • 293
  • 2
  • 8
  • Apologies Michael, as I was just saying to another member in the Comments below I've only just discovered that your plywood isn't what I was thinking of and writing about — all-hardwood multi-ply, which apparently has become as rare as magic dust (unicorn variety) in the US now! – Graphus Jan 15 '17 at 08:23
  • @Graphus Yeah, we get hardwood faced plywood on a variety of cores at the big box stores here, but even the all-veneer ones are typically softwood on the inside. It might be better at a proper lumberyard, but since their web sites fail to call out full hardwood veneers, I assume the worst. (I'm not up to hunting unicorn dust.) – Michael Urman Jan 15 '17 at 14:37
  • Yes best to assume the worst from what I've been reading! It's widely recognised that the quality of ply has dropped but I didn't realise how far it had gone. I wish you could see some of the all-oak, all-beech etc. plywood that's still available over here. It's not cheap but OMG do you get what you pay for. – Graphus Jan 16 '17 at 07:59

3 Answers3

3

My primary question is: am I right about the movement, and if so, is it sufficient to place the fixed shelves in a dado and only glue the front couple inches?

Yes and yes. You could also fix the shelves at the front by toenailing them in from beneath or using pocket screws (could also be used in addition to glue, although there's no benefit to doing both).

Secondarily, can I "hide" the movement at the back, for instance by running a 1/4" dado and only insert the first 1/8" inch during construction (it's winter, so it's probably near my low humidity point)? I didn't plan quite far enough ahead to select my boards for grain direction, so I'm assuming the worst.

Yes this is a good solution, much like the allowances allowed for in some things made from tongue-and-groove. Be generous in your spacing

But, you made an error at the outset in assuming that the solid-wood shelves would be stronger than plywood. Broadly speaking quality plywood will be comparable to most solid woods, of course without any attendant worries about seasonal movement.

If you hadn't already bought your materials plywood would presumably be significantly cheaper as well, so normally this would all be upside. The oak can be put aside for use in a future project of course.

One other reason for going with plywood in your case is I'm not sure there is sufficient allowance for movement. Doing a few quick sums with some guesses on my part I'm getting figures between roughly 1/5" and as high as 1/4". And if your seasonal humidity swings are wider than my assumption the internal moisture change in the wood will be greater than the 8% I've allowed for, obviously meaning that movement could be slightly more than this.

Graphus
  • 65,691
  • 3
  • 50
  • 139
  • I second using plywood for the shelves. Also would normally see the face frame overlay those fixed shelf, which additionally adds strength and covers the edges in place of edgebanding. If you look at older shelves with hardwood shelves, most of them will be split and warped due the to this movement issue over time. – Jacob Edmond Jan 12 '17 at 12:24
  • Thanks! Just the kind additional feedback I hoped for! However I can't substantiate your claim of plywood's strength here. A search for "strength of 3 foot shelves in plywood, solid wood" seems to contradict. In particular Stronger Shelves rates 3/4" plywood (without a lip) at about 41% weight capacity, and/or 75-83% maximum length compared to 3/4" solid lumber. Maybe the complete solid glue-up helps the fixed shelves (the front and back acting like lips), but I didn't plan that for the adjustable ones. – Michael Urman Jan 12 '17 at 13:12
  • @Graphus I did a calculation based upon 3% humidity variance and got less than 1/8". Do you think that assumption is too low? – Ashlar Jan 12 '17 at 13:45
  • @MichaelUrman I started from my knowledge and firsthand experience of plywood v. solid wood, then to check figures I used the data built into the Sagulator. Unfortunately the Sagulator doesn't cover quality plywood, but the fir plywood it does list, sans lipping or edging, compares almost directly with solid oak. So figure you'll easily do better with better ply, then add a lipping and it'll be stronger still. Even if the amount of sag for ply is in the same ballpark it means you don't have to worry about movement, which by itself could be a good enough reason to go with it. – Graphus Jan 12 '17 at 15:45
  • @Ashlar I think it's a bit hopeful, it may not be wrong though. My assumption could be off too. Change in moisture content completely depends on where the OP lives and other factors local to him. But the typical change in EMC used for calculating expected movement is generally 8 or 9%. – Graphus Jan 12 '17 at 15:51
  • Am I misusing the Sagulator? I'm seeing about 150% the deflection in "Plywood (fir)" than in "Oak, Northern Red" on a 34" x 9 1/4" x 3/4" shelf with loads from 30 to 50 lbs/ft. The difference seems irrelevant if the shelf is fixed, and is less pronounced if a 3/2 x 3/4 edging strip is added, but solid always deflects noticeably less. (Per my movement question, though, it may indeed be worth using plywood for the fixed shelves.) – Michael Urman Jan 12 '17 at 20:16
  • @MIchael Urman - you are not misusing Saulator with regard to comparison of solid oak vs plywood. When I read Graphus assertion in his text, I immediately checked what Sagulator would say about various kinds of oak vs pine plywood and found that sag was consistently about 50% more with plywood (There is no choice of veneered plywood, but the thinness of the oak face makes little difference.) – Ast Pace Jan 13 '17 at 05:51
  • Continued - With shelving, the issue is almost always sagging as opposed breaking the shelf. When Sagulator gives a reading of "excessive", it first means too much sag to be aesthetically acceptable, but it never reveals a point that the shelf is likely to break, so it's always best to design the shelving so that it fits within the "borderline" or "acceptable" results. – Ast Pace Jan 13 '17 at 05:52
  • @MichaelUrman No you're not, we just started with different figures. I had to make some guesses on thickness and material which affect the numbers. Using the assumptions I made fir plywood compared almost directly with the oak (again, I guessed at type). Just to be clear, the ply's deflection was higher, but still well within tolerable limits and remember fir ply is softwood ply, which you wouldn't normally want to use for anything where strength is a requirement but it's the only ply they list! Hardwood plywood would be, ah, slightly stronger :-) – Graphus Jan 13 '17 at 09:06
  • @AstPace With the new figures I'm getting deflection for fir ply of 0.01". A hundredth of an inch is more than merely 'acceptable', I think you'd need to check with a straightedge to see it. So it's irrelevant whether the oak's deflection would be superior (same ballpark, 150% worse or not!). And my further point was that you'd use something better than this, naturally some to a lot stiffer. Even if fir ply were what was used it would naturally need the front edge covered, giving the chance to install a lipping which would reduce deflection even further. So why not get the best of both worlds? – Graphus Jan 13 '17 at 09:23
  • You made the general statement that plywood is stiffer than oak. It is not and that is what I have commented about. :) – Ast Pace Jan 13 '17 at 16:22
  • @AstPace "Broadly speaking plywood is stronger and stiffer than most solid woods" is not the same as saying "Generally plywood is stiffer than oak" which is what you think I said. If you compare ply more like for like with wood it IS stiffer, e.g. get a piece of flat-sawn birch with cathedral grain, plane it down to the same thickness as a piece of quality birch ply and then compare those. Not textbook straight-grained, knot-free pieces of oak (Sagulator's assumption, note 1).compared with fir ply! – Graphus Jan 14 '17 at 08:42
  • @graphus Silly me, I thought we were discussing a question about a shelf unit made from oak and plywood. Except for good old Baltic birch, virtually all plywood is fir; if it's called e.g. walnut plywood, it has thin veneers of walnut with all interior plies of fir. – Ast Pace Jan 14 '17 at 20:03
  • @AstPace To be fair what I described in my last Comment is a fairer like-for-like comparison..... however. I have only just discovered that we've actually been arguing about very different products. When you buy hardwood plywood over there it should more rightly be labelled as "hardwood plywood" because it's 1) not all hardwood and 2) not plywood as I would understand it (might not even be all-veneer FGS)! I've been assuming from the outset an all-hardwood multi-ply, or as I'm henceforth going to refer to it, proper plywood :-) Off to edit! – Graphus Jan 15 '17 at 08:18
2

Looks good! Your calculation and assumptions about movement appear to be spot on. The plywood faces with solid oak edges will have almost no movement over the year, but the shelves will. Your solution to leave a gap in the dado joint at the rear should work fine. Glue them up on to the front styles and the first several inches of the side panels. Leave a small gap in the side dadoes as well (1/16") so that the shelves move freely in those slots as well.

Ashlar
  • 8,454
  • 5
  • 23
  • 47
  • Thanks! About the extra gap in the side dados, should that go the whole length of the dado or just where the shelf is not glued? I expected the fixed shelves to add rigidity, and a gap the whole way seems likely to undermine that strength. If they go just part way, can I reasonably add that by sanding? – Michael Urman Jan 12 '17 at 04:40
  • I am talking about a gap in the depth, not the height. Think of it as a glue gap. The sides of the dado which contact the top and bottom surface of the shelf can remain snug. – Ashlar Jan 12 '17 at 13:42
1

After getting general confirmation from the answers that others provided here, here's what I did to actually size the bookcase. And after going though the low and high points of our year's humidity cycle, I'm satisfied with the results.

Step 1: Identify the amount of wood movement expected

This first requires identifying the wood in question (solid oak), and the expected range of humidity swings in the area. I searched for something like "moisture content of interior woodwork on usda maps", and found this chart. From that I was able to estimate a 5% swing between the extremes.

But how much movement does that mean? In a search for something like "wood movement oak" I found this Wood Movement Chart, where you can see that 12 inches of red oak will move 0.9" radially to 0.22" tangentially across a 5% relative humidity range. Since my dimension was only about 9 inches, I only need to allow for 3/4 of that, and I'll assume I chose lumber really badly so will get the larger movement. So that means my shelves' depth will move by up to 3/16".

Step 2: Figure out where my lumber is right now

I was able to use an old moisture meter and take a reading of my lumber's moisture content. This old meter only reports a number on an analog readout, and you need to then adjust that number per a correction table. But it turns out that red oak is a direct mapping.

Unfortunately I lost my notes on exactly where I was. I think it was about 1/3 of the way through the humidity range I expect to see, nearer the low end. So that would mean I could expect up to 1/16 shrinkage or 1/8 growth. I adjusted my cut according to those calculations.

How did it work out?

So far so good. Of course real life dictated that as soon as I finished the bookcases, the shelves in question were completely covered. Thus any gap will never be seen. That may be the most important lesson to take away: as much as you want something to be perfect, it's worth considering how visible it will really be, and err on the side of whatever can't make your project fall apart.

Also, it's worth noting that my calculations here are very much worst case for my design. Since it has some solid lumber in the sides as well, those could effectively cut the depth of the shelves by another couple inches. So I may have overcompensated in multiple ways.

Regardless, I'm happy with the result, and so is my wife, and that's what matters.

Michael Urman
  • 293
  • 2
  • 8