Building shelves, need advice

I had that problem with my Baltic Birch. Never lean large pieces regardless of how dry they are. If you are going to leave pieces out, either put them on a roll cart where the weight is on the bottom edge, or lay them flat(takes up that precious floorium space). The gravity alone warps the piece when the weight is leaning on the top.

Oh wow. Never thought it would warp like that. Lesson learned. Thank you for the advice.

I knew hardwood warps easily, but I always though plywood was built to resist it, especially 13 layer Baltic birch. Lesson learned. LOL

There are a lot of variables, of course, but it’s mostly about moisture, and is affected by grain orientation. And the moisture levels in DFW change dramatically in DFW in the spring.

What was the size of the lumber before you started? Was it kiln dried? I presume it was rift sawn. Did you thickness plane both sides by the same amount (to keep moisture levels constant on both sides)?

Got any pics?

-- Barry, http://BarrysWorkshop.com/

This is a little confusing. The OP said the boards cupped. Then he was advised to not stand up the boards because they will warp. A cup is across the board, and warping, or bowing is along the board. Standing a board on end will not cause cupping. It was likely caused by the factors Barry laid out.

As for leaning boards becoming warped, I’ve left lumber standing against the wall for long periods, and I’ve never had a problem with warping. Take a look at some great woodworker’s inventories, and the boards are often standing on end. Dave Marks comes to mind. He did a show focusing on his wood storage shed. The wood was vertical. Woodworker’s Source, where I shop here, has all of their exotics standing in racks.

In general, gravity doesn’t affect wood like that — it’s the environment. If gravity could do it, then the warped cherry board that I supported on the ends and put 40 lb of bricks in the middle of, would have straightened out. It didn’t.

True, but he stated he leaned them against the wall, not upright. My apologies, I don’t always get the nomenclature between bowing, cupping, warping and twisting correct. My plywood is the standard 3×5 Baltic birch, so I think you can consider that it bent when leaned against the wall.

We have had some wicked storms in DFW lately so there’s been plenty of moisture as well as the temp being up and down.

I bought FAS White Ash, Flat sawn
$3.98 per board foot at a local supplier…not a Home Depot kind of store. I think it was kiln dried.
Width ranged from 8 to 10 inch
Length ranged from 9 to 10 foot boards

I researched flat, quarter and rift sawn and learned that moisture affects flat sawn most.

I did not thickness plane evenly on both sides. I planed one side on my jointer to flatten then ran it through the thickness planer to get the other face flat. I’d say the cut was 60/40 so not exactly same on both faces.
Thank you for the lesson in thickness planing, Barry. Much appreciated.

Rich, I used the words “warped”, “cupped” and “bowed” synonymous. Sorry, I’m not very astute with the terminology. I can design and engineer projects…cutting, gluing, jointing…I get the mechanics, but I am learning I lack the deeper knowledge of the physical properties that can affect my projects. MDF shelves and plywood based cabinets are rather different than solid hardwood table tops.

So to clarify…
The boards that I glued together are flat long ways (16 to 24 inches)
They finished shelves are bowed
After watching several videos on proper glue up dos and dont’s, I am certain I applied too much pressure across the boards and for way too long.
I made it worse by leaning the boards at 60 to 75 degree angle against the wall.

PICS!! As requested.

Thank you for sharing your knowledge gents. I appreciate it.

Those panels look to be around 30” wide, give or take. That’s a tiny cup for that width. I’m not sure what construction techniques you plan to use, but if those are going in dados, they’ll bend to fit. In other words, they’ll only be a problem if they are unrestrained. Not sure if that makes sense, but 3/16” over 30” or so is easy to flatten out.

They will be 24 inches finished. So binding a spine under it will flatten it and keep it flat?
When I say “spine” I am thinking of a solid 1 inch square piece of solid Ash.
I thought about that at the beginning but thought the shelf might just bend again.
SO the spine is a valid idea?

I agree with rich, a small cup in that width can be “forced” out.

It would help us understand if we could see a sketch of the design.

If you flatten with a “spine” be sure to attach in a way that allows for wood movement.

Leaning against the wall, by itself, was probably not the cause of the cup, but leaning one board against the other with no air space in between could easily cause a cup. Moisture exchange between air and wood will be different on the exposed face than on the “covered” face. To store flat panels, and have them stay flat, best practice is to stack and sticker on a flat surface.

The term “warp” applies to all of these types:

-- Jerry, making sawdust professionally since 1976

TheNewGuy,

I am guessing that the explanation for the currently cupped panels is either moisture or a jointer fence not set perfectly square to the jointer table. Like jerryminer, I doubt that leaning the panels against a wall, assuming exposure to air on both faces, caused the cup. However, whenever I lean panels against a wall, I like to set the end of the panels on some scrap wood stickers to keep the panel off the concrete floor.

If the cupping is due to moisture, there would be more moisture on crowned face versus on the opposite face (concave face). Assuming both faces were exposed to air when leaning upright, I wonder whether the position of clamps kept in position for so long might be at least part of the explanation. If the clamping forces were not perfectly parallel to the panel faces, slightly more clamping force could have existed along one face than the other face. These unbalanced forces could have compressed the wood fibers a little more along the face where clamping pressure was slightly greater. This could in turn reduce the ability of the wood fibers to absorb and possibly even release moisture at the same rate as the opposite face. If this guess is correct, then removing the clamps when the glue was fully cured (about 24 hours) could have prevented the cupping. I also suppose that this is why people often recommend alternating the positioning of the clamps; one on top, the next on the bottom, then next clamp back over the top of the panel.

If the problem is due to a moisture imbalance, the concave side (the dryer side) of the panel could be wetted with water and left to soak into the wood. A dripping-wet clean rag placed on the concave surface for a few hours could be enough to flatten the panel though careful monitoring of the panel would be required. If too much moisture is introduced, the cup could reverse to the opposite side.

With some luck, enough moisture would penetrate into the drier wood fibers, causing the drier wood fibers to swell and in so doing, cause the panel to flatten. Once moisture levels on both faces are equalized, keeping the panels stickered, as shown by jerryminer, would help maintain the moisture balance.

As an aside, if the cup is due to moisture, I doubt that positioning the flattening cauls (or as you said splines) edge-wise rather that flat-wise, would have made much difference. Moisture induced forces in wood are significant; enough to split granite.

If the jointer fence is not set perfectly square to the jointer table, a panel could appear to have a cup when in fact it is a more like a segment of a polygon. This can be checked by placing a straight edge across the width the each of the planks on the crowned side. Planks that are themselves flat suggests a problem with the jointer set up; fence set out of 90 degrees to the table. If this is the case, then there are several method for flattening the “cupped” panel, none of which are very good solutions. The easiest as already suggested is fasten some rails that would act to flatten the panel. However, since the cup occurred despite flattening cauls while the panel glue-up cured, this rail-flattening method may not work very well.

To avoid problems with jointer set up, the adjoining edges of planks that will be glued together can be jointed so that upside face of one plank is held against the jointer fence. The second plank is jointed with downside face against the jointer fence. This produces complementary edge to face angles that sum to 180 degrees at the joint line and thus produce flat panels even if the jointer fence is not set at a perfect 90 degrees.

...I thought the Pine wood spines would prevent that. I used them flat wise, not edgewise. They seemed pretty stiff and it looked flat. I guess I should have used them edgewise?...

Any tips on keeping boards flat?...

- TheNewGuy

I like to use cauls to keep my panels flat. Not a flat piece of pine, but a purpose-made caul, with a built-in curve that flattens out under clamping pressure. Mike Henderson, a fellow Jock, wrote a tutorial for his method of making these. You may want to read it here

In use, the system looks like this:

-- Jerry, making sawdust professionally since 1976

Thank you, gents, for the education. Been gone a few days but back and ready to get tame these shelves.

Mike Henderson’s guide on cauls is awesome! Definitely adding those to my shop.
Thank you for that jerryminer.

I’m thinking 1” spines. Shelves are 5/8 so do you think 1” is sufficient?
Picture doesn’t show it but I am thinking I need another spine halfway to the back corner (tip of the pie so to speak)

Should I dado the shelf 1/8 to sink the spine or surface mount?
Perhaps deeper at a 1/4?
Use biscuits or just glue?

1” is likely enough. (I would orient the grain to run as close to vertical as possible for added stiffness)

How are you attaching the shelves to the uprights?

DO NOT glue the “spines” to the shelves. You will be creating an issue with wood movement. Instead, use screws in slotted holes or table clips/buttons (so the shelf can expand/contract without splitting)

If this were my project, I think I would have oriented the grain to run diagonally across the corner, rather than 90 degrees to the wall. The flatness issue would have pretty much disappeared in the shelf-to-upright connection.

You’re going to be OK, though. The ash shelves look great. Very nice job in grain matching.

-- Jerry, making sawdust professionally since 1976

A lot of good advise already given, and I hope I’m not repeating anything already posted. Just a couple things after looking at your pictures, IMHO …
1. Don’t leave your stock flat on the concrete floor …. its like a sponge in water
2. With panel glue ups, if at all possible, alternate the growth rings. One board crown up, next board crown down and on and on.
3. It appears you have not yet cut your panels to final size … if you’d like, you could rip n flip the panels on your table saw, trim the freshly cut edges and re-glue. Remember what JBrow mentioned before about the orientation of your stock during jointing and complimentary angles …. well it also works for table saw cuts. You can compensate for the smallest deviation from a perfectly square blade to table saw top.
4. After you’ve glued your panels with biscuits wait several days, a week is better, before final surfacing. If you sand to soon you might end up with football shaped depressions, that will show themselves after you’ve applied finish. The wood, and more so the biscuits, absorb moisture from the glue, and swell. You sand the panel smooth and flush. The moisture evaporates with time, and the wood and biscuits shrink, to reveal a nice football shape where you placed your biscuits.
Good luck with your project :)