Moisture and Cupping with Glued-Up Solid Poplar Panels

This thread about poplar panels that cupped while stored sheds light on wood moisture movement issues in general. June 30, 2009

I have been working at a small cabinet shop for about a year and a half. My boss is kind enough to let us employees work on small personal projects after work. Since I am still a novice cabinet builder, I consider myself an even more novice furniture maker. I received a good deal on some kiln dried poplar (free) for a mobile cabinet I've been wanting to build. The downside is that it was narrow strips left after ripping out trim blanks, so I had to glue up all of my panels out of many pieces. I face jointed, then edge jointed, and ripped the pieces before gluing. I didn't plane all the pieces to a common thickness because I planed to thickness the panels on our widebelt sander afterwards. All the panels glued up fine, even though some of them ended up thinner than I wanted as I have been working on this project mostly evenings and weekends for a few hours at a time.

I would unstack and then stack the pieces onto a cart before and after working on it. After I had thickness the panels they were all flat, but when I returned to the project after a few days break, one of the shelves and one of the side panels cupped-the side is pretty bad. What did I do wrong?

The sides were stacked on the bottom of the cart with the other pieces on top. I have already had to scrap the poplar bottom for a plywood bottom because of a layout error. I would hate to have to replace both sides with plywood, both because I would have to buy said plywood and because I will have wasted a lot of time (I won't say how much time for my pride's sake) and it will change the appearance of the piece. Does anyone think it will be possible for me to pull the cup out of the sides by using screws through the bottom into the sides as well as using pocket screws to attach the top stretchers to the sides? This is before screwing the top on through oversize holes to allow for top movement.

Forum Responses
(Furniture Making Forum)
From Gene Wengert, forum technical advisor:
Wood only changes size or shape over time for one reason and that is that the MC is changing. If it changes right when you machine it, then you have drying stress. So, you need to check the MC yourself before using wood. There is no guarantee that kiln-dried means a correct MC (usually 7.0% MC). Read up on how to use a meter - search the archives here.

From contributor C:
Given the common price of poplar and the fact that you would prefer to have wood panels, I would suggest you just remake the parts. One of the hardest lessons to learn is when to turn around and throw away a part that you already have quite a bit of time invested in, because it can't be used, or the fix isn't worth the effort involved. I understand it is aggravating, but I have never done this, and said I wish I had tried to fix a piece, rather than just stopping and producing a new one.

From contributor J:

If the whole glued-up panel cupped then one side has lost or gained moisture faster than the other. Either the concave side has shrunk because it has dried out or the convex side has expanded because it has absorbed moisture. These changes can happen in all sorts of ways, such as one side being exposed to especially dry or especially humid air, left lying on a damp surface, or exposed to direct sunlight. Think carefully about what you did and it may become obvious where you went wrong.

In any case, the most conservative thing you can do is to stack these parts indoors, out of direct sun, with stickers in-between so that both sides of each part are exposed to the air, and let them stabilize for a while - maybe a week or so. After that, if the cup is mild then you may be able to pull it flat with screws. If it's severe then you'll have no choice but to either re-make the panel or cut it into a few pieces, re-joint the edges so it goes together with less cup, re-glue and re-flatten on the sander.

However, backing up a few steps, you said, "I have already had to scrap the poplar bottom for a plywood bottom" and also "before screwing the top on through oversize holes to allow for top movement." It sounds as if you don't quite grasp where it's necessary to use joints that accommodate movement and where it is not. If you use a plywood bottom with solid wood sides, you'll have sides that expand and contract with seasonal humidity fluctuations firmly attached to a bottom that is relatively stable. This is a no-no. I don't know whether you really need oversize fastener holes between the cabinet and the top surface because you haven't indicated what material you'll be using for that top, but that joint between the solid wood sides and plywood bottom is problematic.

From contributor L:
Before you give up on the panels, I have something fun for you to try. I'd like to agree with all of the above comments and add one or two. It's almost always better to start with new wood when fixing an error for a host of reasons. Second, I think it's a good practice to plane everything the same thickness prior to gluing, not mill it afterward.

Now on to your cupped panels. On a nice sunny summer day, lay the panel in the grass in full sun. Place the concave side down on the grass and the convex side toward the sun. Now sit in the shade and have a glass of iced tea. By the time you've finished your tea, the panel will have begun to straighten. Keep a close eye on it. Let it go just past straight. Then let it stabilize for a day or two in the shop placed where air can circulate all around it. If the panel is straight after a couple of days, you're finished. If not, you know just what to do.

From Gene Wengert, forum technical advisor:
When I read what I posted, I might have not been clear enough about stress (also called casehardening). If you have wood pieces that have stress from drying (or casehardening stress), the pieces will look great, but if you plane one face more than the other, you unbalance these stresses and get immediate warp.

Otherwise, if the warp occurs over time, it is moisture related. Note that wood always cups toward the bark of the tree. For this reason, some people will alternate the individual pieces that are glued into a panel. Although this sounds like a good idea, it ignores the basic problem - the moisture content is not at the correct level. Note that if you do alternate the pieces and the panel still cups, then one side of the panel is at a different moisture level than the other side (or rarely, there was a wetter core and you planed one face more than the other exposing the wetter core to the drying air).

Oftentimes, the MC after a problem occurs if uniform and at the correct level, therefore, it is important to measure the MC before any problems are seen. Incidentally, it is not common to see moisture problems in the summertime as wood MC and air EMC are usually very close. It is in the wintertime when air EMCs can drop under 5% EMC with wood at 8% MC or more that we have greater problems.

From contributor J:
It seems fairly obvious that this only applies when the MC is dropping. If the board is flattened when dry and the MC then rises, it cups towards the pith; if there's a sudden change in environment then, as you pointed out, you can find yourself with a gradient that warps the panel in ways that have nothing to do with the orientation of the annual rings.

Also, I disagree with the idea that alternating ring orientations in a glued-up panel ignores the basic problem that the MC isn't at the correct level. The practice looks, to me, like a rational way of moderating anticipated potential deformation, done with the understanding that the MC might change, regardless of whether the change is because the wood was wet to begin with or because the object might someday be exposed to a less than ideal environment.

From contributor D:
97% of the cupped panels I have encountered were cupped by environmental changes only. Along the lines of what has been posted, environmental changes are the big culprit. Stressed wood, bad clamping, improper drying, all can wreak havoc, but are relatively rare when working in a professional shop with professionally processed lumber. Gluing strips to make wider panels will even help minimize the effect any single bad board will have on the panel.

I once had the occasion over a five year period to observe white pine six panel doors stacked horizontally in racks in an open, unheated warehouse. When the humidity increased, the top door on each stack bowed up in the middle - along the 80" length - often as much as 3/4". If the humidity dropped, then the door would flatten, then cup up at the ends, again up to 3/4". This could all happen in 36 hours or less. The effect was so strong that it caused the wood doors to behave in ways contrary to normal experience with wood.

The "warped" doors were originally culled out to be returned to the vendor for credit. We also often got doors back from a jobsite that were "warped". In vertical storage for a few days, with good air circulation on all sides, the doors would miraculously flatten themselves out and be usable. We sized 1/2" luan panels to the size of the doors, and kept these on top of each stack, and our problems were gone.

Today, if we have a stack of panels, I advise a piece of cardboard or plywood on top if the weather is changing to prevent the inevitable. The cardboard cups, and often the ply/MDF tries to, but the panels stay flat.

From Gene Wengert, forum technical advisor:
Indeed, you are correct. I should have said that drying wood always cups toward the bark. The reverse happens when being rewetted. It is rare however, to see wood that is too dry, compared to the frequent problem with wet wood in a dry wintertime environment. Also, one typo - "Oftentimes, the MC after a problem occurs is uniform" (not if uniform).

I did not mean to imply that alternating rings is not a good idea. Rather, it does not address the root of the problem which is moisture change, which can be solved by getting the correct MC (wrong MC is the number one cause of warp in my years of experience). A kiln operator typically measures 10 pieces out of 10,000 to determine when the load is dry and having the plant EMC correct for the lumber and the customer's place. A good finish will also help minimize the rate of moisture change, allowing the wood to "absorb" some moisture change seasonally without much warping, especially when compared to a poor finish.

From the original questioner:
Thanks to everyone who has responded so far. I checked the panels, and to my suprise the cup was less severe than two days ago. They had been standing up against the wall rather than stacked on the cart.

Gene - you brought up the finish. Would it be a bad idea to leave the interior unfinished or should I finish it as well? I will probably use a precat lacquer. Contributor J - thanks for the heads up on the plywood bottom. Should I use oversize holes there? Also the top is solid poplar.

From contributor J:
If the top is solid wood then you don't need slotted holes for mounting it because it will shrink and expand ("move") with the solid-wood sides. When you hear advice about mounting solid wood tops with screws through slotted holes, it's because the adviser assumes the cabinet is made out of some engineered sheet good like plywood, MDF or particle board, because those materials are standard for kitchen and bath cabinetry. The way you are building this is, well, it's wrong, but the point is that it's not like standard
cabinetry so the standard advice won't always apply.

When we get to the joint between the plywood bottom and the solid-wood sides, things get stickier. It will be tough to make a corner joint that's strong enough yet accommodates movement. Even if you could come up with a workable joint, the changing width of the side next to the relatively fixed-size bottom would create an odd step at some times of the year. Never, ever build a cabinet for a customer this way.

However, since this cabinet is for yourself, go ahead. Those bottom joints are going to look funny to any experienced eye and will probably loosen after a few years, but by then you'll be amazed at how much you've learned and this little cabinet will be fun to look back on.

From Gene Wengert, forum technical advisor:
We do need slotted holes in solid wood when the grain of the side piece runs perpendicular to the grain of the face. We would, for example, see this with a table skirt or cleat that has the grain (longitudinal or lengthwise direction) running the length of a table while the lengthwise direction of the table is running in the width direction (a common construction technique). As wood does not change length at all lengthwise, we have a large contrast between the movement of the top and the non-movement of the skirt or cleat.