Question: Why does a cabinet door warp? That is, we ship them flat but then after installation, mainly one corner is  not flat? What can we do? We only have it happen rarely, but a repair is not cheap. The warp seems permanent, as the door never returns to flatness.

Answer: This is a great question that needs a detailed explanation. Although a door can cup or crown, almost all objectionable warp is twist. Twist is where if we put the door on a flat surface, one corner is raised off the surface. The question is “Why does a door twist?”

The initial answer is that the warp could be the rails and/or stiles of the door that are causing the twist, or it could be the center pieces (the inserted panel in some cases) that are causing the twist. The first step is to use a handheld jig saw and cut out the center. Now, is the center warped and the frame returns to flatness?  Or vice versa? I have seen cases where it is the center that was warped (even worse than when in the door) and I have seen cases where it is the outside door frame that warps more when the center is removed.

So, now we know which pieces of wood are warping. What makes wood warp, especially twist? The answer is twofold:  wood grain AND moisture. (Note: it is not “wood OR moisture.”)

WOOD GRAIN. When making pool cues, when making long, skinny broom handles, when making large picture frames, when making wood roof shingles, and a few other products, all of these must be perfectly straight lengthwise, both when made and when in use. Yet, we know that wood shrinks and swells when the moisture changes. But, lengthwise, there is very little movement. 

So, we need to make sure that the grain of the wood is parallel to the lengthwise direction of the pieces of wood. That is, we see worse warping behavior when the lengthwise grain direction of the wood is NOT the lengthwise direction of the product. When these two are NOT parallel, we say the wood piece we are working with has slope of grain.

Indeed, there is steep slope of grain around every knot, so we prefer to avoid knots AND the area right around the knots when desiring flat, straight products. (Unfortunately, in an attempt to increase yield, some supervisors instruct the employees to cut as close to a knot as possible, ignoring the swirly grain around the know. Bad idea when flat products are needed.)

Slope of grain can occur because the tree grew the wood cells in a spiral manner, which means that the vertical direction of the tree is not the vertical direction of the cells, giving slope of grain. This is most common in softwoods, especially in the first 20 rings from the center of the log. Not common in hardwoods, however.
Slope of grain can also occur (most often occurs) when the sawmill cuts lumber from the log, but does not cut parallel to the bark. (In sawmill language, the sawmill does not use taper sets.) I do appreciate that oftentimes we have no contact with the sawmill that is producing our lumber. If we do not, then our only hope is that we can control the moisture content perfectly (next section).

For pool cues and other products that absolutely must have no slope of grain, it is common practice at the sawmill to split the log first and then, because a split follows the grain, cut parallel to the split. More expensive? Yes, but worth it because the grain is so straight, or parallel to the sides, that the product stays straight. For cabinet doors, we need to request that the sawmill saw parallel to the bark using taper sets as needed.

MOISTURE CONTENT. Now, understand that warp does not occur one day because the wood decided it wanted to warp, or there was an eclipse of the moon, or the _____ won the election. All warp occurs because the moisture content changed and moisture changed because the relative humidity changed; temperature is not a factor, only relative humidity.

The first step in controlling moisture content to control twist is to make sure every piece of wood is at the same moisture content as it will be in the cabinet door in the home or office. For most of the North America, this means 6.8% MC average with nothing wetter than 7.7% MC and nothing drier than 5.9% MC. (Note: saying 6 to8% MC is too nebulous.) Certainly, along the coast of Florida, New Orleans, and Seattle would be wetter, while Denver and Phoenix would be drier. Be careful if you market to these wetter or drier areas.

Plus, twist is worse when the moisture change is large and fast. Stated another way, slow, small moisture changes will not be a problem even with some slope of grain. Of course, the moisture in our manufacturing facility, or shipping container, or warehouse storage facility is often not close to the moisture in the home or office. So, there is likely a shock to the cabinet wood when it is first installed in the customer’s home UNLESS we have a good moisture vapor resistant finish. Such a finish, even if it costs a bit more, is worth the cost. A good finish retards moisture changes very well; the slow change allows the wood to stay relaxed and warp very little, if any.

(Note: Covering the newly made cabinet with a plastic bag or plastic wrap can be very effective in preventing any moisture change between manufacturing and installation. No moisture change means no warp. So further, if the plant humidity and the home humidity match well, then there will be no moisture change and no warp. Most homes are between 30% RH and 50% RH, wintertime to summertime. A good target for the plant is 37% RH during the heating season. This is equivalent to 7.0% moisture content in the wood.)