It is common practice to edge glue several narrow strips (called staves in this article) into a wide panel. From time to time, this panel will warp after gluing. Unfortunately, there is a lot of misinformation about the causes and cures for such warp.
Let's begin our discussion by considering some basics about the behavior of wood.
Moisture issues
First and foremost, wood shrinks when it loses moisture and swells when it gains moisture. The approximate change in size for most hardwoods is 1 percent size change for a 4 percent moisture content (MC) change. This means that for a 24-inch-wide panel, a 4-percent MC change will cause the face to change by about 1/4-inch. Although a 4-percent MC change occurs when the wood is not dried to a low enough MC initially and is then exposed to a dry environment, it would be rare to have a 4-percent MC change occur in use.
It is key to understand that if the moisture does not change, there will never be any shrinkage or swelling. For this reason, it is important to measure the MC of the staves used in construction. Further, the relative humidity (RH) in the shop and in the home must be measured to assure that no MC change will occur in manufacturing or in use. Wood edge-glued panels will average about 7-percent MC in most homes and offices.
As a corollary, if the MC change is small, the shrinkage or swelling will be small. Further, slow changes in moisture content seem to cause fewer problems than fast changes.
Balanced changes
Second, if one face of an edge-glued panel is exposed to drier or wetter air and the MC of this one face changes while the other face does not change (because it is not exposed to this air or because it has a protective finish on it, for example), then the panel will warp as the exposed face is shrinking or swelling in size while the other face is not changing. A 1-percent size change on one face and not the other will result in about 1/4-inch of cup; that is, the edges will be elevated ¼ inch above the center. It is this differential shrinkage that causes many problems in warping of panels.
Example #1: Consider a pile of recently manufactured panels with an average of 8-percent MC, with only the top panel face exposed to the dry air (5-percent EMC) in the work area. The exposed face will dry and will shrink causing the top panel to cup upward nearly 1/4-inch.
Example #2: Consider a coffee table top that has several coats of moisture-vapor-resistant finish on the top, but has an unfinished bottom side. When the heat comes on in the home lowering the RH and EMC, the bottom side will quickly dry and shrink while the top finish is retarding any moisture change. The net effect is that the table will crown.
Wood location in tree
Avoiding warp - By watching out for basic moisture content and relative humidity factors you can avoid most causes of warp in glued up panels.
Third, the face of a stave that is closer to the bark will shrink or swell slightly more than the side closer to the center or heart of the log, but (here is the key!) this difference is so slight that it is hard to measure. It is for this reason that some people suggest that the individual staves should have their ring pattern reversed so the face closest to the bark alternates from one face of the panel to the other face in adjacent staves. Some people even go so far as to say that the ends of the staves should also be alternated, end for end. The practical truth is that such effort is OK to do, but does not help warping of the panel.
The potential effect of such a difference is greater in staves sawn very close to the center of the log, but for annual rings that are gently curved, the effect is not important.
The truth is that a moisture change that affects both faces equally will rarely cause any warp in a glued-up panel. The effect from the second item above is large - hundreds of times larger than this effect.
Measuring moisture
Fourth, when checking the moisture of the individual staves (in many cases, the best would be staves between 6.0-percent MC and 7.5-percent MC), keep in mind that for some pin-less meters and for pin meters with uninsulated needles, the surface MC can affect the reading greatly. So, 65-percent RH in the early morning in the shop before the dust system is started can give higher MC readings than the true average MC.
Conversely, a 9-percent MC piece that dries on the surface a bit due to dry shop conditions will often read lower than the true average MC. Know the limitations of your moisture measuring equipment.
Also, keep in mind that a dry kiln may have 10,000 pieces of lumber in it, yet the operator often checks only 10 or 12. The MC indicated by a supplier can sometimes (accidentally) be in error.
Incidentally, always measure the moisture content BEFORE the panel warps. After the panel dries and warps, the moisture values will always be correct.
Wet panel staves
Fifth, by far, the most common reason for warp is that the panel staves are wetter than the environment, and so when the panel faces dry unevenly, warp results. Some manufacturing plants have found that by raising the RH in the plant to 50-percent RH (or so), then warp problems disappear (and so do end cracking problems). In truth, warp is often less, as 50-percent RH is equivalent to 9-percent MC, and incoming lumber often is 8- to 9-percent MC. The problem with this approach is that the panels will dry and shrink when they reach the customer's home or office. So, high RH in the plant is merely postponing the warp problem (and end splits too), not curing it.
Stave factors
Sixth, stave width is often cited as a problem. Many people suggest that staves should be no wider than 2-1/2 inches. Indeed, sometimes there can be a stave that misbehaves, shrinkage-wise. By making staves narrow, chances of getting a lot of bad shrinkage in one panel will be lessened because of the variety of staves used.
It is more likely to see bad behavior when the wood in a stave is close to a knot. Sometimes, yield improvement programs encourage cutting too close to a knot (from a shrinkage standpoint). Lower grade lumber also tends to have more undesirable shrinkage behavior. Staves made from flat, straight grain lumber that is at the correct MC can be many inches wide with no warp problems.