How stress warps lumber, plus other FAQs

Question: When we rip a piece of oak lumber, sometimes we see one piece curve to the right and another piece in the same board curve to the left. We have to shorten these pieces to make them straight enough to use. What is causing this movement?

Answer:  Assuming that the lumber was dried normally and that there are no existing moisture gradients, the movement or warp immediately when we machine is caused by lengthwise stress, sometimes called longitudinal stress in textbooks.

In hardwoods, it results for three reasons:  growth stress, slope-of-grain stress (SOG) or drying stress.

GROWTH STRESS. When the tree is growing, there is a lot of weight on the stem from the branches. To keep from collapsing into a blob, the tree resists the force from this weight. If you recall great-grandfather Isaac Newton’s saying, “For every stress there is an equal but opposite stress.” So, the tree stem has stress in it.

The weight stress is minimized when we cut the tree into logs, but the counteracting stress within the wood still is there. Usually, this stress is fairly uniformly distributed, so it does not cause a great deal of problems, but sometimes the stress is large and uneven across the stem. At a sawmill, the lumber will come off the “normal” stem and be straight, but sometimes, when a board is sawn, it will develop immediate lengthwise curvature called bow. Any time you see bowed lumber, you know you are going to have warping problems when you cut the lumber into smaller pieces, especially longer pieces. Thank goodness that most North American wood species have very little stress within the wood overall. But, many foreign species, including eucalyptus, have so much stress that, as the tree falls to the ground when harvesting it, the end of the log will immediately develop two or three huge end cracks or splits.

Growth stress has been studied and solutions looked at for many decades. A practical solution still eludes us. The best idea is to eliminate any pieces that are warped when they come to us. In fact, the NHLA grading rules are very strict on warp in the middle and upper grades. So, growth stress is seldom a problem if the rules are properly applied.

SOG STRESS. When drying the lumber, shrinkage occurs, but in most cases, the lengthwise shrinkage is so small that it can be ignored. However, if there is slope of grain, the wood cell length direction or wood grain does not agree with the lumber’s length direction. A split follows the grain, so lumber with splits at an angle has an excessive slope of grain. During drying, the lengthwise shrinkage is attempted, but often the “normal” wood prevents most or all of this attempted shrinkage. Consequently, stress exists within the wood. It is most common around knots. As soon as we cut the wood into smaller pieces, this stress results in an immediate warp.

Some SOG occurs in softwood trees. Very little occurs in hardwoods. Crooked logs mean lumber will have some SOG. Sawing parallel to the bark at the sawmill means little if any SOG.

DRYING STRESS. Lengthwise shrinkage of lumber is not common, but in some pieces, we do have some stress after drying due to attempted, normal lengthwise shrinkage. To minimize this stress, the lumber is usually steamed or given a fine mist treatment to swell the wood, with the attempted swelling offsetting the attempted shrinkage. This process is called conditioning. Always insist on proper conditioning of any lumber you buy.

IDENTIFYING STRESS. Growth stress will affect green and kiln-dried lumber by causing the lumber to be warped. When ripping lumber, in addition to a sideways warp (side bend or crook), the ripped pieces will bend every which way. With SOG stress, the green lumber will twist, and the ripped pieces will twist. Drying stress will be uniform within the lumber, meaning that if a 6-inch-wide piece of lumber is ripped into three 2-inch pieces, the center will stay straight, while the left piece bows to the left and the right piece is a mirror image and bows to the right. That is, the lengthwise drying stress is uniform and opposite left to right.

Drying stress is preventable by conditioning in the kiln. SOG is controllable by proper sawing. Growth stress is controlled by eliminating warped lumber before drying or ripping.

Q: We have a little wood left over from a job and want to know the best way to store it. Thanks. Love your column, etc.

A: There are two concerns when storing lumber —insects and moisture.

INSECTS. The insect that comes into dry hardwood, the lyctid powderpost beetle, cannot fly very far, so the key to controlling insects in storage is twofold. First, keep the storage area free of all wood debris. This includes free of dust and small particles, as well as 4x4s, and any wood trash. This wood can be used as a breeding ground. Second, do not store the wood near any wood that could be already infected. Foreign wood has a very high risk of carrying the lyctid powderpost beetle indeed, but even some U.S. or Canadian wood can harbor this insect.

The big issue with insects is that it may be 10 months or longer between when the eggs are laid and when the insect emerges from the wood. During this time, there is no visual evidence of their presence. So, control is achieved by being really careful BEFOREHAND. Control, once an infestation is present in furniture, cabinets, flooring, etc., is expensive and difficult.

For 4x4s or other wood spacers, we can heat this wood in a small chamber to more than 133 F throughout. This kills any insects and their eggs. Of course, it does not prevent re-infestation after this sterilization treatment.

MOISTURE. For most of North America, the ideal Moisture Content (MC) of wood used for furniture, cabinets, flooring, etc. is 6.8% MC. This is achieved at approximately 37% RH. A household dehumidifier (I prefer a name brand made in the U.S. such as Honeywell, as they will perform better and last longer) will work very well to maintain dry wood at the desired RH and MC. You can use some 6-mil plastic to enclose the area you want to keep at 37% RH. In the wintertime, you may actually have to add a little moisture to the air to achieve this value. I do believe that a smoke alarm in the chamber would be prudent.

Q: We have been doing some tests with various wood adhesives and are surprised that the “Old fashioned” adhesives seem to be stronger than the modern PUR adhesives and even stronger than epoxy. Can you comment?

A: It is indeed a good idea to test various adhesives to see how they perform in your operation. Being honest, I am concerned that you are testing the adhesive strength because virtually every glue joint we make has the potential to be stronger than the wood itself. This means that a joint should never fail, but instead, the wood should fail. So, reading between the lines of your email, I do wonder if there is something being done wrong with the preparation of the wood surfaces prior to gluing.

Certainly, one requirement is that the surfaces be freshly prepared with sharp tools. Obviously, there must be adequate adhesive. Finally, the clamping pressure should squeeze the adhesive into the nooks and crannies as well as squeeze out the excess adhesive.

I am guessing that you have way too much clamping pressure. In short, if you squeeze too much, there is not enough adhesive left in the joint to do its job. We all know that there is a maximum thickness for a glue joint, but there is also a minimum thickness. A PUR joint must be a little thicker than a PCA joint.

Check with the adhesive manufacturer on the minimum thickness. Epoxy must be even thicker; with epoxy, we need enough adhesive to generate heat so the chemical reaction will occur.

Overall, the pressure is a critical variable, but often pressure is not checked or controlled well. This also means that once pressure is applied, it is extremely important not to let the pressure drop at all until the glue sets.

Q: We are edge gluing some 5/4 oak. We have developed more glue failures than we want. When we switched to 4/4, the joints were perfect. Why would we have trouble with 5/4 and not 4/4? Samples enclosed.

A: There should be no difference based on thickness, so something else changed when you switched thicknesses. When I looked at the sample you sent, I can see that the 5/4 joint failures are due to dull sandpaper on the surfaces to be jointed. As a result, the wood surface is not very strong. The joint is only as strong as the weakest link, which in your case is the wood surface.

Special note to our readers: When I spoke with the people sending the samples after they got my response above, they confirmed that in addition to changing thickness, they also changed the sandpaper on their drum sander that prepares the wood surfaces. They asked how to determine when the sandpaper was too dull.

Three ways I know of are: 1) by feel, making sure the particles still feel sharp; 2) by monitoring the current draw on the belt motors, as dull paper requires slightly more current; and 3) by measuring the wood temperature immediately after sanding, as the temperature will be higher with dull paper. Maybe someone else has another idea. Let us know.


Have something to say? Share your thoughts with us in the comments below.

Profile picture for user genewengert
About the author
Gene Wengert

Gene Wengert, “The Wood Doctor” has been training people in efficient use of wood for 45 years. He is extension specialist emeritus at the University of Wisconsin-Madison.