Cracks between Boards
Almost every wood floor endures some expansion and contraction as seasons and humidity levels change. When homes are heated, humidity levels plummet, boards shrink and spaces appear between the boards. In dry months, cracks can easily develop to the thickness of a dime on a typical solid 2 1/4″ oak floor, with light-colored woods making the cracks appear larger. Plank floors will also show cracks more. These spaces are to be expected and usually close up as the season changes and moisture returns to the air. To reduce the degree of change, homeowners can add moisture to the air during the dry months, ideally by installing a humidifier in the furnace.
As with cracks between boards, both cupping and crowning are natural reactions to moisture and should not be a concern if they occur only to a minor extent. More severe cases, however, indicate a serious moisture problem.
“Cupping” describes a condition in which the edges of a board are high and its center is lower. Humidity is usually the culprit, although cupping also can happen after water has been spilled on the floor and absorbed into the wood. The moisture causes the wood to swell, crushing the boards together and deforming them at the edges. In order to repair the floor, the cause of the moisture must be identified.
Most often, indoor humidity will have to be controlled. Other causes could include situations such as a plumbing leak in the basement, which can allow moisture to migrate up into the subfloor and the wood flooring. Once the cause of the moisture is controlled, cupping usually can be reversed. Oftentimes the floor may naturally dry out and improve over time. Fans may be necessary to speed the drying process. After the floor has dried, it may be necessary to re-coat the floor with the finish, or to sand and refinish the floor.
Crowning is the opposite of cupping: The middle of the board is higher than the edges of the board. This can occur when the surface of the floor encounters moisture. More often, it results when a floor has been sanded too soon after it has cupped. When this happens, the top edges of the board are sanded off and thus are lower than the rest of the board when it returns to a normal moisture content (MC).
Buckling is one of the most extreme reactions to moisture that can occur with a hardwood floor. It happens when the floor literally pulls away from the sub-floor, up to heights as high as several inches. Fortunately, buckling is an uncommon occurrence, usually happening only after a floor has been flooded. Even in such cases, it is possible that a floor can be repaired instead of being totally replaced.
Sunken Joints in Panel Manufacturing
Sunken joints are slight depressions at the glue lines in a finished panel (sometimes noted before veneering or finishing, but often most obvious after finishing).
The most common cause of a sunken joint is that wood around the joint swells when it absorbs water from the glue itself. If the wood is planed or sanded flat before the swelling goes down (that is, before the MC equilibrates), the wood at the joint will shrink later, creating a depression at the joint.
The remedy for a sunken joint is to wait for the moisture around the joint to dissipate before planing or sanding. Generally, seven days at warm temperatures is sufficient.
In a few cases, a very dry wood panel machined flat in the dry condition and then permitted to regain moisture during storage will have very little swelling in thickness at the joint area (the glue is rigid) while the wood away from the joint will swell in thickness normally, creating a sunken joint appearance. The cure, in this case, is to make sure that the MC of the lumber at the time of manufacturing is within 2% of the EMC of the air, even in the humid summer months.
Uneven Surfaces in Panel Manufacturing
Occasionally, several days after sanding, a glued-up panel seems to have been made up of pieces of different thicknesses. This is often referred to as planking.
Assuming that the panel was even when it was sanded or planed, the wood could change shape only if its MC changed. Therefore, planking is caused because individual strips or pieces were at varying moisture contents when the panel was glued up, and these MCs were not equalized before sanding or planing. For example, a 1″ thick panel could experience a noticeable thickness change of up to 0.01″ with only a 2% to 4% MC difference between individual strips. (That is, planking can be obvious when one strip is 6% MC at the time of gluing and the adjacent one is 8% MC.)
Panel End Splits
Within hours to a week after edge gluing of a panel, end splits may be noted, usually, but not always, at the glue line. End splits develop only because wood is shrinking, and the only reason wood shrinks is because it’s drying. In short, end splits are caused because the panels are too wet for the dry air in the plant.
Remember that the end grain of wood dries up to 50 times faster than the edge or the face. The ends, therefore, dry rapidly and shrink, while the center of the panel dries and shrinks more slowly. This difference leads to stress build-up and, potentially, end splits.
To correct this problem, keep the MC of the lumber within 2% of the EMC of the surrounding air. Because the EMC in the plant and in the home or office is lower in the winter, lumber should be dried to a lower MC in the winter (typically 5% to 7.0% MC), and/or the plant must be humidified slightly.
Cupped or Curved Panels
A panel, after manufacturing, may show cup or curl across its width. There are three possible diagnoses for this problem, but all three are caused by moisture changes in the panel that, in turn, result in shrinkage.
A. Because the face of the lumber closest to the bark shrinks more than the other face, a glued-up panel with most or all of the pieces with same grain orientation will shrink more on the bark side as it dries, thus causing cupping. Most important, lumber should be dried to the correct MC to minimize all shrinkage. As an additional safety factor, individual pieces should be varied in their grain orientation-bark side up, then pith side, and so on. (Note: The reverse situation occurs if the lumber is too dry for the EMC.)
B. If the finish on both sides of the panel isn’t identical, the two faces can swell at varying rates, resulting in warpage. Therefore, the finish must be applied equally to both sides, especially water-based finishes. Again, all wood should have similar initial MC.
C. If panels are designed with a cleat or rail which is securely fastened to the bottom of the panel and if the panel changes MC (resulting in shrinking or swelling), the top face is free to move but the bottom face can’t. Warping will result. Therefore, the cleat must be free to move slightly as the panel shrinks or swells. I recommend using slotted screw holes and not gluing the cleat to the panel. But the real cure is using lumber at the correct initial MC.
Splitting and Chipped Grain When Machining
Over-dried lumber is stronger and less bendable, so it is more prone to developing splitting and chipped grain when it is machined. Splitting is often noted with pines; chipped grain with dense hardwoods like oak. (On the other hand, lumber that is too wet will develop panel splits, open glue joints, and warping, as mentioned previously.) The rule is that in addition to sharp knives and properly adjusted machines, avoid excessive over-drying and under-drying. Get the wood at the correct MC. Measure the MC yourself-do not trust an invoice or word of mouth.
Fuzzy Grain after Machining
This condition is characterized by wood fibers not cutting cleanly; instead, they bend over. Then, when they pick up a little moisture, they spring back and stand up above the surface, looking like “peach fuzz.” If the stock is being run slowly, planer knives are sharp, and sandpaper is fresh, this fuzzing is caused by too high an MC in the lumber.
Gluing, Laminating, and Veneering Failure
One of the major areas in furniture manufacturing that causes problems is gluing-edge gluing, laminating, and veneering. There are at least 100 things that could be wrong when a glue joint is below par. But, as a general rule, a gluing problem is a breakdown of one of the five “links” in a glue joint.
The adhesive, or middle link (#1), is determined by the strength of the adhesive itself. If the proper adhesive has been chosen and handled correctly, this link shouldn’t be a problem. The next links (#2 and #3) represent the bond between the wood and the glue. In practice, weakness of this bond is the most common cause of glue-line failure. Weakness results because the adhesive cannot attach itself to the wood. Two common reasons are: (1) the surfaces to be glued are several hours old and have become less active because the glue-bonding sites are occupied with dust or with moisture from the air; and (2) the wood MC and the EMC aren’t close enough, so the wood changed MC, swelled (or shrank), and now the surfaces are no longer perfectly flat (0.002″ to 0.006″ is the preferred gap between surfaces-no more or no less for maximum joint strength). Links #4 and #5 are the wood surfaces themselves.
There are five stages in forming a good glue bond. The key to several of these stages is proper wood moisture. That’s why metering is important.
After application to the wood face, the adhesive must flow to form a fairly smooth, continuous film. It won’t flow if it’s too thick, the wood’s too hot or too cold, if the wood surface is dirty, or if there isn’t enough adhesive.
The adhesive must transfer to the opposite, mating surface. This requires pressure and sufficient adhesive. Excessive assembly time or wood that is too dry can cause a failure in this stage. Lack of sufficient pressure and pre-curing are also common causes.
The adhesive penetrates the microscopic nooks and crannies of the wood surfaces with the application of pressure. Pressure also pushes the two pieces within the required 0.002″ to 0.006″ gap. Too much pressure squeezes out too much glue which is a weak glue line; too little results in a thick glue line, which is also a weak glue line (except for gap-filling adhesives).
Good glue joints are characterized by good “molecular” bonds between the molecules of the wood and the adhesive. For this bonding to occur, the wood and adhesive must be in intimate contact. However, sometimes the wood surface is contaminated or is chemically unable to be bonded to. This is called a “non-wetting” surface. (Imagine trying to glue two pieces of wood that have oil on their surfaces.)
The final stage is the solidifying of the adhesive. Failure to solidify may be caused by cold temperatures, pH problems, or adhesive/catalyst problems. The rate of solidifying is influenced by the wood’s MC (the drier, the faster) and by temperature (the hotter, the faster). Lumber that is too dry withdraws water before all five stages are completed.
A final note on gluing. There are a considerable number of factors in the gluing operation that affect the strength, durability, and appearance of a glue joint: (1) factors relating to the adhesive itself; (2) factors relating to the handling of the adhesive from its arrival in the glue room until it is applied to wood, pressed, and heated; and (3) factors relating to the wood itself.
Cracked or Checked Finishes
This is a frequent complaint. When today’s finishes dry, they become quite rigid. Any expansion or contraction of the wood underneath causes stress which can lead to cracking and checking of the finish. Correcting this problem requires maintaining an EMC in the plant equal to the in-use EMC, and achieving and consistently maintaining the MC of the wood that is equal to the EMC in-use.
The above information was originally prepared for the report entitled:
10 Ways of Eliminating Wood Problems
Dr. Eugene M. Wengert
Extension Specialist, Wood Processing
The University of Wisconsin-Madison, Department of Forestry