A summary of an article published on WoodWeb, of an excerpt from the book "The Wood Doctor's Rx" by Gene Wengert, a retired Professor and Extension Specialist in Wood Process, Department of Forestry, University of Wisconsin-Madison.
"… although wood bending is an art, it is solidly backed by engineering theory."
If we bend wood, we want to operate in the plastic range, otherwise in the elastic range the bend would not be permanent. Of course, too much force results in failure, aka "creep".
When wood is steamed (high moisture and high heat), wood will allow for severe bends. The plastic range in compression is extended and becomes extremely large (up to 10x the dry range), while not affecting the tension properties.
Steaming does cause an overall loss in strength after the piece is dried, so steaming time should be kept to a necessary minimum.
A flat piece of wood when bent can normally absorb a difference between outside and inside radius of 2 to 3% (without using end pressure or steaming) before breaking. This means a 1" thick piece, 20" long, can bend on a 24.7" radius. The rule of thumb is that the radius cannot be less than 20 to 30 times the wood's thickness without steaming and end pressure.
Even though wood becomes more plastic in compression, it does not become much more plastic in tension, therefore the outside radius (the tension side) will be the limiting factor in any bend, even after steaming-although after steaming, the severity of the bends may, in certain cases, be increased without failure.
The most effective method of reducing the tension in the outer radius is to push on the ends, generating a compression force that will offset the tension force from the bend. How hard to push depends on the outer radius elongation being less than 3%… in other words, just try to keep the outside radius at nearly a constant length, as if it was flat.
The traditional method of keeping the outer radius a constant length is to use end blocks like in figure 71-3 or a variation of this idea with the distance between the end blocks exactly the length of the piece.
Leading cause of bend failures are:
- poor material selection for bending stock
- miscut lengths (too long or too short between the end blocks
- improper moisture content
There are two ways to cut the end stock so it will fit between the block: Flush and Bevel (figure 71.3). Either two are acceptable. An additional consideration in bending wood is time. It does take a moment or two for the wood to absorb the stress being generated in bending. Bends must be made relatively slowly.
US Forest Service found 17 out of 25 hardwood species suitable for bending. They are, in order of bending quality:
- Hackberry (Best)
- White Oak
- Red Oak
- Chestnut Oak
- Magnolia
- Pecan
- Black walnut
- Hickory
- Beech
- Elm
- Willow
- Birch
- Ash
- Sweetgum
- Soft maple
- Yellow-poplar
- Hard maple (Worst)
of course results may vary from tree to tree, forest to forest.
Density and the rate of growth has no effect on bending failure. End pressure (in excess of that required) was not related to the number of bending failures. (In other words, a piece that is going to fail will fail even if additional end pressure is exerted.) Bending quality was related to certain strength-reducing defects:
- Cross-grain wood is more likely to break than straight grain. The grain should slope no more than 1:15 along the piece. Occasionally, there can be "local cross grain" (e.g. near a knot) that can result in failure. A special form of cross grain is the interlocked grain, which is especially weak.
- Knots are accompanied by distorted grain and represent weak areas.
- Shake is a separation within the wood parallel to the annual rings. This separation encourage shear failures.
- The pith(or exact center of the tree) is very weak and failures there are likely.
- Surface checks, due to improper drying conditions early in drying are small separations in the wood. In bending, they can result in small bumps or slivers.
- Brash wood is probably found in all species and results in very short (along the grain) breaks rather than longer slivers. Wood that is exceptionally light in weight, compared to the average for species is especially prone to brashness. Brashness can, therefore, be a natural phenomenon. On the other hand, brashness can result because of overheating wood, because of decay, or because of earlier failure.
- Decayed wood is weaker than normal wood, even if decay has not progressed very far.