"The Mechanics of Woods," Feature Article, July 2004

the mechanics
of woods

A U.S. lab is trying to design economic incentives for removing some of the tinder from the country's forests.

This article was prepared by staff writers in collaboration with outside contributors.

Analysis tools developed for working with metals and machine assemblies are getting a tryout in the deep woods. The U.S. Department of Agriculture's Forest Products Laboratory is using engineering software to see if a liability in the wild can be turned into a commercially useful commodity.

John F. Hunt, a research mechanical engineer at the lab in Madison, Wis., is applying finite element analysis to the problem of finding economic uses for small-diameter trees and for the slash that loggers leave behind in forests.

There's no commercial appeal in the trees, with trunks less than 4 inches (10 cm) in diameter, or in the trim that loggers leave on the ground after they take timber. But the rejects can be costly. Stands of small trees and accumulations of slash serve as kindling for highly intense forest fires.

The aim of Hunt's research is to give companies a reason to take the small stuff out of the forest.

The waste wood left by commercial logging operations can provide fuel to intensify forest fires.

Through its Forest Service, the U.S. Department of Agriculture manages over 190 million acres of forest and rangeland, much of it wilderness in the 15 western states. Many more millions of acres are the responsibility of the Bureau of Land Management, which is part of the U.S. Department of the Interior; most of that land is in the western states, too.

These states contain 236 million acres of forests (privately and publicly owned). The Forest Service considers 28 million acres to be at high risk. Among the causes for concern are years of drought, decades of fire suppression, bark beetles that have killed tens of millions of trees, historical logging practices, and more recently, the lack of logging.

Hunt's job is to analyze solid wood and pressed-wood-fiber composites to develop new materials and products, such as laminated wood I-beams and three-dimensional engineered fiberboard. He is doing his analyses with Ansys mechanical design simulation and modeling software from Ansys Inc. in Canonsburg, Pa.

The Forest Products Lab has a vast amount of quantitative data on the properties of wood and wood composites. Its testing and evaluation history spans over 75 years. Its work has contributed to the standards that formed the country's building codes.

A 20-year veteran of the lab, Hunt's specialty is performance-engineered composites, and he is working to characterize their mechanical properties and those of wood.

Driving this segment of the Forest Products Laboratory's work is the USDA's National Fire Plan, mandated by Congress after the disastrous fires of spring and summer 2000. One fire forced the temporary evacuation of Los Alamos, N.M., and the national laboratory there. The summer of 2002 was almost as disastrous, with millions more acres burned and thousands of homes destroyed. More destructive fires are expected in the future.

Hunt is working on two National Fire Plan projects. One is looking at commercially low-grade wood, from small or young trees, and trees with curved trunks. Analysis focuses on conventional axisymmetric and orthogonal microwave heat transfer for uniform drying control when using conventional heating and straightening.

The second considers wood fiber in the parts that loggers lop off the timber trees. The lab hopes to transform the waste material into engineered fiberboard.

The U.S. Department of Agriculture is studying products that would turn the waste into a profitable commodity.

The biggest analytical challenge for Hunt is the highly variable nature of wood. To deal with it, he is using parametric methods for modeling the many geometric variables and advanced probabilistic techniques to characterize wood's properties along with conventional deterministic approaches.

Until probabilistic techniques became practical, the use of FEA in solid wood analyses was held back by the variability in fundamental properties data. The probabilistic technique uses ranges of values for wood's properties, and requires more calculations than problems based on more homogeneous materials like steel.

Many factors come to bear, including the age of the wood, difference in yearly growth rates, dry or wet growing conditions, wood fiber's non-linear viscoelastic properties, and the presence of knots and bark. For wood composites, variability can be caused by mixing tree species in the raw material, the proportion of virgin to recycled or processed to non-processed material, moisture content, and amount or type of adhesive.

"This is as much an economic problem as a technological problem," Hunt explained. Few sawmills can profitably use low-grade trees. Moreover, Congress wants the Forest Service to help preserve the economic viability of rural communities, so any technology the lab develops should fit within the means of rural enterprises.

Hunt sees potential markets for forest materials now considered to have little or no value. According to Hunt, the combined markets for wood pallets, office furniture, partitions and fixtures, and doors add up to billions of dollars. "But new uses for these materials won't take hold until there is verification of their performance characteristics," Hunt said.

PROBLEMS AND PROJECTS

The basic analytical challenge is to determine fundamental properties and apply them correctly in computer models. For instance, there is the study of drying, pressing, and straightening wood with microwave ovens.

The Forest Products Laboratory wants to find ways to use low-grade curved trees that can be cut into 2x4 studs by sawing along the grain of the wood. Hunt is conducting research to explore methods to straighten and dry 2x4 stud wood with penetrating microwave heating. Heat and mass transfer effects in lumber have been the initial focus of much of the work by Hunt and his post-doctoral researcher, Hongmei Gu. An expert in wood drying, she is working to accurately model heat transfer effects, while accounting for the factors that make each piece of wood unique.

Mechanical analysis of a beam made of scrap wood allows for variability in natural material.

The factors include ring orientation, ring density, and porosity. Gu describes wood as "an anisotropic, axisymmetric, porous material with complicated cellular and macro scale structure features and material properties." Porosity, the percentage of openings in wood cells, ranges from 10 percent to 90 percent. A further complication is that cells' radial alignments in the tree ring structure may be offset as much as 50 percent.

For the slash-to-fiberboard project, Hunt is analyzing a new class of sandwich construction material made from fiberized wood. Analysis is needed to account for several geometrical variables that affect strength and physical properties.

"We've developed materials from the fiberized treetops that have a 30 percent greater tensile modulus and a 50 percent increase in tensile strength over industrial hardboard standards," Hunt said.

Potential uses include pallets, office furniture, partitions, containers, and construction panels. "After all this computer simulation data has been collated," he said, "we will fabricate panels and test them to establish whether the Ansys model matches the actual results."

One product under study at the Forest Service Laboratory is an engineered fiberboard that would put wood chips and debris to commercial use.

According to Hunt, he had finished much of the conventional analysis—using standard equations to estimate flat crush tests and bending tests—by the end of 2003. "Since the roots of Ansys are in the world of homogenous materials where linear responses predominate, this work is important for validating the finite element analyses, which are mostly nonlinear," he noted.

"With wood products, the envelope of material properties is huge," Hunt pointed out, "and they must be thoroughly quantified before these materials can gain their rightful roles in the forest-products marketplace. Wood's properties probably vary by one or two orders of magnitude more than the properties of metals. And wood's properties also vary by density, knottiness, and tree species: oak versus lodgepole pine, for example.

"We are developing ways to use engineering tools on problems that have, until now, been mostly art," Hunt said.