INTRODUCTION

For more than two decades researchers in the Department of Forestry, Wildlife and Fisheries at the University of Tennessee have utilized a radiation-based laboratory instrument for making mass-based density measurements of wood and wood products. Wood density is an integral variable of interest in forestry and wood products research. The development and evolution of this particular instrument has played a key role in influencing the direction of several research programs in the department by providing a unique approach to measuring density of wood and wood products.

Gamma densitometry is an example of long-term development and utilization of a technique by researchers in the Department of Forestry, Wildlife and Fisheries that is unique to UT. Only a few forestry or wood products programs in this country possess densitometry capabilities. This review article is intended as a benchmark to describe the more-than-two decade-long gamma densitometry work completed at UT.

BACKGROUND

Density is one of the most important properties used to characterize and classify wood and wood products. Defined as weight per unit volume, wood density is a relative measure of the amount of cell wall material per unit volume. Commercially important domestic tree species characterized as being high density include the oaks (Quercus spp.), hickories (Carya spp.), and hard pines (Pinus spp.) to name a few. Low density wood is typical of spruces (Picea spp.), basswood (Tilia Americana) and the soft pines (Pinus spp.), but includes many other deciduous and coniferous species. The terms softwood and hardwood are misnomers as the wood of some conifer species is of greater density than the wood of some deciduous species.

Wood density is the single most important property of the wood related to all strength properties. Density of the wood is also related to nail and screw holding ability, machinability, gluability, and heat value when burned. Wood density is an important property of the wood when converting wood into remanufactured wood products such as paper, plywood, particleboard, fiberboard and flakeboard.

The determination of density usually involves an independent determination of volume and weight and is commonly referred to as gravimetric determination. Volume might be determined from linear measurements or more commonly from displacement by submersion in water or other fluids. Traditionally, this method of determination has been destructive in that a sample is removed from a tree, a board, or other type of wood product that is being evaluated. Destructive determination precludes further analysis of the particular sample.

RADIATION DENSITOMETRY

Radiation densitometry, the determination of density by radiation absorption, has been at the forefront of alternative methods of determining wood density since the 1940s. Much of the early development of radiation densitometry, as applied to wood, was done by Polge (1978, 1970) using x-ray techniques. Cameron and others (1959) developed a system for using beta rays on thin wood samples. Echols (1970, 1971) helped develop the x-ray method of density determination. Further developmental work with the x-ray method was done by Ellis (1971) and Fletcher and Huges (1970).

Radiation densitometry, as developed by these researchers, consisted of using an x-ray or beta-ray source and photographic film to measure the transmitted radiation; much like a medical x-ray taken today. Limitations abounded; the source strength had to be matched to the wood thickness. If the source strength was too high, little radiation absorption occurred. If the source strength was too low for the sample thickness, the photographic film did not respond linearly. The mass absorption coefficient (the term necessary to correlate radiation absorption and density) varies with source energy, making these source energy changes difficult to calibrate. In addition, the photodensity of the developed film had to be measured with a photodensitometer, a device that was expensive and difficult to calibrate. These limitations, coupled with unique characteristics of wood samples such as removing and preparing increment cores, were serious obstacles to early adaptation of the technique.

The late Dr. Frank Woods initiated the technique of radiation densitometry at the University of Tennessee in the late 1960s. In addition to recognizing the immense potential of applying the technique of radiation densitometry to silvicultural work, Dr. Woods was familiar with modern radiation sources and counting techniques (Woods et al., 1965). Woods developed the first radiation densitometer using a collimated beam from a low energy gamma source coupled to digital radiation counting equipment to eliminate the photographic film step. The gamma source (Fe⁵⁵) provided a constant low energy radiation source at a narrow energy width; the electronic counting equipment provided instantaneous reading of the density data. The result was a machine that could rapidly scan a strip of wood prepared from an increment core and determine density on a scale of several data points per mm. As a result tree rings could be identified and measured directly from the density data, thus giving both growth rate data and wood density data. Unfortunately, modern equipment at this time included a punch-tape teletype to record data and the university mainframe computer to analyze the data (when the paper tape could actually be read!).

In the early 1980s the original gamma densitometer was updated by the addition of a digital stepping motor drive, a precise ball-bearing carriage and drive screw, and a personal computer. The carriage and drive mechanisms allowed precise movement of the sample in variable increments and the computer allowed for data acquisition at a much greater rate than the teletype. This improved version of the gamma densitometer is still being used in almost the same form today (Figure 1). An Americium²⁴¹ source was added to allow density determination of thick samples. Work has been done to improve both the technique and the calibration of the instrument (Moschler and Woods, 1975; Moschler and Dougal, 1988 and; Moschler and Winistorfer, 1990).

RADIATION DENSITOMETRY APPLICATIONS IN FORESTRY RESEARCH AT UT

One of the first studies completed with the gamma densitometer was a comprehensive effort to study the effects of pollution on tree growth and wood density (Lawhon 1973; Woods and Lawhon, 1974). This study involved growth and density determinations for a number of individual years on hundreds of increment cores. This work pioneered both the "gamma densitometer" and the concept of determining wood density as well a growth rate when studying environmental influence on forests.

Woods and Hughes (Hughes 1985) related wood density and growth rate of three oak species before and after the start-up of a coal fired electrical generator near the trees sampled. Woods did additional environmentally related research, one study involving the effect of waste effluent on the growth and density of a loblolly pine plantation near Chattanooga (literate cite?).

A number of unpublished, service projects were initiated and completed using the gamma densitometer. These included the application to dendrochronology (barn timber dating), to detect decay in failed wood components (axe handles, ladder rails, and roof trusses), to identify snow loading in pine plantations, to identify thin egg shells, and to aid in identification of charcoal found on archeological sites.

A second, major area of application of the gamma densitometer at UT has been in silviculture studies. A study by Ross utilized the instrument to determine growth and wood density response to irrigation and fertilization (Ross 1975; Ross et al. 1979). McRae studied the effects of different levels of thinning on density and growth rate of loblolly pine (McRae 1981). McRae found variations in density within growth rings between thinned and unthinned stands (Moschler et al. 1989). During the course of this work McRae developed improved programming to count and classify growth rings from the density data.

In the 1980s there was increased cutting of plantation grown southern pine. This shift in raw material was brought about more quickly than previously expected by a number of factors, including the limits on cutting imposed in the western US, perhaps to a certain extent the "greening" of America, and the rapid rate of development of demand for composite board products, and an increase in location of forest based industries in the southeastern US.. This influx into the manufacturing system of young trees characterized by rapid growth and increased amounts of less dense juvenile wood prompted interest in density surveys by both private companies and government research units. A number of companies are now using some form of the type of radiation densitometry that we helped developed at UT as an evaluation tool for their growing plantations.

In the late 1970's and early 1980's improvements were made in the development and marketing of X-ray tubes and power supplies. In 1990 we added a 15 keV X-ray source and built the first direct reading X-ray densitometer. This change dramatically improved the speed and accuracy of densitometry methods. Woods and Moschler completed a density measurement contract using this new device for the U.S. Forest Service in 1991 (Woods 1991), shortly before his retirement from the University of Tennessee. Dr. Wood's retirement effectively ended the wood anatomy application of densitometry at UT.

A third, major area of application of the gamma densitometer at UT began in the 1980s. Dr. Winistorfer adapted the densitometer to produce vertical density profiles of composite wood panel products (Winistorfer et al. 1986). A vertical density profile (VDP) depicts the change in panel density through the thickness of the panel (Figure 2). The VDP is an unavoidable byproduct of hot-pressing a loose mat of wood particles into a flat, rigid panel and results from the complex phenomena of heat and mass transfer during pressing. This nondestructive evaluation of density of flat-pressed wood panels has become an extremely important measure to the wood industry. This measure of density variation through the panel thickness reflects on manufacturing quality, and is related to panel strength, edge machinability and surface characteristics important for overlaying with plastic laminates and thin wood veneers. In recent years several commercial manufacturers have developed gamma densitometers dedicated to density profile measurement. For a limited period of time UT contracted with several wood products manufacturers in order to provide this important density profile service. The technique proved so valuable and the demand so high that several manufacturers now produce and distribute density profiling instruments and services.

A major advantage of gamma densitometry for density profile analysis is that it is a nondestructive measure and other physical tests can be conducted on the same sample. This technique was used under contract with the Consumer Product Safety Commission to help characterize medium density fiberboard used as chainsaw kickback test material (Winistorfer and Moschler, 1987, 1989, 1989a).

Winistorfer and Davis completed several studies in which this densitometry technique was pivotal to examining manufacturing process variable effects on thickness swell of oriented strandboard (Davis 1989). Winistorfer and McFarland expanded on Davis' work and developed a technique to produce 3-dimensional surface response mapping of density variation for entire laboratory-made composite panels (Figure 3) (McFarland 1992; Winistorfer and McFarland 1992). This technique provides needed information with regard to the development and movement of density within a panel during pressing.

This early work at UT is partially responsible for a surge of interest in using the VDP as a measure of composite board quality. Several companies began marketing gamma based density profiling devices to the board industry. Recent developments in the manufacture and marketing of X-ray tubes have lead to the development of a new generation of density profilers, similar to the one first developed at UT. These profilers are based on the same principle of direct measurement of the transmitted radiation, but with greatly increased accuracy and speed of measurement because of the much higher radiation flux generated by the X-ray tube. Today the "density profiler" is an important tool in most medium density fiberboard plants quality control room and is gaining usage in the particleboard, oriented strandboard, and wheat straw board industries.

The measurement and control of the vertical density profile during pressing of composite wood panels is a promising future application of gamma densitometry. Winistorfer, DePaula and Bledsoe and fabricated a device to move multiple sources and detectors in concert with the closing of an experimental laboratory hot-press (Figure 4) (DePaula 1992).

Winistorfer and Moschler designed and installed a three source density profiler system on the carriage built by Bledsoe and DePaula. This system consists of 3 collimated Cs¹³⁷ sources, associated shielding and safety interlocks, 3 scintillation radiation detectors, and a PC based density monitoring and recording system. In addition to the density determination, the system measures and records 16 channels of temperature data and press position and pressure.

Initial laboratory measurements made at the J. T. Mengel Forest Products Laboratory at UT are the first ever recorded measurements of density during pressing of a composite wood panel. These initial measurements were central to obtaining recent funding from the Cooperative State Research Service (USDA CSRS National Research Initiative Grant) for a two-year research project aimed at providing real-time measures of density during pressing. The National Particleboard Association has also recently provided funding for this research project (Winistorfer 1992).