In order to examine meter readings from two major moisture meter types, the Delmhorst Instrument Company and Wagner Electronic Products provided significant support for securing up-to-date meters for the project. A Delmhorst meter, model RDM-2S and a Wagner meter, model L612, were used.
Ten pieces of 8 foot 2
6 rough green Southern pine lumber were donated
by each of nine sawmills in Georgia. Three mills were located in North
Georgia, three in
middle Georgia and three in South Georgia. It was requested that each piece
have at least two feet clear of knots and wane near its center.
The lumber was picked up and brought back to a research lab at the University of Georgia Warnell School of Forest Resources, where it was stacked and stickered for air drying. The initial plan was to determine moisture content readings at three different levels of drying. However, due to reorganization of program priorities at the Forestry School, lab space became unavailable for producing low levels of moisture content, and then limited time allowed examining only one moisture level within the meters' accuracy ranges. This level was the lab equilibrium moisture content which was achieved by drying the lumber in front of a fan until equilibrium was reached.
The two-foot ``clear'' section of each lumber piece was used for moisture determination. It was felt this section should dry as if it were in the middle of at least a six-foot piece of lumber, which required cutting the 8 foot pieces as in the drawing below:
During drying, the two-foot measuring section was held in position by wooden tabs nailed onto the sides of the end sections.
After drying, meter readings were made near each end and in the middle of the measuring section (as indicated on the drawing), and averaged for the piece. Three calibration settings were used for the Delmhorst meter; Southern pine, longleaf pine and shortleaf pine. Five calibration settings were used for the Wagner meter; SYP, loblolly pine, longleaf pine, shortleaf pine and slash pine. The three readings for each calibration setting for each meter resulted in 24 meter readings per measuring section, for a total of 2,160 meter readings in all.
Each measuring section was weighed after all meter readings were made. The sections were then oven-dried at 103C. for 48 hours and weighed again. Several checks were made to assure this drying time was sufficient to achieve near 0% moisture content. (All tests indicated sections had reached 1% moisture content or less.) At this time, section length and the average of three thickness readings were recorded too. All data were saved on a spreadsheet template for further calculations.
Using spreadsheet math functions, for each measuring section, oven-dry moisture content and oven-dry-volume density were calculated. Then, the difference between each meter reading and its equivalent oven-dry moisture contents were figured. Overall average moisture content and wood density were calculated, as well as average, maximum and minimum differences between moisture meter reading and equivalent oven-dry moisture content values.
Applying statistical functions of the spreadsheet, t-tests of differences between meter readings and oven-dry moisture content were determined. Regression equations were also determined for relationships among the various variables.
After all original measurements were completed and calculations made, seven of the 90 measuring sections were found to have reacquired enough moisture to provide valid meter readings. Consequently these pieces were reweighed and examined again with the Delmhorst meter set for Southern pine, and the Wagner meter for SYP. Two final regression equations were developed with these data added.