Comparisons of interlaboratory swellometer testing of two water-repellent preservative formulations for millwork

Abstract

Water-repellency of preservative formulations used in the millwork industry has long been evaluated by measurement of the dimensional changes in wood treated and then submerged in water according to guidelines published by the millwork industry. Perceptions that this swellometer test was highly variable led to a round-robin test of one solvent-borne and one waterborne commercial formulation by five independent laboratories in order to assess the variation in results among labs. These data were then used to address the sample size needed to achieve a 90 percent confidence level in the mean water-repellent value. The standard deviation for both formulations was near 25 percent of the mean for all labs. The range in average water-repellent values from the labs was approximately 10 percentage units. Not surprisingly, parent boards for swellometer samples were the main source of variation. The data analysis suggested that 12 (for waterborne) or 18 (for solvent-borne) individual test wafers would be needed to provide a 90 percent confidence that an average result from any single lab would be within 5 percentage units of an overall average value from five independent labs using the same sample materials. The new Window and Door Manufacturers Association standard reflects this allowance by establishing the lower limit of water repellency at 55 percent.

Citation

Schmidt, Elmer L.; Murphy, Timothy P.; Cheeks, Charles N.; Ross, Alan S.; Chiu, T. S. (Eugene); Williams, R. Sam. 2002. Comparisons of interlaboratory swellometer testing of two water-repellent preservative formulations for millwork. Forest products journal. Vol. 52, no. 3 (Mar. 2002).:p. 82-84.