Effect of weathering of new wood on the subsequent performance of semitransparent stains

Transcription

1 Effect of weathering of new wood on the subsequent performance of semitransparent stains Martin Arnold William C. Feist R. Sam Williams Abstract To quantify the effect of weathering of new wood surfaces on subsequent coating performance, uncoated planed western redcedar beveled siding and roughsawn Douglas-fir plywood panels were exposed outdoors for 1 to 16 weeks. Following this weathering (called preweathering for convenience), the specimens were coated with an oil-based semitransparent stain and placed outdoors again for 5 years. The stained specimens were evaluated annually to determine the effect of preweathering on finish performance. Evaluations based on visual techniques were compared with those using image analysis instrumentation. Stained substrates weathered differently depending on species and surface morphology. The roughsawn Douglas-fir maintained a higher rating than the smooth-planed western redcedar over the 5 years because the rough surface absorbed 2-1/2 times more stain than did the smooth surface. For both types of substrate, the 1 to 16 weeks of preweathering had no effect on the durability of the stain. However, preweathering increased the amount of stain that could be applied to the surface. Although weathering of new wood caused a weaker coating-wood interface, the increased coating absorption of weathered wood compensated for the surface degradation and resulted in similar protection. Therefore, preweathering has a negative effect from the aspect of cost, that is, more stain is required to obtain the same durability. Outdoor weathering of unprotected wood can cause severe surface degradation (6, 12, 13). Wood siding is often exposed to many weeks or months of weathering before being coated with paints, stains, or other finishes (coatings). This weathering before coating (hereafter called preweathering to differentiate it from weathering of the stained wood) can lead to chemical and physical changes on the wood surface that weaken the future coating-wood interface. This interface is crucial for adhesion of film-forming finishes; decreased adhesion and shortened paint service life were shown for wood preweathered for several months before painting (3,4,8,11,14). Prior to our investigations (15,16), nothing had been published describing the consequences of shortterm preweathering (days to weeks) before coating, and previous investigators had not linked film adhesive strength to durability of applied coatings. Our studies showed that short-term outdoor preweathering of wood caused up to a 50 percent drop in the adhesive strength of film-forming finishes. The severity of the preweathering on paint adhesion depended on the type of paint and wood species. Adhesive strength tests with two film-forming paint systems (acrylic latex and alkyd oil primer paints) demonstrated decreased paint adhesion for low density wood species. With higher density wood species, only the acrylic latex primer paint showed lower adhesive strength after 4 to 8 weeks of preweathering. Although paint failures normally occurred preferentially on the latewood of unweathered wood, preweathered substrates showed more frequent adhesive failure for the earlywood rather than the latewood bands. This change The authors are, respectively, Wood Technologist, Swiss Federal Laboratories for Materials Testing and Research (EMPA), Überlandstrasse 129, CH-8600 Düebendorf, Switzerland; and Supervisory Research Chemist and Research Chemist, USDA Forest Serv., Forest Prod. Lab., One Gifford Pinchot Dr., Madison, WI The authors gratefully acknowledge the laboratory assistance of Peter G. Sotos and Mark Knaebe. This paper was received for publication in May Forest Products Research Society Forest Prod. J. 42(3): MARCH 1992

2 Figure 1. Example of test panels, assembled from three individual specimen boards and coated with a semitransparent oil-based stain, after 4 years of outdoor weathering. The main criterion to evaluate stain performance was the amount of erosion (visible as light-colored spots). Above, western redcedar, below, Douglas-fir plywood. in failure mode was attributed to the more rapid degradation of the earlywood than of the latewood during the preweathering. The objectives of this study were to determine effects of short-term preweathering of wood on spreading rates and durability (outdoor exposure) of semitransparent penetrating stains applied to preweathered wood. A secondary objective was to evaluate image analysis techniques for measuring stain durability. Materials and methods Four replicate test panels 406 mm (16 in.) wide by 343 mm (13.5 in.) long were assembled using three individual wood specimens per panel for each preweathering period (Fig. 1). Exposure conditions for this preweathering and test panel assembly were described previously (16). Heartwood from two species was used: freshly planed, vertical-grained western redcedar (Thuja plicata Donn) beveled siding (12 mm (0.5 in.) thick at the bottom), and roughsawn flatgrained Douglas-fir (Pseudotsuga menzieseii (Mirb.) Franco) plywood (12 mm (0.5 in.) thick). The western redcedar specimens were attached to the test panel with the grain oriented horizontally and the plywood specimens with the grain oriented vertically. The freshly prepared uncoated panels were preweathered outdoors, oriented vertically facing south near Madison, Wis., for 1,2,4,8, or 16 weeks following the same staggered time schedule as described previously (16). Preweathering was completed from May through August Four control panels of each substrate were not preweathered before coating but were stored in a dark room at 65 percent relative humidity and 27 C. Following preweathering, the panels were lightly cleaned using a soft brush wetted with distilled water, air-dried, and coated with a semitransparent oil-based stain (1). This brown, linseed oil-based stain had a 76 percent nonvolatile content and a 0.95-g/cm 3 (7.9- lb./gal.) density. The weight of coating applied per panel was measured, and the resulting spreading rates were calculated (Table 1). After drying for 1 week under ambient laboratory conditions, the coated panels were reinstalled vertically on the Madison test fence in a random arrangement facing south. The erosion of the semitransparent stains was evaluated annually using American Society for Testing

3 2 weeks in June, July, or August. The specimens used in this previous work were matched with those used in the present work. Thus, in this analysis, erosion comparisons were done on the basis of exposure time in weeks regardless of the month or amount of sunlight during that time. More distinct seasonal effects might be expected between the summer and winter months because of greater differences in temperature, rainfall, and sunlight. Image analysis Visual rating values for stain erosion correlated with measurements made with the image analyzer. However, image analysis underestimated the amount of stain erosion. Areas on the specimen surface where the stain eroded appeared light colored (Fig. 1). These light-colored areas measured gray in the image analyzer. However, the contrast of the gray values between eroded and noneroded areas was fairly low. Therefore, defining an exact separation between the eroded and the noneroded areas of the specimen surface was difficult. The erosion results using image analysis were no more precise than the conventional visual method; therefore, the additional time and effort in using the image analysis method (cleaning the specimens, taking pictures, and measuring) were not justified. Thus, instrumental analysis confirmed the efficiency and reliability of the visual rating method, even though the visual method was strongly dependent on the judgment of the evaluator. Figure 2. Coating spreading rates of semitransparent oilbased stain on western redcedar and Douglas-fir as a function of preweathering. Each of the four numbers corresponds to a panel and is the mean of the three specimens comprising that panel. (1 ft. 2 /gal. = m 2 /I). and Materials standard methods (ASTM D ) for stain erosion, expressed on a scale of 10 to 1 (10 being perfect condition; 1 being totally failed) (5). In addition to visual analysis of erosion, we used instrumental image analysis methods to evaluate coating performance for the 1988 evaluation. The analysis was done on black and white negatives of the test panels. All photographs were taken under identical indoor conditions using cleaned panels (Fig. 1). Results and discussion Seasonal climate changes In our previous report, seasonal climate changes during the preweathering did not have a significant influence on paint adhesion (16). For example, paint adhesion for those specimens preweathered for 2 weeks in May was not different from those preweathered for Spreading rates Stain spreading rates (coverage of the substrate in square meters per liter (square feet per gallon)) decreased as the absorption of the stain increased (Table 1). The roughsawn, flat-grained Douglas-fir plywood absorbed up to 2-1/2 times as much semitransparent stain as did the smooth, vertical-grained western redcedar siding. Increased stain absorption on roughsawn wood was reported previously (5,7,9,10). These sources also reported increased absorption of penetrating coatings (water-repellent preservatives and semitransparent stains) on wood preweathered for several years. The change in spreading rate for short periods of preweathering (weeks) had not previously been reported. The amount of penetrating stain absorbed by the wood increased with longer weathering times (Table 1; Fig. 2). A Tukey multiple comparison of the spreading rate with the preweathing time showed significant differences for both types of substrate. In Table 1, the different letters (A through D for western redcedar and A through C for Douglas-fir) indicate that the means are different at the 95 percent confidence level (α = 0.05). Increased absorption of stain (decreased spreading rate) was obtained for western redcedar after only 1 week of preweathering. Significant differences were also shown at 8 and 16 weeks. The Douglas-fir spreading rates were significant only for 16 weeks of preweathering. The effect of preweathering at the 1- through 8-week exposures was probably overshadowed by the increased absorption caused by 12 MARCH 1992

4 the roughsawn surface. The spreading rate for Douglas-fir was less than half that of western redcedar for all the preweathering times. Even with the large amount of mechanical damage to the surface by resawing, the relatively short 16-week exposure to sunlight and rain changed the surface enough to cause increased absorption (decreased spreading rate). Other ongoing studies using penetrating water-repellent preservative and oil-based primer paint have also shown increased absorption of these coatings after several weeks of preweathering. These results will be reported in subsequent publications. Semitransparent stain durability The erosion rating of the stain is a measure of its durability. The variability of these erosion ratings was typical of data obtained from other outdoor exposure studies (Fig. 3). The evaluations for other years were similar. The mean erosion ratings of the 12 observations (4 panels of 3 boards each) are summarized in Table 2. The data indicate two trends. The durability of the stain was better on roughsawn Douglas-fir than on smooth western redcedar, and slightly lower ratings were obtained for the panels as the preweathering time increased. Both of these trends were tested for statistical significance. Analysis of variance (ANOVA) of the erosion ratings by year substantiated the difference in stain durability on the two types of substrate (Table 3). This was expected. Improved durability of penetrating wood stains on roughsawn wood is well known (2,5,7). It is primarily a function of greater stain absorption by the roughsawn wood. Although the means in Table 2 indicate a slight deerease in durability with increased preweathering, the trend could not be substantiated using ANOVA (Table 3). The statistical analysis indicated that the weathering rates (the change in erosion rating with time) for all preweathered panels of a substrate type were the same. A linear fit of the data for the various erosion ratings shows similar trends for all preweathering times for a particular substrate type (Fig. 4). The difference in erosion rating between the two different types of substrate is obvious. The stain erosion rates of the western redcedar and Douglas-fir panels showed similar trends regardless of the amount of preweathering. The increased stain absorption ameliorated the effects of the preweathering for both species. As the spreading rate decreased (more stain applied) with increased weathering, the stained surface had greater durability. The results 13

5 tion usually gives greater durability. However, the increased stain absorption by the preweathered wood does not give longer durability. The degradation of the surface is compensated by the increased stain absorption to give the same overall durability but at an increased cost. We expect similar trends with other low-density softwoods. Evaluation of weathering of panels using image analysis techniques showed no better precision than did that of visual techniques and required substantially more time and expense. However, image analysis techniques would have the advantage of computer storage of data for later evaluation. showed that more finish was required to achieve similar weathering performance. This would result in increased costs to the consumer for achieving the same durability or stain lifetime. Conclusions Short-term (1 to 16 weeks) preweathering of both smooth-planed western redcedar siding and roughsawn Douglas-fir plywood degrades the wood surface such that it accepts an increased amount of semitransparent oil-based stain. Increased stain absorp- 14 Printed on Recycled Paper MARCH 1992