The Effects Of Surface Roughness On The Performance Of .

ASTM D 662-92 (erosion rate of semitransparent stain) after 600, 1,200,1,800, and 2,400 hours of exposure. Thedegradation modes were rated on a 10 to1 scale where 10 original conditionand 1 total failure.STATISTICALEVALUATIONA two-way analysis of variance wascarried out on all data to determinewhether surface roughness (RC or selected roughness parameter) significantlyinfluenced the experimental data (spreading rate and paint performance). Differences between the means of independentvariables were tested for significance using Tukey’s Studentized Range Test. Significant differences were recorded at the 5percent probability level. All statisticalFORESTPRODUCTSJOURNALcalculations were performed with theSAS software package (19).RESULTSR OUGHNESSAND D I S C U S S I O NEVALUATlONStylus tracing was a suitable methodfor roughness determination. Althoughmainly used for polished and smoothsurfaces, the commercial device wasable to measure the coarse substratesafter some mechanical modifications.This is consistent with recent studieswhere the value of stylus tracing systems for wood surface measurementswas shown (9, 18). The five calculatedroughness parameters selected werecompared statistically in a correlationanalysis, manifesting a high correlationbetween all parameters (Table 4). Thehighest and most homogeneous coeffi-VOL. 45, NO. 7/8cients were found for Ra, representingthe arithmetic mean of the absolute values of the profile deviation. Because itscalculation is standardized and the parameter is used in other studies forroughness characterization (8, 15), Rawas selected in our study to quantifysurface roughness.In Figure 1, Ra mean values and thestandard deviation of all 573 profilesscanned in our study (small boards weretraced once, larger ones twice) are plotted. The standard deviation was lowerfor the smooth samples where thesanded surfaces showed a better homogeneity than the planed substrates.Sanding eliminates deviations better because elements found in the anatomicstructure (such as resin ducts or earlywood/latewood differences) are morepronounced in the roughness profiles ofthe knife-planed surfaces, resulting inthe higher standard deviations.The rougher surfaces were characterized by a much higher variabilitywithin the individual categories. TheTukey procedure proved that the differences between the means, except forRCs 3 and 4, were significant from eachother. Thus, stylus tracing roughnessevaluation allowed the characterizationand splitting up between the surfaces offour roughness grades: RC 1, RC 2, RCs3 and 4, and RC 5. Factory processing(RC 3) and the wood Mizer rougheningwith the normal saw blades in the laboratory (RC 4) both gave similar surfacetopographies.The effects of wood species andgrain orientation on surface topographyin the five RCs are shown in Figure 2,where the mean Ra values for the threesubsets are depicted (CF WRC/flatgrained, CV WRC/vertical grained,and PF SYP/flat grained). To determine the variability within each speciesgroup and roughness class, the standarderror of the mean was calculated. Connetted bars within the roughness groupsmark the means that were not differentat the 95 percent significance level. Forthe planed surfaces (RC 1), all meanswere significantly different from eachother with the highest value for the flatsawn WRC and the lowest roughnessfor the SYP. The flat-grained WRC wasa low-quality grade and the factoryplaning resulted in uneven and irregularsurfaces caused by the poor wood quality and worn or dull knives. Addition93

against the cutting forces of the band-saw with normal set (RC 4) and is lesscompressed, so that the cutting qualityis more homogeneous. When the set ofthe blades is increased, as with the second bandsaw, more fibers are tom offthe earlywood tissue, whereas the latewood zones are sheared off, resulting ina high roughness variability of the RC5-SYP surfaces.R OUGHNESSSPREADINGFigure 1. — Mean values and standard deviations of average roughness (Ra)(expressed in micrometers) within the five roughness categories. The numbersquote the sample size. All means, except for roughness categories 3 and 4, weredifferent from each other.Figure 2. --- Mean values of average roughness (Ra) and the standard error ofmean for the species/grain subgroup within the five roughness categories. Barsconnected by lines beneath the graph within each roughness group marks meanvalues not different from each other at a significance level of 5 percent.ally, the differences in earlywood/latewood density caused higher swelling inthe tangentially cut latewood, so thatsome of the profiles of flatsawn woodlooked like densitograms. The fact thatthe SYP surfaces had the lowest roughness was due to the fresh planing in thelaboratory right before roughness wasmeasured.No significant differences betweenthe substrates were found within the RC2 and RC 3 and 4 categories (for the94WRC specimen). Obviously, sanding aswell as factory rough sawing reducedthe influences of grain and wood speties. Within the WRC group, the lowerstandard deviation of the vertical grainsubstrates shows that a homogeneouswood quality results in a more consistent surface quality even after sandingand planing. Compared to WRC, roughSYP showed significantly lower Rameans. It can be assumed that the higherdensity SYP builds up more resistanceANDRATERa and spreading rate numbers of the66 exposure panels subjected to accelerated weathering were compared in acorrelation analysis. The results manifest the inverse relationship betweensurface roughness and spreading ratesreported in previous studies (12,23).An inverse relationship within eachof the three clusters presented in Figure3 exists between the roughness and thespreading rate with correlation coefficients of r -0.79 (smooth surfaces,normal application), r -0.78 (roughsurfaces, normal application), and r -0.42 (rough surfaces. reduced application). The slope of the regression linesdepicts that within the samples paintedwith normal spread, the relationshipwas stronger for the smoother surfacescompared to the rougher samples.As was intended by the study design,the reduced spreading rate on therougher surfaces generated significantlydifferent results from the normalspreading rates on similar rough samples. The spreading rates ranged in thesame order as the coverages for thesmooth surfaces. The slope of the regression line is nearly the same as forthe surfaces painted with normal spread.Tukey groupings based on all datapoints showed there was no significantinfluence on spreading rate either bywood species (one exception) and grainorientation.Figure 3 also allows the visualization of variation within the RCs. It washighest for the sanded specimen withspreading rates between 440 and 250ft.2/gal. In this group, the SYP specimens needed significantly more paintthan the WRC samples, 373 versus 268ft.2/gal., although no significant differences in the roughness of the twosanded species group had been found(Fig. 2).The sanding process seems to revealthe best possible individual characJULY/AUGUST1995

teristics of the substrates as they refer tothe ability to accept coatings. To a lesserextent, the same is true for the planedsurfaces, where the spreading rates forthe RC 1 surfaces were grouped between 490 and 390 ft.2/gal. With increasing roughness, the deviation in thespreading rates became smaller. Obviously, anatomic structural elements important for the spreading rate (e.g., celldimensions and earlywood/latewooddistribution in growth rings) are overlapped by roughness effects (e.g., increased surface area caused by valleys,crevices, and upstanding fiber bundles).A CCELERATEDWEATHERINGThe results of stain erosion and discoloration evaluation after four 600hour periods of accelerated weatheringare shown in Table 3. The numbers represent the averages of three replicates.Afler 2,400 hours, only small differences between erosion and discoloration were found. In Figure 4, the resultsfor vertical grain WRC erosion aregraphed as an example to visualize thestain performance during the test on thedifferent RCs.All boards finished with a reducedspreading rate (SR 2) showed significantly lower ratings in both stain erosion and discoloration as compared tothe normally finished boards. The onlyexception was in the SYP specimenswith reduced coverage, where discoloration had increased during the lastexposure period and was rated similar tothe normally finished boards. The failure in erosion appeared after the firstexposure cycle and lasted until the endof the test, when most of the stain waseroded completely (Table 3).Within the specimens with a minimum finish spread, no clear influence ofthe roughness group was visible. Thisbehavior clearly demonstrated that thegood performance of stain on roughsawn surfaces seen in previous studies(5-7) is primarily due to the higherquantity of finish spread on the roughersurfaces and not an effect of the increased surface roughness. The rapiderosion indicates that rough wood, ifpainted insufficiently, might degradefaster because the moisture uptake andretention of roughened surfaces arehigher, as are swelling and shrinkagemovements in the loosened fibers.On normally finished wood, thesanded surfaces (RC 2) showed similarFOREST PRODUCTS JOURNAL(SYP) or better performance thanroughsawn specimens. Sanded verticalgrain WRC had superior ratings (7.0 forboth erosion and discoloration), but thedifferences in the low-quality flatgrained WRC were surprisingly small(6.3 and 6.0, respectively).SYP, a species with poorer paintholding characteristics, was rated 5.0 inboth erosion and discoloration for RC 2,which is the same rating as for the RC 5surfaces. This result is particularly interesting because spreading rates of thesanded boards were 50 to 60 percenthigher than those of the bandsawn samples. This indicates it is feasible to reachan equal (SYP) or improved long-termperformance (WRC) with less than halfof the quantity of stain used on roughwood, which was thought to be the bestsurface texture until now. Sandingwould make stain application more efficient, not only economically (less material use), but also from an ecologicalstandpoint (fewer volatile organicchemical emissions).The reason for the improved stainperformance on sanded boards can beseen in the substrate conditions realizedby a slight sanding. A microscopicevaluation of all surface grades showedthat the stain uniformly covered thesanded specimens (Fig. 5). It has beenreported that on a microscopic level,sanding or abrasive planing causes surface and subsurface damage when com-Figure 3. — Spreading rates of stain in relation to surface roughness.Figure 4. — Erosion rating of vertically grained, stained, western redcedar afterfour 600-hour periods of accelerated weathering. The numbers identify the corre-paspending spreading rates.VOL. 45, NO. 7/895

red to knife planing (4,14,21). Surfacedamage (broken or crushed cell walls) isthe reason for a poor bonding quality inshear and adhesion tests. ,According to our results, this phenomenon proved to be beneficial forfinish performance. We assume that theupstanding and lightly crushed earlywood fibers created an increased superficial surface and exerted adsorptionforces that prevented the stain frompenetrating in deeper wood zones. Further, normal anatomical penetrationpaths (e.g., wood ray tissue) had beencrushed so that an uptake of stain indeeper cell layers was prevented. It isknown that sanding alters the cellularstructure so that no anatomical roughness (first-degree roughness) is detectable (22). On the other hand, the denselatewood bands were roughened sufficiently so that an increased finish adsorption was possible in these zones,too.Possibly, the lightly loosened andupstanding cell wall material provided amechanical reinforcement for the finishlayer when covered and immersed bythe stain. The finish was concentrated atthe sanded surface layer where it nearlyforms a uniform film and gives protection against radiation and water Thissupports earlier results, which reporteda complete masking of grain in paintedabrasive-planed samples after accelerated weathering when compared toknife-planed samples (10). Accordingto Gaby’s study, knife planers or matchers may cause physical surface changes,especially on flatsawn southern pine,which subsequently can affect paintperformance, whereas abrasive planersdo not compress or burn the surface inthe way that may adversely affect paintand stain life,The reinforcement also appears onthe bandsawn surfaces, but the higherroughness allows more penetration inthe low-density earlywood cells, whichstabilized its weak thin cell walls, butresulted in higher finish uptakes. Theselow-density earlywood zones becamethe starting points for the severe failureon planed flatsawn surfaces because thesurfaces were not covered and protectedsufficiently with finish substance.It will be interesting to evaluate theboards installed at the exposure site inMadison, Wis., to see if the positiveinteraction of sanded surface and finishcharacteristics is limited to the stainedsamples or if painted boards will show asimilar performance. Because of the different viscosities and wetting characteristics of stain and paints, it might bereasonable that another roughness grademay perform better. Conversely, allacrylic paint systems showed the bestperformance on both roughsawn andscratch-sanded boards after 7 years inoutdoor exposure at the USDA ForestService’s test fence (5), so that anequally good behavior might be expected from painted specimens. However, the approach outlined in this studyshould be replicated in additional studies using other grit sizes and sandingprocesses to test their influence on finish performance.CFigure 5.— Microphotograph of a flat-grained, sanded, western redcedar specimens (top) and a flat-grained, planed, southern yellow pine specimens (bottom)after 2,400 hours of accelerated weathering. The darker areas on the left werecovered in the weather-o-meter and allow a comparison to the exposed surfaces.96ONCLUSIONSSurface roughness produced bysanding or sawing can affect the performance of finishes in several ways. Ithas been quantified that finish spreadingrates and surface roughness are relatedand that rough surface substrates needmore finish coverage per area thansmooth surface substrates. Very roughstained wood performed well in longterm exposure mainly because of theJULY/AUGU5T 1995

high spread necessary to cover the surface, whereas rough wood finished withthe same amount of stain as applied tosmooth surfaces (limited amounts)failed completely after only short-timeweathering. The best stain performancewas found on sanded surfaces, with performance ratings even better than thosefound for very rough wood, but withless than half the amount of finish applied.In contrast to what is concluded innumerous studies on adhesive performance, sanding seems to be a perfect surface preparation for coatings because itlevels off inherent differences in woodsurface properties resulting in an equaland homogeneous finish spread. Thelight roughening allows sufficient finishpenetration in dense latewood zonesand avoids an over penetration in earlywood tissue, areas that are usually severe failure zones on planed wood. Astabilization effect and reinforcementfor the finish layer by upstanding fibersand cell wall material is conceivable.Surface sanding has proved to be anadvantageous processing step prior topaint application. Sanded surfacesneeded a relatively low quantity of paintfor coverage and showed best paint performances even on low-grade wood,which can improve the furture competitiveness of wood siding.L ITERATUREFORESTPRODUCTSCITEDJOURNALV OL . 45, No. 7/897

Three standardized (DIN 4768/ISO 4287) roughness parameters: average roughness (Ra), average roughness depth (Rz), and maximum roughness depth (Rt); and two derived numbers: peak roughness (Pr) and peak index (Pi) were calculated and compared statisti-cally. Ra is the arithmet