The first Spanish visual grading standard, UNE 56525:1972 (19) was published in December 1972 for structural timber. Seven visual grades were defined (Extra/100, I/80, II/70, III/60, IV/50, V/40 and VI) and no restriction of species was applied. Based on British standard CP112 Part 2:1971 (20) Argüelles and Arriaga (21) published a visual grading proposal with four visual grades for sawn timber (75, 65, 50 and 40) and three for glulam lamellas (LA, LB and LC). UNE 56544 was published in 1997, first covering softwood and hardwood species: radiata, Scots and maritime pine, black poplar and southern blue gum (standard names of the species according to EN 13556:2003 (22)) with two visual grades (ME-1, ME-2) using the characterization work performed in the INIA-CIFOR Structural Timber Laboratory (23) Two years later Salzmann pine was also included (24) and afterwards black poplar was excluded. The results from Íñiguez-González et al. (25) made it possible to introduce the new visual grade MEG for large cross-section timber (thickness > 70 mm) in the standard. That same year, a specific standard only for hardwoods was published as UNE 56546 (26), and since 2013 the scope of this standard includes two species southern blue gum (27) and sweet chestnut (28-29), so that UNE 56544:2011 (4) was now only for softwoods. Furthermore, UNE 56547:2018 (30) is a visual grading standard for Scots and Salzmann pine overhead poles.

In 1912 the “Illustrierte schweizerische Handwerker-Zeitung” [Illustrated Swiss artisan newspaper] (31) published information about the southwest German timber trade, showing prices and products in the market for wood for paper making, oak, beech, fir and round wood from Baden, Jura and southwest Germany. Before 1938 timber quality grading and the market was organized in Germany by “die Tegernseer Gebräuche” [Tegernsee’s customs]. This was based on regional commercial practices such as “die Gebräuche südwestdeutschen Holzhandelsverkehr” [southwest German timber trade usages] from 1922. DIN 4074 was first published in 1938, and it established 3 grades for “Kantholz” structural timber (I high strength, II medium strength and III low strength) based on, among others in knottiness parameters. According to Glos and Richter (32) a reviewed version of DIN 4074 was published in 1958 including specific visual criteria for “Brett, Bohle und Latte” [board, plank and batten] different from the criterion for “Kantholz” [square timber]. In 1958 DIN 4074-2 (33) was also published for softwood round timber. Machine strength grading was included in 1989. In 2003 DIN 4074-5 (34) was published for hardwoods. The last version of DIN 4074-1 (10) for coniferous sawn timber was published in June 2012.

Each European country developed its own visual grading standards, adapted to their specific timber species’ particularities. Nowadays national visual grading standards should meet the minimum requirements established by European standard EN 14081:1:2016 (35). Despite this, differences still exist between national standards. The main differences between UNE 56544:2011 (4) and DIN 4074-1:2012 (10) are: (1.) Scope: UNE 56544:2011 (4) is applied to Spanish-sourced radiata, Scots, Salzmann and maritime pine. DIN 4074-1:2012 (10) is applied to European larch, Norway spruce, Scots pine and silver fir from Central, North and Eastern Europe (CNE), and German-sourced Douglas fir. (2.) Visual grades: UNE 56544:2011 (4) established two visual grades (ME-1, ME-2) for small cross-section timber up to 70 mm thickness (b), and only one visual grade (MEG) for large cross-section timber (b>70 mm). DIN 4074-1:2012 (10) established 3 visual grades (S13, S10, S7) for square timber (Kantholz) with b>40 mm and b≤h≤3b and separate requirements for boards and battens. (3.) Knot evaluation: According to UNE 56544:2011 (4) knot size is the width of the knot measured at right angles to the longitudinal axis of the piece, and knot clusters are also considered. According to DIN 4074-1:2012 (10) only single knots are considered in Kantholz, and the smallest diameter of the biggest knot is measured. Furthermore, the smallest knot diameter that should be measured is 10 mm in the UNE standard and 5 mm in the DIN standard. (4.) Admissible fissure maximum depth values are 2/5 (ME-1) and 3/5 (ME-2, MEG) according to the UNE standard, and 2/5 (S13) and 1/2 (S10, S7) of the thickness dimension according to the DIN standard. Furthermore, fissure depth is measured using a 0.2 mm feeler gauge in the UNE standard and 0.1 mm feeler gauge in the DIN standard. (5.) The maximum admissible value of bow distortion is less restrictive according to the UNE standard (10 mm / 2 m) than it is according to the DIN standard (8 mm / 2 m).

Several research works compare visual grading by the UNE and DIN standards. Díez et al. (11) graded 776 50x150x3000 mm³ Spanish-sourced Salzmann pine specimens using both standards. When the results of two Spanish visual grades (ME-1, ME-2) were compared with those of three DIN grades (S13, S10, S7) the yields were considered similar. Conde (12) graded Spanish-sourced Salzmann pine and found higher yields using the DIN standard. However, more pieces were graded in the highest UNE visual grade (ME-1) than was the case in the highest DIN visual grade (S13). Esteban (36) and Arriaga et al. (37) compared the visual grading of Scots and maritime pine pieces from existing structures according to the UNE and DIN standards, concluding that there are significant differences in output and that a new assignment of mechanical properties is necessary to complement the strength class system. Adell et al. (13) graded 201 specimens of German-sourced Scots pine, finding higher yields when using the DIN standard. The method used to measure knots was the main reason why rejection rates were higher using the UNE standard.

In this study 100 large cross-section specimens from four different Spanish-sourced species were graded with three main objectives: (1.) to compare visual grading in wet and dry conditions for the same batch of timber, (2.) to compare the results using two visual grading standards: the Spanish UNE 56544:2011 (4) standard and the German DIN 4074-1:2012 (10) standard, and (3.) to evaluate the accuracy of both standards in estimating mechanical properties according to allocation in EN 1912:2012 (1).

One hundred planed large cross-section specimens were tested. They are from four species (25 each): radiata pine (Pinus radiata D. Don), Scots pine (Pinus sylvestris L.), Salzmann pine (Pinus nigra Arnold ssp. salzmannii (Dunal) Franco) and maritime pine (Pinus pinaster Ait. ssp. mesogeensis Fieschi & Gaussen). They are Spanish-sourced and have nominal dimensions of 3000 mm in length with a 100x150 mm² cross-section. 44% of radiata pine, 68% of Scots pine, 64% of Salzmann pine and 56% of maritime pine pieces contained pith. Wet specimens’ ends were sealed in the sawmill to promote uniform drying.

Visual grading according to standards UNE 56544:2011 (4) and DIN 4074-1:2012 (10) was performed in wet condition (around FSP) at reception in the laboratory. The global Modulus of Elasticity in bending (MOEg) was obtained in a four point bending test at the same time (around FSP) according to EN 408:2010+A1:2012 (38). Air-drying took 116–180 days depending on the species until a final MC of around 10% was obtained. Visual grading according to both standards was performed again in dry condition and a four point bending test was applied until failure (obtaining MOEg and bending strength, MOR). The Modulus of Elasticity parallel to the grain (MOE) was calculated from the MOEg previously adjusted to 12% reference MC, using the expression [1], according to EN 384:2016+A1:2018 (3)

Where: MOE, modulus of elasticity parallel to grain (N mm^-2); MOE_g12%, global modulus of elasticity in bending adjusted to 12% MC (N mm^-2).

Wet specimens were not tested to failure. Therefore, density was determined from the mass and volume of the test specimen and adjusted to the density of a small free-defect prism, by dividing by 1.05 according to clause 5.3.4 of EN 384:2016+A1:2018 (3). Characteristic values (5%) were calculated according to ranking method, as study was done with relatively few pieces, some of the values were obtained from only one piece being more informative than accurate.

Wet MC was estimated by the electrical resistance method according to EN 13183-2:2002 (39). Mean dry MC was determined by the oven dry method after mechanical testing according to standard EN 13183-1:2002 (40), using specimen slices free of knots and resin pockets according to EN 408:2010+A1:2012 (38).

Average wet MCs were 32.2%, 27.5%, 35.8% and 34.4% respectively, for radiata, Scots, Salzmann and maritime pine. Average dry MCs were 9.0%, 10.6%, 9.7% and 9.2%, respectively. figure 1 shows the percentage of pieces in each visual grade and rejected using the Spanish UNE 56544 and German DIN 4074-1 standards.

Figure 1 shows major differences between wet and dry grading. High percentages of rejection in dry condition were found using both standards (from 36% to 84%, depending on the species). The low percentages of rejection in wet condition lead us to think that the parameters that may give rise to a high rate of rejection when dry will be those which are directly linked to MC, such as fissures and distortion. Table 2 shows specimen grades according to visual parameters in dry condition.

The results shown in Table 2 confirm that the high percentages of rejection are mainly due to distortion (mainly by twist parameter). According to EN 14081-1:2016 (35) “dry-graded structural timber shall have at the time it is graded for fissures and distortion, a mean MC of 20% or less with no individual measurement exceeding 24%”. Although the other parameters where measured at higher MC, if fissures and distortion were measured according to the standard it is considered to be dry-graded timber (17). As distortion (mainly twist) has the highest influence on rejection of dry graded timber, results from wet grading (mainly based on knots) are not comparable with the ones expected when timber is dry-graded. However, according to Montón et al. (41) distortion has almost no influence in mechanical properties, but reduces the end-use possibilities because the lack of straightness makes difficult the structural high quality uses. EN 14081-1:2016 (35) also shows that in clause A.2.2 “Even if warp of timber does not directly influence strength, it is strongly recommended that timber for building purposes should be subject to some restrictions in this respect”.

Table 2 shows the dry graded and rejected specimens by several visual parameters. In case of the knot visual parameter, the UNE 56544:2011 (4) standard led to a higher rejection percentages (from 0% to 20%) than the DIN 4074-1:2012 (10) (from 0% to 12%). Furthermore, rejection according to knot parameter using DIN 4074-1:2012 (10) standard was lower than expected. Stapel and van-de-Kuilen (6), using the same standard for Norway spruce grading, reported really low rejection percentages by knot evaluation when timber thickness is higher than 50 mm, which was the case in the current study (100 mm). Regarding fissures and slope of grain, as DIN is more restrictive than UNE in these visual parameters, rejection was higher using DIN than UNE. Concerning distortion visual grading parameters (bow, spring and twist), rejection by bow (from 0% to 20%) and spring (from 0% to 12%) parameters were low and similar in both standards. However, rejection by twist parameter was really high for Scots, Salzmann and maritime pine (from 44% to 80%) and low only for radiata pine (8%). A broader study of the visual grading and mechanical properties of radiata pine revealed that twist is the key singularity for the visual grading yield, mainly in small cross-sections (7). This is in agreement with current results where twist was the key singularity but large cross-section radiata pine were the less affected specimens.

The current twist limit value is the same in both standards (1 mm / 25 mm width) according to the maximum value allowed by the European standards. Nevertheless, European standard EN 14081-1:2016 (35) increased, with respect to the previous version, the maximum permissible twist that can be adopted by national standards from 1 mm to 2 mm (/ 25 mm width) for strength classes above C18. Following the possibility to increase the twist limit value up to 2 mm in future versions of the UNE 56544 standard, Figure 2 shows a large decrease in rejection for radiata pine (from 36% to 12%), Scots pine (from 48% to 16%) and Salzmann pine (from 84% to 60%). There was a moderate decrease in case of maritime pine (from 84% to 80%) and similar rejection decreases were found when defining a 2 mm twist limit value in the DIN 4074-1 (Figure 2). According to Montón et al. (41) studying the high influence of distortion in the rejection rate for radiata pine, moderating twist specification from 1 mm to 2 mm has no important consequence for mechanical properties.

Therefore, if this new maximum permissible twist limit of 2 mm, according to EN 14081-1:2016 (35), is included in future versions of both studied standards, rejection will be reduced without significant reduction of the mechanical properties. However, twist limit definition is exclusive competence of national standard committees and could follow different criteria than yield or mechanical properties. For example, the twist limit in DIN 4074-1:2012 (10) standard is based on consensus between producers and end-users, taking into account that timber with big distortion cannot be used in high-quality structures. In order to avoid timber fulfilling mechanical rejected only by distortion and combine it with industry consensus, 2 mm twist limit could be adopted only in the lowest grade (S7), using for high-quality structures the higher grades (S13 and S10). Furthermore, any change in national standards should be reported to the TG1 of CEN/TC124/WG2 including the fulfilling of EN 1912 allocations with the new criteria.

As it was explained in Materials and Methods, in the bending test the overall MOEg should be adjusted to 12% reference MC. According to clause 5.4.2 of EN 384:2016+A1:2018 (3) MC shall be taken to be 18% when it is higher than 18%, and the value should be adjusted by 1% for each 1% of MC. This means the MOEg obtained in wet condition (around FSP) should be adjusted as if MC were 18% instead of its actual level. In this work, MOEg values were obtained by testing in wet and dry conditions, so that it is possible to compare actual dry values with those adjusted to 12% MC wet values. Table 3 shows these wet MOEg test values together with adjusted MOEg values using different adjustment criteria (from 18% MC, from 30% MC, from original wet MC) and dry MOEg adjusted to 12% MC and the ratio with respect to dry MOEg expressed in percentage (100%).

Table 3 shows that MOEg obtained wet without MC adjustment were from 74% to 83% of MOEg obtained dry and adjusted to 12%. When these wet MOEg were adjusted according to EN 384:2016+A1:2018 (3) MOEg values from 79% to 88% of the dry values were found. However, when they were adjusted from 30% the specimens above this MC value and the others adjusted from their original MC, values from 88% to 98% were found. Finally, adjustment from wet MC reported closer values, which in the case of maritime pine was higher than its dry values being an unsafe adjustment. Adjustment according to the standard looks too conservative, but adjustment from the original MC is risky. However, adjustment taking MC as 30% for MC higher than 30% is accurate and safe.

Analysis of variance (ANOVA) was applied to visual grade batches according to their mechanical properties adjusted to 12% according to EN 384:2016+A1:2018 (3). Figure 3 shows UNE 56544 values for Salzmann pine while Figure 4 for DIN 4074-1 shows values for Scots pine, which is the only one of the four species within the German standard but from different source.

Figure 3 is an example for Salzmann pine. Similar results were found for the other three species studied (radiata, Scots and maritime). Figure 3 shows that although visual grades can discern by MOE wet, MOE values corrected to 12% MC according to EN 384:2016+A1:2018 (3) from 18% to 12% do not fulfill the required values for the allocated strength class according to EN 1912. For example, the mean MOE corrected to 12% for MEG wet Salzmann pine is 7364 N mm^-2, which is less than the 9500 N mm^-2 (10000x0.95) required for C22. Furthermore, if instead of correcting MOE wet from 18% to 12% according to EN 384:2016 (3) it is corrected from the original MC to 12% at 1% change per each MC%, the corrected MOE will be 8967 N mm^-2. Visual grades discern MOE and MOR in dry condition. However, it is not possible to discern density values using visual grades.

Figure 4 shows that visual grades discern according to MOE values wet at 95% confidence level. However, the values of MOE corrected to 12% MC according to EN 384:2016+A1:2018 (3) (from 18% to 12%) only fulfill the required values in the case of S13 for the allocated strength class according to EN 1912. Mean MOE corrected to 12% for S13 wet is 13030 N mm^-2, which is more than the 11400 N mm^-2 required for C30. However, S10 10348 N mm^-2 is less than the 10450 N mm^-2 required for C24, and for S7 7890 N mm^-2 is less than 8550 N mm^-2 for C18. When dry, the MOE of rejected batch is higher than the S7 batch. The same pattern was found in the case of MOR. In case of density grades, density cannot be distinguished according to visual grade.

Although both standards generally more or less differentiate between the MOE and MOR values of visual grades for graded specimens in wet and dry condition, the batch rejected by the DIN standard presented higher values than expected. However, neither of the standards can differentiate between the density of visual grades wet or dry.

Tables 4 and 5 show the values actually obtained and the ones that are required as characteristic values for visual grades according to allocation in the EN 1912:2012 (1) standard, together with the percentages of fulfillment.

Table 4 shows that the batch of radiata pine specimens graded as MEG does not achieve the required C20 MOR value (19.73 < 20). However, the number of graded specimens (16) is too low for further conclusions to be drawn, and the achieved MOR value (19.73) is very close to the one required (20). Scots and Salzmann pine fulfill the required values of C22. The MEG maritime pine batch could be assigned to C22 strength class according to the achieved values. Rejected batches from radiata, Salzmann and maritime pine achieved values lower than those required, as was expected. However, the rejected batch of Scots pine (12 specimens) fulfills the required values of strength class C22. But once again, the number of specimens (12) is too low for further conclusions to be drawn, such as the underestimation reported by Adell et al. (13) for German Scots pine graded using UNE 56544. Furthermore, MEG batch property values are higher than those for the rejected batches. In general UNE 56544:2011 (4) seems to work more or less appropriately in dry conditions for these four Spanish-sourced species, except in the case of Scots pine for which it is underestimated, reducing the yield but in a safe way.

Table 5 shows that the S10 graded batch fulfills the requirements of C30 and could be graded as S13. Furthermore, the rejected batch meets C24 strength class values and could be graded at the same level as S10. Rejected specimens fulfilling the C24 (S10) required values was also found by Stapel and van-de-Kuilen (6) for Norway spruce at thicknesses of 100 mm or more, and could be safely assigned to S10 at thicknesses above 120 mm. In this case DIN 4074-1:2012 (10) does not work properly for grading Spanish Scots pine being more accurate the Spanish standard as was expected. The reason for this could be the high level of variability between sources due to environmental and silvicultural factors, since national standards optimize grading by taking into account growth conditions, typical cross-sections and silviculture factors. DIN 4074-1:2012 (10) was specifically designed for German Scots pine and not for Spanish wood. In any case, it is safe to grade Scots pine from Spain using DIN 4074-1:2012 (10), although a large amount of timber is underestimated.