Visual grading: features, national rules and their limits

How a visual grader works

In visual grading a trained person inspects each piece and assigns a grade by the size and position of defects. Typical features checked (Finnish and Polish sources) are knots, grain deviation, cracks, resin and bark pockets, rot, insect tunnels, wane, curvature, twist and mechanical damage, and optionally growth ring width. Knots matter because they cause stress concentrations and deflect the grain; sloping grain causes tension perpendicular to the grain.

Visual grading is a rule-based decision on what can be seen from the surface; its predictors correlate only moderately with strength (lesson 4).

An example of limits: DIN 4074-1 (Germany)

The portal reproduces limits for softwood squared timber from an Austrian school textbook excerpt that cites DIN 4074-1 without an edition year. Boards and planks are graded on edge, and bow and twist values were not usable. Treat the table as an illustration of how a rule is built.

Higher grades have tighter limits on every strength-reducing feature. A piece gets the highest grade whose limits it meets on all features; one failing feature downgrades the whole piece.

Worked example: grading a squared timber

A squared softwood piece shows: knots occupying 1/4 of the cross-section, slope of grain 10%, average ring width 5 mm, wane 1/5, no rot or shake.

• S13: knots limit 1/5 = 0.20, the piece has 0.25. Fails (also slope 10% is above 7% and ring width 5 mm is above 4 mm).
• S10: knots 0.25 is within 2/5 = 0.40; slope 10% is within 12%; ring width 5 mm is within 6 mm; wane 1/5 is within 1/3. Passes.

The piece is S10. S7 also accepts it; in this exercise we assign the highest grade for which every limit is met.

Other national rules (Europe)

• INSTA 142 (Nordic countries): classes T0, T1, T2 and T3, approved as corresponding to EN 338 classes C14, C18, C24 and C30. Stapel notes that it is the only standard in his comparison that uses absolute knot sizes in millimetres rather than ratios. The numerical knot limits were not found in an open source and are not given here.
• PN-D-94021 (Poland): grades KW, KS and KG plus reject. The only limits the portal confirmed are for average ring width: at most 4 mm for KW, 6 mm for KS and 10 mm for KG. Research papers treat C24 as equivalent to KS and C18 as KG for comparison, a research convention. One paper states that the Polish grades were not included in EN 1912; whether this changed with the 2024 edition was not checked.

Rate of growth is a weak proxy

Ring width is a visual stand-in for density and stiffness. In a Polish Scots pine study (210 boards, Silesia), the coefficient of determination between ring width and bending strength was only 0.18 for the whole batch. Log quality class said more: mean ring widths were similar in classes A, B and C (1.88, 1.73 and 1.92 mm), but mean bending strength was 61, 54 and 38 MPa. This is one species from one region; do not generalise it to spruce.

What 0.18 means

A coefficient of determination of 0.18 means that ring width accounts for 18% of the variation in strength; the remaining 1 - 0.18 = 0.82, or 82%, is explained by other things.

Do rules deliver what they declare?

A Holzforschung paper (abstract only) concluded that, depending on origin, species, cross-section and rule, yields or mechanical properties can be low compared with declared values. The practical lesson: a visual rule calibrated in one region does not automatically transfer to timber from elsewhere.

Key points

• Visual grading assigns grades from visible features against written limits; a piece takes the highest grade whose limits it meets on every feature.
• DIN 4074-1 (as excerpted), INSTA 142 and PN-D-94021 are national rules mapped to classes; the mapping of national grades to EN 338 classes belongs to EN 1912.
• Ring width alone is a weak predictor of strength: R2 of 0.18 in the Polish pine study (one species, one region).
• Visual rules calibrated in one region may not deliver the declared properties for timber from another.