What graders should know about wood's mechanical properties: a 2023 review of testing, grading and NDT

Introduction

Note: only part of the original text could be read for this summary; sections not covered are marked as such below.

Anyone who grades, sells or specifies sawn timber relies on a body of knowledge about how wood behaves mechanically: why two boards of the same species differ in strength, why a knot matters more at the edge than in the middle, and why a test result on a small clean specimen cannot simply be applied to a 150 mm structural joist. In 2023 the journal Forests published a long open-access review, "Mechanical Properties of Wood", by researchers from the Universidad Politecnica de Madrid, the USDA Forest Products Laboratory and ETH Zurich, that pulls this knowledge together for engineers and timber users.

The review is aimed at construction, where wood competes with steel, concrete, brick and stone, and it starts with the argument that wood's mechanical behaviour must be understood on its own terms because the material is orthotropic, meaning it has different properties along different directions. It then moves through the history of testing, the difference between small clear wood and structural timber, visual and machine grading, non-destructive testing, and the factors that modify strength and stiffness. This article summarises the review for a sawmill audience, with an important disclosure: the later parts of the review (the sections on load duration, moisture, temperature, size effect and the conclusions) could not be read in the text available for this summary, so those topics are described only at the level of the abstract and figure captions.

Research methodology

This is a narrative literature review, not an experiment. The authors state that it draws mainly on published articles and on European and North American standards, and that it includes historical background and current trends. There is no systematic search protocol, inclusion criteria or meta-analysis described in the portions read. The authors are Francisco Arriaga, Xiping Wang, Guillermo Iniguez-Gonzalez, Daniel F. Llana, Miguel Esteban and Peter Niemz. The paper was received on 11 April 2023 and published on 9 June 2023 in a special issue on the structure and physical and mechanical properties of wood.

The structure follows an engineer's logic. After an introduction on wood as a sustainable building material, the review treats the orthotropic structure of wood and its elastic law, then (section 3) the determination of mechanical properties: early testing history and standardisation, small clear wood versus structural timber, the effect of quality and defects, visual and machine strength grading, and non-destructive testing. A final group of sections addresses influencing factors: duration of load, moisture content, temperature and size. Keywords listed include duration of load, non-destructive testing, orthotropy, size effect, standardisation and timber.

Key findings

Why wood matters in construction. The introduction frames wood as a modern, sustainable and efficient material. It states, for example, that the EU-28 produces about 110 million cubic metres of sawn timber a year, that a cubic metre of wood weighs roughly 600 kg and fixes about 1.5 tonnes of CO2, and it works through a seven-storey timber building in Vaxjo, Sweden, to illustrate resource demand (about 28 cubic metres of wood per 125 square metre apartment). It also cites a product environmental footprint study of a Madrid-area house built with cross-laminated timber and glulam in which foundations and the concrete structure caused 37 percent of the impacts against 9 percent for the timber structure. Cross-laminated timber is described as the main driver of growth, and the review mentions that timber construction can cost about 10 percent more while giving about 7 percent more usable area and cutting 2 to 3 months from medium-rise construction time. These are the authors' figures from the literature they cite.

Structure drives properties. At the small-specimen level, bulk density is the dominant variable, and the review also stresses the influence of microfibril angle: roughly 0 to 5 degrees in spruce versus 15 to 20 degrees in yew, a difference that strongly affects stiffness, strength, swelling and shrinkage. The review shows a density-versus-bending-strength relationship for 103 species.

Orthotropic behaviour. In tension the proportional limit is about 50 to 60 percent of the maximum stress under normal climate, and ultimate tensile strain both parallel and perpendicular to the grain is low (0.7 to 1 percent). For compression perpendicular to the grain, strength is conventionally defined at 2 or 5 percent strain, and radial compression can densify spruce from about 450 to 1200 kg per cubic metre. The review then sets out the generalised Hooke's law for the longitudinal, radial and tangential axes.

Testing and standardisation. Early scientists tested timber in the 17th and 18th centuries; the review notes, for example, that Buffon's team tested more than 1,000 green oak specimens and concluded that small specimens were too variable to predict the strength of large members. National standards bodies appeared from 1901 onward, with ISO following in 1947 and CEN in 1961. Small clear specimens are typically 20 to 50 mm in cross-section, and bending tests use a span of 14 times the depth with a centre load (ISO 13061, ASTM D143). Structural-size bending (ISO 8375, EN 408, ASTM D198) uses a span of 18 times the depth with two loads 6 times the depth apart. EN 408 takes the 5th percentile of bending strength and density and the mean modulus of elasticity (MOE) as main properties. Results from different specimen sizes are not directly comparable, and sawing can reduce load capacity by up to five times relative to round wood, while the strongest ungraded piece may be up to ten times stronger than the weakest.

Knots and grading. Knots reduce strength mainly by deflecting the grain around them; edge knots and knots in the tension zone are most harmful, and clusters can be worse than single knots. For European spruce the review reports correlations with bending, tension and compression strength of about 0.4 to 0.6 for knots, 0.1 to 0.2 for slope of grain, about 0.5 to 0.6 for density and 0.7 to 0.8 for MOE, and states that knots combined with MOE exceed 0.8. It also reports that knots explained about 57 percent of strength variation in pine but only about 26 percent in spruce, so rules should be species-specific. Visual grading rests on the assumption that growth features reduce strength relative to clear wood; the review dates the first formal rules to Sweden in 1764 and to Maine in the 1830s, and cites EN 14081-1 and ISO 9709 for basic principles. Machine strength grading uses non-destructive measurements, mainly density and dynamic MOE, with a visual override, and it can be controlled either through output control (daily production tests adjust machine settings) or machine control (settings from approved grading reports). Grading machines date from the 1960s, and the first Swedish machine was approved in 1974.

Non-destructive testing. The review describes static bending MOE (an "apparent" MOE because shear deflection is included), machine stress rating that senses the lowest flatwise MOE over spans of about 1.2 m in most US machines, and stress-wave methods based on the relationship that wave velocity equals the square root of MOE divided by density. Wave velocity can be measured by time of flight between two sensors or by resonance, where the fundamental frequency of a member is found by Fourier analysis and velocity equals twice the frequency times the length. The resonance method excites many reverberations and gives accurate, repeatable velocity, and it is used in industrial log sorting and lumber machine grading, with portable tools such as a log stiffness tester named as examples.

Influencing factors. According to the abstract and the figure captions, the review closes by covering duration of load (creep and relaxation, with load-duration curves attributed to Madison, Wood and Madsen), moisture content (hygroscopic isotherms and the effect of moisture on tension, bending and compression), temperature (MOE, bending and compression strength relative to 20 degrees C at two moisture levels) and size (length and depth effects, and a depth factor). The numerical content and conclusions of these sections were not available to the writer of this summary and are therefore not reported.

Practical applications for sawmills

From the review: The material directly supports several familiar mill practices. Strength grading is more reliable when stiffness and density are measured than when only visible features are assessed, because MOE correlates more strongly with strength than knots or slope of grain do. Grading rules should be species-specific. Resonance-based acoustic tools are established for sorting logs and grading lumber. Test results depend on specimen size and test configuration, so declared properties must be tied to the standard and the size used.

Portal interpretation (not claims of the review):

• Mills that grade visually might consider whether adding a dynamic MOE measurement, for example by resonance, would reduce downgrading of stiff, strong pieces with large knots, or would catch weak pieces that look clean. The review's correlations point in that direction but it does not provide a business case.
• Log-stage acoustic sorting may give an early indication of stiffness classes, so that logs are directed to products where stiffness matters. The review mentions industrial log sorting but does not quantify yield gains.
• Customers who specify strength classes may find the explanation of characteristic values, size effects and moisture conditions useful when discussing why a given board is graded as it is. Moisture and conditioning at 20 degrees C and 65 percent relative humidity are standard references for test results, so drying and handling records matter.

Limitations

This is a broad narrative review, and the portion read does not describe a systematic literature search, so it may not cover all relevant studies and has no quantitative synthesis. Its focus is on European and North American standards, so readers working with other standards or with tropical or plantation species may need additional sources. Because it is a review, numbers quoted are drawn from earlier studies and standards, not generated in the paper itself.

An important limitation concerns this summary. The text available covered the paper only up to the middle of the non-destructive testing section. The sections on duration of load, moisture, temperature, size effect and the authors' conclusions were known only from the abstract and figure captions, and this summary therefore does not report their numerical content. Readers who need factors such as moisture or size adjustments should read the original.

Conclusions

The review presents wood as a material whose mechanical performance can be understood and predicted, provided its orthotropy, natural variability and the testing configuration are respected. Density, microfibril angle and defects explain much of the variation; knots matter but their predictive value differs by species; and stiffness measured non-destructively, including by stress-wave and resonance methods, is a stronger predictor of strength than most visible features. The authors summarise the field by pairing historical development with current standards, giving engineers and graders a shared reference. For sawmills the main message is that grading, measurement method and specimen size are not neutral technicalities: they shape the properties that end up on a strength-class declaration.

Original research

• Authors: Francisco Arriaga, Xiping Wang, Guillermo Iniguez-Gonzalez, Daniel F. Llana, Miguel Esteban, Peter Niemz
• Title: Mechanical Properties of Wood: A Review
• Journal: Forests (MDPI), volume 14, issue 6, article 1202
• Year: 2023
• DOI: 10.3390/f14061202
• Link: https://www.mdpi.com/1999-4907/14/6/1202
• Licence: Creative Commons Attribution 4.0 (CC BY 4.0), as recorded by the ETH Research Collection

This article is an independent summary written by Sawmill Network Europe; the original paper should be consulted for full content, especially the later sections not covered here.