|| Tutorials home | Decreasing risk exposure | Safety tour | Emergencies | Meteorology | Flight Theory | Navigation | Communications ||
Builders guide to safe aircraft materials
Properties of wood
Rev. 4 — page content was last changed 6 January 2010
(If you wish to learn more about the microstructure of trees google the terms 'tracheid parenchyma rays vessels'.)
Trees expand their trunks by addition of new peripheral growth layers. In softwoods this expansion growth can generally be discerned on the end surfaces of a cut log as a series of concentric annual growth rings. Each ring may display the growth as a lighter colour representing the faster growth during the earlier growing season (early wood). A darker colour indicates slower, denser growth during the less favourable part of the growing season (late wood).
Environmental events will also affect growth rings. Because of favourable year-round conditions the rings in tropical — and possibly sub-tropical — trees may not exist, or may be difficult to discern. The same applies to eucalypts from Australia's warm temperate climate areas.
The honeycomb-like cross-section photomicrograph (100× magnification) shows the band of thin-walled, large cavity early wood cells of one season on the right, with the stronger, thick-walled, small cavity cells of the prior season's late wood on the left. The timber from a fast-grown tree will not be as strong as that from a slow-grown tree of the same species. On the other hand, the timber from a tree that has grown too slowly will not be as strong as that from the optimally grown tree.
The rate of growth is shown by the width of the annual rings, or the number of growth rings per 25 mm. Generally, for those softwoods typically used in aircraft construction, the number of radial growth rings appearing in the end or on the face of sawn boards, should be at least six, possibly eight, per 25 mm but fewer than 15 to 18 — and with a high percentage (50%?) of the stronger late wood. (If the tree is grown too slowly, the strong late wood bands are too narrow; if grown too fast, the weaker early wood bands are too wide [in softwoods], or the late wood cell walls are too thin [in hardwoods].)
One aspect to note is that moisture and minerals extracted from the soil by the roots are delivered to the crown of the trees through a form of capillary attraction within the cells. When seasoned wood is used for mechanical structures, a similar action will draw adhesive away from end-grain joint surfaces unless appropriate application methods are used. See 'Wood joints and adhesives'.
When looking at the machined faces of a board, visible lines from the growth rings may indicate the direction of the board grain. For aircraft-grade timber straight grain is vital (see 'Strength and stiffness'). The general lines of the grain along the longitudinal axis of the board should be reasonably straight and the maximum grain slope on all sides should not deviate from parallelism with the long axis by more than 25 mm in 400 mm, i.e. a ratio around 1:16, or 1:20 if the timber is for wing spars. (Sloping grain has many causes; spiral growth, growth around knots or just the sawmilling process.) Wood will split along the lay of the fibres, so splitting a sample length is accepted as a normal method of detecting grain slope; but there are other less destructive methods.
Two distinct zones can often be seen in the cross-section of a cut log. The inner darker zone, possibly more than 70% of the surface area, is the heartwood, which provides structural strength. The outer zone is the sapwood, which provides the tree's nutrient storage and sap flow. As the trunk expands, the inner sapwood cells are gradually converted to thicker-walled heartwood. Seasoned sapwood is not as dense as heartwood and the nutrients contained make it more prone to insect attack.
In Australia, boards are classed as 'back-sawn' if the growth rings lie at an angle less than 45° to the longer cross-section dimension, or 'quarter-sawn' if the rings lie at an angle greater than 45° to the longer dimension. In the image above you can see that all boards, from the log on the right, would be quarter-sawn when cut this way. However, if quarter-sawn boards are specified by the aircraft designer the expectation may be that the rings are 90° to the longer cross-section dimension ±10°. Quarter-sawn boards are less likely to distort or crack during the drying process and are more stable in service. An advantage with quarter-sawn softwood is that the rate of growth is readily seen when selecting boards. effects of humidity and heat. Timber with MC between 15% and 25 % is sometimes regarded as partially seasoned.
In the drying process the wood first loses the free moisture to reach the fibre saturation point [FSP], where no moisture is contained in the cell cavities but the cell walls are still saturated with bound moisture. The FSP occurs at 30–35% MC in hardwoods and 25–30% MC in softwoods. Timber does not shrink during drying until the FSP is reached, then it begins to shrink at a roughly proportionate rate until an equilibrium MC is attained. Board shrinkage from FSP to 12% MC is usually insignificant longitudinally, quite significant in the tangential direction along the rings (6% to 8%) but not so much in the radial direction — maybe half the tangential shrinkage. However, there are some timbers that do not conform with that generalised statement; Australian hoop pine for example, where the tangential shrinkage from FSP to 12% MC is only about 3.5% and the radial shrinkage is around 2.5%.
The specific gravity of the cell wall material is about the same in most timbers (about 1.5, which is 50% heavier than water) but the density (mass per unit volume) of seasoned wood is governed chiefly by the relationship between cell wall and the cell cavity volume — cavities lack mass — so density is an indicator of strength and stiffness. A single density value for a particular species is often quoted in the literature but there is considerable variation in the density of boards from the same species, which varies according to the maturity of the tree, the part of the trunk (base or top, inner or outer) from which the board is cut, the growth conditions and whether it is plantation or naturally grown — the latter generally being more dense.
(If the apparent specific gravity of a particular wood is stated then multiply that by 1000 to obtain the density in kg/m³. For example, apparent specific gravity 0.55 density = 550 kg/m³.)
Measuring density: the only way to really determine the density of a particular seasoned board is to cut a piece from the board, carefully measure both volume and weight, and convert to kg/cubic metre. If you dry it in the microwave for 20 or 30 minutes before measuring and weighing you will have the density at 0% MC and the density at 12% MC will be about 6% greater. When drying perhaps it may be better if you check the weight after 15 minutes in the microwave then every five minutes or so until the weight stops reducing. To check MC see below.
Density classification: the density of seasoned timber is usually measured for classification purposes at 12% air-dry MC. The density classification is typically:
• exceptionally light — under 300 kg/m³
• light — 300 to 450 kg/m³
• medium — 450 to 650 kg/m³
• heavy — 650 to 800 kg/m³
• very heavy — 800 to 1000+ kg/m³
Aircraft weight restraints mandate the use of wood having low weight but ample strength. Such timber is most likely contained in the lower band of the medium density classification; except perhaps timber for propellers. For 90 years top-quality North American Sitka spruce (average density perhaps 440 kg/m³) has been the aircraft designer's timber of choice for most of the airframe, and regarded as the benchmark for comparison purposes.
An aircraft, like Leo Pownings Jodel D18, requires around 0.15 m³ of wood, so if built from Sitka spruce it might contain 66 kg of timber. Hoop pine (at left) is the preferred Australian timber, but at an average 510 kg/m³ at 12% MC it is about 15% heavier than Sitka spruce — although stiffer and generally stronger, and somewhat more stable. The difference in empty aircraft weight between Sitka spruce and hoop pine (using material of the same dimensions) in the Jodel would be about 10 kg, equivalent to nearly 14 litres of fuel — a significant weight increase when there is a legally imposed upper limit to recreational light aircraft MTOW, and there might be a slight change in cg position. However, although heavier, hoop pine is probably a superior timber for home builder construction; in weight and strength it is similar to Douglas fir but easier to work.
For a comparison of the strength properties of hoop pine, Sitka spruce and Douglas fir see 'Comparison of three recommended softwoods'.
Both MR and ME are expressed in units of pressure — megapascals; MPa [ = N/mm² ]. To convert pounds per square inch to megapascals multiply by 0.007.
The graph is an approximation of compressive strength and stiffness relative to the density of seasoned wood.
Other aspects of wood strength generally considered are:
Commercial strength groups: laboratories ascertain the mechanical properties of species by testing large numbers of 'clear' specimens (i.e. free of sloping grain and other defects) with standard dimensions — possibly 50 mm × 50 mm × 300 mm for the MR test — and generally at 12% MC and in a 20 °C environment. Like density, the mechanical properties of defect-free material of the same species vary considerably, so you must be wary when the MR or ME of a particular species is published as a single value. That value may be the mean value of all tests or it could be the '5th percentile' value — meaning that 95% of the samples tested had an MR or ME higher than that value — or it could be any other of the standard deviation values associated with the normal distribution curve.
To overcome this problem of strength definition when there is so much variation within species, it is convenient to assign timber within strength groups. In the Australian standard grading for seasoned structural timber, each species is allocated a ranking of SD1 (strongest) to SD8 (weakest) according to the following values.
The diagram at left shows a board under tension (the load is trying to stretch it) and the angle of the grain to the axis of the load is about 45°. At this angle the tensile/compressive strength of the board is probably reduced to less than 25% of its available strength at the 0° angle, i.e. when the grain is parallel to the long axis of the board and to the load.
In softwoods grain angles greater than a few degrees produce a markedly disproportionate reduction in tensile/compressive strength — maybe 25% reduction at just 15° and 50% reduction at 30°. The decrease in stiffness is even greater. Hence, as stated previously, the need to use boards that allow all structural members cut from them to have a maximum grain slope better than 1:15 (4°), or perhaps 1:20 (3°) for critical structures, throughout the component. (The structural member can be sawn from the board in a manner that produces minimum grain slope within that member.)
hygroscopic so the MC of seasoned wood will adjust to the relative humidity of its environment — either absorbing water vapour from the atmosphere or evaporating moisture into it — until the wood reaches an MC that is in equilibrium with the atmosphere [EMC]. Read the 'atmospheric moisture' section in the meteorology guide and note vapour partial pressure, and how relative humidity changes with temperature.
So although the components of a fully seasoned wooden structure may have 12% MC when in the temperate coastal zone of Australia, the same structure's MC might fall to 7–8% in dry inland conditions or rise to 18–20% in monsoonal conditions in northern Australia.
The density of wood changes by about 0.5% for each percentage point variation from 12% MC; i.e. a board at 18% MC will be 3% heavier than when at 12%.
The strength of wood is inversely proportional to the temperature; if the temperature of wood at 12% MC is increased from 20 °C to 40 °C, the modulus of rupture will decrease by around 15%. (Rule of thumb: there is about a 1% reduction in the ultimate strength and stiffness values for each 1 °C increase in wood temperature; and the converse applies for temperature decrease.) Short-term heat soaking will not permanently affect strength but long periods at high temperatures will reduce the ultimate MR and ME values.
(The colour of the surface finish has a very significant effect on the temperature of aircraft surfaces: in 40° C ambient temperature the temperature of a white surface can reach 68° C, light green 84° C, red 100° C and black 110° C. An aluminium finish is about 75° C, so it is best to paint your aircraft white. The figures are from an aircraft maintenance publication that was specifically referring to fibre-reinforced polymer surfaces.)
If long-term MC exceeds 20% the wood's susceptibility to decay or dry rot is greatly increased — particularly so in warmer temperatures, and in conditions where free moisture is trapped within the structure and oxygen can be absorbed from the atmosphere. MC changes also affect strength; a change from 12% MC to 18% MC will decrease the modulus of rupture by perhaps 25%.
Increasing bendability: plasticisation (softening) of wood can be achieved chemically, by microwave irradiation, by steaming or by boiling; the latter methods are the most appropriate for the home workshop. If pieces of wood of smaller cross-sections are soaked in a home-made steam bath, or near-boiling water, for sufficient time for the wood to reach an internal temperature of 90–95° C and MC of 18–20%, the piece becomes ductile and thus able to be bent into a permanent shape without fracturing. If then clamped in a shaped form of the desired curve, the piece will maintain somewhere near that bend after cooling and drying out — a very useful property when forming wing rib cap strips and similar curved components. (By experimenting with an exaggerated bending form profile, it is possible to compensate for the springback and achieve the exact curve after release).
Stability: unless all surfaces and joints are coated with an impermeable barrier, the absorption or release of water vapour will prompt dimensional movement (swelling/contraction) in structural members which, in effect, changes tension/compression in the structure.
Movement is insignificant longitudinally but generally quite significant in the tangential dimension (along the rings) — often being more than twice the movement in the radial dimension — so it is important to align the cross-section grain of a structural member so that the extra tangential movement causes the least stress. The green rectangle in the image shows expansion due to moisture intake, though greatly exaggerated. If it is desirable that minimum all-round movement should occur then the longest dimension of a rectangular section member should be aligned in the radial direction. The tangential movement in hoop pine is much less than the norm for aircraft softwoods and only about 40% more than the radial movement, thus the overall dimensional movement in a quarter-sawn hoop pine board will be quite small.
The timber for solid beams should normally be back-sawn, particularly in softwoods, as the strength of the growth rings when roughly parallel to the beam depth provides additional bending resistance. However, stability considerations would dictate the reverse, as the movement in the tangential dimension along the rings may be about twice the movement in the radial dimension, as explained above; thus solid aircraft spars are normally quarter-sawn, though with hoop pine one might opt for a back-sawn spar because of the relatively small difference in radial/tangential movement.
An online wood shrinkage and expansion calculator is available at www.woodbin.com/calcs/shrinkulator.htm. If you input the initial MC (say 8%) and the EMC (say 16%) plus a nominal dimension (say 100) the calculator will provide the expansion radially and tangentially.
Painting, varnishing, epoxy coating or other moisture barriers slow the rate of adjustment to the environment and the wood takes some time to adjust to EMC. So, daily humidity changes probably have no significant effect but seasonal changes certainly will. Moisture is more readily absorbed through end grain because of the capillary action provided, so special measures may need to be taken to minimise that absorption. When constructing an aircraft it is important to ensure that no spaces within the structure are completely sealed off from the atmosphere; not just to ensure moisture movement but also to ensure that all compartments/cells readily adjust to in-flight atmospheric pressure changes.
Measuring MC: you can measure the MC of a board by cutting off a sample, weighing it carefully and immediately, then microwaving it for sufficient time to dry the sample completely. Weigh it again then calculate the initial MC%, which will be:
([initial weight − dry weight] / dry weight) × 100
For example: initial weight 87 grams, dry weight 77 grams:
([87 − 77] / 77) × 100 = 13% MC
Effects of fuel and other liquids. Avgas, kerosene and most lubricating oils do not react with wood and have no significant effect on strength. However alcohols (as contained in mogas) and ethylene glycol (antifreeze) will cause wood to swell and will reduce strength while present, about the same effect as water absorption.
commercial strength groups table. For values of workability, steam bendability and stability I have just used the indicators 'good', 'ok' and 'unsatisfactory' but again it has not been possible to find a satisfactory laboratory comparison.
Although a number of Australian hardwoods have been listed alpine/mountain ash would probably be the hardwood of choice because of its availability.
1. Klinki pine may not be obtainable as most production from PNG is exported to Japan.
2. Douglas fir from NZ plantations is less dense than that from North America and is downgraded to the SD6 strength group. However, some North American Douglas fir is also comparable to the NZ timber; there is a variation between lowland and upland-grown timber.
3. The tangential shrinkage for hoop pine from FSP to 12% MC is about 3.5% while the radial shrinkage is around 2.5%.
The next module in this group is 'Properties of plywoods'
Builders guide to safe aircraft materials – wood, plywood and adhesives modules
| Guide contents | [Properties of wood] | Properties of plywoods |
| Wood joints and adhesives | Wood beams in aircraft | Selecting aircraft timber |
| Basic strength and elastic properties of wood |
AC 43.13-1B FAA advisory circular chapter 1-1; wood structure — materials and practices. Read section 1.2. (pdf document 99 kb)
Why not wood? An article from Flight Safety Australia. (html document)
Aircraft wood An article about North American timber suitable for aircraft construction. [external html document)
ANO 108-29 Specification: timber for use in aircraft propellers. (pdf document 21 kb)
ANO 108-22 Specification: klinki pine for aircraft use. Interesting document but Klinki is probably unobtainable. (pdf document 24 kb)
Copyright © 2005–2010 John Brandon [contact information]