| Tutorials home | Decreasing risk exposure | Safety tour | Emergencies | Meteorology | Flight Theory | Navigation | Communications |

Tutorials home page

Builders guide to safe aircraft materials

Properties of metals

Rev. 16 — page content was last changed 4 October 2010
Google logo


9.1 Structure
At normal temperatures the atoms of solid metals, such as iron and aluminium, are arranged as crystals in the form of a regular matrix or lattice and are held in that arrangement by strong bonding forces associated with free electrons.

Metals have a fairly well defined melting point (1450° C for steel) at which temperature the entire matrix loses cohesion and becomes liquid. As the liquid metal subsequently cools, transformation to a solid starts with the growth of many small crystals forming into clusters, or grains. These grow until all the material consists of small grains having identical crystal matrices but differing orientations — consequently there are irregular boundaries formed between the grains. These boundaries strengthen the metal by resisting 'slip' between grain 'layers'; consequently the finer the grains, the stronger the material.

Alloying is the introduction of atoms of other elements into the crystal matrix of metals to increase strength and to provide particular characteristics. During the conversion from alloy ingots/billets to merchant products, other methods (for example heat treatments, hot working/cold working in rolling mills) are employed to achieve further improvements in strength and other properties.
9.2 Alloy steels
Steel consists of iron plus up to 2% carbon, although generally well below 1% — carbon increases hardness but also brittleness. When other metals (ignoring existing traces) are added the product is an alloy steel. A chromium-molybdenum alloy steel, listed in the American Iron and Steel Institute numbering system as AISI 4130, has been widely used for many years and is the premier material for welded tube airframes and engine mounts for light aircraft. It's also manufactured in sheet, rod, bar and angle forms and is heat treatable to improve properties. The chromium (Cr) and molybdenum (Mo) both increase tensile strength, hardness and toughness. Mo also improves resistance to atmospheric corrosion and intensifies the effects of Cr.

The carbon content of the 4130 alloy is about 0.30% (signified by the last two digits); the other elements vary but are less than 1% Cr and 0.5% Mo, thus 4130 is regarded as a low-alloy steel. If the carbon content of steel exceeds about 0.30% it may contribute to weld quality difficulties. The carbon component places 4130 'chrome-moly' on the boundary between the 'low' and 'medium' carbon content steels. But high-quality welds are still achievable, using oxy-acetylene or TIG welding, and the material can be cut and formed easily, despite the chromium content.

The AN3-20 series general-purpose hexagon head aircraft bolts are usually made from AISI 8740 Ni, Cr, Mo alloy or AISI 4037 Mo alloy. These are high-strength alloys suitable for low temperature conditions — as found at height in winter conditions.

Heat treatments in manufacture. Heating and cooling the near finished merchant product in a carefully controlled process reduces grain size (thus increasing strength), distributes the localised stresses induced during manufacture and increases toughness. The three manufacturing heat treatments associated with 4130 are normalising, water quenching plus tempering and annealing.

The normalising process entails uniform heating to a temperature slightly above the point at which grain structure is affected (known as the critical range) and holding it there for maybe 15 minutes. This is followed by slow cooling in a still, normal temperature atmosphere to produce a fine, uniform grain structure.

Quenching increases tensile strength, yield point and hardness. It involves heating metal above the critical temperature, then hardening by immersion in water or a solution. The rapid cooling ensures that the crystal clusters don't have time to grow in an organised way, thus the grains are small with more irregular boundaries.

Tempering is the reheating of the steel, after quenching, to a temperature below the critical range, then air cooling.

Annealing increases ductility and removes stresses from formed or part-formed material, producing a large-grained structure, and making the metal less tough and easier to machine or cold-form. The process involves heating to a suitable temperature, soaking at that temperature to create a stable structure then allowing the metal to cool very slowly.

The heat treatment condition is often identified by the letter 'A' (annealed) or 'N' (normalised) after the alloy designation (4130 N) or a code like '4130 QT900F', which indicates that the alloy has been quenched and then tempered at 900° F, this provides an ultimate tensile strength around 1150 MPa. Scratch-builders use 4130 in the normalised condition with an ultimate tensile strength around 730 MPa, but welding will reduce this by perhaps 20%.

If it is necessary to recover full normalised strength in the joint areas affected by welding members into an airframe, 4130 N tube structures have to be re-normalised. This is done by reheating those heat-affected zones to a temperature around 800/850° C and then allowing the metal to cool slowly. The process must be carried out in still air at normal temperature and some jigging may be required to prevent distortion in the frame. Even then it is unlikely that the pre-welding strength of 4130N can be recovered. Re-normalising will also relieve the heat expansion/contraction stresses induced in the tubing near the weld.

Stress relieving, rather than renormalising, of homebuilt airframes is usually done after welding. Stress relieving occurs at temperatures around 650/700° C. Tempilstik marking crayons are used to assess when required temperatures are reached. A mark on the metal, made with a particular crayon, will melt at its nominated temperature.

For further information google 'steel austenite pearlite ferrite normalised'.

Mild steels are plain, low-carbon, general-use unalloyed steels that are easily formed but cannot be heat-treated to increase strength. The carbon content is in the range 0.10% to 0.25%, for example AISI 1020 and AISI 1025. Mild steel can be welded easily and the results can be of very high quality. To reduce cost, relatively inexpensive (compared to 4130) mild steel tubing might be used in those parts of light aircraft truss structures where there are no significant tension or bending loads.

AISI 1025 has an ultimate tensile strength around 450 MPa, which is about 60% of 4130N — but expect that the welding process will reduce that by perhaps 20% to 30%.

Corrosion resistant steel [CRES] alloys. Chromium in steel combines with oxygen to form a Cr2O3 barrier that inhibits (but doesn't prevent) corrosion. When the Cr content of CRES alloys exceeds about 12% (in combination with a substantial amount of some other element such as nickel [Ni]) they are classed as stainless steels. Many metal fittings, bolts, rivets and other aircraft fasteners, cables, wires and engine exhaust systems are manufactured from CRES alloys. But because these are difficult to work, they are not in general use as a structural raw material for scratch-builders except for those grades suitable for firewall material.

ASTM 301-304 series (known as 18-8 for the Cr-Ni content percentages, or ISO A2) are the most used — for example, for fuel injector lines. But the acid-resistant/marine grade stainless steel, ASTM 316 (ISO A4) is used for engine exhausts, wire ropes/cables and associated fittings; it may also be specified for use near batteries.
9.3 Nickel alloys
Alloys of nickel [Ni] and copper [Cu] have high strength, and excellent corrosion and acid resistance. The Monel alloys, for example, are used in aircraft rivets. Inconel are more complex Ni-Cr-Fe-Mo-Cu alloys used in high-temperature applications such as engine exhaust systems and perhaps firewalls. (Inconel and Monel are trademark names.)

9.4 Aluminium alloys
Pure aluminium has a tensile strength around 50 MPa but this can be increased perhaps ten-fold by alloying plus thermal treatment and/or mechanical working. The major alloying elements are manganese [Mn], magnesium [Mg], copper [Cu], zinc [Zn] and silicon [Si]. These attributes of alloying, heat treatment and cold working produce a selection of the most versatile and easily formed materials available to the homebuilder.

When pure aluminium surfaces are exposed to the atmosphere, a thin invisible oxide skin forms, permanently protecting the metal from further oxidation; this resistance to normal atmospheric corrosion also applies to the alloys. However, there are other types of corrosion and the alloys' resistance to these depend on the alloying elements; see the section on 'corrosion'.

There are a number of national systems for designating the many available aluminium alloys but the American Aluminum Association's four-digit numbering system is universally recognised. In this system, and except for the near-pure 99%+ aluminium, the major alloying element is indicated by the first digit thus:
      1nnn     Aluminium content greater than 99%
      2nnn     Copper
      3nnn     Manganese
      4nnn     Silicon
      5nnn     Magnesium
      6nnn     Magnesium-silicon
      7nnn     Zinc
      8nnn     Other

In the 1nnn group the last two digits represent the purity above 99% thus alloy 1030 is 99.30% pure aluminium.

Some aluminium alloys in sheet metal form are manufactured with an added surface foil of near-pure aluminium to provide a sacrificial corrosion-resistant surface. For the thin (maybe 0.6 mm/0.025 inch) aluminium sheet used for light aircraft skins the layer is very thin, perhaps 25 micron/0.001 inch. Generally the layers form about 10% of the total thickness (i.e. 5% on each side) and impart a mirror-like finish. Such material is identified as 'Alclad' (a trade name) or 'clad'; a non-clad form may be identified as 'bare'. Some sheets of less corrosion-resistant alloys may be clad with a more resistant alloy rather than the near-pure aluminium.

The following alloys, although far from specific to the aircraft industry, are prominent in light aircraft construction but there are many others and selection is generally a matter of balancing particular needs, availability, strength, weight and cost. Very thin aluminium sheet (perhaps 0.5 mm and below) is easily damaged, even by rather light hail. The lower-strength materials are probably more readily available, and certainly cheaper when measured in $/kg, but then thicker dimensions will be needed to meet flight loads — thus more weight and more cost.

One square metre of 1.0 mm thick aluminium alloy sheet weighs about 2.7 kg, so 0.5 mm sheet would weigh 1.35 kg/m². If an aircraft's wing area is 8 m² then about 16 m² of sheet is used in the skin (top and bottom). Using an expensive high strength 0.5 mm alloy the skin weight is 21.6 kg; if the builder opted for a weaker, less-expensive alloy and increased the thickness to 0.8 mm to compensate then there would be a 13 kg weight penalty. Unnecessary additions to airframe weight detract from aircraft performance (range and rate of climb, for example) and add to ongoing operating costs associated with the increased fuel consumption; so those factors must also be taken into consideration.

Alloy 2024: an alloy (heat-treatable to high strength) of copper-magnesium-manganese (Cu 4.5%, Mg 1.5%, Mn 0.6% plus a number of other elements) which, many years ago, was introduced to replace 2017 (Duralumin) in aircraft structures and is available in forms including bars, plates, rolled shapes, and both bare and Alclad sheet. The bare 2024 material has poor corrosion resistance so it is usually purchased in the Alclad form, which can be recognised by a mirror-like finish; but slight skin damage will expose the 2024 metal to corrosion. Alloy 2024 is the standard commercial aircraft structural material and often used in high-stress structural elements in home-builts. It is also used for stock components like (expensive) spring aluminium undercarriage legs.

Alloy 6061: a magnesium-silicon alloy (Mg 1.0%, Si 0.60%, Cu 0.25%, Cr 0.25%) that develops strength through heat treatment and has good corrosion resistance. This is a very versatile alloy preferred for shapes such as extrusions, angles and channels for spar caps and longerons, drawn tubes (see following note), bars and skin material. The high silicon content provides good fatigue resistance and the 6061-T6 alloy is easy to work in the home workshop. Though expensive, 6061-T6 is probably the most common alloy used by homebuilders because of the availability of forms and sizes. Alloy 6082 is similar and probably the strongest of the readily available 6000 series alloys. Aluminium can be built into an airframe structure using metal fasteners, welding and even epoxy bonding. But simple hand air-gun riveting predominates in all-metal aircraft, so ease of drilling very large numbers of accurate holes in very thin sheets and thicker angles is an important characteristic.

Note: the more expensive drawn tubing generally has higher strength and tighter tolerances than plain extruded or rolled tubing. After extrusion the material is 'drawn' through one or more die-and-mandrel stations to strengthen the grain structure, and to ensure uniform diameter and wall thickness.

Alloy 7075: a very high-strength zinc, copper and magnesium alloy with an exceptional strength/weight ratio. It is difficult to work and very difficult to weld, and has poor corrosion resistance unless clad. The alloy is mainly used in the aerospace industries in plate, rod and bar forms for high-quality machining of parts, but not normally used in home-built aircraft — except perhaps for purchased components. Factory-built hang-glider and trike structures may be formed from 7075 tube.

Alloy 3003: a widely used manganese alloy, sometimes used for aircraft skins, and for pitot and similar fluid lines because it is readily flared and bent. It is about two-thirds the price of 6061-T6 and half the price of 2024-T3 clad but not heat-treatable and thus is weaker.

Alloy 5005: a magnesium-based alloy with similar properties to 3003.

Alloy 5052: another magnesium-based alloy, which is the highest strength alloy of the common non heat-treatable grades, and has particularly good resistance to the sea atmosphere and salt water corrosion. It is highly workable, fatigue-resistant and finds application in fluid lines, fairings, engine cowlings and similar formed parts.

Heat treatment in manufacture. The 2nnn, 6nnn and 7nnn alloys are heat-treatable; that is, they develop strength through various thermal treatments during manufacture. The basic heat treatments for these alloys are identified by the designation 'T' plus a number; this temper condition is added to the alloy designation, thus '6061-T6'. Additional numbers may be appended to identify variations on the basic processes. The heat-treatable alloys are generally unsuitable for welding.

The term solution heat treatment refers to a thermal hardening process where the material is soaked at fairly high temperatures for a number of hours. During this time the alloying elements are put into solid solution; i.e. some of the alloying element's atoms replace aluminium atoms within the normal aluminium crystal lattice. Usually the temperature is then rapidly lowered (quenched) which leaves an unstable structure. Aging is the process of allowing the material to rest without load at room temperature while the internal structure stabilises.
  • T1naturally aged to a substantially stable condition. This applies to products for which the rate of cooling from an elevated temperature-shaping process, such as casting or extrusion, is such that their strength is increased by room-temperature aging; rather than artificial aging in a low temperature oven for perhaps 24 hours.

  • T2 — annealed (cast products only). This designates a type of annealing treatment used to improve ductility and increase dimensional stability of castings.

  • T3 — solution heat-treated and then cold-worked. This applies to products that are cold-worked to improve strength. For example 2024-T3 cold-rolled sheet material.

  • T4 — solution heat-treated and naturally aged to a substantially stable condition. The most common form of driven rivet is manufactured from 2117-T4 wire but the riveting process is a cold-working or strain-hardening process which results in the strength of the driven rivet being equivalent to T3.

  • T5 — artificially aged only. This applies to products that are artificially aged after an elevated-temperature rapid-cool fabrication process, such as casting or extrusion.

  • T6 — solution heat-treated and then artificially aged by sequential heating to around 175° C. This applies to products that are not cold-worked after solution heat-treatment and are readily available in many forms at affordable prices. For example 6061-T6 bars, rods, angles, tubes and sheet.

  • T7 — solution heat-treated and then stabilised.

  • T8 — solution heat-treated, cold-worked, and then artificially aged.

  • T9 — solution heat-treated, artificially aged, and then cold-worked.

  • T10 — artificially aged and then cold-worked.

When welded, the heat-treated alloys lose maybe 30–40% of their strength in the weld area, which cannot be recovered in a home workshop environment and must be compensated for in the joint design.

Non heat-treatable alloys: there are different designations for those alloys that have their mechanical properties adjusted by strain hardening (cold-working) rather than thermal treatment. Cold-working entails such processes as rolling (stretching), compressing or drawing to change the shape of the material. The letter 'H' is followed by two or more digits — the first indicates the particular method used to obtain the temper as follows:
  • H1 — strain hardened
  • H2 — strain hardened, then partially annealed
  • H3 — strain hardened, then stabilised.
The second digit ndicates the degree of strain hardening:
  • 2 — quarter hard
  • 4 — half hard
  • 6 — three quarter hard
  • 8 — full hard
  • 9 — extra hard.

Thus 5052-H32 indicates the alloy has been strain hardened then stabilised and is quarter hard.

9.5 Stress and strain — definition
If an external force is applied to a loose metallic object, such as a length of tube, the tube will move and there is no net change in the internal bonding forces. If a similar force is applied to the same tube while it is a member of an airframe structure — and thus fully or partly restrained from movement — there will be a reaction within the metal as the clusters resist being pulled away from each other, squashed together or their layers sliding apart; these internal resisting forces are the stress. The related strain is a measure of the resulting deformation.

The metals used in aircraft structures are elastic to some extent. That is, they will deform (elongate, bend, flex, compress, twist) under load and when that load is released they will return to their original condition. However there is a limit to the elasticity and if the load is increased beyond that elastic limit into a plastic stage, some of the deformation will remain after the load is released; i.e. there is permanent rather than temporary deformation. If the load is increased beyond that causing permanent deformation, a point will be reached where the material fails. For each metal alloy and form there is a relationship between externally applied forces, the reactive stresses and the associated strains. This is often illustrated in the form of a stress-strain diagram.

The following four definitions associated with the mechanics of materials are from the Oxford Dictionary of Physics published in paperback by Oxford University Press.

" 1. Stress. The force per unit area on a body that tends to cause it to deform (see strain). It is a measure of the internal forces in a body between particles of the material of which it consists as they resist separation, compression, or sliding in response to externally applied forces. Tensile stress and compressive stress are axial forces per unit area applied to a body that tend either to extend it or compress it linearly. Shear stress is a tangential force per unit area that tends to shear a body. [i.e. Stress is the load divided by an area.]

2. Strain. A measure of the extent to which a body is deformed when it is subjected to a stress. The linear strain or tensile strain is the ratio of the change in length to the original length. The bulk strain or volume strain is the ratio of the change in volume to the original volume. The shear strain is the angular distortion in radians of a body subjected to a shearing force.

shearing force  diagram3. Shearing force. A force that acts parallel to a plane rather than perpendicularly, as with a tensile or compressive force. A shear stress requires a combination of four forces acting over (most simply) four sides of' a plane and produces two equal and opposite couples. It is measured as the ratio of one shearing force to the area over which it acts, F/(ab) in the diagram. The shear strain is the angular deformation, theta, in circular measure. The shear modulus is the ratio of the shear stress to the shear strain.

stress-strain diagram4. Elasticity. The property of certain materials that enables them to return to their original dimensions after an applied stress has been removed. In general, if a stress is applied to a wire, the strain will increase in proportion (see OA on the illustration) until a certain point called the limit of proportionality is reached. This is in accordance with Hooke's law. Thereafter there is at first a slight increase in strain with increased load until a point L is reached. This is the elastic limit; up to this point the deformation of the specimen is elastic, i.e. when the stress is removed the specimen returns to its original length. Beyond the point L there is permanent deformation when the stress is removed, i.e. the material has ceased to be elastic and has become plastic.
(Of course plastic deformation is used to shape metals; bending, swaging, beading and flaring of tubing for example). In the plastic stages individual materials vary somewhat; in general, however, at a point B there is a sudden increase in strain without a further increase of stress - this is the yield point. Beyond the point C, the breaking stress, the wire will snap (which occurs at point D)."

Tensile stress is measured as the force per unit area in a plane normal (i.e. at right angles or orthogonal) to the application of the force. Imagine an aluminium strap a metre or so in length hanging vertically from a ceiling beam and supporting a weight of 100 kg. The plane normal to the load will be horizontal, so the area to be measured will be the strap's thickness times its width, say 5 mm × 20 mm = 100 mm² and those dimensions are constant throughout its length; thus the stress will be 1 kg per mm². The area for the stress calculation is the minimum cross-section occurring at any part of the length. Now suppose a gouge or scratch one mm deep is made right across the strap width, having the effect of reducing local thickness to 4 mm and thus area to 80 mm²; consequently the stress will now be 1.25 kg per mm². So although the force (weight) has remained the same, the gouge has raised the stress 25%, which is why such imperfections are known as stress risers.

The same rise in stress would occur if a 4 mm diameter hole was drilled anywhere in the strap length because such a hole would reduce the local cross-section area by 20 mm². Both the hole and the gouge also tend to form stress concentrations close to their peripheries, but we will look at the effect of stress concentrators/risers later. In this example I have ignored the bearing stress, a type of compression loading, which might occur at the upper and lower strap fastenings; but we will look at that in the module 'AN, MS hardware — rivets, bolts and locking devices'.

Material under tensile stress elongates and thus the cross-section area will decrease somewhere. Similarly, material under compressive stress shrinks in length and thus the cross-section area must expand somewhere. This combined with the yield strength has application in riveting, as we will see in the module 'AN, MS hardware — rivets, bolts and locking devices'.

9.6 Application in design
An aircraft designer must be sure that the airframe design will perform its function and that its proportions are sufficient to carry the functional forces applied to it, in any stage of flight, without any part of the structure undergoing permanent deformation — metals become useless in load bearing structures long before they break. See the following:
To accomplish this, the designer must complete a stress analysis and know how the airframe structure will react under the calculated stresses. For that purpose the designer must have access to quantitative mechanical data for all materials and structural components. These data are originated by materials testing laboratories associated with manufacturers, or other organisations, using internationally established standard test methods. The mechanical properties data from the laboratories is developed by others into reference tables for designers covering the multitude of forms, shapes, dimensions, thicknesses and conditions in which that material may be commercially available.

Note that the design will always include the regulated limit load × 1.5 ultimate load safety factor. But of course the designer's ultimate load reserve factor for individual members, or indeed the airframe as a whole, is going to be greater than that regulated ultimate load safety factor.

The mechanical properties of most interest to the aircraft designer — apart from qualities like density, workability, weldability, corrosion resistance and fatigue life — are the following.

  general stress-strain  diagramElastic limit and yield strength [ Sy ] — the elastic limit is the maximum stress the structural member can handle and still return to its original state when the load is released, while the yield strength approximates the stress at the elastic limit.
The term 'proof load' may be substituted for yield strength, particularly in the context of 'pre-loading' in bolted joints.

Note that this particular curve (to the right of the Sy point) shows the metal yielding (i.e. the strain increasing) without any further increase in the load.

A 0.2% offset yield strength is usually specified. This is determined from a stress-strain diagram by offsetting the straight-line portion of the curve by 0.2% from the zero-strain position, reflecting the stress at which larger dislocations occur in the crystals. It is sometimes referred to as the 0.2% proof stress.

Yield strength for metals used in aircraft is generally around 70% of ultimate tensile strength but varies between 65% and 90%. It's also the strength property the designer would normally use in comparing strength/weight ratios.

general stress-strain  diagram

The stress-strain diagram above is generalised and not representative of any particular metal or alloy, but the enlarged portion of a stress-strain diagram at left is representative of aluminium alloys.

Ultimate tensile strength [ Su ] — not as important as yield strength except, perhaps, if the designer is including a 'crash cage' structure for occupant protection. Compressive strength hasn't been mentioned but it can assumed that metal in compression can carry a load equivalent to 70% of Su without buckling and about the same value when subject to torsion stress — a combination of tensile and compressive loads.

Elongation — the plasticity region between Sy and Su is an indication of ductility; the amount of plastic or permanent deformation (stretching, bending, buckling) before the testpiece fails under load. Elongation is the amount of plastic extension at the failure point measured as a percentage of the original length of the gauged portion of the testpiece. The elongation of aircraft structural metals should be not less than 8% and possibly as high as 20%. It's also a measure of brittleness (low or no elongation before breaking) and toughness (high elongation before cracking or breaking). Toughness is a particular property possessed by normalised 4130, which absorbs a lot of energy in permanent deformation before cracking or breaking so it is sometimes used as the centre fuselage (cockpit) structure in an otherwise all-aluminium airframe. The plastic deformation properties of metals when cold-worked is utilised in riveting, flaring or beading tube ends or when swaging fittings to cable ends to form an attachment eye.

Stiffness and modulus of elasticity [E] — an elastic metal under load stores energy and returns that energy when the load is reduced or released. Stiffness is the ratio of stress applied to the strain produced, within the elastic limit, and is apparent from the slope of the initial linear portion of the curve in a stress-strain diagram — the extended modulus line in the diagram above. The greater the angle that line subtends with the strain axis (i.e. the steeper the slope), the stiffer the basic material. Stiffness or rigidity is expressed as the modulus of elasticity (modulus = measure) — otherwise known as Young's modulus which, for tension stress and strain, is calculated as:

E = [force × initial length] / [change in length × cross-section area]

The resultant is a very high number; the 4130 modulus of elasticity is 30 500 000 psi / 210 000 MPa, about three times that of 6061-T6 at 10 000 000 psi / 69 000 MPa. Alloying or heat treatment doesn't have much effect on the stiffness of aluminium (or steel) below yield strength, so all the aluminium alloys mentioned have a similar modulus of elasticity as do all steel alloys.

The bulk modulus is the ratio of the pressure on a body to its fractional decrease in volume. The shear (or rigidity) modulus is the tangential force per unit area divided by the angular deformation in radians.

All-metal light aircraft are sometimes of 'semi-monocoque' construction. Semi-monocoque [monocoque=one shell] fuselage construction consists of a framework of light ring-type formers, and perhaps a few heavier bulkheads, together with longitudinal longerons and light stringers. A skin of thin aluminium sheets is riveted to this framework, taking advantage of a thin skin's good resistance to shear and tension loads, without subjecting it to compression loads which would cause the thin panels to buckle. The aircraft skin and the framework form a very light single structure.
9.7 Effect of shape on stiffness
Structural stiffness in bending or torsion is as important as strength; for example if the aft fuselage bends under changing aerodynamic loads on the tailplane surfaces, the angle of attack of those surfaces will change, thus again changing the load ... and so the cycle continues. In the turbulent atmospheric conditions often encountered near the surface, lack of structural stiffness substantially increases pilot work-load and adds to airframe fatigue — see below.

The modulus of elasticity expresses the stiffness value of standard test pieces, but stiffness of formed metal is also dependent on the geometry of the material. For example the stiffness of a solid metal bar of a particular cross-section area can be changed substantially by reforming it as an angle section, channel section, 'T' section, 'I' section or a round, rectangular or streamlined tube, of the same length and mass. An aircraft designer will be looking for the most efficient material shape for each structural element; i.e. the shape that provides the required stiffness and strength with the least weight.

Structural elements loaded only in tension or compression are, except for long columns, generally very stiff and that stiffness is dependent on the cross-section area (as is the strength) not the cross-section shape. However, the resistance to bending and torsion loads (which are combinations of tension and compression stress acting opposite to each other) is allied to the shape of the element; for example:
  • round or rectangular tubes and other enclosed shapes resist torsion well, and are more rigid against torsion than an 'I' section
  • an 'I' section resists bending better than round or rectangular tubes and that bending stiffness is proportional to the section depth cubed; see properties of beams
  • a hollow circular shaft is the most efficient shape for carrying a torque — hence torque tubes
  • a round tube has the greatest resistance to buckling in compression — and tubing trusses nearly always fail in compression by buckling; see long column buckling.

The bending and torsional stiffness of a tube is a function of diameter and wall thickness. If a round tube of particular outside diameter is reformed into a larger-diameter tube of the same length but reduced wall thickness, the wider tube will be stiffer than the original tube even though the mass is the same. (The increase in stiffness is in proportion to the square of the diameter.) Or a round aluminium tube of 50 mm outside diameter and 2 mm wall thickness will be eight times stiffer than a tube half its diameter with the same wall thickness, even though the larger tube has only slightly more than twice the mass. (The increase in stiffness is in proportion to the cube of the diameter if the wall thickness is maintained — or if the wall thickness is doubled the bending stiffness is also doubled.)

However, there is a limiting relationship between outside diameter and wall thickness before buckling potential might become a problem. Also very thin-walled steel tubing is difficult to weld so there is a minimum wall thickness associated with welded tube structures.

Round, rather than square metal tubing is the material of structural choice because a given mass of metal can be formed into a larger diameter (and thus stiffer) tube of the same wall thickness and length if it is round rather than square. The mass of metal contained in a 10 mm × 10 mm × 2 mm wall thickness square tube would produce a 12 mm outside diameter × 2 mm wall thickness round tube of the same length, which would be considerably stiffer and more resistant to long column buckling. Conversely a round tube can provide the same stiffness as a square tube but with considerably less weight.

Square steel tube greatly simplifies welded truss fabrication, as parts can be mated with simple straight-angled cuts. Whereas, round tube requires more complex radius cutting (a fish mouth) to properly mate the parts prior to welding. This advantage may be appealing to someone who is prepared to trade the long term weight penalty (for example, reduction in payload and rate of climb, increase in fuel bills) for the short time saved in construction. Round tubing is the primary material in most factory-built steel truss fuselages. There may be advantages in using square tubing for fuselage longerons and round tubing for the other members.

Long column buckling. Long, slender structural members that are loaded axially in compression (for example, compression struts in wings and fuselage truss members) have an elastic instability that causes them to reach a critical stress and fail by buckling (i.e. bending sideways) well before the ultimate stress. Such elements are known as long columns. Aluminium skins also exhibit some of the aspects of long column buckling when, for example, wing bending causes compression buckling (oil canning) of skin sections.

The critical stress is:
  • directly proportional to the modulus of elasticity of the material
  • inversely proportional to the column length squared: i.e. halve the length and the critical load is increased four times
  • proportional to the moment of inertia* of the column cross-section, which depends on the cross-sectional area and its distribution from the centroid* of the shape. The farther the area is from the centroid the larger the moment of inertia becomes, which is another reason for round tubing being the favoured material for steel truss fuselages.

    *The centroid is the 'centre of gravity' of a two-dimensional shape; for shapes like the cross-section of aluminium angle sections, the centroid will be outside the cross-section area. The moment of inertia refers to the surface distribution about the centroidal axis and measures the capacity of a cross-section to resist bending.

9.8 Metal fatigue in airframes
Metals in airframes undergo fluctuating cyclic stresses during every flight. If these stresses are great enough they may, over time, cause microscopic fractures to form between grains, both within the metal and on the surface. (This is particularly significant for those aircraft whose fuselages have to be designed as pressure vessels.) Continuing cyclic stresses cause these weaknesses to grow rapidly and strength could deteriorate, maybe to the point where a structural member will fracture. To ensure this does not happen, the designer must be aware of the fatigue/endurance limit of structural materials, design the structure accordingly and perhaps specify a design life/service life for critical members.

stress-strain diagramMetals are laboratory tested to ascertain their fatigue life by subjecting a number of standard test pieces to repetitive positive and negative loadings for thousands/millions of cycles. Initially the test force used is very high, well above Sy but below Su. The elapsed number of cycles at which the test piece fails is plotted on a 'stress/number of cycles' [S-N] diagram. The load is then reduced a little for the next test and the result again plotted; and so on until a large number of tests at reducing loads have been plotted, which might look like one of the curves in the S-N diagram at left. The vertical axis shows stress applied and the horizontal axis shows the total number of cycles — or cumulative flight hours if you like — and it is apparent that the greater the cyclic stress, the shorter the time to failure.

But note the test result for steel. This reveals that once the load is reduced to a particular level during the test process, the metal no longer fractures — no matter how many repetitive loading cycles, at that stress level, are applied. This is the fatigue limit [ Se ] — or endurance limit — and steel will only fail (eventually) if the fatigue limit is exceeded. The fatigue limit is well below the metal's yield strength, perhaps at the 30–60% level. Obviously if the designer expects that loads applied — within the defined flight envelope — will be below fatigue limit there is no need to be concerned about fatigue life or a design life limit for the member.

However, the S-N curve for aluminium shows no fatigue limit; i.e. test failures are still occurring regardless of the reducing application of stress, although the number of cycles to failure increases as stress applied decreases. Thus aluminium has a fatigue strength [ Sf ] that is relative to the number of cycles experienced during its life. The airframe must be designed so that the loading is always less than the fatigue strength relative to a design life limit; i.e. the stress load must always remain under the curve until the airframe reaches its design life limit. Thus there might be critical components within an airframe (or engine) that have a design life limit as low as a few hundred flight hours; these components must be replaced at, or before, that time.

Effect of stress concentrators/raisers: all of this presupposes that the aircraft is not operated outside its limits (for example read Flight at excessive speed) and that all accidental damage is identified quickly and repaired properly. The effect of surface damage in increasing tensile stress was mentioned above, but when a structural member like the main wing spar is under torsional or bending loads the stress is not uniformly distributed throughout the material — it tends to be concentrated at the surface(s); see beam properties.

(The same surface stress flows also apply to some extent to members under tension, which is why aluminium tubing has such a good strength/weight ratio — if the tube were solid metal that ratio would be greatly reduced, because the core metal doesn't contribute much in the way of load sharing.)

Thus surface defects like scratches from over zealous surface preparation, burrs, tool marks, nicks, dents, corrosion, badly formed or burred rivet holes and so on are in an area of stress flow and will act as concentrators, raising the stress in an area where failure is already most likely to occur. Incorrectly fitted rivets, and incorrectly fitted or incorrectly torqued bolts, also act as stress raisers. Design faults that promote discontinuities in the stress flow (square rather than radiused corners — inside and outside — for example) also produce stress concentrations. The edges of all sheet material must be de-burred and rounded.

In practically all cases metal fatigue manifests itself visibly as surface cracks in areas subject to high tensile loads. The cracks propagate at right angles to the tension force — starting from a rivet or bolt hole, an edge notch or a surface defect — and, once started, develop quickly. Some of the airframe areas particularly subject to high tensile stresses are:
  • any fuselage carry-through structures associated with the wings/struts
  • wing strut attachments both at the fuselage and the spar/s
  • upper engine mounting structures
  • attachment point of the vertical tail post
  • under surfaces of the main undercarriage legs.
Fatigue cracking of pipes/tubes subject to engine vibration is often associated with incorrect or insufficient clamping.

Propellers and their blade retaining systems are particularly responsive to surface damage and incorrect retainer fitment. Metal components may fail very quickly if they are subject to stresses outside the design parameters, caused by modification to the original design; see 'The Fox story — gyroscopic loads.
9.9 Fatigue cycles in light aircraft
A cycle doesn't necessarily mean a load reversal; i.e. a structural member in tensile stress changing to compressive stress or vice versa. One cycle is any variation in applied stress — the application/increase of stress followed by the release/decrease of that stress. We saw above that to cause failure, the applied stress has to be of sufficient magnitude and the number of cycles sufficiently high. A vibration in the engine cowling, emanating from the engine or propeller, is a highly repetitive stress fluctuation but of very low magnitude. It is high cycle/low stress, and is not of great concern as a potential fatigue problem, except perhaps as a potential fretting corrosion problem. However, a resonant engine/propeller vibration might have 10 cycles per revolution; thus the number of high-magnitude stress cycles may accumulate rather quickly. At the other extreme a 3.5g wing loading caused by a hard pull-up should be an infrequent event (in a non-aerobatic aircraft) so it is low cycle/high stress and is not significant.

While an aircraft remains airborne, the mean bending loads on the wing and horizontal stabilizer spars is +1g. Superimposed on this are mainly the irregular loads caused by mild turbulence plus, to a much smaller extent, the regular smooth variations caused by the manoeuvring loads. The greater the frequency and extent of turbulence encounters and/or manoeuvring, the greater the number and the amplitude of the bending load cycles during a flight. The more extreme the amplitude of those cycles, the greater the fatigue effect. Similarly, manoeuvring, turbulence and changes in airspeed continually change the torsional loads on wings and empennage.

The cyclic combinations that are cause for concern in fatigue development are the high cycle/high stress events and the most common of these is caused by persistent flight at normal cruise speed in choppy atmospheric conditions — cobblestoning. Training aircraft are also subject to an abnormal number of take-off, circuit and landing cycles, which can make a significant contribution to fatigue damage. The undercarriage and airframe will encounter many stress changes during taxying, take-off and landing — the magnitude depending on the roughness of both the surface and the students/pilots.

Nor does an aircraft have to be flown to develop metal fatigue. There have been some recent examples of aircraft being tied down in the open and exposed to wind loads for long periods — thus developing fatigue failures at a wing strut attachment point, which was also the attachment point for the tethering cable.

Pre-tensioning of structural fasteners greatly reduces the frequency of stress cycles to which the bolts in critical joints are subject. See pre-loading.

9.10 Strength comparison — aluminium and steel alloys
The table below is a comparison of the steel and aluminium alloys mentioned above, showing ultimate tensile strength plus a calculation of the minimum necessary cross-section area for a rod supporting a 10 000 pounds load in tension, with no permanent deformation; i.e. the limit load for that material and cross-section area. In addition, the last column is a calculation of the weight of 10 inches of that alloy section which, in effect, provides a direct comparison of the (tensile) yield strength/weight ratios of the metals; note that 7075-T6 is superior to the quenched and tempered 4130 steel, and 6061-T6 is superior in strength/weight to 4130 N.

I have included the ultimate tensile strengths of the non-tempered versions of the aluminium alloys so that the (roughly) 250% improvement in UTS, provided by the heat treatment, can be seen. I have also added the 6063 alloy which is similar, but not preferable, to 6061-T6 and commonly used for extrusions.

tensile strength
Load carrying capacity
[10 000 lbs]
Alloy psi MPa Area
4130 N 106 000 730 0.149 0.2833 0.42
QT 900° F
166 000 1150 0.0621 0.2833 0.18
2024-0 27 000 185 - - -
2024-T3 70 000 480 0.20 0.1003 0.20
3003-H16 26 000 180 - - -
6061-0 18 000 125 - - -
6061-T6 45 000 310 0.25 0.0975 0.24
6063-0 13 000 90 - - -
6063-T6 35 000 240 ? ? ?
7075-0 32 000 220 - - -
7075-T6 76 000 520 0.137 0.1014 0.14

The next module in this group is 'Metal corrosion'

Builders guide to aircraft materials – metals and hardware modules

| Guide contents | [Properties of metals] | Metal corrosion | Hardware fittings in aircraft structures |

| AN, MS hardware — rivets, bolts and locking devices | Safetying |

Copyright © 2005–2010 John Brandon     [contact information]