01Executive summary
Five materials milestones spread over fifty years share one property: each removed a constraint that designers had previously treated as fixed.
Stainless steel removed corrosion. Age-hardened light alloys removed the weight penalty of metal structure. Nylon removed the dependence on natural fibre supply. Float glass removed the cost and optical distortion of flat glazing. Carbon fibre removed the assumption that stiffness must be paid for in mass. In each case the underlying science was interesting but the engineering achievement was a reproducible industrial process.
02Stainless steel: a film, not a coating
Harry Brearley was investigating erosion in rifle barrels at Sheffield in 1913 when he noticed that a high-chromium experimental steel resisted the etchants used for metallography. The observation was not unique to him — chromium steels had been studied in France and Germany — but Brearley pursued the corrosion behaviour rather than the mechanical properties, and took it into cutlery production.
The mechanism matters for design. Stainless steel is not protected by a coating that can be scratched off. Chromium above roughly 10.5 per cent forms a chromium-rich oxide film a few nanometres thick that reforms immediately when damaged, provided oxygen is available. This has two immediate engineering consequences that are still routinely missed.
It needs oxygen to stay passive
In a crevice, under a gasket, beneath a deposit or inside a stagnant dead leg, oxygen is consumed and not replenished. The film cannot repair, and localised attack proceeds far faster than general corrosion ever would. Crevice and pitting corrosion are therefore design and detailing problems, not material selection problems.
Chlorides break the film
Chloride ions penetrate the passive layer locally. This is why austenitic grades that perform indefinitely inland can pit within months in coastal or chlorinated service, and why molybdenum-bearing and duplex grades exist. Specifying “stainless” without a grade is not a specification.
| Family | Structure | Typical strengths | Typical limitations |
|---|---|---|---|
| Austenitic | Face-centred cubic, non-magnetic | Excellent formability and weldability, tough at low temperature | Susceptible to chloride stress corrosion cracking, higher thermal expansion, not hardenable by heat treatment |
| Ferritic | Body-centred cubic, magnetic | Lower cost, good resistance to chloride stress corrosion cracking | Lower toughness, limited weldability in thick section |
| Martensitic | Hardenable by quench and temper | High hardness and strength, holds an edge | Lowest corrosion resistance of the families, hydrogen embrittlement risk |
| Duplex | Mixed austenite and ferrite | High strength, good chloride cracking resistance, thinner sections for the same duty | Narrow permissible weld heat input band, embrittlement if held at intermediate temperature |
Fabrication decides service life as much as grade selection does. Heat tint left after welding is chromium-depleted and is not passive; it must be removed by pickling, passivation or mechanical means. Carbon steel contamination from grinding media or handling will rust on a stainless surface and initiate attack beneath. Australian practice for stainless in building and marine exposure is addressed in AS/NZS 4673 and AS/NZS 1554.6; verify currency before use.
03Light alloys and the stressed skin
Aluminium in the pure state is too soft to be structural. The change came from precipitation hardening — discovered when a copper-bearing aluminium alloy was found to gain strength over several days at room temperature after quenching. The strengthening comes from a fine dispersion of second-phase particles that impede dislocation motion. It is a heat treatment, not an alloy composition alone, which means the property depends on process history and can be destroyed by subsequent welding heat.
The structural consequence appeared in aircraft. The Junkers J 1 of 1915 flew as a cantilever all-metal aeroplane with no external bracing wires. Over the following two decades this developed into the stressed-skin monocoque: the outer skin carries shear and a substantial share of the bending, rather than merely providing shape over a braced frame.
Why specific properties, not absolute ones, govern
An aircraft structure is sized by stiffness and strength per unit mass, because every kilogram of structure displaces a kilogram of payload or fuel for the life of the airframe. Aluminium alloy has roughly one third the density of steel and roughly one third the elastic modulus, so on a simple tension member the two are comparable. The advantage appears in bending and buckling, where a thicker, lighter section of the same mass has a far greater second moment of area.
Specific stiffness (stiffness per unit mass)
The near equality of the first three is one of the more useful facts in materials selection, and one of the most counter-intuitive. Metals do not differ greatly in specific stiffness. Choosing between them is therefore decided by corrosion, temperature capability, fatigue behaviour, joinability and cost — not by a stiffness-to-weight advantage that largely does not exist. Composites break the pattern because the stiffness is directional.
04Nylon and the designed molecule
Wallace Carothers’ group at DuPont synthesised a polyamide in 1935 that could be drawn into a fibre stronger than silk. What distinguishes it from the earlier plastics covered in the preceding series is intent: earlier polymers were discovered and then found applications, whereas nylon was the product of a programme aimed at understanding polymerisation and then building a molecule to specification.
Drawing is the step that makes the fibre. As-spun polyamide has randomly oriented molecular chains. Cold drawing to several times the original length aligns the chains along the fibre axis, raising tensile strength and modulus substantially and lowering extensibility. The strength is therefore a processing outcome, not a material constant — the same lesson as precipitation hardening, in a completely different material system.
05Float glass: a process worth more than a product
Before 1959, flat glass was either drawn — cheap but optically poor, with visible distortion — or ground and polished from cast plate, which was optically excellent and very expensive because most of the material became waste. Alastair Pilkington’s insight was to form the ribbon on a bath of molten tin: the tin is denser than glass, perfectly flat, and molten across the required temperature range, so the glass floats and both surfaces form flat and fire-polished without any grinding at all.
- Batch and meltSilica sand, soda ash, limestone and cullet melt continuously in a regenerative furnace at around 1,500 °C.
- Float bathMolten glass flows onto molten tin under a reducing atmosphere; gravity and surface tension give a uniform, parallel, fire-polished ribbon.
- Annealing lehrThe ribbon is cooled on a controlled gradient to relieve residual stress that would otherwise cause spontaneous fracture.
- Inspection and cutOptical scanning detects inclusions and distortion; the ribbon is cut on the move into stock sizes.
The concept is simple enough to explain in a sentence. Making it work took about seven years and a large sustained investment, with the plant producing unsaleable glass for months. This gap between concept and reproducible process recurs throughout this series, and is the single most common reason that technically sound ideas fail commercially. The engineering risk in a process innovation is almost never in the physics; it is in the years of iteration needed before yield becomes acceptable.
06Carbon fibre and directional properties
Work at the Royal Aircraft Establishment in 1963 produced carbon fibre with a modulus high enough to be structurally interesting, by carefully controlling the oxidation and carbonisation of a precursor fibre under tension. Comparable work proceeded independently in Japan and the United States, and the modern industry draws on all three lines.
Composites differ from metals in a way that changes the design method rather than merely the numbers. A metal is isotropic: one modulus, one yield strength, and geometry is the only directional decision. A laminate has properties that depend on fibre direction, so the layup is part of the design. This is simultaneously the great advantage — put material only where the load path needs it — and the great difficulty, because strength across the fibres and between plies is governed by the matrix and is far lower.
The laminate is designed, not selected
Ply angles, stacking sequence and symmetry are design variables. An unsymmetric laminate will warp on cooling from cure. Analysis requires laminate theory rather than a single allowable stress.
Damage is often invisible
An impact that leaves no visible surface mark can delaminate plies internally and remove a large fraction of compressive strength. Certification therefore works to a damage-tolerance philosophy with defined inspection intervals and methods.
Joints dominate
Composites cannot be welded and bolt holes cut fibres precisely where load must transfer. Bonded joints, ply drop-offs and metallic fittings usually govern the design rather than the panel itself.
The process is the material
Fibre volume fraction, void content and cure history vary with manufacturing route, so properties are established by coupon testing of the actual process. Two parts of identical geometry from different processes are different materials.
07Takeaways for current practice
- Specify grade and condition, never family. “Stainless”, “aluminium” and “composite” are not specifications. The temper, grade and process route determine whether the part survives.
- Detailing usually beats material selection. Crevices, dead legs, dissimilar metal contact and trapped moisture will defeat a correctly chosen alloy. Most corrosion failures are geometry failures.
- Metals differ far less in specific stiffness than intuition suggests. Choose between them on corrosion, temperature, fatigue, joinability and cost.
- Treat processing history as part of the specification. Precipitation hardening, cold drawing, annealing and cure schedule each set the delivered properties, and downstream heat can undo them.
- Budget for the process, not the invention. The gap between a working demonstration and a reproducible yield is where innovation programmes usually fail.
Australian and international references relevant to this part include AS/NZS 4673 for cold-formed stainless structures, AS/NZS 1554.6 for stainless welding, AS/NZS 1664 for aluminium structures, AS 1288 for glass in buildings, and ISO 2768 and related standards for general tolerances. Cited by number for orientation only — confirm currency and applicability before use.