Why there are thousands of alloys
Count the metals in any room: steel hinges and screws, an aluminium window frame, copper wiring, a brass or stainless-steel tap. Outside, the bridge is steel, the aircraft overhead is aluminium, titanium and composite, and its engine runs on nickel blades in gas hotter than the melting point of the metal they are made of.
Steel dominates by a wide margin. The world makes about 1.9 billion tonnes of crude steel a year (worldsteel), roughly ten times the tonnage of all other metals combined. Yet engineers do not choose "steel" or "aluminium". They choose S355 for a building frame, 42CrMo4 for a crankshaft, 316L for a chemical plant, 6061-T6 for a bicycle frame and CMSX-4 for a turbine blade. Each is a recipe — a base metal, a few per cent of other elements, a processing route and a heat treatment — tuned to sit at a particular point among strength, toughness, weight, corrosion resistance, ease of making and cost.
This lesson tours those recipes with tools you already have — phase diagrams, heat treatment and strengthening mechanisms — then compares alloys by the kilogram and ends with the trade-offs every alloy designer faces.
One idea behind every alloy
A pure metal is soft because its dislocations glide easily: pure iron yields at well under 100 MPa, pure aluminium at a few tens. Almost every engineering alloy answers one question: how do we put obstacles in the dislocations' way without making the metal brittle? The obstacles are the ones you have met — solute atoms, fine particles of a second phase, grain boundaries and other dislocations — plus, above all in steel, a phase transformation that traps carbon in the strained lattice of martensite. Each family below uses a different mix, chosen to suit its crystal structure and chemistry, and the obstacles that raise strength also tend to lower ductility and toughness: the first of the trade-offs at the end.
Steels and cast irons
Iron alloys owe their versatility to two facts. Iron changes crystal structure on heating, from body-centred cubic ferrite (α) to face-centred cubic austenite (γ) at 912 °C, and austenite dissolves up to 2.14 wt% carbon while ferrite dissolves almost none. Cool a carbon-rich austenite and the carbon must go somewhere: into iron carbide, cementite (Fe₃C), in the fine plates of pearlite if cooling is slow, or trapped in the distorted, very hard tetragonal lattice of martensite if it is fast.
Definition— steel and cast iron
A steel is an iron alloy with less than about 2 wt% carbon (in practice, almost always under 1.5 %), usually with manganese and often other elements. A cast iron contains roughly 2 to 4.5 wt% carbon, plus silicon, and is shaped by casting rather than by rolling or forging.
Carbon steels are classed by carbon content: low-carbon (below about 0.25 %), soft, weldable and cheap, for car bodies, cans, pipes and beams; medium-carbon (0.25–0.6 %), quenched and tempered for shafts, gears and rails; high-carbon (0.6–1.4 %), hard, for springs, wire and cutting tools. The lever rule tells us how much pearlite each contains when cooled slowly.
# Phase fractions in slowly cooled plain-carbon steels, just below the eutectoid (727 °C),
# from the lever rule on the iron–cementite diagram.
C_ALPHA, C_EUT, C_CEM = 0.022, 0.76, 6.70 # wt% C: ferrite limit, eutectoid, cementite (Fe3C)
def microstructure(c):
"""(proeutectoid phase, its mass fraction, pearlite fraction, total Fe3C fraction)."""
total_cem = (c - C_ALPHA) / (C_CEM - C_ALPHA) # all the cementite, in pearlite or not
if c <= C_EUT:
pearlite = (c - C_ALPHA) / (C_EUT - C_ALPHA)
return "ferrite", 1 - pearlite, pearlite, total_cem
pro_cem = (c - C_EUT) / (C_CEM - C_EUT)
return "cementite", pro_cem, 1 - pro_cem, total_cem
print(" wt% C proeutectoid pearlite all Fe3C")
for c in [0.10, 0.20, 0.45, 0.76, 1.00, 1.40]:
phase, f_pro, f_p, f_cem = microstructure(c)
print(f"{c:6.2f} {f_pro:5.2f} {phase:<10} {f_p:9.2f} {f_cem:9.3f}")
# Pearlite itself is about one-ninth cementite by mass; its thin plates make it hard.
print(f"cementite in pearlite: {(C_EUT - C_ALPHA) / (C_CEM - C_ALPHA):.3f}")A 0.20 % C structural steel is 76 % soft ferrite and 24 % pearlite, a 0.45 % C shaft steel is 58 % pearlite, and a eutectoid rail steel is all pearlite. Even then cementite is only 11 % of the mass: pearlite's hardness comes from the fine spacing of its carbide plates, not from the amount of carbide. Above 0.76 % C, brittle proeutectoid cementite outlines the old austenite grains unless working or heat treatment breaks it up.
Alloy steels add a few per cent of chromium, molybdenum, nickel, manganese or vanadium, mainly to raise hardenability: they slow the diffusional transformations, so that martensite forms right through a thick section on quenching. High-strength low-alloy (HSLA) steels take the opposite route: under 0.1 % of niobium, vanadium or titanium pins grain boundaries during controlled rolling, and the fine grains give strength and toughness with little carbon, so the steel stays weldable.
Stainless steels contain at least 10.5 % chromium, which forms a chromium-rich oxide film a few nanometres thick on the surface. The film is passive — it stops further corrosion — and heals itself in air if scratched. The families are named after their crystal structure:
| Family | Typical grade | Composition (wt%) | Character | Uses |
|---|---|---|---|---|
| Austenitic | 304, 316 | 18 Cr, 8–10 Ni (316: + 2–3 Mo) | fcc; tough even at cryogenic temperatures | sinks, food and chemical plant |
| Ferritic | 430 | 16–18 Cr | bcc, magnetic, cheaper | appliances, exhausts |
| Martensitic | 420 | 12–14 Cr, 0.15–0.4 C | hardened by quenching | knives, surgical instruments |
| Duplex | 2205 | 22 Cr, 5 Ni, 3 Mo | ferrite + austenite; twice the yield strength of 316 | offshore and desalination plant |
| Precipitation-hardening | 17-4PH | 17 Cr, 4 Ni, 4 Cu | martensite with copper precipitates | pump and aerospace parts |
Nickel stabilises austenite; molybdenum resists pitting in seawater.
Cast irons sit near the eutectic, so they melt some 300 °C below steel, flow well into moulds and shrink little. Most of their carbon comes out as graphite (silicon promotes this), and the graphite's shape decides everything. In grey iron it forms flakes whose edges act as internal cracks: weak in tension, but cheap, easy to machine and excellent at damping vibration, so it makes engine blocks, brake discs and machine-tool beds. In ductile iron, a trace of magnesium or cerium makes the graphite grow as spheres; the cracks become round holes, and the iron is tough enough for crankshafts, water mains and wind-turbine hubs — a discovery of the 1940s, made in the United States and Britain independently. White iron, with its carbon kept as cementite, is hard and brittle and used for wear parts.
Aluminium, copper, titanium and nickel alloys
Aluminium
Aluminium is a third of the density of steel, protected by its oxide film, a good conductor, and easily rolled, extruded and cast. Pure aluminium is soft, so almost all of it is alloyed. Wrought alloys carry a four-digit number registered by The Aluminum Association, whose first digit names the main alloying element:
| Series | Main additions | Strengthened by | Typical uses |
|---|---|---|---|
| 1xxx | none (≥ 99 % Al) | work hardening | foil, electrical conductors |
| 2xxx | Cu (Mg) | precipitation (heat treatable) | aircraft skins and fittings |
| 3xxx | Mn | work hardening | drinks-can bodies, cookware |
| 5xxx | Mg | solid solution, work hardening | ships, tanks, can ends |
| 6xxx | Mg + Si | precipitation | extrusions, frames, car panels |
| 7xxx | Zn (Mg, Cu) | precipitation | aircraft wings and spars, sports gear |
A suffix gives the temper: O annealed, H strain-hardened, T4 solution-treated and naturally aged, T6 artificially aged to peak strength by the precipitation sequence you have studied. Age hardening was found by accident in 1906 by Alfred Wilm in Germany; his aluminium–copper–magnesium alloy, Duralumin, went on to frame airships and the first all-metal aircraft. Casting alloys, mostly aluminium–silicon near the eutectic at 12.6 % Si, make engine blocks, wheels and, increasingly, large sections of car underbodies in one piece. Primary aluminium is made by electrolysis (the Hall–Héroult process of 1886), using roughly 13 to 15 MWh of electricity per tonne; remelting scrap needs only about 5 % of that, which is why aluminium recycling matters so much (see sustainable materials).
Copper
Copper is chosen first for conduction: it defines the International Annealed Copper Standard (100 % IACS, 58 MS/m), second only to silver, and almost every motor winding, cable and circuit-board trace is copper (see printed circuit boards in Lumenary EE). Every alloying element lowers conductivity, so alloys are used where strength, wear or corrosion matter more: brasses (copper–zinc) for taps, valves and instruments; bronzes (copper–tin, and aluminium bronzes) for bearings, propellers and bells; cupronickels for seawater pipework and coins; and precipitation-hardened copper–beryllium for springs, machined only under industrial controls because its dust is toxic.
Titanium
Titanium has the highest strength-to-weight ratio of the common structural metals, an oxide film that resists seawater and body fluids, and half the modulus of steel. It is expensive because extraction is hard: in the Kroll process, titanium tetrachloride is reduced by molten magnesium in batches, and the porous "sponge" is then remelted under vacuum. Titanium is hexagonal α below 882 °C and body-centred cubic β above; aluminium and oxygen stabilise α, vanadium and molybdenum stabilise β. Commercially pure (α) grades serve chemical plant; the α + β alloy Ti‑6Al‑4V, about half of all titanium used, makes airframes, fan blades and implants (see biomaterials); β alloys, the strongest, make landing gear and springs.
Nickel
Nickel is face-centred cubic at every temperature, dissolves large amounts of other elements, and keeps a protective oxide in hot, corrosive gases. Nickel–copper alloys resist seawater; nickel–chromium–molybdenum alloys survive hot acids that dissolve stainless steel; nickel–chromium alloys line furnaces and nuclear steam generators. And nickel is the base of the superalloys.
Superalloys: strength in the hottest part of the engine
A jet engine's efficiency rises with the temperature of the gas entering its turbine. In modern engines that gas is well above the 1300 °C or so at which nickel superalloys begin to melt, and each first-stage blade, spinning at around ten thousand revolutions per minute, feels a centrifugal pull of several tonnes. Power-station gas turbines face the same problem (see power plants and energy conversion in Lumenary ME). Three layers of materials science solve it.
The alloy. A modern blade alloy is about 60 % nickel with cobalt, chromium, aluminium, tantalum, tungsten, rhenium and a dozen other elements. Heat treatment precipitates γ′, the ordered intermetallic Ni₃Al, as cubes about half a micrometre across, filling 60 to 70 % of the volume, separated by thin channels of the nickel-rich solid solution γ. The two phases have almost the same lattice spacing, so the interfaces are coherent and the cubes resist coarsening even after thousands of hours at temperature.
A dislocation entering ordered γ′ leaves behind a sheet of wrongly ordered bonds, an antiphase boundary, so dislocations can only cross it in pairs, the second restoring the order the first destroyed. Stranger still, Ni₃Al gets stronger as it gets hotter, up to about 700–800 °C, because the paired dislocations cross-slip into locked configurations — the main reason superalloys stay strong where steels have long since softened.
The crystal. Creep at these temperatures happens largely at grain boundaries, so the industry got rid of them. Directional solidification, developed in the 1960s, freezes the blade from one end so that its grains run along it; a spiral "grain selector" at the root lets only one grain through, and since the early 1980s the hottest blades have been single crystals. Without grain boundaries, the elements added to strengthen them could be removed, raising the melting point, and a few per cent of slow-diffusing rhenium (in alloys such as CMSX-4) slows creep further.
The coatings and the cooling. The blade is hollow: compressor air flows through it and out of rows of tiny holes, wrapping the surface in a film of cooler air. Over a bond coat that grows a protective alumina scale lies a thermal barrier coating of yttria-stabilised zirconia a few tenths of a millimetre thick, a ceramic you will meet in ceramics and glasses. Together they keep the metal several hundred degrees cooler than the gas. Turbine discs, cooler but more highly stressed, use fine-grained polycrystalline alloys; the widely used Inconel 718, strengthened by a related precipitate, Ni₃Nb (γ″), serves up to about 650 °C.
High-entropy alloys
For thousands of years, alloys had one principal element with smaller amounts of others. In 2004 two groups, Brian Cantor's in Oxford and Jien-Wei Yeh's in Hsinchu, Taiwan, published a different idea: mix five or more elements in nearly equal amounts. Most metallurgists expected a brittle tangle of intermetallic compounds; instead, CrMnFeCoNi, now the Cantor alloy, solidified as a single face-centred cubic solid solution. Yeh's explanation gave the field its name. For an ideal solution, the configurational entropy of mixing per mole of atoms is
which for elements in equal amounts is . The more elements, the larger the entropy, the more negative the term in the Gibbs energy of the solid solution, and the more the disordered solution is favoured over ordered compounds at high temperature. An empirical companion rule uses the valence electron concentration (VEC), the average number of outer s and d electrons per atom: alloys with VEC of 8 or more tend to be face-centred cubic, and below about 6.9, body-centred cubic.
import numpy as np
R = 8.314 # J/(mol K)
# Valence electrons per atom (s + d electrons), as used in the VEC rule.
VEC = {"Al": 3, "Ti": 4, "Zr": 4, "Hf": 4, "V": 5, "Nb": 5, "Ta": 5, "Cr": 6, "Mo": 6,
"W": 6, "Mn": 7, "Fe": 8, "Co": 9, "Ni": 10, "Cu": 11, "Zn": 12}
def mixing_entropy(x):
"""Ideal configurational entropy of mixing per mole of atoms, in units of R."""
x = np.asarray(x, dtype=float) / np.sum(x)
return -np.sum(x * np.log(x))
def vec(alloy):
return sum(n * VEC[el] for el, n in alloy.items()) / sum(alloy.values())
def vec_rule(v): # empirical: fcc if VEC >= 8, bcc if VEC < 6.87, otherwise both
return "fcc" if v >= 8.0 else ("bcc" if v < 6.87 else "fcc + bcc")
alloys = { # atomic proportions, and the structure reported for the cast alloy
"Cu-30Zn brass": ({"Cu": 70, "Zn": 30}, "fcc"),
"Ni-22Cr": ({"Ni": 78, "Cr": 22}, "fcc"),
"CrMnFeCoNi (Cantor)": ({"Cr": 1, "Mn": 1, "Fe": 1, "Co": 1, "Ni": 1}, "fcc"),
"Al0.3CoCrFeNi": ({"Al": 0.3, "Co": 1, "Cr": 1, "Fe": 1, "Ni": 1}, "fcc"),
"AlCoCrFeNi": ({"Al": 1, "Co": 1, "Cr": 1, "Fe": 1, "Ni": 1}, "bcc + B2"),
"NbMoTaW": ({"Nb": 1, "Mo": 1, "Ta": 1, "W": 1}, "bcc"),
"HfNbTaTiZr": ({"Hf": 1, "Nb": 1, "Ta": 1, "Ti": 1, "Zr": 1}, "bcc"),
}
print(f"{'alloy':20} {'dS/R':>5} {'J/(mol K)':>9} {'VEC':>5} {'VEC rule':>10} {'observed':>9}")
for name, (comp, seen) in alloys.items():
s, v = mixing_entropy(list(comp.values())), vec(comp)
print(f"{name:20} {s:5.2f} {s * R:9.1f} {v:5.2f} {vec_rule(v):>10} {seen:>9}")
for n in (2, 3, 5, 10):
print(f"{n:2d} equal elements: dS = R ln {n} = {R * np.log(n):5.1f} J/(mol K);"
f" T dS at 1200 K = {R * np.log(n) * 1200 / 1000:4.1f} kJ/mol")Five equal elements give J/(mol K), and at 1200 K the entropy term is worth about 16 kJ/mol. The VEC rule gets the clear cases right — the Cantor alloy (VEC 8.0) is fcc, the refractory alloys NbMoTaW and HfNbTaTiZr (VEC 5.5 and 4.4) are bcc — but puts Al₀.₃CoCrFeNi, which is fcc as cast, in its mixed band, and cannot tell that equiatomic AlCoCrFeNi forms a disordered bcc phase interwoven with an ordered B2 phase.
Some of these alloys are remarkable. The Cantor alloy gets both stronger and tougher as it is cooled, with fracture toughness above 200 MPa√m at liquid-nitrogen temperature (Gludovatz et al.), and its relative CrCoNi is tougher still, down to 20 K (Liu et al.), because deformation twinning keeps work hardening going as strain builds up. Refractory high-entropy alloys of niobium, molybdenum, tantalum and tungsten keep useful strength above 1200 °C, beyond nickel superalloys, though most are brittle at room temperature and oxidise badly — the problems their designers are working on now.
Comparing alloys by the kilogram
For a bridge or a ship, the cost per tonne decides. For anything that moves — an aircraft, a car, a bicycle, a robot arm, a satellite — mass is what matters, and the right comparison is not strength but strength per kilogram. Two simple ratios start the comparison: the specific stiffness and the specific strength . But the shape of the part matters too, and that leads to the idea of a material index.
Example— The material index for a light, stiff beam
A beam of length and square cross-section must have a bending stiffness of at least (force per unit deflection); its length is fixed, but we may choose its material and its size . From beam theory, , with and a constant that depends on how the beam is supported and loaded. The mass is .
Solve the stiffness condition for the free variable: . Substitute into the mass:
The first bracket is fixed by the design; the second depends only on the material. The lightest beam is made of the material with the largest . Repeat the argument for a tie in tension of fixed strength () or a flat panel of fixed width and length with free thickness (), and a different exponent appears each time.
The exponent changes the ranking. Here are representative values for eleven alloys, and the mass of each part relative to one made of ordinary structural steel:
import numpy as np
# Representative room-temperature values, rounded: density (kg/m^3), Young's modulus (GPa),
# 0.2 % proof (yield) strength (MPa). Strengths are standard minimums or typical values for the
# temper named; real values depend on product form, section size and supplier.
alloys = [
("S355 structural steel", 7850, 210, 355),
("42CrMo4 steel, Q&T", 7850, 210, 750),
("304 stainless, annealed", 8000, 193, 205),
("ductile iron GJS-400-15", 7100, 169, 250),
("Al 6061-T6", 2700, 69, 276),
("Al 7075-T6", 2810, 72, 503),
("Mg AZ31B-H24", 1770, 45, 220),
("Ti grade 2 (CP)", 4510, 105, 275),
("Ti-6Al-4V, annealed", 4430, 114, 880),
("Cu C11000, annealed", 8940, 115, 70),
("Inconel 718, aged", 8190, 200, 1030),
]
names = [a[0] for a in alloys]
rho, E, sy = (np.array([a[i] for a in alloys], dtype=float) for i in (1, 2, 3))
r = rho / 1000 # Mg/m^3, so E/r is in MJ/kg and sy/r in kJ/kg
spec_E = E / r # specific stiffness, MJ/kg (= GPa per Mg/m^3)
spec_s = sy / r # specific strength, kN m/kg
tie_mass = (rho / sy) / (rho[0] / sy[0]) # mass of a tie of given strength, relative to S355
beam_mass = (rho / E**0.5) / (rho[0] / E[0]**0.5) # light stiff beam: maximise E^(1/2)/rho
panel_mass = (rho / E**(1/3)) / (rho[0] / E[0]**(1/3)) # light stiff panel: E^(1/3)/rho
print(f"{'alloy':26} {'E/rho':>6} {'sy/rho':>7} | mass relative to S355 for a")
print(f"{'':26} {'MJ/kg':>6} {'kJ/kg':>7} | {'strong tie':>10} {'stiff beam':>10} {'stiff panel':>11}")
for i in np.argsort(-spec_s):
print(f"{names[i]:26} {spec_E[i]:6.1f} {spec_s[i]:7.0f} | {tie_mass[i]:10.2f} {beam_mass[i]:10.2f} {panel_mass[i]:11.2f}")Three lessons are in that table. First, specific stiffness is almost the same for every common structural metal, between 23 and 27 MJ/kg (copper, at 12.9, is the outlier), so for a stiff tie the choice of alloy saves nothing. Second, specific strength varies twenty-five-fold, from 8 kJ/kg for annealed copper to about 200 for Ti‑6Al‑4V, whose strong tie weighs under a quarter of one in S355. Third, shape changes the winner: for a stiff panel magnesium wins (0.38 of the steel mass) and aluminium comes second, because a less dense metal can be made thicker, and bending stiffness grows with the cube of the thickness. That is why laptop cases are magnesium or aluminium.
Ashby's property charts plot these quantities for every family of materials on logarithmic axes, where each index becomes a straight line of fixed slope; they are the standard first step of materials selection.
Alloy design trade-offs
An alloy is a compromise, and knowing the compromises is most of the art. Five come up again and again.
Strength against toughness and ductility. Blocking dislocations reduces how far a metal can deform before it breaks: plot elongation against strength for car-body steels and you get a falling "banana". Much of modern steel research climbs above it, with dual-phase steels (martensite islands in ferrite), TRIP steels whose retained austenite transforms as it is strained, TWIP steels that deform by twinning, and boron steels hot-formed and quenched in the die to about 1500 MPa.
Strength against weldability. A weld cools fast, so a high-carbon or highly alloyed steel forms brittle martensite beside it, which can crack as hydrogen from welding diffuses in. Welding engineers summarise the risk in a carbon equivalent; the International Institute of Welding's is
By a common rule of thumb, steels below about 0.40 weld without special precautions; higher values need preheating and controlled procedures. So structural steels get their strength from fine grains and microalloying rather than from carbon.
Strength against corrosion and cracking. Peak-aged 7075-T6 is prone to stress-corrosion cracking, so aircraft use over-aged tempers such as T73, trading some strength for resistance. Stainless steel held between about 450 and 850 °C, as beside a weld, precipitates chromium carbides on its grain boundaries and corrodes along the chromium-depleted zones; the low-carbon "L" grades (at most 0.03 % C) prevent this. And steels stronger than about 1000 MPa grow sensitive to hydrogen embrittlement.
Performance against cost and supply. Most niobium comes from Brazil and most cobalt from the Democratic Republic of the Congo, and rhenium, a by-product of copper mining, is produced at a few tens of tonnes a year worldwide. Alloys that need them carry a supply risk; superalloy makers have worked for two decades to use less rhenium.
Performance against recyclability. Copper and tin in scrap steel cannot be refined out and make steel crack during hot rolling, so mixed scrap must be diluted with clean iron; mixed aluminium scrap can usually only become casting alloys. Impurity-tolerant alloys and better scrap sorting are central to sustainable materials.
How alloys are specified, designed and made today
Names and certificates. Every industrial alloy is defined by a standard: EN 10027 steel names (S355, 42CrMo4), AISI/SAE numbers (1020, 4140), the Unified Numbering System, the Aluminum Association's registry, ASTM specifications, and the Japanese, Chinese and Indian standards that parallel them. Each delivery comes with a certificate of the melt's measured composition and tensile results, and aircraft design values come from statistically based handbooks of allowables, not from single tests.
Computational alloy design. The CALPHAD method fits thermodynamic models of each phase to measurements and first-principles calculations, then predicts the phases of alloys with ten or more elements. Commercial packages (Thermo-Calc, FactSage, Pandat) and open-source ones (pycalphad, OpenCalphad), coupled with diffusion and precipitation models, form integrated computational materials engineering, which has cut alloy development from decades to years. You will meet the methods in multiscale modelling and AI-assisted materials discovery.
Decarbonising metals. Iron and steel making emits roughly 7 to 9 per cent of the world's fossil CO₂ (worldsteel), mostly from reducing ore with coke in blast furnaces. As of 2026 two routes are growing: electric arc furnaces melting scrap (about a quarter to a third of world output), and hydrogen direct reduction, which removes the ore's oxygen as water instead of CO₂. The Swedish HYBRIT pilot delivered the first hydrogen-reduced steel in 2021, and larger plants are being built; their cost depends on cheap low-carbon electricity.
Making parts. Casting, forging, rolling, extrusion, powder metallurgy and metal additive manufacturing (see manufacturing processes and additive manufacturing in Lumenary ME) each leave their own microstructure, and an alloy designed for one route often fails in another: alloys such as 7075 that crack during laser powder-bed fusion are an active problem of 2026.
Common misconceptions
Misconception
"Stainless steel never rusts." Anything that breaks its passive film faster than the film heals can corrode it: chlorides (pitting on a seaside railing, crevice corrosion under a gasket), sensitised grain boundaries beside welds, or iron particles smeared on from a carbon-steel tool. Molybdenum-bearing and duplex grades resist chlorides better.
Misconception
"Titanium is the strongest metal." It has one of the highest specific strengths, not the highest strength. Quenched and tempered steels, maraging steels and nickel superalloys all have higher yield strengths than Ti‑6Al‑4V; titanium wins only when strength is divided by density.
Misconception
"Cast iron is always brittle." Grey iron, with its flakes of graphite, has little tensile ductility. Ductile iron, which differs only in the shape of its graphite, stretches 10 to 20 per cent before breaking and is used for crankshafts and pressure pipes.
Misconception
"High-entropy alloys are single-phase because of their entropy." Entropy favours a solid solution, but enthalpy often wins: many alloys of five or more elements are multiphase, and several famous single-phase ones decompose on long annealing. Some of the most useful are deliberately multiphase.
Exercises
1. Metals and alloys
Warm-upHumanA plain-carbon steel with 0.40 wt% C is cooled slowly from the austenite region to just below 727 °C. Using the lever rule with ferrite at 0.022 wt% C and the eutectoid at 0.76 wt% C, what percentage of its mass is pearlite? Give the nearest whole number.
2. Metals and alloys
CoreHumanA pipe of type 304 stainless steel (0.06 wt% C) carries a mildly corrosive process liquid. After a year, it has corroded along a narrow band a few millimetres to each side of every weld, while the weld metal itself and the rest of the pipe are untouched. Under the microscope, the attack follows the grain boundaries. What is the most likely cause, and the remedy?
3. Metals and alloys
CoreHumanWhich of these help a nickel superalloy turbine blade survive in gas hotter than its own melting point? Select all that apply.
4. Metals and alloys
Warm-upHumanUsing the ideal configurational entropy of mixing, what is for an equiatomic four-element alloy such as NbMoTaW, in J/(mol K)? Take J/(mol K) and give the nearest whole number.
5. Metals and alloys
CoreHumanA steel tie rod (S355: GPa, kg/m³) in a machine frame is to be replaced by one of aluminium alloy 6061-T6 ( GPa, kg/m³) with the same length and the same axial stiffness . How does the mass of the rod change?
6. Metals and alloys
CoreComputationalWrite two functions for a welding engineer.
carbon_equivalent(comp)takes a dict of composition in wt% (for example{"C": 0.16, "Mn": 1.40}) and returns the IIW carbon equivalent . Elements missing from the dict count as 0; elements not in the formula (such as Nb or Si) are ignored.needs_preheat(comp)returnsTruewhen the carbon equivalent is above 0.40, the rule of thumb used in the lesson, andFalseotherwise.
7. Metals and alloys
StretchComputationalWrite lightest(materials, part), which returns the name of the material that makes the lightest part of the given kind. materials is a list of tuples (name, density_kg_m3, E_GPa, yield_MPa), and part is one of
"tie": a tie of fixed length and required strength (mass ∝ ρ/σ_y),"beam": a beam of fixed length and required bending stiffness, square section free to scale (mass ∝ ρ/E^(1/2)),"panel": a flat panel of fixed length and width and required bending stiffness, thickness free (mass ∝ ρ/E^(1/3)).
Raise ValueError for any other part.
Further reading and resources
- ASM InternationalPublishes the ASM Handbooks (properties, phase diagrams and heat treatment of every major alloy family) and the Alloy Phase Diagram Database.
- worldsteel (World Steel Association)Production statistics, steel's carbon footprint and the industry's decarbonisation data.
- The Aluminum AssociationKeeps the international registry of wrought aluminium alloy designations and tempers.
- Copper Development AssociationProperty data and applications of copper and its alloys, by UNS number.
- Nickel InstituteTechnical guides to stainless steels and nickel alloys.
- Microstructural development in equiatomic multicomponent alloys · B. Cantor, I. T. H. Chang, P. Knight and A. J. B. Vincent · 2004One of the two papers that started the field of high-entropy alloys, with the five-element alloy CrMnFeCoNi.
- Nanostructured high-entropy alloys with multiple principal elements: novel alloy design concepts and outcomes · J.-W. Yeh et al. · 2004The other founding paper, which named the field.
- High-entropy alloys · E. P. George, D. Raabe and R. O. Ritchie · 2019A critical review of what high-entropy alloys do and do not offer.
- A fracture-resistant high-entropy alloy for cryogenic applications · B. Gludovatz et al. · 2014
- Nickel-based superalloys for advanced turbine engines: chemistry, microstructure and properties · T. M. Pollock and S. Tin · 2006
- The Superalloys — Fundamentals and Applications · R. C. Reed · 2006Cambridge University Press; the standard monograph on nickel superalloys.
- Materials Selection in Mechanical Design · M. F. AshbyThe book that introduced material indices and property charts.
- pycalphadAn open-source Python library for computational thermodynamics (CALPHAD), used to design alloys.
- HYBRITThe Swedish project making iron with hydrogen instead of coke.