
Engineering Materials and Metallurgy
Build unit cells, cool alloys through their phase diagrams, quench a steel through its TTT diagram, carburise a gear and grow a crack until the plate breaks
20 chapters, 23 interactives, few words. Every chapter is mapped to the Anna University, VTU, SPPU, Mumbai University and KTU units it covers: classes of materials and bonding, crystal structures, Miller indices and X-ray diffraction, imperfections, diffusion, slip and twinning, strengthening and annealing, solidification, phase diagrams and the lever rule, eutectic and eutectoid reactions, the iron-iron carbide diagram, heat treatment with TTT and CCT diagrams, hardenability and surface hardening, alloy steels and cast irons, non-ferrous alloys, polymers, ceramics and composites, powder metallurgy and materials selection, mechanical testing, fracture, fatigue and creep. SPPU rows follow 211503 (SE Automation and Robotics), whose unit list we could check; Mechanical's 202044 covers the same subject.
- AUME3392Engineering Materials and Metallurgy
- VTUBME303Material Science and Engineering
- SPPU211503Material Science and Engineering Metallurgy
- MUMEC304Materials and Metallurgy
- MU2403114Material Science
- KTUMET205Metallurgy and Material Science
By Skill-Lync · Updated 17 Nov 2026 · 20 chapters · 23 interactives · free, no sign-up
Classes of engineering materials and atomic bonding
Which syllabus covers thisAU not in syllabusVTU BME303 Module 1SPPU 211503 Unit IMU MEC304 Module 1; 2403114 Module IKTU MET205 Module 1
What are the main classes of engineering materials?
Metals, ceramics, polymers and composites. Metals (steel, aluminium, copper) are strong, tough, conduct heat and electricity and can be bent without breaking. Ceramics (alumina, silicon carbide, glass, cement) are hard, stiff and heat-resistant but brittle. Polymers (polyethylene, nylon, epoxy) are light, easy to mould and corrosion-proof, but soft and limited to modest temperatures. Composites combine two of these, such as glass or carbon fibres in epoxy, to get properties none has alone. Semiconductors, biomaterials and smart materials are often listed as extra groups.
What are ionic, covalent and metallic bonds?
The three primary bonds, all strong. In an ionic bond one atom gives electrons to another, and the positive and negative ions attract in every direction (NaCl, MgO). In a covalent bond neighbours share electrons in fixed directions (diamond, silicon, the chains in polymers). In a metallic bond the outer electrons leave their atoms and form a 'sea' shared by all the positive ion cores. That free electron sea is why metals conduct and why their atoms can slide past each other without the bond breaking, which makes metals ductile.
Why do strongly bonded materials melt at higher temperatures and expand less?
Plot the energy of two atoms against their spacing: attraction pulls the energy down, repulsion pushes it up at close range, and the sum has a well. The bottom of the well is the equilibrium spacing r0 and its depth is the bonding energy E0. A deeper well needs more thermal energy to pull atoms apart, so the melting point rises: tungsten (849 kJ/mol) melts at 3,410 °C, argon (7.7 kJ/mol) at −189 °C. The well is steeper on the close side, so a vibrating atom spends more time further out and the material expands as it heats.
What are secondary (van der Waals) bonds?
Weak attractions between molecules or atoms that already have full shells, caused by electric dipoles: fluctuating dipoles in argon, permanent dipoles in HCl, and the strongest kind, the hydrogen bond, in water and between nylon chains. They are typically about 10 kJ/mol against hundreds for primary bonds. Secondary bonds hold the chains of thermoplastics together, which is why those plastics soften on heating while the covalent chains themselves stay intact. They also hold graphite's layers together, so the layers slide and graphite works as a lubricant.
How does bonding decide a material's properties?
Bond strength sets the melting point and the stiffness: Young's modulus follows the curvature of the energy well at r0, which is why diamond and ceramics are stiff and polymers are not. Bond direction and charge decide ductility: metallic bonds let planes of atoms slip and re-bond, while in ionic crystals slipping one plane brings like charges face to face and in covalent crystals it breaks directional bonds, so both are brittle. Free electrons explain electrical and thermal conductivity. Structure-property correlation, the theme of this book, starts here.
Argon's bonding energy is 7.7 kJ/mol, about 1% of iron's. Solid argon exists only below about:
Weak van der Waals bonds need very little thermal energy to break, so argon melts at −189 °C. Iron, with metallic bonds of 406 kJ/mol, melts at 1,538 °C.
Worked examples
Bonding energy per atom · Bonding energies: NaCl 640 kJ/mol, W 849 kJ/mol, Ar 7.7 kJ/mol; 1 eV per particle = 96.485 kJ/mol
- NaCl: 640/96.485 = 6.63 eV per ion pair
- W: 849/96.485 = 8.80 eV per atom
- Ar: 7.7/96.485 = 0.080 eV per atom, about a hundredth of tungsten's
Answer: 6.63, 8.80 and 0.080 eV
Equilibrium spacing and bond energy of an ion pair · E_N = −A/r + B/rⁿ with A = 1.436 eV·nm, B = 5.86 × 10⁻⁶ eV·nm⁹, n = 9
- At r0 the force is zero: dE/dr = A/r² − nB/rⁿ⁺¹ = 0, so r0 = (nB/A)^(1/(n−1))
- r0 = (9 × 5.86×10⁻⁶ / 1.436)^(1/8) = 0.279 nm
- E_A = −1.436/0.279 = -5.15 eV; E_R = B/r0⁹ = 0.57 eV (exactly 1/n of |E_A|)
- E0 = -5.15 + 0.57 = -4.57 eV
Answer: r0 = 0.279 nm, E0 = -4.57 eV
Crystal structures: SC, BCC, FCC and HCP, packing factor, density and polymorphism
Which syllabus covers thisAU not in syllabusVTU BME303 Module 1SPPU 211503 Units I and IIMU MEC304 Module 1; 2403114 Module IKTU MET205 Module 1
What is the difference between crystalline and non-crystalline (amorphous) materials?
In a crystal the atoms sit in a pattern that repeats in three dimensions over long distances: long-range order. All metals, most ceramics and some polymers solidify this way. In an amorphous solid such as window glass, or a polymer cooled quickly, the atoms have only short-range order: each atom's nearest neighbours are arranged sensibly, but the pattern does not repeat. Amorphous solids have no sharp melting point; they soften gradually over a range around the glass transition temperature.
What is a unit cell, and what are the seven crystal systems?
The unit cell is the smallest block of atoms that builds the whole crystal when stacked in three directions, like a brick in a wall. It is described by three edge lengths a, b, c and three angles α, β, γ. The possible combinations give seven crystal systems: cubic, tetragonal, orthorhombic, rhombohedral, hexagonal, monoclinic and triclinic. Adding body-, face- and base-centred versions gives the 14 Bravais lattices. Most engineering metals are cubic or hexagonal, so this book works mainly with those.
How do SC, BCC, FCC and HCP structures differ?
In the number of atoms per cell, the number of touching neighbours (coordination number) and how tightly the spheres pack. Simple cubic: 1 atom, 6 neighbours, 52% full; only polonium uses it. BCC (α-iron, chromium, tungsten): 2 atoms, 8 neighbours, 68%. FCC (copper, aluminium, nickel, γ-iron): 4 atoms, 12 neighbours, 74%. HCP (magnesium, titanium, zinc): 12 neighbours and 74% too, because FCC and HCP are the two ways of stacking close-packed layers (ABCABC and ABAB). Pick a metal and shrink the atoms to see inside its cell.
How do you calculate the theoretical density of a metal?
Mass of the atoms in one unit cell divided by the cell's volume. The cell holds n atoms, each of mass A/N_A (atomic weight over Avogadro's number), and its volume is a³ for a cubic cell, with a found from the atomic radius: a = 2R for SC, 4R/√3 for BCC and 2√2 R for FCC. The answer usually lands within 1% of the measured density, which is a quick check on the structure you assumed. A wrong n or a wrong a-R relation shows up immediately.
What are polymorphism and allotropy?
A material with more than one crystal structure is polymorphic; for a pure element the word is allotropy. Iron is the one that matters most: BCC α-iron (ferrite) up to 912 °C, FCC γ-iron (austenite) from 912 to 1,394 °C, then BCC δ-iron up to melting at 1,538 °C. Every heat treatment of steel in this book depends on that BCC to FCC change. Carbon is another example: diamond and graphite are the same element with different structures and completely different properties. Tin turns brittle grey below 13.2 °C.
What are interstitial voids, and why does γ-iron dissolve more carbon than α-iron?
The gaps between touching atoms. FCC has 4 octahedral voids per cell, able to hold a sphere of 0.414R, and 8 tetrahedral voids of 0.225R. BCC has more voids (6 octahedral, 12 tetrahedral) but they are smaller and flattened: the largest sphere fits at 0.291R. Carbon (radius about 0.071 nm) is bigger than any of them, so it strains the lattice wherever it sits, but much less in FCC's large octahedral void. That is why austenite dissolves up to 2.14 wt% carbon and ferrite only 0.022 wt%.
Iron changes from BCC to FCC when heated through 912 °C. Its volume:
FCC packs 74% of space, BCC only 68%, so the same atoms take less room after the change, about 1% less even though the atoms are slightly larger in FCC. A dilatometer sees iron contract on heating through 912 °C, the opposite of normal thermal expansion.
Worked examples
Packing factor of BCC · Two atoms per cell, touching along the body diagonal
- Body diagonal √3·a = 4R, so a = 4R/√3
- Volume of atoms: 2 × (4/3)πR³
- APF = 2 × (4/3)πR³ / (4R/√3)³ = π√3/8 = 0.680
Answer: APF = 0.68: BCC is 68% full
Theoretical density of copper · FCC, R = 0.128 nm, A = 63.5 g/mol, N_A = 6.022 × 10²³ /mol
- a = 2√2 R = 0.3620 nm; V_C = a³ = 0.04745 nm³ = 4.745e-23 cm³
- ρ = nA/(V_C N_A) = 4 × 63.5 / (4.745e-23 × 6.022×10²³)
- ρ = 8.89 g/cm³ (measured 8.94)
Answer: 8.89 g/cm³, within 1% of the measured value
Lattice parameter and density of α-iron · BCC, R = 0.1241 nm, A = 55.85 g/mol
- a = 4R/√3 = 4 × 0.1241/1.732 = 0.2866 nm
- V_C = a³ = 0.02354 nm³
- ρ = 2 × 55.85 / (2.3540e-23 × 6.022×10²³) = 7.88 g/cm³
Answer: a = 0.2866 nm, ρ = 7.88 g/cm³ (measured 7.87)
Miller indices, planar and linear density, and X-ray diffraction
Which syllabus covers thisAU not in syllabusVTU BME303 Module 1SPPU 211503 Unit IMU MEC304 Module 1; 2403114 Module IKTU MET205 Modules 1 and 2
How do you write the Miller indices of a direction?
Take a vector through the cell, subtract the tail coordinates from the head coordinates (in units of the cell edges a, b, c), clear fractions to the smallest integers and write them in square brackets: [uvw]. A negative index gets a bar over it. The body diagonal from the origin to the far corner is [111]; the face diagonal in the base is [110]. Directions related by symmetry form a family written <uvw>: all twelve face diagonals of a cube are <110>.
How do you find the Miller indices of a plane?
Find where the plane cuts the three axes in units of a, b, c, take the reciprocals, clear fractions and write (hkl). A plane parallel to an axis cuts it at infinity, so that index is 0. If the plane passes through the origin, shift the origin to another corner first; a negative intercept gives a barred index. Equivalent planes form a family {hkl}: the six faces of a cube are {100}. Switch between planes and directions, change h, k and l, and watch the section and the lattice points on it.
What are planar and linear density, and why do they matter?
Planar density is the number of atoms centred on a plane per unit area; linear density is the number of atoms per unit length along a direction. The densest planes and directions are where slip happens, because their atoms are packed closest and the planes are furthest apart. In FCC the densest planes are {111} and directions <110>; in BCC there are no fully close-packed planes, and {110} with <111> are the densest. That pairing gives the slip systems of chapter 6.
How are planes written in hexagonal crystals?
With four Miller-Bravais indices (hkil), using three axes a1, a2, a3 at 120° in the basal plane plus the c axis. The third index is redundant, i = −(h + k), but it makes equivalent planes look alike: the six prism faces are all {101̄0}. The basal plane, the close-packed layer in HCP, is (0001). Converting a three-index plane (hkl) to four indices just inserts i = −(h + k).
How does X-ray diffraction identify a crystal structure?
X-rays scattered by parallel planes reinforce only when the extra path between neighbouring planes, 2d sin θ, is a whole number of wavelengths: Bragg's law. For a cubic crystal d = a/√(h² + k² + l²), so each family of planes gives a peak at its own angle. The structure decides which peaks appear: BCC shows only planes with h + k + l even (110, 200, 211...), FCC only all-odd or all-even (111, 200, 220, 311...). Read the peak sequence and you know the structure; measure the angles and you know a.
Which plane of an FCC metal has the most atoms per unit area?
(111) is the close-packed plane: 4/(√3 a²) = 2.31/a² atoms per unit area, against 2/a² on (100) and 1.41/a² on (110). That is why FCC metals slip on {111}.
Worked examples
Indices of a plane from its intercepts · A plane cuts x at 1/2, y at 1 and runs parallel to z
- Intercepts: 1/2, 1, ∞
- Reciprocals: 2, 1, 0
- Already integers, so no clearing is needed
Answer: (210)
Planar density in aluminium · FCC aluminium, a = 0.4049 nm
- (111): the triangle inside one cell holds 3 × 1/6 + 3 × 1/2 = 2 atoms in an area √3a²/2, so PD = 4/(√3 a²)
- PD(111) = 4/(1.732 × 0.4049²) = 14.09 atoms/nm²
- (100): 4 × 1/4 + 1 = 2 atoms in a², so PD = 2/a² = 12.20 atoms/nm²
Answer: (111) 14.1 atoms/nm², (100) 12.2 atoms/nm²
Diffraction angle for the (220) planes of iron · α-iron (BCC), a = 0.2866 nm; cobalt Kα X-rays, λ = 0.1790 nm; first order
- 2 + 2 + 0 = 4 is even, so (220) diffracts in BCC
- d = 0.2866/√8 = 0.1013 nm
- sin θ = λ/(2d) = 0.1790/(2 × 0.1013) = 0.8833
- θ = 62.04°, so 2θ = 124.1°
Answer: 2θ = 124.1°
Crystal imperfections: point, line, surface and volume defects, and how to see them
Which syllabus covers thisAU not in syllabusVTU BME303 Module 1SPPU 211503 Unit IMU MEC304 Module 1; 2403114 Module IKTU MET205 Module 2
How are crystal imperfections classified?
By how many dimensions they extend. Point defects (zero-dimensional): vacancies, self-interstitials, and impurity atoms sitting in place of a host atom (substitutional) or between host atoms (interstitial); ionic crystals also have Schottky (a cation-anion vacancy pair) and Frenkel (a displaced ion) defects. Line defects: dislocations. Surface or planar defects: free surfaces, grain boundaries, twin boundaries and stacking faults. Volume defects: pores, voids, cracks and inclusions. Nearly every useful property of a metal, from strength to diffusion, is controlled by its defects.
How many vacancies does a metal have, and why do they increase with temperature?
A vacancy costs energy to make (Qv, about 0.9 eV in copper) but adds disorder, so at any temperature there is an equilibrium number that rises exponentially with temperature. In copper near its melting point roughly one site in a few thousand is empty; at room temperature the fraction is around 10⁻¹⁵. Vacancies are how atoms diffuse in metals, so diffusion speeds up just as fast. Move the temperature and the formation energy and read the fraction on a log scale.
What are edge and screw dislocations, and what is the Burgers vector?
A dislocation is a line around which the atoms are out of register. An edge dislocation is the edge of an extra half-plane of atoms pushed into the crystal; a screw dislocation is a spiral ramp, as if the crystal were cut part way and sheared by one atom spacing. Most real dislocations are mixed. The Burgers vector b is the closure failure of a loop of atom-to-atom steps drawn around the line: perpendicular to an edge line, parallel to a screw line. In metals b is the shortest lattice repeat: a/√2 in FCC, a√3/2 in BCC.
What is dislocation density, and how does a Frank-Read source multiply dislocations?
Dislocation density is the total line length per unit volume (or lines crossing a unit area): about 10³ mm⁻² in carefully solidified crystals, 10⁵ to 10⁶ mm⁻² in annealed metals and up to 10⁹ to 10¹⁰ mm⁻² after heavy cold work. They multiply during deformation at sources such as the Frank-Read source: a segment pinned at both ends bows out under shear stress, loops round, pinches off a complete loop and is ready to bow again. The stress to operate it is about Gb/L for a pinned length L.
What are grain boundaries and twin boundaries?
A grain boundary separates two crystals (grains) of different orientation in a polycrystal. Low-angle boundaries, under about 15° of misorientation, are rows of dislocations; high-angle boundaries are a few atoms thick and badly matched. Boundary atoms have extra energy, so boundaries etch faster, are faster diffusion paths, block dislocations (chapter 7) and are where cracks and corrosion often start. A twin boundary is a special, low-energy boundary across which the atoms are mirror images, common in annealed brass and copper. Stacking faults are mistakes in the ABC stacking of close-packed layers.
How are grains and defects observed?
Metallography: grind the specimen flat, polish it to a mirror, then etch it with a chemical such as nital (nitric acid in alcohol) for steel. Grain boundaries and different phases are attacked at different rates, so they show up under an optical microscope, which resolves features down to roughly 0.2 μm. A scanning electron microscope (SEM) scans an electron beam over the surface and gives deep focus and much higher magnification, ideal for fracture surfaces. A transmission electron microscope (TEM) sends electrons through a thin foil and can image individual dislocations. X-ray diffraction (chapter 3) gives structure and lattice parameter.
Copper at 1,000 °C (Qv = 0.9 eV). Roughly what fraction of its lattice sites are empty?
Nv/N = exp(−0.9/(8.62×10⁻⁵ × 1273)) = 2.74e-04, about one site in 3,600. Close to the melting point, yet still only a small fraction of a percent.
Worked examples
Vacancies in copper at 1,000 °C · Qv = 0.9 eV/atom, density at 1,000 °C 8.4 g/cm³, A = 63.5 g/mol, k = 8.62 × 10⁻⁵ eV/K
- N = N_A ρ / A = 6.022×10²³ × 8.4 / 63.5 = 7.97e+22 atoms/cm³ = 7.97e+28 atoms/m³
- Nv/N = exp(−0.9 / (8.62×10⁻⁵ × 1273)) = 2.74e-04
- Nv = 7.97e+28 × 2.74e-04 = 2.2e+25 vacancies/m³
Answer: 2.2e+25 vacancies per m³, one in about 3,600 sites
Burgers vectors of copper and iron · Cu FCC a = 0.3615 nm; α-Fe BCC a = 0.2866 nm
- FCC: b = a/2 [110], |b| = a/√2 = 0.2556 nm (equal to 2R: neighbouring atoms touch along <110>)
- BCC: b = a/2 [111], |b| = a√3/2 = 0.2482 nm (2R along <111>)
Answer: |b| = 0.256 nm (Cu), 0.248 nm (Fe)
Stress to operate a Frank-Read source · Copper, G = 46 GPa, b = 0.2556 nm, pinned segment L = 1 μm
- τ ≈ Gb/L: the line tension (about Gb²/2) must be balanced by the stress bowing the segment to a semicircle of radius L/2
- τ ≈ 46,000 MPa × 0.2556×10⁻⁹ m / 1×10⁻⁶ m = 11.8 MPa
Answer: About 12 MPa; halving L doubles it
Diffusion in solids: mechanisms, Fick's laws and carburising
Which syllabus covers thisAU not in syllabusVTU BME303 Module 2SPPU not in syllabusMU not in syllabusKTU MET205 Module 2
What is diffusion in solids?
The movement of atoms through a solid from where they are concentrated to where they are scarce, one atomic jump at a time. Atoms vibrate about their sites and every so often one has enough energy to jump to a neighbouring empty site. Interdiffusion is the mixing of two different metals placed in contact; self-diffusion is the same random jumping among atoms of one metal, measured with radioactive tracers. Diffusion is slow at room temperature and fast near the melting point, which is why most heat treatments happen hot.
What are the vacancy and interstitial mechanisms of diffusion?
In vacancy diffusion an atom jumps into a neighbouring vacancy, so the atom and the vacancy swap places. It needs a vacancy next to the atom, so it depends on the vacancy concentration (chapter 4); substitutional atoms such as nickel in copper and iron in iron move this way. In interstitial diffusion a small atom (carbon, nitrogen, hydrogen) hops between interstitial voids. There are many empty voids and no vacancy is needed, so interstitial diffusion is far faster. Carbon in α-iron at 500 °C has D of about 1.4 × 10⁻¹² m²/s.
What is Fick's first law?
In steady state, when the concentration at each point no longer changes, the flux of atoms through a plane (kg or atoms per m² per second) is proportional to the concentration gradient and flows down it. The constant is the diffusion coefficient D, in m²/s. It applies to purifying hydrogen through a palladium sheet or carbon passing through a steel plate held between a carburising and a decarburising gas.
What is Fick's second law, and how does it describe carburising?
When the concentration changes with time, Fick's second law applies. For a thick piece of steel at C0 whose surface is held at Cs by a carbon-rich gas, the solution is an error-function profile. The depth reached at a given concentration grows with √(Dt): to double the case depth you need four times the time. Change the temperature, time and carbon levels and watch the profile and the case depth move.
What factors affect the rate of diffusion?
Temperature most of all: D rises exponentially, D = D0 exp(−Qd/RT), so for carbon in austenite going from 925 to 1,000 °C multiplies D by 2.4. Then the diffusing species (small interstitials are fast, large substitutional atoms slow), the host structure (more open BCC is faster than close-packed FCC for the same atom), and the path: grain boundaries, surfaces and dislocations are short-circuits, so fine-grained metals diffuse faster at low temperature. Plot ln D against 1/T and you get a straight line whose slope is −Qd/R.
Where is diffusion used in mechanical engineering?
Case hardening of gears and camshafts by carburising, nitriding and carbonitriding (chapter 13); homogenising cast ingots to even out segregation; sintering powder-metal parts (chapter 17); diffusion bonding of titanium and dissimilar metals; and doping of silicon wafers. It also causes problems: decarburisation of steel surfaces in a furnace, which softens them, creep at high temperature (chapter 20), and the growth of brittle intermetallic layers at solder joints.
Carburising for 4 hours gives a 0.6 mm case. At the same temperature, 8 hours gives about:
Case depth grows with √(Dt). Doubling the time multiplies it by √2 = 1.41, so 0.6 × 1.41 = 0.85 mm. To double the depth you need four times the time.
Worked examples
Flux through a steel plate (Fick's first law) · Steel plate at 700 °C, D = 3 × 10⁻¹¹ m²/s; carbon 1.2 kg/m³ at 5 mm and 0.8 kg/m³ at 10 mm below the carburising surface
- Gradient: (0.8 − 1.2)/(10 − 5)×10⁻³ = −80 kg/m⁴
- J = −D dC/dx = −3×10⁻¹¹ × (−80) = 2.4e-09 kg/(m²·s)
Answer: J = 2.4e-09 kg/(m²·s)
Carburising time for a 1 mm case · Steel with 0.20 wt% C, surface held at 1.00 wt% C, 925 °C; want 0.40 wt% C at 1.0 mm. Carbon in γ-iron: D0 = 2.3 × 10⁻⁵ m²/s, Qd = 148 kJ/mol
- D = 2.3×10⁻⁵ exp(−148,000/(8.314 × 1198)) = 8.11e-12 m²/s
- (Cx − C0)/(Cs − C0) = (0.40 − 0.20)/(1.00 − 0.20) = 0.25 = 1 − erf(z), so erf(z) = 0.75 and z = 0.813
- z = x/(2√(Dt)) → t = (x/2z)²/D = (0.001/1.627)² / 8.11e-12
- t = 46,562 s = 12.9 h
Answer: About 12.9 h at 925 °C
Same case, hotter furnace · As above but at 1,000 °C
- D(1,000 °C) = 1.95e-11 m²/s, 2.4 times D at 925 °C
- Same depth and same concentrations → same z, so D·t is constant
- t = 12.9 h / 2.40 = 5.4 h
Answer: 5.4 h: 75 °C hotter saves more than half the time, at the cost of grain growth
Plastic deformation: slip systems, Schmid's law and twinning
Which syllabus covers thisAU ME3392 Unit VVTU BME303 Module 1SPPU 211503 Unit IMU MEC304 Module 1; 2403114 Module IKTU MET205 Module 1
How do metals deform plastically?
By dislocations gliding on slip planes, not by whole planes of atoms sliding at once. Shearing a perfect crystal plane over its neighbour would need a stress of roughly G/10 to G/30, thousands of megapascals; real metals yield at a few hundred megapascals or less. A dislocation moves by breaking and remaking one row of bonds at a time, like moving a rug by pushing a ruck along it. Each dislocation that crosses the crystal shifts the top over the bottom by one Burgers vector, and the steps appear on a polished surface as slip lines.
What is a slip system, and how many do FCC, BCC and HCP metals have?
A slip plane plus a slip direction in it: the densest plane and the densest direction. FCC: {111}<110>, 4 planes × 3 directions = 12 systems, all easy to start, so copper, aluminium and austenitic stainless steel are very ductile. BCC: {110}<111> gives 12, and {211} and {321} add more, but they need a higher stress, which rises steeply in the cold. HCP: basal slip (0001)<112̄0> gives only 3, so zinc and magnesium are less ductile at room temperature unless other systems or twinning help.
What is Schmid's law?
A tensile stress σ on a single crystal resolves into a shear stress on each slip system: τR = σ cos φ cos λ, where φ is the angle between the load axis and the slip plane normal and λ the angle between the load axis and the slip direction. Slip starts on the system with the largest Schmid factor cos φ cos λ when τR reaches the critical resolved shear stress τCRSS, a property of the metal. The factor is at most 0.5, at φ = λ = 45°. Pick a crystal and a load axis to see which of the 12 systems slips first.
What is twinning, and how is it different from slip?
In twinning a region of the crystal shears so that its atoms become a mirror image of the untwinned part across a twin plane. In slip each block moves by whole atom spacings and the orientation does not change; in twinning every plane moves by a fraction of a spacing in proportion to its distance from the twin plane, and the twinned region is reoriented. Twinning gives little strain by itself but happens fast, in HCP metals and in BCC metals at low temperature or high strain rate, and the reorientation can bring new slip systems into play. Annealing twins form during heat treatment of FCC brass and copper.
Why is a polycrystal stronger than a single crystal of the same metal?
In a polycrystal every grain has a different orientation, so for any load direction many grains are badly oriented for slip, and each grain must change shape to stay joined to its neighbours. That needs at least five independent slip systems working in each grain (the von Mises condition), which FCC and BCC metals have and HCP metals usually do not. On average the yield stress of an FCC polycrystal is about 3.06 times the critical resolved shear stress (the Taylor factor), against a factor of 2 for the best-oriented single crystal.
A single crystal is pulled along [001]. A slip system whose slip plane normal is perpendicular to the load (φ = 90°) feels a resolved shear stress of:
cos 90° = 0, so τR = 0 however hard you pull: the load lies in the slip plane and has no shear component across it. The same happens when the slip direction is perpendicular to the load (λ = 90°).
Worked examples
Resolved shear stress in an iron single crystal · BCC iron pulled along [010] at 52 MPa; slip system (110)[1̄11]; τCRSS = 30 MPa
- cos φ = ([010]·[110])/(1 × √2) = 1/√2 → φ = 45°
- cos λ = ([010]·[1̄11])/(1 × √3) = 1/√3 → λ = 54.7°
- τR = 52 × 0.707 × 0.577 = 21.2 MPa (below 30 MPa: no slip yet)
- Yield: σy = 30/(0.707 × 0.577) = 73.5 MPa
Answer: τR = 21.2 MPa; slip starts at σ = 73.5 MPa
Orientation changes the yield stress of an FCC crystal · FCC crystal, critical resolved shear stress τc; load along [001], then along [111]
- [001]: the best systems have cos φ = 1/√3 and cos λ = 1/√2, m = 1/√6 = 0.408 (8 systems at once), σy = τc/0.408 = 2.45 τc
- [111]: the best systems have cos φ = 1/3 and cos λ = 2/√6, m = 0.272 (6 systems), σy = 3.67 τc
- Ratio 3.67/2.45 = 1.5
Answer: Along [111] the crystal yields at 1.5 times the stress needed along [001]
Strengthening mechanisms, cold and hot working, recovery, recrystallisation and grain growth
Which syllabus covers thisAU ME3392 Unit IIVTU BME303 Module 3SPPU 211503 Unit IMU MEC304 Modules 1 and 4; 2403114 Module IIKTU MET205 Modules 1, 2 and 4
How do you make a metal stronger?
Make it harder for dislocations to move, because yielding is dislocation motion. Four ways: make the grains smaller so boundaries block them (grain refinement), dissolve foreign atoms whose strain fields pin them (solid-solution strengthening), add more dislocations so they tangle with each other (strain hardening), or scatter fine hard particles in their path (precipitation and dispersion hardening, chapter 15). All except grain refinement cost ductility; grain refinement is the one mechanism that raises strength and toughness together, which is why controlled-rolled steels aim for fine grains.
What is the Hall-Petch relation?
Yield stress rises with the inverse square root of grain diameter: halve d and the grain-boundary term rises by 41%. Boundaries stop dislocations because the slip planes do not continue across them, so dislocations pile up until the stress starts slip in the next grain; smaller grains mean shorter pile-ups and a higher applied stress. Grain size is reported as the ASTM number G, where a count of N grains per square inch at 100× gives N = 2^(G−1). Move G and watch the micrograph and the yield stress together.
What is solid-solution strengthening?
Dissolved atoms of a different size distort the lattice around them: a smaller substitutional atom pulls its neighbours in, a larger one pushes them out, and an interstitial such as carbon in iron strains it strongly. Dislocations have strain fields of their own, so they are attracted to and held by these sites, and more stress is needed to move them. Brass (zinc in copper) and Monel (copper in nickel) are stronger than their pure base metals for this reason. The effect grows with misfit and with the amount dissolved.
What is strain hardening, and what is the difference between cold and hot working?
As a metal is deformed, dislocations multiply (from about 10⁶ to 10¹⁰ mm⁻²) and get in each other's way, so each further increment of strain needs more stress. Cold working, below the recrystallisation temperature, uses this: cold-drawn wire and cold-rolled sheet are stronger but less ductile, with a good finish and close tolerances. Hot working, above the recrystallisation temperature, recrystallises the metal as it is shaped, so large deformations are possible with low forces and no hardening, at the price of scale and looser tolerances. Cold work is measured by the reduction in area.
What happens in recovery, recrystallisation and grain growth?
Heating a cold-worked metal releases its stored energy in three stages. Recovery: dislocations rearrange into low-energy walls and some internal stress and conductivity recover, with little change in strength. Recrystallisation: new strain-free grains nucleate and grow until they consume the deformed ones, and strength and ductility return to the annealed values. Grain growth: with more time or temperature the boundaries move to reduce their area and the grains coarsen. The recrystallisation temperature is typically 0.3 to 0.5 of the absolute melting point, and falls with more cold work and higher purity.
Why does lead not work-harden at room temperature?
Because room temperature is above its recrystallisation temperature, about −4 °C, so lead recrystallises as fast as it is deformed: bending it at room temperature is hot working. Tin is similar. Aluminium recrystallises at about 80 °C, copper about 120 °C, iron about 450 °C and tungsten about 1,200 °C, so for them room-temperature forming is cold working and they harden. Hot and cold working are defined by the metal's recrystallisation temperature, not by how hot the workshop feels.
A lead strip and a copper strip are bent back and forth at room temperature. Which one gets harder?
Lead recrystallises at about −4 °C, so at room temperature it anneals itself as it deforms. Copper recrystallises at about 120 °C, so at room temperature it is cold worked and strain hardens.
Worked examples
Percent cold work in drawing a rod · Copper rod drawn from 15.2 mm to 12.2 mm diameter
- %CW = (A0 − Ad)/A0 × 100 = (d0² − dd²)/d0² × 100
- %CW = (15.2² − 12.2²)/15.2² × 100 = (231.04 − 148.84)/231.04 × 100 = 35.6%
Answer: 35.6% cold work
Grain size for a target yield stress · Iron: lower yield 135 MPa at d = 0.05 mm and 260 MPa at d = 0.008 mm. Find d for 205 MPa
- k = (260 − 135)/(1/√0.008 − 1/√0.05) = 125/(11.180 − 4.472) = 18.63 MPa·mm^½
- σ0 = 135 − 18.63/√0.05 = 51.7 MPa
- d = (k/(205 − σ0))² = (18.63/153.3)² = 0.0148 mm
- ASTM: G = −6.644 log10(0.0148) − 3.288 = 8.9 (taking d as the mean intercept)
Answer: d ≈ 15 μm, ASTM grain size about 8.9
ASTM grain size number from a count · 45 grains per square inch counted on a micrograph at 100×; what would you count at 250×?
- N = 2^(G−1) at 100× → G = 1 + log2 45
- G = 1 + 5.49 = 6.5
- At 250× each square inch shows (100/250)² of the area: 45 × 0.16 = 7.2 grains per square inch
Answer: G = 6.5; about 7 grains per square inch at 250×
Solidification: nucleation, growth, cooling curves, dendrites and grain size
Which syllabus covers thisAU not in syllabusVTU BME303 Module 3SPPU 211503 Unit IIMU MEC304 Module 2; 2403114 Module IIIKTU MET205 Module 1
Why does a liquid metal have to cool below its melting point before it solidifies?
Because a tiny crystal costs energy before it saves any. Turning liquid into solid releases free energy in proportion to the volume (ΔGv, negative below Tm), but every new crystal also creates a solid-liquid surface, which costs energy in proportion to its area (γ). For a very small cluster the surface term wins, so it melts again. Only clusters bigger than a critical radius r* can grow. At exactly Tm, ΔGv is zero and r* is infinite, so the melt has to undercool first.
What are the critical radius and the activation barrier for nucleation?
Setting dΔG/dr to zero gives the radius where the curve peaks, r*, and the height of that peak, ΔG*, the energy barrier a cluster must climb to become a stable nucleus. ΔGv grows in proportion to the undercooling ΔT, so r* falls as 1/ΔT and ΔG* as 1/ΔT². That is why a little more undercooling makes nucleation suddenly much easier. Slide the undercooling and watch the peak shrink and move left.
What is the difference between homogeneous and heterogeneous nucleation?
Homogeneous nucleation happens in the bulk of a pure liquid with nothing to help it, and needs a large undercooling: a couple of hundred degrees for most metals in the laboratory. In practice nuclei form on mould walls, oxide films or added inoculant particles. A nucleus that wets the surface (contact angle θ) is a spherical cap with less surface to pay for, so the barrier drops to S(θ)·ΔG*, with S between 0 and 1. Real castings solidify with only a few degrees of undercooling for this reason.
What does the cooling curve of a pure metal and of an alloy look like?
A pure metal cools, undercools slightly, then holds a flat thermal arrest at its melting point while latent heat is released, and only cools again once it is all solid. An alloy of two metals that dissolve in each other (like Cu-Ni) freezes over a range instead: the slope changes at the liquidus, where the first solid appears, and again at the solidus, where the last liquid freezes. A eutectic alloy freezes at one temperature, like a pure metal. Plotting arrests and slope breaks for many alloys is how phase diagrams were first drawn.
Why do metals solidify as dendrites, and how is grain size controlled?
In an undercooled melt any bump on a growing crystal pokes into colder liquid and grows faster, so the solid branches into tree-like dendrites along preferred crystal directions until the branches meet. Each nucleus becomes one grain. More nuclei mean more, smaller grains: cool faster (chill moulds), add inoculants (titanium-boron in aluminium, ferrosilicon in cast iron) or stir the melt. Fine grains make a casting stronger and tougher (the Hall-Petch effect) and spread impurities more evenly.
Doubling the undercooling of a pure metal changes the nucleation barrier ΔG* to:
ΔGv is proportional to ΔT and ΔG* = 16πγ³/(3ΔGv²), so ΔG* scales as 1/ΔT². Twice the undercooling gives a quarter of the barrier, and r* halves.
Worked examples
Critical nucleus of gold · Gold: ΔHf = −1.16×10⁹ J/m³, γ = 0.132 J/m², Tm = 1064 °C (1337 K), undercooling 230 °C; lattice parameter at Tm 0.413 nm (FCC)
- r* = −2γTm/(ΔHf ΔT) = 2 × 0.132 × 1337/(1.16×10⁹ × 230) = 1.32 nm
- ΔG* = 16πγ³Tm²/(3ΔHf²ΔT²) = 9.68 × 10⁻¹⁹ J
- Nucleus volume (4/3)π(r*)³ = 9.70 nm³; unit cell 0.413³ = 0.0704 nm³ → 138 unit cells
- 4 atoms per FCC cell → about 551 atoms
Answer: r* = 1.32 nm, ΔG* = 9.68×10⁻¹⁹ J, about 550 atoms (method of Callister Example 10.1)
Copper nucleating on a mould wall · Copper: ΔHf = −1.77×10⁹ J/m³, γ = 0.177 J/m², Tm = 1085 °C, ΔT = 236 °C; contact angle on the wall 60°
- Homogeneous: r* = 2 × 0.177 × 1358/(1.77×10⁹ × 236) = 1.15 nm; ΔG* = 9.82 × 10⁻¹⁹ J
- S(60°) = (2 + 0.5)(1 − 0.5)²/4 = 0.1562
- Heterogeneous barrier = 0.1562 × 9.82 = 1.53 × 10⁻¹⁹ J (same r*, much smaller cap)
Answer: The wall cuts the barrier to 15.6%: 1.53×10⁻¹⁹ J instead of 9.82×10⁻¹⁹ J
Undercooling for a 1 nm nucleus in nickel · Nickel: ΔHf = −2.53×10⁹ J/m³, γ = 0.255 J/m², Tm = 1455 °C (1728 K); target r* = 1 nm
- Rearrange r* = −2γTm/(ΔHf ΔT) for ΔT
- ΔT = 2 × 0.255 × 1728/(2.53×10⁹ × 1×10⁻⁹) = 348 °C
Answer: ΔT ≈ 348 °C, about the largest undercooling seen for nickel (319 °C), so homogeneous nuclei stay above ~1 nm
Solid solutions, Hume-Rothery rules, Gibbs phase rule, the lever rule and isomorphous systems
Which syllabus covers thisAU ME3392 Unit IVTU BME303 Module 2SPPU 211503 Unit IIMU MEC304 Module 2; 2403114 Module IIIKTU MET205 Module 3
What are substitutional and interstitial solid solutions?
A solid solution keeps the host's crystal structure with the second element dissolved in it. In a substitutional solution the solute atoms take the place of host atoms: nickel in copper, zinc in copper (brass), chromium in iron. In an interstitial solution small atoms squeeze into the gaps between host atoms: carbon, nitrogen and hydrogen in iron. Either way the misfit strains the lattice, which is why alloys are stronger than the pure metal (solid-solution strengthening).
What are the Hume-Rothery rules for solid solubility?
Four conditions for two metals to dissolve in each other in all proportions: atomic radii within about 15% of each other; the same crystal structure; similar electronegativity (otherwise they form compounds); and similar valence (a metal dissolves a higher-valence metal more readily than a lower one). Copper and nickel pass all four, so they form one solid solution at every composition. Break one rule and solubility becomes limited, which gives the eutectic and other diagrams of the next chapter.
What is the Gibbs phase rule?
It counts how many variables you can change freely without changing the number of phases. F = C − P + 2, where C is the number of components and P the number of phases; for metals at atmospheric pressure, pressure is fixed and F = C − P + 1. For a binary alloy, a single phase region has F = 2 (vary both temperature and composition), a two-phase region F = 1 (pick the temperature and both phase compositions are fixed), and a three-phase invariant point F = 0.
How do you use the tie line and the lever rule?
In a two-phase region, draw a horizontal tie line through the alloy's point. Where it meets the liquidus is the liquid's composition C_L; where it meets the solidus is the solid's composition C_α. The mass fraction of each phase is the length of the opposite arm over the whole tie line, like a balanced lever. Drag the alloy and temperature and watch the tie line, the compositions and the fractions change. The curves are an ideal-solution shape fitted through the measured 1250 °C tie line, so read other temperatures as approximate.
What is coring and how is it removed?
During slow equilibrium cooling the solid keeps adjusting its composition along the solidus by diffusion. In a real casting cooling is too fast for diffusion in the solid, so the first solid to form (rich in the high-melting element, nickel in Cu-Ni) stays at the centre of each dendrite and the last solid at its edges is leaner. That composition gradient is coring. It lowers the effective solidus, so reheating can melt the grain boundaries. A homogenising anneal below the solidus lets diffusion even it out.
A 35 wt% Ni Cu-Ni alloy sits at 1250 °C in the L + α region. Its solid α is:
The tie line ends on the solidus at 42.5 wt% Ni. Nickel melts higher, so the solid that freezes first takes more of it; the liquid left behind is leaner, 31.5 wt% Ni.
Worked examples
Lever rule in a Cu-Ni alloy · 35 wt% Ni at 1250 °C; tie line meets the liquidus at 31.5 and the solidus at 42.5 wt% Ni
- Phases: liquid + α (two phases, so F = 2 − 2 + 1 = 1)
- W_L = (42.5 − 35)/(42.5 − 31.5) = 0.68
- W_α = (35 − 31.5)/(42.5 − 31.5) = 0.32
Answer: 68% liquid of 31.5 wt% Ni and 32% α of 42.5 wt% Ni (Callister Example 9.2)
Hume-Rothery check: nickel in copper, carbon in iron · Atomic radii: Cu 0.1278, Ni 0.1246, Fe 0.1241, C 0.071 nm; Cu and Ni both FCC, electronegativity 1.9 and 1.8, both valence +2
- Size difference Cu-Ni: (0.1278 − 0.1246)/0.1278 = 2.5%, well inside 15%
- Same structure, similar electronegativity and valence → all four rules pass: complete solubility
- Carbon in iron: radius ratio 0.071/0.1241 = 0.57, far outside 15%, so carbon cannot substitute; it sits in interstitial sites, and only a little fits
Answer: Cu-Ni: unlimited substitutional solubility; C in Fe: limited interstitial solubility
Degrees of freedom across a Cu-Ni diagram · Binary Cu-Ni at 1 atm; points in the liquid, in L + α, and pure copper at its melting point
- Liquid only: F = 2 − 1 + 1 = 2 (temperature and composition both free)
- L + α: F = 2 − 2 + 1 = 1 (fix temperature and both phase compositions follow)
- Pure Cu melting: C = 1, P = 2 → F = 1 − 2 + 1 = 0, so melting happens at one fixed temperature
Answer: F = 2, 1 and 0: why a pure metal shows a flat arrest and an alloy freezes over a range
Invariant reactions: eutectic, eutectoid, peritectic and peritectoid, and the binary eutectic system
Which syllabus covers thisAU ME3392 Unit IVTU BME303 Module 2SPPU 211503 Unit IIMU MEC304 Module 2; 2403114 Module IIIKTU MET205 Module 3
What are the eutectic, eutectoid, peritectic and peritectoid reactions?
Four reactions where three phases meet at one temperature and one composition (F = 0). Eutectic: one liquid freezes into two solids, L → α + β. Eutectoid: the same with a solid on the left, γ → α + β (austenite to pearlite in steel). Peritectic: a liquid and a solid react to give a new solid, L + δ → γ. Peritectoid: two solids give a third, α + β → γ. On cooling, 'eutectic' splits one phase into two; 'peritectic' joins two into one.
How do you read a binary eutectic diagram such as lead-tin?
Lead and tin dissolve each other only partly, so the diagram has a solid α (lead-rich) and β (tin-rich) on each side, two liquidus lines falling to a V, and a horizontal eutectic line at 183 °C from 18.3 to 97.8 wt% Sn. The 61.9 wt% Sn alloy melts at 183 °C, lower than either metal, which is why it was the classic electronic solder. Pick a composition and temperature to read the phases, the tie line and the fractions. Lines here are straight-line approximations through the key points.
What microstructure does a hypoeutectic alloy have?
Between 18.3 and 61.9 wt% Sn, primary α crystals freeze first while the alloy crosses the α + L region. When it reaches 183 °C the liquid that is left has exactly the eutectic composition and freezes all at once into fine alternating layers of α and β: the eutectic microconstituent. So the casting is primary α grains sitting in a lamellar eutectic. The amounts follow from the lever rule just below 183 °C using 18.3 and 61.9 as the tie-line ends. Hypereutectic alloys show primary β instead.
What are the five common types of binary phase diagram?
Grouped by how the two metals behave: (1) complete solubility in liquid and solid, the isomorphous lens (Cu-Ni); (2) complete liquid solubility but no solid solubility, a simple eutectic (Bi-Cd is close); (3) partial solid solubility with a eutectic (Pb-Sn, Al-Si); (4) partial solid solubility with a peritectic (parts of Fe-C, Cu-Zn); (5) systems that form an intermediate compound such as Mg₂Pb, which splits the diagram into two simpler ones. Real diagrams like Fe-C or Cu-Zn combine several of these.
What happens in the peritectic reaction of steel?
At 1493 °C in the iron-carbon system, liquid of 0.53 wt% C reacts with δ-ferrite of 0.09 wt% C to form austenite (γ) of 0.17 wt% C. The new γ forms as a shell around the δ crystals, so the reaction slows down because atoms have to diffuse through that shell. In continuous casting of steels near 0.1-0.15 wt% C this causes uneven shrinkage and surface cracks, which is why steelmakers treat 'peritectic grades' with care.
A 50 wt% Sn solder cools slowly to just below 183 °C. Most of it is:
Primary α = (61.9 − 50)/(61.9 − 18.3) = 0.27, so 73% is eutectic. 50 wt% is closer to the eutectic point (61.9) than to the α limit (18.3).
Worked examples
Lead-tin phases at 150 °C · 40 wt% Sn-60 wt% Pb at 150 °C; at this temperature α holds 11 and β 98 wt% Sn
- The point lies in the α + β region
- W_α = (98 − 40)/(98 − 11) = 0.67
- W_β = (40 − 11)/(98 − 11) = 0.33
Answer: 67% α (11 wt% Sn) and 33% β (98 wt% Sn) (Callister Example 9.2)
Microconstituents of a 50 wt% Sn solder · 50 wt% Sn, just below 183 °C; α 18.3, eutectic 61.9, β 97.8 wt% Sn
- Primary α = (61.9 − 50)/(61.9 − 18.3) = 0.273; eutectic = (50 − 18.3)/(61.9 − 18.3) = 0.727
- Total α (primary + in the eutectic) = (97.8 − 50)/(97.8 − 18.3) = 0.601; total β = 0.399
- α inside the eutectic = 0.601 − 0.273 = 0.328
Answer: 27% primary α + 73% eutectic; 60% α and 40% β in total
Just above the peritectic in steel · Fe-0.17 wt% C just above 1493 °C: δ 0.09 and liquid 0.53 wt% C
- W_L = (0.17 − 0.09)/(0.53 − 0.09) = 0.182
- W_δ = 1 − 0.182 = 0.818
- At 1493 °C all of it reacts, L + δ → γ (0.17 wt% C), because 0.17 is the peritectic composition
Answer: 18% liquid + 82% δ react completely to austenite
The iron-iron carbide diagram: phases, invariant reactions, steels and cast irons
Which syllabus covers thisAU ME3392 Unit IVTU BME303 Module 2SPPU 211503 Unit IVMU MEC304 Module 2; 2403114 Module IVKTU MET205 Module 3
What phases appear in the iron-iron carbide diagram?
Four. α-ferrite: BCC iron that dissolves at most 0.022 wt% carbon (at 727 °C), soft and ductile. γ-austenite: FCC iron, stable above 727 °C, dissolving up to 2.14 wt% C at 1147 °C; every hardening treatment starts from it. Cementite, Fe₃C: a hard, brittle carbide with 6.70 wt% C. δ-ferrite: BCC again, only above 1394 °C. Pure iron changes from BCC to FCC at 912 °C. The diagram stops at 6.70 wt% C because cementite is treated as the second component.
What are the three invariant reactions of the Fe-C system?
Peritectic at 1493 °C: L + δ → γ. Eutectic at 1147 °C: liquid of 4.30 wt% C → γ (2.14) + Fe₃C, a mixture called ledeburite. Eutectoid at 727 °C: austenite of 0.76 wt% C → α (0.022) + Fe₃C, as fine alternating plates called pearlite. The eutectoid matters most for steels, because every steel passes through it on cooling. The A1 line is the eutectoid temperature, A3 the boundary where ferrite starts to form, Acm where cementite starts to form.
How does the microstructure of hypo- and hypereutectoid steels form?
A steel below 0.76 wt% C (hypoeutectoid) crosses A3 on cooling and grows proeutectoid ferrite at the austenite grain boundaries; the remaining austenite enriches in carbon until, at 727 °C, it is 0.76 wt% C and turns to pearlite. Above 0.76 wt% C (hypereutectoid) proeutectoid cementite forms first, as a brittle network on the grain boundaries, then the rest becomes pearlite. Pick a carbon content and temperature to read phases and fractions, and the slowly cooled structure.
How are steels and cast irons classified by carbon?
Up to 0.022 wt% C is commercially pure iron. Steels run from 0.022 to 2.14 wt% C: low carbon (below about 0.25%) for sheet and structures, medium carbon (0.25-0.6%) for shafts, gears and rails, high carbon (0.6-1.4%) for springs, tools and wire. Above 2.14 wt% C the alloy passes the eutectic on freezing and is a cast iron, usually 3.0-4.5% C. Steels are hot-worked and heat-treated; cast irons melt at lower temperatures and are cast into shape.
Is cementite a stable phase?
No. Fe₃C is metastable: given enough time at high temperature, especially with silicon present, it decomposes into iron and graphite. So there are really two diagrams, iron-cementite and iron-graphite, very close to each other. Steels follow the cementite diagram because their carbon is low and cooling fast enough. Grey and ductile cast irons follow the graphite one, which is why they contain graphite flakes or nodules instead of cementite.
A 0.40 wt% C steel is slowly cooled. Roughly how much of it is pearlite?
W_pearlite = (0.40 − 0.022)/(0.76 − 0.022) = 0.51. The pearlite fraction is not the carbon percentage; it is the lever-rule fraction across 0.022-0.76.
Worked examples
Phases and microconstituents of a 0.35 wt% C steel · Fe-0.35 wt% C, slowly cooled, just below 727 °C
- Total ferrite = (6.70 − 0.35)/(6.70 − 0.022) = 0.95; total Fe₃C = 0.05
- Proeutectoid ferrite = (0.76 − 0.35)/(0.76 − 0.022) = 0.56
- Pearlite = (0.35 − 0.022)/(0.76 − 0.022) = 0.44
Answer: 56% proeutectoid ferrite + 44% pearlite (Callister Example 9.4)
A 1.2 wt% C tool steel · Fe-1.2 wt% C, slowly cooled, just below 727 °C
- Pearlite = (6.70 − 1.2)/(6.70 − 0.76) = 0.926
- Proeutectoid cementite = (1.2 − 0.76)/(6.70 − 0.76) = 0.074
- The cementite sits as a network on the old austenite grain boundaries: brittle until it is spheroidised
Answer: 92.6% pearlite + 7.4% grain-boundary cementite
Carbon content from a micrograph · A slowly cooled plain carbon steel shows 25% pearlite and 75% ferrite by area (taking area fraction = mass fraction)
- Hypoeutectoid, so W_pearlite = (C₀ − 0.022)/(0.76 − 0.022)
- C₀ = 0.022 + 0.25 × 0.738 = 0.206 wt% C
Answer: About 0.21 wt% C, a low-carbon steel such as 1020 or 20C8
Heat treatment of steel: annealing, normalising, hardening, tempering, TTT and CCT diagrams, austempering and martempering
Which syllabus covers thisAU ME3392 Unit IIVTU BME303 Module 3SPPU 211503 Unit VMU MEC304 Modules 2-3; 2403114 Modules IV-VKTU MET205 Module 3
What are the types of annealing and normalising?
All start by heating, then differ in temperature and cooling. Full annealing: about 15-40 °C above A3 (hypoeutectoid) or A1 (hypereutectoid), then furnace cooling, giving soft coarse pearlite for machining. Process annealing: below A1 to soften cold-worked parts by recrystallisation. Stress relieving: a few hundred degrees, removing residual stresses from welding or machining. Spheroidising: hours just below A1 (around 700 °C) to ball up cementite, the softest structure. Normalising: about 55 °C above A3 or Acm, then air cooling, for finer pearlite and a refined, uniform grain.
What does a TTT diagram show?
Time-temperature-transformation: quench austenite to a temperature, hold it there, and record when transformation starts and finishes. Both curves form a C shape with a nose, for a eutectoid steel near 540 °C, where transformation is fastest. Just below A1 you get coarse pearlite after a long wait; nearer the nose, fine pearlite; below the nose, bainite; below Ms, martensite, which forms without diffusion. Pick a treatment and watch its cooling path cross the curves. The diagram here is schematic, shaped after the published one for a 1080 steel.
What is martensite, and why does some austenite stay behind?
When austenite is cooled faster than the critical cooling rate (about 140 °C/s for a eutectoid steel), carbon has no time to diffuse out and the FCC lattice shears into a body-centred tetragonal one with the carbon trapped: martensite, very hard and brittle. It forms between Ms and Mf as the temperature drops, not with time. The Koistinen-Marburger relation gives the fraction. In high-carbon steels Mf lies below room temperature, so some retained austenite remains; a sub-zero treatment converts more of it.
Why is hardened steel tempered?
As-quenched martensite is too brittle to use and full of residual stress. Tempering reheats it below A1, usually 250-650 °C: carbon precipitates as fine carbides and the martensite relaxes into tempered martensite, trading some hardness for a lot of toughness. Higher temperature or longer time gives softer, tougher steel; the Hollomon-Jaffe parameter says which combinations are equivalent. Some alloy steels lose toughness if held around 375-575 °C (temper embrittlement), so they are cooled through it quickly.
What are austempering and martempering?
Two quenches that pause in a salt bath. Austempering: quench into a bath above Ms (about 260-400 °C), hold until the austenite has fully turned to bainite, then cool. Bainite is tough, and there is no martensite, so distortion and cracking are low. Martempering (marquenching): quench into a bath just above Ms, hold only until the surface and centre reach the same temperature, then air cool so martensite forms everywhere at once. Hold too long and bainite starts; try both in the interactive.
How is a CCT diagram different from a TTT diagram?
Parts are cooled continuously, not held at one temperature, so a continuous-cooling-transformation diagram is the one to read for real quenches. Its curves sit lower and to the right of the TTT curves, because the steel spends only part of its time at each temperature. For a plain eutectoid steel the CCT diagram shows no bainite at all: pearlite stops forming below about 450 °C and any austenite left becomes martensite. The critical cooling rate is the slowest cooling curve that misses the pearlite nose.
A eutectoid steel bar is quenched in oil, slower than the critical cooling rate but much faster than air. It most likely ends up as:
The cooling curve cuts through the pearlite nose for part of the austenite; the rest reaches Ms and turns to martensite. On continuous cooling a plain eutectoid steel forms no bainite.
Worked examples
Retained austenite and sub-zero treatment · Eutectoid steel, Ms = 220 °C, quenched to room temperature (25 °C), then to −80 °C; Koistinen-Marburger f = 1 − exp(−0.011(Ms − T))
- At 25 °C: f = 1 − exp(−0.011 × 195) = 0.883 → retained austenite 11.7%
- At −80 °C: f = 1 − exp(−0.011 × 300) = 0.963 → retained austenite 3.7%
- M50 = Ms − ln 2/0.011 = 157 °C; M90 = 11 °C
Answer: Sub-zero treatment cuts retained austenite from 12% to 4%
Equivalent tempering by the Hollomon-Jaffe parameter · Tempered 1 h at 600 °C; the shop wants the same hardness at 550 °C (C = 20)
- P = (600 + 273) × (20 + log 1) = 17463
- At 550 °C (823 K): log t = 17463/823.15 − 20 = 1.215
- t = 16.4 h
Answer: About 16.4 h at 550 °C equals 1 h at 600 °C
Reading the schematic TTT diagram · This chapter's schematic eutectoid diagram: hold at 600 °C
- Start curve at 600 °C: about 1.3 s; finish: about 10 s
- 600 °C lies above the nose (540 °C) and below 650 °C, so the product is fine pearlite
- Holding at 400 °C instead (below the nose) would give upper bainite, but only after several seconds
Answer: Fine pearlite, done in about 10 s (schematic times; real 1080 diagrams are similar in shape)
Hardenability, the Jominy end-quench test and surface hardening (carburising, nitriding, cyaniding, induction, flame, laser)
Which syllabus covers thisAU ME3392 Unit IIVTU BME303 Module 3SPPU 211503 Unit VMU MEC304 Modules 2-3; 2403114 Modules IV-VKTU MET205 Module 3
What is the difference between hardness and hardenability?
Hardness is how hard the steel gets where it does turn to martensite, and depends almost only on carbon. Hardenability is how deep martensite forms in a quench, and depends on how slowly the steel transforms to pearlite: alloying elements (Mn, Cr, Mo, Ni) and a coarser grain push the TTT nose to the right, so slower cooling still misses it. A 1045 and a 4140 at similar carbon reach similar surface hardness, but only the 4140 hardens through a thick shaft.
How is the Jominy end-quench test done?
A standard bar 25 mm in diameter and 100 mm long is austenitised, hung in a fixture and quenched by a water jet on one end only. The quenched end cools fastest and the cooling rate falls along the bar. After cooling, two flats are ground along the length and hardness is measured every 1.5 mm (1/16 inch) from the quenched end. A steel whose hardness stays high far along the bar has high hardenability. Steels sold as 'H' grades are guaranteed to fall inside a hardenability band on this curve.
What is the ideal critical diameter, and how is it calculated?
DI is the largest bar that gets 50% martensite at its centre in a perfect quench, one that holds the surface at the bath temperature. Grossmann showed it can be estimated from chemistry: a base value for carbon and grain size, multiplied by one factor per alloying element. A real quench of severity H (still oil about 0.3, still water 1.0, brine 2) hardens only a smaller bar, the actual critical diameter D. Pick a steel and quench to see both. The method is empirical and other DI equations differ slightly.
How do carburising, nitriding, carbonitriding and cyaniding differ?
All give a hard, wear-resistant case on a tough core. Carburising: a low-carbon steel (0.1-0.25% C) is held around 900-950 °C in a carbon-rich gas, pack or liquid until carbon diffuses in, then quenched and tempered; cases are typically 0.5-2 mm. Nitriding: nitrogen diffuses in at about 500-550 °C, below A1, forming hard nitrides with Al, Cr, Mo; no quench, so very little distortion, but thin cases. Carbonitriding adds both carbon and nitrogen in gas; cyaniding does it in a molten cyanide bath.
How do induction, flame, laser and electron-beam hardening work?
They heat only the surface of a medium-carbon or alloy steel into austenite, then quench it, so only the case turns to martensite and the composition does not change. Induction hardening uses an AC coil: eddy currents crowd into a skin whose depth falls with frequency, so a high frequency gives a thin case. Flame hardening uses an oxy-acetylene torch, cheap for large parts. Laser and electron beams heat very thin, precise tracks that self-quench into the cold metal below. Vacuum and plasma (ion) processes carburise or nitride cleanly without oxidising the part.
What is a thermo-mechanical treatment?
Combining deformation and heat treatment in one cycle so the steel gets the benefits of both. Controlled rolling finishes rolling at a low temperature so austenite is fine-grained and gives fine ferrite. Ausforming deforms metastable austenite in the bay of the TTT diagram (around 500 °C) and then quenches it, giving very strong martensite. TMT rebars are quenched by water sprays straight off the mill so the surface turns to martensite, then self-tempered by the hot core: strong, weldable bars without expensive alloying.
The same 4140 shaft is quenched in still oil instead of agitated water. The diameter it can harden at the centre:
DI depends only on chemistry and grain size, but the actual critical diameter depends on quench severity H. Still oil (H ≈ 0.3) removes heat more slowly than agitated water (H ≈ 1.5), so a smaller bar hardens.
Worked examples
Ideal diameter from chemistry · 0.30 C, 0.20 Si, 0.40 Mn, 0.15 Ni, 0.95 Cr, 0.15 Mo (wt%), ASTM grain size 8
- Base DI for 0.30% C, grain size 8: 0.170 in
- Factors: Mn 1 + 3.333 × 0.40 = 2.333; Si 1.14; Ni 1.055; Cr 1 + 2.16 × 0.95 = 3.052; Mo 1.45
- DI = 0.170 × 2.333 × 1.14 × 1.055 × 3.052 × 1.45 = 2.11 in = 54 mm
Answer: DI ≈ 2.1 in (54 mm): no quench can give 50% martensite at the centre of a bigger bar (Herring's worked example)
4140 against 1045 · Typical analyses, grain size 7. 4140: 0.405 C, 0.875 Mn, 0.25 Si, 0.95 Cr, 0.20 Mo. 1045: 0.465 C, 0.75 Mn, 0.25 Si
- 4140: DI = 0.2143 × 3.917 × 1.175 × 3.052 × 1.60 = 4.82 in (122 mm)
- 1045: DI = 0.2295 × 3.500 × 1.175 = 0.94 in (24 mm)
- Agitated oil (H 0.5) on 4140: D = 3.33 in = 85 mm; still water (H 1.0) on 1045: D = 0.41 in = 10 mm
Answer: 4140 hardens a 85 mm bar even in oil; 1045 manages only about 10 mm in water
Induction frequency for a case depth · Steel above its Curie point (μr = 1), resistivity about 1.2 μΩ·m; 10 kHz and 200 kHz generators
- δ = 503 √(ρ/(μr f)) = 503 × √(1.2×10⁻⁶/10⁴) = 5.5 mm at 10 kHz
- At 200 kHz: δ = 1.23 mm
- Case depth scales as 1/√f: 20 times the frequency gives about 1/4.5 of the depth
Answer: About 5.5 mm at 10 kHz (large shafts) and 1.2 mm at 200 kHz (gear teeth, small pins)
Alloy steels and cast irons: alloying elements, stainless, tool, HSLA and maraging steels, grey, white, malleable and SG iron, designations
Which syllabus covers thisAU ME3392 Units I, IIIVTU BME303 Modules 2, 5SPPU 211503 Unit IVMU MEC304 Module 3; 2403114 Module VKTU MET205 Module 4
What does each common alloying element do to steel?
Manganese deoxidises, ties up sulphur as MnS and raises hardenability. Silicon deoxidises and strengthens ferrite (electrical steels use up to about 3%). Chromium raises hardenability, forms hard carbides and, above about 11%, makes steel stainless. Nickel toughens, especially at low temperature, and stabilises austenite. Molybdenum raises hardenability, creep strength and resistance to temper embrittlement. Vanadium and titanium form fine carbides that keep grains small. Tungsten keeps tools hard when hot.
What makes stainless steel stainless, and what is sensitisation?
At least about 11% chromium forms a thin, self-healing chromium oxide film. Ferritic grades (Cr only, BCC) are cheap and magnetic; martensitic grades add carbon so they can be hardened (knives, valves); austenitic grades such as 304 (18% Cr, 8-10% Ni) are FCC, non-magnetic, tough and the most common. If an austenitic grade is held around 500-800 °C, for example beside a weld, chromium carbides form on the grain boundaries and strip the chromium next to them: sensitisation, leading to intergranular corrosion (weld decay). Low-carbon 'L' grades or Ti/Nb-stabilised grades avoid it.
What are tool steels, HSLA and maraging steels?
Tool steels are high-carbon steels with carbide formers. High-speed steel T1 is the classic 18-4-1: 18% tungsten, 4% chromium, 1% vanadium, about 0.7% carbon; it stays hard at the red heat of fast cutting. M2 swaps much of the tungsten for molybdenum. HSLA (high-strength low-alloy) steels are low-carbon steels with tiny additions of Nb, V or Ti for fine grains: strong and weldable for pipelines and car bodies. Maraging steels have about 18% Ni, cobalt, molybdenum and titanium with almost no carbon, aged near 480 °C to very high strength.
How do grey, white, malleable and SG cast irons differ?
By the form of their carbon. Grey iron: carbon as graphite flakes (silicon and slow cooling help), cheap, easy to machine, damps vibration, weak in tension because flakes act as cracks; engine blocks and machine beds. White iron: carbon as cementite (low silicon, fast cooling), hard and brittle, for wear liners. Malleable iron: white iron annealed for a long time at 800-900 °C so cementite becomes graphite rosettes; ductile. SG (ductile) iron: magnesium or cerium added before casting makes graphite form spheres, giving steel-like ductility; crankshafts and pipes.
How are steels designated in IS, SAE/AISI and EN systems?
IS 1570: 40C8 means 0.40% carbon (×100), C for plain carbon steel, and 0.8% manganese (×10). SAE/AISI four digits: the first two give the family (10 plain carbon, 41 Cr-Mo, 43 and 86 Ni-Cr-Mo) and the last two the carbon in hundredths, so 4140 is a 0.40% C Cr-Mo steel. EN: C45 is non-alloy 0.45% C; 42CrMo4 has 0.42% C and the first element's content times 4 (Cr 1%); high-alloy names start with X and give real percentages, as in X5CrNi18-10. Pick a grade to decode it and see its carbon equivalent.
Which is the easiest to weld without preheat?
IIW carbon equivalents at mid-range composition: 20C8 0.33, 4140 0.78, 1080 0.94. Below about 0.40 a steel is usually welded without preheat; higher CE means a harder, crack-prone weld zone.
Worked examples
Decode 40C8 and check its weldability · IS 1570 grade 40C8: C 0.35-0.45, Mn 0.60-0.90 wt%
- 40 → 0.40% C; C → plain carbon steel; 8 → 0.8% Mn
- Mid-range CE = 0.40 + 0.75/6 = 0.525
- 0.40 < CE ≤ 0.60: preheat is usually needed for thick sections
Answer: Medium-carbon steel, CE ≈ 0.53: weldable with preheat
Carbon equivalent of 4140 and 8620 · Mid-range: 4140 C 0.405, Mn 0.875, Cr 0.95, Mo 0.20; 8620 C 0.205, Mn 0.80, Ni 0.55, Cr 0.50, Mo 0.20
- 4140: CE = 0.405 + 0.875/6 + (0.95 + 0.20)/5 = 0.781
- 8620: CE = 0.205 + 0.80/6 + (0.50 + 0.20)/5 + 0.55/15 = 0.515
- Above 0.60 (4140) needs preheat and post-weld heat treatment; 8620 sits in the preheat band
Answer: 4140 CE 0.78 (poor), 8620 CE 0.51 (fair)
Is this grey iron hypo- or hypereutectic? · Grey cast iron: 3.4% C, 2.1% Si, 0.1% P; cast-iron carbon equivalent CE = C + (Si + P)/3
- CE = 3.4 + (2.1 + 0.1)/3 = 4.13
- Eutectic CE is 4.3: below it the iron is hypoeutectic, so primary austenite dendrites freeze first, then eutectic with graphite flakes
Answer: CE 4.13: hypoeutectic grey iron, a typical engine-block composition
Non-ferrous alloys: copper, aluminium and precipitation hardening, titanium, magnesium, nickel superalloys and shape memory alloys
Which syllabus covers thisAU ME3392 Unit IIIVTU BME303 Module 5SPPU 211503 Unit VIMU MEC304 Module 5 (shape memory alloys)KTU MET205 Modules 4, 5
What is the difference between brass and bronze?
Brass is copper with zinc; bronze is copper with tin (and, by extension, copper with aluminium, silicon or phosphorus). Up to about 35% zinc, brass stays a single α phase that cold-works well: 70/30 cartridge brass is deep-drawn into cartridge cases and radiator tubes. Muntz metal (60/40) adds the harder β phase and is hot-worked into valves and fittings. Tin bronzes resist wear and sea water; phosphor bronze makes springs and bushes, and gun metal (Cu-Sn-Zn) makes pump and valve castings. Cupronickel (Cu-Ni, 90/10 or 70/30) resists sea-water corrosion in condenser tubes and coins.
How does precipitation hardening make an aluminium alloy strong?
In three steps. Solution treatment heats the alloy into the single-phase α field so all the copper dissolves. Quenching in water traps the copper in a supersaturated solid solution. Ageing, at room temperature (natural) or at 150-200 °C (artificial), lets the copper cluster into GP zones and then θ″ and θ′ precipitates only a few nanometres across, so closely spaced that they block dislocations. Ageing too long or too hot coarsens them towards equilibrium θ (CuAl₂) and the alloy softens: overageing. Pick the copper content and the ageing temperature and step through the treatment.
What do aluminium alloy designations like 2024-T3 and 6061-T6 mean?
The four digits give the main alloying element: 1xxx pure aluminium, 2xxx copper, 3xxx manganese, 5xxx magnesium, 6xxx magnesium and silicon, 7xxx zinc. The temper after the dash says how it was processed: O annealed, H strain-hardened (non-heat-treatable alloys), T heat-treated. T4 is solution-treated and naturally aged, T6 solution-treated and artificially aged to peak strength, T3 solution-treated, cold-worked and naturally aged. So 2024-T3 is an aircraft-skin Al-Cu-Mg alloy and 6061-T6 is the common peak-aged Al-Mg-Si extrusion alloy. 1xxx, 3xxx and 5xxx cannot be precipitation-hardened.
Where are titanium, magnesium and nickel-based superalloys used?
Titanium alloys (Ti-6Al-4V is half of all titanium used) give steel-like strength at 56% of steel's density, resist corrosion and are body-compatible: compressor blades, airframe forgings, implants. Magnesium is the lightest structural metal, about 1.8 g/cm³, used in die-cast housings, steering wheels and laptop frames; it corrodes easily and is hard to cold-form. Nickel-based superalloys such as Inconel 718 and Nimonic alloys keep their strength above 650 °C because γ′ (Ni₃Al) precipitates resist creep: turbine discs and blades, exhaust parts.
What are bearing alloys, solders and shape memory alloys?
Bearing metals pair hard particles that carry load with a soft matrix that embeds dirt and conforms to the shaft: tin- or lead-based white metals (Babbitt), leaded bronzes and aluminium-tin. Solders are low-melting alloys that wet and join metals: Sn-Pb eutectic melts at 183 °C, and lead-free Sn-Ag-Cu near 217 °C. Shape memory alloys such as nickel-titanium (Nitinol) can be deformed in their low-temperature martensite and spring back to a trained shape when heated into austenite; the same transformation gives superelasticity. They make stents, orthodontic wires, actuators and pipe couplings.
An Al-4% Cu alloy is solution-treated, quenched and then left on the bench at room temperature for four days. Its hardness:
The quenched alloy is supersaturated, so copper starts clustering into GP zones even at room temperature. Natural ageing over a few days gives the T4 temper; rivets of this alloy are kept in a freezer until they are driven, to stop it.
Worked examples
Solution and ageing of Al-4% Cu · C₀ = 4 wt% Cu; solvus fit through 5.65 wt% at 548 °C and 0.45 wt% at 300 °C; θ ≈ 53 wt% Cu; age at 190 °C
- Solvus temperature for 4 wt%: 1/T = 1/821.15 + ln(5.65/4)/B with B = 4,801 K, so T = 775.4 K = 502 °C
- Solidus (straight line 660 → 548 °C): 660 − 19.82 × 4 = 581 °C, so solution-treat between about 505 and 575 °C (2024 is treated near 495 °C)
- Copper still dissolved at 190 °C: C_α = 0.062 wt%
- Lever rule: w_θ = (4 − 0.062)/(53 − 0.062) = 0.0744
Answer: Solution window about 502-581 °C; at most 7.4 wt% θ can precipitate at 190 °C
Specific stiffness and strength of steel, titanium and aluminium · Steel E = 200 GPa, ρ = 7.85 g/cm³, σy = 250 MPa; Ti-6Al-4V 114 GPa, 4.43 g/cm³, 880 MPa; Al 6061-T6 70 GPa, 2.70 g/cm³, 275 MPa
- E/ρ: steel 25.5, titanium 25.7, aluminium 25.9 GPa per g/cm³
- σy/ρ: steel 31.8, titanium 198.6, aluminium 101.9 MPa per g/cm³
- So for a stiffness-limited tie all three weigh about the same; for a strength-limited one titanium is about six times lighter than mild steel
Answer: Specific stiffness is almost identical (≈ 26); specific strength is where titanium and aluminium win
Non-metallic materials: polymers, engineering ceramics, intermetallics, composites, nano and smart materials
Which syllabus covers thisAU ME3392 Unit IVVTU BME303 Module 5SPPU 211503 Unit VIMU MEC304 Modules 5, 6KTU MET205 Module 5
What is the difference between thermoplastics and thermosets?
Thermoplastics are long chains held together by weak secondary bonds, so they soften when heated and can be remoulded and recycled: polyethylene (PE) and polypropylene (PP) for packaging and tanks, PVC for pipes, PS, PMMA (acrylic) for clear covers, PET for bottles and fibres, polycarbonate (PC) for tough transparent parts, nylon (PA) for gears and bushes, ABS for housings, and high-performance PEEK, PPS and PTFE. Thermosets cross-link into one network when they cure and char instead of melting: epoxy, phenol-formaldehyde (Bakelite), urea-formaldehyde, polyester resins. Elastomers are lightly cross-linked chains that stretch and recover.
What is the glass transition temperature of a polymer?
The temperature at which the amorphous part of a polymer changes from a hard, glassy solid to a soft, rubbery one, because the chains gain enough energy to slide past each other in segments. Its stiffness can drop a thousandfold across Tg. Below Tg, PMMA and polystyrene (Tg near 100 °C) are rigid and brittle; above it, polyethylene (Tg far below room temperature) is soft and flexible. Crystalline regions also melt at Tm, above Tg. A plastic part is designed to work below Tg if it must be stiff, or above it if it must flex.
What are engineering ceramics and where are they used?
Ionic or covalent compounds that are hard, stiff, light and stable at high temperature, but brittle because their bonds do not let dislocations move. Alumina (Al₂O₃) makes spark-plug insulators, cutting inserts and wear liners. Silicon carbide (SiC) makes abrasives, seals and kiln parts. Silicon nitride (Si₃N₄) makes bearing balls and cutting tools. Partially stabilised zirconia (PSZ) resists cracking because the stress at a crack tip transforms its grains and squeezes the crack shut; it makes dies and implants. SiAlON combines Si₃N₄ toughness with alumina's oxidation resistance. Intermetallics such as Ni₃Al and TiAl sit between metals and ceramics.
How do you calculate the stiffness of a fibre composite by the rule of mixtures?
Loaded along the fibres, fibre and matrix stretch by the same strain, so the modulus is the volume-weighted average (Voigt): stiff fibres carry almost all the load. Loaded across the fibres, both carry the same stress and their strains add, so the soft matrix controls the modulus (Reuss). A carbon/epoxy lamina is about 16 times stiffer along its fibres than across them, which is why real parts stack plies at 0°, ±45° and 90°. Density is a simple average too. Pick a fibre and change the volume fraction.
What are the types of composite materials?
By matrix: polymer matrix composites (PMC, glass or carbon in epoxy or polyester: boat hulls, wind blades, aircraft and car panels), metal matrix composites (MMC, SiC particles or fibres in aluminium: brake discs, pistons) and ceramic matrix composites (CMC, SiC fibres in SiC: turbine shrouds, brake discs that survive 1,000 °C+, where the fibres make an otherwise brittle ceramic tough). By reinforcement: particulate (concrete, cermets, carbide tools), short fibre, continuous fibre, and structural forms such as laminates and sandwich panels. Nano composites add particles below 100 nm, such as nanoclay or carbon nanotubes.
What are nanomaterials and smart materials?
Nanomaterials have at least one dimension below 100 nm, so a large share of their atoms sit at surfaces or grain boundaries and their properties change: nanocrystalline metals are stronger (Hall-Petch), nanoparticles are more reactive, carbon nanotubes and graphene are extremely stiff and strong. They are made top-down (ball milling, lithography) or bottom-up (chemical vapour deposition, sol-gel). Smart materials respond to a stimulus with a useful change: shape memory alloys to temperature, piezoelectrics such as PZT to voltage or force, magnetostrictive alloys to magnetic field, and electro- or magneto-rheological fluids that stiffen on command.
A unidirectional carbon/epoxy lamina (60% fibre) is loaded across the fibres instead of along them. Its stiffness is roughly:
Across the fibres the soft epoxy carries the full stress: E₂ ≈ 8.6 GPa against E₁ ≈ 139 GPa along them, a ratio of about 16.
Worked examples
Carbon/epoxy lamina by the rule of mixtures · Carbon fibre E = 230 GPa, ρ = 1.76 g/cm³; epoxy E = 3.5 GPa, ρ = 1.20 g/cm³; Vf = 0.6
- E₁ = 0.6 × 230 + 0.4 × 3.5 = 139.4 GPa
- E₂ = 1/(0.6/230 + 0.4/3.5) = 8.55 GPa
- ρ = 0.6 × 1.76 + 0.4 × 1.20 = 1.536 g/cm³
- E₁/ρ = 90.8, against steel 210/7.85 = 26.8: 3.4 times
Answer: E₁ ≈ 139 GPa, E₂ ≈ 8.6 GPa, 3.4 times the specific stiffness of steel along the fibres
How much load the glass fibres carry · E-glass 72 GPa, epoxy 3.5 GPa, Vf = 0.5, load along the fibres
- Equal strain, so each phase's share of the load is E·V over the total
- E₁ = 0.5 × 72 + 0.5 × 3.5 = 37.75 GPa
- Fibre share = 36/37.75 = 0.954
Answer: The fibres carry 95.4% of the load although they are half the volume
Degree of polymerisation of polyethylene · Number-average molar mass 100,000 g/mol; repeat unit C₂H₄ (C = 12.011, H = 1.008 g/mol)
- Repeat-unit mass = 2 × 12.011 + 4 × 1.008 = 28.054 g/mol
- n = 100,000/28.054 = 3,565
Answer: About 3,565 repeat units per chain
Powder metallurgy, surface coatings and materials selection: property charts and performance indices
Which syllabus covers thisAU ME3392 Unit II (sintering); Unit III (materials standards)VTU BME303 Modules 4, 5SPPU 211503 Unit IIIMU not in syllabusKTU not in syllabus
What are the steps of the powder metallurgy process?
Make the powder, blend it, compact it, sinter it, and finish it. Powders are made by atomising molten metal with water or gas jets, by mechanical methods (milling, crushing, machining chips), by chemical reduction of oxides, or by electrolysis. They are characterised by particle size and its distribution, shape, flow rate and apparent density. Blending adds a lubricant (zinc stearate) and alloying powders. Compaction in a die at 100-800 MPa makes a fragile green part. Sintering below the melting point (about 1,120 °C for iron) bonds the particles by diffusion and the part shrinks a little. Sizing, infiltration or oil impregnation finish it.
What are the advantages and applications of powder metallurgy?
Near-net shapes with almost no scrap, controlled porosity, and materials that cannot be melted and cast together. Self-lubricating bronze bushes hold oil in 20-30% porosity. Cemented carbide tools are tungsten carbide grains bonded by cobalt; cermets bond ceramic grains with metal. Tungsten and molybdenum filaments and electrical contacts of silver-tungsten or copper-tungsten combine refractory metals with good conductors. Friction materials for clutches and brakes, diamond-impregnated tools, and car parts such as gears, sprockets and connecting rods are made this way. Limits: die cost needs volume, parts are size-limited and residual porosity lowers toughness.
What are the main surface coating processes?
Electroplating deposits nickel, chromium, zinc or copper from a salt solution onto a part made the cathode: cheap, for corrosion protection and appearance. Chemical vapour deposition (CVD) reacts gases at 800-1,000 °C so a hard film such as TiC, TiN or Al₂O₃ grows on carbide inserts. Physical vapour deposition (PVD) evaporates or sputters the coating material in vacuum at 200-500 °C, low enough for hardened steel tools and moulds: the gold TiN on drills. High-velocity oxy-fuel (HVOF) spraying fires molten or semi-molten powder at supersonic speed to build thick, dense WC-Co or chromium carbide wear coatings on shafts and valves.
How do you select a material using an Ashby chart and a performance index?
Write down the function, objective, constraints and free variables. For a light, stiff beam of fixed length and stiffness, with the section size free, the mass is lowest for the material with the biggest E^½/ρ. On a log E against log ρ chart, every line of slope 2 has a constant E^½/ρ, so slide the line up: the materials left above it are the best. A light stiff tie uses E/ρ (slope 1), a plate E^⅓/ρ (slope 3). Then screen on cost, temperature and corrosion, and look up full data for the shortlist. VTU's lab does the same with GRANTA EduPack. Typical mid-range values are used here.
Where does materials data for design come from?
Early in design, from broad property charts and generic databases that give ranges; once a material is shortlisted, from standards and supplier datasheets with guaranteed minimum values (IS 2062 for structural steel, IS 1570 and ASTM A29 for carbon and alloy steels, the Aluminum Association for aluminium tempers); and for critical parts, from your own tests on the actual batch, because a datasheet's typical value is not a design allowable. Databases such as GRANTA, MatWeb and MMPDS (for aerospace allowables) organise this. Always note the condition: temper, heat treatment, product form and test direction change the numbers.
For the lightest beam of a given stiffness (section size free), which wins: mild steel, aluminium 6061 or spruce wood?
The index is E^½/ρ: steel 1.85, aluminium 3.08, spruce 7.03. A spruce beam of the same stiffness weighs about 26% of the steel one; it is just much deeper.
Worked examples
The lightest stiff beam · Index E^½/ρ; typical E (GPa), ρ (g/cm³): mild steel 210, 7.85; Al 6061 69, 2.70; CFRP 70, 1.6; spruce 10, 0.45
- Steel √210/7.85 = 1.846; aluminium 3.077; CFRP 5.229; spruce 7.027
- Mass for the same stiffness ∝ 1/index
- Relative to steel: aluminium 0.60, CFRP 0.35, spruce 0.26
Answer: Spruce beam 26%, CFRP 35% and aluminium 60% of the steel beam's mass
The lightest stiff tie · Index E/ρ, same data plus Ti-6Al-4V 114, 4.43 and alumina 370, 3.9
- Steel 26.8, aluminium 25.6, titanium 25.7: the three common metals are almost equal
- CFRP 43.8; alumina 94.9
Answer: Alumina and CFRP win in tension stiffness; steel, aluminium and titanium tie
Sintering shrinkage of an iron compact · Green density 6.6 g/cm³, sintered density 7.2 g/cm³, wrought iron 7.87 g/cm³; mass unchanged
- Volume scales as 1/density, length as its cube root
- Linear shrinkage = 1 − (6.6/7.2)^⅓ = 2.86%
- Porosity left = 1 − 7.2/7.87 = 8.5%
Answer: The die must be about 2.9% oversize; the part keeps 8.5% porosity
Mechanical testing: tension and compression, Brinell, Vickers and Rockwell hardness, Izod and Charpy impact, and non-destructive testing
Which syllabus covers thisAU ME3392 Unit VVTU BME303 practical componentSPPU 211503 Unit IMU MEC304 Module 6 (NDT)KTU not in syllabus
What properties does a tensile test give?
A standard specimen is pulled in a universal testing machine while load and extension are recorded, giving the engineering stress-strain curve. From it: Young's modulus (slope of the elastic line), yield strength (or 0.2% proof stress when there is no clear yield point), ultimate tensile strength at the peak, percentage elongation and reduction in area at fracture (ductility), and toughness as the area under the curve. Mild steel shows an upper and lower yield point; aluminium yields gradually; cast iron fractures with almost no plastic strain. Pull each specimen and watch the regions and the energy absorbed.
How is a compression test different from a tension test?
The specimen is a short cylinder (height about 1.5-3 times the diameter, so it does not buckle) squeezed between platens. Ductile metals flatten and barrel because friction at the platens holds the ends, and they never break, so only yield strength is meaningful. Brittle materials such as cast iron, concrete and ceramics are tested in compression because that is how they are used: grey cast iron is three to four times stronger in compression than in tension, and fails on an inclined shear plane. The Erichsen cupping test is a related sheet-metal check of formability.
How are Brinell and Vickers hardness numbers calculated?
Both divide the test load by the area of the permanent indent. Brinell presses a 10 mm hard-metal ball with 3,000 kgf for steel (500-1,500 kgf for softer metals) and measures the indent diameter; the large indent averages over coarse structures such as cast iron. Vickers presses a 136° diamond pyramid and measures the diagonals; one scale covers soft to fully hardened steel, and light loads give micro-hardness of single phases or thin cases. Change the load and indent size and watch the number. For steels, UTS in MPa is roughly 3.45 times HB, an approximation only.
How does the Rockwell hardness test work?
Rockwell measures depth, not area, so it reads directly on a dial with no microscope. A minor load of 10 kgf seats the indenter, the major load is added and removed, and the remaining extra depth gives the number. Rockwell C uses a diamond cone with 150 kgf total for hardened steel; Rockwell B uses a 1/16-inch ball with 100 kgf for softer steel, brass and aluminium. It is fast and suits production checks. Nano-indentation records load against depth with forces in millinewtons to measure hardness and modulus of thin films and individual grains.
What is the difference between the Izod and Charpy impact tests?
Both break a notched bar with a swinging pendulum and measure the energy it absorbs from the height lost: E = mgL(cos β − cos α). In Charpy the 55 mm bar lies horizontally as a simply supported beam and is struck behind the notch; in Izod it is clamped vertically as a cantilever and struck on the notch side. Charpy is easy to cool or heat, so it is used to find the ductile-to-brittle transition temperature of steels. Results compare materials and batches; they are not design values like fracture toughness.
What are the common non-destructive testing methods?
Dye penetrant: a coloured or fluorescent liquid seeps into surface-breaking cracks and a developer draws it out, for any non-porous material. Magnetic particle: a magnetised ferromagnetic part leaks flux at surface and near-surface cracks, which collect iron powder. Ultrasonic: a 1-10 MHz pulse reflects from internal flaws and the echo time gives their depth, for welds, forgings and thickness gauging. Eddy current: a coil's impedance changes near cracks in conductive parts, for tubes and aircraft skins. Radiography: X-rays or gamma rays record porosity and inclusions in castings and welds on film or a detector.
Two steels are Brinell tested at 3,000 kgf with a 10 mm ball. Steel A leaves a 3.5 mm indent, steel B a 4.5 mm indent. Which is harder?
Same load, smaller area of indent, higher pressure carried: A is about 302 HBW and B about 179 HBW.
Worked examples
Brinell hardness and estimated UTS of a steel · F = 3,000 kgf, D = 10 mm, indent d = 4.0 mm
- √(D² − d²) = √(100 − 16) = 9.165 mm
- HB = 6,000/(π × 10 × 0.835) = 228.8
- Indent depth = (10 − 9.165)/2 = 0.417 mm
- UTS ≈ 3.45 × 229 = 789 MPa (approximate)
Answer: 229 HBW 10/3000; UTS roughly 789 MPa
Vickers hardness from two diagonals · F = 30 kgf; diagonals 0.45 mm and 0.47 mm
- Mean diagonal d = 0.46 mm
- HV = 1.8544 × 30/0.46² = 262.9
- Depth = d/7 = 65.7 µm, so the specimen (or case) should be at least 1.5d = 0.69 mm thick
Answer: 263 HV30
Energy absorbed in a Charpy test · Pendulum 22 kg, arm 0.8 m to the striker, released at 140° from the hanging position, rises to 60° after breaking the bar
- Height at release = 0.8(1 − cos 140°) = 1.413 m; after = 0.8(1 − cos 60°) = 0.400 m
- E = 22 × 9.81 × (1.413 − 0.400) = 218.6 J
Answer: The bar absorbed about 219 J (friction and windage are usually checked by a free swing and subtracted)
Fracture: ductile and brittle fracture, the ductile-to-brittle transition, Griffith theory, Orowan's modification and fracture toughness
Which syllabus covers thisAU ME3392 Unit VVTU not in syllabusSPPU not in syllabusMU MEC304 Module 4; 2403114 Module VIKTU MET205 Modules 2, 5
What is the difference between ductile and brittle fracture?
Ductile fracture comes after large plastic deformation: the bar necks, voids form at inclusions in the neck, grow and join, and the last ring shears at 45°, leaving a cup-and-cone surface with dimples under the microscope. It gives warning and absorbs a lot of energy. Brittle fracture runs fast with little plastic flow, on a flat surface, often by cleavage across grains (chevron or river marks pointing back to the origin) or along grain boundaries. It gives no warning. Glass, cast iron and steels below their transition temperature break this way.
What is the ductile-to-brittle transition temperature (DBTT)?
The temperature range over which the Charpy impact energy of a material falls from a high ductile shelf to a low brittle one. Body-centred cubic metals, ordinary ferritic steels above all, show it, because their yield strength rises steeply at low temperature until cleavage wins over slip. Face-centred cubic metals such as aluminium, copper and austenitic stainless steel do not, so they are used for cryogenic tanks. Liberty ships cracked in two in cold North Atlantic water during the Second World War for this reason. Fine grains, nickel and low carbon lower the DBTT.
What is Griffith's theory of brittle fracture?
Griffith explained why glass breaks at a hundredth of its theoretical strength: it contains tiny cracks, and a crack grows when the elastic energy it releases is enough to create its two new surfaces. That gives a fracture stress that falls with the square root of crack length, so doubling the crack lowers the strength by √2. Orowan noticed that metals absorb far more energy in plastic work at the crack tip, γp about a thousand times γs, and added it to the surface energy. Pick glass and compare Griffith's curve with the measured toughness.
What is fracture toughness KIc and how is it used?
The stress intensity factor K = Yσ√(πa) measures how severely a crack of size a concentrates an applied stress σ; Y depends on geometry (1 for a centre crack in a wide plate, 1.12 for an edge crack). Fracture happens when K reaches the material's plane-strain fracture toughness KIc. So for a given stress you can find the largest crack the part tolerates, and size inspection intervals around it. 7075-T651 aluminium has KIc about 24 MPa√m; 4340 steel tempered at 260 °C about 50, and at 425 °C about 87 MPa√m, trading strength for toughness.
How does drilling a hole at a crack tip stop the crack?
A sharp crack tip has a tiny radius ρ, so the local stress is many times the applied stress: σmax = σ(1 + 2√(a/ρ)) by Inglis. Drilling a round hole at the tip replaces that radius with the hole's radius, millimetres instead of micrometres, and the peak stress falls by an order of magnitude, often below what is needed to restart the crack. It is a temporary repair on aircraft skins, castings and bridges until the part is replaced; the crack must be found to its true end first, usually with dye penetrant.
A brittle plate breaks at 100 MPa with a crack of length a. If the crack grows to 4a, it breaks at:
Fracture stress scales as 1/√a. Four times the crack, half the stress: 100/√4 = 50 MPa.
Worked examples
Largest surface flaw a glass plate can carry · Glass E = 69 GPa, specific surface energy γs = 0.3 J/m², tensile stress 40 MPa
- Griffith: a = 2Eγs/(πσ²)
- a = 2 × 69×10⁹ × 0.3/(π × (40×10⁶)²) = 8.24 µm
- Equivalent toughness √(2Eγs) = 0.20 MPa√m, below the measured ≈ 0.75 MPa√m for glass, so the real tolerance is a little larger
Answer: Any surface scratch deeper than about 8.2 µm breaks the plate at 40 MPa
Tempering 4340 steel for crack tolerance · Edge crack (Y = 1.12), design stress 800 MPa; 4340 tempered at 260 °C: KIc = 50 MPa√m; at 425 °C: 87.4 MPa√m
- 260 °C temper: a_c = (1/π)(50/(1.12 × 800))² = 0.99 mm
- 425 °C temper: a_c = (1/π)(87.4/896)² = 3.03 mm
- Both tempers have a yield strength well above 800 MPa (1,640 and 1,420 MPa)
Answer: The tougher temper tolerates a 3.0 mm crack instead of 1.0 mm, three times easier to find before it fails
Crack-arrest hole · Edge crack 10 mm long, tip radius about 0.01 mm; a 6 mm hole (ρ = 3 mm) is drilled at the tip
- Sharp tip: σmax/σ = 1 + 2√(10/0.01) = 64.2
- With the hole: σmax/σ = 1 + 2√(10/3) = 4.65
Answer: Peak stress drops from 64 to 4.7 times the applied stress, about 14 times lower
Fatigue and creep: S-N curves, endurance limit, mean stress, creep stages, Larson-Miller parameter and superplasticity
Which syllabus covers thisAU ME3392 Unit VVTU not in syllabusSPPU 211503 Unit I (fatigue and creep tests)MU MEC304 Module 4; 2403114 Module VIKTU MET205 Module 5
What is fatigue failure and how does it start?
Failure under repeated stress well below the yield strength. Slip under cyclic load builds persistent slip bands that form intrusions and extrusions at the surface; a crack starts there or at a notch, scratch or inclusion, grows a little each cycle and leaves beach marks and microscopic striations; finally the remaining section breaks suddenly. Most broken shafts, springs, axles and welded joints fail this way. Fatigue life improves with a smooth surface, generous fillets, compressive residual stress from shot peening, surface hardening and clean steel; corrosion and tensile residual stress shorten it.
What is an S-N curve and the endurance limit?
Specimens are cycled at different stress amplitudes, often in a rotating-beam machine, and the cycles to failure are plotted on a log scale. Steels and titanium flatten out: below the endurance limit, roughly half the ultimate strength for a polished steel specimen, they survive indefinitely (10⁷ cycles and beyond). Aluminium, copper and magnesium alloys keep falling, so they are quoted as a fatigue strength at a stated life, such as 138 MPa at 5×10⁸ cycles for 2024-T3. A mean tensile stress shortens life; the Goodman, Gerber or Soderberg line converts it to an equivalent reversed stress.
What is creep and what are its three stages?
Slow, permanent deformation under constant stress at high temperature, important above about 0.4 of the melting point in kelvin (around 400 °C for steels, room temperature for lead and solder). Primary creep starts fast and slows as the metal work-hardens. Secondary, steady-state creep runs at a constant minimum rate where hardening and recovery balance; designers use this rate. Tertiary creep accelerates as voids form at grain boundaries and the section necks, ending in rupture. Change the stress and temperature and watch the rate and the life move by orders of magnitude.
What is the Larson-Miller parameter used for?
To trade temperature against time in creep. It combines the absolute temperature T and rupture time tr into one number, P = T(C + log tr) with C about 20 for steels, and for a given material and stress P is roughly constant. So a short test at a higher temperature predicts the life at a lower service temperature: 100 hours at 700 °C has the same P as about 33,000 hours at 600 °C. Plotting stress against P puts all test temperatures on one master curve for design.
What makes a material creep resistant, and what is superplasticity?
Creep runs by dislocation climb, by diffusion through grains and along grain boundaries, and by grain-boundary sliding. Resistant alloys block these: a high melting point, stable fine precipitates such as γ′ in nickel superalloys, solid-solution elements like Mo and W, and coarse grains or single crystals with no grain boundaries across the load (turbine blades). Superplasticity is the opposite: very fine, stable grains (under 10 µm) at about half the melting point and a slow strain rate let grains slide past each other, so alloys such as Ti-6Al-4V and some aluminium alloys stretch several hundred percent without necking, used to form complex aerospace panels.
One steel shaft and one aluminium shaft are each cycled at a stress amplitude below the steel's endurance limit for 10⁹ cycles. Which is more likely to fail?
Steel has a true endurance limit; aluminium has none, so given enough cycles any amplitude eventually cracks it. Aluminium aircraft structure is therefore given a finite life and inspected.
Worked examples
Fatigue life with and without a mean stress · Polished steel, Su = 600 MPa; Shigley estimate: S = 0.9Su at 10³ cycles, Se = 0.5Su at 10⁶; σa = 400 MPa, then σm = 100 MPa added
- b = −(1/3) log(540/300) = -0.08509; a = 540²/300 = 972 MPa
- Zero mean: N = (400/972)^(1/b) = 34,017 cycles
- Goodman: σar = 400/(1 − 100/600) = 480 MPa
- N = (480/972)^(1/b) = 3,992 cycles
Answer: A 100 MPa mean stress cuts the life from about 34,017 to 3,992 cycles, more than eight times
Accelerated creep test by Larson-Miller · C = 20; a specimen ruptures in 100 h at 700 °C; service at 600 °C at the same stress
- P = 973.15 × (20 + log 100) = 21,409 K·h
- At 873.15 K: log tr = 21,409/873.15 − 20 = 4.520
- tr = 33,084 h
Answer: About 33,084 h, nearly 3.8 years in service
Test temperature for a 100,000 h design life · Service 600 °C for 100,000 h; test wanted to finish in 1,000 h; C = 20; and Norton exponent n = 5
- P = 873.15 × (20 + 5) = 21,829 K·h
- T = P/(20 + 3) = 949.1 K = 676 °C
- Separately, with n = 5 doubling the stress raises the steady creep rate 32 times
Answer: Test at about 676 °C; and stress, not only temperature, must be held to service values
Keep going
- How heat treatment transforms steel
- What makes alloys so strong? The role of solid solutions
- How do composite materials boost engineering efficiency?
- Strength of Materials interactive book
- Machining Science, Machine Tools and Metrology interactive book
- Learn Strength of Materials: a step-by-step path
- All Skill-Lync interactive books
Recommended reading
- Materials Science and Engineering: An Introduction, W. D. Callister and D. G. Rethwisch
- Introduction to Physical Metallurgy, S. H. Avner
- Materials Science and Engineering: A First Course, V. Raghavan
- Material Science and Metallurgy for Engineers, V. D. Kodgire and S. V. Kodgire
- Engineering Materials: Properties and Selection, K. G. Budinski and M. K. Budinski
- Mechanical Metallurgy, G. E. Dieter
- Engineering Metallurgy, Part I: Applied Physical Metallurgy, R. A. Higgins
- Materials Selection in Mechanical Design, M. F. Ashby
How this book was made
Chapters follow the official syllabi listed above. Every interactive's formulas and every worked-example number were checked by script. Explanations, animations and examples are original to Skill-Lync. Topics informed by MIT OpenCourseWare 3.091 (Introduction to Solid-State Chemistry), CC BY-NC-SA. All explanations, animations and examples are original to Skill-Lync; MIT does not endorse this page.