Ceramics — alumina, zirconia, silicon carbide, silicon nitride; properties at high T: Question bank

Definition Symbols you'll need before you start

These questions reuse three ideas from the parent's bond-potential model. Meet them once, in plain words, so no symbol appears "cold" below.

$x$ is how far an atom sits from its equilibrium bond length $x_{0}$ — the spacing where the bond is happiest, the bottom of the energy valley in Figure 1. So $x = 0$ means "resting spot", $x > 0$ means "stretched", $x < 0$ means "squeezed".
$c$ is the bond stiffness (the springiness): a big $c$ makes a steep, narrow valley — atoms barely move. $g$ measures the lopsidedness (anharmonicity) of that valley: it is steeper on the squeeze side than the stretch side. Both are drawn in Figure 1.
$k_{B}$ is Boltzmann's constant, the fixed number ( $1.38 \times 1 0^{- 23}$ J/K) that converts a temperature $T$ into a typical thermal energy $k_{B} T$ — literally "how much jiggle energy one degree of temperature buys an atom". It shows up whenever heat meets a single atom.

With those, the parent's expansion formula $α = \frac{3 g k _{B}}{4 c ^{2} x _{0}}$ reads plainly: more lopsidedness $g$ → more expansion; stiffer bonds (big $c$ ) or longer bonds (big $x_{0}$ ) → less expansion.

True or false — justify

A ceramic with the highest melting point automatically has the best thermal-shock resistance.

False. Shock resistance is

R = \frac{σ _{f} ( 1 - ν )}{E α}

— strength, stiffness and expansion — while melting point comes from bond-well depth; these are different physics. SiC has a huge

T_{m}

yet dense alumina (also high

T_{m}

) cracks under modest shock.

Ceramics are "weak materials" because they shatter.

False. They have enormous compressive strength and stiffness; they only fail easily in tension, where a surface flaw opens into a running crack (see Thermal Stress and Fracture Mechanics).

Because they are already oxidised, oxide ceramics like alumina and zirconia cannot degrade in hot air.

True for oxidation specifically — they can't burn further — but they can still fail by thermal shock, creep, or phase change; being oxide-safe is only one hazard removed.

Adding ~3% yttria to zirconia raises its melting point.

False. Yttria stabilises the tetragonal phase so transformation toughening works and the destructive 4% volume jump on cooling is suppressed — it is phase/toughness control, not a melting-point trick.

Silicon carbide cannot be used in air because it oxidises.

False. In hot air it undergoes passive oxidation, growing a protective, self-healing

S i O_{2}

glass skin up to ~1600 °C; only in low-oxygen "active oxidation" does it form gaseous SiO and erode.

Metals bend and ceramics crack for essentially the same reason.

False. Metals flow because a non-directional electron sea lets planes slide (dislocation motion); ceramics can't slide planes without pushing like charges together or snapping rigid covalent bonds, so cracks propagate instead. See Crystal Bonding — Ionic vs Covalent.

Zirconia is the toughest common ceramic because it has the strongest bonds.

False. Its toughness comes from transformation toughening — a stress-triggered tetragonal→monoclinic change at the crack tip whose 4% expansion squeezes the crack shut, adding

Δ K_{t r an s f or ma t i o n}

K_{I c}

(see the crack-tip sketch, Figure 2). Bond strength alone doesn't explain it.

A low thermal expansion coefficient

α

makes a ceramic a good thermal barrier.

False. A barrier needs low thermal conductivity

k

(blocks heat flux

q = k Δ T / L

); low

α

helps it survive shock and stick to the blade, but zirconia coats blades because of low

k

Spot the error

"Alumina's

k \approx 30

W/m·K and SiC's

k \approx 120

, so SiC is the better thermal barrier coating."

Error: barriers want LOW $k$ . High conductivity means heat pours straight through; that's why zirconia (

k \approx 2

) — not SiC — is used for TBCs.

"We picked silicon nitride for the coating because it has the highest melting point."

Error.

S i_{3} N_{4}

is chosen for thermal-shock resistance (tiny

α \approx 3 \times 1 0^{- 6}

/K → large

R

); it also doesn't melt but decomposes/sublimes, so "melting point" isn't even the right concept.

"The

- g x^{3}

term in the bond potential is a small correction we can ignore for expansion."

Error. That lopsidedness (

g

) term is thermal expansion — a purely harmonic well

\frac{1}{2} c x^{2}

(the symmetric valley in Figure 1) gives

⟨ x ⟩ = 0

at every

T

, so a symmetric bond would not expand at all.

"Zirconia is toughened, so we can freely thermally cycle pure

Z r O_{2}

blades."

Error. Pure zirconia undergoes the 4% volume tetragonal↔monoclinic jump on every cool-down and self-destructs; you need partially stabilised (yttria-doped) zirconia for the toughening to be safe. See Phase Transformations.

"Ceramics have low

α

, so thermal stress inside them is always negligible."

Error. Thermal stress is

σ_{t h} = \frac{E α Δ T}{1 - ν}

; ceramics have huge

E

, so even small

α

times a large

Δ T

can exceed

σ_{f}

(that's exactly why alumina cracks at ~77 K jumps).

"A CMC is just a solid ceramic block, so it fails as brittly as monolithic alumina."

Error. In a CMC, fibres bridge and deflect cracks, giving graceful, non-catastrophic failure — the whole point is to defeat monolithic brittleness.

Why questions

Why do deep, directional bonds give ceramics both high melting points and low thermal expansion?

A deep valley (Figure 1) needs more jiggle energy

k_{B} T

to escape (high

T_{m}

), and a stiff (large

c

) valley makes

α = \frac{3 g k _{B}}{4 c ^{2} x _{0}}

small — the same bond stiffness that raises

T_{m}

shrinks

α

Why does the

S i O_{2}

skin on SiC protect it rather than keep eating into it?

The glass layer is dense and adherent, sealing the surface from oxygen (passivating), so oxidation is self-limiting — much like aluminium's oxide film.

Why does a low-

α

TBC still need its

α

close to the blade's, not simply as low as possible?

On heating/cooling the coating and metal must expand together; a large

α

mismatch builds interfacial stress and the coating spalls (peels) off the superalloy.

Why is

R^{'} = R \cdot k

used for severe thermal shock instead of

R

alone?

High conductivity

k

smooths temperature gradients, lowering the actual

Δ T

across the part; for fast, severe shocks how quickly heat spreads matters, so

k

enters the merit figure.

Why does the crack-tip transformation add toughness rather than weaken zirconia?

Look at Figure 2 — grains near the tip swell 4%, but the surrounding solid resists that swelling, so it clamps those grains in compression. That compression pushes back on the very stress trying to open the crack, cancelling part of the applied stress-intensity

K

; the outside load must climb higher to reach the fracture threshold, which is extra toughness

Δ K_{t r an s f or ma t i o n}

Why can't we just make turbine blades entirely of ceramic to skip cooling?

Ceramics' tension weakness and shock sensitivity make a monolithic blade risk catastrophic fracture under vibration and transients; instead a cooled superalloy carries the load and a thin zirconia TBC blocks the heat.

Edge cases

What happens to SiC in a low-oxygen (fuel-rich) hot zone?

It switches to active oxidation, forming gaseous SiO instead of solid

S i O_{2}

; no protective skin forms, so the surface erodes and degrades.

Limiting case: a perfectly symmetric (harmonic) bond well — what is its thermal expansion?

Exactly zero for all

T

(set

g = 0

in Figure 1's valley), because

⟨ x ⟩

of a symmetric potential stays at equilibrium

x_{0}

; expansion needs the asymmetric

- g x^{3}

term.

Boundary case:

Δ T

raised until

σ_{t h} = σ_{f}

— what is physically happening?

The constrained thermal stress just reaches the fracture strength, so

Δ T = R

is the largest survivable sudden jump; any hotter shock cracks the part.

Degenerate case: a ceramic with zero applied tensile load but full compression — is a surface flaw dangerous?

No — compression tends to close cracks, so ceramics tolerate huge compressive loads; danger arises only when a stress field opens the flaw in tension.

What if you fully stabilise zirconia (enough yttria to lock cubic phase at all

T

You remove the tetragonal→monoclinic transformation entirely — no volume jump but also no transformation toughening, so you trade toughness for dimensional stability.

Edge case: two ceramics have identical

α

and

σ_{f}

but one has double the stiffness

E

— which resists shock better?

The lower-

E

one, since

R = \frac{σ _{f} ( 1 - ν )}{E α}

falls as

E

rises; a stiffer bond converts the same thermal strain into more stress.

Recall One-line summaries to self-test

Shock resistance depends on ::: $R = σ_{f} (1 - ν) / (E α)$ , not on melting point. Zirconia toughness mechanism ::: stress-triggered tetragonal→monoclinic transformation, 4% expansion closes the crack. Why zirconia coats blades ::: lowest $k$ (~2 W/m·K) blocks heat flux, and $α$ matched to the blade prevents spalling. SiC in air survives because ::: it grows a passive, self-healing $S i O_{2}$ glass skin. Ceramics are brittle because ::: sliding planes forces like-charge repulsion / breaks rigid covalent bonds, so cracks run instead. The symbols $x_{0}$ , $c$ , $g$ , $k_{B}$ mean ::: equilibrium bond length, bond stiffness, well lopsidedness, and Boltzmann's constant (energy per degree).