Cryogenic propellants — handling, insulation, boil-off: Question bank

This is a rapid-fire trap-hunting deck for Cryogenic Propellants — handling, insulation, boil-off. Each item below is a claim, an error, a "why", or a boundary case that the topic invites you to get wrong. Read the prompt, commit to an answer out loud, then reveal.

Nothing here needs a calculator — the arithmetic-heavy drills live elsewhere. This is about spotting the misconception.

The five symbols this deck uses

Before the traps, meet every symbol on this page in plain words, so nothing appears unearned. The two pictures below carry the whole story.

Definition The symbols, defined from zero

$\dot{Q}$ — the heat leak rate: how many joules of heat sneak into the cold tank each second (unit: watts, W). The dot on top is shorthand for "per second". Split into $\dot{Q}_{cond}$ (through solid supports), $\dot{Q}_{rad}$ (as radiation across the gap), and $\dot{Q}_{conv}$ (carried by moving air, if any).
$Δ T$ — the temperature difference between the warm outside and the cold cryogen (unit: kelvin, K). The $Δ$ just means "the gap in". Big $Δ T$ = strong push for heat to flow.
$k$ — thermal conductivity of a material: how easily heat walks through it. Copper: high $k$ . Titanium/plastic: low $k$ .
$A$ — the cross-sectional area heat flows through (unit: m²). More area = more parallel lanes for heat.
$L$ — the length of the heat path along a support strut (unit: m). Longer = heat has farther to travel.
$L_{v}$ — the latent heat of vaporization: the joules needed to boil one kilogram of liquid into gas without changing its temperature (unit: J/kg). This is the "toll" each kilogram pays to escape as vapour — see Latent Heat and Phase Changes.
$\overset{m}{˙}_{boil-off}$ — the boil-off mass rate: kilograms of liquid turning to gas each second (unit: kg/s). It equals the heat leak divided by the toll: $\overset{m}{˙}_{boil-off} = \dot{Q} / L_{v}$ .
$ε$ — emissivity: a number from 0 to 1 saying what fraction of "perfect" radiation a surface actually emits. Polished mirror ≈ 0.02, dull black ≈ 0.9.

Figure — Cryogenic propellants — handling, insulation, boil-off

Why conduction scales as $A / L$ . Picture heat as walkers crossing a corridor from the warm wall to the cold wall (left figure). Widen the corridor (bigger $A$ ) and more walkers cross at once — flow rises with $A$ . Lengthen the corridor (bigger $L$ ) and each walker takes longer, so the crowd thins out — flow falls with $L$ . That is exactly why Fourier's law reads $\dot{Q}_{cond} = k A Δ T / L$ : the $A$ on top, the $L$ underneath. The steeper the temperature drop ( $Δ T / L$ ), the harder heat is pushed.

Why radiation scales as $T^{4}$ . A hot surface throws out glowing "photon" packets (right figure). As it gets hotter two things grow at once: it throws more packets per second, and each packet carries more energy. Two growing effects multiply, and the honest bookkeeping (Stefan-Boltzmann Law) makes the total climb as the fourth power of temperature. The net flow between two surfaces is the difference $T_{hot}^{4} - T_{cold}^{4}$ — because the cold surface throws packets back. A radiation shield (a middle layer) intercepts packets partway, cutting the effective difference the coldest wall ever feels — the idea behind multilayer insulation (MLI).

True or false — justify

Cryogens boil at their storage temperature, so heating them makes them boil faster at a higher temperature.

False. Adding heat converts more liquid to gas at the same temperature (that is what

L_{v}

, latent heat, means); the temperature stays pinned at the boiling point while the phase change absorbs the energy.

Better insulation eventually stops boil-off completely.

False. No real insulation has zero heat leak, so some

\dot{Q}

always arrives and some liquid always boils; you can only slow it, never zero it, unless

Δ T = 0

A tank left sealed with no venting is safer than one that vents.

False. Sealed boil-off gas has nowhere to go, so pressure climbs until the tank ruptures — venting is a safety feature, not a leak to be fixed. See Structural Design - Pressure Vessels.

LH₂ has a higher latent heat than LOX, so an LH₂ tank always loses less mass per day.

False. Higher

L_{v}

(bigger toll per kg) resists boil-off per unit heat, but LH₂ is far colder, giving a bigger

Δ T

and thus more heat ingress

\dot{Q}

; in practice LH₂ boils off faster. Two competing effects — the toll

L_{v}

(Latent Heat and Phase Changes) versus the driving

Δ T

Evacuating the gap between tank and shell removes convection and helps with radiation.

Half true. Vacuum kills conduction/convection through gas, but radiation crosses a vacuum unhindered (the photon packets need no medium) — that is exactly why radiation still needs MLI shields. See Vacuum Technology.

Doubling the number of support struts doubles the conduction heat leak.

True (roughly). Struts are parallel heat corridors, so total conductive

\dot{Q}_{cond}

scales with the total cross-sectional area

A

— twice the struts, twice the

A

, twice the leak, via Fourier's Law of Heat Conduction.

Radiation heat leak depends mostly on the cold surface temperature.

False. Because

\dot{Q}_{rad} \propto T_{hot}^{4} - T_{cold}^{4}

and the hot side is far larger,

T_{hot}^{4}

dominates; the cold term is nearly negligible (e.g.

8 0^{4} ≪ 30 0^{4}

Aluminium struts are a good choice because aluminium is light.

False (for insulation). Aluminium has high thermal conductivity

k

, so it channels heat straight into the cryogen; low-

k

titanium or composites are chosen despite weight penalties.

Spot the error

"To reduce conduction we should make the struts shorter so heat has less material to fight through."

Wrong direction.

\dot{Q}_{cond} = k A Δ T / L

: a longer path (

L

large, denominator) spreads the temperature drop out and lowers the leak. Shorter struts leak more.

"MLI works by adding a thick blanket that traps air like a sweater."

Wrong mechanism. MLI is many thin reflective layers in vacuum; it fights radiation by making each layer re-radiate (each is a shield in the s02 picture), not by trapping air (air would add conduction/convection).

"Since exhaust is just water vapour, LH₂/LOX combustion releases no useful energy."

Confuses product with process. The water vapour is the result of an energetic reaction; the high exhaust velocity is precisely what gives the ~450 s specific impulse — see Propellant Mass Fraction.

"Emissivity

ε = 0.8

means the surface reflects 80% of radiation."

Backwards.

ε

is the emitted/absorbed fraction, so

ε = 0.8

means it emits 80% of a perfect blackbody; a low

ε \approx 0.02

(polished/aluminized) is what reflects and insulates.

"Boil-off percentage per day is a property of the propellant alone."

No — it depends on tank size, insulation, supports, and environment (all of

\dot{Q}

's inputs). The same LH₂ boils off at wildly different rates in a lab dewar versus a launch vehicle.

"Convection is always part of the heat budget, so we must include

\dot{Q}_{conv}

in space."

In vacuum (space or an evacuated jacket) there is no fluid to carry heat, so

\dot{Q}_{conv} = 0

; only conduction through solids and radiation remain.

"Newton's law of cooling gives an exact

h

we can derive from theory."

The convective coefficient

h

is empirical — it depends on flow speed, geometry and fluid, and is measured/correlated, not derived from first principles like

σ

Why questions

Why does the radiation term use

T^{4}

and not

T

T^{2}

Because both the number of emitted photon packets and the energy each carries grow with temperature (see the s02 figure); two growing effects multiply, and the bookkeeping gives the fourth power (Stefan-Boltzmann Law).

Why do we prefer long, thin struts over short, thick ones?

Long thin struts have large

L

and small

A

, and

\dot{Q}_{cond} \propto A / L

(the corridor picture in s01), so both choices minimise the conductive leak while still carrying the structural load.

Why does the vapour need to be vented rather than just cooled back?

Re-cooling requires a refrigerator adding heat/work and mass; on most vehicles it is simpler and lighter to vent the small continuous boil-off overboard (though "zero boil-off" active cooling exists on long missions).

Why is a temperature shield at an intermediate temperature (say 80 K) useful?

A cold shield intercepts most of the

T_{hot}^{4}

radiation before it reaches the cryogen (the middle layer in s02), dropping the effective

Δ (T^{4})

the coldest surface ever sees — the principle behind stacking many MLI layers.

Why must the outer surface avoid moisture/condensation?

A cold outer wall condenses and freezes atmospheric water, which then conducts heat inward and adds mass; a vapour barrier keeps the insulating vacuum/gap dry.

Why does LOX tolerate simpler insulation than LH₂?

LOX is much warmer (

- 18 3^{\circ}

C vs

- 25 3^{\circ}

C), so its

Δ T

to the environment is smaller, giving less heat ingress

\dot{Q}

for the same insulation.

Edge cases

Reason about these as limits — imagine sliding one input to its extreme and watching the formula respond, like reading the ends of a graph.

What is the boil-off rate the instant

Δ T \to 0

(tank at ambient temperature)?

Zero — with no temperature difference, both

\dot{Q}_{cond} \propto Δ T

and

\dot{Q}_{rad} \propto (T_{hot}^{4} - T_{cold}^{4})

vanish, so no heat arrives and

\overset{m}{˙}_{boil-off} = \dot{Q} / L_{v} = 0

. (But then it is not a cryogen either.)

If emissivity

ε \to 0

(perfect mirror), what happens to radiation leak?

\dot{Q}_{rad} \to 0

; a perfect reflector emits/absorbs nothing (it throws back every packet). Real surfaces bottom out near

ε \approx 0.02

, not exactly zero.

What happens to

\dot{Q}_{rad}

T_{cold} = T_{hot}

It vanishes:

T_{hot}^{4} - T_{cold}^{4} = 0

, so equal temperatures mean zero net radiative exchange even though both surfaces still radiate (equal packets each way).

A tank is completely full with no ullage (gas) space and is sealed. What is the danger as heat arrives?

The first bit of boiling vapour has no room to expand, so pressure spikes almost instantly — full sealed cryo tanks are extremely dangerous; ullage volume and venting are mandatory (Structural Design - Pressure Vessels).

In deep space with no atmosphere and only the Sun, which heat term dominates?

Radiation — no fluid means

\dot{Q}_{conv} = 0

, and with careful low-

k

struts conduction can be minimised, leaving solar and thermal radiation as the main ingress (Rocket Engine Cooling deals with the opposite extreme).

If the latent heat

L_{v}

were much larger, what happens to boil-off mass rate for fixed

\dot{Q}

It drops, since

\overset{m}{˙}_{boil-off} = \dot{Q} / L_{v}

; a larger

L_{v}

means each kilogram pays a bigger toll before evaporating, so fewer kilograms are lost per second.

Recall One-line self-test

The single fact that resolves half the traps ::: Cryogens boil at constant temperature; heat controls the rate of phase change, not the temperature — everything else follows from the heat-balance $\dot{Q} = \overset{m}{˙}_{boil-off} L_{v}$ .