Expander cycle — hydrogen-cooled nozzle drives turbine: Question bank

Before you start, remember the one-sentence picture: the cold liquid hydrogen is used as a coolant, it absorbs nozzle heat and boils into high-pressure gas, and that gas spins the turbopumps before being burned. Nothing is burned to feed the turbine.

Symbols and sign conventions (build these first)

Before any trap makes sense, we must earn every symbol. Read this once; then the reveals below use them freely.

Definition Every symbol on this page, in plain words

$\overset{m}{˙}$ — mass flow rate, "how many kilograms of fluid pass a point each second" (units kg/s). The dot means "per second". Picture a stopwatch counting kilograms streaming past.
$c_{p}$ — specific heat capacity at constant pressure: the energy (in joules) needed to warm one kilogram by one kelvin. Big $c_{p}$ = the fluid is a greedy heat sponge. Hydrogen's is ~ $14000 J/kg\cdotK$ .
$ρ$ (Greek "rho") — density, kilograms packed into one cubic metre (kg/m³). Dense = heavy for its size (oxygen), light = fluffy for its size (hydrogen).
$Δ p$ (the triangle $Δ$ means "change in") — pressure rise a pump adds, i.e. exit pressure minus inlet pressure, in pascals (Pa).
$γ$ (Greek "gamma") — specific-heat ratio $c_{p} / c_{v}$ , a pure number (~1.4 for diatomic gases like H₂). It measures how "springy" the gas is when compressed or expanded; it appears whenever we expand a gas without adding heat.
$η_{p}$ , $η_{t}$ (Greek "eta") — pump efficiency and turbine efficiency, pure numbers between 0 and 1 saying what fraction of the ideal work is really achieved. Real machines waste some, so these are $< 1$ .
$p_{1}, p_{2}$ — pressures at turbine inlet and outlet; $T_{1}, T_{2}$ the matching temperatures.

Figure — Expander cycle — hydrogen-cooled nozzle drives turbine

The figure above traces the hydrogen loop with energy arrows: heat in at the jacket (green, $+ Q$ ), work out at the turbine to the shaft, and work in at the pumps from the same shaft. Notice the single shaft couples turbine and pumps — one spinning rod, so they all turn at the same rotational speed; the turbine must produce exactly the torque the two pumps demand at that speed.

Where the formulas come from (so the traps are obvious)

The square-cube limit is worth seeing rather than asserting — the plot below shows heat supply ( $\propto L^{2}$ ) and pump demand ( $\propto L^{3}$ ) crossing at a definite size, beyond which no closed expander can close.

True or false — justify

The expander cycle burns a small amount of fuel in a preburner to drive the turbine.

False. A closed expander burns nothing for the turbine; the turbine gas is hydrogen heated by nozzle-wall heat only. It is burned afterward in the main chamber. Preburners belong to Staged combustion cycle and Gas generator cycle.

The heat that drives the turbine is "free" in the sense that it costs no extra propellant.

True. The nozzle must be cooled anyway (see Regenerative cooling); routing fuel through the walls recycles waste heat that would otherwise melt the wall. No propellant is spent to create that heat.

The hydrogen leaves the turbine at higher pressure than it entered.

False. A turbine extracts work by letting the gas expand, so pressure drops from

p_{1}

p_{2} < p_{1}

. That lower-pressure gas must still be above chamber pressure to be injected.

In a closed expander cycle, all the heated hydrogen eventually reaches the combustion chamber.

True. "Closed" means nothing is dumped: 100% of the hydrogen passes turbine → chamber. In an open (bleed) expander a fraction is vented overboard, costing a little Specific impulse.

Because the turbine adds energy to the hydrogen, the hydrogen is hotter leaving the turbine than entering it.

False. The turbine removes energy (

W_{t u r b} > 0

leaves the gas), so expansion cools it:

T_{2} = T_{1} (p_{2} / p_{1})^{(γ - 1) / γ} < T_{1}

. The heating happened earlier, in the cooling jacket.

An expander cycle can be scaled to arbitrarily high thrust just by making the nozzle bigger.

False. Heat supply grows with surface area (

\propto L^{2}

) but pump demand grows with propellant flow (

\propto L^{3}

). By the Square-cube law the demand eventually outruns the supply, capping closed expanders near a few hundred kN.

The oxidizer pump is generally cheaper (less power) per kilogram than the fuel pump.

True. Pump power scales with volume flow

\overset{m}{˙} / ρ

. Liquid oxygen has far higher density

ρ

than hydrogen, so the same

Δ p

costs less power per kg despite the larger mass flow.

Kerosene would work just as well as the coolant/working fluid if you accept slightly lower efficiency.

False. Kerosene cokes (carbonizes and clogs channels) when heated and has low

c_{p}

, so it cannot absorb enough clean heat to run the turbine at all. Hydrogen's huge

c_{p}

(~14 kJ/kg·K) is what makes the cycle possible — see Liquid hydrogen properties.

Spot the error

Q = \overset{m}{˙} c_{p} T_{o u t}

is the heat absorbed by the jacket."

Error: it must be the temperature rise,

Q = \overset{m}{˙} c_{p} (T_{o u t} - T_{in})

. Heat depends on how much the temperature changes, not the absolute exit temperature.

"The turbine work formula uses

p_{1} / p_{2}

raised to

(γ - 1) / γ

Error: the ratio is

(p_{2} / p_{1})

, i.e. exit over inlet, which is less than 1. Using

p_{1} / p_{2}

would give a term

> 1

and a negative work — physically wrong for expansion. See Isentropic flow relations.

"Pump power is

W = \overset{m}{˙} Δ p ρ / η_{p}

Error: density is in the denominator:

W = \overset{m}{˙} Δ p / (ρ η_{p})

. Denser fluid means smaller volume flow

\overset{m}{˙} / ρ

, hence less power — so

ρ

must divide, not multiply.

"An efficiency

η_{t} = 0.75

should multiply the whole bracket and add a bit more since real turbines beat ideal."

Error: real turbines are worse than ideal, so

η_{t} < 1

scales the ideal work down. You never get more than the isentropic maximum.

"The self-sustaining condition is

W_{t u r b} \leq W_{p u m p}

Error: the inequality points the other way. The turbine must supply at least the pump demand:

W_{t u r b} \geq W_{p u m p, f u e l} + W_{p u m p, o x}

, otherwise the shaft slows and the engine stalls.

"Since hydrogen absorbs heat in the jacket, it must be flowing outward from the chamber toward the tank."

Error: the flow is tank → pump → cooling jacket → turbine → chamber. Hydrogen moves toward the chamber the whole time; it just heats up along the way.

"The turbine and each pump can spin at whatever speed they like."

Error: they share one shaft, so turbine and both pumps turn at the same rotational speed. The design must match the turbine's torque-vs-speed curve to the pumps' demand at that common speed — see Turbopump design.

Why questions

Why is hydrogen chosen over any other fuel as the working fluid?

Its very large specific heat

c_{p}

lets a small mass pick up a lot of enthalpy per kelvin, and it boils cleanly without coking, so nozzle heat converts into abundant, usable high-pressure gas.

Why does the expander cycle waste no propellant, unlike the gas generator cycle?

The turbine gas is hydrogen heated by the wall, not gas made by combustion. It flows on into the main chamber and is burned there — nothing is dumped overboard, so Specific impulse stays high.

Why does density

ρ

appear in the denominator of pump power?

Because the work to raise pressure acts on volume, and the volume flow is

\overset{m}{˙} / ρ

— the same mass of a denser fluid occupies less volume, so less work per kilogram is needed to push it across

Δ p

Why must the turbine use only a modest pressure drop rather than expanding the gas fully?

The hydrogen leaving the turbine still has to be injected into the combustion chamber, which sits at high pressure. If the turbine expanded the gas too far, its exit pressure would fall below chamber pressure and it couldn't enter.

Why does the exponent

(γ - 1) / γ

(not 1) govern how temperature falls in the turbine?

Because expansion with no heat added obeys

p V^{γ} =

const, not simple proportionality; combining it with the ideal-gas law makes temperature drop less steeply than pressure, which is exactly what that fractional exponent encodes.

Why does the square-cube law limit expander cycles specifically, and not gas generators?

Expanders depend on wall area (

L^{2}

) to gather heat, which lags behind the

L^{3}

propellant demand. A gas generator makes its turbine power by burning propellant, so its supply scales with volume too and doesn't hit the geometric ceiling.

Why is the same

Q

described as both "cooling the nozzle" and "powering the turbine"?

They are one and the same energy (

Q > 0

leaves the wall and enters the fluid): it simultaneously stops the wall melting and raises the gas enthalpy the turbine later extracts. One heat flow does two jobs.

Why does an open (bleed) expander allow higher turbine power than a closed one?

Dumping the turbine hydrogen overboard removes the constraint that it must survive at chamber-injectable pressure, so a larger flow and bigger pressure drop can be used — at the cost of the vented propellant's contribution to thrust. (See the bleed-port branch in the schematic.)

Edge cases

If the hydrogen entered and left the cooling jacket at the same temperature (

T_{o u t} = T_{in}

), what turbine power is available?

Q = \overset{m}{˙} c_{p} \cdot 0 = 0

, so no enthalpy is added and the turbine can produce no work — the cycle cannot start. Some temperature rise is essential.

What happens at engine startup, before the nozzle is hot?

There is little wall heat, so the turbine can't yet make full power — expander cycles need a startup transient (e.g. tank pressure, tapped stored energy, or the margin built into the power balance) to bootstrap the flow until the walls heat up.

If pump efficiency

η_{p}

drops toward zero, what does the power balance say?

Pump demand

\overset{m}{˙} Δ p / (ρ η_{p})

blows up toward infinity, so the fixed turbine power can no longer cover it and the self-sustaining inequality fails — the engine cannot close.

For a fixed turbine, if you demand a much higher chamber pressure (larger

Δ p

across the pumps), what limits you?

Pump power rises linearly with

Δ p

while turbine power is capped by available wall heat. Beyond some

Δ p

the demand exceeds supply, which is one reason expanders favour moderate chamber pressures.

In the limit of a vanishingly small pressure ratio

p_{2} / p_{1} \to 1

, what is the turbine work?

The bracket

1 - (p_{2} / p_{1})^{(γ - 1) / γ} \to 0

, so

W_{t u r b} \to 0

. No pressure drop means no expansion and no extractable work — the turbine must be given a genuine pressure ratio to do anything.

Recall Quick self-test

The turbine gas in a closed expander is heated by ::: heat conducted through the nozzle/chamber walls (regenerative cooling), not by combustion. Pump power scales with volume flow, so denser propellant is ::: cheaper to pump per kilogram (oxygen costs less power than hydrogen for the same $Δ p$ ). Expanders are size-limited because heat supply grows as area while demand grows as ::: volume/flow — the square-cube law. The turbine and both pumps share one shaft, so they all spin at ::: the same rotational speed, and the turbine torque must match the combined pump demand.

See also: Turbopump design · Regenerative cooling · Isentropic flow relations · Square-cube law · Specific impulse · Liquid hydrogen properties