Nuclear thermal propulsion — NTR Isp ~900 s concept: Visual walkthrough

Everything below assumes you know only two everyday facts: hot things store energy, and you can push gas through a hole. We earn every symbol from there.

Definition Our two modelling assumptions

(1) Ideal gas. Throughout we treat the propellant as an ideal gas: the molecules are tiny, don't stick to each other, and obey $p V = n R T$ for $n$ moles — pressure $p$ times volume $V$ equals the mole-count $n$ times the gas constant $R$ times absolute temperature $T$ . This law is what later lets us write enthalpy as $h = c_{p} T$ and expose the $1/ M$ . (2) Isentropic flow. We assume the flow down the nozzle is isentropic = adiabatic (no heat leaks in or out through the walls) and reversible (no friction, shocks, or turbulence wasting energy). This is the crucial assumption that makes energy bookkeeping clean: with no heat added and nothing wasted, all the thermal energy the gas gives up must reappear as directed motion. Real nozzles are close enough to this ideal that it is a fair starting point.

Step 1 — Picture the rocket as a machine that turns "hot & still" into "cold & fast"

WHAT. Imagine a closed chamber full of gas, with one shaped opening at the back. Inside the chamber the gas is hot and barely moving (it just jostles in place). By the time it leaves the opening it is cold and racing straight backwards.

WHY. A rocket produces thrust by throwing mass backwards. So the only number we ultimately care about is how fast the gas leaves — the exhaust speed, which we name $v_{e}$ (the little $e$ just means "exhaust"). If we can find $v_{e}$ , everything else follows.

PICTURE. Look at the figure: the chamber on the left is a fog of slow dots (hot, chaotic). The nozzle on the right squeezes then flares, and the dots come out as neat parallel amber arrows — slow-random motion converted into fast-directed motion.

Figure — Nuclear thermal propulsion — NTR Isp ~900 s concept

Step 2 — Where does the speed come from? Trade heat for motion

WHAT. The gas has no giant spring pushing it. What it has is thermal energy — the energy of its random jiggling when hot. As the gas flows down the nozzle it cools, and the jiggle-energy it loses reappears as forward speed.

WHY. This is just energy conservation applied to a flowing stream, using assumption (2) above: because the flow is isentropic (no heat leaks, nothing wasted), every joule of thermal energy the gas releases becomes orderly kinetic energy — none escapes through the walls or dissipates in turbulence. We track energy per kilogram of gas so the answer doesn't depend on how big our rocket is.

Before writing the balance we must first name the "thermal energy per kg" quantity. A moving parcel of gas is not free — the gas behind it is pushing it forward, and it in turn pushes the gas ahead. That pushing is real work, called flow-work. The honest bookkeeping of "energy carried by a flowing gas" must include both the internal jiggle energy and this push-work. The sum has its own name, enthalpy, written $h$ — a quantity measured per kilogram. (Step 3 unpacks why enthalpy is the right currency and gives its formula; here we only need the name.)

PICTURE. The figure shows a single parcel of gas sliding down the nozzle: on the left a fat amber "thermal energy" bar (its enthalpy) and a tiny motion arrow; on the right the thermal bar has shrunk and the motion arrow has grown. The total (bar + arrow) is the same height at both ends — that equality is conservation.

Step 3 — Why the energy currency is enthalpy, not plain heat

WHAT. We now unpack the enthalpy $h$ named in Step 2 and write its formula. First we need the two ways a gas can be warmed, because $h$ is built from them.

WHY THIS TOOL. If you used only the internal jiggle energy $c_{v} T$ , you'd forget the push-work a flowing gas does and get an answer too small by a factor $γ$ . Enthalpy is defined precisely to include both: $h = internal jiggle c_{v} T + flow-work per kg \frac{p}{ρ} = c_{p} T$ (here $p / ρ$ is the pressure divided by density = the push-work carried per kg; for an ideal gas the two pieces collapse neatly into $c_{p} T$ ). Enthalpy $h = c_{p} T$ is the tool built exactly to answer "how much energy does a kilogram of flowing gas carry?" — that is why it, and not internal energy, belongs in Step 2.

PICTURE. The figure shows a plug of gas in a pipe with a piston of gas behind it doing work $p V$ to shove it along. The label reads: enthalpy $=$ (jiggle energy $c_{v} T$ ) $+$ (this shove energy).

Substituting $h = c_{p} T$ into the Step-2 balance: $\frac{1}{2} v_{e}^{2} = c_{p} (T_{c} - T_{e})$ Every symbol here is now earned. Next we crack open $c_{p}$ itself.

Step 4 — Crack open $c_{p}$ : why lighter molecules win

WHAT. $c_{p}$ is measured per kilogram, so it secretly depends on how heavy each molecule is. One kilogram of a light gas contains far more molecules than one kilogram of a heavy gas. More molecules per kg = more little energy-buckets per kg = more energy stored per kg for the same temperature.

WHY the formula has this exact shape. Two facts, both flowing from the ideal-gas law of the intro box, pin down $c_{p}$ :

$c_{p} - c_{v} = R_{s}$ . When you heat a gas at constant pressure it must expand and do push-work $p Δ V$ ; from $p V = n R T$ that extra work per kelvin per kg is exactly $R / M$ . So warming at constant pressure always costs $R / M$ more per kg than warming at constant volume ( $c_{v}$ ). We give this recurring group its own name, the specific gas constant $R_{s} \equiv R / M$ (units J/(kg·K)) — it is just the universal constant $R$ shared out over the mass of one mole. That is where the $R / M$ is born.

$c_{p} / c_{v} = γ$ . By definition the ratio of the two specific heats is named $γ$ .

Combine them: from (2), $c_{v} = c_{p} / γ$ ; substitute into (1): $c_{p} - c_{p} / γ = R_{s}$ , i.e. $c_{p} (1 - 1/ γ) = R_{s}$ , so $c_{p} = \frac{γ}{γ - 1} R_{s} = \frac{γ}{γ - 1} \frac{R}{M}$ . The $γ / (γ - 1)$ is not magic — it is just algebra tidying those two facts together.

PICTURE. Two boxes of equal mass (1 kg each): the left box holds a swarm of tiny light molecules; the right holds a few fat heavy ones. Same total kg, same temperature — but the left box has many more energy-carriers, shown as more little glowing dots.

Formula

$c_{p}$ written through molar mass $c_{p} = \frac{γ}{γ - 1} \frac{R}{M}$

$R = 8.314 J/(mol\cdotK)$ — the universal gas constant, the same for every gas; it fixes how much energy a mole holds per kelvin.
$M$ — molar mass in kg/mol (this is the SI unit we must use in every plug-in; the familiar "g/mol" of a periodic table is $1000 \times$ larger, so always convert: hydrogen $2 g/mol = 2 \times 1 0^{- 3} kg/mol$ ). Dividing by $M$ turns "per mole" into "per kg." This is the $1/ M$ that makes light gases powerful.
$R_{s} = R / M$ — the specific gas constant just named; it bundles $R$ and $M$ into one per-kg number.
$γ = c_{p} / c_{v}$ — the adiabatic index, a pure number (no units) describing how many ways a molecule can store energy (spin, vibrate...). For simple gases $γ \approx 1.3$ – $1.4$ .
$\frac{γ}{γ - 1}$ — a dimensionless efficiency dial that converts "energy stored per mole" into "energy available to become directed flow."

Plug in: $\frac{1}{2} v_{e}^{2} = \frac{γ}{γ - 1} \frac{R}{M} (T_{c} - T_{e})$

Step 5 — Let the nozzle finish the job: full expansion

WHAT. A long, well-shaped nozzle keeps expanding the gas, cooling it more and more, so the exit temperature $T_{e}$ drops far below the chamber temperature. In the ideal limit we set $T_{e} \to 0$ , meaning almost all the thermal energy got converted to speed.

WHY, and how a real nozzle approaches it. The exit temperature is not free — it is locked to the exit pressure by the isentropic relation $T_{e} / T_{c} = (p_{e} / p_{c})^{(γ - 1) / γ}$ . A converging–diverging nozzle first accelerates the gas to the speed of sound at its narrowest point (the throat "chokes"), then the flaring diverging cone keeps expanding and cooling it as the pressure falls toward the outside pressure $p_{e}$ . To reach $T_{e} \to 0$ you would need $p_{e} \to 0$ — i.e. expand all the way down to vacuum, which is exactly the case in space and requires an infinitely large exit area. So $T_{e} \to 0$ is the best-case ceiling a very large space nozzle approaches but never quite reaches; the bigger the diverging bell, the lower $p_{e}$ , the lower $T_{e}$ , the closer to ideal. We take the clean limit to expose the physics fingerprint (we revisit finite $T_{e}$ as an edge case in Step 7).

PICTURE. A temperature-versus-position plot down the nozzle: $T$ starts high at $T_{c}$ and slides down a curve toward the axis; the shaded area under the drop is the energy that became motion. The ideal (vacuum-expansion) case shades all the way to zero.

Setting $T_{e} = 0$ and solving for $v_{e}$ (multiply both sides by 2, take the square root):

The fingerprint: stripping constants, $v_{e} \propto \frac{T _{c}}{M}$ . Hot helps (top), heavy hurts (bottom).

Step 6 — Read the fingerprint: why NTR beats chemical

WHAT. Compare two rockets using only $T_{c} / M$ . (Molar masses below are quoted in the everyday g/mol for readability; when you actually compute $v_{e}$ you convert to kg/mol as in Step 8 — but for a ratio the factor of 1000 cancels, so g/mol is fine here.)

	$T_{c}$ (K)	$M$ (g/mol)	$T_{c} / M$
Chemical (H₂+O₂ → water)	3600	18	200
NTR (pure H₂)	2700	2	1350

WHY. Notice the NTR is cooler ( $2700 < 3600$ ) yet its ratio is $6.75 \times$ larger — because its molecule is $9 \times$ lighter. The light molecule wins the tug-of-war.

PICTURE. Two bars for $T_{c} / M$ : the chemical bar short despite a tall temperature flag; the NTR bar tall despite a shorter temperature flag, with a big "÷ light M" arrow lifting it.

$\frac{v _{e}^{NTR}}{v _{e}^{chem}} = \frac{1350}{200} = 6.75 \approx 2.6$

Roughly double the exhaust speed → roughly double the Specific Impulse. This is the ~900 s vs ~450 s story.

Step 7 — Edge and degenerate cases (never let the reader fall off the map)

WHAT & WHY — five boundaries of the formula:

The figure overlays the cases on the $v_{e}$ curve: a real-nozzle dip, an $M \to 0$ asymptote, an $M \to \infty$ collapse, a melting-point wall, and the $γ$ push.

Step 8 — Put a number on it (the ~900 s check)

WHAT. March the formula to a value using NTR hydrogen numbers: $T_{c} = 2700$ K, $M = 2 \times 1 0^{- 3}$ kg/mol (converted from 2 g/mol), $γ = 1.4$ .

WHY. A derivation you can't turn into a number is a story, not physics. This closes the loop with the parent note.

Prefactor: $\frac{2 γ}{γ - 1} = \frac{2.8}{0.4} = 7.0$ .
Thermal energy per kg: $R_{s} T_{c} = \frac{R T _{c}}{M} = \frac{8.314 \times 2700}{2 \times 1 0 ^{- 3}} = 1.122 \times 1 0^{7}$ J/kg.
$v_{e} = 7.0 \times 1.122 \times 1 0^{7} = 7.86 \times 1 0^{7} \approx 8860$ m/s.
$I_{s p} = v_{e} / g_{0} \approx 8860/9.81 \approx 903$ s. ✔

The one-picture summary

One figure, left to right, compresses all eight steps: hot still gas → enthalpy currency → cracked open by $1/ M$ → nozzle converts it → square root gives speed → the $T / M$ fingerprint → ~900 s.

Recall Feynman retelling of the whole walkthrough

A rocket only cares about one thing: how fast it can throw gas out the back. Inside the chamber the gas is hot and jiggling but going nowhere. As it slides through the nozzle it cools down and speeds up — the jiggle-energy turns into a straight fast rush, and because the flow is smooth and loses no heat to the walls (isentropic), all of that released energy becomes motion. To count that energy honestly we can't use plain heat, because moving gas also has to shove the gas in front of it; the right bookkeeping is called enthalpy, which is just internal jiggle energy ( $c_{v} T$ ) plus that shove — together $c_{p} T$ . When we write enthalpy per kilogram, a $1/ M$ pops out: a kilogram of light gas is a huge crowd of tiny molecules, so it stores more speed-making energy than a kilogram of a few fat molecules. Solve for the speed and — because energy goes like speed squared — a square root appears, leaving the clean rule $v_{e} \propto T / M$ . Hot helps, heavy hurts. A nuclear rocket is actually cooler than a chemical flame, but it burns the lightest gas there is, hydrogen, nine times lighter than the steam a chemical rocket makes. Nine times lighter beats the lower temperature easily, so the hydrogen flies out about 2.6× faster — and $2700$ K of hydrogen lands you right at that famous ~900 seconds.

Active-recall

Why must the nozzle flow be assumed isentropic?

Adiabatic (no heat leaks through walls) + reversible (no friction/shocks) guarantees every joule of thermal energy released becomes directed kinetic energy, making

\frac{1}{2} v_{e}^{2} = h_{c} - h_{e}

exact.

Why is enthalpy

c_{p} T

, not

c_{v} T

, the right energy currency?

A flowing gas also does push/flow-work; enthalpy = internal energy (

c_{v} T

) + flow-work, so it captures all the energy a moving kilogram carries.

What is the specific gas constant

R_{s}

R_{s} = R / M

, the universal constant shared over one mole's mass; units J/(kg·K). It is the

R / M

group that appears throughout.

Where does the

1/ M

come from in

v_{e}

c_{p}

is per-kg, and one kg of light gas has more molecules (energy-carriers), so its per-kg heat scales as

R / M

How does a real nozzle approach

T_{e} \to 0

Via

T_{e} / T_{c} = (p_{e} / p_{c})^{(γ - 1) / γ}

: a bigger diverging bell drops the exit pressure toward vacuum, cooling

T_{e}

toward zero — reached only by an infinite space nozzle.

Why does temperature sit under a square root?

Kinetic energy

\propto v^{2}

; solving energy conservation for

v

requires a square root.

What is

g_{0}

in the

I_{s p}

definition?

Standard gravity

9.81 m/s^{2}

— a fixed conversion constant (not local gravity) that makes

I_{s p}

come out in seconds.

What happens to

v_{e}

M \to \infty

It goes to zero — a heavy molecule barely moves however hot the gas; this is why heavy exhaust is bad.

What actually caps solid-core NTR

I_{s p}

near 900 s?

Reactor fuel elements melt near 3000 K — a materials limit, not a physics limit.

The scaling fingerprint of exhaust velocity?

v_{e} \propto T_{c} / M

Connections

Parent topic — NTR ~900 s
Specific Impulse — what $v_{e} / g_{0}$ becomes.
De Laval Nozzle — the machine of Steps 1 & 5.
Adiabatic Flow & Enthalpy — the energy currency of Step 3.
Nuclear Fission — the heat source behind $T_{c}$ .
Chemical Rocket Propulsion — the ~450 s baseline of Step 6.
Tsiolkovsky Rocket Equation — where high $v_{e}$ pays off.
Nuclear Electric Propulsion — the alternative that trades thrust for even higher $I_{s p}$ .