CEA (Chemical Equilibrium with Applications) — using NASA-CEA tool to compute Isp, Tc, products: Visual walkthrough

We are going to answer one question: how fast does the exhaust come out, and why does a hot, light gas come out fastest?

Step 1 — What is a rocket, stripped to a picture?

WHAT. A rocket is a box (the chamber) full of very hot gas, with one open end (the nozzle). The gas rushes out the open end; the box gets pushed the other way. That push is thrust.

WHY. Before any algebra we must agree on the objects. Everything below is about one lump of gas travelling from "sitting in the hot chamber" to "screaming out the exit". If you can picture that lump, you can follow the derivation.

PICTURE. The magenta arrow is our gas lump. It starts slow and hot on the left (the chamber), speeds up and cools as it passes through the pinch (the throat), and exits fast on the right.

Figure — CEA (Chemical Equilibrium with Applications) — using NASA-CEA tool to compute Isp, Tc, products

Step 2 — Enthalpy: the "energy purse" a moving gas carries

WHAT. Each kilogram of gas carries a bundle of energy we call its specific enthalpy, written $h$ (units: joules per kilogram). Think of $h$ as the purse of thermal energy the gas can spend.

WHY. We need a bookkeeping quantity that stays constant as gas flows and can be traded for speed. Plain internal energy is not enough because the gas also pushes on the gas ahead of it; enthalpy is precisely "internal energy $+$ the push-work", so it is the right currency for flowing gas. That is why we use $h$ and not just temperature.

PICTURE. A hot, slow lump has a fat purse (large $h$ , tiny speed). As it accelerates, coins leave the purse and become motion — the purse shrinks, the speed grows.

Step 3 — The trade: enthalpy in the chamber becomes speed at the exit

WHAT. For steady flow with no heat added or lost through the walls (adiabatic), the sum "purse $+$ kinetic energy per kilogram" is the same everywhere: $h_{c} = h_{e} + \frac{1}{2} v_{e}^{2}$

WHY. This is just conservation of energy for our lump, written per kilogram. Nothing leaks in or out through the walls, so whatever leaves the purse ( $h_{c} - h_{e}$ ) must reappear as motion ( $\frac{1}{2} v_{e}^{2}$ ). We use this balance because it directly links the thing we can compute (a temperature drop) to the thing we want (exit speed).

PICTURE. A see-saw. On the left pan: purse energy $h$ . On the right pan: kinetic energy $\frac{1}{2} v_{e}^{2}$ . As the gas moves down the nozzle, weight slides from left pan to right pan; the total never changes.

Rearranging for the quantity we care about: $v_{e} = 2 (h_{c} - h_{e})$

$v_{e}$ — exit speed we are hunting.
$h_{c} - h_{e}$ — the enthalpy spent going from chamber to exit (the coins that turned into motion).
The square root comes from kinetic energy being $\frac{1}{2} v^{2}$ : to undo the "square" we take a root.

Step 4 — How much purse gets spent? (deriving the pressure-ratio bracket)

WHAT. We now compute $h_{c} - h_{e}$ from Step 2's $h \approx c_{p} T$ , but we do not yet know $T_{e}$ . We will get it from the rule that a smooth nozzle expands the gas at constant entropy. The result is $h_{c} - h_{e} = \frac{γ}{γ - 1} \frac{R _{u} T _{c}}{M} [1 - (\frac{p _{e}}{p _{c}})^{\frac{γ - 1}{γ}}]$

WHY. We need $T_{e}$ to know the temperature drop. A perfectly smooth, loss-free nozzle neither wastes energy to friction nor swaps heat with the walls — that "no-loss, no-heat" condition is what physicists call constant entropy, and it is the best-case the nozzle aims for.

Definition Meet the newcomers

$S$ — entropy: a number measuring how "spread out / disordered" the gas's energy is. A loss-free, no-heat-swap (adiabatic + reversible) process keeps $S$ constant — we call that isentropic. That constancy is the extra fact that pins down $T_{e}$ .
$c_{v}$ — specific heat at constant volume: joules to warm 1 kg by 1 K while the gas is not allowed to expand (no push-work done). Always $c_{v} < c_{p}$ , because at constant pressure some heat also does expansion work.
$γ$ (gamma) — the heat-capacity ratio $c_{p} / c_{v}$ . It measures how "springy" the gas is. For hot rocket exhaust $γ \approx 1.2$ .
$R_{u}$ — the universal gas constant, $8.314$ J/(mol·K). A fixed conversion between temperature and energy-per-mole.
$M$ — the average molar mass of the exhaust (kg/mol): how heavy one mole of the gas soup is.

The three algebra moves (each earns its place):

Rewrite $c_{p}$ using gas constants. For an ideal gas, $c_{p} - c_{v} = R_{u} / M$ (per kilogram) and $γ = c_{p} / c_{v}$ . Solving these two together gives $c_{p} = \frac{γ}{γ - 1} \frac{R _{u}}{M} .$ Why: this trades the awkward $c_{p}$ for the two numbers a designer actually reports — the springiness $γ$ and the exhaust weight $M$ .
So the spent purse is $c_{p} (T_{c} - T_{e})$ . Using $h \approx c_{p} T$ from Step 2, $h_{c} - h_{e} = c_{p} (T_{c} - T_{e}) = \frac{γ}{γ - 1} \frac{R _{u}}{M} T_{c} (1 - \frac{T _{e}}{T _{c}}) .$ Why: we have isolated the temperature ratio $T_{e} / T_{c}$ , the only unknown left.
Constant entropy links temperature ratio to pressure ratio. The isentropic rule for an ideal gas says $\frac{T _{e}}{T _{c}} = (\frac{p _{e}}{p _{c}})^{\frac{γ - 1}{γ}} .$ Why: this is the payoff of choosing constant entropy — it converts the temperature we can't measure into the pressure ratio we set with the nozzle. Substituting gives the boxed bracket.

Term-by-term inside the bracket:

$(\frac{p _{e}}{p _{c}})^{\frac{γ - 1}{γ}}$ — the temperature ratio $T_{e} / T_{c}$ , i.e. the fraction of the purse left over at the exit.
$1 - (\dots)$ — therefore the fraction actually spent into motion.
If $p_{e} = p_{c}$ (no expansion), the bracket is $0$ : nothing spent, no speed.

Step 5 — Assemble the exit velocity

WHAT. Put Step 4 into Step 3 ( $v_{e} = 2 (h_{c} - h_{e})$ ): $v_{e} = \frac{2 γ}{γ - 1} \frac{R _{u} T _{c}}{M} [1 - (\frac{p _{e}}{p _{c}})^{\frac{γ - 1}{γ}}]$

WHY. This is the whole point: one formula that turns chamber conditions into ideal exit speed. Reading it tells us what a rocket designer should chase.

PICTURE. The formula as a machine: three input pipes ( $T_{c}$ , $M$ , pressure ratio) feed a box, one output pipe ( $v_{e}$ ) comes out. Notice $T_{c}$ is on top and $M$ is on the bottom — pull them apart in your mind.

Step 6 — From speed to $I_{s p}$ : what actually pushes, and the fair scorecard

WHAT. The $v_{e}$ above is the momentum jet. But there is a second push: if the exit pressure $p_{e}$ differs from the outside (ambient) pressure $p_{a}$ , the pressure mismatch across the exit area $A_{e}$ also shoves the rocket. Total thrust is $F = \overset{m}{˙} v_{e} + (p_{e} - p_{a}) A_{e}$ where $\overset{m}{˙}$ is the mass flow (kg/s) and $A_{e}$ the exit area. We fold both terms into one effective exhaust velocity $c$ : $c = v_{e} + \frac{( p _{e} - p _{a} ) A _{e}}{m ˙}, F = \overset{m}{˙} c .$ Then specific impulse uses $c$ , not raw $v_{e}$ : $I_{s p} = \frac{c}{g _{0}}, g_{0} = 9.81 m/s^{2}$

WHY. Real thrust is momentum flux plus the pressure-area term. Dividing by $g_{0}$ (Earth's standard gravity) turns metres-per-second into seconds, a size-free number, so a tiny model engine and a giant booster with the same propellant get the same $I_{s p}$ . That is why the rocket equation uses $I_{s p}$ .

PICTURE. A "speedometer" $v_{e}$ , a small pressure-term add-on, combined into effective $c$ , then stamped through a divide-by- $g_{0}$ press into a clean "seconds" gauge $I_{s p}$ .

Combining with Step 5, at the matched-pressure design point: $I_{s p} \propto \frac{T _{c}}{M}$ This is the parent note's headline — now derived, not asserted.

Step 7 — The twist: fuel-rich beats stoichiometric (the fork)

WHAT. The naive move is to burn at the stoichiometric ratio (exactly enough oxidizer to consume all fuel), because that gives the hottest flame — the largest $T_{c}$ . But peak $I_{s p}$ sits on the fuel-rich side, where $T_{c}$ is slightly lower.

WHY. Because $I_{s p} \propto T_{c} / M$ , not $T_{c}$ . Running fuel-rich for $H_{2} / O_{2}$ leaves leftover $H_{2}$ (molar mass $\approx 2$ g/mol) in the exhaust, which drags the average $M$ down fast. The drop in $M$ outweighs the small drop in $T_{c}$ , so the ratio $T_{c} / M$ actually rises. Understanding why $T_{c}$ drops when you leave stoichiometric belongs to Adiabatic Flame Temperature and Dissociation at High Temperature.

PICTURE. Two curves against mixture ratio $O / F$ : $T_{c}$ (magenta) peaks at stoichiometric ( $O / F = 8$ for hydrolox); but $I_{s p}$ (violet) peaks earlier, on the fuel-rich side (~ $O / F = 6$ ). The gap between the two peaks is the whole lesson.

Step 8 — Edge and degenerate cases (never hit a wall you didn't show)

WHAT / WHY / PICTURE, one panel per case:

No expansion ( $p_{e} = p_{c}$ ): bracket $= 0 \Rightarrow v_{e} = 0$ . A rocket with no nozzle throat that never drops the pressure produces no jet — the purse is never spent.
Vacuum limit ( $p_{e} \to 0$ ): bracket $\to 1$ . Maximum possible spend; $v_{e}$ hits its ceiling $\frac{2 γ}{γ - 1} \frac{R _{u} T _{c}}{M}$ . This is why $I_{s p}$ is quoted highest "in vacuum".
Very heavy exhaust ( $M$ large, e.g. metal-oxide smoke): $T_{c} / M$ collapses, $I_{s p}$ tanks even if the flame is blazing hot. Heavy exhaust is the enemy.
Cold chamber ( $T_{c} \to 0$ ): $v_{e} \to 0$ . No thermal purse to spend — nothing to convert.
Frozen vs equilibrium: if we freeze the composition, no recombination heat is added downstream, so $h_{c} - h_{e}$ is smaller and $v_{e}$ is a touch lower. Real engines sit between frozen and equilibrium (finite reaction rates). This ties back to Chemical Equilibrium and Kp and Gibbs Free Energy and Equilibrium.

The one-picture summary

Everything compressed: hot chamber (fat purse) → nozzle trades purse for speed via the see-saw → the spend depends on $T_{c}$ up, $M$ down, and how far pressure falls → add the pressure-thrust term → divide by $g_{0}$ → $I_{s p} \propto T_{c} / M$ , peaking fuel-rich.

Recall Feynman retelling (plain words, no symbols)

Picture a lump of very hot gas sitting in a box. Heat is just hidden motion, and a hot gas carries a big "purse" of energy per kilogram. Open one end of the box and shape it like a funnel: as the gas squeezes out, it cools — and every bit of coolness it loses turns into forward speed, like sliding weight across a see-saw. How much cooling you get is set by how far the pressure falls: a smooth funnel loses no energy to rubbing or heat leaks, and that "no-loss" rule is exactly what fixes the exit temperature from the pressure drop. Two things decide the final speed: how hot it started (a fatter purse to spend) and how light the gas molecules are (light things are easy to fling fast for the same energy). There's a small extra shove if the jet's own pressure differs from the air outside — biggest in vacuum, where there's nothing pushing back. We then divide the whole effective speed by gravity so the score comes out in seconds and doesn't care whether the engine is tiny or huge. The last surprise: don't burn perfectly. Burning a bit fuel-heavy makes the flame slightly cooler but leaves lots of feather-light hydrogen in the exhaust, and that lightness wins — so the best rocket runs a little richer than "perfect".

Recall Self-check

Why does $v_{e}$ depend on the square root of energy? ::: Kinetic energy grows as speed squared ( $\frac{1}{2} v^{2}$ ), so recovering speed from energy needs a square root. Which fact turns the temperature ratio $T_{e} / T_{c}$ into a pressure ratio? ::: The constant-entropy (isentropic) rule $T_{e} / T_{c} = (p_{e} / p_{c})^{(γ - 1) / γ}$ . Why must $p_{e} \leq p_{c}$ ? ::: Otherwise the bracket goes negative and $v_{e}$ would be a square root of a negative number — a nozzle only expands, never compresses. What is the effective exhaust velocity $c$ and how does it differ from $v_{e}$ ? ::: $c = v_{e} + (p_{e} - p_{a}) A_{e} / \overset{m}{˙}$ includes the pressure-thrust term; $c = v_{e}$ only when $p_{e} = p_{a}$ . Why is peak $I_{s p}$ fuel-rich, not stoichiometric? ::: Fuel-rich lowers $M$ (leftover light $H_{2}$ ) faster than it lowers $T_{c}$ , so $T_{c} / M$ rises.