Specific impulse Isp = v_e - g₀ — definition, physical meaning, units: Visual walkthrough

Before we touch a single equation, let us agree on the cast of characters — the physical things we will draw over and over.

Step 1 — A rocket is a machine that throws mass

WHAT. Picture a rocket floating in empty space. It opens a nozzle and throws a small blob of gas out the back at speed $v_{e}$ . Nothing else happens — no air, no ground, no outside push.

WHY. We start here because this single act is the entire source of a rocket's motion. If we understand one thrown blob, we understand the whole engine — the rest is just adding up blobs. This is the Impulse-Momentum Theorem in its rawest form: give momentum to the gas, get equal-and-opposite momentum back.

PICTURE. The blue arrow (gas, going left at $v_{e}$ ) and the amber arrow (rocket, nudged right) are the whole story. They are equal in "momentum-strength," opposite in direction — Newton's third law drawn.

Figure — Specific impulse Isp = v_e - g₀ — definition, physical meaning, units

Step 2 — Turn "throwing mass" into a force (thrust)

WHAT. In one second the engine throws out $\overset{m}{˙}$ kilograms of gas, each kilogram leaving at speed $v_{e}$ . The momentum leaving per second is $\overset{m}{˙} \times v_{e}$ . A force is exactly "momentum delivered per second," so:

$F = kg thrown each second \overset{m}{˙} \times speed of each kg v_{e}$

WHY THIS TOOL — why "momentum per second"? We could ask "what is the force?" many ways, but force is defined as the rate at which momentum changes. Since the only thing changing here is the gas's momentum ( $\overset{m}{˙} v_{e}$ every second), the rocket must feel that same amount as a forward push $F$ . No new physics — just the definition of force applied to the gas stream.

PICTURE. The conveyor belt of little gas packets: each second a packet of mass $\overset{m}{˙}$ shoots left carrying momentum $\overset{m}{˙} v_{e}$ ; the recoil stacks up into a steady amber thrust arrow $F$ on the rocket.

Step 3 — Add up the whole burn to get total impulse

WHAT. The engine does not fire for one second — it fires for the whole burn, lasting $t_{b}$ seconds. Impulse $J$ is the total push added up over time: force multiplied by how long it acts. When force can change moment to moment, "add up over time" is written as an integral $\int F d t$ (the $\int$ sign literally means "sum the tiny slices $F d t$ ").

$J = \int_{0}^{t_{b}} F d t = \int_{0}^{t_{b}} \overset{m}{˙} v_{e} d t$

WHY THIS TOOL — why an integral? Because we are summing infinitely many tiny time-slices $d t$ , each contributing a sliver of push $F d t$ . That "sum of infinitely many thin slices" is precisely what an integral computes. If $F$ were constant we could just multiply $F \times t_{b}$ ; the integral is the honest version that works even while the engine's flow varies.

PICTURE. The area under the thrust-versus-time curve. Each thin amber strip is one $F d t$ slice; the total shaded area is $J$ .

Step 4 — Pull $v_{e}$ out; what's left is just the fuel used

WHAT. Assume $v_{e}$ stays constant during the burn (a well-designed engine). A constant can slide outside the sum:

$J = v_{e} = m_{p} \int_{0}^{t_{b}} \overset{m}{˙} d t = v_{e} m_{p}$

The leftover integral $\int \overset{m}{˙} d t$ = "kilograms-per-second added up over all the seconds" = the total propellant mass $m_{p}$ that got burned.

WHY. Adding up "how much left each second" across the whole burn can only give the total amount that left — the whole tank of fuel. That is $m_{p}$ , in kilograms. Pulling $v_{e}$ out is legal only because we assumed it constant; that assumption is what makes the algebra clean.

PICTURE. The tank drains from full to empty; the striped region is the total propellant $m_{p}$ that flowed out — the same shaded area as Step 3, now re-labelled as mass.

Step 5 — Divide by the weight of that fuel (the key trick)

WHAT. Specific impulse is defined as impulse per unit weight of propellant. Weight is mass times gravity, and here we deliberately use the fixed constant $g_{0}$ :

$W_{p} = m_{p} g_{0} ⟹ I_{s p} = \frac{J}{W _{p}} = \frac{v _{e} m _{p}}{m _{p} g _{0}}$

The $m_{p}$ on top and bottom cancel:

$I_{s p} = \frac{v _{e}}{g _{0}}$

WHY divide by weight, and why $g_{0}$ not local $g$ ? Dividing by the amount of fuel removes engine size — a tiny thruster and a giant booster are now judged on the same scale (pure efficiency). We use the fixed $g_{0}$ (not the local gravity where the rocket flies) so the number is the same on Earth, Mars, or deep space — it is a property of the engine, not the place.

PICTURE. Top row (impulse) and bottom row (weight) both carry the same $m_{p}$ block; the amber "cancel" stroke deletes it, leaving the clean ratio $v_{e} / g_{0}$ .

Step 6 — Why the answer comes out in seconds

WHAT. Check the units. Impulse has units of newton·seconds (N·s); weight has units of newtons (N). Dividing:

$[I_{s p}] = \frac{N\cdots}{N} = s$

The newtons cancel. Equivalently, from $v_{e} / g_{0}$ : $\frac{m/s}{m/s ^{2}} = s$ .

WHY care? A "second" here is not how long the burn lasts. It means: the engine can produce a thrust equal to the weight of one unit of its propellant, for that many seconds, per unit of propellant weight it spends. It is a normalized score — same idea as "kilometres per litre" being a fixed rating of a car, not a trip length.

PICTURE. The two cancellation ladders side by side: N/N and (m/s)/(m/s²) both collapsing to the single unit s.

Step 7 — The degenerate and edge cases

Never leave a scenario undrawn. Here is what happens at the extremes.

Case A — $v_{e} \to 0$ (gas barely leaves). Then $I_{s p} = 0/ g_{0} = 0$ . An engine that "throws" gas at zero speed gives zero push per kilogram — a useless engine. The formula agrees.

Case B — huge $v_{e}$ but tiny $\overset{m}{˙}$ (ion thruster). $I_{s p} = v_{e} / g_{0}$ becomes very large (thousands of seconds) — superb efficiency. But thrust $F = \overset{m}{˙} v_{e}$ can still be minuscule because $\overset{m}{˙}$ is tiny. High $I_{s p}$ does not mean high thrust. See the Tsiolkovsky Rocket Equation for where high $v_{e}$ pays off.

Case C — non-constant $v_{e}$ . If $v_{e}$ varies, we cannot pull it out in Step 4. Then we must use the full definition $I_{s p} = \frac{\int F d t}{g _{0} \int m ˙ d t}$ — an average effective $v_{e}$ weighted by mass flow. The simple $v_{e} / g_{0}$ is the constant- $v_{e}$ special case.

PICTURE. Three mini-panels: (A) flat gas, zero thrust; (B) fast thin jet, big $I_{s p}$ tiny $F$ ; (C) a wiggly $v_{e} (t)$ curve that refuses to factor out.

The one-picture summary

Here is the entire chain in one figure: throw mass → get thrust → sum over the burn → factor out $v_{e}$ → divide by propellant weight → the mass cancels → seconds fall out.

Recall Feynman retelling — the whole walkthrough in plain words

A rocket moves by throwing stuff backward. Throw $\overset{m}{˙}$ kilograms every second, each at speed $v_{e}$ , and the backward momentum you hand out each second ( $\overset{m}{˙} v_{e}$ ) is exactly the forward push $F$ you feel. Run the engine for a whole burn and add up all those pushes over time — that total is the impulse $J$ . Because $v_{e}$ stays the same, you can pull it out front, and the pile of "kilograms thrown each second, added over all seconds" is just the whole tank of fuel, $m_{p}$ . So $J = v_{e} m_{p}$ . Now here is the clever move: divide the total push by the weight of the fuel you spent, $m_{p} g_{0}$ . The fuel amount $m_{p}$ appears on top and bottom, so it cancels — which is why the answer no longer cares whether your engine is tiny or huge. What survives is $v_{e} / g_{0}$ : exhaust speed over a fixed constant. Do the units and the newtons (or the metres-per-second) cancel too, leaving plain seconds — a universal report card for the engine, the same on Earth or Mars.

Recall

The three-step spine of the derivation ::: (1) $F = \overset{m}{˙} v_{e}$ , (2) $J = v_{e} m_{p}$ , (3) $I_{s p} = J / (m_{p} g_{0}) = v_{e} / g_{0}$ . Why does $m_{p}$ cancel and what does that guarantee ::: Impulse carries $m_{p}$ , weight carries $m_{p}$ ; dividing cancels it, so $I_{s p}$ is size-independent. Why seconds ::: N·s ÷ N = s, equivalently (m/s)/(m/s²)=s.

Connections

Rocket Propulsion
Thrust and Mass Flow Rate — Step 2 built $F = \overset{m}{˙} v_{e}$ here
Effective Exhaust Velocity — the $v_{e}$ in every step
Impulse-Momentum Theorem — Steps 1–3 are this theorem for a gas stream
Standard Gravity g0 — the fixed $g_{0}$ of Step 5
Tsiolkovsky Rocket Equation — where high $v_{e}$ (high $I_{s p}$ ) buys $Δ v$