Tsiolkovsky rocket equation — full first-principles derivation from momentum: Question bank

The one equation everything here orbits: $Δ v = v_{e} ln (\frac{m _{0}}{m _{f}})$ where $v_{e}$ = exhaust speed relative to the rocket, $m_{0}$ = initial mass (rocket + all fuel), $m_{f}$ = final mass (rocket + leftover fuel), and $R = m_{0} / m_{f}$ is the mass ratio. If any of those words feel shaky, go back to the parent note first.

Before you start, pin down the symbols and sign conventions that trip everyone up:

Definition Symbols & sign conventions used on this page

$v$ = the rocket's own speed measured in the ground frame (relative to a fixed observer on the ground). It starts small and grows as fuel burns.
$v_{e}$ = exhaust speed measured relative to the rocket (fixed by the engine's nozzle, taken constant). Because the gas leaves backward at $v_{e}$ relative to a rocket moving at $v$ , its ground-frame velocity is $v - v_{e}$ .
$m$ = the rocket's mass at any instant. As fuel burns, $m$ shrinks, so its tiny change $d m$ is negative ( $d m < 0$ ). The little chunk of gas thrown out is therefore $- d m$ , a positive amount.
$\overset{m}{˙} = - \frac{d m}{d t} > 0$ = the mass-flow rate, i.e. how many kilograms per second the engine throws out. The minus sign is exactly what makes this positive, since $d m / d t < 0$ . This is the variable used in thrust.
$R = m_{0} / m_{f} > 1$ = the mass ratio; since you start heavier and burn down, $R$ is always bigger than 1 and $ln R$ is always positive.

The two facts you will lean on repeatedly, both from Conservation of Momentum applied to a variable-mass system: $d v = - v_{e} \frac{d m}{m}, Δ v = v_{e} ln R .$ The first says a small speed gain $d v$ is bought with a small fractional mass loss $d m / m$ . Keep the reference figure below in view while you work through the traps.

Figure — Tsiolkovsky rocket equation — full first-principles derivation from momentum

True or false — justify

The rocket equation says total momentum is conserved, so the rocket cannot gain speed.

False. Momentum of the whole system (rocket + all ejected gas) is conserved, but conservation constrains the total, not each piece. The gas is thrown backward and carries negative momentum; to keep the sum fixed the rocket must carry an equal positive momentum, i.e. speed up. That trade is the whole mechanism — see Conservation of Momentum.

Doubling the mass of propellant loaded (keeping the empty rocket

m_{f}

the same) doubles the final

Δ v

False, and note the phrasing: if

m_{f}

is fixed and you double the fuel, then

m_{0} = m_{f} + fuel

does not double — you only enlarge the added part. Even in the best case where the ratio

R

itself doubled,

Δ v = v_{e} ln (2 R) = v_{e} ln R + v_{e} ln 2

, which merely adds a fixed

v_{e} ln 2

, not multiplies

Δ v

by two. The logarithm converts multiplying the ratio into adding a constant.

The rocket can never travel faster than its own exhaust speed

v_{e}

False. Since

Δ v = v_{e} ln R

and

ln R

grows without bound as

R

grows, you get

Δ v > v_{e}

the moment

ln R > 1

, i.e.

R > e \approx 2.718

. Physically the rocket receives a fresh push from every kilogram of gas, and each push acts on an ever-lighter vehicle, so the boosts accumulate past

v_{e}

If a rocket ejects gas but its mass stays the same, it still gains

Δ v

False. Speed is bought by

d v = - v_{e} d m / m

; with no change in mass there is no

d m

, so

d v = 0

. "Ejecting mass" and "losing mass" are literally the same physical event — you cannot have one without the other, and it is the mass loss that pays for the speed.

In the ground frame, the exhaust gas always moves backward.

False. The gas's ground-frame velocity is

v - v_{e}

(rocket's own ground speed minus the relative exhaust speed). While

v < v_{e}

this is negative (backward), but once the rocket's ground speed

v

exceeds

v_{e}

the quantity

v - v_{e}

turns positive — the gas drifts forward in the ground frame, just slower than the rocket that left it behind.

Two rockets with identical mass ratios but different exhaust speeds get the same

Δ v

False. With the same

R

the factor

ln R

is identical, but

Δ v = v_{e} ln R

scales directly with

v_{e}

. Twice the exhaust speed means twice the

Δ v

for the same fuel fraction — which is exactly why engineers chase high Specific Impulse engines rather than just loading more fuel.

A heavier rocket (larger

m_{0}

) always reaches a higher

Δ v

False. The equation contains only the ratio

m_{0} / m_{f}

, never the absolute masses. A 20 kg model and a 200 t launcher that share the same

R

and the same

v_{e}

reach identical

Δ v

. Scale cancels out entirely.

The equation holds during launch straight off the pad.

False. The clean form

Δ v = v_{e} ln R

assumes free space — no gravity, no drag. On the pad, gravity continuously robs momentum, subtracting a

\int g d t

gravity-loss term (Gravity Loss and Drag Loss), so the achieved

Δ v

falls short of

v_{e} ln R

Newton's Second Law in the familiar

F = ma

form is enough to derive the equation.

False (subtly).

F = ma

silently assumes fixed mass. Here mass leaves the vehicle every instant, carrying its own momentum, so you must use the momentum form

F = d p / d t

and bookkeep the ejected gas explicitly. Skip that and the

v d m

terms go missing. See Newton's Second Law (momentum form) and Variable Mass Systems.

Spot the error

"Ejected gas mass is

d m > 0

because mass is leaving the rocket."

Error:

m

denotes the rocket's mass, which decreases, so its change

d m < 0

. The chunk of gas thrown out is

- d m

, which is positive. It feels like "mass leaving = positive amount," but the sign attaches to whose mass we are tracking; flip it and the whole result's sign flips.

"The gas leaves at

v_{e}

, so in the ground frame the gas moves at

v_{e}

Error:

v_{e}

is measured relative to the rocket, which is itself moving at

v

in the ground frame. The ground-frame gas velocity is

v - v_{e}

, not

v_{e}

. Using the constant

v_{e}

where a ground-frame value belongs breaks the clean separation of variables and corrupts the integral.

"We keep the

d m d v

term to be more accurate."

Error:

d m d v

is second-order — a product of two infinitesimally small quantities — so it shrinks faster than the first-order terms and vanishes in the limit

d t \to 0

. Keeping it adds nothing real; the calculus limit requires dropping it.

\int \frac{d m}{m} = m

, so integrating gives

Δ v = - v_{e} (m_{f} - m_{0})

Error:

\int \frac{d m}{m} = ln m

, not

m

. The entire logarithm — and with it every "diminishing returns" insight — lives precisely in the fact that you are integrating

1/ m

, not

1

"Thrust equals

v_{e}

alone."

Error: Thrust

= v_{e} \overset{m}{˙}

, where

\overset{m}{˙} = - d m / d t > 0

is the mass-flow rate defined at the top of this page (see Thrust and Mass Flow Rate). Force is a rate, so you need both how fast you throw the gas (

v_{e}

) and how much per second (

\overset{m}{˙}

);

v_{e}

by itself has the wrong units for a force.

"Since

m_{f} > m_{0}

, the log is negative, so

Δ v < 0

Error:

m_{0} > m_{f}

— you start heavier (full of fuel) and burn down to

m_{f}

. So the ratio

R = m_{0} / m_{f} > 1

and

ln R > 0

. Always write big-over-small to keep the log positive.

"A multistage rocket is just one big rocket, so staging can't beat the equation."

Error: The equation still governs each stage, but by dropping empty tanks you shrink

m_{f}

for every remaining stage, raising its ratio

R

. Staging doesn't break the equation — it exploits the log by resetting the mass ratio favourably at each drop. See Multistage Rockets.

Why questions

Why is the relationship logarithmic instead of linear?

Because each speed gain obeys

d v = - v_{e} d m / m

: the cost is measured as a fractional mass change

d m / m

, not an absolute one. Summing up fractional changes across the burn is precisely the operation

\int 1/ m d m = ln m

. So equal fractions of mass burned buy equal chunks of speed, and that "equal-ratio-gives-equal-boost" behaviour is the defining fingerprint of a logarithm.

Why must

v_{e}

be constant for the clean formula to hold?

v_{e}

varied with mass or time it could not be pulled outside the integral

\int v_{e} d m / m

, and the result would no longer collapse to a single

v_{e} ln R

. Constant

v_{e}

is exactly the property that lets the exhaust speed factor out and leave a pure logarithm behind.

Why does the derivation need the momentum form of Newton's law rather than

F = ma

Because the system's mass changes each instant.

F = ma

assumes

m

is fixed and would miss the momentum carried off by the departing gas (the

v d m

terms). The momentum form

F = d p / d t

tracks the total

p = m v

correctly even as

m

drops, which is what a variable-mass problem demands.

Why does burning half of what remains give the same

Δ v

boost as burning the first half?

Because each halving multiplies the mass ratio by a factor of 2, and the log converts that multiplication into an addition:

ln (2 R) - ln R = ln 2

. So every halving adds the same fixed

v_{e} ln 2

, no matter how much mass is left in absolute terms — equal ratios, equal boosts.

Why is roughly 88% of an orbital rocket typically propellant?

To reach

Δ v \approx 9.4

km/s with

v_{e} \approx 4.5

km/s you need

R = m_{0} / m_{f} = e^{9.4/4.5} \approx 8

. That means

m_{f} / m_{0} \approx 1/8

, so the leftover (structure + payload) is only about an eighth and the propellant is the other

\approx 7/8 \approx 88%

. Because

Δ v

grows only as

ln R

, pushing

R

from 8 upward costs enormous extra fuel — the log punishes a modest

v_{e}

hard.

Why does a rocket in free space keep accelerating even though total momentum is conserved?

Conservation fixes only the combined momentum of rocket-plus-exhaust, not the momentum of either part alone. The rocket continually hands backward momentum to the gas and, to keep the sum unchanged, gains an equal forward momentum itself. Every ejected kilogram is another such trade, so the rocket keeps speeding up.

Edge cases

What happens to

Δ v

m_{f} = m_{0}

(no fuel burned)?

The mass ratio is

R = 1

, and

ln 1 = 0

, so

Δ v = 0

. No mass thrown means no momentum traded and no speed gained — the equation degenerates exactly the way physical sense demands.

What does

Δ v

approach as

m_{f} \to 0

(burn everything)?

The ratio

R = m_{0} / m_{f} \to \infty

, so

ln R \to \infty

and

Δ v \to \infty

in the idealised equation. It is physically unreachable — you can't shrink the structural mass to zero — but it shows there is no mathematical speed ceiling, only a payload-and-structure one.

v_{e} = 0

(gas dumped with zero relative speed), what is

Δ v

Zero. With

v_{e} = 0

the ejected mass carries no momentum relative to the rocket, so there is nothing to react against (

Δ v = 0 \cdot ln R = 0

). Dropping mass gently does not propel you; you must throw it.

At the instant the rocket's ground speed

v

equals

v_{e}

, what is the exhaust's ground-frame velocity?

Exactly zero, since the gas moves at

v - v_{e} = v_{e} - v_{e} = 0

. The gas is momentarily "left hanging" motionless in the ground frame — a clean sanity check that

v - v_{e}

is the right ground-frame expression.

Does the equation depend on the direction of gravity or any external field?

In the ideal free-space form, no — it contains no

g

at all. External fields enter only as separate correction terms (Gravity Loss and Drag Loss); the core momentum bookkeeping that produces

v_{e} ln R

is field-free.

If the rocket ejected mass forward instead of backward, what changes?

The relative exhaust velocity reverses sign, effectively flipping the sign of

v_{e}

, so

Δ v

becomes negative — the rocket slows down or reverses. Direction of ejection sets the direction of thrust, and the algebra carries that through the sign of

v_{e}

Recall One-line self-test before you leave

Cover everything: Why a logarithm? Because equal fractions of mass burned buy equal speed boosts, and adding up fractional mass changes ( $\int d m / m$ ) is the definition of $ln$ .