This page hammers the parent equation Δ v = v e ln ( m 0 / m f ) through every kind of situation it can throw at you . We map the space of cases first, then work one example per cell. If you meet a scenario in an exam, it lives somewhere in the table below.
Before anything, one reminder of the meaning of each symbol, because we use them on every line:
Δ v = change in velocity the rocket can buy (m/s), a budget , not a final speed.
v e = effective exhaust velocity (m/s), how fast gas leaves relative to the rocket.
m 0 = wet mass (kg) = everything at the start, including all propellant.
m f = dry mass (kg) = everything left after the burn.
R = m 0 / m f = the mass ratio , a pure number with no units.
ln = the natural logarithm , the "shrinking machine" — see Natural Logarithm and Integration of 1/x .
e ≈ 2.718 = the base of that logarithm; e x undoes ln x .
The equation has only three knobs (v e , m 0 , m f ), but the questions asked about it fall into these classes:
Cell
Case class
What is unknown
Degenerate / edge check
A
Forward: given all masses + v e , find Δ v
Δ v
—
B
Degenerate: no fuel burned, m 0 = m f
Δ v
ln 1 = 0
C
Inverse for fuel: given target Δ v , find R
R , propellant mass
needs e x
D
Via specific impulse: v e not given, I s p is
Δ v
v e = g 0 I s p
E
Limiting behaviour: R → ∞ (all fuel, no dry mass)
ceiling of Δ v
ln R → ∞ slowly
F
Two-stage: chain two burns
total Δ v
Δ v add, ratios multiply
G
Word problem: real vehicle with losses
actual speed gained
subtract gravity loss
H
Exam twist: solve for m f given Δ v , m 0 , v e
m f
invert the log twice
Each example below is tagged with its cell letter. Together they fill every row.
Worked example A · Given masses and
v e , find Δ v
A rocket has m 0 = 500 , 000 kg, m f = 100 , 000 kg, v e = 3000 m/s. Find Δ v .
Forecast: the ratio is 5, and ln 5 ≈ 1.6 , so guess a bit under 3000 × 1.6 ≈ 4800 m/s. Hold that number.
Compute the mass ratio R = m 0 / m f = 500000/100000 = 5 .
Why this step? Δ v depends on the ratio of masses, not their difference, so the ratio is the first thing to nail down.
Take the logarithm: ln 5 = 1.609 .
Why this step? The equation feeds R through ln because velocity gain accumulates as d v = − v e d m / m ; the 1/ m integrates to a log.
Multiply by v e : Δ v = 3000 × 1.609 = 4828 m/s.
Why this step? Each fractional bit of mass lost buys v e times that fraction in speed.
Verify: units are ( m/s ) × ( dimensionless ) = m/s ✔. Our forecast (~4800) matches. Propellant burned = m 0 − m f = 400 , 000 kg, less than the total — sane ✔.
Worked example B · No fuel burned
Same rocket, but the engine never fires: m 0 = m f = 500 , 000 kg. Find Δ v .
Forecast: you threw no mass backward, so Newton's 3rd law gives you no kick. Guess Δ v = 0 .
Mass ratio R = 500000/500000 = 1 .
Why this step? Nothing was ejected, so wet and dry mass are identical.
ln 1 = 0 .
Why this step? ln answers "to what power do I raise e to get this number?" Since e 0 = 1 , the answer for input 1 is 0 .
Δ v = 3000 × 0 = 0 m/s.
Verify: matches the physics — no exhaust, no thrust, no speed change ✔. This is the anchor that tells you the whole formula is built correctly.
Worked example C · Given target
Δ v , find the mass ratio
You must achieve Δ v = 9400 m/s (a typical low-Earth-orbit budget) with v e = 3400 m/s. What mass ratio R do you need, and what fraction of the rocket is propellant?
Forecast: 9400/3400 ≈ 2.8 , and e 2.8 is a big number — guess R around 15, i.e. most of the rocket is fuel.
Isolate the log: divide both sides by v e .
ln R = v e Δ v = 3400 9400 = 2.765
Why this step? We want R , and it is trapped inside ln ; peeling off v e first exposes the log.
Undo the log with e x :
R = e 2.765 = 15.88
Why this step? Because e l n R = R — exponentiation is the exact inverse of the natural log.
Propellant fraction: if R = m 0 / m f = 15.88 , then m f / m 0 = 1/15.88 = 0.063 , so propellant is 1 − 0.063 = 0.937 , about 93.7% .
Why this step? The dry fraction is the reciprocal of the ratio; everything else is fuel.
Verify: plug back — 3400 × ln ( 15.88 ) = 3400 × 2.765 = 9401 m/s ≈ target ✔. Nearly 94% propellant is the famous "tyranny of the rocket equation."
v e not given, only I s p
An engine advertises I s p = 450 s. The dry mass (structure + payload) is m f = 20 , 000 kg and it carries 300 , 000 kg of propellant. Find Δ v . Use g 0 = 9.81 m/s².
Forecast: 450 s is a high-performance hydrogen engine, so v e is around 4400 m/s; with lots of fuel expect Δ v over 10 km/s.
Convert I s p to exhaust velocity: v e = g 0 I s p = 9.81 × 450 = 4414.5 m/s.
Why this step? The equation needs a speed , but I s p is measured in seconds; see Specific Impulse (Isp) . The bridge is v e = g 0 I s p .
Build the wet mass: m 0 = 300000 + 20000 = 320000 kg.
Why this step? m 0 must include all propellant plus everything that stays.
Mass ratio R = 320000/20000 = 16 , and ln 16 = 2.773 .
Why this step? Same core computation as always — feed the ratio through the log.
Δ v = 4414.5 × 2.773 = 12 , 241 m/s.
Verify: units — ( m/s 2 ) ( s ) = m/s for v e ✔, then × dimensionless ✔. Value comfortably exceeds LEO budget, as forecast ✔.
Worked example E · What is the most
Δ v one stage can ever give?
Suppose v e = 3400 m/s is fixed. If you could magically shrink the dry mass toward zero (all fuel, no structure), what happens to Δ v as R → ∞ ?
Forecast: ln grows without bound, so there is no hard ceiling — but it grows painfully slowly . Guess: doubling R only adds a fixed lump of Δ v .
Look at the growth of ln R as R increases. Each multiplication of R by a fixed factor adds a fixed amount of ln .
Why this step? Because ln ( k R ) = ln k + ln R — a constant ln k is added no matter how big R already is. See the figure.
Concretely, each time you double R , Δ v rises by v e ln 2 = 3400 × 0.693 = 2356 m/s — the same jump every doubling.
Why this step? ln ( 2 R ) − ln R = ln 2 , a constant, so the Δ v step per doubling is constant.
Conclusion: Δ v → ∞ as R → ∞ , but only logarithmically. To gain another 2356 m/s you must double the whole mass ratio again — the cost explodes.
Verify: check the doubling law numerically. From R = 8 to R = 16 : 3400 ( ln 16 − ln 8 ) = 3400 × 0.693 = 2356 m/s ✔. From R = 16 to R = 32 : same 2356 m/s ✔. Equal steps confirm log growth.
Worked example F · Two-stage rocket, add the
Δ v s
Stage 1: m 0 = 100 , 000 kg, burns down to m f = 40 , 000 kg, v e = 3000 m/s. Stage 2 (after dropping the empty stage-1 hardware): m 0 = 30 , 000 kg, m f = 8 , 000 kg, v e = 4000 m/s. Find total Δ v .
Forecast: each stage is a modest ratio (2.5 and 3.75), so each gives a couple of km/s; total maybe 6–8 km/s.
Stage 1: R 1 = 100000/40000 = 2.5 , Δ v 1 = 3000 ln 2.5 = 3000 × 0.9163 = 2749 m/s.
Why this step? Each stage obeys the same rocket equation independently.
Stage 2: R 2 = 30000/8000 = 3.75 , Δ v 2 = 4000 ln 3.75 = 4000 × 1.3218 = 5287 m/s.
Why this step? After separation the empty tank is gone, so stage 2 starts with a fresh, lighter wet mass — that is exactly why staging beats the log; see Multistage Rockets .
Add them: Δ v total = 2749 + 5287 = 8036 m/s.
Why this step? Δ v is a change in velocity; sequential changes simply add.
Verify: ratios multiply into an effective 2.5 × 3.75 = 9.375 , but because the two stages have different v e you cannot use one log — you must add the two separately, which we did ✔. Total in the km/s range as forecast ✔.
Worked example G · Real launch, subtract gravity loss
A booster has m 0 = 550 , 000 kg, m f = 150 , 000 kg, v e = 3100 m/s. During its 150-second vertical climb it loses roughly 150 × 9.81 = 1472 m/s to gravity drag. What actual speed gain does the rocket end with?
Forecast: ideal Δ v near 4000 m/s, minus ~1500 loss, so around 2500 m/s of real speed.
Ideal Δ v : R = 550000/150000 = 3.667 , so Δ v ideal = 3100 ln 3.667 = 3100 × 1.2993 = 4028 m/s.
Why this step? The Tsiolkovsky formula gives the budget assuming no external forces.
Subtract gravity loss: real speed gain = 4028 − 1472 = 2556 m/s.
Why this step? Gravity pulls backward the whole climb; that momentum debt comes straight out of the ideal budget. Δ v is a budget, not the final speed — this is the classic mistake the parent note warns about.
Verify: the real gain (2556) is less than ideal (4028) ✔ — losses always shrink the budget, never grow it. Order-of-magnitude matches forecast ✔.
Δ v , m 0 , v e , find m f
A design must hit Δ v = 6000 m/s with v e = 3200 m/s and a fixed wet mass m 0 = 80 , 000 kg. What dry mass m f is allowed, and how much propellant does that imply?
Forecast: 6000/3200 ≈ 1.9 , so R = e 1.9 ≈ 6.6 ; dry mass is m 0 divided by that, roughly 12,000 kg.
Get the required log: ln R = Δ v / v e = 6000/3200 = 1.875 .
Why this step? Same isolation as Cell C — strip off v e to expose the log.
Exponentiate for the ratio: R = e 1.875 = 6.521 .
Why this step? e x undoes ln , giving the pure mass ratio.
Solve for dry mass: m f = m 0 / R = 80000/6.521 = 12 , 268 kg.
Why this step? Since R = m 0 / m f , rearranging gives m f = m 0 / R .
Propellant needed: m 0 − m f = 80000 − 12268 = 67 , 732 kg.
Why this step? Propellant burned is always wet minus dry.
Verify: plug back — 3200 ln ( 80000/12268 ) = 3200 ln ( 6.521 ) = 3200 × 1.875 = 6000 m/s ✔. Dry mass matches forecast (~12,000 kg) ✔.
Recall Which cell is each exam question?
"Find Δv from masses" ::: Cell A.
"Engine off, mass unchanged" ::: Cell B, answer 0.
"How much fuel for this Δv?" ::: Cell C (invert with e x ).
"Only I s p given" ::: Cell D (use v e = g 0 I s p ).
"Why does more fuel help so little?" ::: Cell E (log grows slowly).
"Two stages, total Δv?" ::: Cell F (add the Δv's).
"Real launch speed?" ::: Cell G (subtract gravity loss).
"Find the allowed dry mass" ::: Cell H (solve m f = m 0 / e Δ v / v e ).
Mnemonic The inversion pair
Forward burns a log: Δ v = v e ln R . Backward un-logs it: R = e Δ v / v e . Whenever the unknown is inside the log, reach for e .
Parent topic — the equation itself.
Conservation of Momentum — the derivation these examples rest on.
Newton's Third Law — why throwing mass gives thrust.
Specific Impulse (Isp) — used in Cell D.
Multistage Rockets — the physics behind Cell F.
Thrust and Mass Flow Rate — differential cousin behind Cell G's climb.
Natural Logarithm and Integration of 1/x — why ln and e pair up.