The Relativistic Rocket Equation and Traveler Time

Andromeda is $2.5$ million light-years away. Fly there at a steady $1\,g$ — accelerating for the first half, decelerating for the second — and you arrive having aged 28.6 years. Not 2.5 million. Twenty-eight. A crew that leaves in their twenties walks out onto another galaxy in their fifties, and the Milky Way behind them is two and a half million years older. That is the seduction, and it is entirely correct. Constant proper acceleration converts distance into rapidity, and rapidity accumulates linearly with the traveller's own clock — so that clock grows only logarithmically with the distance covered. The galaxy is a short walk, if you are willing to arrive very late. Then you ask what it costs. The fuel budget is governed by the same rapidity, and it enters as an exponential: even a perfect photon rocket, converting matter to a collimated beam of light at one hundred per cent efficiency, needs a mass ratio of $e^{\Delta\varphi}$ . For Andromeda that is $\mathbf{6.9\times10^{12}}$ — seven trillion kilograms of fuel per kilogram delivered. The rapidity that makes the time trivial is the same rapidity that makes the fuel impossible; it is simply sitting on the wrong side of the exponential. That is this article.

Key takeaways

Signature $(+,-,-,-)$ : $\eta_{\mu\nu} = \mathrm{diag}(+1,-1,-1,-1)$ , $ds^2 = c^2dt^2 - dx^2$ , timelike intervals positive.
The worldline under constant proper acceleration is hyperbolic motion — derived in `aq-sr-relativistic-uniform-acceleration`, recalled here, not re-derived: $x = \frac{c^2}{g}\big(\cosh\frac{g\tau}{c} - 1\big)$ , $\;t = \frac{c}{g}\sinh\frac{g\tau}{c}$ , $\;\varphi = g\tau/c$ .
Owned result — traveller time. Inverting the worldline gives $\;\boxed{\tau = \frac{c}{g}\,\mathrm{arccosh}\!\left(1 + \frac{gd}{c^{2}}\right)}$ against a coordinate time $t = \frac{c}{g}\sqrt{\left(1 + \frac{gd}{c^{2}}\right)^{2} - 1}$ . For large $d$ the coordinate time grows linearly and the proper time only logarithmically. That gap is the whole promise of relativistic flight.
Owned result — the rapidity budget, the bridge between fuel and motion: $\;\boxed{\Delta\varphi = \frac{v_{\text{ex}}}{c}\,\ln\!\frac{M_i}{M_f}}$ . Rapidity is what adds under successive boosts, so the Tsiolkovsky structure survives relativity intact — with $\Delta\varphi$ in the place of $\Delta v$ . Velocity does not add; rapidity does.
Owned result — the reality check that kills the fantasy. A photon rocket ( $v_{\text{ex}} = c$ ) needs $\;\boxed{M_i/M_f = e^{\Delta\varphi}}$ . Proxima Centauri, there and stopped: $\mathbf{38.7:1}$ . Andromeda: $\mathbf{6.9\times10^{12}:1}$ . A fusion drive at $v_{\text{ex}} = 0.1c$ needs $7.6\times10^{15}:1$ for Proxima alone; a chemical rocket needs $10^{108{,}000}:1$ .
The natural scales: $c/g = 3.06\times10^{7}\ \mathrm{s} = 0.97\ \mathrm{yr}$ and $c^{2}/g = 9.16\times10^{15}\ \mathrm{m} = 0.97$ light-year.
Key idea: proper time is logarithmic in the distance and the mass ratio is exponential in the rapidity. The same $\Delta\varphi$ appears in both — once inside a $\log$ , once inside an $\exp$ . Relativity gives the traveller the stars and hands the engineer the bill, and the bill is written in the exponential.

The core insight

There is exactly one number in this problem: the total rapidity $\Delta\varphi$ the ship must expend. Everything else is a function of it. The distance goes as $\cosh\Delta\varphi$ — exponentially large. The time you age goes as $\Delta\varphi$ itself, linearly, since $\varphi = g\tau/c$ . And the fuel goes as $e^{\Delta\varphi\,c/v_{\text{ex}}}$ — exponentially large again, with $c/v_{\text{ex}} \ge 1$ and enormous for any real engine. The traveller's ageing is the logarithm of the distance; the mass ratio is a power of it. That asymmetry is why interstellar flight looks easy from the cabin and impossible from the fuel tank.

Where this lives today

Nothing built or seriously proposed comes within many orders of magnitude of the mass ratios above, and the honest engineering response has been to abandon the rocket. Light sails carry no propellant and so have no mass ratio: Breakthrough Starshot accelerates gram-scale sails with a ground-based laser array, reaching $0.2c$ and a Proxima flyby in about twenty years — no deceleration, no crew, no fuel aboard. That is not a relativistic rocket; it is the deliberate refusal to be one, and equation (3) is exactly why. The rapidity budget itself, $\Delta\varphi = (v_{\text{ex}}/c)\ln(M_i/M_f)$ , is also the cleanest demonstration that rapidity — not velocity — is the physically additive measure of motion: the same fact that makes hyperbolic motion linear in $\varphi$ , and the Lorentz group a one-parameter additive group along a boost axis.

The physical picture

Tsiolkovsky's rocket equation says a ship throwing mass backwards at speed $v_{\text{ex}}$ gains speed logarithmically in the mass it throws: $\Delta v = v_{\text{ex}}\ln(M_i/M_f)$ . It is the most quoted result in astronautics and the reason chemical rockets are mostly fuel tank. It also cannot survive relativity: it is unbounded. Burn enough fuel and $\Delta v$ exceeds $c$ . The repair is not a correction factor. It is a change of variable. Velocities do not add — that is the whole point of the composition law — but rapidities do: $\varphi = \operatorname{artanh}(v/c)$ , and successive boosts compose by $\varphi_1 + \varphi_2$ (`aq-sr-rapidity-hyperbolic-boosts`). A rocket burn is exactly a sequence of infinitesimal boosts, each performed in the frame the ship is in at that moment. So the quantity that accumulates additively across the burn is not $v$ but $\varphi$ , and the rocket equation reappears, structurally identical, with $\Delta\varphi$ in place of $\Delta v$ :

\Delta\varphi = \frac{v_{\text{ex}}}{c}\,\ln\!\frac{M_i}{M_f}.

Nothing is bounded any more, because $\varphi$ is unbounded — but $v = c\tanh\varphi$ is, automatically, so the speed limit takes care of itself. The engineering content is unchanged: fuel buys rapidity, logarithmically. Now turn it round, which is what nobody does. Solve for the mass ratio:

\frac{M_i}{M_f} = \exp\!\left(\frac{c}{v_{\text{ex}}}\,\Delta\varphi\right).

The rapidity for a real trip is of order a few — $1.83$ for a one-way $1\,g$ hop to Proxima, $14.8$ for Andromeda. Tame-looking. But it sits in an exponent, multiplied by $c/v_{\text{ex}}$ : $68{,}000$ for a chemical rocket, $10$ for an optimistic fusion drive. Only a photon rocket, throwing exhaust backwards at $c$ itself, gets $c/v_{\text{ex}} = 1$ — and even then the mass ratio is $e^{\Delta\varphi}$ , seven trillion for Andromeda. Meanwhile the traveller's clock does something else. Under constant proper acceleration $x = \frac{c^2}{g}(\cosh\varphi - 1)$ with $\varphi = g\tau/c$ , so distance is exponential in the traveller's own time — and inverting, the traveller's time is logarithmic in the distance. Ten times further costs $\ln 10$ of rapidity, about $2.2$ years of extra ageing. That is the seduction; the bill for those two years is a factor of ten in the mass ratio.

Figure

PUNCHLINE — the traveller's clock is LOGARITHMIC in the distance; the fuel is EXPONENTIAL in the traveller's clock. Both axes are log. Earth's clock t (red) grows linearly with distance — you cannot beat light. The crew's clock τ (blue) barely moves: Proxima 3.5 yr, the galactic centre 19.8 yr, Andromeda 28.6 yr. Twenty-eight years of crew time to cross 2.5 million light-years. Then read the numbers in the boxes: the photon-rocket mass ratio, the BEST any engine can do, for each of those trips. Proxima 38.7 to 1. Galactic centre 720 million to 1. Andromeda 6.9 trillion to 1. The same rapidity Δφ that enters the crew's clock LINEARLY enters the fuel bill as an EXPONENT — and that, not the speed of light, is what stops us

Figure

The exponential wall. For the Proxima trip (Δφ = 3.657, there and stopped) the mass ratio is exp(Δφ · c/v_ex), plotted here as log₁₀ of the mass ratio against the exhaust speed. The curve is not steep — it is a HYPERBOLA IN THE EXPONENT, and it goes vertical as v_ex → 0. A photon rocket sits at v_ex = c and pays 38.7 to 1: that is the FLOOR, the best relativity permits, because nothing can be thrown backwards faster than light. Halve the exhaust speed to 0.5c and the mass ratio SQUARES to 1,500. An optimistic fusion drive at 0.1c pays 7.6×10¹⁵. A chemical rocket at 4.4 km/s — v_ex/c = 1.5×10⁻⁵ — pays 10^108,000, a number with more digits than the observable universe has atoms. The wall is not technological. It is the rocket equation

The core equations

The worldline is not derived here. Constant proper acceleration gives hyperbolic motion, and that is `aq-sr-relativistic-uniform-acceleration`'s result; we take it and put it to work. Starting from rest at the origin:

\varphi = \frac{g\tau}{c}, \qquad x(\tau) = \frac{c^{2}}{g}\left(\cosh\frac{g\tau}{c} - 1\right), \qquad t(\tau) = \frac{c}{g}\sinh\frac{g\tau}{c} \qquad (1)

where: $\tau$ is the crew's clock, $t$ Earth's, $\varphi$ the rapidity; $x$ is measured from the launch point, which is what the $-1$ is for. Invert the first of these for $\tau$ as a function of the distance flown — the article's central object:

\boxed{\;\tau(d) = \frac{c}{g}\,\mathrm{arccosh}\!\left(1 + \frac{g\,d}{c^{2}}\right)\;} \qquad\qquad \boxed{\;t(d) = \frac{c}{g}\sqrt{\left(1 + \frac{g\,d}{c^{2}}\right)^{2} - 1}\;} \qquad (2)

where: $d$ is the distance flown under continuous acceleration. For $gd/c^2 \gg 1$ , $\mathrm{arccosh}(1+u) \to \ln(2u)$ , so $\tau \to \frac{c}{g}\ln\!\big(2gd/c^2\big)$ — logarithmic — while $t \to d/c$ , linear and never less than the light-travel time. The rapidity a burn buys, from momentum conservation in the momentarily comoving frame:

\boxed{\;\Delta\varphi = \frac{v_{\text{ex}}}{c}\,\ln\!\frac{M_{i}}{M_{f}}\;} \qquad\Longleftrightarrow\qquad \boxed{\;\frac{M_{i}}{M_{f}} = \exp\!\left(\frac{c}{v_{\text{ex}}}\,\Delta\varphi\right)\;} \qquad (3)

where: $v_{\text{ex}}$ is the exhaust speed relative to the ship. This is Tsiolkovsky with $\Delta v \to c\,\Delta\varphi$ , and the substitution is exact, not approximate: rapidity is the additive quantity, so it is what the engine accumulates. For a photon rocket — the theoretical maximum, exhaust at $c$ — the factor $c/v_{\text{ex}}$ becomes $1$ :

\boxed{\;\frac{M_{i}}{M_{f}} = e^{\Delta\varphi}\;} \qquad (\text{photon rocket, } v_{\text{ex}} = c) \qquad (4)

where: $\Delta\varphi$ is the total rapidity expended, counting every burn — a trip that accelerates to $\varphi_{\max}$ and then stops spends $2\varphi_{\max}$ . No engine can do better than this, whatever it is made of, because nothing can be thrown backwards faster than light. Equation (4) is a floor imposed by relativity itself. And the scales, with $g = 9.81\ \mathrm{m/s^2}$ :

\frac{c}{g} = 3.06\times10^{7}\ \mathrm{s} = \mathbf{0.97\ yr}, \qquad \frac{c^{2}}{g} = 9.16\times10^{15}\ \mathrm{m} = \mathbf{0.97\ light\text{-}year} \qquad (5)

Deriving the two halves

Derivation AThe rapidity budget: Tsiolkovsky survives, with rapidity in place of velocity

The goal is equation (3), and it deserves care because it is the bridge of the whole article — the only place where the fuel tank and the worldline touch. Everything about the trip's time follows from hyperbolic motion; everything about its cost follows from this one relation. The derivation is three lines, and the crucial move is to do the bookkeeping in the momentarily comoving frame, then recognise that what accumulates across successive such frames is rapidity, not velocity.

Step 1. Work in the MCIF. The ship has rest mass $M$ and is momentarily at rest; it ejects exhaust of rest mass $dm_{\text{ex}}$ backwards at $v_{\text{ex}}$ and recoils to $dv$ . To first order, energy and momentum conservation read $Mc^{2} = (M + dM)c^{2} + \gamma_{\text{ex}}dm_{\text{ex}}c^{2}$ and $0 = (M+dM)dv - \gamma_{\text{ex}}dm_{\text{ex}}v_{\text{ex}}$ . Why: $dM$ is negative, and the energy equation says the exhaust's total energy, not its rest mass, comes out of the ship's rest mass. That is the relativistic content: the tank gets lighter by more than the mass of the propellant, because the binding energy leaves too.

Step 2. Eliminate the exhaust. Energy gives $\gamma_{\text{ex}}dm_{\text{ex}} = -dM$ ; substituting into momentum and dropping second-order terms, $dv = -v_{\text{ex}}\,dM/M$ . Why: Tsiolkovsky's differential relation, exactly Newtonian in form. Nothing relativistic is lost, because in the MCIF the ship is at rest and everything is first order.

Step 3. Recognise what $dv$ is. In the MCIF the ship goes from rest to $dv$ , which is a boost of rapidity $d\varphi = \operatorname{artanh}(dv/c) = dv/c$ to first order. And rapidities add across successive boosts. So the ship's total rapidity in the launch frame is the sum of these increments: $d\varphi = -\dfrac{v_{\text{ex}}}{c}\dfrac{dM}{M}$ . Why: the whole derivation. Add the $dv$ 's and you are adding velocities, which is illegal, and you get a rocket that exceeds $c$ . Rapidity is the additive quantity, so rapidity is what the engine banks.

Step 4. Integrate. $\displaystyle\int_{0}^{\Delta\varphi} d\varphi = -\frac{v_{\text{ex}}}{c}\int_{M_i}^{M_f}\frac{dM}{M}$ gives $\Delta\varphi = \dfrac{v_{\text{ex}}}{c}\ln\dfrac{M_i}{M_f}$ — equation (3). Exponentiating, $M_i/M_f = \exp\big(\tfrac{c}{v_{\text{ex}}}\Delta\varphi\big)$ . Why: note the exact Newtonian limit — for $v_{\text{ex}} \ll c$ , $\Delta v \approx c\,\Delta\varphi = v_{\text{ex}}\ln(M_i/M_f)$ , Tsiolkovsky recovered without a scratch. Relativity did not modify the rocket equation. It told us which variable it was always about.

Step 5. Note the floor. Since $v_{\text{ex}} \le c$ for any physically possible exhaust, $c/v_{\text{ex}} \ge 1$ , so $M_i/M_f \ge e^{\Delta\varphi}$ for every conceivable engine. Why: a theorem, not an engineering estimate. It does not care what the fuel is, how the reactor works, or what the year is. The photon rocket saturates it; nothing beats it. *Verified in verify_aq-sr-relativistic-rocket-equation-s6.py: the MCIF conservation equations are solved symbolically to first order and yield $dv = -v_{\text{ex}}dM/M$ ; the integration is checked; the Newtonian limit $c\,\Delta\varphi \to v_{\text{ex}}\ln(M_i/M_f)$ is recovered; and $M_i/M_f \ge e^{\Delta\varphi}$ is asserted for a scan of $v_{\text{ex}} \le c$ .*

Go deeper · Derivation B · 2 minTraveller time: logarithmic in the distance, and the arccosh that makes it so

Derivation BTraveller time: logarithmic in the distance, and the arccosh that makes it so

Equation (2) is where the promise of relativistic flight lives, and inverting the hyperbolic worldline is all that stands between a reader and it. The goal is $\tau(d)$ and $t(d)$ , and then — the part that matters — the asymptotics, because the headline claim is not any single number but a statement about growth rates: the coordinate time is linear in $d$ , the proper time logarithmic. That is what licenses "twenty-eight years to Andromeda".

Step 1. Invert the position. From (1), $\cosh(g\tau/c) = 1 + gd/c^{2}$ , so $\tau = \frac{c}{g}\,\mathrm{arccosh}(1 + gd/c^{2})$ — equation (2). Why: the dimensionless group is $gd/c^{2}$ , the distance in units of the $0.97$ -light-year scale. Everything in this article is a function of that one number.

Step 2. The coordinate time follows from $\sinh = \sqrt{\cosh^{2}-1}$ : $t = \frac{c}{g}\sqrt{(1 + gd/c^{2})^{2} - 1}$ . Why: no separate integration is needed — $\cosh^2 - \sinh^2 = 1$ does the work, the standing dividend of the hyperbolic parametrisation.

Step 3. Take the far limit, where the physics is. For $u \equiv gd/c^{2} \gg 1$ : $\mathrm{arccosh}(1+u) = \ln\!\big(1+u+\sqrt{u^2+2u}\big) \to \ln(2u)$ , so $\boxed{\tau \to \dfrac{c}{g}\ln\!\dfrac{2gd}{c^{2}}}$ — logarithmic in $d$ . And $t \to \dfrac{c}{g}\,u = \dfrac{d}{c}$ — linear in $d$ , and asymptotically equal to the light-travel time. Why: the entire result. Every factor of ten in distance costs the crew a fixed $\frac{c}{g}\ln 10 = 2.23$ years, forever, while Earth pays ten times more each time. The universe is small from the cabin and large from the ground, and both descriptions are exactly right.

Step 4. Sanity-check against causality. $t > d/c$ always, with equality only asymptotically — the ship never beats light, even though its crew age arbitrarily little. Why: the crew's short ageing is not superluminal travel — it is length contraction, seen from the ship. In the ship's momentary frame the distance ahead has shrunk by $\gamma$ , and it is that shortened distance the crew fly. Nothing outruns a photon.

Step 5. Assemble the real trip. A practical voyage accelerates for the first half and decelerates for the second, arriving at rest, so each half covers $D/2$ and the totals are $\tau_{\text{tot}} = 2\,\tau(D/2)$ and $t_{\text{tot}} = 2\,t(D/2)$ , with a peak rapidity $\varphi_{\max} = g\tau(D/2)/c$ reached at the midpoint and a total rapidity expended of $\Delta\varphi = 2\varphi_{\max}$ — because the ship must throw away as much rapidity as it gained. Why: the deceleration is not free. It costs exactly as much rapidity as the acceleration did, and therefore squares the mass ratio — the most commonly dropped factor in back-of-envelope starship arithmetic, worth $38.7$ rather than $6.2$ for Proxima. *Verified in verify_aq-sr-relativistic-rocket-equation-s6.py at 50-digit precision (mpmath): $\mathrm{arccosh}(1+u)$ is evaluated in the numerically stable logarithmic form, never by naive float64 subtraction; $\tau$ , $t$ , $\varphi_{\max}$ and $\Delta\varphi$ are recomputed for every trip in the article and cross-checked against $\cosh\varphi_{\max} = 1 + gD/2c^{2}$ ; $t > D/c$ is asserted for all three; and $g$ and $c$ are perturbed to confirm the printed figures actually move.*

Go deeper · Derivation C · 2 minThe photon rocket, and the number that ends the discussion

Derivation CThe photon rocket, and the number that ends the discussion

The two halves now meet. The kinematics says the crew age only $\Delta\varphi\,c/g$ ; the dynamics says the fuel costs $\exp(c\Delta\varphi/v_{\text{ex}})$ . Substituting one into the other produces the article's owned structural result — a number rather than a formula, because the point is not that the fuel bill is large but that it is large for the best engine physics permits. This card computes that floor for real destinations, then asks what the fuel would have to be made of.

Step 1. Put the two results together. From (4) and $\Delta\varphi = 2\varphi_{\max}$ , the photon-rocket mass ratio for a there-and-stop trip is $M_i/M_f = e^{2\varphi_{\max}}$ with $\cosh\varphi_{\max} = 1 + gD/2c^{2}$ . Why: this is a floor, not an estimate. Any engine with $v_{\text{ex}} < c$ does exponentially worse, and no engine has $v_{\text{ex}} > c$ .

Step 2. Evaluate it for Proxima Centauri, $D = 4.2465$ ly. $gD/2c^{2} = 2.1233/0.9684 = 2.1926$ , so $\cosh\varphi_{\max} = 3.1926$ , $\varphi_{\max} = 1.8285$ , $\Delta\varphi = 3.6570$ , and $M_i/M_f = e^{3.6570} = \mathbf{38.7}$ . Why: thirty-nine kilograms on the pad for one kilogram at Proxima — with a perfect photon drive. A fusion drive at $v_{\text{ex}} = 0.1c$ needs $e^{36.57} = 7.6\times10^{15}$ ; a chemical rocket at $4.4\ \mathrm{km/s}$ needs $e^{249{,}000} \approx 10^{108{,}000}$ , exceeding the atom count of the observable universe by a hundred thousand orders of magnitude.

Step 3. Evaluate it for Andromeda, $D = 2.537$ Mly. Now $gD/2c^{2} = 1.310\times10^{6}$ , so $\varphi_{\max} = 14.78$ , $\Delta\varphi = 29.56$ , and $M_i/M_f = e^{29.56} = \mathbf{6.9\times10^{12}}$ . Why: the crew age $28.6$ years — the trip is comfortable — and the mass ratio is seven trillion. The obstacle to intergalactic travel is not time, and it is not the speed of light. It is fuel, and the wall is exponential.

Step 4. Ask what the fuel must be. A photon rocket needs its propellant converted, with unit efficiency, into a collimated beam moving at exactly $c$ . The only process in nature that converts rest mass to radiation completely is matter–antimatter annihilation — and even that is imperfect, since $p\bar{p}$ annihilation yields mostly charged pions, which decay to neutrinos and muons and are anything but a clean directed beam. Take the ideal case anyway: to deliver a $100$ -tonne payload to Proxima you need $M_i = 38.7 \times 10^{5}\ \mathrm{kg}$ , so the fuel is $3.77\times10^{6}\ \mathrm{kg}$ , of which half must be antimatter — about $1{,}900$ tonnes of it. Why: total world antimatter production to date is measured in nanograms. The gap is thirty orders of magnitude, and it is not an engineering gap.

Step 5. Price it in energy. Annihilating $3.77\times10^{6}\ \mathrm{kg}$ releases $E = mc^{2} = 3.39\times10^{23}\ \mathrm{J}$ . World primary energy consumption is about $6.2\times10^{20}\ \mathrm{J}$ per year ( $\approx 620$ EJ), so the trip costs roughly $\mathbf{550\ years}$ of the entire planet's output — to send $100$ tonnes to the nearest star. Why: the honest summary. Relativity does not forbid interstellar travel; it makes the journey short for the traveller. What forbids it is the exponential, and the exponential is not a technological problem. It is the rocket equation. *Verified in verify_aq-sr-relativistic-rocket-equation-s6.py: every mass ratio, energy and rapidity above is recomputed at 50 digits and greped from this article; the chemical, fusion and photon cases are computed from one common $\Delta\varphi$ ; and the inequality $M_i/M_f \ge e^{\Delta\varphi}$ is verified over a scan of exhaust velocities.*

Limiting cases

Case	Result	Meaning
$v_{\text{ex}} \ll c$ , $\Delta\varphi \ll 1$	$\Delta v \to v_{\text{ex}}\ln(M_i/M_f)$	Tsiolkovsky, recovered exactly. Relativity did not change the equation — it changed the variable
$gd/c^{2} \ll 1$	$\tau \to t \to \sqrt{2d/g}$	the Newtonian answer; crew and Earth clocks agree
$gd/c^{2} \gg 1$	$\tau \to \frac{c}{g}\ln\frac{2gd}{c^{2}}$ , $\;t \to d/c$	log versus linear — the entire promise, and the entire trap
$v_{\text{ex}} = c$	$M_i/M_f = e^{\Delta\varphi}$	the photon rocket: the floor no engine can go below
decelerate to rest	$\Delta\varphi = 2\varphi_{\max}$ , so $M_i/M_f$ is squared	arriving is as expensive as leaving

Common trapThe deceleration doubles the rapidity and therefore SQUARES the mass ratio

A ship that accelerates to $\varphi_{\max}$ and coasts spends $\Delta\varphi = \varphi_{\max}$ . One that also stops at the far end spends $2\varphi_{\max}$ — and since the mass ratio is $e^{\Delta\varphi\,c/v_{\text{ex}}}$ , stopping squares it. For Proxima that is $6.2:1$ versus $\mathbf{38.7:1}$ for a photon rocket; for Andromeda, $2.6\times10^{6}$ versus $\mathbf{6.9\times10^{12}}$ . A round trip — out, stop, back, stop — spends $4\varphi_{\max}$ and needs $M_i/M_f = \mathbf{1{,}500:1}$ to Proxima. Every manoeuvre is a withdrawal and the exponential compounds them. Quoting a one-way, no-braking mass ratio as "the cost of the trip" is the standard way this calculation is made to look survivable.

Worked examples

1 · Proxima Centauri, there and stopped

A ship accelerates at $1\,g$ for half the way to Proxima Centauri ( $D = 4.2465$ ly), then decelerates and arrives at rest. How long does the crew age, how long does Earth wait, and what is the minimum possible mass ratio?

Step 1 — identify. Accelerate–decelerate, so each half covers $D/2 = 2.1233$ ly: use equation (2) on the half and double. The natural unit is $c^{2}/g = 0.9684$ ly.

Step 2 — the dimensionless group. $\dfrac{gD}{2c^{2}} = \dfrac{2.1233}{0.9684} = \mathbf{2.1926}$ , so $\cosh\varphi_{\max} = 1 + 2.1926 = \mathbf{3.1926}$ and $\varphi_{\max} = \mathrm{arccosh}(3.1926) = \mathbf{1.8285}$ . The peak speed is $\beta = \tanh\varphi_{\max} = \mathbf{0.9497}$ , with $\gamma = 3.193$ .

Step 3 — the clocks. With $c/g = 0.9684$ yr:

\tau_{\text{tot}} = 2\,\frac{c}{g}\,\varphi_{\max} = 2(0.9684)(1.8285) = \mathbf{3.54\ yr}, \qquad t_{\text{tot}} = 2\,\frac{c}{g}\sinh\varphi_{\max} = 2(0.9684)(3.0319) = \mathbf{5.87\ yr}.

Earth waits $5.87$ years — comfortably more than the $4.25$ -year light-travel time, as causality demands.

Step 4 — the boundary number. The rapidity spent is $\Delta\varphi = 2\varphi_{\max} = \mathbf{3.657}$ , so the photon-rocket mass ratio is $M_i/M_f = e^{3.657} = \mathbf{38.7}$ . For a $100$ -tonne payload that is $3.77\times10^{6}\ \mathrm{kg}$ of fuel — some $1{,}900$ tonnes of it antimatter — releasing $3.39\times10^{23}\ \mathrm{J}$ : roughly 550 years of world energy consumption. A fusion drive at $v_{\text{ex}} = 0.1c$ needs $e^{36.57} = \mathbf{7.6\times10^{15}}$ ; a chemical rocket at $4.4\ \mathrm{km/s}$ , $\mathbf{10^{108{,}000}}$ . The nearest star, with a perfect engine, is a civilisation-scale energy project. With a real engine it is arithmetic that has stopped meaning anything.

2 · Andromeda in a working lifetime

The same $1\,g$ programme, to the Andromeda galaxy: $D = 2.537$ million light-years.

Step 1 — identify. Same equations, six orders of magnitude more distance. Here $gD/2c^{2} = 1.3099\times10^{6} \gg 1$ , so the asymptotic forms apply.

Step 2 — the rapidity. $\varphi_{\max} = \mathrm{arccosh}\big(1 + 1.3099\times10^{6}\big) = \mathbf{14.779}$ , so $\Delta\varphi = \mathbf{29.557}$ and the peak Lorentz factor is $\gamma_{\max} = \cosh\varphi_{\max} = 1.31\times10^{6}$ .

Step 3 — the clocks. $\tau_{\text{tot}} = 2(0.9684)(14.779) = \mathbf{28.6\ yr}$ . Earth's clock: $t_{\text{tot}} = 2(0.9684)(1.3099\times10^{6}) = \mathbf{2.54\times10^{6}\ yr}$ — the light-travel time, essentially exactly, as it must be.

Step 4 — the boundary number. Twenty-eight point six years of crew time to cross two and a half million light-years: a crew leaving at twenty-five arrives at fifty-four. And the photon-rocket mass ratio is $e^{29.557} = \mathbf{6.9\times10^{12}}$ — seven trillion kilograms of perfectly annihilated matter per kilogram that arrives. The journey is a career; the fuel is a galaxy. Going ten times further adds only $\frac{c}{g}\ln 10 = 2.23$ years to the crew's ageing — and multiplies the fuel by ten. The exponential, not the light-speed limit, is what stops us.

Physics of the portrait

The photon rocket is the only engine that saturates the relativistic bound, and antimatter the only fuel that could conceivably feed it. Both are consequences of equation (3), not engineering opinions.

Origins · 2 minAntimatter and how this was discovered

Antimatter

1928–1995 · Dirac's equation demands it · Anderson finds the positron · Sanger proposes the photon rocket · CERN makes the first atoms of antihydrogen

Antimatter enters physics as an unwanted root. Dirac's 1928 relativistic wave equation for the electron had negative-energy solutions, and rather than discard them he proposed, by 1931, that they described a particle of the same mass and opposite charge; Carl Anderson found it in a cloud chamber in 1932 and called it the positron. What made it interesting to rocketry is a single number: matter and antimatter annihilate with complete conversion of rest mass to energy, a specific energy of $c^{2} = 9\times10^{16}\ \mathrm{J/kg}$ — ten orders of magnitude above any chemical fuel, three above fission. Eugen Sänger spent the 1950s on the consequences, lecturing on the photon rocket at Zurich in 1953 and publishing Zur Mechanik der Photonen-Strahlantriebe in Munich in 1956; his problem, which remains the problem, is that annihilation makes gamma rays and pions radiating in all directions, and a rocket needs a beam. Bussard's 1960 interstellar ramjet tried to sidestep the mass ratio by scooping fuel from the interstellar medium — an idea whose drag defeats it. Antimatter meanwhile stayed laboratory-scale: the first antihydrogen atoms were made at CERN in 1995, nine of them, and world production since is measured in nanograms. The physics is settled and the arithmetic is merciless. A perfect antimatter photon rocket is the best engine the universe permits, and it still needs $38.7$ kilograms on the pad per kilogram delivered to the nearest star.

History

1903

— Tsiolkovsky publishes the rocket equation, delta-v equals exhaust speed times the logarithm of the mass ratio, the founding result of astronautics.

1928

— Dirac's relativistic equation for the electron admits negative-energy solutions, and by 1931 he interprets them as antiparticles.

1932

— Anderson observes the positron in cloud-chamber tracks, confirming antimatter experimentally.

1946

— Ackeret derives the relativistic rocket equation, showing that the Tsiolkovsky structure survives with rapidity in place of velocity (Helv. Phys. Acta 19, 103).

1953

— Sanger lectures at Zurich on the theory of the photon rocket; he publishes Zur Mechanik der Photonen-Strahlantriebe in Munich in 1956.

1960

— Bussard proposes the interstellar ramjet, scooping interstellar hydrogen to escape the mass-ratio wall (Astronautica Acta).

1995

— CERN produces the first atoms of antihydrogen; total world antimatter production to date remains of order nanograms.

Going further

Texts. Rindler, Introduction to Special Relativity, 2nd ed., on hyperbolic motion and the relativistic rocket · Taylor & Wheeler, Spacetime Physics, 2nd ed. · Ackeret, J., "Zur Theorie der Raketen", Helv. Phys. Acta 19, 103–112 (1946) · Sänger, E., Zur Mechanik der Photonen-Strahlantriebe (Oldenbourg, Munich, 1956) · Bussard, R. W., "Galactic Matter and Interstellar Flight", Astronautica Acta 6, 179 (1960).

Provenance. Signature $(+,-,-,-)$ . Every equation and number above is recomputed in verify_aq-sr-relativistic-rocket-equation-s6.py at 50-digit precision (mpmath), which greps this article for each printed value. $\mathrm{arccosh}(1+u)$ is evaluated in its numerically stable logarithmic form throughout, never by float64 subtraction of near-equals; $g$ , $c$ and $D$ are perturbed to confirm every headline figure moves under its inputs, so no printed number is a floating-point artefact. Base status: THIS ARTICLE WAS AUTHORED FROM SCRATCH. The AllQFT base for this slug is displaced — all eight of its equation cards are the Rindler hyperbola of `aq-sr-relativistic-uniform-acceleration`, with zero exhaust velocity, zero mass ratio, zero rocket physics. Reproducing it faithfully would have shipped a duplicate of that article. The one thing salvaged is its scale figures, $c/g \approx 0.97$ yr and $c^2/g \approx 0.97$ ly, which are correct and re-verified here.

Scope. The hyperbolic worldline is recalled, not re-derived — it belongs to `aq-sr-relativistic-uniform-acceleration`; rapidity and its additivity to `aq-sr-rapidity-hyperbolic-boosts`. A companion treatment of the rocket problem in the dynamics chapter is `aq-sr-relativistic-rocket-equation`, linked by slug; no claim is made here about its contents. Owned by this article and nothing else: the traveller-time inversion $\tau(d)$ , the rapidity budget as the bridge from fuel to motion, and the photon-rocket mass ratio for real destinations.

Builds on. Proper Acceleration and the Comoving Frame · Relativistic Uniform Acceleration and the Rindler Hyperbola · Rapidity. Leads to. Rindler Coordinates · The Rindler Horizon · The Equivalence Principle.

If the crew age only 28.6 years crossing Andromeda, are they not travelling faster than light?

No. Earth's clock records $2.54$ million years for the same trip, and the ship's coordinate time always satisfies $t > D/c$ — it never beats a photon that left with it. The two statements are reconciled by asking whose distance. In the ship's momentary rest frame the distance ahead is length-contracted by $\gamma$ , and at the midpoint $\gamma \approx 1.3\times10^{6}$ , so Andromeda is a couple of light-years off as the crew measure it. They cross that gap at nearly $c$ in a couple of years of their own time. Nothing is superluminal in any frame; the crew simply have a shorter journey to make.

Why is the mass ratio exponential when the Newtonian one is also exponential in delta-v?

It is the same exponential — that is the point. Tsiolkovsky already says $M_i/M_f = e^{\Delta v/v_{\text{ex}}}$ ; the relativistic version says $M_i/M_f = e^{c\Delta\varphi/v_{\text{ex}}}$ . What changes is what goes in the exponent. Newtonian $\Delta v$ for a fast trip is a few times $c$ , which is nonsense; relativistic $\Delta\varphi$ is of order $2$ to $30$ for interesting trips. Multiply by $c/v_{\text{ex}}$ — $1$ at best, $68{,}000$ for a chemical rocket — and the exponent does the rest. Relativity did not introduce the exponential; it told us which quantity it acts on, and imposed the floor $v_{\text{ex}} \le c$ that Newton never had.

Is a mass ratio of 38.7 for Proxima really impossible? Chemical rockets manage 20 to 1.

The number is not the problem; the fuel is. A mass ratio of $38.7$ requires $v_{\text{ex}} = c$ — the propellant converted entirely into a collimated beam of light. The only complete matter-to-energy process is annihilation, so the tank must be half antimatter: some $1{,}900$ tonnes of it for a $100$ -tonne payload, against a total world production so far of a few nanograms. And real annihilation does not oblige — proton–antiproton annihilation yields mostly charged pions flying off in all directions, so the effective exhaust is neither collimated nor at $c$ . Drop $v_{\text{ex}}$ to a still-fantastical $0.1c$ and the mass ratio becomes $7.6\times10^{15}$ . The $38.7$ is not a target. It is a proof that no target exists.

Why does stopping at the destination cost so much more than flying past?

Because the mass ratio is exponential in the rapidity and the rapidity is spent, not returned. Accelerating to $\varphi_{\max}$ costs $\varphi_{\max}$ ; shedding it again costs another $\varphi_{\max}$ , because the engine must throw mass the other way and cannot recover what it threw before. So $\Delta\varphi$ doubles and the mass ratio squares: $6.2 \to 38.7$ for Proxima, $2.6\times10^{6} \to 6.9\times10^{12}$ for Andromeda. This is why almost every serious interstellar proposal is a flyby — the science is thin, but the fuel bill is the square root of the alternative.

Does any of this need general relativity, given that the ship is accelerating the whole time?

No, and the belief that it does is the most persistent myth in the subject. Every calculation here is in flat Minkowski spacetime, with the Riemann tensor identically zero. Special relativity handles accelerated objects — and accelerated frames — perfectly well; the machinery for the latter is Rindler coordinates (`aq-sr-rindler-coordinates`), a coordinate change on flat spacetime and nothing more. General relativity is needed for gravity: curvature, and tidal effects no coordinate change can remove. A rocket at $1\,g$ has no curvature anywhere near it.

The distance to Proxima has an uncertainty. Does the answer really pin down to 38.7?

The mass ratio is remarkably insensitive to distance, being exponential in the arccosh of it. Change $D$ by ten per cent and $\varphi_{\max}$ moves by about $0.05$ , so the mass ratio moves by about five per cent. What it is sensitive to is $c/v_{\text{ex}}$ , which sits in the exponent undivided: halve the exhaust speed and you square the mass ratio. The honest error bars belong on the engine, not on the star chart.