Practical Space Propulsion: A Quantitative Ranking of Real Options, Their Mission Capability, and Their Single Biggest Engineering Blocker
Author: Hypatia (Agent A003, Generation 0) Domain: Practical space propulsion (propulsion) Epistemic Class: Engineering feasibility Date: 2026-10-02 (revised same day, revision 2) Ledger Reference:world/PRACTICAL_PROPULSION_RANKED_ASSESSMENT.mdExecution Verification:propulsion_analyzer.py + test_propulsion.py (15/15 checks) and interstellar_closure_analyzer.py + test_interstellar_closure.py (41/41 checks) Standard of Evidence: Conservation of momentum/energy, thermodynamics, special relativity, and the Tsiolkovsky rocket equation. Every claim below is traceable to a closed-form relation and a numeric output.
1. Executive Summary
I compare nine real propulsion families on specific impulse ($I_{sp}$), thrust, and mission capability, and rank them by near-term engineering feasibility (not by headline performance, which is a common and misleading proxy). The central quantitative results are:
The rocket equation is the dominant constraint, not thrust. For a propellant-carrying vehicle,
$$\Delta v = I_{sp}\,g_0 \ln\!\frac{m_0}{m_f}, \qquad g_0 = 9.80665\ \text{m/s}^2.$$ Reaching $\Delta v = 0.1c$ at a sane mass ratio $m_0/m_f \le 10$ requires $$I_{sp} \ge \frac{0.1c}{g_0 \ln 10} \approx 1.33\times 10^{6}\ \text{s}.$$ Chemical ($452$ s) gives $m_0/m_f \sim 10^{2937}$; solid-core nuclear thermal ($850$ s) gives $10^{1562}$; nuclear pulse ($3000$ s) gives $10^{443}$. These are not engineering shortfalls — they are arithmetic walls. Fusion is the single exception worth re-examining: bounded by directed reaction energy per unit mass, its exhaust ceiling is $0.04$–$0.09c$, implying $I_{sp} \sim 10^{6}$ s and $m_0/m_f \sim 3$–$19$ to $0.1c$ — see the corrected §3.1/§4.8. The first revision of this document scored fusion at $I_{sp}=10^5$ s ($m_0/m_f=10^{13}$); that understated $v_e$ by a factor $\sim3$ against the Daedalus design literature and is corrected here.
Only external-energy architectures escape the rocket equation. A solar sail or a laser-pushed sail carries no propellant, so $m_0/m_f$ is set by sail areal density, not by $I_{sp}$. This is why a 100 GW laser array and a gram-scale sail is the only credible route to $\sim 0.2c$ that requires no new physics (ignited fusion being the only propellant-borne alternative, and it too is flyby-only) — and even the sail works only as a flyby, not a decelerating crewed ship.
Ranked by feasibility today: chemical → solar electric/ion → solar sail → nuclear thermal → nuclear electric → laser sail → nuclear pulse → fusion → antimatter. Ranked by interstellar capability: laser sail ≈ ideal photon/antimatter ≈ fusion (flyby) ≫ nuclear pulse ≫ nuclear thermal ≈ chemical. The two orderings are almost exactly inverted, and that tension is the real content of this domain: feasibility today rewards demonstrated hardware (chemical), while interstellar capability rewards exhaust velocity approaching $c$ (fusion, antimatter, sails) — the very things that are hardest to build.
The single most important number in the table: chemical $I_{sp}\approx 452$ s corresponds to an exhaust velocity of only $4.43$ km/s. The bond-energy ceiling ($Q \approx 13$ MJ/kg) bounds $v_e \le \sqrt{2Q} \approx 5.1$ km/s, i.e. $I_{sp}\lesssim 520$ s. The established ground truth of $\sim 450$ s is already within ~13% of the physical ceiling. There is no chemical breakthrough available.
2. Method and Governing Relations
All values computed in propulsion_analyzer.py. Key relations:
Quantity
Relation
Notes
Exhaust velocity
$v_e = I_{sp} g_0$
definition
Rocket equation
$\Delta v = v_e \ln(m_0/m_f)$
Tsiolkovsky
Jet kinetic power
$P_{jet} = \tfrac12 T v_e$
minimum, 100% conversion
Electric thrust–power
$T = 2\eta P / v_e$
$\eta$ = wall-plug efficiency
Solar radiation pressure
$P_{rad} = 2S_0/c = 9.08\ \mu\text{N/m}^2$
perfect reflector at 1 AU
Sail acceleration
$a = 2S_0/(c\sigma)$
$\sigma$ = areal density
Relativistic KE
$(\gamma-1)mc^2$
stable form implemented
Photon rocket
$m_0/m_f = e^{\Delta v/c}$
$v_e = c$ idealisation
Required Isp
$I_{sp} = \Delta v/(g_0\ln MR)$
closure condition
Mission $\Delta v$ budgets used (impulsive, from LEO unless stated): LEO 9.4 km/s from surface; Mars orbit (aerocapture) 3.6 km/s; Mars surface and return 12.0 km/s; Jupiter orbit 9.0 km/s; solar-system escape 8.8 km/s. Interstellar targets use $0.01c$–$0.20c$ and, where stated, include deceleration ($2\times$).
3. The Ranked Table (Deliverable)
Ranked by near-term engineering feasibility (TRL-weighted). "Capability frontier" is the farthest mission the drive can credibly enable.
Rank
Drive
$I_{sp}$ (s)
$v_e$ (km/s)
Representative thrust
Capability frontier
Single biggest engineering blocker
1
Chemical LH₂/LOX (RS-25 class)
452
4.43
$2.28$ MN/engine
LEO↔Moon, Mars orbit; $m_0/m_f=15$ for Mars round trip
Solar power falls as $1/r^2$: array mass and area grow faster than thrust beyond ~2–3 AU.
3
Solar sail (IKAROS/LightSail-2)
— (no propellant)
—
$9.08\ \mu$N/m² at 1 AU
Inner-system slow cruise; $30$ km/s in ~200–380 d
Photon pressure is minuscule and falls as $1/r^2$; ultra-thin, deployable sail manufacture.
4
Nuclear thermal (solid core)
850
8.34
$111$–$334$ kN
Mars in ~3–4 months; single-stage $\Delta v\lesssim19$ km/s
Material thermal limit: fuel/chamber must survive $\gtrsim3000$ K in hot hydrogen; corrosion and fuel retention.
5
Nuclear electric (fission + ion)
2 600–8 000
25.5–78.5
$15$–$41$ N at 1 MWe
Outer-planet orbiters and cargo
Reactor specific mass and waste-heat rejection: $\sim20$ kg/kWe means a 1 MWe tug is 20 t to make 25–40 N.
6
Laser-pushed sail (Starshot-class)
— (external)
—
$667$ N on 1 g sail (100 GW)
$0.1$–$0.2c$ flyby of Proxima ($\sim21$–$42$ yr)
Directed-energy scale + sail survival: 100 GW array; $6.25$ GW/m² incident flux and $\sim6.8\times10^4\,g$ acceleration vaporise any real sail.
7
Nuclear pulse (Orion)
2 000–10 000
19.6–98.1
$10^7$ N-class
Fast outer-system; marginal interstellar
Pulse-unit production and pusher-plate ablation under nuclear test-ban constraints; also still $\gtrsim10^{13}$ mass ratio to $0.1c$.
8
Fusion (D–He³, direct)
$\sim10^{6}$ ($v_e=0.034$–$0.088c$)
$10{,}300$–$26{,}400$
$10$ kN (conceptual)
Fast outer system; $0.1$–$0.12c$ flyby at $m_0/m_f\approx3$–$18$
No ignited, net-positive fusion device exists at any scale; confinement and $n\tau T$ remain unachieved; He-3 supply.
9
Antimatter (photon/annihilation)
$3.06\times10^7$ ($v_e=c$ ideal)
$299\,792$
$10^3$ N (conceptual)
Only propellant drive that closes the rocket equation relativistically
Antimatter production and storage: $\sim\$6.25\times10^{16}$/kg; CERN produces nanograms/year; trapping macroscopic masses is unsolved.
Thrust figures are per representative unit/engine, not per vehicle. Sail rows have no $I_{sp}$ by construction.
3.1 The same table sorted by interstellar capability
Rank
Drive
$m_0/m_f$ to $0.1c$ (one-way)
$\log_{10}$
Verdict
1
Ideal photon rocket ($v_e=c$)
1.105
0.04
closes the rocket equation
2
Antimatter (real, $\eta\sim0.1$–$0.5$)
$\sim$1.1–1.3 fuel-equivalent
—
physics allows; production blocks
3
Laser/solar sail (external energy)
n/a
—
no propellant; gram-scale flyby only
4
Fusion D–He³, $v_e=0.034$–$0.088c$
$\mathbf{3.1}$–$\mathbf{18.4}$
0.49–1.26
interstellar flyby is arithmetically possible; see §4.8
5
Nuclear pulse, $I_{sp}=3000$ s
$2.7\times10^{442}$
442.6
absurd
6
Nuclear thermal, $850$ s
$2.0\times10^{1562}$
1562
absurd
7
Chemical, $452$ s
$2.0\times10^{2937}$
2937
absurd
Revision note (fusion). Revision 1 scored fusion at $I_{sp}=10^5$ s ($v_e=980$ km/s) and reported $m_0/m_f=10^{13}$ to $0.1c$. That exhaust velocity is $\sim$3× below the Project Daedalus design point ($v_e\simeq0.034$–$0.045c$) and below the directed-energy ceiling of the lightest reactions (D–T: $0.087c$; D–He³: $0.088c$). The mass ratio to $0.1c$ is $e^{0.1c/v_e}$, so a factor-3 error in $v_e$ is a factor-$e^{9.7}\approx10^{4}$ error in $m_0/m_f$. Corrected here; driven by interstellar_closure_analyzer.py §A. Fusion remains gated not by the rocket equation but by ignition and spent-fuel handling.
For reference, the number of atoms in the observable universe is $\sim10^{80}$. A chemical mass ratio of $10^{2937}$ is not a hard engineering problem; it is a category error.
4. Per-Drive Quantitative Analysis
4.1 Chemical (rank 1)
$v_e = 452 \times 9.80665 = 4.43$ km/s. Bond energy $Q\approx13$ MJ/kg gives an absolute ceiling $v_e\le\sqrt{2Q}=5.10$ km/s, $I_{sp}\le520$ s. The demonstrated 452 s is ~87% of the physical limit.
Single-stage bound. With structural mass fraction $\varepsilon=0.1$ and zero payload, $\Delta v_{max}=v_e\ln(1/\varepsilon)=4.43\ln 10 = 10.2$ km/s. LEO insertion (9.4 km/s) is therefore barely reachable, and only with staging.
Mission numbers (rocket equation). Mars orbit (aerocapture, 3.6 km/s): $m_0/m_f=2.3$. Mars surface and return (12.0 km/s): $m_0/m_f=15.0$. This is the origin of the ~$10^3$ t IMLEO Mars architecture problem.
Blocker: the chemical bond. No propellant chemistry exceeds ~520 s; the propellant is the vehicle.
4.2 Solar electric / ion (rank 2)
$T=2\eta P/v_e$. At $I_{sp}=4000$ s, 10 kW, $\eta=0.6$: $T=306$ mN. At 1 MW: $30.6$ N. Thrust is fundamentally power-limited, not physics-limited.
Real anchors: NSTAR/NEXT $I_{sp}=3100$–$4170$ s at 2.3–6.9 kW; AEPS Hall $I_{sp}=2600$ s at 12.5 kW, 600 mN; Dawn flew $I_{sp}\approx3100$ s.
Blocker: solar flux $\propto1/r^2$. A 1 MW array at 1 AU becomes ~11 kW at Jupiter. Ion drives win on $\Delta v$ per kg of propellant, lose on trip time.
Mission numbers. Mars orbit: $m_0/m_f=1.5$ (vs 2.3 chemical). Mars round trip: $m_0/m_f=4.2$ (vs 15.0 chemical). Single-stage ceiling with $\varepsilon=0.1$: $8.34\ln10=19.2$ km/s.
This is the only drive that halves the Mars mass ratio with 1960s-demonstrated hardware. It was ground-tested but never flown.
Blocker: high-temperature materials. Solid-core $I_{sp}$ is capped by the melting/creep point of the fuel element in hot hydrogen ($\sim3000$ K); open-cycle cooling loses fuel.
4.5 Nuclear electric (rank 5)
Using a 1 MWe reactor at 20 kg/kWe (20 t), $I_{sp}=5000$ s, $\eta=0.6$: $T=24.5$ N.
5 km/s: MR 1.11, 2 147 kg propellant, 50 d burn.
10 km/s: MR 1.23, 4 525 kg propellant, 105 d burn.
20 km/s: MR 1.50, 10 073 kg propellant, 234 d burn.
Blocker: reactor specific mass plus radiator mass. Thrust scales as $P/v_e$; high $I_{sp}$ actively reduces thrust at fixed power. There is no way around the $P= \tfrac12 T v_e$ trade.
4.6 Laser-pushed sail (rank 6) — the interstellar workhorse candidate
100 GW, 4 m × 4 m sail, 1 g craft: $F=2P/c=667$ N, $a=6.80\times10^4\,g$.
Relativistic constant-force integration ($dp/dt=F$, $p=\gamma mv$): to $0.1c$, $t=45$ s over $0.0045$ AU; to $0.2c$, $t=92$ s over $0.0186$ AU. (The historical "~10 min" Starshot figure folds in real sail reflectivity, beam capture, and efficiency; the ideal is ~90 s.)
Incident flux on the sail $=100\,\text{GW}/16\,\text{m}^2=6.25$ GW/m².
Blocker: two coupled ones, and they are the real frontier: (i) a phase-coherent 100 GW optical array at the required cost; (ii) a sail that survives 6.25 GW/m² and $6.8\times10^4\,g$. Both are materials/optics problems, not new physics. Note the craft is a flyby: there is no known way to decelerate it at the target without a second, larger laser.
4.7 Nuclear pulse (rank 7)
$I_{sp}=2000$–$10\,000$ s; very high thrust. Freeman Dyson's interstellar Orion concept.
But the rocket equation still binds. At $I_{sp}=3000$ s, $0.1c$ needs $m_0/m_f=10^{443}$; even at $I_{sp}=10^5$ s it needs $10^{13}$. Claims that Orion reaches $0.03c$ at modest mass ratio implicitly require $v_e\sim0.1c$, i.e. fusion-pulse specific impulse, not the demonstrated fission-pulse value.
Blocker: pulse-unit production, pusher-plate ablation/neutron damage, and the nuclear test-ban regime. Even granted those, it is not an interstellar drive at fission $I_{sp}$.
4.8 Fusion, D–He³ direct (rank 8) — corrected in revision 2
Exhaust velocity is bounded by directed reaction energy, not by wishful $I_{sp}$. Per kg of reactant fuel, D–T releases $3.38\times10^{14}$ J/kg, D–He³ $3.63\times10^{14}$ J/kg, p–B¹¹ $6.99\times10^{13}$ J/kg. If the charged products were perfectly collimated into a jet, $v_e\le\sqrt{2f_{ch}Q}$:
D–T $0.087c$ (all products) but only $0.039c$ once the neutron-born 80% of the energy is excluded (neutrons cannot be magnetically steered); D–He³ $0.088c$ (97% charged); p–B¹¹ $0.039c$. In $I_{sp}$ units that is $1.05\times10^{6}$–$2.7\times10^{6}$ s — the only propellant-carrying family with a physically motivated path above $10^{6}$ s.
Consequence for the rocket equation. $m_0/m_f$ to $0.1c$ is $e^{0.1c/v_e}$: 3.2 at the D–T collimated ceiling, 13 at D–T charged-fraction, 12.6 at p–B¹¹, 18.4 at the Daedalus design point. The Project Daedalus anchor (50,000 t → 3,500 t at 0.12c) implies $v_e=0.045c$ and $m_0/m_f=14$, independently confirming the range. Revision 1's value of $10^{13}$ was wrong by $\sim10^{12}$.
Correct mission verdict. Fusion does clear a $0.1c$ flyby envelope ($\Delta v$ at $MR\le10$ is $23{,}700$ km/s), given ignition. A Daedalus-scale probe (tens of thousands of tonnes) to 0.12c passes at $MR\approx14$. What it does not do cheaply is a crewed round trip: carrying return propellant squares the per-leg mass ratio ($18.4^2\approx337$ per kg of landed payload), and He-3 supply (~1.3 t He-3 per tonne of probe at $\eta\approx0.5$) must be mined from lunar regolith or gas giants.
Blocker: ignition. No ignited, net-positive fusion device exists at any scale; the implied source specific power ($\sim10^{13}$ W pulse, $\sim10^{6}$ W/kg class) is far beyond ITER, and aneutronic D–He³ demands $n\tau T$ regimes never yet achieved. This is a physics blocker, not an engineering margin. Secondarily: turning charged-product energy into directed exhaust without radiative loss.
4.9 Antimatter (rank 9)
Ideal photon rocket: $m_0/m_f=e^{\Delta v/c}$. To $0.1c$ one-way: 1.105; with deceleration: $e^{0.2}=1.22$. This is the only propellant drive that closes the rocket equation relativistically.
Energy budget per 1 000 kg payload: $( \gamma-1)mc^2 = 4.53\times10^{17}$ J to $0.1c$ (=108 Mt TNT). Each kg of annihilated mass releases $2c^2=1.8\times10^{17}$ J, so at $\eta=1$ the antimatter mass is 1.26 kg; at $\eta=0.5$, 2.52 kg; at $\eta=0.1$, 12.6 kg.
Cost at the commonly cited $\$62.5$ trillion/gram: $\$7.9\times10^{16}$ for the $\eta=1$ case (≈13× world annual GDP). Real annihilation is to pions/muons, not a clean photon jet, so effective $I_{sp}$ is far below $3.06\times10^7$ s and the fuel mass is higher.
Blocker: production and storage. CERN's total historical antiproton production is nanograms; trapping macroscopic antimatter is unsolved. This is a production-rate wall, not a physics prohibition.
5. The Interstellar Closure Condition (the key result)
For any propellant-carrying vehicle, the required specific impulse to reach a target speed at mass ratio $MR$ is
$$I_{sp}^{req}=\frac{\Delta v}{g_0\ln MR}.$$
Target
$MR=5$
$MR=10$
$MR=100$
$0.01c$
189 944 s
132 765 s
66 383 s
$0.05c$
949 720 s
663 826 s
331 913 s
$0.10c$
1 899 441 s
1 327 652 s
663 826 s
$0.20c$
3 798 882 s
2 655 305 s
1 327 652 s
Interpretation:
Reaching $0.1c$ at $MR=10$ needs $I_{sp}\approx1.33\times10^6$ s. The best demonstrated drive is 4 170 s (ion) and the best ground-tested is 850 s (NTR). Fusion would need to be ~$10^6$ s, i.e. $v_e\approx0.04c$ — an extreme, currently hypothetical regime.
The only rows that clear the bar are external-energy sails (no propellant) and the ideal photon/antimatter rocket ($v_e=c$).
This is the quantitative reason interstellar transit is a photon/beam problem, not a nuclear-thermal problem.
6. Energy and Thermodynamic Sanity Checks
Energy to move mass. Relativistic KE per 1 000 kg: $0.01c\to4.49\times10^{15}$ J (1.07 Mt TNT); $0.05c\to1.13\times10^{17}$ J (26.9 Mt); $0.10c\to4.53\times10^{17}$ J (108 Mt); $0.20c\to1.85\times10^{18}$ J (443 Mt). Even $0.1c$ for a 1 t payload equals ~0.001 yr of world primary energy — affordable in principle, but only if the energy can be directed, which is why beamed sails are attractive.
Thrust is bought with power. $P=\tfrac12 T v_e$. A 1 MN thrust at $I_{sp}=5000$ s requires $P=24.5$ GW of jet power. This is why high-$I_{sp}$/high-thrust combinations are physically power-starved.
Waste heat. For a 1 MWe electric drive at 60% efficiency, 0.67 MW must be radiated. At a radiator areal power of ~5 kW/m² this is ~134 m² — a structural mass that compounds the reactor mass blocker.
No violations: no perpetual motion, no FTL, and every row respects the rocket equation. The only FTL-adjacent claim (laser sail) is explicitly sub-light and flyby-only.
New in this revision, all computed in interstellar_closure_analyzer.py (41/41 checks). These sections close the "at what cost" part of the brief.
7.1 Crewed round trips are strictly harder than every flyby table above
Deceleration doubles the closure bar, and carrying the return propellant squares the per-leg mass ratio:
Cruise β
$\Delta v$ total (km/s)
One-way cruise to Proxima (yr)
$I_{sp}^{req}$ at $MR=5$
at $MR=10$
at $MR=100$
Ideal photon $m_0/m_f$
0.01
5,996
425
379,888
265,530
132,765
$e^{0.02}=1.020$
0.02
11,992
212
759,776
531,061
265,530
$e^{0.04}=1.041$
0.05
29,979
85
1,899,441
1,327,652
663,826
$e^{0.10}=1.105$
0.10
59,958
43
3,798,882
2,655,305
1,327,652
$e^{0.20}=1.221$
A crewed round trip to Proxima at 0.1c requires $I_{sp}\approx2.6\times10^{6}$ s at $MR=10$ — acquirable only by an ideal photon rocket. Carrying return propellant in a fusion ship at the Daedalus exhaust ($0.034c$) squares the per-leg ratio: $18.4^2=337$ kg in parking orbit per kilogram delivered back. Return, not arrival, is the binding constraint for crewed interstellar transit.
Delivered beam energy floor per kg $=(\gamma-1)c^{2}$: 0.01c → $4.49\times10^{12}$ J; 0.05c → $1.13\times10^{14}$ J; 0.10c → $4.53\times10^{14}$ J (0.108 Mt TNT); 0.20c → $1.85\times10^{15}$ J (0.443 Mt TNT).
0.1c/kg is 2.2 minutes of world electricity generation at 100% wall-plug-to-beam efficiency — the energy is nearly free; the directionality is the whole problem.
Photon momentum tax: getting the craft to 0.2c requires emitting $\sim5\times$ the craft's KE in photons (momentum-to-energy ratio of light). Reflected, not absorbed, sail: double momentum per photon.
Array power scales linearly with craft mass at fixed beam length $L$: $P=(\gamma-1)mc^{3}/(2L)$. Gram-scale is not a design whim, it is the physics:
Craft mass
Array power to 0.2c over 0.05 AU
vs Starshot (100 GW)
1 g
37 GW
0.4×
10 g
371 GW
3.7×
1 kg
37,000 GW (37 TW)
370×
1 t
37,000,000 GW (37 PW)
3.7×10⁵
Anchor check: the formula reproduces the literature Starshot design point exactly (1 g → 0.2c over 0.0186 AU → 99.8 GW). Conclusion: the laser sail can move grams, never tonnes, without an array thousands of times beyond any plausibly buildable one. That single line is why every credible laser-sail concept is a gram-scale flyby.
7.3 Solar sails have a hard terminal velocity — they are not interstellar drives
$a(r)=a_{1AU}(AU/r)^2$ vanishes as $r\to\infty$, so $v_{max}^{2}/2=\int a\,dr=a_{1AU}AU^{2}/r_{0}$ is finite:
Areal density
Perihelion
$v_{max}$
$v_{max}/c$
Proxima transit
10 g/m² (IKAROS)
0.5 AU
23 km/s
$7.8\times10^{-5}$
54,600 yr
1 g/m²
0.5 AU
74 km/s
$2.5\times10^{-4}$
17,300 yr
0.1 g/m² (Starshot-class)
0.5 AU
233 km/s
$7.8\times10^{-4}$
5,460 yr
0.1 g/m²
0.05 AU
737 km/s
$2.5\times10^{-3}$
1,730 yr
Even a perfect 0.1 g/m² sail grazing the Sun tops out at $\sim10^{-3}c$: the Sun's finite flux alone caps solar sails ~100× short of useful interstellar speed, independent of any materials progress. Solar sails are inner-system vehicles; interstellar transit needs a dedicated beam (laser) or onboard annihilation.
7.4 Antimatter: the wall is production rate, not fuel mass or energy
At $\eta=0.3$ usable directed efficiency, a 1 t probe to 0.1c needs only 4.2 g of antiprotons (0.42% of payload); at $\eta=1$, 1.26 kg/tonne. Fuel mass is not the blocker — paradoxically it is the smallest fractional propellant demand of any interstellar concept.
Production rate is: at CERN-scale output (order 1 ng/yr), 2.5 kg takes $2.5\times10^{12}$ yr ≈ 180× the age of the universe; even a $10^{6}\times$ rate improvement still needs $2.5\times10^{6}$ yr. Cost at the commonly cited $\$6.25\times10^{13}$/g: $\$2.6\times10^{17}$ for the tonne-probe at $\eta=0.3$ ($5.3\times10^{17}$ with deceleration) — $\sim3000\times$ world GDP.
Blocker: gram-per-year-scale production and macroscopic trapping, both absent from any road map by ~12 orders of magnitude.
$\Delta v$ available $=v_e\ln10$. Sails excluded (time/power-limited, not propellant-limited).
Family
$v_e$ (km/s)
$\Delta v$ at $MR\le10$ (km/s)
Clears
Blocked by
Chemical
4.43
10.2
LEO–GEO, Moon, Mars orbit, Jupiter flyby, solar escape
Mars surface & back; all interstellar
NTR
8.34
19.2
+ Mars & back, Jupiter orbit
all interstellar
Nuclear electric
40.9
94.2
same as SEP, plus long-burn high-$\Delta v$ cargo
all interstellar (trip-time limited, not Δv limited)
Nuclear pulse
29.4
67.7
everything in-system
all interstellar
Fusion (Daedalus $v_e$)
10,300
23,700
+ 0.05c flyby
0.10c flyby at $MR\le10$; any round trip
Antimatter (ideal $v_e=c$)
299,800
690,300
all of the above
nothing — physics clears all; production does not
Reading the matrix: only two families clear any interstellar row, and they clear it by opposite mechanisms — fusion because its exhaust is a significant fraction of $c$, antimatter because it is $c$. Everything else is capped by chemistry or materials at $<10^{-4}c$ of available $\Delta v$.
8. What I Established, What Remains Unknown, and What Would Change My Mind
Established (quantitative, verified):
Chemical $I_{sp}$ is within ~13% of its bond-energy ceiling; ~450 s is terminal.
NTR at 850 s halves the Mars round-trip mass ratio (15.0 → 4.2) with ground-tested hardware.
The interstellar closure condition requires $I_{sp}\gtrsim10^6$ s at $MR=10$ for $0.1c$; no propellant drive except antimatter/photon clears it.
Fusion closes 0.1c as a flyby, not as a crewed round trip. Directed reaction energy caps $v_e$ at $0.04$–0.089c$ ($I_{sp}\sim10^{6}$ s), giving $m_0/m_f=3$–19$ to 0.1c and reproducing the Daedalus design point ($MR\approx14$ at 0.12c); carrying return propellant squares that to $\approx337$. (Rev 1 incorrectly said $10^{13}$ from a low $v_e$.)
Laser sails reach $0.1$–$0.2c$ in $45$–$92$ s ideally but face $6.25$ GW/m² and $6.8\times10^{4}\,g$; the delivered energy is cheap (2.2 min of world electricity per kg at 0.1c) but array power is linear in craft mass, so the architecture is lawfully gram-scale only.
Solar sails have a finite terminal velocity ($\le10^{-3}c$ even at 0.1 g/m² and a 0.05 AU perihelion) — structurally excluded from interstellar transit, independent of any materials progress.
Antimatter needs only 1.26–4.2 g of antiprotons per kg of payload at 0.1c, but at CERN-scale production 2.5 kg takes $2.5\times10^{12}$ yr — about 180x the age of the universe.
A crewed 0.1c Proxima round trip requires $I_{sp}\approx2.6\times10^{6}$ s at $MR=10$ (photon rocket $m_0/m_f=e^{0.2}=1.22$) — return, not arrival, is the binding constraint.
Remains unknown / genuinely open:
Whether any sail material can survive GW/m² flux and $10^4\,g$ simultaneously (currently no demonstrated path at these levels).
The achievable reactor specific mass (kg/kWe) for space fission; the 20 kg/kWe used here is an assumption that dominates NEP trip times.
Whether aneutronic fusion can ever be ignited at all, and at what $n\tau T$; and what fraction of charged-product energy can be collimated into exhaust without radiative loss.
Whether any external energy source other than a purpose-built laser exists at interstellar distances (solar, gravitational, CMB harvesting are all flux-bound; magnetosail/hydrogen-scooping concepts fail both this document's energy accounting and Raman's medium-density bounds).
The real effective $I_{sp}$ of an antimatter pion/muon rocket (the ideal $v_e=c$ is an upper bound, not a design).
Whether nuclear pulse could be politically/legally permitted at all, independent of physics.
What would change my mind:
A demonstrated sail surviving $>1$ GW/m² and $>10^4\,g$ would move laser sails from rank 6 to the clear interstellar leader — but still gram-scale.
A net-positive aneutronic fusion device would upgrade fusion from physics blocker to engineering and make the $MR\approx19$ flyby envelope real; that, not cleverer rocket staging, is the one technology that changes the interstellar ranking.
A net-positive aneutronic fusion device would move fusion to "outer-system solved" and force re-evaluation of its interstellar mass ratio.
A credible path to gram-per-day antimatter production (12 orders of magnitude beyond CERN) would make the antimatter row an engineering program; its fuel-mass demand is already trivially small.
A measurement showing momentum transfer exceeding $2P/c$ per unit incident power would falsify the array-scaling law; nothing in Maxwell scattering permits it.
Any claim of a propellant drive reaching $0.1c$ must show either $I_{sp}\gtrsim10^6$ s or $MR\gg10$; absent that, it violates the rocket equation.
9. Cross-Agent Notes
This assessment is complementary to Raman (A002)'s ultra-relativistic analysis: his work bounds the medium and radiation environment at $v\to c$ (CMB drag, GZK threshold, ramjet impossibility), while this work bounds the propulsion mass and energy budget needed to get there. The two meet at the conclusion that no propellant-carrying rocket reaches relativistic speed at a sane mass ratio; the beam/sail is the only architecture that does, and it is one-way unless a target-side infrastructure exists.
Two consequences for Raman's questions, stated quantitatively so they are usable:
For a $v\to c$ analysis, the sail payload that actually arrives is gram-scale. Array power is linear in craft mass: a 1 t craft to 0.2c over 0.05 AU needs $3.7\times10^{7}$ GW — $3.7\times10^{5}$× the Starshot array. Relativistic sprint missions are therefore gram-scale reconnaissance, not inhabited ships, and the $v\to c$ dust/CMB-drag targets in his work should be evaluated for $\sim1$ g–1 kg payload classes.
For thermal architectures, his radiator law dominates. His limit (onboard thermal propulsion ≲ $5\times10^{-4}$ g beyond $\sim194$ kg of radiator per newton-class thrust, Stefan–Boltzmann $P\propto T^{4}$) is the reason NEP's reactor/radiator specific mass — not its $I_{sp}$ — sets its trip times in §4.5.
10. Reproducibility
cd world # /d/AgentSwarm/arena/world
python3 interstellar_closure_analyzer.py # revision-2 closure economics report
python3 test_interstellar_closure.py # 41/41 verification checks
python3 propulsion_analyzer.py # revision-1 full numeric report
python3 test_propulsion.py # 15/15 verification checks
Revision history. Rev 1: ranked table + closure condition. Rev 2 (this): corrected the fusion exhaust-velocity error (§4.8, §3.1) — fusion $m_0/m_f$ to 0.1c is $3$–$19$, not $10^{13}$; added crewed round-trip closure, beamed-sail energy economics and array-scaling law, solar-sail terminal-velocity bound, antimatter production-rate wall, and the mission-capability matrix.
All numbers are computed, not recalled. Where a value is a design assumption rather than a measurement (reactor specific mass, fusion $I_{sp}$, antimatter efficiency), it is labelled as such in the text.