Chapter 40: Capstone — Your Complete Space Mission

"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away." — Antoine de Saint-Exupéry, Wind, Sand and Stars

Overview

Thirty-nine chapters ago, we quoted a single number — about $9.4\ \text{km/s}$ of velocity change to reach low Earth orbit — and called it the reason spaceflight is hard. Everything since has been an answer to that number and to the exponential equation that governs it. You have derived the rocket equation, budgeted delta-v across the solar system, sized engines and nozzles, chosen orbits, designed transfers, pointed spacecraft, radiated away heat, closed links across billions of kilometers, and learned why rockets fail and how reuse is rewriting the economics of all of it. Along the way — if you took the Design Your Mission track — you have been quietly assembling something: a real design, for a real mission, one chapter at a time.

This chapter is where it all comes together. There is no new physics here, and that is the point. The capstone does not teach you a new tool; it hands you every tool you already own and asks you to build something with them. You will complete your Mission Design Review — the MDR document that a real program must pass before anyone spends money on hardware — and then you will defend it, because a design that cannot survive a hard question in a review room will not survive contact with space.

To show you how, we will walk four complete worked missions, one for each track: a communications satellite to geostationary orbit, a cargo lander to the Moon, a science orbiter at Mars, and a probe that rendezvouses with a near-Earth asteroid. Each is a mini-MDR — requirements, orbit, delta-v budget, sized vehicle, launch vehicle, systems, timeline, risk — carried end to end with real numbers. Read the one that is yours in full; skim the others to see how the same machine, pointed at four very different objectives, produces four very different vehicles. That universality — one process, one astrotools package, four missions — is the deepest lesson of the book.

In this chapter, you will learn to:

• Inventory everything your MDR must contain, and see how each chapter fed one row of it.
• Size and defend a GEO comsat, a lunar lander, a Mars orbiter, and an asteroid probe — completely.
• Read each vehicle's mass ratio as a direct reading of the tyranny of the rocket equation.
• Present a design the way a review board wants to hear it: driving requirement, margins, and risks first.

Learning Paths

🚀 Space Enthusiast: Read 40.1 for the shape of a complete mission, then read the one track that most excites you (40.2–40.5) as a story with numbers. You do not need to reproduce the arithmetic to feel the punchline of each: the mass ratio tells you how hard the mission is. Finish with 40.6 — it is the most human section in the book.

📐 Engineering Student: Do the whole chapter, and reproduce every number in your own track by hand and in astrotools before reading ours. This is your comprehensive final: if you can size any one of these four missions from a blank page, you have earned the book. Then attempt a second track cold.

🎮 KSP Player: You have flown versions of all four of these. Now put the numbers under them. Build each track's vehicle in the game to its sized wet mass and see whether your delta-v readout matches the budgets here — the game is the most honest review board you will ever face.

🛰️ Industry Prep: 40.1 is the table of contents of a real MDR package; 40.6 is the review itself. Learn to present in the order 40.6 gives — driving requirement, margins, risks — because that is the order a program manager listens in. Your own track's mini-MDR is a writing sample worth keeping.

40.1 The Mission Design Review document

A Mission Design Review is the moment a mission stops being an idea and becomes a design that someone will either fund or reject. In the lifecycle of Chapter 29, the MDR (also called the Mission or System Definition Review) closes the concept phase: it answers one question — does this mission close? — by laying every piece of the design on one table and checking that the pieces are consistent, sized, margined, and buildable. It is not the end of design; it is the gate that authorizes the expensive detailed design to begin. But for you, the reader who has been building toward it since Chapter 1, it is the finish line, because assembling it proves you can design a space mission.

Definition (Mission Design Review). The MDR is the first formal review gate of a space program, at which the requirements, orbit, delta-v budget, spacecraft architecture, launch vehicle, operations concept, and risk assessment are presented together and judged for internal consistency and closure — that is, for whether a vehicle sized to the budget actually fits its launcher, its power, its mass, and its money. It introduces no new engineering; it is the disciplined act of checking that all the engineering agrees with itself.

Here is the full inventory of an MDR, with the chapter that gave you each row. If you kept a Design Your Mission notebook, you have already written most of this; this table is the order to assemble it in.

Notice the spine that runs down the middle of this table — objective → orbit → delta-v budget → propulsion → mass → launch vehicle. That is the load-bearing chain of every mission, and it is the chain we worked in Chapter 29: the delta-v budget converts, through the rocket equation $m_0/m_f = e^{\Delta v/v_e}$, into a propellant mass, which sets the wet mass, which picks the launch vehicle, which sets the cost. Everything else on the table — power, thermal, comms, pointing — is a subsystem that hangs off that spine and adds mass to it. When we walk the four tracks below, we will always work the spine first, in full arithmetic, and then hang the subsystems on it.

🚪 Threshold Concept. A mission is not a collection of parts; it is a chain of consequences that begins with a single sentence of objective and ends with a number of dollars. Once you can see that chain — once you can trace how "provide broadband to a continent" becomes "1,800 m/s of orbit-raising" becomes "2,076 kg of propellant" becomes "one Falcon 9 to GTO" becomes a price — you are no longer a student of the parts. You are a mission designer, because you can hold the whole chain in your head and see where a change at one end ripples out the other. That view is what this book has been building toward, and it is what you will demonstrate in your MDR.

🔧 Engineering Reality: A real MDR package is hundreds of pages and dozens of people — a requirements database, trajectory analyses, subsystem trade studies, a risk register, a cost model, a schedule. What you are building is the skeleton of that: the ten rows above, each with a defensible number and a stated margin. That skeleton is exactly what a program manager sketches on a whiteboard in the first week, before the hundreds of pages exist, to decide whether the mission is even worth studying. Getting the skeleton right — and honest — is the whole game at this stage. The pages come later; the closure is decided now.

🔄 Check Your Understanding 1. In one sentence, what does it mean for a mission design to "close"? 2. Which single row of the MDR table, if you got it badly wrong, would most amplify into an error in all the rows below it — and why?

Answers

1. A design closes when every budget — delta-v, mass, power, cost, schedule — balances at once, with margin: a vehicle sized to the delta-v budget actually fits within its launch vehicle's capacity, its power supply, its mass allowance, and its money. 2. The delta-v budget. Through the rocket equation it converts exponentially into propellant mass, so a small error there is amplified into a large error in wet mass, which then picks the wrong launch vehicle and blows the cost — it sits at the top of the load-bearing chain, and everything downstream inherits its mistake.

40.2 Track A worked walkthrough: a communications satellite to GEO

Our first mission is the one that pays for much of the space economy: a communications satellite in geostationary orbit, hanging over a fixed longitude, relaying signals to a whole hemisphere. We will build its mini-MDR end to end. (If this is your track, check every number against your own notebook; if it is not, watch how the spine of 40.1 turns one sentence into one sized vehicle.)

Objective and driving requirement. Provide fixed broadband coverage to one continent for 15 years. From Chapter 29 we identified the driving requirement as that 15-year lifetime, because lifetime sets the station-keeping propellant, which is a large slice of the wet mass, which picks the launch vehicle, which dominates the cost.

Orbit. To appear motionless over one spot on Earth, the satellite must orbit once per sidereal day at the equator: geostationary orbit, radius $42{,}164\ \text{km}$, altitude $35{,}786\ \text{km}$, inclination $0^\circ$ (from Chapter 9). Its circular speed there is $v = \sqrt{\mu/r} = \sqrt{3.986\times10^{5}/42{,}164} = 3.075\ \text{km/s}$ (Earth $\mu = 3.986\times10^{5}\ \text{km}^3/\text{s}^2$).

Delta-v budget. The launch vehicle drops the satellite into a geostationary transfer orbit (GTO); the satellite's own propulsion must do the rest. Launching from Cape Canaveral ($28.5^\circ$ latitude) means the transfer plane is inclined $28.5^\circ$ to the equator, and that plane change is folded into the apogee burn (cheapest done high, where the satellite is slow — Chapter 10). The spacecraft-own budget (Tier 2 figures, consistent with the delta-v map of Appendix G):

Propulsion and vehicle sizing. Take a dry mass (bus + payload, no propellant) of $1{,}500\ \text{kg}$. Now the trade study of Chapter 29, rerun here as the sizing step, using $v_e = I_{sp}\,g_0$ and $m_0 = e^{\Delta v/v_e}\,m_f$:

• Chemical apogee engine, $I_{sp} = 320\ \text{s}$: $v_e = 320 \times 9.80665 = 3{,}138\ \text{m/s}$. Mass ratio $m_0/m_f = e^{2726/3138} = e^{0.869} = 2.38$. Wet mass $= 2.38 \times 1{,}500 = 3{,}576\ \text{kg}$; propellant $= 2{,}076\ \text{kg}$.
• Electric (xenon Hall) thruster, $I_{sp} = 1{,}800\ \text{s}$: $v_e = 17{,}652\ \text{m/s}$. Mass ratio $= e^{2726/17652} = e^{0.154} = 1.17$. Wet mass $= 1{,}750\ \text{kg}$; propellant (xenon) $= 250\ \text{kg}$.

The electric satellite is less than half the launch mass of the chemical one, for the identical mission, because its six-times-faster exhaust collapses the mass ratio toward 1. That is the rocket equation's exponential, for once, working for us. The price is time: all-electric orbit-raising is a months-long spiral, not a days-long series of burns (see Chapter 20).

Worked Example: does it fit its launch vehicle? A Falcon 9 can place roughly $5{,}500\ \text{kg}$ into GTO (Tier 2; Appendix H). The chemical comsat at $3{,}576\ \text{kg}$ fits one-to-a-rocket with margin. The electric comsat at $1{,}750\ \text{kg}$ fits two to a rocket — $2 \times 1{,}750 = 3{,}500\ \text{kg} < 5{,}500\ \text{kg}$ — halving the launch cost per satellite. This is exactly the real 2015 decision that put two all-electric Boeing 702SP satellites on one Falcon 9. The design closes, and the trade study has just paid for itself. $\blacksquare$

Spacecraft systems (hung on the spine). With the vehicle sized, the subsystems follow — each a mass that must be inside the $1{,}500\ \text{kg}$ dry budget:

Launch, timeline, and risk. One Falcon 9 to GTO; chemical orbit-raising over ~1 week (or electric over ~4–6 months); 15 years of station-keeping; disposal to a graveyard orbit ~$300\ \text{km}$ above GEO at end of life (Chapter 35). The dominant risk is not physics but duration: 15 years in the radiation environment with no repair means redundancy in every critical string (Chapter 32), and the driving requirement — that lifetime — is exactly what the review board will interrogate hardest.

🔄 Check Your Understanding 1. The chemical and electric comsats have the same delta-v budget (2,726 m/s) yet very different wet masses. Which single quantity is responsible, and through which equation? 2. Why is the plane change folded into the apogee burn rather than done at GEO or in LEO?

Answers

1. The engine's specific impulse: it sets $v_e$, hence the mass ratio $e^{\Delta v/v_e}$ (2.38 chemical vs 1.17 electric), through the rocket equation. Same delta-v, different exhaust velocity, ~2× different wet mass. 2. A plane change costs $\Delta v = 2v\sin(\Delta i/2)$, proportional to orbital speed $v$. At GEO altitude the satellite is slow (~3.1 km/s) versus LEO (~7.8 km/s), so turning the plane there is far cheaper — and combining it with the circularization burn (vector addition, law of cosines) is cheaper still than doing the two separately.

40.3 Track B: a lunar lander

Now a very different objective: deliver two tonnes of cargo to the surface of the Moon. The physics is unforgiving in a new way — the Moon has no atmosphere, so every bit of slowing, from lunar-orbit speed all the way to a gentle touchdown, must be bought with propellant. There is no free braking, no parachute, no aerocapture. This mission is where the tyranny of the rocket equation shows its teeth.

Objective and driving requirement. Land $2{,}000\ \text{kg}$ of cargo softly on the lunar surface, one way. The driving requirement is the landed cargo mass, because — as we are about to see — on a high-delta-v mission every kilogram of payload drags an enormous multiple of itself in propellant.

Orbit / destination. The route is Earth → trans-lunar injection (TLI) → low lunar orbit (LLO, ~$100\ \text{km}$, circular speed $\sqrt{4{,}903/1{,}837} = 1.63\ \text{km/s}$ using the Moon's $\mu = 4{,}903\ \text{km}^3/\text{s}^2$) → powered descent to the surface.

Delta-v budget. Tracing the route across the delta-v map (Appendix G), the spacecraft's own budget (the launch vehicle separately pays the ~$9.4\ \text{km/s}$ to LEO):

The descent leg carries the largest margin (15%) because the final approach — hover, hazard avoidance, and a divert to a safe spot — is the least predictable and least forgiving maneuver in the whole mission.

Propulsion and vehicle sizing. Landing demands high thrust to counter gravity during the hover, so this is chemical territory (electric thrusters cannot hover — Chapter 20). Take a storable or methalox engine at $I_{sp} = 320\ \text{s}$ ($v_e = 3{,}138\ \text{m/s}$), and a "landed mass" (dry structure + cargo) of $2{,}200\ \text{kg}$. Sizing the whole trip as a single stage:

$$ \frac{m_0}{m_f} = e^{5980/3138} = e^{1.906} = 6.72, \qquad m_0 = 6.72 \times 2{,}200 = 14{,}800\ \text{kg}. $$

Propellant $= 14{,}800 - 2{,}200 = 12{,}600\ \text{kg}$. The lander is 85% propellant by mass — for every kilogram you land on the Moon, you must launch nearly six kilograms of propellant to LEO to move it. Compare this mass ratio, 6.72, with the comsat's 2.38: the only difference is the delta-v (5,980 vs 2,726 m/s), and because it sits in the exponent, that roughly 2× more delta-v produces a nearly 3× larger mass ratio. This is theme 1, undiluted.

🐛 Find the Error. A student proposes to cut the lander's mass by doing the whole descent as a single stage — "why carry the dead weight of a separate stage?" — and is surprised the vehicle comes out at 14.8 tonnes. Where is the leverage they are missing?

Answer

They have the argument exactly backwards, and it is the same lesson as Falcon 9 in Chapter 3. A single stage carries all its dry structure and empty tanks through every leg, including the final descent, so its propellant works against dead mass the whole way. Staging the trip — a trans-lunar injection stage jettisoned after TLI, so the descent stage need not haul empty TLI tanks down to the surface — dramatically shrinks the vehicle. This is precisely why Apollo used lunar-orbit rendezvous with a dedicated, two-stage lunar module (Chapter 36, Chapter 22): the rocket equation forced it. The single-stage number here is not a design; it is the motivation for staging.

Launch, systems, and risk. At ~$14{,}800\ \text{kg}$ (single-stage) the fueled lander needs a heavy launcher — a Falcon 9 in expendable mode (~$22.8\ \text{t}$ to LEO), Falcon Heavy, Vulcan, or New Glenn (Chapter 30); staging the vehicle would cut this substantially. Power is modest (a short mission, solar or battery); thermal must survive the two-week lunar day–night cycle; GN&C needs terrain-relative navigation for the autonomous final descent, the lunar cousin of Mars's "seven minutes of terror" (Chapter 27). The dominant risk is the landing itself — a single, un-reversible, autonomous event with no abort.

🔗 Connection: The lander's brutal 6.72 mass ratio is the same arithmetic that shaped the entire Apollo program. Faced with a direct-ascent lander that would have needed a rocket far larger than the Saturn V, NASA chose lunar-orbit rendezvous — leaving the return vehicle in orbit and landing only a small, staged module. The rocket equation did not just influence that decision; it made it. Every lunar architecture since, from Apollo to Artemis, is a response to the number we just computed.

40.4 Track C: a Mars orbiter

Our third mission leaves Earth entirely: a science orbiter to map Mars from a low, near-polar orbit. Here the climax of the Hohmann-to-Mars thread (Chapter 11, Chapter 34) becomes one row of a design, and a beautiful trade appears: whether to brake into orbit with propellant or with the thin Martian air.

Objective, orbit, and window. Map Mars globally for one Mars-year from a ~$400\ \text{km}$ near-polar orbit. The interplanetary leg is the minimum-energy Hohmann transfer of Chapter 11, which we computed there and reuse now (Tier 2):

The launch vehicle (with an upper stage) supplies the departure $C_3$; the spacecraft's own delta-v begins at Mars. The question is how to stop.

The capture trade. Arriving at Mars with $v_\infty = 2.65\ \text{km/s}$, the spacecraft is on a hyperbola. At the periapsis of a $400\ \text{km}$ pass (radius $r = 3{,}790\ \text{km}$, Mars $\mu = 4.283\times10^{4}\ \text{km}^3/\text{s}^2$) its speed is

$$ v_{\text{hyp}} = \sqrt{v_\infty^2 + \frac{2\mu}{r}} = \sqrt{2.65^2 + \frac{2(4.283\times10^{4})}{3{,}790}} = \sqrt{7.02 + 22.60} = 5.44\ \text{km/s}. $$

The circular speed at that altitude is $v_c = \sqrt{\mu/r} = \sqrt{42{,}830/3{,}790} = 3.36\ \text{km/s}$. Two architectures compete:

• Propulsive capture straight to the low circular science orbit: burn $\Delta v = 5.44 - 3.36 = 2.08\ \text{km/s}$ at periapsis. Fast (one burn) but expensive.
• Aerobraking: burn only enough to capture into a loose elliptical orbit ($\Delta v \approx 0.9\ \text{km/s}$), then dip repeatedly into the upper atmosphere over ~6 months, letting drag lower the apoapsis to the science orbit for free. Cheap in propellant, expensive in time.

Building each budget (with cruise trajectory-correction maneuvers, TCMs, and margin):

Aerobraking saves ~$700\ \text{kg}$ of propellant — nearly a third of the wet mass — which is why almost every real Mars orbiter (Mars Global Surveyor, Odyssey, MRO, MAVEN) has used it. The cost is ~6 months of delicate, daily atmospheric passes before science can begin, and the risk of a pass that dips too deep. The trade is time and risk against mass, and for an orbiter with years of science ahead, mass usually wins.

💡 Intuition: Aerobraking is the interplanetary version of the comsat's electric-propulsion trade in 40.2: both buy a huge mass saving with a large chunk of time. The pattern is worth internalizing — in spaceflight, propellant and time are the two currencies you are forever trading between, and the rocket equation sets the exchange rate. Whenever a mission can afford to be patient, patience is almost always cheaper than propellant.

Systems, launch, and risk. At Mars the Sun is weaker — solar flux $S = 1{,}361/1.524^2 = 586\ \text{W/m}^2$, so arrays must be ~$2.6\times$ larger than at Earth for the same power (Chapter 25). Communications run through the Deep Space Network at X-band, with the ~$14\ \text{minute}$ one-way light time forcing autonomy at the critical moments (Chapter 26, Chapter 31). The launch vehicle need only be modest — the sized spacecraft (~$1.4\text{–}2.1\ \text{t}$) to $C_3 = 8.7\ \text{km}^2/\text{s}^2$ is an Atlas V- or Falcon 9-class launch, as MRO and MAVEN both were. The dominant risk is the single Mars-orbit-insertion burn: it happens once, behind the planet, out of contact, and it must work.

🔄 Check Your Understanding 1. Why does the launch window for Mars open only every ~26 months? 2. Aerobraking cut the propellant by ~700 kg. In one sentence, what did the mission pay for that saving, and why is that price often worth it for an orbiter?

Answers

1. Because a minimum-energy Hohmann transfer requires Earth and Mars to be in a specific relative geometry (a specific phase angle), and that geometry recurs only once per synodic period — the time for the faster Earth to lap the slower Mars — which is about 780 days, or 26 months (Chapter 11). 2. It paid ~6 months of slow, risky atmospheric passes before science operations could start; that is usually worth it because a lighter spacecraft means a cheaper launch, and an orbiter designed for years of science can afford to spend its first few months braking.

40.5 Track D: an asteroid rendezvous

The last mission is the most patient and, in one sense, the most extreme: a probe that matches orbits with a near-Earth asteroid to study it from close formation. You cannot "orbit" a body a few hundred metres across — its escape speed is centimetres per second — so rendezvous here means matching the asteroid's heliocentric orbit around the Sun and flying alongside it. Matching velocity, not just arriving, is what makes this mission expensive, and it is where high-specific-impulse electric propulsion stops being a nice option and becomes the only way the mass budget closes.

Objective and target. Rendezvous with and characterize a near-Earth asteroid. Take an illustrative accessible target on a low-eccentricity, low-inclination orbit with semi-major axis $a \approx 1.2\ \text{AU}$ (Tier 3 — a real target's eccentricity and, above all, its inclination dominate the true cost).

Delta-v budget. The heliocentric Hohmann from Earth's $1.0\ \text{AU}$ to $1.2\ \text{AU}$ is gentle — using $\mu_\odot = 1.327\times10^{11}\ \text{km}^3/\text{s}^2$, the departure and arrival excess speeds work out to only $v_{\infty,\text{dep}} \approx 1.32\ \text{km/s}$ and $v_{\infty,\text{arr}} \approx 1.27\ \text{km/s}$. But a rendezvous must cancel that arrival excess (a flyby would not), and a real target forces a few degrees of plane change — and at heliocentric speeds near $28\ \text{km/s}$, even $3^\circ$ costs $2(28)\sin(1.5^\circ) \approx 1.5\ \text{km/s}$. Add the low-thrust spiral's inefficiency, and the electric engine's job totals on the order of:

For scale: NASA's Dawn, which orbited two bodies (Vesta then Ceres) with plane changes, needed on the order of $11\ \text{km/s}$ — far beyond any chemical stage's reach (Chapter 20).

Why electric is mandatory. Size the $5{,}300\ \text{m/s}$ budget two ways for a $700\ \text{kg}$ dry probe:

• Gridded ion engine, $I_{sp} = 3{,}000\ \text{s}$ ($v_e = 29{,}420\ \text{m/s}$): mass ratio $= e^{5300/29420} = e^{0.180} = 1.20$. Wet mass $= 838\ \text{kg}$; xenon $= 138\ \text{kg}$.
• Chemical, $I_{sp} = 320\ \text{s}$ ($v_e = 3{,}138\ \text{m/s}$): mass ratio $= e^{5300/3138} = e^{1.689} = 5.41$. Wet mass $= 3{,}789\ \text{kg}$; propellant $= 3{,}089\ \text{kg}$.

The ion probe carries $138\ \text{kg}$ of xenon; the chemical one would need $3{,}089\ \text{kg}$ of propellant — more than twenty times as much — turning an $840\ \text{kg}$ spacecraft into a nearly $3{,}800\ \text{kg}$ one. For a high-delta-v deep-space mission, the exponential in the rocket equation is so punishing that only a high exhaust velocity can beat it. Electric propulsion does not merely help here; it is the difference between a mission that closes and one that does not.

🔗 Connection: Notice the same exponential, theme 1, appearing in every track with a different face. For the comsat (40.2) it made electric propulsion attractive; for the lunar lander (40.3) it made a single-stage vehicle enormous and forced staging; for the Mars orbiter (40.4) it made aerobraking worth six months; and here it makes electric propulsion mandatory. One equation, four missions, four different survival strategies — all responses to $m_0/m_f = e^{\Delta v/v_e}$.

Systems, power, and time. An ion engine drinking kilowatts needs a large solar array — the array, not the propellant, is the driving mass. At $1.2\ \text{AU}$ the Sun gives $S = 1{,}361/1.2^2 = 945\ \text{W/m}^2$, so the arrays are large but manageable (a far-out asteroid would instead demand an RTG — Chapter 25). Communications lean on the DSN and optical navigation — imaging the faint target against known stars — for the final approach (Chapter 13, Chapter 26). The launch is cheap (an $840\ \text{kg}$ probe to a slightly positive $C_3$ — a small vehicle or a rideshare), which reframes the whole mission: its scarcity is not launch mass but time — years of continuous, quiet ion thrusting, the way Dawn took four years just to reach Vesta. The dominant risk is duration and the single-string ion system running for those years without a repair truck.

40.6 Presenting and defending your design

You have a completed MDR. The last skill — the one that separates a design that flies from a design that sits in a drawer — is defending it. A real Mission Design Review is not an admiring audience; it is a board of experienced engineers whose job is to find the flaw you have grown too close to see (Chapter 32). They are on your side — better they find the flaw than space does — but they will not be gentle, and the way you present decides whether the review sharpens your design or shreds it.

Lead with the driving requirement. A board wants to know, in your first two sentences, what single requirement sizes your mission and why. "This is a 15-year GEO comsat; the 15-year lifetime is the driving requirement, because it sets the station-keeping propellant that dominates the wet mass." Now every reviewer knows what to listen for. A presenter who buries the driving requirement on slide 40 has already lost the room.

Show your margins, and show they are honest. The sharpest question at any review is some version of "show me your margins." Have them ready — mass, power, delta-v — and be able to say why each is the size it is (generous early, per the schedule of Chapter 29) and how you will buy it down. A margin you cannot justify reads as either padding or wishful thinking, and both cost you credibility.

Name your single points of failure. For each track above there is one un-abortable, un-repeatable event: the comsat's 15-year endurance, the lander's autonomous touchdown, the orbiter's do-or-die insertion burn, the asteroid probe's years-long single-string ion system. A board will find these whether you name them or not; naming them first, with your mitigation (redundancy, testing, reserve), turns an ambush into a strength.

📜 From History: The value of an adversarial review is written in the missions that skipped one. The Mars Climate Orbiter was lost in 1999 not to a physics error but to an unmanaged interface — one team's software used imperial units, another's metric — that a rigorous review of the navigation interface would have caught (Chapter 29). The board that asks "how do your two teams agree on units?" is not being pedantic; it is doing the one thing that saves missions. When you defend your design, welcome that question. The reviewer trying hardest to break your mission on the ground is the one most trying to save it in flight.

The six themes, one last time. Every mission you just designed is a meeting place of the book's six threads, and a good defense speaks to all of them:

1. The tyranny of the rocket equation. Every mass ratio you computed — 2.38, 6.72, 5.41 versus 1.20 — is a direct reading of the exponential. It is why you chose staging, aerobraking, or electric propulsion. It is the master constraint, and you now design against it by reflex.
2. Space is unforgiving. Every single point of failure you named exists because there is no repair shop in orbit. Redundancy, testing, and margin are not caution; they are survival.
3. Orbital mechanics is beautiful. The same $\mu$ and vis-viva sized a GTO apogee burn, a lunar capture, a Mars insertion, and a heliocentric match. One gravity, applied carefully, took you everywhere.
4. Mass is the enemy. Every subsystem kilogram you fought for was a kilogram multiplied by the mass ratio into wet mass. You learned to see a gram of structure as costing several grams of propellant.
5. Reusability is changing everything. The launch that carries your mission may cost a tenth of what it would have a generation ago (Chapter 38) — which is why missions like your smallsat comsat or asteroid probe are now thinkable at all.
6. History matters. You chose lunar-orbit rendezvous because Apollo showed why; you chose aerobraking because Mars orbiters proved it; you chose electric propulsion because Dawn flew it. Every decision stood on someone's earlier one.

🔧 Engineering Reality: The hardest thing to present is not a number; it is an honest uncertainty. A mature designer says "this delta-v is Tier 2 — good to a few percent — and here is the margin that covers my being wrong," not "this is 2,726 m/s" with false precision. Boards trust the presenter who flags what they do not know over the one who pretends to know everything, because the second presenter is the one whose mission fails on the thing they glossed over. Confidence in a review is not the absence of doubt; it is knowing exactly where your doubt is and having carried margin against it.

🔄 Check Your Understanding 1. Why should a design defense open with the driving requirement rather than a description of the spacecraft? 2. A reviewer finds a real flaw in your architecture at the MDR. Why is this the best possible outcome, and not a failure?

Answers

1. Because the driving requirement tells the board what single number sizes the entire mission, so every subsequent choice can be judged against it; without it, reviewers cannot tell which of your decisions are load-bearing and which are incidental. 2. Because a flaw caught at the MDR costs a line in a document to fix, while the same flaw caught later costs redesign, or scrapped hardware, or — if space finds it — the mission. The cost of a flaw rises about tenfold at each gate, so catching it at the earliest gate is the cheapest and safest place it can possibly be found.

Mission Design Checkpoint: your completed MDR and the assembled astrotools

This is the last checkpoint, and it is not an increment — it is the completion. Everything you have added since Chapter 1 now stands as one document and one program.

The design — your finished MDR. Assemble the ten rows of 40.1 for your track into a single package, and confirm it closes: a vehicle sized to your margined delta-v budget must fit within your launch vehicle's capacity, your power supply, your mass budget, and (in the real world) your money. Write the one-paragraph elevator summary that opens a review: driving requirement → delta-v budget → mass ratio → wet mass → launch vehicle → cost. If you can say that paragraph and defend every number in it, you have designed a space mission.

The code — the whole astrotools package, working together. Every module you built — rocket, orbits, maneuvers, interplanetary, propulsion, attitude, power, thermal, comms, and the mission synthesizer — now composes into one end-to-end sizing tool. The capstone driver assembles them and runs a mission from orbit geometry to a confirmed, sized vehicle (the full package is reproduced in Appendix I):

# capstone: size the Track-A comsat end to end using the assembled astrotools package.
from astrotools.orbits    import circular_velocity                 # Ch. 6
from astrotools.maneuvers import hohmann, plane_change             # Ch. 10
from astrotools.mission   import roll_up_dv, size_vehicle          # Ch. 29

MU = 3.986e5                          # km^3/s^2  (work in km, km/s)
R_LEO, R_GEO = 6771.0, 42164.0

# 1. geometry -> the transfer legs (km/s), plane change folded into the apogee burn
dv1, dv2, _ = hohmann(MU, R_LEO, R_GEO)                # LEO->GTO, GTO->GEO
gtogeo = 1.800                                          # km/s incl. 28.5 deg plane change (Ch.10)

# 2. delta-v budget (m/s) -> roll up with per-leg margins  (Ch. 29)
budget = [("GTO->GEO", 1800, 0.05), ("SK 15yr", 750, 0.10), ("disposal", 11, 0.0)]
ideal, margined = roll_up_dv(budget)

# 3. size the vehicle from the margined budget (Ch. 3 rocket equation, inverted)
m_prop, m_wet = size_vehicle(margined, isp=320, payload=1500)      # chemical option

# 4. confirm closure against the launch vehicle (Ch. 30, Appendix H)
F9_TO_GTO = 5500.0                                                  # kg, Tier 2
print(f"GEO speed:  {circular_velocity(MU, R_GEO):.3f} km/s")
print(f"budget:     {ideal} ideal, {margined:.0f} margined m/s")
print(f"vehicle:    propellant {m_prop:.0f} kg, wet {m_wet:.0f} kg")
print(f"closes on Falcon 9 to GTO? {m_wet < F9_TO_GTO}")
# Expected output:
# GEO speed:  3.075 km/s
# budget:     2561 ideal, 2726 margined m/s
# vehicle:    propellant 2076 kg, wet 3576 kg
# closes on Falcon 9 to GTO? True

The four numbers in that output are the spine of the entire Track-A MDR — the orbit it must reach, the delta-v it must supply, the vehicle that supplies it, and the confirmation that a real rocket can lift it. The full worked driver, plus the three other tracks, is in code/project-checkpoint.py and the example-*.py files. Point the same package at your own mission's numbers, and it sizes your vehicle too. That is the whole point of astrotools: not this one comsat, but any mission you will ever want to design.

Summary

The capstone introduces no new physics; it assembles the book into a complete Mission Design Review. Carry these forward:

Worked numbers to remember: Track A comsat — 2,726 m/s margined, wet ~3,576 kg (chemical) / ~1,750 kg (electric), one/two per Falcon 9. Track B lander — 5,980 m/s, wet ~14,800 kg single-stage (why we stage). Track C Mars orbiter — 259-day cruise, 26-month window; aerobraking (~1,400 kg) vs propulsive (~2,090 kg). Track D asteroid — ~5,300 m/s, ion wet ~838 kg (138 kg xenon) vs chemical ~3,789 kg. All vehicle masses are Tier 2/3 — sized to illustrate the method, not to fly.

Spaced Review

This is the book's final retrieval set, and it is deliberately broad: five questions spanning Parts I–V. Answer from memory before checking — if you can, you have integrated the whole book.

1. (Ch. 3) The lunar lander's mass ratio (6.72) is far larger than the comsat's (2.38), though both use the same $I_{sp}=320\ \text{s}$ engine. What single quantity explains the difference, and why does a roughly 2× change in it produce a nearly 3× change in mass ratio?
2. (Ch. 6 & Ch. 10) The comsat's plane change is folded into the apogee (GEO-altitude) burn. Using $\Delta v = 2v\sin(\Delta i/2)$, why is turning the plane cheaper at GEO than in LEO?
3. (Ch. 11) Why can a Mars mission launch only in a window that opens every ~26 months, and what is that interval called?
4. (Ch. 20 & Ch. 16) For the asteroid probe, ion propulsion needed 138 kg of xenon where chemical would need 3,089 kg. What property of the ion engine causes this, and what does the mission pay for it?
5. (Ch. 23 & Ch. 29) A structures engineer saves you 8% of dry mass and your mission's mass ratio is 3.0. Why is that 8% dry saving worth much more than 8% of the wet mass?

Answers

1. The delta-v (5,980 vs 2,726 m/s). Because it sits in the exponent of $m_0/m_f = e^{\Delta v/v_e}$, a change in delta-v is amplified: doubling the exponent squares the ratio, so ~2× more delta-v yields far more than 2× the mass ratio. 2. Plane-change cost is proportional to orbital speed $v$; the satellite is much slower at GEO (~3.1 km/s) than in LEO (~7.8 km/s), so the same angle costs less than half as much up high — and combining it with the circularization burn saves more still. 3. Because a minimum-energy Hohmann transfer needs Earth and Mars in a specific relative geometry, which recurs only once per synodic period (~780 days ≈ 26 months) as the faster Earth laps the slower Mars. 4. Its far higher specific impulse (~3,000 s vs ~320 s) means a ~9× higher exhaust velocity, which collapses the mass ratio toward 1; the mission pays for it in time (years of low-thrust cruise) and in the electrical power to run the engine. 5. Because dry mass and wet mass are linked by the rocket equation: every kilogram of dry mass drags $m_0/m_f = 3.0$ kilograms of wet mass, so an 8% dry saving becomes a ~24% wet-mass saving — the mass ratio is the amplifier.

What's Next: beyond the book

There is no forty-first chapter. There is the rest of your life with this material, and it is worth saying plainly what you can now do that you could not forty chapters ago: you can design a space mission. Not hand-wave one — design it. Given an objective, you can pick an orbit, build a delta-v budget, choose an engine, size the vehicle with the rocket equation, select a launch vehicle, sketch the subsystems, and defend the whole thing in a review. That is a real, rare, and durable skill, and it is yours now whether your next step is a career, a degree, or a very serious hobby.

Where you point it is up to you. If you want the professional version of everything here, Space Mission Engineering: The New SMAD is the book the industry actually uses, and your astrotools package is a toy model of the tools inside it. If you want to feel the rocket equation in your hands, Kerbal Space Program is a mission-design spiral you can run in an evening, and it will lie to you less than your intuition does. If you want to go deeper on the mechanics, Curtis and Vallado await; on propulsion, Sutton and Biblarz. And if you want to build the real thing, the field has never in its history been more open to newcomers — the reusability revolution of Chapter 38 and the small-satellite democratization of Part V have thrown the doors wide, and the future we mapped in Chapter 39 will be built by people who, right now, know exactly what you now know.

This is one book in the DataField series; there are others, on other hard and beautiful subjects, built the same way — physics first, then engineering, then the confidence to do something with both. But this one was about rockets, and rockets are, in the end, a kind of argument: that a species clever enough to work out $\Delta v = v_e \ln(m_0/m_f)$ on paper is clever enough to ride it off the planet. Tsiolkovsky made that argument from a deaf schoolteacher's desk in provincial Russia, decades before the technology existed, because he trusted the physics. You have now followed the same physics from a thrown ball to a complete mission to Mars. The cradle, as he said, is not a place to stay.

Go design something. The equation is on your side.