Chapter 4: Getting to Orbit

"There is an art to flying, or rather a knack. The knack lies in learning how to throw yourself at the ground and miss." — Douglas Adams, Life, the Universe and Everything

Overview

A rocket stands on the pad pointing straight at the sky, and it lifts off going straight up, so almost everyone's first mental model of spaceflight is the same: to get to space, you go up. That model is not just incomplete — it is wrong in a way that hides the entire physics of orbit. Going up gets you to space. It does not get you to orbit. And the difference between those two things is the difference between a firework and a satellite.

Here is the sentence this whole chapter exists to install in your head: orbit is sideways, not up. An object is in orbit not because it is high, but because it is moving sideways so fast that as gravity pulls it down, the round Earth curves away beneath it just as quickly — so it falls, and falls, and never arrives. Astronauts on the International Space Station are not "beyond gravity." Gravity out there is nearly as strong as on the ground. They are falling, continuously, at almost eight kilometers every second, and missing the planet. That is what an orbit is.

In Chapter 3 we built the rocket equation, $\Delta v = v_e \ln(m_0/m_f)$, and learned that delta-v is the currency of spaceflight. We quoted a price for reaching low Earth orbit — about $9.4\ \text{km/s}$ — and promised to explain where that number comes from, and in particular why it is larger than the roughly $7.8\ \text{km/s}$ at which a satellite actually circles the Earth. That promise comes due now. The extra kilometers per second are eaten by two thieves that the rocket equation, sitting in an imaginary vacuum with no planet under it, never sees: gravity, which pulls you back down during every second of the climb, and drag, the resistance of the air you punch through on the way up. Getting to orbit is the art of paying those thieves as little as possible while turning a vertical launch into a horizontal, orbital velocity.

In this chapter, you will learn to:

• See orbit correctly — as sideways speed, not altitude — using Newton's own thought experiment.
• Follow a real rocket through its ascent: the vertical rise, the pitch-over, and the gravity turn.
• Quantify gravity losses and drag losses, and finally account for the $9.4\ \text{km/s}$ to orbit.
• Understand the trade-off that shapes every ascent trajectory, and why thrust-to-weight matters.
• Tell suborbital from orbital flight, and see why reaching orbit is about 30 times harder than reaching space.

Learning Paths

🚀 Space Enthusiast: Section 4.1 is the one that will change how you watch a launch forever — read it slowly. Then skim 4.2 and 4.6; the "missing the ground" idea in 4.6 is the payoff. You can treat the loss integrals in 4.3–4.4 as "gravity and drag steal delta-v" without the algebra.

📐 Engineering Student: Read everything. The gravity-loss integral in 4.3 and the ascent trade-off in 4.5 are the conceptual foundation of trajectory optimization; the ⭐⭐/⭐⭐⭐ exercises and both case studies are where the real learning is. You will use the launch delta-v budget for the rest of the book.

🎮 KSP Player: You have flown a hundred gravity turns without naming them. Sections 4.2 and 4.5 put the physics under your thumb: why you pitch east right after liftoff, why "straight up" wastes fuel, and why a too-heavy or too-light rocket both fly badly. Try the exercises on thrust-to-weight.

🛰️ Industry Prep: Section 4.3's loss accounting and 4.5's optimal-trajectory framing are the vocabulary of ascent analysis. The Mission Design Checkpoint adds the single largest line item in any Earth-launched mission's delta-v budget — the launch itself.

4.1 Orbit is sideways, not up

Let us do the experiment that Isaac Newton did on paper in the 1680s, three centuries before anyone could do it for real. It is the clearest thing ever written about orbits, and it needs no equations to feel.

🧩 Productive Struggle. Imagine a cannon on top of an impossibly tall mountain — tall enough to poke above the atmosphere, so there is no air. The cannon fires horizontally, parallel to the ground far below. Fire it gently and the ball arcs down and lands a few kilometers away. Fire it harder and it lands farther. Now keep increasing the muzzle speed. What happens to where it lands? Is there a special speed at which something qualitatively new occurs? Sit with this before reading on.

Fire the cannonball slowly and it traces the familiar arc of a thrown stone: it falls as it flies, and buries itself in the ground a short distance away. Fire it faster and it travels farther before landing — but notice why it lands. It lands because the ground stays flat under it while the ball curves down. On a flat Earth, every cannonball eventually hits the ground, no matter how fast, because "down" and "the ground" are the same direction everywhere.

But the Earth is not flat. It is a ball about $6{,}371\ \text{km}$ in radius, and its surface curves away from any straight line. Fire the cannonball fast enough and something wonderful happens: the ball still falls toward the Earth exactly as before — gravity has not changed — but now, in the time it takes to fall, the round Earth has curved out from under it by the same amount. The ball is always falling and the ground is always dropping away, and the two exactly cancel. The cannonball never gets any closer to the surface. It has entered orbit. It will come all the way around the planet and, if you duck, hit you in the back of the head.

🚪 Threshold Concept: orbit is sideways, not up. An orbit is not a place you reach by going high. It is a speed you reach by going sideways. A satellite stays up not because something holds it up — nothing does — but because it is moving horizontally so fast that its constant fall toward Earth is perfectly matched by the curvature of the planet falling away beneath it. Height alone buys you nothing: lift a ball to the altitude of the Space Station and let go, and it drops straight back down. Give that same ball enough horizontal velocity and it will circle the Earth forever. Once you truly see this, every launch looks different. The rocket is not climbing a ladder to space; it is turning a vertical lift-off into a horizontal sprint, and the sprint is the hard part.

How fast is "fast enough"? We can find the sideways speed for a low, circular orbit with nothing more than Newton, which is to say with the physics of Chapter 2. For a circular orbit, gravity is the centripetal force — the inward pull that bends a straight-line motion into a circle. Newton's second law for circular motion says the required inward force per unit mass is $v^2/r$, and gravity supplies an inward acceleration of $\mu/r^2$, where $\mu = GM$ is Earth's gravitational parameter (Chapter 2), $\mu = 3.986\times10^{5}\ \text{km}^3/\text{s}^2$. Set them equal:

$$ \frac{v^2}{r} = \frac{\mu}{r^2} \quad\Longrightarrow\quad v = \sqrt{\frac{\mu}{r}}. $$

That is the circular orbital velocity. (We are deriving it here as a pure force balance; the general energy treatment for any orbit — the vis-viva equation — waits for Chapter 6, and Kepler's full geometry for Chapter 8.) Put in the numbers for an orbit skimming just above the surface, $r \approx R_\oplus = 6{,}371\ \text{km}$:

Worked Example: how fast is sideways? With $\mu = 3.986\times10^5\ \text{km}^3/\text{s}^2$ and $r = 6{,}371\ \text{km}$, $$v = \sqrt{\frac{3.986\times10^5}{6{,}371}} = \sqrt{62.6\ \text{km}^2/\text{s}^2} = 7.91\ \text{km/s}.$$ A body must move sideways at about $7.9\ \text{km/s}$ — roughly $28{,}500\ \text{km/h}$, or Mach 23 — to circle the Earth just above the ground. At the more realistic altitude of a real orbit, say $200\ \text{km}$ ($r = 6{,}571\ \text{km}$), the speed is a touch lower, $v = \sqrt{3.986\times10^5/6{,}571} = 7.79\ \text{km/s}$, which is the $\sim 7.8\ \text{km/s}$ we quoted in Chapter 3. Higher orbits are slower, a beautiful and counterintuitive fact we will make precise in Chapter 6. For now, hold onto the headline number: orbital speed is about $7.8\ \text{km/s}$, and it is sideways.

Now let us make the "falling and missing" picture quantitative, because the numbers are almost too neat. A body in low orbit moves about $7.9\ \text{km}$ sideways every second. Over that same second, how far does the round Earth's surface curve away beneath a straight horizontal line? A little geometry (the "sagitta" of a circular arc) gives the drop as $d^2/(2R)$ for a horizontal distance $d$:

$$ \text{drop} = \frac{d^2}{2R} = \frac{(7{,}900\ \text{m})^2}{2\,(6.371\times10^6\ \text{m})} = \frac{6.24\times10^7}{1.274\times10^7}\ \text{m} = 4.9\ \text{m}. $$

And how far does an object fall under gravity in one second? From Chapter 2, $\tfrac12 g t^2 = \tfrac12 (9.81)(1)^2 = 4.9\ \text{m}$. The two numbers are identical. In one second the orbiting body falls $4.9\ \text{m}$ toward Earth, and in that same second the Earth's surface curves $4.9\ \text{m}$ away. It falls forever and never lands. That coincidence is not a coincidence at all — it is the orbit, expressed in meters.

Here is that whole argument in code you can keep:

import math
MU_EARTH = 3.986e5   # km^3/s^2, Earth's gravitational parameter (Ch. 2)
R_EARTH  = 6371.0    # km

def circular_velocity(mu, r):
    """Circular orbital speed (km/s) at radius r (km): v = sqrt(mu/r)."""
    return math.sqrt(mu / r)

v = circular_velocity(MU_EARTH, R_EARTH)          # sideways speed near the surface
d = v * 1.0                                        # km travelled horizontally in 1 s
drop = (d * 1000)**2 / (2 * R_EARTH * 1000)        # metres the ground curves away
fall = 0.5 * 9.81 * 1.0**2                          # metres a free body falls in 1 s
print(f"orbital speed near surface: {v:.2f} km/s")
print(f"in 1 s: move {d:.1f} km sideways; ground curves away {drop:.1f} m; body falls {fall:.1f} m")
# Expected output:
# orbital speed near surface: 7.91 km/s
# in 1 s: move 7.9 km sideways; ground curves away 4.9 m; body falls 4.9 m

⚠️ Common Misconception: "Space starts at 100 km, so I just need to climb 100 km." Altitude and orbit are almost unrelated problems. Crossing the boundary of space (conventionally the Kármán line, about $100\ \text{km}$ up) takes only a modest vertical hop, as we will compute in §4.6. Staying there — orbiting — takes $\sim 7.8\ \text{km/s}$ of sideways speed, which is a vastly larger investment. A vehicle that goes straight up to $100\ \text{km}$ with no sideways velocity comes straight back down a few minutes later. It visited space; it never orbited. The height is the cheap part; the sideways speed is the whole ballgame.

🔄 Check Your Understanding 1. Two objects are both at the altitude of the Space Station ($\sim 400\ \text{km}$). One is moving sideways at $7.7\ \text{km/s}$; the other was carried up by a balloon and released at rest. What happens to each? 2. Using $v = \sqrt{\mu/r}$, does orbital speed increase or decrease as you go to a higher orbit?

Answers

1. The first is in orbit — it falls toward Earth but the surface curves away beneath it, so it circles the planet indefinitely. The second simply falls straight down and crashes; altitude without sideways speed is not an orbit. 2. It decreases: $r$ is in the denominator under the square root, so larger $r$ gives smaller $v$. Higher orbits are slower — we will see the surprising consequences in Chapter 6.

4.2 The gravity turn

We have established the destination: not a height, but a horizontal velocity of about $7.8\ \text{km/s}$. The launch vehicle's whole job is to rotate a standing-still rocket into that sideways sprint, as cheaply as possible. The elegant maneuver that does it is called the gravity turn, and the schedule of orientations it follows is the pitch program.

Definition (gravity turn). A gravity turn (or zero-lift turn) is an ascent maneuver in which the rocket lifts off vertically, pitches over by a small angle shortly after launch, and then holds its thrust pointed straight along its own velocity vector, letting gravity gradually bend that velocity from vertical toward horizontal. The vehicle applies no sideways steering force; it simply falls forward as it climbs, and the falling does the turning.

Definition (pitch program). The pitch program is the pre-planned schedule of the vehicle's pitch angle (its orientation relative to the local vertical) as a function of time or altitude during ascent — the sequence of commanded attitudes that carries it from a vertical lift-off to a horizontal orbital insertion. In a pure gravity turn the pitch program is mostly hands-off: after one small initial kick, the rocket lets physics choose its angle.

Why turn this way, rather than simply steering the rocket over with brute force? Two reasons, one about stress and one about efficiency, and they happen to agree.

The stress reason is decisive. A launch vehicle is a slender tube full of propellant under pressure, built as light as physics allows (theme #4, mass is the enemy). If it flies through thick air at an angle of attack — nose pointed even a few degrees away from the oncoming airflow — the air pushes sideways on the whole body with enormous force, and the vehicle can bend or break. So during the part of the flight where the air is dense, the rocket must keep its angle of attack essentially at zero: it must point exactly into its own airflow, which means exactly along its velocity vector. A gravity turn does precisely this by construction. The thrust is always aligned with the velocity, so the vehicle never presents its flank to the wind. (Aerodynamic forces, the angle of attack you avoid, and the pressure the vehicle feels — which peaks at a point called max-Q — are the subject of the next chapter, Chapter 5.)

The efficiency reason we will make quantitative in §4.3, but the essence is simple: any thrust you spend pushing sideways to steer is thrust not spent adding speed, and any moment you spend pointed upward is a moment gravity is stealing your delta-v. Letting gravity do the turning for free, while you point your engine straight along the direction you are already going, wastes neither.

Here is how the maneuver actually unfolds on a real vehicle:

1. Vertical rise. The rocket lifts off straight up, to clear the launch tower and to get out of the thickest air as directly as possible. This lasts only a few seconds.
2. The pitch-over kick. A few seconds to a few tens of seconds after lift-off, while still moving slowly, the vehicle tips its nose a small angle (a few degrees) in the direction it wants to go — downrange, usually toward the east. This tiny, deliberate maneuver is the only active steering the gravity turn needs.
3. The gravity turn proper. Now the vehicle holds zero angle of attack — thrust along velocity — and simply flies. Because it has a small horizontal component of velocity and gravity pulls straight down, the velocity vector rotates, second by second, a little further from vertical toward horizontal. No steering force is applied; the arc is carved by gravity alone.
4. Insertion. By the time the vehicle is high and fast, near the edge of the sensible atmosphere, it is flying almost horizontally, pouring the last of its delta-v into pure sideways speed until it reaches the $\sim 7.8\ \text{km/s}$ that means orbit.

📜 From History: The first vehicle to reach space, Germany's V-2 in 1944, did not fly a proper gravity turn — it followed a programmed tilt and arced over as a ballistic missile, aimed to come back down on a city. The gravity turn as we now fly it was refined for orbital launch, where the goal is the opposite of the V-2's: not to come back down, but to convert the climb into enough horizontal speed to never come down. The same physics of a tilted, falling-forward trajectory serves a bomb and a satellite; only the final speed, and the intent, differ. Engineering is often like this — the mathematics is neutral, and history decides what it is for.

💡 Intuition: Picture riding a bicycle and letting go of the handlebars on a gentle downhill curve. You do not wrench the wheel; you lean, and gravity and momentum carry you smoothly around. The gravity turn is the rocket "leaning" into its climb and letting the planet's pull ease it from pointing up to pointing sideways — no yanking on the controls, no stress on the frame, just a falling forward that happens to trace the most graceful possible path to orbit.

4.3 Gravity losses

Now we confront the first thief. In the ideal world of the rocket equation, every bit of exhaust velocity becomes vehicle velocity. But a rocket climbing away from Earth is not in that ideal world: for every second its engines burn, gravity is pulling it back down, and some of the delta-v the equation promises is spent merely holding the vehicle up against that pull instead of speeding it along. That stolen delta-v is the gravity loss.

Definition (gravity loss). Gravity loss (also gravity drag) is the reduction in the velocity a rocket actually gains, compared with the ideal delta-v the rocket equation would deliver in gravity-free space, because a component of the vehicle's acceleration is continuously canceled by gravity during a burn. Quantitatively, if $\gamma$ is the flight-path angle — the angle of the velocity vector above the local horizontal (careful: this $\gamma$ is not the specific-heat ratio that shares the symbol in Chapter 19) — then the gravity loss accumulated over a burn of duration $t_b$ is $$\Delta v_{\text{grav}} = \int_0^{t_b} g\,\sin\gamma \; dt .$$

Read that integral physically. Only the component of gravity along the direction of flight slows you down; that component is $g\sin\gamma$. When the rocket flies straight up ($\gamma = 90^\circ$, $\sin\gamma = 1$), gravity fights the motion head-on and the loss rate is the full $g \approx 9.81\ \text{m/s}$ of delta-v drained every second. When the rocket flies horizontally ($\gamma = 0^\circ$, $\sin\gamma = 0$), gravity pulls perpendicular to the motion — it bends the path (that is what keeps you in orbit!) but takes nothing off your speed, so the gravity loss rate is zero. This single fact is the deep reason for the gravity turn: the faster you can get from vertical to horizontal, the less delta-v gravity steals. Every second spent pointed up is expensive; the art of ascent is to stop pointing up as soon as the atmosphere and the need to gain some altitude allow.

Let us bound the loss and then estimate it. Take a first stage that burns for $t_b = 162\ \text{s}$ (a Falcon-9-class number). If it somehow flew straight up the entire time — the worst case — the gravity loss would be the full

$$ \Delta v_{\text{grav}}^{\text{(max)}} = g\,t_b = 9.81\ \text{m/s}^2 \times 162\ \text{s} = 1{,}589\ \text{m/s} \approx 1.6\ \text{km/s}. $$

That is a colossal penalty — more than a fifth of orbital speed, gone. A real gravity turn does far better because it spends the later part of the burn tilted well over. If the average value of $\sin\gamma$ over the first-stage burn is about $0.6$ (the vehicle is vertical at first, then increasingly horizontal), the first stage's gravity loss is roughly

$$ \Delta v_{\text{grav}} \approx g\,\langle\sin\gamma\rangle\,t_b = 9.81 \times 0.6 \times 162 \approx 954\ \text{m/s}. $$

The second stage, flying nearly horizontal through near-vacuum, adds only a few hundred more. Summed over the whole ascent, a typical orbital launch pays a gravity loss of roughly $1.2$ to $1.6\ \text{km/s}$. (These are representative figures; the exact value depends on the vehicle's thrust-to-weight and its pitch program, and I would flag any single number here as illustrative rather than exact.)

G0 = 9.81  # m/s^2

def gravity_loss(g, sin_gamma_avg, t_burn):
    """Gravity loss (m/s) = g * <sin(flight-path angle)> * burn time."""
    return g * sin_gamma_avg * t_burn

print(round(gravity_loss(G0, 1.0, 162)))   # worst case: straight up the whole burn
print(round(gravity_loss(G0, 0.6, 162)))   # a real gravity turn, stage 1
# Expected output:
# 1589
# 954

Now for the moment this chapter has been building toward — the payoff of a promise made back in Chapter 3. In Chapter 3 we computed, from real masses, that Falcon 9's first stage delivers an ideal delta-v of about $3.86\ \text{km/s}$. Yet when that first stage shuts down (an event called MECO, main-engine cutoff) roughly $2.5$ minutes into flight, the vehicle is traveling at only about $2.3\ \text{km/s}$. The rocket equation handed the stage $3.86\ \text{km/s}$ of delta-v; the speedometer reads $2.3$. Where did the missing $\sim 1.5\ \text{km/s}$ go? It went to the thieves — overwhelmingly to gravity loss, with a smaller slice to drag (§4.4). That $\sim 1.5\ \text{km/s}$ gap, between the delta-v a stage spends and the speed the vehicle actually has, is exactly the extra the rocket must carry beyond orbital velocity. It is the concrete answer to Chapter 3's puzzle: the $9.4\ \text{km/s}$ to orbit is the $7.8\ \text{km/s}$ you end up going, plus the delta-v the thieves took while you were getting there.

🔧 Engineering Reality: thrust-to-weight and the hover trap. Gravity loss depends brutally on how fast you climb, which is set by the vehicle's thrust-to-weight ratio (the ratio of engine thrust to vehicle weight at lift-off; we treat it rigorously in Chapter 16). Consider the extreme: a rocket with thrust exactly equal to its weight, thrust-to-weight $= 1$. It hovers. Its engines burn furiously, propellant streams out, and the vehicle goes nowhere — every bit of its delta-v is being spent at rate $g$ just to not fall, gaining zero speed and zero height. That is a $100\%$ gravity loss, the hover trap. It is why real launch vehicles lift off with thrust-to-weight around $1.2$ to $1.5$: high enough to climb briskly and get the gravity-loss clock stopped, but not so high that acceleration loads and drag (§4.4) become punishing. If you ever wondered "why can't a rocket just rise gently to save fuel?" — this is the answer. Rising gently means burning against gravity for longer, which costs more delta-v, not less.

🔄 Check Your Understanding 1. A rocket flies straight up for $30\ \text{s}$. Ignoring drag, how much delta-v does it lose to gravity during that time? 2. Two identical rockets fly to orbit; one pitches over aggressively early, the other stays vertical longer. Which pays more gravity loss, and why might the aggressive one nonetheless get in trouble?

Answers

1. $\Delta v_{\text{grav}} = g\,t = 9.81 \times 30 \approx 294\ \text{m/s}$ — nearly $0.3\ \text{km/s}$ gone in half a minute of vertical flight. 2. The one that stays vertical longer pays more gravity loss (larger $\sin\gamma$ for longer). But pitching over too aggressively drives the vehicle into thicker air at higher speed and can build a dangerous angle of attack, raising drag and structural loads — the tension we resolve in §4.5.

4.4 Drag losses (intro)

The second thief is the atmosphere. A rocket does not climb through empty space; for its first minutes it tears through air, and that air pushes back. The delta-v spent overcoming that push is the drag loss.

Definition (drag loss). Drag loss is the delta-v a launch vehicle expends against aerodynamic drag during atmospheric ascent — the retarding drag force integrated over the climb and divided by the vehicle's mass: $$\Delta v_{\text{drag}} = \int_0^{t_b} \frac{D}{m}\; dt, \qquad D = \tfrac12\,\rho\,v^2\,C_d\,A ,$$ where $D$ is the drag force, $\rho$ is the local air density, $v$ the speed, $A$ the vehicle's frontal area, and $C_d$ its drag coefficient. The quantity $q = \tfrac12\rho v^2$ is the dynamic pressure; it and its famous peak, max-Q, are developed in Chapter 5.

The striking thing about drag loss is how small it is — typically only about $0.1\ \text{km/s}$ (sometimes up to a few tenths for a bulky or slow vehicle), a tenth or less of the gravity loss. That seems impossible at first: a rocket reaches Mach 20-plus, and drag grows as the square of speed, so surely the drag near orbital velocity is ruinous? The resolution is the beautiful conspiracy hidden in $D = \tfrac12\rho v^2 C_d A$: the two factors that matter, air density $\rho$ and speed $v$, are large at opposite times. Low down, where the air is thick, the rocket is still slow ($v$ small, so $v^2$ tiny). By the time the rocket is fast, it is high up, where the air has thinned to almost nothing. Earth's atmosphere falls off roughly exponentially with a scale height of about $8\ \text{km}$ — every $8\ \text{km}$ of altitude cuts the density by a factor of $e \approx 2.7$ — so by $50\ \text{km}$ the density is under a thousandth of sea level, and above the Kármán line ($100\ \text{km}$) it is negligible. The rocket does the vast majority of its accelerating in a near-vacuum it has already climbed above. Drag only gets a brief, low-speed window to bite.

Worked Example: how hard does drag actually push? Estimate the drag deceleration at the worst moment, max-Q, for a Falcon-9-class vehicle. Take the peak dynamic pressure as $q \approx 30\ \text{kPa} = 30{,}000\ \text{Pa}$ (a typical value), a body of diameter $3.7\ \text{m}$ so frontal area $A = \pi(1.85)^2 = 10.75\ \text{m}^2$, a supersonic drag coefficient $C_d \approx 0.4$, and a vehicle mass at that instant of about $360\ \text{t} = 3.6\times10^5\ \text{kg}$. The drag force is $$D = q\,C_d\,A = 30{,}000 \times 0.4 \times 10.75 = 1.29\times10^5\ \text{N} \approx 130\ \text{kN},$$ and the deceleration it causes is $$\frac{D}{m} = \frac{1.29\times10^5\ \text{N}}{3.6\times10^5\ \text{kg}} = 0.36\ \text{m/s}^2.$$ Compare that with gravity's $9.81\ \text{m/s}^2$: even at its very worst, drag robs the vehicle about thirty times more slowly than gravity does. And max-Q lasts only briefly; integrated over the whole ascent, drag loss comes to roughly $0.1\ \text{km/s}$. (All figures here are representative — Tier 2/3 — and meant to establish the order of magnitude, not a precise value for any specific flight.)

So the two thieves are wildly unequal. Gravity loss is the giant, $\sim 1.5\ \text{km/s}$; drag loss is the pickpocket, $\sim 0.1\ \text{km/s}$. That imbalance matters, because it tells you where to spend your cleverness. A rocket engineer will accept a bit more drag to cut gravity loss, almost every time — which is exactly the trade-off that shapes the optimal trajectory, and the subject of §4.5.

🔗 Connection: Drag loss is small, but drag itself is not harmless. The same aerodynamic pressure that steals a little delta-v also crushes and shakes the vehicle, and its peak — max-Q — is one of the most stressful moments of any launch, the point at which controllers watch most anxiously and at which many rockets throttle their engines down to ease the load. That structural story, along with the payload fairing that shelters the cargo and the aerodynamic heating of ascent, is Chapter 5. Here we only needed drag's small bite out of the delta-v budget; there we meet its full violence.

4.5 The optimal ascent trajectory

We can now assemble the whole picture. The path a rocket flies from the pad to orbit — the ascent trajectory — is the solution to a genuine optimization problem, a tug-of-war between the two thieves and the vehicle's own fragility.

Definition (ascent trajectory). The ascent trajectory is the complete path, together with its pitch and throttle program, that a launch vehicle follows from lift-off to orbital insertion. The optimal ascent trajectory is the one that reaches the target orbit for the least total delta-v — that is, the one that minimizes the sum of gravity loss and drag loss (plus small steering losses) while respecting the vehicle's structural and heating limits.

The tug-of-war has two sides, and they pull in opposite directions:

• To minimize gravity loss, you want to go horizontal as soon as possible. Every second spent pointed upward bleeds delta-v at up to $g$ per second. The gravity-minimizing extreme would be to pitch over almost immediately and fly nearly sideways from the start.
• To minimize drag loss (and structural stress), you want to go up first and pitch over late. Rising steeply gets the vehicle out of the thick lower atmosphere by the most direct route, spending as little time as possible in the dense air where drag and dynamic pressure are worst. The drag-minimizing extreme would be to climb straight up until the air is thin, then turn.

You cannot satisfy both. Pitch over too early and you drag the vehicle sideways through thick air at building speed — high drag, high dynamic pressure, dangerous loads. Pitch over too late and gravity robs you blind while you point uselessly at the sky. The optimal ascent threads between them, and it looks like exactly what real rockets do: rise nearly vertically through the dense lower atmosphere, begin a gentle gravity turn while still low and slow, and then bend progressively toward horizontal as the air thins, arriving nearly sideways by the time the engines are pushing you the last few km/s to orbital speed. It is a compromise trajectory, and the compromise is legible in the flight of every launcher you will ever watch.

Two more levers shape the optimum, both tied to thrust-to-weight:

• Too little thrust-to-weight and the vehicle climbs sluggishly, lengthening the burn and running up the gravity-loss clock (recall the hover trap of §4.3).
• Too much thrust-to-weight and the vehicle accelerates so hard low down that it hits high speeds while still in thick air, spiking drag and max-Q, punishing the structure, and subjecting the payload to heavy g-loads. It may even be worth throttling down through max-Q, as Falcon 9 does, to tame the peak.

The sweet spot — lift-off thrust-to-weight around $1.2$–$1.5$ — is not a coincidence or a tradition; it falls out of minimizing total losses subject to not breaking the rocket. This is theme #1, the tyranny of the rocket equation, wearing a trajectory: because delta-v is so precious and so exponentially costly (Chapter 3), shaving even $100\ \text{m/s}$ off the losses is worth real engineering effort, and the entire shape of an ascent is bent around that thrift.

🐛 Find the Error. A student reasons: "Gravity loss is the big one, $\sim 1.5\ \text{km/s}$, and drag loss is tiny, $\sim 0.1\ \text{km/s}$. So the optimal trajectory must be the one that kills gravity loss entirely: pitch over to horizontal immediately after lift-off and never point up at all." What is wrong with this?

Answer

The student is right that gravity loss dominates the delta-v budget, but wrong that you may therefore ignore drag and structure. Pitching to horizontal immediately would drive the vehicle at rapidly building speed through the densest part of the atmosphere, where dynamic pressure and drag scale as $\rho v^2$. That does two things the delta-v accounting alone misses: it can raise drag loss far above $0.1\ \text{km/s}$, and — far more dangerously — it subjects the airframe to a large angle of attack and a huge aerodynamic load that can simply tear the vehicle apart (Chapter 5). You also need some vertical climb just to gain altitude and get above the sensible atmosphere before you can safely go fast. The optimum is a genuine compromise: minimize gravity loss as much as the atmosphere and the structure allow, not absolutely. A "loss-optimal" trajectory that breaks the rocket has infinite cost.

🔗 Connection: Notice that the launch vehicle usually pitches toward the east. That is not arbitrary: the Earth itself is rotating eastward, and a due-east launch from near the equator gives the rocket a running start of up to about $0.46\ \text{km/s}$ — free velocity subtracted straight off the required launch delta-v. This is why launch sites cluster at low latitudes and why the direction you launch (the azimuth) is a design choice, a story told in Chapter 30. For now, just note that our $9.4\ \text{km/s}$ figure is a no-credit baseline; an eastward equatorial launch can shave it toward $\sim 8.9\ \text{km/s}$.

4.6 Suborbital vs. orbital: "missing the ground"

We are ready to sharpen the distinction that this whole chapter has circled, and to put a number on how much harder orbit is than mere altitude.

Definition (suborbital). A suborbital flight reaches space — it crosses the boundary conventionally placed at the Kármán line, about $100\ \text{km}$ — but does not achieve orbital velocity. Its trajectory is a ballistic arc: it goes up, arcs over, and falls back to the surface. It is in space only briefly, and it never circles the Earth.

Definition (orbital flight). Orbital flight is a trajectory with enough tangential (sideways) speed — about $7.8\ \text{km/s}$ in low Earth orbit — that the vehicle continuously falls toward Earth while the curved surface falls away beneath it, so it circles the planet without descending. It is, in the language of §4.1, perpetual falling that keeps missing the ground.

The gap between them is enormous, and the cleanest way to see it is through energy, because energy does not care about the direction of motion. A suborbital hop that just reaches $100\ \text{km}$ needs, ideally (ignoring the losses that would push the real figure higher), a launch speed of

$$ v_{\text{hop}} = \sqrt{2 g h} = \sqrt{2 \times 9.81 \times 100{,}000} = 1{,}401\ \text{m/s} \approx 1.4\ \text{km/s}, $$

whereas orbit needs $v_{\text{orbit}} \approx 7.8\ \text{km/s}$ sideways. Because kinetic energy grows as the square of speed, the energy to reach orbit compared with the energy to reach space is

$$ \frac{E_{\text{orbit}}}{E_{\text{hop}}} \approx \left(\frac{7{,}800}{1{,}401}\right)^2 \approx 31 . $$

Reaching orbit takes about thirty times the energy of reaching space. And essentially all of that extra energy is kinetic — it is the sideways speed. Space is a place you can touch with a stunt; orbit is a regime you must earn with an enormous, relentless investment in horizontal velocity. This is the quantitative heart of "space is hard" from Chapter 1: the hard part was never the height.

V_ORBIT_LEO = 7800.0    # m/s, orbital speed in low LEO
GRAV_LOSS   = 1500.0    # m/s, representative gravity loss to orbit
DRAG_LOSS   = 100.0     # m/s, representative drag loss

def launch_delta_v(v_orbit, grav_loss, drag_loss, rot_bonus=0.0):
    """Ideal launch delta-v (m/s) a vehicle must supply to reach orbit."""
    return v_orbit + grav_loss + drag_loss - rot_bonus

leo    = launch_delta_v(V_ORBIT_LEO, GRAV_LOSS, DRAG_LOSS)
leo_eq = launch_delta_v(V_ORBIT_LEO, GRAV_LOSS, DRAG_LOSS, rot_bonus=465)  # due east, equator
v_hop  = (2 * 9.81 * 100_000) ** 0.5                                        # coast to 100 km
print(f"delta-v to LEO (no rotation credit): {leo:.0f} m/s")
print(f"delta-v to LEO (equatorial, due east): {leo_eq:.0f} m/s")
print(f"suborbital hop needs ~{v_hop:.0f} m/s; orbit/hop energy ratio = {(V_ORBIT_LEO**2)/(v_hop**2):.0f}x")
# Expected output:
# delta-v to LEO (no rotation credit): 9400 m/s
# delta-v to LEO (equatorial, due east): 8935 m/s
# suborbital hop needs ~1401 m/s; orbit/hop energy ratio = 31x

This distinction is not academic; it is the fault line of the whole launch industry. Vehicles like Blue Origin's New Shepard, Virgin Galactic's SpaceShipTwo, and, historically, the X-15 and the V-2 are suborbital: they reach space with something like $1$–$2\ \text{km/s}$ of delta-v and come back. Vehicles like Falcon 9, the Saturn V, and every satellite launcher are orbital: they must find the full $\sim 9.4\ \text{km/s}$. A suborbital tourist flight and an orbital launch look superficially alike — both are rockets going up — but they live on opposite sides of a roughly thirtyfold energy chasm. That chasm is why "we sent a rocket to space" and "we put something in orbit" are entirely different sentences.

💡 Intuition: the knack of orbiting. Recall the epigraph. Douglas Adams joked that the trick to flying is to throw yourself at the ground and miss. That is not a joke about orbits — it is a definition of them. An orbiting spacecraft is doing exactly that, forever: hurling itself at the Earth at $7.8\ \text{km/s}$ and, because it is going sideways so fast and the planet is round, missing. The rocket's entire job on launch day is to throw the payload sideways hard enough that it misses the ground on the way down. Everything in this chapter — the gravity turn, the loss budget, the optimal trajectory — is bookkeeping in service of that one absurd, beautiful trick.

🔄 Check Your Understanding 1. New Shepard reaches about $100\ \text{km}$ and lands its passengers minutes later. Roughly how does its delta-v requirement compare with an orbital launch, and why does it come back down? 2. A friend says "the Space Station is so high that there's no gravity there." Correct them in two sentences.

Answers

1. A suborbital hop to $100\ \text{km}$ needs only $\sim 1.4\ \text{km/s}$ ideal (a few km/s with losses), versus $\sim 9.4\ \text{km/s}$ for orbit — and in energy about thirty times less. It comes back down because it never gained the $\sim 7.8\ \text{km/s}$ of sideways speed needed to keep missing the ground; with no orbital velocity, its ballistic arc simply falls back. 2. Gravity at the Station's altitude (~$400\ \text{km}$) is about $90\%$ of its value at the ground — nearly full strength. The astronauts float because they and the Station are in continuous free fall together, orbiting sideways fast enough to miss the Earth, not because gravity is absent.

Mission Design Checkpoint: your launch delta-v and rocket.py

In Chapter 3 you began your mission's delta-v budget — the list of maneuvers your spacecraft's own propulsion must perform after a launch vehicle drops it off. Now we add the biggest single line item in any Earth-launched mission: getting off Earth in the first place.

The design. Open your Mission Design Review (MDR) and add a top line to the delta-v budget labeled "launch to LEO ≈ 9.4 km/s." Then annotate it with a crucial distinction: for all four tracks (A: comsat to GEO, B: lunar lander, C: Mars orbiter, D: asteroid rendezvous), this $\sim 9.4\ \text{km/s}$ is normally the launch vehicle's job, not your spacecraft's — the rocket you select in Chapter 30 burns it to place you in low Earth orbit, and your spacecraft's own tanks are sized for everything after that (the GTO/GEO, trans-lunar, or trans-Mars legs you listed in Chapter 3). Recording the launch delta-v explicitly, even though a separate vehicle pays it, keeps the whole energy story honest and drives your launch-vehicle choice later. Note in your MDR whether your target orbit's speed differs from the $7.8\ \text{km/s}$ LEO value, and whether an eastward, low-latitude launch could shave the number.

The code. Extend astrotools/rocket.py (begun in Chapter 3) with a launch estimate:

import math
G0 = 9.80665  # standard gravity, m/s^2

def mass_ratio(dv, ve):
    """Mass ratio m0/mf for a target delta-v (m/s) at exhaust velocity ve (m/s)."""
    return math.exp(dv / ve)

def ascent_delta_v(v_orbit, gravity_loss=1500.0, drag_loss=100.0, rotation_bonus=0.0):
    """Launch delta-v (m/s) to reach an orbit of tangential speed v_orbit (m/s).
    Adds representative gravity and drag losses; subtracts an optional Earth-rotation
    credit (up to ~465 m/s, due east from the equator)."""
    return v_orbit + gravity_loss + drag_loss - rotation_bonus

dv_launch = ascent_delta_v(7800)                 # LEO, no rotation credit
print("launch delta-v to LEO:", round(dv_launch), "m/s")
print("vehicle mass ratio needed (ve ~ 3.0 km/s):", round(mass_ratio(dv_launch, 3000)))
# Expected output:
# launch delta-v to LEO: 9400 m/s
# vehicle mass ratio needed (ve ~ 3.0 km/s): 23

That mass ratio of about $23$ is the punchline that ties this chapter back to the last one. A single kerosene stage, we learned in Chapter 3, tops out near a mass ratio of $1/\varepsilon \approx 12.5$ because structure cannot be made massless. But the losses of ascent demand a mass ratio of $\sim 23$ — nearly double the ceiling. No single stage can do it. The gravity and drag losses of getting to orbit are, quantitatively, part of why staging is not optional but mandatory. In later chapters these same functions will point at your mission: ascent_delta_v sets the launch requirement, and the rocket equation converts it into the vehicle you will need to select and size.

Summary

Getting to orbit is the art of converting a vertical lift-off into $\sim 7.8\ \text{km/s}$ of horizontal speed while paying the two thieves — gravity and drag — as little as possible. Carry these forward:

Numbers worth memorizing: orbital speed $\approx 7.8\ \text{km/s}$; launch delta-v to LEO $\approx 9.4\ \text{km/s}$; gravity loss $\sim 1.5\ \text{km/s}$, drag loss $\sim 0.1\ \text{km/s}$; Earth-rotation credit up to $\sim 0.46\ \text{km/s}$ (eastward, equatorial); orbit is $\sim 30\times$ the energy of a $100\ \text{km}$ hop.

Spaced Review

Retrieval strengthens memory. Answer from memory before checking, then look back at the cited chapter.

1. (Ch. 2) Earth's escape velocity from the surface is about $11.2\ \text{km/s}$ and circular orbital velocity is about $7.9\ \text{km/s}$. What is the ratio, and why does it come out that way?
2. (Ch. 2) By Newton's first law, once a satellite is in orbit in vacuum, why does it need no engine to stay there?
3. (Ch. 3) A launch vehicle must supply $9.4\ \text{km/s}$. With an effective $v_e = 3.0\ \text{km/s}$, what mass ratio does it need — and what does that imply about staging?
4. (Ch. 3) An engine has $I_{sp} = 330\ \text{s}$. What is its exhaust velocity, roughly?

Answers

1. The ratio is $11.2/7.9 \approx 1.41 = \sqrt{2}$. Because escape velocity is $v_{\text{esc}} = \sqrt{2\mu/r}$ and circular velocity is $v_{\text{circ}} = \sqrt{\mu/r}$, so $v_{\text{esc}} = \sqrt{2}\, v_{\text{circ}}$ at any given radius. 2. Newton's first law: with no atmosphere there is no drag to slow it, and gravity acts perpendicular to its motion (bending the path into a circle, not braking it), so it coasts indefinitely — the reason orbits above the atmosphere do not decay. 3. Mass ratio $= e^{9400/3000} = e^{3.13} \approx 23$. That is nearly double the $\sim 12.5$ ceiling a single realistic stage can reach, so a single stage cannot get to orbit — staging is mandatory. 4. $v_e = I_{sp}\,g_0 = 330 \times 9.81 \approx 3{,}240\ \text{m/s} \approx 3.2\ \text{km/s}$.

What's Next

We have treated the atmosphere as a mere thief of a little delta-v, but it is far more than that. The same air that costs $0.1\ \text{km/s}$ of drag also presses on the vehicle with a dynamic pressure that peaks — at the dreaded max-Q — hard enough to crush a poorly built rocket, shakes it with buffeting, heats its skin, and dictates when the payload fairing can be jettisoned. Having budgeted the atmosphere's small bite out of our delta-v, we now turn to its full physical violence. In Chapter 5 we compute dynamic pressure, locate max-Q, anatomize the aerodynamic forces on an ascending vehicle, and see why rockets throttle down and hold their angle of attack near zero as they punch through the sky.