Attitude control modes — spin stabilization, 3-axis active

WHY do we need attitude control at all?

WHAT it is: keeping (or changing) which way the spacecraft is facing — where the camera looks, where the antenna points, where the solar panels catch sunlight.

WHY it's hard: in orbit there is (almost) no friction and no gravity to lean on. Tiny disturbance torques — solar radiation pressure, atmospheric drag, gravity-gradient, magnetic — slowly tumble the craft. Without control it drifts to a random orientation.

HOW we fight this: exploit the physics of rigid-body rotation, governed by Euler's rotational equation (derived below).

First-principles derivation: rigid-body rotation

Derivation of Euler's equation (WHY it takes this form):

Newton's law for rotation in an inertial frame: $\vec{\tau}_{\text{ext}} = \left(\frac{d\vec{L}}{dt}\right)_{\text{inertial}}$

But $\mathbf{I}$ is constant only in the body frame (it rotates with the craft). Using the transport theorem, for any vector $\vec{A}$ : $\left(\frac{d\vec{A}}{dt}\right)_{\text{inertial}} = \left(\frac{d\vec{A}}{dt}\right)_{\text{body}} + \vec{\omega}\times\vec{A}$

Why this step? In the body frame $\mathbf{I}$ is fixed, so the derivative is easy; the cross-term accounts for the frame itself spinning. Setting $\vec{A}=\vec{L}=\mathbf{I}\vec\omega$ :

$\boxed{\;\vec{\tau} = \mathbf{I}\dot{\vec{\omega}} + \vec{\omega}\times(\mathbf{I}\vec{\omega})\;}$

For principal axes ( $\mathbf{I}=\text{diag}(I_1,I_2,I_3)$ ) this splits into three scalar Euler equations: $\tau_1 = I_1\dot\omega_1 + (I_3-I_2)\omega_2\omega_3$ $\tau_2 = I_2\dot\omega_2 + (I_1-I_3)\omega_3\omega_1$ $\tau_3 = I_3\dot\omega_3 + (I_2-I_1)\omega_1\omega_2$

Why we care: the cross-terms are free gyroscopic coupling. Spin stabilization USES them; 3-axis control must FIGHT them.

Mode 1 — Spin Stabilization

HOW it resists disturbance (precession, not tumble): A constant transverse torque $\tau$ makes $\vec L$ precess: $\Omega_{\text{prec}} = \frac{\tau}{L} = \frac{\tau}{I\omega_s}$ Why this step? From $\vec\tau = \vec\Omega\times\vec L$ , a bigger spin $L$ → smaller precession rate → the pointing "wanders" far more slowly. Big $\omega_s$ = stiff.

Limits: you can point only the spin axis; to re-point you must slowly precess it with thrusters. No fast slews. Cameras must "de-spin" or use scanning.

Mode 2 — 3-Axis Active Control

HOW — reaction wheels (momentum exchange): Total angular momentum is conserved: $\vec L_{\text{body}} + \vec L_{\text{wheels}} = \text{const}$ . Spin a wheel up → body counter-rotates. Torque on body: $\vec\tau_{\text{body}} = -\,\dot{\vec h}_{\text{wheel}}$ Why this step? No external torque needed — you just trade momentum with an internal flywheel (Newton's 3rd law for rotation).

Saturation problem: wheels spin faster with every correction and eventually saturate (max RPM). You then desaturate ("momentum dump") using thrusters or magnetorquers to dump wheel momentum to the environment.

Figure — Attitude control modes — spin stabilization, 3-axis active

Side-by-side

Feature	Spin stabilization	3-axis active
Principle	Gyroscopic stiffness (passive $\vec L$ )	Feedback torques (active)
Pointing	One axis only	Any axis, any time
Slew speed	Slow (precession)	Fast
Complexity/cost	Low	High (sensors + wheels + computer)
Failure mode	Flat-spin if $I_{\text{spin}}\neq I_{\max}$	Wheel saturation

Worked example Worked #2 — Reaction wheel slew

Body $I=500\ \text{kg·m}^2$ , must rotate $30°$ . Wheel provides constant $\tau=0.1\ \text{N·m}$ accelerating half the slew, decelerating the other half.

Half angle $=15°=0.262$ rad. $\theta=\tfrac12\alpha t^2$ , $\alpha=\tau/I=2\times10^{-4}\ \text{rad/s}^2$ .
$t=\sqrt{2\theta/\alpha}=\sqrt{2(0.262)/2\times10^{-4}}\approx51\ \text{s}$ per half. Why split? Bang-bang: accelerate then brake so you arrive at rest.
Total $\approx102$ s. Peak wheel momentum $h=I\alpha t=500(2\times10^{-4})(51)=5.1\ \text{N·m·s}$ — check it's below wheel limit.

Common mistake Steel-man the classic errors

1. "Spin about the long, thin axis — it looks most stable, like a pencil balanced tip-up." Why it feels right: a thrown pencil or bullet seems to spin nicely about its long axis. The truth: a bullet is stable only because it has NO energy dissipation and gets gyroscopic support; a real flexing spacecraft dissipates energy and must spin about the maximum-inertia axis. Explorer 1 tumbled for exactly this reason. Fix: max- $I$ axis = min energy at fixed $L$ = the only dissipation-stable state.

2. "Reaction wheels create torque from nothing / forever." Why it feels right: they seem to spin the craft with no fuel. Truth: they only exchange momentum; they saturate and must be desaturated by external torque (thrusters/magnetorquers). Fix: remember total angular momentum is conserved.

3. "More spin = better pointing accuracy on all axes." Why: stiffness scales with $L$ . Truth: it stiffens only the spin axis; the other two axes are useless for independent pointing. Fix: spin gives 1-axis pointing only.

Recall Feynman: explain to a 12-year-old

Imagine you're floating in a swimming pool and want to turn to face the diving board — but there's nothing to grab. Two tricks:

Spin like a top: if you're already spinning fast, you're hard to knock off course — like a spinning bike wheel resists tipping. But you can only "look" along your spinning axis. That's spin stabilization.
Carry a heavy spinning disk: speed the disk up one way and you turn the other way (like a swivel chair spinning when you throw your arms). Point sensors tell a computer how wrong you are, and it nudges the disks to fix it — on all three directions. That's 3-axis active. The catch: the disks spin faster and faster until they max out, so occasionally you use gas jets to "reset" them.

Flashcards

Euler's rigid-body equation (vector form)

\vec\tau = \mathbf{I}\dot{\vec\omega} + \vec\omega\times(\mathbf{I}\vec\omega)

Why does spin give pointing stability?

Large angular momentum

L=I\omega_s

; a disturbance torque only precesses

\vec L

slowly at

\Omega=\tau/L

, giving gyroscopic stiffness.

Which axis must a dissipating spacecraft spin about, and why?

The maximum-inertia axis, because

T=L^2/2I

is minimized there and energy dissipation drives the body to minimum energy → stable.

What was the Explorer 1 lesson?

It spun about its minimum-inertia (long) axis and flat-spun; dissipation grew the wobble.

How does a reaction wheel torque the body without fuel?

Momentum exchange:

\vec\tau_{body}=-\dot{\vec h}_{wheel}

; conservation of total angular momentum.

What is wheel saturation and its fix?

Wheels reach max RPM and can't add more momentum; fix by desaturation/momentum dump using thrusters or magnetorquers.

1-axis PD control law and closed-loop equation

\tau=-K_p\theta-K_d\dot\theta \Rightarrow I\ddot\theta+K_d\dot\theta+K_p\theta=0

(damped oscillator).

Damping ratio for a PD-controlled axis

\zeta=K_d/(2\sqrt{K_pI})

; choose

\approx0.7

for fast, low-overshoot settling.

Precession rate under transverse torque

\Omega=\tau/(I\omega_s)

Main trade: spin vs 3-axis

Spin = cheap/passive/1-axis; 3-axis = costly/complex/any-axis, fast slews.

Connections

Euler's rotational equations of motion
Moment of inertia tensor & principal axes
Reaction wheels and Control-Moment Gyros (CMG)
Momentum dumping / desaturation with magnetorquers
PID control & second-order systems (damping ratio)
Gravity-gradient & solar-radiation disturbance torques
Gyroscopic precession
Explorer 1 flat-spin anomaly

Concept Map

Hinglish (regional understanding)

Intuition Hinglish mein samjho

Dekho, space mein satellite ko turn karna easy nahi hai — koi zameen ya deewar nahi hai jise push karke ghoom sako. Isliye do main tarike hain. Pehla: spin stabilization — pura satellite ek lattu (top) ki tarah tez ghumaao. Jab uska angular momentum $L=I\omega_s$ bada hota hai, to koi bhi disturbance torque use aasani se tilt nahi kar pata — bas dheere-dheere precess karta hai rate $\Omega=\tau/L$ par. Matlab jitna tez spin, utna stiff pointing. Catch yeh hai ki sirf ek axis (spin axis) ko hi point kar sakte ho.

Dusra tarika: 3-axis active control — yahan satellite ghoomta nahi, balki andar reaction wheels lagi hoti hain. Ek wheel ko tez ghumao to Newton ke third law se body ulti taraf ghoom jaati hai (jaise ghumne wali chair par haath phenko to chair ghoom jaati hai). Sensor batata hai ki kitni error hai, computer PD law $\tau=-K_p\theta-K_d\dot\theta$ se torque nikalta hai, aur teeno axes ko independently point kar sakte ho — fast bhi. Problem: wheels ka RPM badhta rehta hai aur ek din saturate ho jaata hai; tab thrusters ya magnetorquers se "momentum dump" karke reset karna padta hai.

Ek super important baat — spin ke liye axis kaunsa chuno? Hamesha maximum moment of inertia wala axis (fat, flat, frisbee jaisa). Kyunki fixed $L$ par energy $T=L^2/2I$ tabhi minimum hoti hai jab $I$ max ho, aur real satellite mein fuel slosh/flexing se energy dissipate hoti hai jo body ko min-energy state mein le jaati hai. Agar tumne patla (min- $I$ ) axis choose kiya to satellite flat-spin mein chala jayega — yahi Explorer 1 ke saath hua tha. Yaad rakho: "Fat and flat is where it's at."

Attitude control modes — spin stabilization, 3-axis active

WHY do we need attitude control at all?

First-principles derivation: rigid-body rotation

Mode 1 — Spin Stabilization

Mode 2 — 3-Axis Active Control

Side-by-side

Flashcards

Connections

Concept Map

Hinglish (regional understanding)

Go deeper — visual, from zero

Test yourself — Guidance, Navigation & Control (GNC)

Connections