Visual walkthrough — Attitude control modes — spin stabilization, 3-axis active

---

Step 1 — What is angular velocity, really?

WHAT. A rigid body (nothing bends — think a solid brick, not jelly) that is turning has an axis it spins around and a rate it spins at. We bundle both into one arrow called the angular velocity, written $\vec{\omega }$ (the little arrow means "this is a vector — it has a direction").

WHY an arrow and not just a number? Because "how fast" alone doesn't tell you which way the body turns. The arrow points along the spin axis; its length is the spin rate (radians per second). Point your right thumb along the arrow — your curling fingers show the direction of turning. That is the whole convention.

PICTURE. The green arrow $\vec{\omega }$ lies along the axis; the blue circle is the actual motion of a point on the body.

---

Step 2 — Spin does not always point where you push

WHAT. We now need angular momentum $\vec{L}$ — the rotational version of "how much oomph the spin carries." For a point mass it is mass × how far it is from the axis × how fast. For a whole body we add up every chunk.

WHY introduce $\vec{L}$ at all? Because the law of the universe for rotation is written in terms of $\vec{L}$, not $\vec{\omega }$. And here is the surprise the picture must show: $\vec{L}$ and $\vec{\omega }$ usually point in different directions. A body that is wide in one direction and thin in another throws its momentum sideways relative to its spin.

PICTURE. Same object spinning about the green $\vec{\omega }$, but the yellow $\vec{L}$ leans away from it — because more mass sits off to one side.

---

Step 3 — Principal axes: the three directions where life is simple

WHAT. For any rigid body there exist three special perpendicular axes where the tilting stops: spin about one of them and $\vec{L}$ points exactly along $\vec{\omega }$. These are the principal axes, and their three "hardness" numbers are $I_1,I_2,I_3$ (the principal moments of inertia).

WHY do we hunt for them? Because along these axes the messy table $\mathbf{I}$ collapses to just three numbers on a diagonal: $\mathbf{I}=\left(\begin{array}{ccc}{I}_1& 0& 0\\ 0& I_2& 0\\ 0& 0& I_3\end{array}\right)$ The zeros mean "no sideways leaking." Every equation from here on is written in this frame, so it stays readable.

PICTURE. The three colored axes of a drum-shaped satellite; the flat "frisbee" axis has the biggest $I$, the two through the rim have smaller $I$.

---

Step 4 — The law of rotation lives in the wrong frame

WHAT. Newton's rotational law is: an external torque $\vec{\tau }$ (a twist — force applied off-axis) changes the angular momentum: $\vec{\tau }={\left(\frac{d\vec{L}}{dt}\right)}_{\text{inertial}}$ Read it as "twist = rate of change of spin-momentum," measured by someone floating still in space (the inertial frame — a frame that doesn't rotate).

WHY is this a problem? The nice three-number $\mathbf{I}$ only holds in the body frame — the frame bolted to the spacecraft, which is itself spinning. So $\vec{L}$'s components are simple in the body frame, but the law demands the rate of change seen from the still frame. Two different observers, one equation. We must connect them.

PICTURE. Same yellow arrow $\vec{L}$: on the left the calm inertial observer, on the right the spinning body observer. The arrow is identical — but how fast each sees it change differs, because the right one is turning underneath it.

---

Step 5 — Assembling Euler's equation

WHAT. Put $\vec{A}=\vec{L}=\mathbf{I}\vec{\omega }$ into the bridge. In the body frame $\mathbf{I}$ is constant, so its body-frame change is just $\mathbf{I}\dot{\vec{\omega }}$ (the dot means "rate of change in time"). Substitute: $\vec{\tau }=\mathbf{I}\dot{\vec{\omega }}+\vec{\omega }\times (\mathbf{I}\vec{\omega })$

WHY this is the master key. Everything in attitude control is this one line. The first term is the twist you apply; the second is a free twist the spin gives you for nothing. Spin stabilization lives off that free term; 3-axis control (see Reaction wheels and Control-Moment Gyros (CMG)) has to overpower it. See also Euler's rotational equations of motion.

PICTURE. The equation drawn as a flow: applied torque → straight speed-up term + curling gyroscopic term.

Along the three principal axes this splits into three plain scalar lines: ${\tau }_1=I_1{\dot{\omega }}_1+\underset{\text{free gyroscopic coupling}}{\underbrace{(I_3-I_2){\omega }_2{\omega }_3}}$ ${\tau }_2=I_2{\dot{\omega }}_2+(I_1-I_3){\omega }_3{\omega }_1$ ${\tau }_3=I_3{\dot{\omega }}_3+(I_2-I_1){\omega }_1{\omega }_2$

---

Step 6 — Precession: why a spun-up craft is "stiff"

WHAT. Give the craft one big spin ${\omega }_s$ about, say, axis 3, so $L=I_3{\omega }_s$ is large. Now a small steady sideways torque $\tau$ hits it. Instead of tipping over, the spin axis slowly sweeps in a cone — it precesses — at rate ${\Omega }_{\text{prec}}=\frac{\tau }{L}=\frac{\tau }{I_3{\omega }_s}.$

WHY it doesn't just fall. From $\vec{\tau }=\vec{\Omega }\times \vec{L}$: the torque can only rotate $\vec{L}$ perpendicular to itself, not shrink it. Bigger $L$ → the same push turns it more slowly → the pointing drifts at a snail's pace. That slowness is the "gyroscopic stiffness."

PICTURE. The yellow $\vec{L}$ traces a shallow cone; a fat $L$ makes a skinny, slow cone (stiff); a thin $L$ makes a fast wide cone (floppy).

---

Step 7 — Energy picks the winner: the max-inertia rule

WHAT. A real spacecraft leaks rotational energy — fuel sloshes, panels flex, joints rub. But it cannot leak angular momentum without an external torque (nothing to push on). So $L$ is fixed while kinetic energy $T$ slowly drains. For spin about a principal axis: $T=\frac{L^2}{2I}.$

WHY this decides the axis. With $L$ locked, $T$ is smallest when $I$ is largest. Dissipation always slides the body downhill toward minimum energy — i.e. toward spinning about the maximum-inertia axis. If you start spinning about the minimum-inertia axis, you're at maximum energy for that $L$; the tiniest wobble releases energy and grows, until the craft flops onto its fat axis. That is the Explorer 1 flat-spin anomaly.

PICTURE. An energy "hill": min-$I$ spin sits on the peak (unstable), max-$I$ spin sits in the valley (stable). Dissipation is the ball rolling down.

---

Step 8 — The degenerate cases (never leave a gap)

WHAT & WHY (each its own scenario):

• Perfect sphere, $I_1=I_2=I_3$. All coupling terms vanish; $\vec{L}\parallel \vec{\omega }$ always. Every axis is equally "fine," but there's no stiffness advantage anywhere — spin does nothing special.
• Symmetric body, $I_1=I_2\ne I_3$ (a drum/frisbee). The two rim axes tie. Spin about $I_3$: if $I_3$ is the largest (frisbee) → stable; if $I_3$ is the smallest (pencil) → unstable under dissipation. This is the shape most satellites actually use.
• Zero spin, ${\omega }_s=0$. Then $L=0$, precession rate $\Omega =\tau /L\to \infty$: no gyroscopic stiffness at all. A non-spinning craft has no passive resistance — it needs active reaction wheels and a PID loop instead.
• Intermediate axis, $I_{\min }<I_{\text{spin}}<I_{\max }$. Unstable even without dissipation — the famous "tennis-racket" flip. Never spin here.

PICTURE. A little map: green valley = max-$I$ (safe), red peak = min-$I$ (Explorer 1), yellow saddle = intermediate axis (tennis-racket flip), grey = sphere (neutral).

---

The one-picture summary

Everything above collapses into one diagram: an arrow $\vec{\omega }$ makes momentum $\vec{L}$ through the tensor $\mathbf{I}$; torque changes $\vec{L}$; the frame-spin adds the gyroscopic curl; big $L$ gives slow, stiff precession; and energy leakage always parks the spin on the fat axis.

Recall Feynman: the whole walkthrough in plain words

Picture a spinning frisbee floating in space. The way it turns is one arrow ($\vec{\omega }$). But because a frisbee is wide, the momentum of its spin is a second arrow ($\vec{L}$) that can lean off to the side — a lopsided body throws its spin crookedly, and the little "hardness table" $\mathbf{I}$ is what does the leaning. There are three magic directions (principal axes) where the two arrows line up; we always do the bookkeeping there.

The law says: a twist changes the momentum arrow. But we're standing on the spinning frisbee, so even a frozen arrow looks like it's sweeping around us — that fake-but-real swirl is the $\vec{\omega }\times \vec{L}$ term, and gluing it to the honest law gives Euler's equation. That extra swirl is a free twist: it's what makes tops magically resist tipping. Push a fast-spinning craft sideways and it barely drifts — it just cones around slowly, because a huge momentum arrow is hard to turn.

Last trick: the craft slowly leaks energy (sloshing fuel, flexing panels) but can't leak momentum. Same momentum, less energy means it wants to spin about its widest axis, where energy is lowest. So a real satellite always ends up spinning like a frisbee, never like a pencil — and the ones that ignored this, like Explorer 1, tumbled.

Recall

What single arrow encodes both spin axis and spin rate? ::: The angular velocity $\vec{\omega }$ — direction is the axis, length is the rate. Why can $\vec{L}$ point in a different direction than $\vec{\omega }$? ::: Because $\mathbf{I}$ is a tensor: resistance to spin differs by direction, so a lopsided body throws momentum sideways. What does the $\vec{\omega }\times \vec{L}$ term in Euler's equation represent? ::: The apparent change in $\vec{L}$ caused by measuring it from the spinning body frame — the gyroscopic term. Why does bigger spin give stiffer pointing? ::: Precession rate $\Omega =\tau /L$; a larger $L=I{\omega }_s$ turns the same torque more slowly. Which axis is dissipation-stable and why? ::: The maximum-inertia axis, because $T=L^2/2I$ is minimized there at fixed $L$, and dissipation drives toward minimum energy.