1.5.6 · D5Rotational Mechanics

Question bank — Parallel axis theorem — I = I_CM + Md² — proof

1,546 words7 min readBack to topic

True or false — justify

The theorem works between any two parallel axes as long as they are the correct distance apart.
False. One axis must pass through the CM. The cross terms only vanish when the origin is the CM; between two off-CM axes they survive and the clean formula breaks.
The moment of inertia about the CM axis is the smallest possible among all axes parallel to it.
True. Since and , any parallel axis adds a non-negative amount. The minimum occurs at , i.e. the CM axis itself.
If you double the distance , the correction term doubles.
False. The correction is , so doubling quadruples it (). Distance enters squared — moving out is punished sharply.
The parallel axis theorem changes the value of depending on which outer axis you pick.
False. is a fixed property of the body about its own CM axis. Only the added term changes as you move the outer axis; never does.
Two different bodies with the same and same will have the same about any parallel axis away.
True. depends only on , , and — nothing about the detailed shape once those three match.
The theorem requires the body to be a symmetric or "nice" shape.
False. The derivation used only at the CM, which holds for any rigid body of any shape. No symmetry is assumed.
You can use the theorem to find about an axis that does not lie in the plane of a flat plate.
True. The proof ignored the coordinate entirely (distance to a -axis uses only ). The theorem is fully 3D; the axis just has to be parallel to the CM axis.

Spot the error

"A student computes of a rod about its end as using (the full length)."
The error is . The CM of a uniform rod is at its midpoint, so the end is only away. Correct: .
"To get about axis B (distance from CM) from about axis A (distance from CM), write ."
Wrong — you cannot connect two off-CM axes directly. Route through the CM: , , so , not .
"Since , I'll measure from the corner of the object to the axis."
is not measured from an arbitrary point. It is the perpendicular distance between the two parallel axes, equivalently CM-to-outer-axis. A corner is irrelevant unless the axis passes through it.
"For a tilted outer axis (not parallel to the CM axis), just plug the shortest distance into ."
The theorem is only valid for parallel axes. If the axes are tilted relative to each other, the whole geometry changes and does not apply at all.
"I subtracted from a rim-axis to get a value smaller than ."
Impossible. is the minimum, so no parallel axis gives less. If you got something smaller, you either mislabelled which term is or used the theorem between two non-CM axes.
"The disc's rim is because that's its known formula."
That is (central axis). The rim axis is away, so . Don't forget the .

Why questions

Why do the cross terms and vanish in the proof?
Because the origin was placed at the CM, and by the very definition of CM, , forcing (same for ). This is the sole reason the formula is clean.
Why must exactly ONE axis pass through the CM — not zero, not two?
You need one to make the cross terms vanish. Two parallel axes generally can't both pass through the CM (only one line through the CM has that fixed direction), and zero would leave the ugly cross terms in.
Why does the correction term look like a point mass at the CM?
Because is literally the moment of inertia of a single point of mass at perpendicular distance from the axis. The theorem splits inertia into "spread about CM" plus "whole mass as a point at the CM."
Why does moving the axis farther from the CM always increase , never decrease it?
The added term is , and both and , so it can only add. The only way to reduce toward its minimum is to move the axis toward the CM.
Why can we ignore the -coordinate of each mass element in the derivation?
The distance from a point to a -axis depends only on its and (it's ). Sliding a point along doesn't change its distance to a vertical axis, so never enters .
Why is the theorem so useful in practice?
Because for standard shapes (rod, disc, sphere) is tabulated once, and the theorem lets you reach any parallel axis by adding one easy term — no re-integration needed. This is why it underpins the Physical pendulum and Radius of gyration.

Edge cases

What does the theorem give when ?
. The "outer" axis coincides with the CM axis — the theorem gracefully reduces to the trivial identity, confirming the CM value is the floor.
For a point mass, is there a meaningful ?
A point mass has about any axis through itself (all its mass is at distance zero). So , recovering the point-mass formula directly — the theorem's building block.
Can ever be negative, making some parallel axis have negative ?
No. is a sum of masses times squares, so it is always . Adding keeps non-negative for every axis.
What happens to the theorem for a massless (idealized) framework, ?
The correction , so , which is itself for zero mass. Everything degenerates to zero inertia — consistent, since there's no mass to resist rotation.
If the CM lies outside the physical body (like a ring or an L-shape), does the theorem still hold?
Yes. The CM is a geometric point defined by ; it need not lie inside the material. The proof only used that mathematical property, so the theorem applies unchanged.
Does the theorem work in the limit where the outer axis is infinitely far away?
Formally as . Physically sensible: swinging the whole mass at an enormous radius demands enormous rotational inertia — the term dominates and becomes negligible.

Recall One-line takeaway

Every trap here reduces to two facts: (1) one axis must pass through the CM, and (2) is the perpendicular distance, entering squared. Guard those two and no version of this question can catch you.

Connections