You are on page 1of 40

PhysicsWOOT (2751)

AoPS Staff

Tuesday
Aug 24, 2021 - Apr 12, 2022
7:30 - 9:30 PM ET (4:30 - 6:30 PM PT)

Overview
Lesson 3 (Sep 28) Class Transcript - Rotational Motion
< Go back to the class overview page
Copyright © AoPS Incorporated. This page is copyrighted material. You can view and print this page for your own use, but you
cannot share the contents of this file with others.
Display all student messages • Show few student messages • Hide student messages
meichenlaub 2021-09-28 19:30:07
Lesson 3: Rotational Motion

meichenlaub 2021-09-28 19:30:11


Today we're going to study Newton's laws in the context of things spinning or turning. There are a bunch of tools for doing this -
angular momentum, moments of inertia, etc. To start off the day, we're going to begin using just Newton's laws, and as we work
through our first problem, we'll have the chance to develop some of those tools, then use them in later problems.

meichenlaub 2021-09-28 19:30:24


You are probably familiar with how ice skaters can spin faster and faster by pulling their limbs inward.

meichenlaub 2021-09-28 19:30:26

meichenlaub 2021-09-28 19:30:33


In our first problem, we'll use a simple model for an ice skater to see how this works.

meichenlaub 2021-09-28 19:30:39


Model an ice skater as a circle of mass M and radius R with a massless stick of length 2L through it, and a mass m on each end of
the stick, a distance L from the center of the circle. (The stick with masses can represent the limbs of the ice skater.) The masses
m can move along the stick without friction. Suppose the skater originally spins at angular frequency ω 0. The masses m move in to
the center of the circle, pulled by the skater. What is the new angular frequency of the skater?

meichenlaub 2021-09-28 19:30:43


meichenlaub 2021-09-28 19:31:06
I suspect that you can solve this problem quickly using angular momentum conservation, but our goal for this first problem isn't to
use all the tools of rotational motion, but to warm up by seeing where those tools come from. So I'd like to do this problem using
Newton's laws.

meichenlaub 2021-09-28 19:31:18


Let's break the model into two systems:

meichenlaub 2021-09-28 19:31:19


"the skater" is the circle of mass M along with the massless rod.

"the limbs" is the two point masses of mass m.

meichenlaub 2021-09-28 19:31:27


The idea is that as the skater pulls the limbs inward, the limbs somehow make the skater spin faster and faster. We'll try to work
out the details of how that works.

meichenlaub 2021-09-28 19:31:35


For the sake of simplicity, let's assume that the skater pulls the limbs in symmetrically, so the distance of each mass m from the
center of the circle is r.

meichenlaub 2021-09-28 19:31:41


What forces act on the mass m as it moves toward the center of the circle?

meichenlaub 2021-09-28 19:32:44


Let's work in an inertial reference frame, so there's no centrifugal force.

Pleaseletmewin 2021-09-28 19:32:59


tension in stick

ethanchen323 2021-09-28 19:32:59


centripetal force, force from skater

EthanTLee23 2021-09-28 19:32:59


Pulling force from arm

mathtiger6 2021-09-28 19:32:59


centripetal

DGA_Student_1 2021-09-28 19:32:59


normal force from the limb

Hiboy 2021-09-28 19:32:59


centripetal force

theshuklas 2021-09-28 19:32:59


centripetal force

meichenlaub 2021-09-28 19:33:10


There's a centripetal force exerted by the skater pulling m towards the circle. There's also a normal force from the stick on the
masses.

meichenlaub 2021-09-28 19:33:39


(Not many people noticed the normal force!)
meichenlaub 2021-09-28 19:33:46
By Newton's third law, there are also forces from m on the skater - a normal force and a centripetal force. Which is responsible for
speeding up the skater's rotation?

Zian2006 2021-09-28 19:34:33


Normal

EL2024 2021-09-28 19:34:33


normal

HappiHippo 2021-09-28 19:34:33


normal force

DGA_Student_1 2021-09-28 19:34:33


the normal force, as it supplies a torque to the skater

meichenlaub 2021-09-28 19:34:38


The normal forces are the ones that speed the skater up. Using the language of rotational dynamics, these forces exert torques on
the skater. But we're going to do this without appealing to torques.

meichenlaub 2021-09-28 19:34:50


Newton's second law, at least as applied to the center of mass of the skater, isn't going to help us either. The center of mass has no
motion in spinning, and Newton's second law will only tell us that the net force on the skater is zero.

meichenlaub 2021-09-28 19:35:08


So if we want to figure out how much the skater speeds up and we're not using Newton's second law and not using special rotation
tools, what approach do we have left?

DGA_Student_1 2021-09-28 19:35:49


energy?

EthanTLee23 2021-09-28 19:35:49


Energy conservation?

Hershey806 2021-09-28 19:35:49


energy

EL2024 2021-09-28 19:35:49


energy?

meichenlaub 2021-09-28 19:35:55


We can try thinking about energy. Specifically, the change in the rate of change of the rotational kinetic energy of the skater is
equal to the power delivered to the skater by the limbs,

meichenlaub 2021-09-28 19:36:04


Ṫ = P = 2 F ⋅ →
v.

meichenlaub 2021-09-28 19:36:12


→ →
where F is the force from a limb on the skater and v is the velocity of the point on the skater that contacts the limb.

meichenlaub 2021-09-28 19:36:24


Let's start with the kinetic energy. What's Ṫ in terms of M, R, ω, and ω̇ ?

meichenlaub 2021-09-28 19:36:53


T is the kinetic energy.

mathtiger6 2021-09-28 19:38:52


MR^2\omega \omega\dot

mathtiger6 2021-09-28 19:38:52


MR 2ωω̇
DGA_Student_1 2021-09-28 19:38:52
2
sorry, I meant Ṫ = MR ωω̇

EthanTLee23 2021-09-28 19:38:52


MR 2ww ′

meichenlaub 2021-09-28 19:39:01


The kinetic energy is

1 1
T= Mv 2 = MR 2ω 2.
2 2

meichenlaub 2021-09-28 19:39:07


Taking the time derivative,

meichenlaub 2021-09-28 19:39:10

Ṫ = MR 2ωω̇.

meichenlaub 2021-09-28 19:39:21


Next we need to take a look at the power. Let's use polar coordinates and call r̂ the direction along the stick and θ̂ the direction
perpendicular to the stick.

meichenlaub 2021-09-28 19:39:23

meichenlaub 2021-09-28 19:40:14



What is v for a point on the skater?

meichenlaub 2021-09-28 19:41:36


(In terms of R, ω, r̂, and θ̂.)

DGA_Student_1 2021-09-28 19:42:17



v = Rω θ̂

mathtiger6 2021-09-28 19:42:17

Rωω̂
Zian2006 2021-09-28 19:42:17
Rωθ̂

QMech 2021-09-28 19:42:17


RωΘ̂

mathtiger6 2021-09-28 19:42:17

Rωθ̂

EL2024 2021-09-28 19:42:17


v=Rw(theta hat)

meichenlaub 2021-09-28 19:42:23


It's the velocity for rotational motion,


v = ωR θ̂.

meichenlaub 2021-09-28 19:42:35



So if we want F ⋅ →
v , we'll need to find F θ, the θ component of the force on the skater. This confirms what we said earlier, that it's
the normal force that matters for speeding the skater up, not the centripetal force.

meichenlaub 2021-09-28 19:42:55


Of course, from Newton's second and third laws, the force on the skater is


F = − m→
a,

where →
a is the acceleration of a limb.

meichenlaub 2021-09-28 19:43:07


Does anyone see what a θ, the acceleration of a limb in the θ direction, is?

QMech 2021-09-28 19:44:38


Rω̇

EL2024 2021-09-28 19:44:38


MRw'

meichenlaub 2021-09-28 19:44:46


Hard question, but this is a good place to start.

meichenlaub 2021-09-28 19:44:50


We saw in the handout that for rotational motion, the acceleration is

a = − ω 2rr̂ + rω̇ θ̂.


So we might guess

a θ = rω̇.

meichenlaub 2021-09-28 19:45:03


Just as a trial, let's see what happens if we say that's the acceleration. Then the force we're looking for would be


F ⋅ θ̂ = − mrω̇.

meichenlaub 2021-09-28 19:45:18


If we go back to our equation relating the time derivative of kinetic energy and power, plug in everything we've found so far, and
simplify, what do you get?
meichenlaub 2021-09-28 19:47:37

(There are three substitutions to make. One for Ṫ, one for F θ, and one for v θ.)

meichenlaub 2021-09-28 19:48:18


Also, keep track of the difference between M and m.

meichenlaub 2021-09-28 19:48:32


And the difference between R and r.

meichenlaub 2021-09-28 19:49:31


Ah, okay. I see what's going on.

meichenlaub 2021-09-28 19:49:52


People are substituting in v = ωR, which makes sense. We said that's the speed of the skater.

meichenlaub 2021-09-28 19:50:11


But that's the speed of a point on the edge of the skater. We actually want the speed of the point on the stick where the force is
applied.

PhysKid11 2021-09-28 19:51:23


-2mr^2 = MR^2

EthanTLee23 2021-09-28 19:51:23


$MR^2 = -2r^2 m

meichenlaub 2021-09-28 19:51:29


That's what I got!

meichenlaub 2021-09-28 19:51:34


That equation was


Ṫ = P = 2 F ⋅ →
v.

Substituting in all our results,

MR 2ωω̇ = − 2mrω̇ ⋅ rω.

And this simplifies to

MR 2 = − 2mr 2.

meichenlaub 2021-09-28 19:51:51


What do you think of that?

DGA_Student_1 2021-09-28 19:52:58


its definitely wrong-- one side is positive and the other is negative

mathtiger6 2021-09-28 19:52:58


i think the negative is strange

QMech 2021-09-28 19:52:58


mass can't be negative

mathisawesome01 2021-09-28 19:52:58


impossible

Zian2006 2021-09-28 19:52:58


We never assumed anything about M, R, r, or m

Yufanwang 2021-09-28 19:52:58


wait M and m have opposite signs then???
meichenlaub 2021-09-28 19:53:03
That's clearly wrong! The skater should be able to have any mass and radius, and this equation is useless for determining how fast
the skater spins, since it doesn't include ω.

meichenlaub 2021-09-28 19:53:26


Evidently, something went wrong in our derivation. That's okay. When you're solving problems at this level, you're going to forget
terms, make algebra errors, and make mistakes. Even the kid who wins IPhO in any given year is going to make mistakes during
the test. So it's a really important skill to learn to realize something is wrong and not get flustered. Instead, we just go back, find
the mistake and try again.

meichenlaub 2021-09-28 19:53:40


Often, what goes wrong is that we have some result we've seen before and we apply it when maybe it isn't totally appropriate. Does
anyone see what we did wrong here?

meichenlaub 2021-09-28 19:54:19


Hint: it's the step where we didn't work out the details, and just assumed we knew it based on something we remembered from the
handout.

DGA_Student_1 2021-09-28 19:54:52


We missed a term in the tangential acceleration of the limb. It should be a θ = rθ̈ + 2ṙθ̇

TRay06 2021-09-28 19:54:52


acceleration is not simply rw'

Hershey806 2021-09-28 19:54:52


mathisawesome01 2021-09-28 19:54:52


the object is being pulled inward, so acceleration is not given by the formula we used

Yufanwang 2021-09-28 19:54:52


We guessed that a θ = rω̇

EthanTLee23 2021-09-28 19:54:52


The tangential acceleration?

rohitk14 2021-09-28 19:54:55


the acceleration step?

meichenlaub 2021-09-28 19:55:02


The problem was when we said

a θ = rω̇.

meichenlaub 2021-09-28 19:55:06


In the handout, we derived that result under the assumption that r is constant, but it's not!

meichenlaub 2021-09-28 19:55:13


No problem, let's just try again.

meichenlaub 2021-09-28 19:55:19


The velocity in polar coordinates is


v = ṙr̂ + ωr θ̂.

meichenlaub 2021-09-28 19:55:30


Okay, so taking the time derivative of the θ part, what do we get for a θ ?

EL2024 2021-09-28 19:56:48


w'r+wr'

PhysKid11 2021-09-28 19:56:48


omega dot * r + omega*r dot?

mathisawesome01 2021-09-28 19:56:48


a θ = ωṙ + ω̇r

EthanTLee23 2021-09-28 19:56:48


(w ′ r + wr ′ )θ̂

meichenlaub 2021-09-28 19:56:55


We can try again with

d
aθ = (rω) = ṙω + ω̇r.
dt

meichenlaub 2021-09-28 19:57:02


Now go back and do the calculation again - set the time derivative of kinetic energy equal to the power. You can say "done" when
you've simplified that as much as is obvious.

olive0827 2021-09-28 19:59:08


done

Yufanwang 2021-09-28 19:59:08


done

EthanTLee23 2021-09-28 19:59:08


Done

PhysKid11 2021-09-28 19:59:08


done

EL2024 2021-09-28 19:59:08


done?

mathtiger6 2021-09-28 19:59:08


2
MR ω̇ = − 2r(ṙω + ω̇r)

meichenlaub 2021-09-28 19:59:13


Just as before, we have

Ṫ = − 2m(ωr)a θ.

Putting in what we know,

MR 2ωω̇ = − 2mωr(rω̇ + ṙω).

This simplifies to

MR 2ω̇ = − 2m(r 2ω̇ + rṙω).

meichenlaub 2021-09-28 19:59:23


The left hand side is clearly the time derivative of MR 2ω. Is the right hand side a total time derivative as well?

DGA_Student_1 2021-09-28 20:01:07


No

AOPS81619 2021-09-28 20:01:07


It seems very close to the derivative of r 2ω

mathtiger6 2021-09-28 20:01:07


it is not the time derivative of r 2ω
EL2024 2021-09-28 20:01:07
I would think -mr^2w but doesnt that come out to be -m(2r'rw+r^2w')

meichenlaub 2021-09-28 20:01:13


No, it's not. It's really close. It looks almost like a time derivative of r 2ω is in there, but the coefficients don't work out quite right.
We were hoping for a simple result, and it came out looking like a bit of a mess. Perhaps there was another mistake! Does anyone
see what it could have been?

mathisawesome01 2021-09-28 20:02:41


d
did not account for r̂
dt

DGA_Student_1 2021-09-28 20:02:41


we forgot that r̂ and θ̂ vary with time

mathtiger6 2021-09-28 20:02:41


r-component

meichenlaub 2021-09-28 20:02:51


We got a θ wrong yet again! Yeah, it's annoying, but it happens. And we're getting really close now.

meichenlaub 2021-09-28 20:02:55


Let's go back to the velocity again,


v = ṙr̂ + ωr θ̂.

meichenlaub 2021-09-28 20:02:58


What did we forget about when we took the time derivative of that and found the θ component?

Yufanwang 2021-09-28 20:03:54


r cap and theta cap aren't constant either

EL2024 2021-09-28 20:03:54


r'(r hat)

Zian2006 2021-09-28 20:03:54


The r̂ component

AOPS81619 2021-09-28 20:03:54


the ṙr̂

Hiboy 2021-09-28 20:03:54


the rhat component

mathisawesome01 2021-09-28 20:03:54


d

dt

Yufanwang 2021-09-28 20:03:54


r̂ and θ̂ aren't constant

DGA_Student_1 2021-09-28 20:03:54


we need to consider the time rates of change of r̂ and θ̂

meichenlaub 2021-09-28 20:03:59


We forgot that there's a time derivative of r̂ ! Specifically,

d
r̂ = ω θ̂.
dt

meichenlaub 2021-09-28 20:04:03


Okay, with that in place, one last time, what's a θ ?
meichenlaub 2021-09-28 20:04:28
(We don't need to worry about the time derivative of θ̂ because that is purely in the r̂ direction.)

DGA_Student_1 2021-09-28 20:05:42


a θ = rθ̈ + 2ṙθ̇

EL2024 2021-09-28 20:05:42


2r'w+w'r

mathtiger6 2021-09-28 20:05:42


a θ = ṙω + ṙω + ω̇r

QMech 2021-09-28 20:05:42


a_\theta = r\dot{\omega} + 2\dot{r}\omega}

EthanTLee23 2021-09-28 20:05:42


2r ′ w + w ′ r

meichenlaub 2021-09-28 20:05:56


Now we have

a θ = ṙω + ω̇r + ωṙ.

The first term there is from the derivative of r̂, and the second two terms are those we found before. So

a θ = 2ṙω + rω̇.

meichenlaub 2021-09-28 20:06:05


All right, one more time. If you put this new a θ into the energy equation, do you now get a total time derivative on the right hand
side?

mathisawesome01 2021-09-28 20:06:30


yes finally

PhysKid11 2021-09-28 20:06:30


yes

taniaban 2021-09-28 20:06:30


yes

Zian2006 2021-09-28 20:06:30


Yes

mathtiger6 2021-09-28 20:06:30


2
yes, of r ω

meichenlaub 2021-09-28 20:06:34


Yes indeed! We now have

MR 2ω̇ = − 2m(r 2ω̇ + 2rṙω),

and we can write that as

d d
dt (
MR 2ω =
dt) (
− 2mr 2ω . )
meichenlaub 2021-09-28 20:06:43
Those two time derivatives are opposites. So what must be a conserved quantity? And what's the name of the quantity?

QMech 2021-09-28 20:07:27


angular momentum

Hershey806 2021-09-28 20:07:27


angular momentum

Zian2006 2021-09-28 20:07:27


Angular momentum

Hiboy 2021-09-28 20:07:27


agnular momentum

DGA_Student_1 2021-09-28 20:07:27


the quantity MR 2ω + 2mr 2ω must be conserved -- it is angular momentum

EL2024 2021-09-28 20:07:27


mr^2w, angular momentum

meichenlaub 2021-09-28 20:07:30


The sum of those two terms is conserved.

d
dt (
MR 2ω + 2mr 2ω = 0.)
meichenlaub 2021-09-28 20:07:34
The name of this quantity is of course the angular momentum. At last we're introduced to it!

L = (MR 2 + 2mr 2)ω.

And the quantity MR 2 + 2mr 2 is called the moment of inertia, I.

meichenlaub 2021-09-28 20:07:40


And finally, what's the answer to the original question? If the ice skater originally has r = L, ω = ω 0 and in the end has r = 0, what is
the final ω f ?

taniaban 2021-09-28 20:09:07


MR 2 + 2mr 2
w0 ∗
MR 2

mathtiger6 2021-09-28 20:09:07


2mL 2
ω0 + ω0
MR 2

Hershey806 2021-09-28 20:09:07


ω ( 2mL 2 + MR 2 )
MR 2

DGA_Student_1 2021-09-28 20:09:11


MR 2 + 2mL 2
ωf = ω0
MR 2

meichenlaub 2021-09-28 20:09:15


The angular momentum is

L = (MR 2 + 2mL 2)ω 0 = MR 2ω f.

Solving for the final angular velocity,

MR 2 + 2mL 2
ωf = ω0 .
MR 2

meichenlaub 2021-09-28 20:09:22


There are a couple of lessons from this problem. One is that we expect coherence in physics. Sure, we learn lots of tools, like
angular momentum conservation, torque and the rotational analog of Newton's second law, the work-energy theorem, etc. But we
should be able to understand a situation via different tools, and see all of them as connected.

meichenlaub 2021-09-28 20:09:34


I don't recommend solving this problem on the USAPhO via work-energy, since angular momentum conservation is clearly the way
to go. But hopefully you have a better appreciation of angular momentum conservation now, and understand some of the tools of
polar coordinates a bit better.

meichenlaub 2021-09-28 20:09:44


The second is that you will indeed make mistakes sometimes, especially when using tools you're less familiar with. That's okay;
just catch the mistakes, go back, think about what assumptions you might be making that aren't justified, and fix up your approach.

meichenlaub 2021-09-28 20:10:00


Finally, let's look for one more connection between concepts. The correct acceleration in the θ direction in this problem involved
the term

2ṙω, which we could also write as 2v rω. Where have you seen an acceleration with that mathematical form before?

QMech 2021-09-28 20:10:38


coriolis acceleration

Hershey806 2021-09-28 20:10:38


coriolis

mathtiger6 2021-09-28 20:10:38


coriolis force

PhysKid11 2021-09-28 20:10:38


coriolis force

DGA_Student_1 2021-09-28 20:10:38


coriolis acceleration!

EthanTLee23 2021-09-28 20:10:38


Coriolis

Zian2006 2021-09-28 20:10:38


Coriolis force

meichenlaub 2021-09-28 20:10:42


That's the mathematical form for the Coriolis acceleration. This suggests one more way to do this problem - go into a rotating
reference frame, rotating along with the skater. Then as the skater pulls the limbs in, there will be a Coriolis force on the limbs, and
this Coriolis force will spin the skater up to faster speeds.

meichenlaub 2021-09-28 20:10:59


Sorry about the double-ruse in this problem, where I introduced a mistake, then pretended to fix it with another mistake! I'll
promise not to do that again in PhysicsWOOT, but it is realistic to how people solve problems, so I thought we should go through it
together at least once.

meichenlaub 2021-09-28 20:11:12


Now that we've used Newton's laws and work-energy to get introduced to the tools of rotational motion, we'll be free to use all
those tools to solve problems for the rest of the class.

meichenlaub 2021-09-28 20:11:20


This next problem is from USAPhO 2010.

meichenlaub 2021-09-28 20:11:23


A thin plank of mass M and length L rotates about a pivot at its center. A block of mass m ≪ M slides on the top of the plank. The
system moves without friction. Initially, the plank makes an angle θ 0 with the horizontal, the block is at the upper end of the plank,
and the system is at rest. Throughout the problem you may assume that θ ≪ 1, and that the physical dimensions of the block are
much, much smaller than the length of the plank.

meichenlaub 2021-09-28 20:11:43


Let x be the displacement of the block along the plank, as measured from the pivot, and let θ be the angle between the plank and
the horizontal. You may assume that centripetal acceleration of the block is negligible compared with the linear acceleration of the
block up and down the plank.

meichenlaub 2021-09-28 20:11:45

meichenlaub 2021-09-28 20:12:06


For a certain value of θ 0, x = kθ throughout the motion, where k is a constant. What is this value of θ 0 ? Express your answer in
terms of M, m, and any fundamental constants you require.

meichenlaub 2021-09-28 20:12:23


This problem statement talks quite a bit about approximations, and many olympiad problems require you to make the right
approximations. We'll take this problem as a good chance to practice making appropriate approximations without, for example, just
approximating everything away to zero.

meichenlaub 2021-09-28 20:12:39


Before we get into all that, though, let's get a rough understanding of what happens in the scenario the question is asking us about.
What happens to the plank when x gets to the middle of the plank?

mathtiger6 2021-09-28 20:13:30


harmonic motion

chyc 2021-09-28 20:13:30


it'll be level

Yufanwang 2021-09-28 20:13:30


it stops accelerating

QMech 2021-09-28 20:13:30


it begins to tip in the other direction

EthanTLee23 2021-09-28 20:13:30


It becomes horizontal

Zian2006 2021-09-28 20:13:30


No angular acceleration

meichenlaub 2021-09-28 20:13:34


From the equation x = kθ, we see that when x = 0, we also have θ = 0. In other words, the plank becomes flat just as the box gets
to the middle of the plank. Then what happens?

mathisawesome01 2021-09-28 20:14:24


goes the other way, symmetric

EL2024 2021-09-28 20:14:24


it starts decelerating when the box goes to the other side

QMech 2021-09-28 20:14:24


tips in other direction

EthanTLee23 2021-09-28 20:14:24


The block moves to the another side, and θ increases in the other direction

chyc 2021-09-28 20:14:24


the plank tilts up and the box slides up the plank
HappiHippo 2021-09-28 20:14:24
the plank tilts up from the angular momentum of the initial movement

Hiboy 2021-09-28 20:14:24


it tips the other way

taniaban 2021-09-28 20:14:24


it keeps going and the plank keeps tilting counterclockwise

mathisawesome01 2021-09-28 20:14:24


symmetrically goes the other way, so we only have to worry about the block stopping at the middle

meichenlaub 2021-09-28 20:14:30


Then the box's inertia carries it to the right, and the plank's rotational inertia means it starts tipping upward to the right. The box
slides all the way to the right hand side of the plank, and when it gets there, the plank stops rising and starts falling. The box slides
back to the left, reaching the center of the plank just as the plank becomes flat, etc.

meichenlaub 2021-09-28 20:14:47


Okay, that's the motion we're looking for. Let's start analyzing what happens with the box.

meichenlaub 2021-09-28 20:14:53


What's the acceleration of the box?

EL2024 2021-09-28 20:15:47


gsin(theta)

Hiboy 2021-09-28 20:15:47


sin(theta)*m*g

taniaban 2021-09-28 20:15:47


g ∗ sinθ

EthanTLee23 2021-09-28 20:15:47


gsinθ

PhysKid11 2021-09-28 20:15:47


gsin(theta)

QMech 2021-09-28 20:15:47


gsinθ

TRay06 2021-09-28 20:15:47


g*sin(theta)

Zian2006 2021-09-28 20:15:47


gsin(θ) ≈ gθ

meichenlaub 2021-09-28 20:15:52


The box is on an incline, so the usual rule for a box on a frictionless incline gives us

d 2x
a= = gsinθ.
dt 2

meichenlaub 2021-09-28 20:16:04


That is not the exact acceleration of the box, of course. Even if the box were not moving relative to the plank, it would have some
acceleration because the plank is rotating.

meichenlaub 2021-09-28 20:16:12


We're not including that acceleration because the problem says not to! The problem says that the centripetal acceleration, which is
the acceleration associated with the rotation of the plank, is negligible compared to the acceleration down the plank that we just
found. So sometimes we make an approximation just because the problem says so.

meichenlaub 2021-09-28 20:16:38


Also, the problem says θ ≪ 1, which means
sinθ ≈ θ

and

meichenlaub 2021-09-28 20:16:40

a ≈ gθ.

mathisawesome01 2021-09-28 20:16:46


We know that sinθ = sin(kx) ≈ kx because rotation is negligible for acceleration

meichenlaub 2021-09-28 20:16:51


Again, the problem simply tells us that we can make this approximation.

meichenlaub 2021-09-28 20:16:56


That's the acceleration of the box. What do you get for the angular acceleration of the plank?

meichenlaub 2021-09-28 20:17:21


1
(The moment of inertia is ML 2. )
12

meichenlaub 2021-09-28 20:18:56


Your answer should definitely depend on x !

Zian2006 2021-09-28 20:20:22


12mgxcos ( θ )
ML 2

EthanTLee23 2021-09-28 20:20:22


12mgx
ML 2

taniaban 2021-09-28 20:20:22

12mgxcosθ
ML 2

Hershey806 2021-09-28 20:20:22


12mgx
whoops, cosθ
ML 2

EL2024 2021-09-28 20:20:22


12mgxcos(theta)/(ML^2) but with approximations 12mgx/(ML^2) ?

mathisawesome01 2021-09-28 20:20:22


mgxcosθ
Iα = −
1
ML 2
12

meichenlaub 2021-09-28 20:20:27


To find the angular acceleration of the plank, we can use the rotational analog of Newton's second law,

τ = Iα.

meichenlaub 2021-09-28 20:20:50


The moment of inertia is

1
I= ML 2 + mx 2.
12

But we'll approximate this as simply


1
I= ML 2.
12

And the torque is

τ = mgxcosθ ≈ mgx.

meichenlaub 2021-09-28 20:21:00


So

1
mgx ≈ ML 2α.
12

or

meichenlaub 2021-09-28 20:21:08

12mgx
α≈ .
ML 2

meichenlaub 2021-09-28 20:21:20


So there were some more approximations in there. For example, we approximated cosθ ≈ 1. How do you know to use that instead of
θ2
cosθ ≈ 1 − ?
2

QMech 2021-09-28 20:22:09


leading order approx

chyc 2021-09-28 20:22:09


it's too complicated, and we don't need a second-order equation in theta for our expression

AOPS81619 2021-09-28 20:22:09


if everything simplifies to become 0 or something then add more terms

DGA_Student_1 2021-09-28 20:22:09


the problem would be too complicated otherwise

EL2024 2021-09-28 20:22:09


because it's simpler?

meichenlaub 2021-09-28 20:22:16


Generally we want to keep the simplest approximation that doesn't just tell us that the motion of the system is zero. If we keep
1
cosθ ≈ 1 − θ 2, we'd have
2

1 2
12mgx(1 − θ )
2
α≈ .
ML 2

meichenlaub 2021-09-28 20:22:28


This results in a needlessly-complicated equation that doesn't look like it's going to satisfy x = kθ, as the problem instructs us.
We're keeping too much detail.

meichenlaub 2021-09-28 20:22:51


On the other hand, suppose we were finding the potential energy of a pendulum, which is

U = − mgLcosθ.

There, if we approximate cosθ ≈ 1 the potential energy becomes


U = − mgL.

The potential energy wouldn't even depend on θ any more, so there's no way we'd finding the oscillatory motion of the pendulum
come out of such an analysis. So there's no single rule about how deep into an approximation we need to go; that changes based
on the situation. We go deep enough to get non-trivial behavior.

meichenlaub 2021-09-28 20:23:38


(Sorry, there is a bug in our new classroom so that math like "\[I = " gets rendered italics instead of as an equation, etc. I'm fixing
them as they show up.)

meichenlaub 2021-09-28 20:23:47


Another approximation we made is worth looking at. In finding the moment of inertia of the plank, we ignored the mass m of the
1
box because m is so small. We simply used I ≈ ML 2 for the moment of inertia, setting m = 0. But when we found the torque, we
12
didn't set m = 0. Why not?

HappiHippo 2021-09-28 20:24:38


because the torque is applied by m

EL2024 2021-09-28 20:24:38


it would mean theirs 0 torque

EthanTLee23 2021-09-28 20:24:38


If we did, there would be no torque

Dhillonr25 2021-09-28 20:24:38


because then we wouldn't have any torque and the problem would be trivial

Hiboy 2021-09-28 20:24:38


because torque is more dependent on m

taniaban 2021-09-28 20:24:38


that would have led to zero angular acceleration which is obviously unhelpful

aj1745 2021-09-28 20:24:38


The mass is applying the torque?

theallpro1 2021-09-28 20:24:38


then our torque would just be 0

DGA_Student_1 2021-09-28 20:24:38


Because then the whole torque would be zero, and we would get a trivial motion.

QMech 2021-09-28 20:24:43


if m = 0 we would get 0 angular acceleration which wouldn't help us progress

meichenlaub 2021-09-28 20:24:48


If we set m = 0 in the torque, we'd find

τ = 0.

In other words, approximating m = 0 doesn't throw out interesting dynamics when we're finding the moment of inertia. The
1
equation I = ML 2 + x 2m is only changed by a very small percentage by approximating m = 0. On the other hand, the equation
2
τ ≈ mxg is changed drastically by substituting in m = 0.

meichenlaub 2021-09-28 20:25:00


So again there's no one rule for what approximation to make. We need to be judicious, and choose the approximation that makes
the system workable, but not trivial.

meichenlaub 2021-09-28 20:25:13


Now we have both a, the acceleration of the box, and α, the angular acceleration of the plank. What can we use to relate them?
meichenlaub 2021-09-28 20:26:25
We definitely cannot just set a = αr. a is the linear acceleration of the box. α is the angular acceleration of the board. They aren't
related so simply kinematically.

meichenlaub 2021-09-28 20:26:36


Besides, what the heck is r ?

Zian2006 2021-09-28 20:26:55


x = kθ

QMech 2021-09-28 20:26:55


x = kθ

DGA_Student_1 2021-09-28 20:26:55


a = kα

mathtiger6 2021-09-28 20:26:55


plug in x = kθ or a = kθ̈ into the equations

EthanTLee23 2021-09-28 20:26:55


Second derivative of x = kθ

meichenlaub 2021-09-28 20:26:59


We can use the information in the problem,

x = kθ.

Taking two time derivatives,

a = kα.

meichenlaub 2021-09-28 20:27:05


What do you get for k ?

meichenlaub 2021-09-28 20:29:24


We can't just leave θ and x in there. We want k to be a constant.

mathtiger6 2021-09-28 20:30:01


ML 2

12m

Zian2006 2021-09-28 20:30:01


ML 2
k=
12m

chyc 2021-09-28 20:30:01


ML 2
k=
12m

EthanTLee23 2021-09-28 20:30:01


2


ML
12m

meichenlaub 2021-09-28 20:30:08


Putting in our approximations for a and α, we have

12mgx
gθ = k .
ML 2
Solving for k, we get

M
k=L
√ 12m
.

meichenlaub 2021-09-28 20:30:24


And what is θ 0 ?

theallpro1 2021-09-28 20:32:12


sqrt(3m/M)

Hershey806 2021-09-28 20:32:12


1 12m
2

√ M

chyc 2021-09-28 20:32:12


3m
√ M

meichenlaub 2021-09-28 20:32:16


We have

x0
θ0 =
k
L/2
=
M
L
√ 12m

3m
=
√ M
.

meichenlaub 2021-09-28 20:32:44


(Remember that L is the length of the entire plank, so the block is initially L / 2 from the center.)

meichenlaub 2021-09-28 20:32:50


Okay, does this result gel with the question's directive to assume θ ≪ 1 ?

HappiHippo 2021-09-28 20:33:14


yes

mathisawesome01 2021-09-28 20:33:14


yes

mathtiger6 2021-09-28 20:33:14


yes, since m ≪ M

HappiHippo 2021-09-28 20:33:14


because m<<M

DGA_Student_1 2021-09-28 20:33:14


Yes, as m << M

meichenlaub 2021-09-28 20:33:18


3m
Yes, the problem tells us m ≪ M, so
√ M
≪ 1, so all the angles are much less than 1. This at least partially justifies our

assumptions.
meichenlaub 2021-09-28 20:38:12
There are a couple more parts to that problem, but let's move on now to something more conceptual.

meichenlaub 2021-09-28 20:38:19


An astronaut on the space station has a cylinder significantly longer than the diameter of the endcaps. The cylinder is mostly filled
with water, although the water can slosh around inside. The astronaut spins the cylinder along its symmetry axis. What happens to
the cylinder over time? Ignore any potential outside forces on the cylinder.

meichenlaub 2021-09-28 20:38:44


This is a weird-looking question! It's going to be impossible to model the precise dynamics of water sloshing around.

meichenlaub 2021-09-28 20:39:01


But suppose water sloshes around inside the cylinder. What's some important physical quantity that changes because of this, and
does it go up or down?

Zian2006 2021-09-28 20:40:11


Energy; it always try to go down

DGA_Student_1 2021-09-28 20:40:11


energy

Dhillonr25 2021-09-28 20:40:11


moment of inertia

Yufanwang 2021-09-28 20:40:11


moment of inertia goes up

Shadow_Heist 2021-09-28 20:40:11


the moment of inertia changes, and it goes up.

A-Aban 2021-09-28 20:40:11


would moment of inertia increase?

meichenlaub 2021-09-28 20:40:18


The momentum can't change because there are no external forces. Likewise, the angular momentum can't change because if there
are no external forces, there are no external torques. But as the water sloshes, it dissipates mechanical kinetic energy. So the
kinetic energy of the cylinder decreases with time.

meichenlaub 2021-09-28 20:40:35


If the cylinder's kinetic energy goes down, we might think it simply slows down its spinning, and maybe even stops. Why isn't that a
possibility?

Hershey806 2021-09-28 20:41:14


no external forces/torques

Zian2006 2021-09-28 20:41:14


Angular momentum conservation

EthanTLee23 2021-09-28 20:41:14


No torque to stop the spinning

DGA_Student_1 2021-09-28 20:41:14


angular momentum conservation

HappiHippo 2021-09-28 20:41:14


then the angular momentum of the cylinder would have disappeared into nowhere

mathtiger6 2021-09-28 20:41:14


no outside force

meichenlaub 2021-09-28 20:41:18


The cylinder can't simply stop spinning because that would require it to reduce its angular momentum. The angular momentum of
the cylinder is conserved.

meichenlaub 2021-09-28 20:41:31


So if the angular momentum of the cylinder is conserved, what has to change for it to lose kinetic energy?

Zian2006 2021-09-28 20:42:07


Moment of inertia

taniaban 2021-09-28 20:42:07


moment of inertia

HappiHippo 2021-09-28 20:42:07


moment of inertia

Latnor 2021-09-28 20:42:07


I must increase

meichenlaub 2021-09-28 20:42:12


The kinetic energy of a rotating rigid body is

L2
T= .
2I

The cylinder's angular momentum L can't change, so to lose kinetic energy, it could increase its moment of inertia, I.

meichenlaub 2021-09-28 20:42:22


The cylinder is almost full, so the distribution of the water can't change much. So how can the cylinder increase its moment of
inertia?

meichenlaub 2021-09-28 20:43:04


The "mostly full" thing is trying to point out that there isn't much space for water to go to the outside. Yeah, I guess that happens a
tiny bit, but think of something else.

meichenlaub 2021-09-28 20:44:02


Like, ignore the distribution of water or where the water is in the cylinder. What else can you think of to change I ?

meichenlaub 2021-09-28 20:44:14


And the cylinder is rigid.

Zian2006 2021-09-28 20:45:04


Spins on a perpendicular axis

AOPS81619 2021-09-28 20:45:04


Does it start spinning along some other axis?

QMech 2021-09-28 20:45:04


axis of rotation?

chyc 2021-09-28 20:45:04


the axis of rotation shifts??

meichenlaub 2021-09-28 20:45:09


The cylinder can rotate about a different axis - in this case it rotates about the axis through its center, parallel to its endcaps. This
way, the mass is, on average, much further from the rotation axis than when the cylinder rotates about its symmetry axis. This
increases the moment of inertia and decreases the energy.

meichenlaub 2021-09-28 20:45:23


Here's a gif of this actually happening:

meichenlaub 2021-09-28 20:45:24


meichenlaub 2021-09-28 20:45:39
Great! Let's do another real USAPhO problem! This one is from 2019, and it illustrates some important points about rotational
motion that we haven't touched on yet.

meichenlaub 2021-09-28 20:45:49


A bead is placed on a horizontal rail, along which it can slide frictionlessly. It is attached to the end of a rigid, massless rod of
length R. A ball is attached at the other end. Both the bead and the ball have mass M. The system is initially stationary, with the
ball directly above the bead. The ball is then given an infinitesimal push, parallel to the rail.

meichenlaub 2021-09-28 20:45:52


meichenlaub 2021-09-28 20:46:17
Assume that the rod and ball are designed in such a way so that they can pass through the rail without hitting it. In other words, the
rail only constrains the motion of the bead. Two subsequent states of the system are shown below.

meichenlaub 2021-09-28 20:46:27

meichenlaub 2021-09-28 20:46:35


Derive an expression for the tension in the rod when it is horizontal, as shown in the top picture above.

meichenlaub 2021-09-28 20:46:49


Let's think just a little about what's going on to get oriented. When the rod is horizontal, what direction is the bead moving? What
direction is the ball moving? What direction is the center of mass moving?

AOPS81619 2021-09-28 20:48:51


The ball is moving down, the bead is stationary

Zian2006 2021-09-28 20:48:51


The bead is stationary, ball downwards, CM downwards
taniaban 2021-09-28 20:48:51
0, down, down

TRay06 2021-09-28 20:48:51


bead isn't moving, ball moves vertically down and center of mass of the whole system moves vertically down

meichenlaub 2021-09-28 20:48:56


The bead isn't allowed to move vertically, and with the rod horizontal, the bead is at its furthest to the left. Since its displacement is
at a maximum, its speed is zero; the bead isn't moving. By contrast, the ball is clearly moving downward because the rod is falling.
The center of mass's motion is the average of these two, so it also moves down, and does so half as fast as the ball.

meichenlaub 2021-09-28 20:49:24


We have a stick that starts out vertical and winds up horizontal. The problem asks for tension, but finding the velocities and
angular velocities involved seems like a good intermediate step. Given all that, what basic approach (what principle) does this
situation scream that we should use?

Hershey806 2021-09-28 20:50:35


energy

Zian2006 2021-09-28 20:50:35


Energy conservation

AOPS81619 2021-09-28 20:50:35


energy conservation

meichenlaub 2021-09-28 20:50:40


Yes, this is a conservation of energy question. We're given two states, and there's no mechanism for dissipating energy, so we
should be able to figure out the details of the second state using energy conservation from the first state.

meichenlaub 2021-09-28 20:51:23


For those saying "angular momentum", don't guess just based on what the lesson topic is! There's torque on this system from
gravity, for example.

meichenlaub 2021-09-28 20:51:28


So what's the total energy of the rod?

meichenlaub 2021-09-28 20:51:41


(plus ball)

EL2024 2021-09-28 20:52:14


mgr

mathtiger6 2021-09-28 20:52:14


MgR

AOPS81619 2021-09-28 20:52:14


MgR

EthanTLee23 2021-09-28 20:52:14


MgR

HappiHippo 2021-09-28 20:52:14


MgR

DGA_Student_1 2021-09-28 20:52:14


MgR

meichenlaub 2021-09-28 20:52:19


Given the starting state, it's

E = MgR.

meichenlaub 2021-09-28 20:52:31


When the rod is horizontal, there's only kinetic energy. How can we calculate the instantaneous motion of the system from this
kinetic energy?

meichenlaub 2021-09-28 20:54:50


Here, we ω, the angular velocity of the rod. So just finding the speed of the ball is not everything we're going for.

DGA_Student_1 2021-09-28 20:55:19


1
it's instantaneously a rotation about the bead, so the kinetic energy is just 2 MR 2ω 2

Hiboy 2021-09-28 20:55:19


kinetic energy = 1/2*I*omega^2

Dhillonr25 2021-09-28 20:55:19


we could try to calculate the velocity of the cm of the rod (or the bead at the end of it)

HappiHippo 2021-09-28 20:55:19


set it equal to the rotational kinetic energy

meichenlaub 2021-09-28 20:55:37


Well it looks like there are two methods being suggested. One is that we consider the rod to be rotating about the bead and use the
1
formula K = I endω 2 to find ω.
2

meichenlaub 2021-09-28 20:55:52


Another method is to think of the rod as having both some translation of the center of mass and some rotation about the center of
1 1
mass. Then we can use the formula K = (M tot)v 2 + I centerω 2 to find the kinetic energy.
2 2

meichenlaub 2021-09-28 20:56:12


If we use the second method, though, we'll have two variables: v is the speed of translation of the center of mass and ω the angular
velocity of rotation about the center of mass. How will we be able to relate those variables to each other?

Hershey806 2021-09-28 20:57:21


v=rw/2

EL2024 2021-09-28 20:57:21


2v=wR

mathtiger6 2021-09-28 20:57:21


v = ωR / 2

Hiboy 2021-09-28 20:57:21


v = w*(R/2)

taniaban 2021-09-28 20:57:21


v = wr / 2

meichenlaub 2021-09-28 20:57:26


1
The bead must stay on the track, so it can't have any vertical velocity. That means that, in terms of magnitudes, v = ωR.
2

meichenlaub 2021-09-28 20:57:31


Okay, choose one of those two methods and find ω. Let me know what you found, and which method you used.

DGA_Student_1 2021-09-28 20:59:27


2g
ω=
√ R
(I used the first method)

mathtiger6 2021-09-28 20:59:27


2g
ω=
√ R
, both

Zian2006 2021-09-28 20:59:27


2g

√ R

Dhillonr25 2021-09-28 20:59:27


sqrt(2g/R)

Hiboy 2021-09-28 20:59:27


sqrt(2*g/R)

EL2024 2021-09-28 20:59:27


sqrt(2g/R) the rod being rotated about the bead

meichenlaub 2021-09-28 20:59:49


It looks like we get the same answer via both methods, which is good!

meichenlaub 2021-09-28 20:59:51


1
If we find the rotation about the bead, the moment of inertia is I end = MR 2, and so K = (MR 2)ω 2 = MgR. Solving, we get
2
2g
ω=
√ R
.

meichenlaub 2021-09-28 21:00:04

If we use the method of translation and rotation about the center of mass, the moment of inertia is I center = 2M ()R 2
2
and we have

the equation

1 1 MR 2 2
MgR = ⋅ 2Mv 2 + ω .
2 2 2

1 2g
Then using v =
2
ωR and solving for ω, we again find ω =
√ R
.

meichenlaub 2021-09-28 21:00:37


Of course, if you found the speed of the ball, then found ω from kinematics, that would also be valid.

meichenlaub 2021-09-28 21:00:48


Now we need to find the tension in the rod. What methods are available for that?

meichenlaub 2021-09-28 21:01:46


Can you be a little more specific than "centripetal force"? Like, where are you going to put the axis of rotation?

HappiHippo 2021-09-28 21:02:47


centripetal force about the bead

Latnor 2021-09-28 21:02:47


the axis of rotation is the bead

DGA_Student_1 2021-09-28 21:02:47


at the bead

Hiboy 2021-09-28 21:02:47


centripetal force about the bead

QMech 2021-09-28 21:02:47


Centripetal force with AOR about bead

EL2024 2021-09-28 21:02:47


centripetal force with the bead as the axis, but i'm not sure if that would find tension since I think the motion isn't circular

Zian2006 2021-09-28 21:02:47


Centripetal force: ball about bead (axis)
meichenlaub 2021-09-28 21:02:58
Okay, everyone wants to make it rotation about the bead.

meichenlaub 2021-09-28 21:03:09


But there are really two methods: first, we can think of the rod as rotating about the bead. Then we can find the centripetal force on
the ball using the formula for centripetal force.

meichenlaub 2021-09-28 21:03:14


The other method is to think of the rod as rotating about its own center and find the centripetal force on either the bead or the ball.

meichenlaub 2021-09-28 21:03:28


Choose one of those two methods, use it to find the tension. Tell me what tension you found and what method you used.

mathisawesome01 2021-09-28 21:04:55


2mg, first method

mathtiger6 2021-09-28 21:04:55


2Mg, 1st

EthanTLee23 2021-09-28 21:04:55


T = 2mg using the bead as the axis of rotation

taniaban 2021-09-28 21:04:55


Mg; method 2

meichenlaub 2021-09-28 21:05:03


Whoa! We're getting different answers here! If we think of the rod as rotating about the bead, the centripetal force on the ball is
Mω 2R = 2Mg.

meichenlaub 2021-09-28 21:05:13


R
On the other hand, if we think of the rod as rotating about its own center, the centripetal force on the ball is Mω 2 because the ball
2
is only half as far away from the center of the rod as it is from the bead. Then we get that the tension is Mg.

meichenlaub 2021-09-28 21:05:32


Which of these two is the right method - rotation about the bead or rotation about the center? How do you know?

AOPS81619 2021-09-28 21:06:48


The second method because in the first method the bead is accelerating

mathtiger6 2021-09-28 21:06:48


probably the second, since the bead accelerates

ethanchen323 2021-09-28 21:06:48


the second, since the bead moves

EthanTLee23 2021-09-28 21:06:48


The one around the center, because tension pulls on both sides

meichenlaub 2021-09-28 21:06:53


We just tried to find the tension by thinking about the centripetal force on the ball. But what if we try thinking about the bead
instead? The tension is uniform throughout the rod, so whatever force is on the ball is on the bead as well (just in the opposite
direction). In the situation with rotation about the bead, we'd say that the bead's acceleration is zero, but there's a force of 2Mg on
it, which doesn't make sense! Only the calculation with rotation about the center of the rod leads to a consistent result.

meichenlaub 2021-09-28 21:07:19


The trick where we think of the motion of the rod as rotation about the bead is a very useful one. As you probably noticed when we
were doing the conservation of energy calculation, that viewpoint is easier and faster to use to find the angular velocity of the rod.

meichenlaub 2021-09-28 21:07:31


What we need to keep in mind, though, is that this is a trick that only works for velocities. When we go to trying to find forces, we're
interested in accelerations. The motion of the rod is instantaneously a rotation about the bead, but in this problem the bead can
move!
meichenlaub 2021-09-28 21:07:58
So suppose we go to find the acceleration of the ball. The way we do this is to find

v(t + dt) − v(t)


a(t) = .
dt

If we think of the rod as rotating about the bead, we'll get the right thing for v(t). But then we wait a small moment of time dt, the
bead itself has picked up some small velocity dv, so our calculation of v(t + dt), which ignored this new velocity of the bead, is
wrong.

meichenlaub 2021-09-28 21:08:18


In other words, the takeaway is that you can consider the motion of a rigid body to instantaneously be rotation about some point
and find the correct velocities, even if that point is moving. But don't use this trick to find accelerations or forces!

meichenlaub 2021-09-28 21:08:29


Instead, you can always think of the motion as translation of the center of mass plus rotation about the center of mass; go back to
that picture when you want to find a force. That means that in this case, the answer was Mg.

meichenlaub 2021-09-28 21:08:41


Derive an expression for the tension in the rod when the ball is directly below the bead, as shown in the lower diagram above.

meichenlaub 2021-09-28 21:09:03


This ought to go much the same way as the previous part - conservation of energy to the rescue!

meichenlaub 2021-09-28 21:09:10


Before we dive into calculations, though, let's get a feel for what's going on. In this situation, the rod is vertical. What direction is
the bead moving? What direction is the ball moving? What direction is the center of mass moving?

mathisawesome01 2021-09-28 21:10:20


bead moving to the right, ball moving to the left, center of mass stationary (but accelerating up)

taniaban 2021-09-28 21:10:20


right, left, 0

EL2024 2021-09-28 21:10:20


right,left,stationary?

EthanTLee23 2021-09-28 21:10:20


Bead is moving right, ball is moving left, com is instantaneously stationary

Hiboy 2021-09-28 21:10:20


right, left, not moving

mathtiger6 2021-09-28 21:10:20


left, right, none

meichenlaub 2021-09-28 21:10:25


The rod has fallen by rotating clockwise. Due to its rotational inertia, it's continuing to rotate clockwise at this point. Additionally,
the center of mass can't move left or right because there are no horizontal forces. The center of mass is at its lowest point,
meaning it has no y velocity, so the center of mass is stationary. With the center of mass stationary and the rod rotating clockwise,
the ball is moving left and the bead is moving right.

meichenlaub 2021-09-28 21:10:59


In this case, can we use the trick of thinking of the motion as instantaneous rotation about the bead?

DGA_Student_1 2021-09-28 21:11:50


No, because the bead has a nonzero velocity

mathtiger6 2021-09-28 21:11:50


no

Dhillonr25 2021-09-28 21:11:50


no
Zian2006 2021-09-28 21:11:50
No

AOPS81619 2021-09-28 21:11:50


No

meichenlaub 2021-09-28 21:11:56


No, we can't. In the case where the rod was horizontal, the bead was momentarily stationary. It had to be because it can only move
horizontally, and with the rod itself being horizontal, the bead was at the extreme left of its trajectory. Since it was at an extremum
of position, its velocity must have been zero.

meichenlaub 2021-09-28 21:12:11


In this case, though, the rod is swinging. The center of mass of the rod can't move left or right. That means that the ball is moving
to the left and the bead is swinging to the right. In fact, the bead has its maximum speed at this point.

meichenlaub 2021-09-28 21:12:32


So we'll have to think about the motion of the center of mass, which is stationary.

meichenlaub 2021-09-28 21:12:37


With that in mind, go ahead and calculate ω. What do you get?

mathisawesome01 2021-09-28 21:14:35


8g
ω=
√ R

Zian2006 2021-09-28 21:14:35

√8g / R
AOPS81619 2021-09-28 21:14:35
2g
ω=2
√ R

Hiboy 2021-09-28 21:14:35


sqrt(8*g/R)

taniaban 2021-09-28 21:14:38

√8g / R
meichenlaub 2021-09-28 21:14:52

1
The potential energy that's been released is 2MgR. The kinetic energy of the rod is Iω 2, with I = 2M
2
R 2
2 ()
. Setting these equal,

2MgR =
1
2
⋅ 2M
2()
R 2 2
ω .

and solving for ω, we get

8g
ω=
√ R
.

meichenlaub 2021-09-28 21:15:06


R
Last time, we found the centripetal force on the ball using F = Mω 2 . It we apply that same formula here, what will we get for the
2
tension in the rod?

taniaban 2021-09-28 21:15:27


4Mg

Nivek 2021-09-28 21:15:27


4Mg

Hershey806 2021-09-28 21:15:27


4Mg

EL2024 2021-09-28 21:15:27


4mg

meichenlaub 2021-09-28 21:15:30


We'd have

8g R
F=M = 4Mg.
R 2

meichenlaub 2021-09-28 21:15:33


Alas, this is wrong! However, unlike previous instances of being wrong, where we were wrong for only one reason, this time we've
managed to be wrong for two different reasons simultaneously! What reasons do you see that this calculation is incorrect?

mathisawesome01 2021-09-28 21:16:51


center of mass is accelerating

mathtiger6 2021-09-28 21:16:51


tension and gravity cause centripetal force

meichenlaub 2021-09-28 21:16:55


One is that we didn't draw a free body diagram and think about all the forces. Instead, we calculated the net force on the ball. But
that's not the same as the tension force, because there's gravity as well.

meichenlaub 2021-09-28 21:17:01

meichenlaub 2021-09-28 21:17:06


So that's part of it. But we're also wrong because we only found the acceleration relative to the center of mass. Even though the
center of mass isn't moving, that doesn't mean it's not accelerating!

meichenlaub 2021-09-28 21:17:12


Can you find the acceleration of the center of mass?

EL2024 2021-09-28 21:19:53


4g?

meichenlaub 2021-09-28 21:19:57


The center of mass's vertical position is

R
y= sinθ,
2
with θ the angle between the rod and the horizontal. So its acceleration is the second time derivative,

R 2
a= − ω sinθ.
2

8g
And plugging in sinθ = − 1 and ω 2 = , we find
R

a = 4g.

meichenlaub 2021-09-28 21:20:16


So the center of mass is accelerating upward at 4g. The ball is accelerating upward, relative to the center of mass, at 4g. Then
what's the tension in the rod?

Hershey806 2021-09-28 21:21:37


9Mg

meichenlaub 2021-09-28 21:21:41


The absolute acceleration of the ball is 8g upward. There's a downward force from gravity of Mg on it, so the total upward force,
which is the tension in the rod, must be 9Mg.

meichenlaub 2021-09-28 21:22:14


There is one last part to this problem which I'll let you try on your own.

meichenlaub 2021-09-28 21:22:18


By now, we have had a lot of practice with rotational dynamics, so we're going to finish up today with a little practice with
kinematics.

meichenlaub 2021-09-28 21:22:23


A wheel of radius R rolls without slipping. A small dot is on the very edge of the wheel. What is the distance moved by the point
during one complete revolution of the wheel?

meichenlaub 2021-09-28 21:22:30


Does anyone know what the curve this problem is talking about is called?

mathisawesome01 2021-09-28 21:22:45


cycloid

HappiHippo 2021-09-28 21:22:45


cycloid!

QMech 2021-09-28 21:22:45


cycloid

AOPS81619 2021-09-28 21:22:45


cycloid

DGA_Student_1 2021-09-28 21:22:45


a cycloid

dagurvich 2021-09-28 21:22:45


cycloid

EthanTLee23 2021-09-28 21:22:47


cycloid

Dhillonr25 2021-09-28 21:22:51


cycloid

meichenlaub 2021-09-28 21:22:56


The trajectory is called a cycloid, and it looks like this:

meichenlaub 2021-09-28 21:22:58


meichenlaub 2021-09-28 21:23:04
We're trying to find the length along one of those humps.

meichenlaub 2021-09-28 21:23:22


Of course, we could try to find the exact equation for a cycloid, i.e. find y(x). Then we could do an integral

2πR

s = ∫ x = 0 dy 2 + dx 2.

meichenlaub 2021-09-28 21:23:32


This is of course exactly what we're not going to do! Instead, let's use some knowledge about the kinematics of rolling without
slipping.

meichenlaub 2021-09-28 21:23:40


One way to view rolling without slipping is that the wheel's center is moving forward at speed v while the wheel is rotating about its
v
own center with angular velocity ω = . But what's another useful instantaneous description of how different parts of the wheel
r
are moving?

Hershey806 2021-09-28 21:24:20


bottom is instantaneously at rest, top moves with speed 2v

mathisawesome01 2021-09-28 21:24:20


oh rotating around bottom

mathtiger6 2021-09-28 21:24:20


they rotate around the contact point

DGA_Student_1 2021-09-28 21:24:20


instantaneous rotation about contact point with angular speed ω

AOPS81619 2021-09-28 21:24:20


The point in contact with the ground has 0 velocity

meichenlaub 2021-09-28 21:24:25


Instantaneously, we can think of the wheel as being pivoted about its point of contact with the ground, and rotating with angular
v
frequency ω = .
r

meichenlaub 2021-09-28 21:24:30


meichenlaub 2021-09-28 21:24:42
We should be able to use this to find the speed of a point anywhere on the rim of the wheel.

meichenlaub 2021-09-28 21:25:03


At that point, we'll have the speed of the point during its entire trip. Integrating v(t) over t should get us the distance traveled.

meichenlaub 2021-09-28 21:25:10


So let's take a point on the edge of the wheel as shown below in blue:

meichenlaub 2021-09-28 21:25:12

meichenlaub 2021-09-28 21:25:20


What distance do we want to find in order to get the speed of the point?

taniaban 2021-09-28 21:25:50


distance to the contact point

AOPS81619 2021-09-28 21:25:50


The distance between that point and the bottom

EL2024 2021-09-28 21:25:50


red to blue?

HappiHippo 2021-09-28 21:25:50


distance to the red point

DGA_Student_1 2021-09-28 21:25:50


the distance from the contact point to the blue point

EthanTLee23 2021-09-28 21:25:50


Distance from blue dot to red

dagurvich 2021-09-28 21:25:50


distance to contact point

Yufanwang 2021-09-28 21:25:50


the distance from the contact point (red dot)

meichenlaub 2021-09-28 21:25:59


We want the distance from the contact point with the ground, labeled d below.

meichenlaub 2021-09-28 21:26:13


meichenlaub 2021-09-28 21:26:30
And what is d in terms of R, the radius of the circle, and θ ?

EL2024 2021-09-28 21:27:08


2sin(theta/2)R

EthanTLee23 2021-09-28 21:27:08


θ
2Rsin
2

AOPS81619 2021-09-28 21:27:08

d= √2R (1 − cosθ)
2

Hiboy 2021-09-28 21:27:08


sin(theta)*2*r

taniaban 2021-09-28 21:27:08


2Rsinθ / 2

HappiHippo 2021-09-28 21:27:08

√2R 2 2
− 2R cosθ

Dhillonr25 2021-09-28 21:27:08


2Rsin(theta/2)

Hershey806 2021-09-28 21:27:08


θ
2Rsin( 2 )

Hiboy 2021-09-28 21:27:08


sin(theta/2)*2*r

meichenlaub 2021-09-28 21:27:12


Drawing another triangle

meichenlaub 2021-09-28 21:27:13


meichenlaub 2021-09-28 21:27:17
we see

d
2
= Rsin
θ
2()
.

or

d = 2Rsin
()
θ
2
.

meichenlaub 2021-09-28 21:27:25


(If you used the law of cosines, you probably got

d 2 = 2R 2(1 − cosθ).

This is also correct, but I find it a little harder of a form to work with for this problem.)

meichenlaub 2021-09-28 21:27:46


So the speed of the point is

meichenlaub 2021-09-28 21:27:51

v(θ) = dω = 2Rωsin ()
θ
2
.

meichenlaub 2021-09-28 21:28:04


The total distance traveled is of course

s = ∫ vdt.
meichenlaub 2021-09-28 21:28:10
How might we rewrite dt so that it plays better with v as we found it?

HappiHippo 2021-09-28 21:28:50


1
dt = dθ
ω

DGA_Student_1 2021-09-28 21:28:50



dt = ω

Hiboy 2021-09-28 21:28:50


dtheta*(v/r)

EthanTLee23 2021-09-28 21:28:50


dt = dθ / w

EL2024 2021-09-28 21:28:50


dt=d(theta)/w?

AOPS81619 2021-09-28 21:28:50



dt = ω

meichenlaub 2021-09-28 21:28:54


We have


dt = .
ω

With that, what do you get for the distance s ?

DGA_Student_1 2021-09-28 21:30:20


8R

mathisawesome01 2021-09-28 21:30:20


8R

mathtiger6 2021-09-28 21:30:20


8R

HappiHippo 2021-09-28 21:30:20


8R

taniaban 2021-09-28 21:30:20


8R

Zian2006 2021-09-28 21:30:20


8R

meichenlaub 2021-09-28 21:30:25


Our integral is

2π dθ
s = ∫ θ = 0 2Rωsin(θ / 2)
ω

= 2R∫ θ = 0 sin(θ / 2)dθ

= 2R ⋅ − 2cos(θ / 2)| 2π
0
= − 4R ⋅ ( − 1 − 1)
= 8R.

meichenlaub 2021-09-28 21:30:36


So the point travels a distance 8R during a revolution.
meichenlaub 2021-09-28 21:30:39
Just a quick check - is that figure the right ballpark?

DGA_Student_1 2021-09-28 21:31:01


yes

mathtiger6 2021-09-28 21:31:01


looks right in the rolling gif

HappiHippo 2021-09-28 21:31:01


sounds about right

Zian2006 2021-09-28 21:31:01


Seems about right

mathisawesome01 2021-09-28 21:31:01


yes

meichenlaub 2021-09-28 21:31:05


Yes, sure. The distance the entire wheel travels is 2πR, and the cycloid clearly travels further than that due to its up-and-down
motion, but not too much more than that. It's a pretty reasonable result.

meichenlaub 2021-09-28 21:31:14


Here's a little bonus to follow on:

meichenlaub 2021-09-28 21:31:19


What is the radius of curvature at the top of the cycloid?

meichenlaub 2021-09-28 21:31:26


By this, we mean that there's a certain circle that just fits into the cycloid at the top of its trajectory.

meichenlaub 2021-09-28 21:31:40


We know that instantaneously, the wheel is rotating about its contact point with the ground. The top of the cycloid is a distance 2R
from the contact point, so it seems like we are tracing out part of a circle with radius 2R.

meichenlaub 2021-09-28 21:31:44


meichenlaub 2021-09-28 21:31:47
And yet this is wrong! What's the mistake being made?

meichenlaub 2021-09-28 21:33:09


It's the same mistake as when we tried to find the tension in the rod in the previous problem by thinking of the motion as rotation
about the bead. In both cases, we're looking for an acceleration; we can't think of the motion as rotation about a moving contact
point.

meichenlaub 2021-09-28 21:33:20


This is a trap people sometimes fall into with rolling without slipping! If we think of the wheel as rotating about its point of contact,
we get the correct velocities of all the points. That means we could, for example, find the kinetic energy using this method as well,
because that just requires knowing the velocities.

meichenlaub 2021-09-28 21:33:28


But one moment later, the wheel is rotating about a new point. So we can't find accelerations this way, and we can't find the radius
of curvature this way. Let's find another method.

meichenlaub 2021-09-28 21:33:38


What's the speed of the point at the top of the wheel, given the wheel is rotating with angular frequency ω ?

mathisawesome01 2021-09-28 21:33:59


2v

DGA_Student_1 2021-09-28 21:33:59


2Rω

AOPS81619 2021-09-28 21:33:59


2Rω

Hershey806 2021-09-28 21:33:59


2wR

EL2024 2021-09-28 21:34:02


2wr

meichenlaub 2021-09-28 21:34:05


The speed is

v = 2Rω.

meichenlaub 2021-09-28 21:34:08


And what's the acceleration of the point at the top of the wheel?

EL2024 2021-09-28 21:34:40


w^2R

meichenlaub 2021-09-28 21:34:45


We can switch back to the picture where the wheel is rotating about its center and also translating to the right. The translation to
the right is steady and so doesn't affect the acceleration, so the acceleration is

a = ω 2R.

meichenlaub 2021-09-28 21:35:00


Using both v and a, what's the radius of curvature?

mathisawesome01 2021-09-28 21:35:43


4R

QMech 2021-09-28 21:35:43


4R

EL2024 2021-09-28 21:35:43


4R
AOPS81619 2021-09-28 21:35:43
4R

pimath 2021-09-28 21:35:43


4R

meichenlaub 2021-09-28 21:35:49


The acceleration is

v2
a=
r
(2ωR) 2
=
r
4ω 2R 2
= .
r

where r is the radius of curvature. Setting this equation to ω 2R, we have

4ω 2R 2
ω 2R = .
r

and solving for r,

r = 4R.

meichenlaub 2021-09-28 21:36:10


So the radius of curvature is actually four times the radius of the wheel at the top of the cycloid.

meichenlaub 2021-09-28 21:36:15

meichenlaub 2021-09-28 21:36:23


Summary

meichenlaub 2021-09-28 21:36:25


Rotational motion is governed by Newton's laws; we don't need any new physics beyond Newton's laws to understand it.

meichenlaub 2021-09-28 21:36:29


There are shortcuts that help us deal with rotational motion problems. These include:


→ dL
- The rotational analog of Newton's second law, τ = .
dt

→ →
- A formula for the angular momentum, L = Iω.

- A formula for the moment of inertia, I = ∭ r 2⊥ ρdV.


1 1
- A formula for the kinetic energy, T = mv 2 + Iω 2.
2 2

meichenlaub 2021-09-28 21:36:36


We can think of the motion of a rigid body in terms of translation of the center of mass plus rotation about the center of mass.

meichenlaub 2021-09-28 21:36:40


We can also sometimes think of it in terms of rotation about a point. However, if they point is moving or accelerating, we have to be
very careful about using this method to find accelerations or forces.

meichenlaub 2021-09-28 21:36:47


Credits
Figure skater: "The Physics of Figure Skating", Science Friday, https://www.sciencefriday.com/segments/the-physics-of-figure-
skating/
Rotating cylinder: "The Bizarre Behavior of Rotating Bodies, Explained", Veritasium. https://youtu.be/1VPfZ_XzisU
Cycloid gif: By Zorgit - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=4552689

© 2022 Art of Problem Solving


About Us • Contact Us • Terms • Privacy

Copyright © 2022 Art of Problem Solving

You might also like