Suppose we have a disk of radius $r$ and mass $m$ travelling at velocity $v$. I want to calculate the instantaneous angular momentum with axis through the edge of the disc (on the circumference).

Angular momentum $= I \omega$. $I = \frac{1}{2}mr^2 + mr^2 = \frac{3}{2}mr^2$ by the parallel axis theorem. $\omega = \frac{v}{r}$. Therefore, angular momentum $= \frac{3mrv}{2}$.

Alternatively, angular momentum $=p\times r= m r \times v = mrv$.

Why do these two methods differ? Which, if any, are correct?

  • 1
    $\begingroup$ Is the disk rolling? If not then there is no angular velocity specified and angular momentum has only the $r \times m v$ term. $\endgroup$ – ja72 Apr 1 '14 at 20:34

From your description, I assume the disk only only translating, not rotating. Is this correct? If so, read on. If not, I'll delete.

I'm uncomfortable with the first method that uses $L=I\omega$. In this equation, it is assumed that every point on the rigid body can be characterized by the same angular velocity $\omega$. From your description of the motion of the disk, it seems like this doesn't apply here. The disk is only translating, not rotating about a point; thus, each point will have a different angular velocity. I don't have a problem with your expression for the moment of inertia $I$, but that would only be applicable if the object were rotating about a point on its edge.

I believe your second method $\vec{L}=m\vec{r}\times\vec{v}$ makes an assumption that the object is a point particle. You can see this because you treat each point in the body as being characterized by the same position vector $\vec{r}$. This may or may not lead to the correct answer. As another poster stated, the form you want to use is $L=\int \vec{r}\times d\vec{p} = \int_A \vec{r} \times \sigma \vec{v}\ dA$, where I've used a two-dimensional integral since you are treating the disk as two-dimensional. The term $\sigma$ is the two dimensional areal mass density $m/(\pi R^2)$. Let's continue with this integral to see where it leads. $$L=\sigma\int_A \vec{r} \times \vec{v} \ dA = \sigma\int_A (\vec{r}_{CM} + \vec{r}_{body}) \times \vec{v} \ dA = \sigma\int_A \vec{r}_{CM} \times \vec{v} \ dA + \sigma\int_A \vec{r}_{body} \times \vec{v} \ dA $$ I've separated the position vector $\vec{r}$ into the sum of the vector to the center of mass and the vector from the center of mass to a general point on the body. $$ L = \sigma\int_A \vec{r}_{CM} \times \vec{v} \ dA + \sigma \underbrace{\int_A \vec{r}_{body} \times \vec{v} \ dA}_{=0?} \stackrel{?}{=} \sigma A \vec{r}_{CM}\times \vec{v}$$ I believe the integral with the underbrace is zero by symmetry. (Someone want to chime in?) If so, this method produces your second result.

  • $\begingroup$ By definition for a rigid body, all points share the same angular velocity vector. Only the linear velocity vector changes from location to location. $\endgroup$ – ja72 Apr 1 '14 at 20:35
  • $\begingroup$ If the rigid body were rotating, I would agree with that. But compare $d\theta/dt$ for an object close to the reference point and far from the reference point. In some time $dt$, both points travel the same distance $dl=v\,dt$, but the closer point's path would subtend a larger $d\theta = dl/r=v\,dt/r$, where I've assumed the motion is perpendicular to the reference point. $\endgroup$ – BMS Apr 1 '14 at 20:39
  • $\begingroup$ I think we're imagining two different scenarios. I assumed from the description that the disk was only translating, not rotating. $\endgroup$ – BMS Apr 1 '14 at 20:42
  • $\begingroup$ If the body is not rotating, all points share the same rotational velocity of zero. You make it sound like some points are rotating while others do not, which is incorrect. $\endgroup$ – ja72 Apr 1 '14 at 20:42

The first method is correct.

The second method is wrong because the equation you use only applies to point particles, not continuous masses with volume (such as a disk). You're incorrectly treating the disk as a point particle located at the disk center. If you want to use the second method, you'll need to use this equation for angular momentum of continuous masses:

$\vec{L} = \int_V dV \, \vec{r} \times \rho(\vec{r})\vec{v}$

  • 1
    $\begingroup$ Is the first method actually correct? The equation $L=I\omega$ assumes every point on the rigid body can be described by the same angular velocity $\omega,$ which is true for an object rotating about a point. Not sure about this case with the object moving with uniform motion. $\endgroup$ – BMS Apr 1 '14 at 18:47
  • $\begingroup$ If you assume that v is in a direction perpendicular to the vector r from the center to the edge of the disk, then the first method is correct because the disk is moving, at a given instant in time, in a rotational direction with respect to the axis through the edge of the disk. $\endgroup$ – Izzhov Apr 1 '14 at 18:53
  • $\begingroup$ Actually $\vec{H} = \vec{r} \times m \vec{v} + I \vec{\omega}$ is the full expression for angular momentum not about the c.m. so the second method is not wrong. They just describe different problems. $\endgroup$ – ja72 Apr 1 '14 at 20:38

Case a)

Body with uniform motion (no rotation), with C the center of the disk and A a point on the edge (for example below, at a distance $R$).

$$ \begin{aligned} \vec{v}_A & = (v,0,0) \\ \vec{\omega} & = (0,0,0) \\ \vec{v}_C & = \vec{v}_A + (0,-R,0) \times \vec{\omega} = (v,0,0) \\ \vec{L} & = m \vec{v}_C = (m v,0,0) \\ \vec{H}_A & = I \vec{\omega} + (0,-R,0)\times \vec{L} = (0,0,R m v) \\ \end{aligned} $$

Where $\vec{L}$ is linear and $\vec{H}_A$ is angular momentum about point A.

Case b)

Body rolling with edge point A motionless, but with rotational speed $\Omega$

$$ \begin{aligned} \vec{v}_A & = (0,0,0) \\ \vec{\omega} & = (0,0,\Omega) \\ \vec{v}_C & = \vec{v}_A + (0,-R,0) \times \vec{\omega} = (\Omega R,0,0) \\ \vec{L} & = m \vec{v}_C = (m \Omega R,0,0) \\ \vec{H}_A & = I \vec{\omega} + (0,-R,0)\times \vec{L} = (0,0,\Omega (I_C+m R^2)) \\ \end{aligned} $$

The angular momentum here uses the parallel axis theorem and it expands to..

$$ v_C = \frac{\Omega}{R} \\ I_C = \frac{m}{2} R^2 $$

$$ \begin{aligned} L & = (m v_C,0,0) \\ H_A & = \frac{3}{2} m v_C R \\ \end{aligned} $$

So you see your two solutions correspond to two different problems.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.