# Cylinder lying on conveyor belt

#### JD_PM

Yes F will have a greater magnitude than f, but that just means the resulting observed acceleration is in the same direction as F,
OK I think that what happens is that $a_o$ has a negative sign as well, because is pointing backwards. Thus:

$$a_o = \frac{F - f}{m}$$

#### JD_PM

Ok, so what about rotational acceleration.
$$\tau = Rf$$

$$f = \frac{\tau}{R}$$

#### haruspex

Homework Helper
Gold Member
2018 Award
$$\tau = Rf$$

$$f = \frac{\tau}{R}$$
That does not mean anything unless you specify the axis.

#### kuruman

Homework Helper
Gold Member
(as we are on a non inertial reference frame we have to add a reaction force F; the action force is clearly f).
It looks like you are confused about action and reaction forces. Only action forces act on the system, here the cylinder. Reaction forces act on entities outside the system and are not part of the system's free body diagram. For example, the horizontal contact force of static friction acting on the cylinder has a reaction counterpart that is in the opposite horizontal direction and is part of the belt free body diagram, the action force of gravity $m\vec g$ acting on the cylinder has a reaction force $-m\vec g$ acting on the Earth, etc. etc.
I don't know what to write here. I would not have taken you along a non-inertial path. Not sure what @kuruman has in mind.
I do not wish to detract @JD_PM from the current path. When this problem has been solved to everybody's satisfaction, perhaps I will post my solution for consideration.

#### haruspex

Homework Helper
Gold Member
2018 Award
I do not wish to detract @JD_PM from the current path.
Yes, I was reluctant to remain involved for the same reason, but we seem to be in different timezones and I did not want to leave the OP dangling. Anyway, it seems to me that the thread has since then stuck with the non-inertial view, so no need for you to hold back.

#### kuruman

Homework Helper
Gold Member
Yes, I was reluctant to remain involved for the same reason, but we seem to be in different timezones and I did not want to leave the OP dangling. Anyway, it seems to me that the thread has since then stuck with the non-inertial view, so no need for you to hold back.
OK, then. We still need to hear from @JD_PM about what reference axis one ought to calculate the torque acting on the cylinder. One choice is better than others in the sense that Newton's 2nd law for rotations is sufficient to determine the acceleration of the CM in the non-inertial frame.

#### JD_PM

OK, then. We still need to hear from @JD_PM about what reference axis one ought to calculate the torque acting on the cylinder. One choice is better than others in the sense that Newton's 2nd law for rotations is sufficient to determine the acceleration of the CM in the non-inertial frame.
To Recap: I am still convinced that the right answer is(see #9 for more details):

$$v(t) = v_{CM} + V(t) = 2R \omega + V(t)$$

Do you agree with me?

#### JD_PM

To Recap: I am still convinced that the right answer is(see #9 for more details):

$$v(t) = v_{CM} + V(t) = 2R \omega + V(t)$$

Do you agree with me?
And I still have the following doubt(stated previously on #10):

Now I am confused on the following:

I know that the translational speed of the centre of mass for pure rolling motion on a non moving floor and measured from an inertial reference frame is:

$$v_{CM} = \frac{ds}{dt} = R\omega$$

Where $s$ is the arc of the cylinder

But how is it possible that I get $2R \omega$ just with the difference that in this case the floor (belt) is moving?

#### JD_PM

My solution is obtained from an inertial reference frame.

@kuruman you were interested in solving the problem from a non inertial reference frame. why?

#### JD_PM

That does not mean anything unless you specify the axis.
I would say that the axis plays a role if we use:

$$\tau = I \alpha$$

The angular acceleration is obtained from the stated equation:

$$\alpha = \frac{\tau}{I}$$

#### kuruman

Homework Helper
Gold Member
To Recap: I am still convinced that the right answer is(see #9 for more details):

$$v(t) = v_{CM} + V(t) = 2R \omega + V(t)$$

Do you agree with me?
I do not agree with you. In the special case $V(t)=0$ for all $t$, your expression predicts that the cylinder will be moving. Does that make sense? Besides, what is $\omega$? Your final answer must be in terms of the given quantities and these are $M,~k,~R$ and $V(t)$.
My solution is obtained from an inertial reference frame.

@kuruman you were interested in solving the problem from a non inertial reference frame. why?
The strategy for finding $v_{cm}(t)$ is to use dynamics to find $a_{cm}(t)$ and integrate. $\vec F_{net}= M\vec a_{cm}$ is insufficient to complete the task because there is also angular acceleration. So you need $\tau=I\alpha$. I am arguing that if you transform to the non-inertial frame, it is sufficient to use $\tau=I\alpha$ in order to find the acceleration of the cm in the non-inertial frame and hence in the inertial frame. However, for that to happen, you need to choose wisely the axis about which you the torque and the moment of inertia are to be expressed.
I would say that the axis plays a role if we use:
$$\tau = I \alpha$$
The angular acceleration is obtained from the stated equation:
$$\alpha = \frac{\tau}{I}$$
It's not "if we use", it's "because we have to use".

#### haruspex

Homework Helper
Gold Member
2018 Award
I would say that the axis plays a role if we use:

$$\tau = I \alpha$$

The angular acceleration is obtained from the stated equation:

$$\alpha = \frac{\tau}{I}$$
The torque of a force is always in respect of an axis. Different axis, different torque. Likewise moment of inertia. Even the angular acceleration changes with axis; if your axis is not through the mass centre then the linear acceleration of the body contributes to its angular acceleration about the axis. You must use the same axis for all three.
E.g. consider a force F applied at the mass centre, but an axis distance s off to one side. The force has torque Fs about the axis. The body does not rotate about its own centre, but the axis sees the body as accelerating around it at angular rate a/s=F/(ms). The moment of inertia about this axis (since the body is not rotating about its own centre) is ms2. Torque Fs=(ms2)(F/(ms)).

Further, it is advisable to use either a point fixed in space or the mass centre of the body as the axis. Anything else can yield the wrong answer.

#### kuruman

Homework Helper
Gold Member
Further, it is advisable to use either a point fixed in space or the mass centre of the body as the axis. Anything else can yield the wrong answer.
Yes! In the inertial frame there is no point on the cylinder that is fixed in space, so one must use the cylinder axis for referencing torques and the moment of inertia. In the non-inertial frame, however, there is a point that is fixed in space and one can use it as reference instead of the cylinder axis.

#### haruspex

Homework Helper
Gold Member
2018 Award
In the inertial frame there is no point on the cylinder that is fixed in space
True, but you do not have to use a point that is fixed to the cylinder. Any point fixed in space will do.

#### kuruman

Homework Helper
Gold Member
True, but you do not have to use a point that is fixed to the cylinder. Any point fixed in space will do.
Yes, of course.

#### JD_PM

I do not agree with you. In the special case $V(t)=0$ for all $t$, your expression predicts that the cylinder will be moving. Does that make sense? Besides, what is $\omega$? Your final answer must be in terms of the given quantities and these are $M,~k,~R$ and $V(t)$.

The strategy for finding $v_{cm}(t)$ is to use dynamics to find $a_{cm}(t)$ and integrate. $\vec F_{net}= M\vec a_{cm}$ is insufficient to complete the task because there is also angular acceleration. So you need $\tau=I\alpha$. I am arguing that if you transform to the non-inertial frame, it is sufficient to use $\tau=I\alpha$ in order to find the acceleration of the cm in the non-inertial frame and hence in the inertial frame. However, for that to happen, you need to choose wisely the axis about which you the torque and the moment of inertia are to be expressed.

It's not "if we use", it's "because we have to use".
Using $\tau=I\alpha$ I got:

$$\tau=I\alpha = Rf$$

$$Mk^2 \alpha = RMa_{CM}$$

$$a_{CM} = \frac{k^2 \alpha}{R}$$

NOTE: I used an axis passing through the centre of mass of the cylinder. You will get the same result using $\tau=I\alpha$ no matter from what reference frame you analyse the rotation. This is because the fictitious force F (in non inertial reference frame case) exerts no torque on the cylinder.

Before proceeding to calculate $v_{CM}$ please tell me how you see the reasoning above.

#### haruspex

Homework Helper
Gold Member
2018 Award
Using $\tau=I\alpha$ I got:

$$\tau=I\alpha = Rf$$

$$Mk^2 \alpha = RMa_{CM}$$

$$a_{CM} = \frac{k^2 \alpha}{R}$$

NOTE: I used an axis passing through the centre of mass of the cylinder. You will get the same result using $\tau=I\alpha$ no matter from what reference frame you analyse the rotation. This is because the fictitious force F (in non inertial reference frame case) exerts no torque on the cylinder.

Before proceeding to calculate $v_{CM}$ please tell me how you see the reasoning above.
From the above I deduce you are assuming $f =Ma_{CM}$. Presumably aCM is inthe reference frame of the belt. What happened to the fictional force?

#### JD_PM

From the above I deduce you are assuming $f =Ma_{CM}$. Presumably aCM is inthe reference frame of the belt. What happened to the fictional force?
Note that the fictitious force F is exerted on the centre of mass of the cylinder i.e. F produces no torque on the cylinder.

See on my diagrams how F acts on the COM.

#### haruspex

Homework Helper
Gold Member
2018 Award
Note that the fictitious force F is exerted on the centre of mass of the cylinder i.e. F produces no torque on the cylinder.

See on my diagrams how F acts on the COM.
I have no issue with your torque equation. As you say, F does not have a torque about the cylinder's centre. The problem is with your second equation in post #41, where you introduce aCM. It seems to me that to get that you must have assumed $f=Ma_{CM}$. I can't see where you define aCM. If it is the acceleration of the mass centre in the belt frame then F should appear in the equation.

#### JD_PM

I have no issue with your torque equation. As you say, F does not have a torque about the cylinder's centre. The problem is with your second equation in post #41, where you introduce aCM. It seems to me that to get that you must have assumed $f=Ma_{CM}$. I can't see where you define aCM. If it is the acceleration of the mass centre in the belt frame then F should appear in the equation.
OK I see what you mean. Actually there is a mistake, as the second line at #41 should be:

$$Mk^2 \alpha = RMa_{o}$$

Having $f =Ma_{o}$, where $a_{o}$ is the tangential acceleration.

@kuruman suggested solving the problem being on a non inertial reference frame and using 2nd Newton's law for rotation. Well, in that scenario we have the picture #20 but with the fraction pointing to the right (I made I mistake when I posted the diagram #20). In such a scenario, using 2nd Newton's law for rotation we have:

$$\tau=I\alpha = Rf$$

From this equation there is no way you get $a_{CM}$ because:

$$F = Ma_{CM}$$

And F exerts no torque on the cylinder, therefore it does not appear on $\tau=I\alpha = Rf$

I made some research and found you can get both $v_{CM}$ and $a_{CM}$ using:

$$v_{CM} = \frac{ds}{dt} = R\frac{d\theta}{dt} = R\omega$$

But for that you need to assume constant acceleration...

And this approach does not use 2nd Newton's law... so I am basically stuck.

#### haruspex

Homework Helper
Gold Member
2018 Award
OK I see what you mean. Actually there is a mistake, as the second line at #41 should be:

$$Mk^2 \alpha = RMa_{o}$$

Having $f =Ma_{o}$, where $a_{o}$ is the tangential acceleration.

@kuruman suggested solving the problem being on a non inertial reference frame and using 2nd Newton's law for rotation. Well, in that scenario we have the picture #20 but with the fraction pointing to the right (I made I mistake when I posted the diagram #20). In such a scenario, using 2nd Newton's law for rotation we have:

$$\tau=I\alpha = Rf$$

From this equation there is no way you get $a_{CM}$ because:

$$F = Ma_{CM}$$

And F exerts no torque on the cylinder, therefore it does not appear on $\tau=I\alpha = Rf$

I made some research and found you can get both $v_{CM}$ and $a_{CM}$ using:

$$v_{CM} = \frac{ds}{dt} = R\frac{d\theta}{dt} = R\omega$$

But for that you need to assume constant acceleration...

And this approach does not use 2nd Newton's law... so I am basically stuck.
I might return to this when I've more time to suggest how you could continue with that line, but for now I'd like to suggest a far simpler approach.
In the inertial frame, can you think of an axis about which there are no torques? What would that allow you to do?

#### JD_PM

I might return to this when I've more time to suggest how you could continue with that line, but for now I'd like to suggest a far simpler approach.
In the inertial frame, can you think of an axis about which there are no torques? What would that allow you to do?
SCENARIO ANALYSED FROM AN INERTIAL FRAME

We have to point out that in an inertial frame of reference we just have one external force: the static friction (no slipping rolling motion).

You asked for a selection of an axis which allows the system having $\tau = 0$ . That would be achievable if we were to select the point of contact cylinder-belt as the one the axis passes through perpendicularly.

With respect to that axis angular momentum is conserved. Is it that your idea?
If that is the case, how could we get $v_{CM}$ from AM conservation?

#### jbriggs444

Homework Helper
SCENARIO ANALYSED FROM AN INERTIAL FRAME
With respect to that axis angular momentum is conserved. Is it that your idea?
If that is the case, how could we get $v_{CM}$ from AM conservation?
If I am understanding the problem correctly and seeing where @haruspex is going, we have a pretty simple known and unchanging quantity to utilize.

1. Starting angular momentum is zero
2. Angular momentum is conserved in the absence of external torques.
3. There are no external torques.

If angular momentum is zero and we know the velocity of the bottom contact surface of the cylinder with the belt, what else can we calculate?

#### haruspex

Homework Helper
Gold Member
2018 Award
select the point of contact cylinder-belt as the one the axis passes through perpendicularly.
It depends exactly what you mean by that. Recall that we must choose a reference axis that is either the mass centre or is fixed in space. Since the belt is moving, the "point of contact" moves.

Your other equation should express that it is rolling contact.

#### JD_PM

If angular momentum is zero and we know the velocity of the bottom contact surface of the cylinder with the belt, what else can we calculate?
I guess you are driving me to the vectorial angular momentum equation:

$$L = r p$$