# Angular Acceleration - Should an unpivoted rod rotate only about the centre of mass?

1. Jan 9, 2010

### sriram_15_93

1. The problem:
A Force F is provided at the top of a rod placed vertically placed on a smooth frictionless surface. The length of the rod is L and its mass, M. Find the angular acceleration of the rod at the instant the force is applied.
- Please tell me how to solve this and why you have solved it in that way.

2. I solved this problem about the centre of mass, but can it be solved about any other point?

3. If I can solve it about any point I wish, why so?

4. Is it necessary that a body such as the one in the above problem must rotate about its centre of mass only? How can we state this? What is the proof? How do we find out about which point the problem must be solved.

5. I don't know if my method of solving was correct, but when I solved it about the centre of mass I got one answer, and when i solved it about the point in contact with the ground, I got a different answer. So, is angular acceleration different about different points? Or is angular acceleration an intrinsic property? If so, why? Are the answers I get (by solving in the aforementioned two ways) the angular acceleration relative to the respective points about which I have solved the question? If they are, how do I get the answer in the ground frame? What would have been the pseudo forces if it was relative acceleration I was calculating?

I solved the problem in the following ways:

a) About the centre of mass:
F*(L/2)=I*angular acceleration=(M*L^2)/12*angular acceleration
=>angular acceleration=6*F/(M*L)

b) About the point at the bottom of the rod, in contact with the frictionless horizontal surface:
F*(L)=I*angular acceleration=(M*L^2)/3*angular acceleration
=>angular acceleration=3*F/(M*L)

2. Jan 9, 2010

### tiny-tim

Welcome to PF!

Hi sriram_15_93! Welcome to PF!

(have an omega: ω and an alpha: α and try using the X2 tag just above the Reply box )
a) is correct.

In b), you've used the moment of inertia about a point other than the c.o.m. or the centre of rotation, and you can't do that, you have to use the moment of inertia about the c.o.m. to get Lspin, and then add on the Lorbital as an extra, to get the total L.

Torque = d/dt (L) = d/dt (Lspin + Lorbital)

= d/dt (Ic.o.m.ω + rc.o.m.x vc.o.m.) = Ic.o.m.α + rc.o.m.x ac.o.m.

"Torque about point P = IPα" (using IP as the moment of inertia about the point P) only works if P is the c.o.m. or the centre of rotation.

(and it works for the centre of rotation, O, only because Ic.o.m.α + rc.o.m.x ac.o.m. = Ic.o.m.α + rc.o.m.x (rc.o.m.x αc.o.m.) = Ic.o.m.α + rc.o.m.2α = IOα by the parallel axis theorem.)

In this case (leaving out all the Fs and Ls …), ac.o.m. = 1, and α = 6, and the torque about a point distant r from the force is r.

If you take moments about the centre of the rod (r = 1/2), dL/dt = 6*1/12 = 1/2 = r

If you take moments about the top of the rod (r = 0), dL/dt = 6*1/12 minus 1*1/2 = 0 = r

If you take moments about the centre of rotation (r = 2/3), dL/dt = 6*1/12 plus 1*1/6 = 2/3 = r (or dL/dt = 6*1/9 = 2/3 = r)

If you take moments about the bottom of the rod (r = 1), dL/dt = 6*1/12 plus 1*1/2 = 1 = r

3. Jan 10, 2010

### sriram_15_93

Re: Angular Acceleration - Should an unpivoted rod rotate only about the centre of ma

Thanks a lot Tiny-Tim! So, spin angular momentum an intrinsic property while orbital angular momentum depends on the point about which we are calculating torque, is it? And about the signs (6*1/12 plus 1/2 or 6*1/12minus 1/2) - how did you get them? I've always been solving sums only about the c.o.m because I've been told it simplifies calculations but I've never known why. So the concept of spin angular momentum and orbital angular momentum is new to me. So please explain how you got a plus or a minus in the equations.
and another doubt:
By rc.o.m I think you mean the distance of the reference point from the centre of mass. but does ac.o.m mean the acceleration of the centre of mass, or the acceleration of the reference point w.r.t the c.o.m?
Thanks once again!

Last edited: Jan 10, 2010
4. Jan 10, 2010

### tiny-tim

Exactly.
I just look at the diagram to see whether the c.o.m. is going clockwise or anti-clockwise.

The c.o.m. in this case moves to the right, so from above it's anti-clockwise, and from below it's clockwise (and the spin is clockwise).
I meant the position or acceleration of the c.o.m from my "origin" (and I would always have ac.o.m = r''c.o.m anyway).

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Threads - Angular Acceleration Should Date
Finding the coefficient of kinetic friction on an incline Jan 13, 2018
Wheel-Hub System and Angular Acceleration Jan 6, 2018
Angular acceleration Dec 12, 2017