I Lagrangian - surface of sphere

AI Thread Summary
A free particle moving on a sphere's surface is analyzed using the Lagrangian approach, leading to equations of motion that describe the particle's dynamics. The Euler-Lagrange equations yield a constant of motion and a second-order differential equation for theta, which is subsequently integrated with respect to time. The integration process reveals difficulties in obtaining a specific solution form, prompting discussions about variable separation and substitution methods. Participants suggest changes in variables to simplify the integral, but challenges persist in solving the resulting expressions. The conversation highlights the complexities of integrating motion equations in spherical coordinates.
rolotomassi
Messages
52
Reaction score
0
I have a free particle moving on the surface of a sphere of fixed radius R. Gravity is ignored and m/2 is left out since its constant.
The lagrangian is L = R^2 \dot{\theta^2} + R^2 sin^2{\theta} \dot{\phi^2}

Using the Euler Lagrange equations I obtain
sin^2{\theta} \dot{\phi} = A = const \ (1) \\ \ddot{\theta} - sin{\theta}cos{\theta} \dot{\phi^2} = 0 \ (2)
and by substituting 1 into 2
\ddot{\theta} = A^2 cos{\theta}/sin^3{\theta}

By integrating w.r.t time and using the fact dt = d{\theta}/\dot{\theta} and that theta and its time derivative are treated as independent coordinates i get
\dot{\theta} + A^2/ 2 \dot{\theta} sin^2{\theta} \ = c_1
integrating w.r.t time again i get. [Using omega instead of theta dot now.]

\theta - A^2 cot{\theta}/\omega^2 \ = t c_1 + c_2

I can't see anything wrong but I am supposed to get this into the form

\theta(t) = arccos ( \sqrt{1 - A^2/\omega^2} cos(\omega t + \theta_0)

and I cant. If anyone can help I would appreciate it a lot. Thanks
 
Physics news on Phys.org
Your integration of the ##\ddot\theta## equation looks suspicious.
 
You're right it was. What I get now which I am pretty sure is correct is this horrible integral

\theta = \frac{1}{\omega} \int \sqrt[]{ c_1 - c_2/sin^2{\theta} } \ d \theta

Which I also have no idea how to solve :S
 
rolotomassi said:
You're right it was. What I get now which I am pretty sure is correct is this horrible integral

\theta = \frac{1}{\omega} \int \sqrt[]{ c_1 - c_2/sin^2{\theta} } \ d \theta

Which I also have no idea how to solve :S
You cannot have ##\theta## as the integration variable and the result!
 
wow I'm being stupid today.

If i separate the variables I get \int \frac{1}{ \sqrt[]{(c_1 - c_2/sin^2{\theta} )} }d \theta \ = \ \int dt
 
Indeed, so I suggest a change of variables in the left integral. (Doing this change of variables from the beginning significantly simplifies the differential equation. You might try it instead.)
 
I start again and get \dot \theta = c_1 \int \frac{u}{(1-u^2)^{3/2}} \frac{dt}{du}du But still abit stuck. If I sub in du/dt I keep getting a theta dot term. But I can't just assume u dot is independent of u and take it out the integral either. Brain is really hitting brick wall after brick wall with this
 
So start by solving the integral then.
 
Solving the integral IS the question. If i could solve the integral I wouldn't be asking for help on solving the integral.
 
  • #10
Well, so use the substitution in it. It is a standard integral.
 
  • #11
The integral is now \int d\theta (c_2 + \frac{c_1^2}{u^2 -1})^{-1/2}= \int dt As I find it. But i get this simplifying to \int du \frac{1}{u^2 - a^2} for the LHS where a^2 is a combination of the other constants. But this is a logarithm when integrating which doesn't make sense for the answer.
 
  • #12
\int 1/(c - 1/2sin^2x)^{1/2} dx is what I've got now
 
  • #13
You are not doing the arithmetics properly then. Please show your work.
 
  • #14
That was the correct integral bar a constant, which has now been solved. Thanks anyway
 
Back
Top