2
$\begingroup$

I am struggling with an assignment from my lecturer, he has asked me to numerically (mathematica) and analytically compute solutions for the trajectory of a point particle in AdS3 space with global coordinates given by metric $$ds^2=R^2(-\cosh^2(\rho)dt^2+d\rho^2+\sinh^2(\rho)d\phi^2)$$ I calculated three geodesic equations of motion

$$\ddot{\rho}+\sinh(\rho)\cosh(\rho)(\dot{t}\dot{t}-\dot{\phi}\dot{\phi})=0$$

$$\ddot{\phi}+\coth(\rho)\dot{\rho}\dot{\phi}=0$$

$$\ddot{t}+\tanh(\rho)\dot{\rho}\dot{t}=0$$

I realise they're coupled but I honestly have no clue how I go finding $\rho(\tau)$, $t(\tau)$ and $\phi(\tau)$.

$\endgroup$
5
  • $\begingroup$ First, I think you might have some factors of 2 missing in the $\phi$ and $t$ equations. $\endgroup$
    – secavara
    Commented Jan 13, 2018 at 20:14
  • $\begingroup$ Second, one case that can be solved easily is when $\rho$ is fixed. $\endgroup$
    – secavara
    Commented Jan 13, 2018 at 20:48
  • $\begingroup$ Hi Secavara, could you elaborate? $\endgroup$
    – Cyphox32
    Commented Jan 13, 2018 at 20:53
  • $\begingroup$ Oh, I made some scribbles so I could be wrong, but for $\phi$ and $t$ I got $\ddot{\phi}+2\coth(\rho) \dot{\rho} \dot{\phi} = 0$ and $\ddot{t}+2\tanh(\rho) \dot{\rho} \dot{t} = 0$ . And then notice that if we take $\dot{\rho} = 0$, then we can find some geodesics easily. $\endgroup$
    – secavara
    Commented Jan 13, 2018 at 20:57
  • $\begingroup$ In that case, $t=a+b\tau$ and $\phi=c \pm b\tau$. $\endgroup$
    – secavara
    Commented Jan 13, 2018 at 21:00

1 Answer 1

1
$\begingroup$

Assuming your equations are correct (I have not checked), but coupled ODEs such as this (and geodesic equations in general) must be solved numerically. The trick is to employ the standard Runge-Kutta solvers, it is most convenient to deal with first-order systems. In general, given the geodesic equations:

$\ddot{x}^{a} + \Gamma^{a}_{bc} \dot{x}^{b} \dot{x}^{c} = 0$,

we can write these as a first-order system:

$\dot{x}^{a} = v^{a}$, $\dot{v}^{a} = - \Gamma^{a}_{bc} v^{b} v^{c}$

So, for your system above, define $x^{a} = \dot{\rho}^{a}$, $y^{a} = \dot{\phi}^{a}$, and $z^{a} = \dot{t}^{a}$. Then, coupled with the remaining equations you have a 6-D system, first-order, and the solutions will depend on specifying 6 such initial conditions.

Hope this helps.

$\endgroup$
4
  • $\begingroup$ It's hardly a "trick" employing RK solvers, since they are among the most popular and often the 'go to' before resorting to more elaborate methods. $\endgroup$
    – JamalS
    Commented Jan 14, 2018 at 11:22
  • $\begingroup$ @JamalS i think you took my sentence way too literally. R-K solvers don't natively work on 2nd-order ODEs, the "trick" is to write 2nd-order ODEs as first-order ODEs, and then apply the R-K methods. $\endgroup$ Commented Jan 14, 2018 at 15:33
  • $\begingroup$ @Dr.IkjyotSinghKohli I have made substantial progress in my problem, however I am having trouble specifying inital conditions that lead to a stable solution according to the solver I am using. Thanks $\endgroup$
    – Cyphox32
    Commented Jan 17, 2018 at 15:53
  • $\begingroup$ @Cyphox32 hi. Yes. This is frequently a problem in such numerical applications. You may want to look at using MATLAB / Octave / NumPY/SciPy in Python for these sorts of things. MATLAB has very powerful numerical solvers: ODe23s OdE45, etc.. and it is easier to experiment with initial conditions. $\endgroup$ Commented Jan 17, 2018 at 16:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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