• Register
PhysicsOverflow is a next-generation academic platform for physicists and astronomers, including a community peer review system and a postgraduate-level discussion forum analogous to MathOverflow.

Welcome to PhysicsOverflow! PhysicsOverflow is an open platform for community peer review and graduate-level Physics discussion.

Please help promote PhysicsOverflow ads elsewhere if you like it.


New printer friendly PO pages!

Migration to Bielefeld University was successful!

Please vote for this year's PhysicsOverflow ads!

Please do help out in categorising submissions. Submit a paper to PhysicsOverflow!

... see more

Tools for paper authors

Submit paper
Claim Paper Authorship

Tools for SE users

Search User
Reclaim SE Account
Request Account Merger
Nativise imported posts
Claim post (deleted users)
Import SE post

Users whose questions have been imported from Physics Stack Exchange, Theoretical Physics Stack Exchange, or any other Stack Exchange site are kindly requested to reclaim their account and not to register as a new user.

Public \(\beta\) tools

Report a bug with a feature
Request a new functionality
404 page design
Send feedback


(propose a free ad)

Site Statistics

174 submissions , 137 unreviewed
4,308 questions , 1,640 unanswered
5,089 answers , 21,602 comments
1,470 users with positive rep
635 active unimported users
More ...

  How to ad-hoc solve Einstein-equation for general metric numerically?

+ 2 like - 0 dislike

Hey folks, feel free to move this post wherever you feel it fits (newbie)

I would like to solve some specific energy-momentum field configuration on a grid (say, 1+1 dim) and I wonder how this would be most efficiently done. Let us assume I have some initial data $T_{\mu \nu} (t,x)$ in some local coordinates $x$ at time $t = 0$, for instance some local energy density $\epsilon(x)$ and 4-speed $v_\mu(x)$, s.t.

$$ T_{\mu\nu}(t=0, x) \equiv  \epsilon(x) v_\mu(x) v_\nu(x) $$

My first guess would be to start now by assuming that the system is slowly changing, i.e. $\partial_0 g_{\mu\nu} \approx 0 $, assuming a minkowsky metric $\eta_{\mu\nu}$ with perturbations, calculating the Christoffel symbols, Curvature Tensor, etc. and finding the difference $\xi(x)$ at $t=0$:

$$\xi(x) = \left \|  8\pi T_{\mu\nu}(x) - R_{\mu\nu}(x) + g_{\mu\nu}(x)R(x)\right \| $$

Now I imagined using some crazy downhill algorithm to minimize this function, something like

$$ g_{\mu\nu} \rightarrow g_{\mu\nu} - \frac{\partial \xi(x)}{\partial g_{\mu\nu}(x)} $$

Of course, this does not use the symmetries (such as Bianchi, etc.) and could fall into a local minimum (worse: Maybe the gauge condition $\nabla_\mu g_{\alpha \beta} = 0$ is not fullfilled?) However, let us imagine this worked, my next step would be to proceed to the next timestep $t \rightarrow t + \Delta t$ and use the geodetic equation $\nabla_\mu T^{\mu \nu} = 0 $ to somehow calculate the new energy tensor $T^{\mu \nu}(t+\Delta t, x)$, i.e. assuming that locally we can do something badly forbidden like

$$\epsilon(x, t + \Delta t) = \frac{1}{V} \int d^d x' \cdot \epsilon(x',t) \delta(x,x' + \Delta t \cdot v(x')) \\ \tilde{v}(x, t + \Delta t) = \frac{1}{V} \int d^d x' \cdot v(x',t) \delta(x,x' + \Delta t \cdot v(x'))\\ v^\mu(x, t+ \Delta t) = \tilde{v}^\mu(x, t + \Delta t) - \Delta t \cdot \Gamma^\mu_{\alpha \beta} \tilde{v}^\alpha(x,t+\Delta t) \tilde{v}^\beta(x,t+\Delta t)   $$

Where in the last equation I have used the geodetic equation for local speeds (?). After having calculated the new speeds and energy densities, the respective tensor follows as above and I can calculate the new metric. I think the main problem with the method here is the usage of the local frames $t$. Probably I have to change the coordinate system to the respective local frame to be able to make the last step and be able to use t?

Let us assume that this method has run through the whole world sheet, I would suggest just to start over and recalculate the metric, this time with changing derivates? Apparently the problem is to solve the partial differential equations with border condition $T_{\mu\nu}(t=0,x) = \text{const} $ and therefore the method described above is very naive. Maybe you have a better idea?

asked Oct 5, 2016 in Computational Physics by vsilv (60 points) [ revision history ]
edited Oct 5, 2016 by vsilv

Please log in or register to answer this question.

user contributions licensed under cc by-sa 3.0 with attribution required

Your rights