physics

General Relativity

This is a high-level introduction to the topic. It assumes familiarity with calculus and special relativity. It’s also very incomplete, I just thought the equations would look cool in the overall aesthetic of this website lol. Finally, it’s also incoherent because general relativity is hard and it would take a lot of effort to put it in a reasonable order, which I will do at some point!

Failure of Newtonian Physics

Newtonian gravitation states that the relationship between mass and the gravitational field it generates is

$$\nabla^2\Phi = 4\pi G\rho$$

The force is, as usual, the negative gradient. For two point masses, this comes out as

$$\begin{aligned} F_G&=-\nabla\Phi\&=-\frac{GMm}{4\pi r^2} \end{aligned}$$

We can calculate the field for an arbitrary mass distribution $\rho(\vec{r})$:

$$\Phi(\vec{r},t)=-G\int{\frac{\rho(\vec{r}’,t)}{\left\lvert\vec{r}-\vec{r}’\right\rvert}d^3r’}$$

However, this clashes with special relativity! If the $t$ on both sides were the same, changing the position of a particle somewhere would instantaneously change the field everywhere, violating the speed of causality.

However, we’ve dealt with something like this before! The electromagnetic wave equation is

$$\square \phi=-\frac{1}{c^2}\frac{\partial^2\phi}{\partial t^2}+\nabla^2\phi=-\frac{\rho}{\varepsilon_0}$$

With solution

$$\phi(\vec{r},t)=\frac{1}{\varepsilon_0}\int{\frac{\rho(\vec{r’}, t-R/c)}{R}d^3r’}, \quad R=\left\lvert\vec{r}-\vec{r}’\right\rvert$$

The times are no longer the same, and the signal travels at the speed of light!

Surely we know how to fix Newtonian gravity now: we just replace $\nabla^2$ with $\square$ to get

$$\square\Phi=4\pi\rho G$$

However, we get two problems:

  1. In EM, $\phi$ and $\rho$ are scalars, while we need $\Phi$ a scalar but $\rho$ a tensor
  2. EM fields do not generate themselves, but gravitational fields do, because of mass-energy equivalence! We have no mechanism for this yet!

We must look for a new theory.

Equivalence

The weak equivalence principle states that it is impossible to detect a gravitational field in a local inertial frame (or equivalently, the ratio of gravitational mass to inertial mass is a constant).

The strong form states that all physics is independent of whether you are accelerating or in a local gravitational field.

This has an interesting quantum point, where phase evolves differently if you’re in a field, but this is only measurable by using a comparison point outside the field, thus not breaking the principle!

Geodesic motion

Newton’s first law can be generalised. In local inertial coordinates, the equation of motion is simply: no acceleration.

$$\frac{d^2\xi^\alpha}{d\tau^2}=0$$

A generalisation of the special relativistic spacetime interval gives us

$$-c^2d\tau^2=\eta_{\alpha\beta}d\xi^\alpha d\xi^\beta$$

If we switch to general coordinates $x^\mu$ using

$$\frac{d\xi^\alpha}{d\tau}=\frac{d\xi^\alpha}{dx^\mu}\frac{dx^\mu}{d\tau}$$

Then the equation of motion becomes

$$0=\frac{d^2\xi^\alpha}{d\tau^2}=\frac{d}{d\tau}\left(\frac{d\xi^\alpha}{dx^\mu}\frac{dx^\mu}{d\tau}\right)=\frac{dx^\mu}{d\tau}\frac{d}{d\tau}\frac{d\xi^\alpha}{dx^\mu}+\frac{d\xi^\alpha}{dx^\mu}\frac{d^2 x^\mu}{d\tau^2}$$

After some wrangling, this simplifies to

$$0=\frac{d^2 x^\lambda}{d\tau^2}+\left(\frac{\partial x^\lambda}{\partial\xi^\alpha}\frac{\partial^2\xi^\alpha}{\partial x^\lambda \partial x^\nu}\right)\frac{dx^\mu}{d\tau}\frac{dx^\nu}{d\tau}$$

If we then define the affine connection, or the Christoffel symbol (the latter is a specialisation of the former to mainfolds with a metric),

$$\boxed{\Gamma^\lambda_{\mu\nu}=\frac{\partial x^\lambda}{\partial \xi^\alpha}\frac{\partial^2 \xi^\alpha}{\partial x^\mu\partial x^\nu}}$$

This becomes

$$\boxed{\ddot{x}^\lambda+\Gamma^\lambda_{\mu\nu}\dot{x}^\mu \dot{x}^\nu=0}$$

This is the equation of motion in a curved spacetime!