What is Special Relativity?
Time and space are the same thing, and it's all relative?
- What is Special Relativity?
When we read What is LHC:ATLAS?, I mentioned that since the speed of the particles approaches the speed of light, we have to account for the relativistic effects.
"However, we do have Galilean relativity which has worked perfectly fine for me" you might inject, and you'd be right. Galilean relativity serves us just well in most of our lives. In fact, for speeds far below the speed of light Einstein relativity and Galilean relativity are nearly indistinguishable from each other. 
However, in certain circumstances such as when you're using a particle accelerator, or even something as mundane as using GPS, Galilean relativity is not enough, and you have to use Einstein's relativity.
"Okay, so I need this Einstein's relativity. What is it?" you ask begrudgingly.
Einstein's special relativity (refered to just as special relativity from here) can be derived entirely from two fundamental postulates.
The Fundamental Postulates
- All intertial reference frames are equivalent.
This version of the postulate, however, follows a somewhat circular argument as an object which does not accelerate does not have a net force, however, the only way we know that it doesn't have a net force is by checking if the object accelerates.
This inconsistency lead Einstein to develop his general theory of relativity which solves the problem by working for any reference system, and not just inertial systems. However, if you ignore this detail, you'll be fine.
- The speed of light in vacuum, , is a universal constant in all reference frames.
While today, we know this to be true as we have measured it empirically, Einstein couldn't possibly have known this at his time, you rebut. And you'd be right. In fact, it's possible to derive that given that light is a electromagnetic wave that light must travel at from Maxwell's equations. However, Einstein describes in his notes from 1949 himself finding it as a solution to an apparent paradox he stumbled upon when he was 16. This approach leads to the stronger claim that is constant in all reference frames.
The standard configuration is a configuration of coordinate systems often used in special relativity to ease, and somewhat normalize the notation.
This setup is illustrated below.
Since handling the individual dimensions separately can quickly become a mess, it's an often used convention to combine the dimensions into a four vector with the unit vectors:
The index of the time dimension can wary between authors. One convention is to use the zeroth dimension as the time dimension, so the unit vectors become
which gives us the four vector
By using Einstein notation, we get
where it's assumed we sum over repeated greek letters.
We have adopted the convention where we denote four vectors with bold face such as, , while we continue to denote three vectors with an arrow such as, .
For some more syntatic sugar, let
In which case
The reason we have a , and not in the time dimension is to avoid messing up causality which the philosophers tell me is a good thing.
One neat thing about using four vectors is that the dot product between any two four vectors (or a chain of them) is invariant. So
In the standard configuration, the componentwise Lorentz transformation is given by
While deriving every formula of special relativity is beyond the scope of this project, deriving time contraction seems worthwhile due to it's broad applicability.
Let's consider a light-clock consisting of two perfect mirrors put parallel with a distance to each other, and let a foton travelling at speed bounce between the mirrors.
Obviously, the time it takes for the foton to bounce from the bottom mirror to the top mirror and back again is going to be
because time equals distance over speed.
This can be rewritten as
Now, let's consider the same system, but this time the mirrors move along the -axis with velocity while the foton continous to bounce between the mirrors.
Let the time it takes for the foton to complete one full bounce be , and let the distance the mirrors have moved before the foton comes back be . Then we know that
So we can write .
We know that the foton travels along twice for a total distance of , so we find that
We now have all the sides in a right-angled triangled, and per pythagoras, we know that
Which gives us
By dividing by , we find
Isolating , we find
Or when ,
Just like classcal mechanics, we still have conservation of momemtum and total energy in relativistic collisions.
Example: Compton Spread
We consider a system consisting of a photon and an electron. In the initial state, the photon has a four momentum vector , and the electron is at rest, .
In the final state, the photon has the four momentum vector , and the electron has the four momentum vector .
The setup is illustrated below.
This is an example of Compton spread. The angle of spreading depends on the wavelength of the incoming photon with the relation
This chapter has leaned on the content from "Speciel Relativitetsteori" by Ulrik Uggerhøj whenever a source has not been cited.
Note: While there'll always be a small factor of difference between Einstein and Galilean relativity, for small velocities, the difference will be within the margin of error. ↩︎
We use special and not general relativity because it's simpler, and sufficient for our purposes, and also because I'm taking a course in special, and not general relativity. ↩︎
Actually, we're so sure about being an universal constant that we have defined the meter to be derived from which has the interesting consequence that every time we attempt to 'measure' , we are actually just measuring the accuracy of the equipment. ↩︎
There are several different ways of measuring including some which can be done at home. One interesting way of measuring is by using the aberration of light first discovered by James Bradley who found that light travels approximately 10,210 times faster than the Earth around the sun. ↩︎
Note: While this is done in dimensions, the exact same principles apply in dimensions; you just tag along a axis. ↩︎
Note: The prime symbol () has nothing to do with differentiation in special relativity. Except when it does, but in those rare cases the author usually marks it clearly to avoid ambiguity. ↩︎