Invaraince

[Spook]

HI guys first post :cool:

I need to show that

B^2-E^2/C^2 is invariant under Lorentz transformation (E and B are electromagnetic fields)

now:

B^2-E^2/C^2=B^2_x+B^2_y+B^2_z-E^2_x/C^2-E^2_y/C^2-E^2_z/C^2)

and

E'_x=E_x
E'_y=\gamma(E_y-\frac{v}{c}B_z)
E'_z=\gamma(E_z-\frac{v}{c}B_y)
B'_x=B_x
B'_y=\gamma(B_y+\frac{v}{c}E_z)
B'_z=\gamma(B_z+\frac{v}{c}E_y)

but i cant manupilate it to give me the correct answer ie

B'^2-E'^2/C^2=B^2_x+B^2_y+B^2_z-E^2_x/C^2-E^2_y/C^2-E^2_z/C^2

Can anyone help me out? Basically because of the \gamma^2 term im tring to factorise out a 1-\frac{v^2}{C^2} ie (1/\gamma^2) but im having no joy.

[turin]

First of all, you can make things less complicated by rotating the coordinate system so that E = exEx (or, just as well, eyEy or ezEz). (EDIT: I TAKE IT BACK; NO YOU SHOULDN'T DO THIS.) I'll give it a more thorough look to see if I can find your problem.

The first suspicion I have is that the problem is stated in a different unit system (SI) than you are using in your solution (Gaussian?). In the SI units, E has the same units as cB (or E/c as B), but your transformed components add E to (v/c)B, so the units don't seem like they agree.

I get a (+) sign in front of the B term in my transformed Ez'. I think this makes sense, because the orientation of z to y is the opposite of the orientation of y to z (wrt rotation about the x axis).

[Spook]

Dont make it anymore complicated than it needs to be :p. Its a direct substitution problem ie change E to E' and B to B' and using the transformed values show that B^2-E^2/C^2=B'^2-E'^2/C^2

[robphy]

In my experience, identities involving \gamma and v are more recognizable if you use the hyperbolic functions.

Write v = \tanh\theta and \gamma =\cosh\theta, where \theta is the rapidity. Of course, \gamma v = \sinh\theta.

\exp\theta has physical significance, as well.

[DW]

HI guys first post :cool:

I need to show that

B^2-E^2/C^2 is invariant under Lorentz transformation (E and B are electromagnetic fields)

now:

B^2-E^2/C^2=B^2_x+B^2_y+B^2_z-E^2_x/C^2-E^2_y/C^2-E^2_z/C^2)

and

E'_x=E_x
E'_y=\gamma(E_y-\frac{v}{c}B_z)
E'_z=\gamma(E_z-\frac{v}{c}B_y)
B'_x=B_x
B'_y=\gamma(B_y+\frac{v}{c}E_z)
B'_z=\gamma(B_z+\frac{v}{c}E_y)

but i cant manupilate it to give me the correct answer ie

B'^2-E'^2/C^2=B^2_x+B^2_y+B^2_z-E^2_x/C^2-E^2_y/C^2-E^2_z/C^2

Can anyone help me out? Basically because of the \gamma^2 term im tring to factorise out a 1-\frac{v^2}{C^2} ie (1/\gamma^2) but im having no joy.

A couple of your equations have wrong signs(and wrong factors of c depending on your choice of units). They should be

E'_x=E_x
E'_y=\gamma(E_{y}-vB_{z})
E'_z=\gamma(E_{z}+vB_{y})
B'_x=B_x
B'_y=\gamma(B_y+\frac{v}{c^2}E_z)
B'_z=\gamma(B_z-\frac{v}{c^2}E_y)

Adjust the signs and try again.

[Spook]

Yes thanks I forgot to report back that I had done it successfully. The annoying thing was I actually looked up the equations on a website so assumed they were correct.

[DW]

Yes thanks I forgot to report back that I had done it successfully. The annoying thing was I actually looked up the equations on a website so assumed they were correct.

LOL. Out of curiosity what was the website? About assuming website physics as accurate, check out the link
http://www.crank.net/physics.html