Register to reply

What is the angular-momentum 4-vector?

by Zarathustra0
Tags: 4vector, angularmomentum
Share this thread:
Zarathustra0
#1
May10-11, 02:51 PM
P: 23
Uh, the title pretty much says it: I'm wondering what the 4-vector analog to the classical 3-angular momentum is. Also, is the definition

L = r [tex]\times[/tex] p

still valid for the 3-angular momentum in special relativity?
Phys.Org News Partner Science news on Phys.org
Fungus deadly to AIDS patients found to grow on trees
Canola genome sequence reveals evolutionary 'love triangle'
Scientists uncover clues to role of magnetism in iron-based superconductors
Bill_K
#2
May10-11, 03:38 PM
Sci Advisor
Thanks
Bill_K's Avatar
P: 4,160
In four dimensions, angular momentum is no longer represented by a vector. The position is a vector, xμ. The momentum is a vector, pν. But when you take the cross product you get a rank two tensor, Jμν = xμ pν - pμ xν.

Why does this happen in four dimensions? Well actually the cross product is supposed to do that. It's only in three dimensions that we can replace it with a vector.

Jμν has six components. If you look at three of them in which μ and ν are both space indices, Jij = xi pj - pi xj, then these values correspond to the three components of what we call the angular momentum vector: Lx = Jyz, Ly = Jzx, and Lz = Jxy.
Zarathustra0
#3
May10-11, 04:37 PM
P: 23
Thanks for the help--I assume then that this tensor transforms by applying the Lorentz-transformation matrix twice?

tiny-tim
#4
May10-11, 04:55 PM
Sci Advisor
HW Helper
Thanks
tiny-tim's Avatar
P: 26,148
Smile What is the angular-momentum 4-vector?

Hi Zarathustra0! Welcome to PF!

to transform it, it's easiest to write it as a 2-form rΛp, with the six (antisymmetric) basis elements tΛx tΛy tΛz yΛz zΛx and xΛy

so yes you apply the Lorentz transformation matrix twice to each basis element (of course, it only makes a difference to t and x)
Zarathustra0
#5
May10-11, 10:52 PM
P: 23
To make sure I understand (and I may not because the notions of forms and wedge products are still a bit new to me), does this mean I should express the tensor as a linear combination of the tensors

[0 1 0 0]
[-1 0 0 0]
[0 0 0 0]
[0 0 0 0],

[0 0 1 0]
[0 0 0 0]
[-1 0 0 0]
[0 0 0 0],

[0 0 0 1]
[0 0 0 0]
[0 0 0 0]
[-1 0 0 0],

[0 0 0 0]
[0 0 1 0]
[0 -1 0 0]
[0 0 0 0],

[0 0 0 0]
[0 0 0 1]
[0 0 0 0]
[0 -1 0 0], and

[0 0 0 0]
[0 0 0 0]
[0 0 0 1]
[0 0 -1 0],

perform the Lorentz transformation on each by two applications of the transformation matrix, and then add up the resulting six matrices? In other words, do the above tensors form the basis in question? Earlier I naively tried to just apply the transformation matrix twice to the J-tensor itself, but the result wasn't antisymmetric, so I don't think it worked.

Then again, given the nature of matrix multiplication, it seems like writing the matrix as a linear combination of other matrices and multiplying each of them by the square of the transformation matrix would have the same effect as just multiplying the square of the transformation matrix directly with the original angular-momentum tensor itself, and I already tried that, so maybe I'm still not understanding.
tiny-tim
#6
May11-11, 03:32 AM
Sci Advisor
HW Helper
Thanks
tiny-tim's Avatar
P: 26,148
Hi Zarathustra0!
Quote Quote by Zarathustra0 View Post
To make sure I understand (and I may not because the notions of forms and wedge products are still a bit new to me), does this mean I should express the tensor as a linear combination of the tensors

perform the Lorentz transformation on each by two applications of the transformation matrix, and then add up the resulting six matrices?
I don't recommend trying to mix up tensors and 2-forms.

Use one or the other.

2-forms, in my opinion, are easier.

There isn't much call for writing out the Minkowski angular momentum but the similar electromagnetic field is regularly used

we can write it either as the antisymmetric tensor (matrix) 0 By -Bz Ex etc (which takes up four lines, and doesn't say anything intuitive), or as a neat 2-form

ExtΛx + EytΛy + EztΛz + BxyΛz + ByzΛx + BzxΛy

but once you get used to the idea, you can just write that 2-form as
(Ex,Ey,Ez,Bx,By,Bz),
and you'll know what that means just as you know what (E,px,py,pz) means without having to write the basis vectors!
(try writing out the Lorentz force expression F* = q(t,x,y,z)Λ(Ex,Ey,Ez,Bx,By,Bz), to get a feel for how it all works )
Zarathustra0
#7
May11-11, 07:50 AM
P: 23
OK, in that case to apply the transformation twice to each basis element, should I simply make the substitutions

ct' --> [tex]\gamma[/tex]ct - [tex]\beta\gamma[/tex]x,

x' --> -[tex]\beta\gamma[/tex]ct + [tex]\gamma[/tex]x,

y' --> y, and

z' --> z

twice to the six basis elements?
tiny-tim
#8
May11-11, 08:00 AM
Sci Advisor
HW Helper
Thanks
tiny-tim's Avatar
P: 26,148
yes!

so for example x'Λt' = ?

(btw, you may find it easier to use the rapidity α, where tanhα = v = β, and coshα = γ)
Zarathustra0
#9
May11-11, 08:22 AM
P: 23
OK, cool--thanks for all the help!

Actually I think I have one last (perhaps trivial) question. In the position 4-vector (in r [tex]\wedge[/tex] p), should the temporal component ct be 0? This seems intuitive since we're just talking about two points in space (the position of the orbiting particle and the center of rotation) and not events at different times.
tiny-tim
#10
May11-11, 10:40 AM
Sci Advisor
HW Helper
Thanks
tiny-tim's Avatar
P: 26,148
I can't remember.

I can't remember ever seeing a useful application of rΛp !
DrGreg
#11
May11-11, 06:32 PM
Sci Advisor
PF Gold
DrGreg's Avatar
P: 1,847
Quote Quote by Zarathustra0 View Post
In the position 4-vector (in [itex]r \wedge p[/itex]), should the temporal component ct be 0?
No, it needs to be ct so that r is a genuine 4-vector (= rank-1 tensor) and hence the angular-momentum is a genuine rank-2 tensor. Note that t contributes to the other 3 components of the tensor that don't correspond to angular momentum (see post 2).
Sam Gralla
#12
May12-11, 10:35 AM
P: 95
The time-space components of the angular momentum tensor give the center of mass of the system (minus t times the momentum, if you are in a frame with non-zero momentum). You can think of the angular momentum tensor as a way of combining the ordinary angular momentum and the center of mass into a single 4D object, much as the relativistic energy-momentum vector combines energy and momentum into a single 4D object.


Register to reply

Related Discussions
Heisenberg Uncertainty Relations - angular momentum and angular displacement Advanced Physics Homework 1
Vector products -Finding the direction of angular momentum Advanced Physics Homework 7
Physics Rotating Disk, Moment of Inertia, Angular Momentum, Angular Velocity Introductory Physics Homework 12
Angles made by angular momentum vector with magnetic field Introductory Physics Homework 2
Angular momentum and orbital angular momentum problems Introductory Physics Homework 3