Bogoliubov transformation 3-mode

  • Thread starter Thread starter Naibaf
  • Start date Start date
  • Tags Tags
    Transformation
Naibaf
Messages
3
Reaction score
0
Hi everyone,

I'm working on spin-wave theory and I have a problem with a bogoliubov transformation.
I must do the transformation with 3 bosons and i have no idea how to do it.
I've only found the transformation for 1 and 2-mode bosons, but not for three...

It exist?

Thanks...



And sorry for my english...
 
Physics news on Phys.org
Maybe you could provide us with your hamiltonian?
 
The hamiltonian is very long, but the idea is that I have, products of bosons operator, for example: a*a + a*b + a*c + b*b + bb + b*c and so on...

Where a, b an c are boson operators.
As there are "non-diagonal" terms, the idea is use Bogoliubov Transformation for make the hamiltonian a diagonal hamiltonian... But for 2 differents bosons, the tranformation is:

alpha_{k} = cosh(theta_{k}) a_{k} - sinh(theta_{k}) b*_{k}, where b* is b-dagger

ans this transformation is like "universal" because everyone use it.
So my question is if there is a known bogoliubov transformation for 3 differents bosons in a hamiltonian...
( I hope you understand my english :) )
 
I can certainly tell you that a solution exists, at least in the fermionic case, because the Bogoliubov transformation is used routinely in nuclear physics, where you can have ~100 protons and neutrons. Sorry not to be able to supply the actual solution you want :-)
 
In this case, the 3 bosons operators are different, I have: a, b and c.
I guess that for ~100 differents protons the transformation must be huge!

I also guess that the bogoliubov transformation for fermions should not be so different from that of bosons......at least I have some hope! :)
thanks
 
So, introducing the vector A=(a_k, a^*_{-k}, b_k, b^*_{-k},c_k, c^*_{-k})^T,
you should be able to write your hamiltonian as H=A^\dagger M(k) A. Then you have to find the unitary transformation U which diagonalizes the matrix M, i.e., you have to solve the eigenvalue problem for M. Then the vector of the alphas is UA.
Not a big deal, in principle.
 
Back
Top