3
$\begingroup$

In discrete models like Ising we have Hamiltonians of the form

$$H(\sigma)=\frac{1}{N}\sum_{i=1}^{N}J_{ij}\sigma_{i}\sigma_{j},$$

where $\sigma_{i}=\pm 1$ , $J_{ij}$ interaction coefficients and N is size of system.

So I was wondering if anyone has studied the following Hamiltonian

$$\sum_{i=1}^{N}J_{ij}e^{ix_{i}}e^{ix_{j}},$$

where $x_{i}\in [0,2\pi]$. So the phase space is $\mathbb{T}^{N}$. Then we can ask similar questions, like what is the Gibb's measure at $\mathbb{T}^{\infty}$ and so the distribution of the free energy $\frac{1}{N}ElogZ_{N}$.

Note that I am not taking a dot product like in the Heisenberg model.

Also, note the interesting picture of the Hamiltonian $\sum_{i=1}^{3}e^{ix_{i}}e^{ix_{j}}$. Eventually it fills the entire 3-disk, but I thought it was pretty cool seeing those cluster points.

enter image description here

Any references will be greatly appreciated.

$\endgroup$
2
  • 6
    $\begingroup$ your "Hamiltonian" is not real-valued, so how are you defining Gibbs measure? I also cannot figure out what your picture represents. $\endgroup$
    – j.c.
    Commented Dec 9, 2015 at 15:00
  • $\begingroup$ Except that your Hamiltonian is not real, this feels like some (possibly frustrated) $O(2)$ compass model. $\endgroup$ Commented Dec 12, 2015 at 18:40

0

You must log in to answer this question.