Specific Heat

[evanberkowitz]

I think the specific heat $c = (\left\langle U^2 \right\rangle - \left\langle U \right\rangle^2)/\Lambda = \Lambda \times$InternalEnergyDensityVariance. Here's evidence that the N-->∞ limit is converging.

[evanberkowitz]

I'm used to thinking about the Ising model, where this observable has a sharp peak at the critical J^-1. Is there an obvious reason there's no equivalent peak here?

[evanberkowitz]

The answer of course is that if there were a peak that would violate the no-local-order-parameters property of BKT. So it's good that it's smooth!

[evanberkowitz]

Actually, the specific heat (density) is $c = (\langle U^2 \rangle - \langle U \rangle^2 ) / \Lambda T^2$, which will wind up being $c = (\langle U^2 \rangle - \langle U \rangle^2 ) \times \kappa^2 / \Lambda $.

When W≠1 we do have real global symmetry breaking, and so maybe there should actually be a spike, were we to multiply the above figure by κ^2.