Phase transitions in 3D Ising model with cluster weight by Monte Carlo method

TL;DR

This study investigates phase transitions in a 3D cluster weight Ising model using Monte Carlo simulations, revealing different universality classes and phase transition orders depending on the cluster weight parameter.

Contribution

It introduces a cluster weight extension to the 3D Ising model and explores its phase diagram and universality classes through advanced Monte Carlo methods.

Findings

Universalities differ between models on 3D lattice.

Phase transition type depends on cluster weight parameter.

Estimated critical exponents and phase diagram details.

Abstract

A cluster weight Ising model is proposed by introducing an additional cluster weight in the partition function of the traditional Ising model. It is equivalent to the O($n$) loop model or $n$-component face cubic loop model on the two-dimensional lattice, but on the three-dimensional lattice, it is still not very clear whether or not these models have the same universality. In order to simulate the cluster weight Ising model and search for new universality class, we apply a cluster algorithm, by combining the color-assignation and the Swendsen-Wang methods. The dynamical exponent for the absolute magnetization is estimated to be $z=0.45(3)$ at $n=1.5$, consistent with that of the traditional Swendsen-Wang methods. The numerical estimation of the thermal exponent $y_t$ and magnetic exponent $y_m$, show that the universalities of the two models on the three-dimensional lattice are different. We obtain the global phase diagram containing paramagnetic and ferromagnetic phases. The phase transition between the two phases are second order at $1\leq n< n_c$ and first order at $n\geq n_c$, where $n_c\approx 2$. The scaling dimension $y_t$ equals to the system dimension $d$ when the first-order transition occurs. Our results are helpful in the understanding of some traditional statistical mechanics models.