Cluster-flip colloidal and atomistic algorithms with background potentials

We introduce an extension to cluster algorithms of colloidal and atomistic models that naturally incorporates nonuniform background potentials. Simulations of these systems are important for studying the statistical mechanics of fluids and myriad liquid, solid, and glassy phase transitions. Nonuniform background potentials like gravitational or trapping fields are typically present in experiments and can reveal new features of the statistical mechanics at play due to, e.g., nonuniform chemical potential. This method takes existing cluster methods and incorporates background potentials without adding a rejection step by treating the orientation of the system as a dynamical degree of freedom. We assess its efficiency in several applications and explore connections with the celebrated swap Monte Carlo for glasses.