I am trying to calculate interactive force between two metal atoms and was wondering if there's any way VASP may be used to calculate vdW forces using higher order techniques such as RPA.
Accordng to the VASP version you are using, vdW non-empirical (i.e. not Grimme-like) corrections are possible. The most benchmarked one is the Tkatchenko-Scheffler method (version 5.3.3 and later). For a more comprehensive list and related keywords see: