Modeling of partial differential equations is an art that leaves the scientists with a choice of what is deemed important, e.g., mathematical well-posendess, conservation laws, thermodynamic consistency, … ,or just simplicity - in multiphysics settings this calls for appropriate modeling strategies. In this talk, a thermodynamics inspired strategy for the modeling of free boundary flows with moving contact lines will be revisited. It will be argued that this mathematical structure is not just useful for modeling but also for the numerical approximation of solutions. Furthermore, possible strategies for their extension to non-smooth settings will be shortly motivated, e.g., hysteresis and particulate flows.