A high-fidelity computational fluid dynamics (CFD)-based model for bubble nucleation of the refrigerant HFE7100 on micrometer-featured surfaces is presented in this work. The single-fluid, incompressible Navier-Stokes equations, along with energy transport and natural convection effects, are solved on a featured surface resolved grid. An a priori cavity detection method is employed to convert raw profilometer data of a surface into well-defined cavities. The cavity information and surface morphology are represented in the CFD model by geometric mesh deformations. Surface morphology is observed to initiate buoyancy-driven convection in the liquid phase, which in turn results in faster nucleation of cavities. Simulations pertaining to a generic rough surface show a trend where smaller size cavities nucleate with higher wall superheat. This local-scale model will serve as a self-consistent connection to larger device scale continuum models where local feature representation is not possible.

This content is only available via PDF.
You do not currently have access to this content.