Direct numerical simulation of nucleate boiling with a resolved microlayer and conjugate heat transfer

TL;DR

This work tackles the challenging DNS of nucleate boiling in the microlayer by extending an open-source phase-change framework to resolve conjugate heat transfer and microlayer dynamics. It combines a Volume-of-Fluid—based interface representation with a ghost-fluid method to stabilize an AMR-driven solver, achieving three orders of magnitude reduction in computational cost while maintaining stability and accuracy. The study validates against MIT pool-boiling experiments, demonstrates grid-convergent predictions, and reveals how the interfacial heat-transfer resistance and contact angle influence microlayer evaporation and bubble growth. Additionally, it reports the first open-literature complete bubble cycle DNS with a resolved microlayer and conjugate heat transfer, highlighting the solver’s potential for broader boiling problems and parametric studies.

Abstract

In this paper, a phase-change model based on a geometric Volume-of-Fluid (VOF) framework is extended to simulate nucleate boiling with a resolved microlayer and conjugate heat transfer. Heat conduction in both the fluid and solid domains is simultaneously solved, with Interfacial Heat-Transfer Resistance (IHTR) imposed. The present model is implemented in the open-source software Basilisk with adaptive mesh refinement (AMR), which significantly improves computational efficiency. However, the approximate projection method required for AMR introduces strong oscillations within the microlayer due to intense heat and mass transfer. This issue is addressed using a ghost fluid method, allowing nucleate boiling experiments to be successfully replicated. Compared to previous literature studies, the computational cost is reduced by three orders of magnitude. The influence of contact angle is further investigated, revealing consistent thermodynamic effects across different contact angles. Finally, a complete bubble cycle from nucleation to detachment is simulated, which, to our knowledge, has not been reported in the open literature. Reasonable agreement with experimental data is achieved, enabling key factors affecting nucleate boiling simulations in the microlayer regime to be identified, which were previously obscured by limited simulation time.

Figures (29)

• Figure 1: Schematic of nucleate boiling with a microlayer (not to scale).
• Figure 2: Schematic for the concept of equivalent conductive resistance: (a) IHTR located at the liquid-vapor interface. (b) IHTR located at the fluid-solid boundary.
• Figure 3: Schematic for the discretization of the diffusion terms in the energy equation.
• Figure 4: Schematic for the implicit discretization of the heat diffusion terms at the fluid-solid boundary.
• Figure 5: Film evaporation with conjugate heat transfer: (a) Schematic of the 1D film evaporation with conjugate heat transfer. (b) Temperature distribution at $t = 0.5$ s. (c) Time history of the interface position. (d) Relative error of the interface position on different grid resolutions. Grid levels 6 to 8 correspond to effective grid resolutions ranging from $64 \times 1$ cells to $256 \times 1$ cells, resulting in minimum grid sizes from $0.05$ to $0.0125$.