Direct Numerical Simulation (DNS) of boiling flows requires a substantial number of grid points to precisely resolve the thermal boundary layer surrounding the phase interface. This resolution is vital for accurate calculation of the temperature gradient, which directly influences the mass transfer rate. However, the thermal boundary layer is typically three orders of magnitude smaller than the bubble's diameter, leading to an impractical number of required grid points for computational resources available on affordable PCs or workstations.

To address this challenge and enable bubble-growth boiling flow simulations without relying on a supercomputer, we propose a novel numerical method within the framework of the Volume-Of-Fluid (VOF) approach. This method employs a coarse grid, where the grid size may exceed the thickness of the thermal boundary layer. By adopting this approach, we aim to achieve accurate simulation results while reducing the computational requirements associated with grid resolution.

In our coarse grid approach, we model the thermal boundary layer and “artificially” maintain the interface temperature above the saturation temperature in the solution of the temperature field by incorporating a temperature-profile sharpening coefficient ${\mathcal{K}}_s$. To validate the effectiveness of this approach, we conducted two validation cases: the Scriven bubble-growth problem and an experimental measurement of single bubble growth on a heated surface. Encouragingly, both cases showed good agreement with the simulation results. In the latter case, we introduced additional subgrid scale models, i.e., a microlayer model and a model of contact-angle hysteresis. These models enabled us to evaluate the bubble force balance accurately. The comprehensive approach described here represents an advancement in the development of sharp-interface phase-change simulation methods that can be applied to larger-scale problems and parametric investigations.

沸腾流的直接数值模拟 (DNS) 需要大量网格点来精确解析相界面周围的热边界层。该分辨率对于精确计算温度梯度至关重要，温度梯度直接影响传质速率。然而，热边界层通常比气泡直径小三个数量级，导致经济实惠的 PC 或工作站上可用的计算资源所需的网格点数量不切实际。

为了应对这一挑战并在不依赖超级计算机的情况下实现气泡生长沸腾流模拟，我们在流体体积（VOF）方法的框架内提出了一种新颖的数值方法。该方法采用粗网格，其中网格尺寸可能超过热边界层的厚度。通过采用这种方法，我们的目标是获得准确的模拟结果，同时减少与网格分辨率相关的计算要求。

在我们的粗网格方法中，我们对热边界层进行建模，并通过结合温度分布锐化系数，“人为地”将界面温度维持在温度场解中的饱和温度之上${\mathcal{K}}_s$。为了验证这种方法的有效性，我们进行了两个验证案例：Scriven 气泡生长问题和加热表面上单个气泡生长的实验测量。令人鼓舞的是，这两种情况都与模拟结果非常吻合。在后一种情况下，我们引入了额外的亚网格尺度模型，即微层模型和接触角滞后模型。这些模型使我们能够准确评估气泡力平衡。这里描述的综合方法代表了锐界面相变模拟方法开发的进步，可应用于更大规模的问题和参数研究。