稀薄气体流动中气-固界面作用过程的数值模拟及物理模型研究

Accurately simulating and modeling the gas-surface interaction process is an indispensable part for the rarefied gas flow study, which is highly concerned in the modern industry involving high-altitude, high-speed, vacuum and micro-scale. Combining molecular dynamics simulation and simplified physical model provides an effective way to study and model the gas-surface interaction process. This project will investigate the features of gas-surface interaction process in typical applications of rarefied gas flows by molecular dynamics simulation, and improve the existing physical model to provide an accurate and feasible boundary conditions for rarefied gas flow simulation. Molecular dynamics simulation will reveal two important features of gas-surface interaction process: the correlation between the incident and reflected velocities, and the dependence of momentum/energy accommodation coefficients on the gas incident condition, which provides guidance for developing more accurate boundary conditions; Based on existing simplified physical models, by modifying the thermal velocity distribution of the effective solid cube and ensuring the time reversibility of consecutive collisions and attractive potential well effect, we can improve the model to satisfy the reciprocity relation and establish a simplified physical model complying with the basic constraints of boundary condition; We will study the relation between the effective parameters in the simplified model with the inherent system properties. Reliable methods to obtain the effective parameters from the system properties will be explored. It will make the simplified physical model parameter-free and greatly improve the feasibility of the model.

准确模拟并描述气-固界面作用过程是研究稀薄气体流动不可或缺的重要一环，备受高空、高速、真空、微尺度等现代工业领域的关注。将分子动力学模拟和简化物理模型相结合，是研究并描述该过程的有效途径。本项目采用分子动力学方法研究典型稀薄气体流动应用中的气-固界面作用机理，并改进已有的简化物理模型，为模拟稀薄气体流动提供准确而实用的边界条件。分子动力学模拟气-固界面作用过程将揭示气体分子入射与反射速度的相关性、以及动量和能量适应系数随入射条件的变化特性，为发展更优的边界条件提供指导；基于已有的简化物理模型，将调整等效固体块热振动速度分布，并保证连续碰撞和吸引势井作用的时间可逆性，使模型满足互易性条件，建立符合边界基本约束条件的简化物理模型；研究简化物理模型中所含等效参数与系统固有属性间的关联，探索从系统中直接获得等效参数的可靠方法，从而使简化物理模型不再含有任何未知参数，提高模型的可用性。

发展准确且实用的气-固界面边界条件，是稀薄气体流动研究领域备受关注的科学问题。本项目以发展适用于全流域的气-固界面作用简化物理模型为目标，立足于国际上边界条件领域的最新进展，结合分子动力学模拟和物理建模方法，对气-固界面作用的微观机理和规律开展仿真模拟、理论分析和模型构建。研究揭示了传统气-固界面唯象模型（如Maxwell和Cercignani-Lampis模型）在散射核函数和适应系数等方面的诸多缺陷，以及其对Knudsen扩散流等典型稀薄气体流动问题的显著影响；通过构建新的局部碰撞核和可逆碰撞轨迹计算方法，研究建立了一个新型的气-固界面作用简化物理模型，使其满足所有边界基本约束条件，即非负性、归一化以及互易性条件，可以作为稀薄气体流动模拟的边界条件使用；在准确性方面，该物理模型克服了传统唯象边界模型的缺陷；在实用性方面，该物理模型仅包含三个输入参数，研究探索了这些输入参数的确定方法，从系统的基本属性直接获得表面粗糙度、表面吸引强度和局部碰撞能量适应系数，使得此物理模型中不含有任何未知参数；开发了实现该物理模型的程序包，可与开源DSMC软件SPARTA兼容，用于模拟大规模稀薄气体流动问题。典型算例显示，本项目发展的气-固界面作用物理模型在准确性、计算效率和实用性方面非常具有优势。.本项目研究成果，已发表一作SCI论文2篇，包括国际计算流体力学领域顶级期刊Journal of Computational Physics 1篇和Journal of Applied Physics 1篇；另有合作发表Nature Nanotechnology 1篇和正在撰写的SCI论文1篇。获批软件著作权1项。本项目已完成项目任务书的预定研究内容，达到预期成果要求。