How to accurately predict nanoscale flow: Theory of single-phase or two-phase?

• Published in: Physics of fluids (1994) 2023-01, Vol.35 (1)
• Main Authors: Sun, Chengzhen, Zhou, Runfeng, Bai, Bofeng
• Format: Article
• Language: English
• Subjects: Adsorption; Annular flow; Bulk density; Flow velocity; Fluid flow; Fluid mechanics; Molecular dynamics; Single-phase flow; Two phase flow; Velocity distribution
• ISSN: 1070-6631; 1089-7666
• DOI: 10.1063/5.0134510

Accurate evaluation and recognition of nanoscale flow is the premise of the extension of classical theories of fluid mechanics to nanoscales. Despite the widely reported nonuniform characteristics of nanoconfined fluids, nanoscale flow is still considered as a single-phase flow in general, resulting in large deviations in theoretical predictions of velocity profile and flow rate. Considering the significant characteristics of a two-phase flow in nanoscales and the similarity between nanoscale flow and gas–liquid two-phase annular flow, we put forward a novel viewpoint that nanoscale flows should be described based on the theory of a two-phase flow. To support this idea, nanoscale flows under different fluid types, densities, temperatures, fluid–solid interactions, and driving pressures are extensively tested using molecular dynamics simulations. The results demonstrate that nanoscale flows can be divided into an adsorption phase and a bulk phase, and the characteristics of a two-phase flow are especially obvious under low fluid density, strong fluid–solid interaction, and low fluid temperature. The reasonability is further demonstrated by systematically analyzing the interphase density difference, interphase velocity difference, interphase mass exchange, and interfacial fluctuation, which are typical characteristics of a two-phase flow at macroscales. Finally, we present a series of theoretical descriptions of nanoscale flow from the perspective of a two-phase flow. By adopting different viscosity and density in the adsorption phase and bulk phase, the new model can better capture the physical details of nanoscale flow, such as velocity distribution and flow rate.