{"title":"Comparisons between full molecular dynamics simulation and Zhang’s multiscale scheme for nanochannel flows","authors":"Chuntao Jiang, Yongbin Zhang","doi":"10.1007/s00894-024-06115-8","DOIUrl":null,"url":null,"abstract":"<div><h3>Context</h3><p>In a very small surface separation, the fluid flow is actually multiscale consisting of both the molecular scale non-continuum adsorbed layer flow and the intermediate macroscopic continuum fluid flow. Classical simulation of this flow often takes over large computational source and is not affordable owing to using molecular dynamics simulation (MDS) to model the adsorbed layer flow, if the flow field size is on the engineering size scale such as of 0.01–10 mm or even bigger like occurring in micro or macro hydrodynamic bearings. The present paper uses full MDS to validate Zhang’s multiscale flow model, which yields the closed-form explicit flow equations respectively for the adsorbed layer flow and the intermediate continuum fluid flow. Here, full MDS was carried out for the pressure-driven flow of methane in the nano slit pore made of silicon respectively with the channel heights 5.79 nm, 11.57 nm, and 17.36 nm. According to the number density distribution, the flow areas were respectively discriminated as the adsorbed layer zone and the intermediate fluid zone. The values of the characteristic parameters for Zhang’s multiscale scheme were extracted from full MDS and input to Zhang’s multiscale flow equations respectively for calculating the flow velocity profile and the volume flow rates of the adsorbed layers and the intermediate fluid. It was found that for these three channel heights, the flow velocity profiles calculated from Zhang’s model approximate those calculated from full MDS, while the total flow rates through the channel calculated from Zhang’s model are close to those calculated from full MDS. The accuracy of Zhang’s multiscale flow model is improved with the increase of the channel height.</p><h3>Method</h3><p>The recent modification of the optimized potential for liquid simulation (MOPLS) model was used to calculate the interaction force between two methane molecules. In order to calculate the interaction force between the wall atoms and the methane molecules accurately, our previous non-equilibrium multiscale MDS was used. The interaction forces between the methane molecule and the wall atom were obtained from the coupled potential function by the L-B mixing rule when the fluid molecules arrived at near wall. The methane molecule diameter was obtained from the radial distribution function by using equilibrium MDS under the same initial conditions. The local viscosities across the adsorbed layer were obtained from the local velocity profile by using the Poiseuille flow method. The motion equation of the methane molecule was solved by the leapfrog method. The temperature of the simulation system was checked by Bhadauria’s method, i.e., the system temperature was rectified by the velocities in the <i>y</i>- and <i>z</i>-directions. The flow velocity distributions across the channel height and the volume flow rates through the channel were also calculated from Zhang’s closed-form explicit flow equations respectively for the adsorbed layer flow and the intermediate fluid flow. The results respectively obtained from full MDS and Zhang’s multiscale flow equations were then compared.</p></div>","PeriodicalId":651,"journal":{"name":"Journal of Molecular Modeling","volume":null,"pages":null},"PeriodicalIF":2.1000,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Molecular Modeling","FirstCategoryId":"92","ListUrlMain":"https://link.springer.com/article/10.1007/s00894-024-06115-8","RegionNum":4,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"BIOCHEMISTRY & MOLECULAR BIOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Context
In a very small surface separation, the fluid flow is actually multiscale consisting of both the molecular scale non-continuum adsorbed layer flow and the intermediate macroscopic continuum fluid flow. Classical simulation of this flow often takes over large computational source and is not affordable owing to using molecular dynamics simulation (MDS) to model the adsorbed layer flow, if the flow field size is on the engineering size scale such as of 0.01–10 mm or even bigger like occurring in micro or macro hydrodynamic bearings. The present paper uses full MDS to validate Zhang’s multiscale flow model, which yields the closed-form explicit flow equations respectively for the adsorbed layer flow and the intermediate continuum fluid flow. Here, full MDS was carried out for the pressure-driven flow of methane in the nano slit pore made of silicon respectively with the channel heights 5.79 nm, 11.57 nm, and 17.36 nm. According to the number density distribution, the flow areas were respectively discriminated as the adsorbed layer zone and the intermediate fluid zone. The values of the characteristic parameters for Zhang’s multiscale scheme were extracted from full MDS and input to Zhang’s multiscale flow equations respectively for calculating the flow velocity profile and the volume flow rates of the adsorbed layers and the intermediate fluid. It was found that for these three channel heights, the flow velocity profiles calculated from Zhang’s model approximate those calculated from full MDS, while the total flow rates through the channel calculated from Zhang’s model are close to those calculated from full MDS. The accuracy of Zhang’s multiscale flow model is improved with the increase of the channel height.
Method
The recent modification of the optimized potential for liquid simulation (MOPLS) model was used to calculate the interaction force between two methane molecules. In order to calculate the interaction force between the wall atoms and the methane molecules accurately, our previous non-equilibrium multiscale MDS was used. The interaction forces between the methane molecule and the wall atom were obtained from the coupled potential function by the L-B mixing rule when the fluid molecules arrived at near wall. The methane molecule diameter was obtained from the radial distribution function by using equilibrium MDS under the same initial conditions. The local viscosities across the adsorbed layer were obtained from the local velocity profile by using the Poiseuille flow method. The motion equation of the methane molecule was solved by the leapfrog method. The temperature of the simulation system was checked by Bhadauria’s method, i.e., the system temperature was rectified by the velocities in the y- and z-directions. The flow velocity distributions across the channel height and the volume flow rates through the channel were also calculated from Zhang’s closed-form explicit flow equations respectively for the adsorbed layer flow and the intermediate fluid flow. The results respectively obtained from full MDS and Zhang’s multiscale flow equations were then compared.
期刊介绍:
The Journal of Molecular Modeling focuses on "hardcore" modeling, publishing high-quality research and reports. Founded in 1995 as a purely electronic journal, it has adapted its format to include a full-color print edition, and adjusted its aims and scope fit the fast-changing field of molecular modeling, with a particular focus on three-dimensional modeling.
Today, the journal covers all aspects of molecular modeling including life science modeling; materials modeling; new methods; and computational chemistry.
Topics include computer-aided molecular design; rational drug design, de novo ligand design, receptor modeling and docking; cheminformatics, data analysis, visualization and mining; computational medicinal chemistry; homology modeling; simulation of peptides, DNA and other biopolymers; quantitative structure-activity relationships (QSAR) and ADME-modeling; modeling of biological reaction mechanisms; and combined experimental and computational studies in which calculations play a major role.