DSMC Simulations of Nanoscale and Microscale Gas Flow

DSMC Simulations of Nanoscale and Microscale Gas Flow, Fig. 1 Flowchart of the DSMC algorithm... 

DSMC Simulations of Nanoscale and Microscale Gas Flow, Fig. 9 (a) Pressure distributions, and (b) pressure deviation from the linear distribution, along the channel for different values of wall heat flux rate... 

DSMC Simulations of Nanoscale and Microscale Gas and (d) Kn Flow, Fig. 4 Conductive Heat flux lines overlaid on the Kn = 0.1 temperature contour, top row NS, bottom row DSMC, (a)... 

However, the length and time scales that molecular-based simulations can probe are still very limited (tens of nanosecond and a few nanometers), due to computer memory and CPU power limitations. On the other hand, nanoscale flows are often a part of larger scale devices that could contain both nanochannels and microfluidic domains. The dynamics of these systems depends on the intimate connection of different scales from nanoscale to microscale and beyond. MD simulation cannot simulate the whole systems due to its prohibitive computational cost, whereas continuum Navier-Stokes simulation cannot elucidate the details in the small scales. These limitations and the practical needs arising from the study of multiscale problems have motivated research on multiscale (or hybrid) simulation techniques that bridge a wider range of time and length scales with the minimum loss of information. A hybrid molecular-continuum scheme can make such multiscale computation feasible. A molecular-based method, such as MD for liquid or DSMC for gas, is used to describe the molecular details within the desired, localized subdomain of the large system. A continuum method, such as finite element or finite volume based Navier-Stokes/Stokes simulation, is used to describe the continuum flow in the remainder of the system Such hybrid method can be applied to solve the multiscale phenomena in gas, liquid, or solid. 

---