Session E30: Nanofluids: Computations II

Fluid flows in nano/micro network configurations: a multiscale molecular-continuum approach

We present a new hybrid molecular-continuum methodology for resolving multiscale flows emergent in nano-/micro-scale networks, in particular for NEMS/MEMS applications. The method models junction and channel components of the network using independent MD micro elements. Long channels with uniform or gradually varying nano-scale sections along the direction of flow, contribute the most towards the highest computational savings, by replacing them with much smaller MD simulations. Junction components, however, do not exhibit any length-scale separation and are modelled in their entirety. All micro elements are coupled together in one hybrid simulation using standard continuum fluid-dynamics equations, that dictate the overall macroscopic flow in the network. In the case of isothermal, incompressible, low-speed flows we use the conservative continuity and momentum equations. An iterative algorithm is presented that computes at each iteration the new constraints on the pressure differences applied to individual micro elements, in addition to enforcing overall continuity in the network. We show that the hybrid simulation of various small network cases converge quickly to the result of a full MD simulation over just a few iterations, with significant computational savings.   

Realistic molecular dynamics simulations of water flow through carbon nanotube membranes

Water transport through (7,7) carbon nanotubes (CNTs) is investigated using non-equilibrium molecular dynamics simulations. In particular, we are interested in how the CNT length and the prevalence of defects affects the internal flow dynamics. Pressure-driven water flow through CNTs ranging from 2.5 nm to 50 nm long is simulated. Structural defects are modeled as vacancy sites (missing carbon atoms). We demonstrate that under the same applied pressure difference an increase in CNT length has a negligible effect on the resulting mass flow rate and fluid flow velocity. Flow enhancements over hydrodynamic expectations are in fact directly proportional to the CNT length. Axial profiles of fluid properties demonstrate that CNT inlet and outlet effects are significant: large viscous losses in these regions contrast with central ``developed flow'' regions in longer CNTs where the flow is almost frictionless. Increasing the numbers of defects in the CNT structure does, however, lead to significant reductions in the fluid velocity and mass flow rate through the CNTs.   

Combined Temperature/Momentum Boundary Conditions for Molecular Dynamics Simulations of Flow in Nanofluidic Systems

Molecular dynamics (MD) is a useful technique for scientific investigations of nanofluidic processes as it explicitly represents the dynamics of every atom in a system. In order to model systems of interest, e.g. nanochannels, it is necessary to constrain atomic motions to conform to conditions corresponding to large-scale information, e.g. thermodynamic variables. However, many engineered configurations involve complex interactions between the system and its environment, and take place in non-trivial geometries. To accurately simulate these phenomena, methods to apply boundary conditions to MD systems are required that simultaneously regulate the temperature (i.e., energy) and momentum of the atoms in a local manner. This work uses an atomistic-to-continuum formulation to generate boundary conditions by using finite elements (FE) and their associated shape functions to define ``boundaries'' for a particle system. By projecting onto the FE basis, coarse-scale observables are identified for regulation based on separating the mean and fluctuating velocity components defining the momentum and temperature. Regulating the MD system is achieved by applying constraints posed on the coarse-grained variables. The method is illustrated by application to several nanofluidic systems.   

Comparing Molecular Dynamics Models for Electrolyte Solutions in Nanochannels

In electrolyte modelling, it is common to simplify the solvent using the three-component model (3CM), i.e.~a single-site, chargeless Lennard-Jones atom as the solvent component. To account for the dielectric nature of typical solvents, a relative permittivity value is applied to all Coulombic interactions, thus weakening ion-ion interactions as if each ion is surrounded by a solvation shell. Fluid Density Functional Theory, Monte Carlo simulation, and molecular dynamics (MD) simulation all commonly employ the 3CM to facilitate calculations, but the consequences are not well characterized. We used MD to compare the 3CM electrolyte to a molecular solvent model (MSM) where the solvent is a three-site H$_2$O) molecule. Special care was taken to compare cases with the same thermodynamic state by having a quantifiable reference state, and cases covered a range of applied surface charge in a nanochannel configuration. At a glance, the two models give qualitatively similar density profiles. However, we find that many profile features, physical quantities such as electric field and potential, as well as ionic packing structure near the surface evolve quite differently as the load is varied.   

Bouncing, splashing and disintegrating nanodrops

The impact of nanometer-sized drops on solid surfaces is studied by molecular dynamics simulations. The surfaces are atomically smooth, dry and non-wetting, and both volatile and non-volatile liquids are considered. At low impact velocities drops distort on contact but bounce off the surface and relax back to a spherical shape. At higher velocities drops form a prompt splash on impact and subsequently disintegrate, while at still higher velocities drops disintegrate immediately on impact. In contrast to macroscopic drops, the presence or absence of vapor plays no role at all in nanodrop splashing.   

Nanodrop impact on rough and textured surfaces

We use molecular dynamics simulations to investigate the impact of a nanometer-sized drop onto structured atomic surfaces. Rough surfaces with Gaussian or power-law correlations are constructed using a Fourier synthesis algorithm. At low impact velocity drops spread into a lamella, and we study its shape and maximum extension as a function of surface roughness and wettability. At higher impact velocities a prompt splash occurs, and we examine the effects of the surface and external vapor on the behavior of the lamella rim. We also consider the effect of surface wettability patterns on splashing and spreading, and compare the results to lattice-Boltzmann simulations in the same geometry.