Thermostating highly confined fluids

Abstract

In this work we show how different use of thermostating devices and modeling of walls influence the mechanical and dynamical properties of confined nanofluids. We consider a two dimensional fluid undergoing Couette flow using nonequilibrium molecular dynamics simulations. Because the system is highly inhomogeneous, the density shows strong fluctuations across the channel. We compare the dynamics produced by applying a thermostating device directly to the fluid with that obtained when the wall is thermostated, considering also the effects of using rigid walls. This comparison involves an analysis of the chaoticity of the fluid and evaluation of mechanical properties across the channel. We look at two thermostating devices with either rigid or vibrating atomic walls and compare them with a system only thermostated by conduction through vibrating atomic walls. Sensitive changes are observed in the xy component of the pressure tensor, streaming velocity, and density across the pore and the Lyapunov localization of the fluid. We also find that the fluid slip can be significantly reduced by rigid walls. Our results suggest caution in interpreting the results of systems in which fluid atoms are thermostated and/or wall atoms are constrained to be rigid, such as, for example, water inside carbon nanotubes.

Water flow in carbon nanotubes: The effect of tube flexibility and thermostat

“…[37][38][39][40] The velocities in the flexible tubes are significantly higher than those in the rigid tubes, which can be due to several reasons as follows: (i) The vibration of wall atoms about their lattice sites smoothens the potential energy surface felt by the water molecules on the wall surface. (ii) Fluid molecules can slip past the flexible wall more favorably than past the rigid wall, since the flexible atoms can move backward due to steric repulsion to facilitate the flow, while the rigid wall atoms lead to specular reflection which slows down the flow, 18 as schematically depicted in Fig. 2. (iii) Momentum transfer from the excited phonon modes of the CNTs to the fluid has been suggested to contribute to the larger flow rates of water in flexible CNTs.…”

“…16 A study on polymer melts sheared by Lennard-Jones walls found that the slip length increased with the shear rate when the walls were kept rigid, while the slip length was independent of the shear rate for flexible walls. 17 Bernardi et al 18 observed significant differences in temperature, density, velocity, and stress profiles between the two thermostatting approaches (i.e., flexible or rigid walls) and advocated thermostatting walls and caution in interpreting the slip obtained from NEMD simulations. Yong and Zhang 19 simulated the Couette flow of a Lennard-Jones fluid by thermostatting only the fluid, only the walls, or both, and they compared different thermostatting algorithms.…”

“…In fact, the different thermostatting approaches were suggested to be partially responsible for the large variation in the reported water flow rates and slip lengths in CNTs. 18,21,[23][24][25] The large variation suggests that these properties may be very sensitive to the simulation details, making it a suitable test case to study the influence of temperature control on nanoconfined fluid transport, with implications also for other nonequilibrium systems.…”

Non-equilibrium phase behavior and friction of confined molecular films under shear: A non-equilibrium molecular dynamics study

“…[51][52][53] Thermostatting the wall atoms rather than the confined sample is the most physically realistic option. The implementation of bulk system thermostats without modification may lead to spurious particle dynamics.…”

Molecular dynamics pre-simulations for nanoscale computational fluid dynamics

“…The temperature is maintained at T = 292.8 K by coupling molecules to a velocity-unbiased Berendsen thermostat (Berendsen et al 1984) with a time constant of s T ¼ 21:6 fs, applied within 36 independent bins placed in the y-direction, with each bin being 0.34 nm thick. It has been shown that using a thermostat on a confined fluid may affect the flow properties (Bernardi et al 2010); however, a thermostat is used in all the MD simulations in this work so that we can compare with isothermal CFD simulations. The same thermostat is used in both the pre-simulations and the full MD simulations; therefore, the same error is present in both-the verification of the simulation approach is thus not undermined by any physical uncertainty introduced by the thermostat.…”