0
Research Papers: Micro/Nanoscale Heat Transfer

# An Atomistic Simulation Study of the Role of Asperities and Indentations on Heterogeneous Bubble Nucleation

[+] Author and Article Information
Brian R. Novak, Edward J. Maginn, Mark J. McCready

Department of Chemical and Biomolecular Engineering,  University of Notre Dame, Notre Dame, IN 46556

J. Heat Transfer 130(4), 042411 (Mar 21, 2008) (9 pages) doi:10.1115/1.2818771 History: Received February 22, 2007; Revised July 16, 2007; Published March 21, 2008

## Abstract

Heterogeneous bubble nucleation was studied on surfaces having nanometer scale asperities and indentations as well as different surface-fluid interaction energies. Nonequilibrium molecular dynamics simulations at constant normal stress and either temperature or heat flux were carried out for the Lennard–Jones fluid in contact with a Lennard–Jones solid. When surface defects were of the same size or smaller than the estimated critical nucleus (the smallest nucleus whose growth is energetically favored) size of $1000–2000Å3$, there was no difference between the defected surfaces and atomically smooth surfaces. On the other hand, surfaces with significantly larger indentations had nucleation rates that were about two orders of magnitude higher than the systems with small defects. Moreover, nucleation was localized in the large indentations. This localization was greatest under constant heat flux conditions and when the solid-fluid interactions were weak. The results suggest strategies for enhancing heterogeneous bubble nucleation rates as well as for controlling the location of nucleation events.

<>

## Figures

Figure 1

A necessary condition for trapping vapor in a crevice when a surface is flooded with liquid is that the angle between the liquid-vapor interface, α, is greater than the crevice angle, θ(12).

Figure 2

The configuration of a heterogeneous simulation box. The light colored atoms are the fluid atoms and the dark colored ones are solid. The z directions were bounded by wall potentials or a fixed layer of solid, while periodic boundary conditions were used in the x and y directions.

Figure 3

The defected solids with small defects. The defects were approximately 1.5nm wide and two layers deep.

Figure 4

The solid used in the large indentation simulations. The indentation dimensions were approximately 4.5×4.5×3nm. The upper picture is a top view and the lower picture is a diagonal cross section.

Figure 5

Diagonal cross section of the solid for the large system with constant heat flux. The bottom layer is fixed, kinetic energy is added to the second to the bottom layer each time step (49), and the other layers have just microcanonical dynamics.

Figure 6

Box height as a function of time for a simulation of the large indentation with a weak surface using constant heat flux, and the third order polynomial from a fit to the data from 0to756ps. The nucleation time was taken as the first time when the data deviated from the polynomial by more than 3.5Å.

Figure 7

Nucleation rates (m−3s−1) versus inverse temperature (K−1) on a weakly attractive solid surface. Dotted lines are to guide the eye. A small indentation, asperity, and a flat surface with the same simulation box size were simulated. A large indentation was also simulated. The nucleation rates were not significantly different when comparing the small defects to a flat surface, but the rate for the large indentation was about two orders of magnitude greater than for the small defects at 131.0K. The error bars are 68.3% confidence intervals.

Figure 8

Relative void probability for the large indentation as a function of x and y averaged over the z direction, over multiple simulations, and over time up until a void percolated the box. The top figure is for the constant temperature case (131.0K, weak surface, 30 simulations). The bottom left figure is for the constant heat flux case (149MW∕m2, 122.0K starting temperature, 39 simulations) with a weak surface. The bottom right figure is for the constant heat flux case (149MW∕m2, 122.0K starting temperature, 39 simulations) with a neutral surface. In all three cases, void formation was favored in the indentation whose approximate boundaries are shown by the white line. The dark regions in the lower left and upper right corners are due to the fact that one point in those corners was set to either 0 or 1 to get the scale correct.

Figure 9

These figures represent the ways in which atoms can collapse into a 2-D void on a weakly attractive surface. Fluid atoms are represented as circles, and the solid surface as a line. The arrows point into the void. The arc length that can collapse into a void of the same radius and contact angle is smaller in the indentation (right) compared with the flat part of the surface (left). The volume of a critical nucleus in the indentation is also smaller for a given critical radius.

Figure 10

Time average until a void had a volume greater than 1000Å3 and simulation average over 39 simulations of local temperature (T) minus bulk fluid temperature (Tfluid) as a function of z position for the large indentation with constant heat flux. Two separate regions were considered, above the indentation and above the flat part of the surface. In both the neutral and weak cases, the temperature in the indentation was considerably higher than in the rest of the fluid.

Figure 11

Nucleation times as a function of the negative logarithm of the ratio of non-nucleated (N) to total simulations (N0) for all constant heat flux cases. The times for the neutral cases were longer because the adsorbed layers had to be heated up before nucleation could occur. Nucleation was slightly faster for the flat surface compared to the indented one for the neutral case and there was little difference for the weak case.

Figure 12

These figures represent the ways in which atoms can collapse into a 2-D void on a neutral surface. Fluid atoms are represented as light gray (yellow) and dark gray (green) circles, and the solid surface as a solid line. The “edge” of the adsorbed fluid layers is the dashed line. The arrows point into the void. Nucleation cannot occur in the adsorbed fluid layers between the solid line and the dashed line. These layers effectively reduce the the size of the indentation and make the nucleation more like homogeneous nucleation, since the light gray (yellow) atoms in the adsorbed fluid layers can collapse into the void, but are not as likely to collapse as the dark gray (green) bulk fluid atoms since they are not as mobile.

## Discussions

Some tools below are only available to our subscribers or users with an online account.

### Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related Proceedings Articles
Related eBook Content
Topic Collections