Hostname: page-component-7b7567c8c-z4zkx Total loading time: 0 Render date: 2026-11-16T06:12:17.625Z Has data issue: false hasContentIssue false

Binary collision dynamics of equal-sized nanodroplets

Published online by Cambridge University Press:  11 January 2024

Yi-Feng Wang
Affiliation:
State Key Laboratory of Alternate Electrical Power System withd Renewable Energy Sources, North China Electric Power University, Beijing 102206, PR ChinaResearch Center of Engineering Thermophysics, North China Electric Power University, Beijing 102206, PR China
Yi-Bo Wang
Affiliation:
State Key Laboratory of Alternate Electrical Power System withd Renewable Energy Sources, North China Electric Power University, Beijing 102206, PR ChinaResearch Center of Engineering Thermophysics, North China Electric Power University, Beijing 102206, PR China
Zhi-Hui Cai
Affiliation:
State Key Laboratory of Alternate Electrical Power System withd Renewable Energy Sources, North China Electric Power University, Beijing 102206, PR ChinaResearch Center of Engineering Thermophysics, North China Electric Power University, Beijing 102206, PR China
Qiang Ma
Affiliation:
State Key Laboratory of Alternate Electrical Power System withd Renewable Energy Sources, North China Electric Power University, Beijing 102206, PR ChinaResearch Center of Engineering Thermophysics, North China Electric Power University, Beijing 102206, PR China
Yan-Ru Yang
Affiliation:
State Key Laboratory of Alternate Electrical Power System withd Renewable Energy Sources, North China Electric Power University, Beijing 102206, PR ChinaResearch Center of Engineering Thermophysics, North China Electric Power University, Beijing 102206, PR China
Shao-Fei Zheng*
Affiliation:
State Key Laboratory of Alternate Electrical Power System withd Renewable Energy Sources, North China Electric Power University, Beijing 102206, PR ChinaResearch Center of Engineering Thermophysics, North China Electric Power University, Beijing 102206, PR China
Duu-Jong Lee
Affiliation:
Department of Mechanical Engineering, City University of Hong Kong, Kowloon Tong 999077, Hong KongDepartment of Chemical Engineering & Materials Science, Yuan Ze University, Chung-Li 32003, Taiwan
Xiao-Dong Wang*
Affiliation:
State Key Laboratory of Alternate Electrical Power System withd Renewable Energy Sources, North China Electric Power University, Beijing 102206, PR ChinaResearch Center of Engineering Thermophysics, North China Electric Power University, Beijing 102206, PR China
*
†Email addresses for correspondence: shaofeizheng56@gmail.com, wangxd99@gmail.com
†Email addresses for correspondence: shaofeizheng56@gmail.com, wangxd99@gmail.com

Abstract

Binary nanodroplet collisions have received increasing attention, whilst the identification of collision outcomes and the viscous dissipation mechanism have remained poorly understood. Using molecular dynamics simulations, this study investigates binary nanodroplet collisions over wide ranges of Weber number (We), Ohnesorge number (Oh) and off-centre distances. Coalescence, stretching separation and shattering are identified; however, bouncing, reflexive separation and rotational separation reported for millimetre-sized collisions are not observed, which is attributed to the enhanced viscous effect caused by the ‘natural’ high-viscosity characteristics of nanodroplets. Intriguingly, as an intermediate outcome, holes form in retracting films at relatively high We, arising from the vibration and thermal fluctuation of the films. Due to the combined effects of inertial, capillary and viscous forces, binary nanodroplet collisions fall into the cross-over regime, so estimating viscous dissipation becomes extremely important for distinguishing outcome boundaries. Based on the criterion that stretching separation is triggered only when the residual off-centre kinetic energy exceeds the surface energy required for separation, the boundary equation between coalescence and stretching separation is established. Here, viscous dissipation is calculated by the extracted flow feature from simulations, showing that the ratio of viscous dissipation to the initial kinetic energy depends only on Oh, not on We. Because of complex viscous dissipation mechanisms, the same boundary equation in the cross-over regime has also not been satisfactorily revealed for macroscale collisions. Therefore, the proposed equation is tested for wide data sources from both macroscale and nanoscale collisions, and satisfying agreement is achieved, demonstrating the universality of the equation.

JFM classification

Information

Type
JFM Papers
Copyright
© The Author(s), 2024. Published by Cambridge University Press
Figure 0

Figure 1.(a) A phase diagram of binary droplet collisions (Oh = 0.0044 and D0 = 0.6 mm) in an atmospheric environment (Pan et al. 2019). (b) Schematics of outcomes, including coalescence (CO), bouncing (BO), reflexive separation (RES), stretching separation (SS), rotational separation (ROS) and shattering (SH). The Weber numbers in the original figure of Pan et al. (2019) are four times the Weber numbers in panel (a) because the relative velocity, 2V0, between two droplets is used by Pan et al. (2019) for calculating Weber numbers, whereas the velocity, V0, of one droplet is considered in this work.

Figure 1

Figure 2.Schematic of the simulation system containing two nanodroplets, where two views for observing the dynamic behaviour of nanodroplets are shown.

Figure 2

Figure 3.Phase diagram of binary nanodroplet collisions with Oh = 0.35, including coalescence (CO), stretching separation (SS) and shattering (SH). It should be emphasised that the hole is an intermediate outcome marked by a red circle based on the final outcomes (CO, SS and SH).

Figure 3

Figure 4.Snapshots of head-on binary nanodroplet collisions at We = (a) 24, (b) 54, (c) 74 and (d) 109. The numbers shown in panel (d) at t = 30 and 40 ps highlight the positions of jetted fingers. It is worth noting that when the thickness of the peripheric rim is larger than the one at the centre of liquid films, only the outline of the rim can be seen in the front view. To show the relative thickness between the rim and the centre of liquid films, supplemented sliced snapshots near the beginning of forming holes for panel (c,d) are shown in figure S6(e,f).

Figure 4

Figure 5.Snapshots of off-centre binary nanodroplet collisions at We = (a) 24, (b) 54, (c) 74 and (d) 109.

Figure 5

Figure 6.(a,b) Phase diagrams of binary nanodroplet collisions with Oh = (a) 0.39 and (b) 0.45, where the solid line represents the proposed scaling law. (c,d) Boundary data of the stretching separation regime from (c) Al-Dirawi et al. (2021) and Sommerfeld & Pasternak (2019), and (d) Sommerfeld & Kuschel (2016). (e) Values of prefactor (C) obtained from figures 3, 6(a–d) and figure S8 at various Oh, where the values of C regressed from figures 3 and 6(a,b) are marked by black squares, and the value regressed from figure S8 is marked by a green square. Here, the dashed line represents a constant value of 3.8, and the solid line represents the predicted value by the scaling law of C ~ (1 − α)−1, where the prefactor of this scaling is adopted as 1, and α is obtained by the model of βmax, i.e. (3.20).

Figure 6

Figure 7.(a) Schematics of a head-on collision process and a symmetry plane, where, due to the symmetry, the collision can be considered as a nanodroplet impacting the imaginary symmetry plane, bringing the convenience of modelling βmax. (b–d) Velocity contours of head-on binary nanodroplet collisions at We = 85 at t = (a) 8, (b) 13 and (c) 18 ps. The left side shows the velocity component in the impact direction, Vz, whereas the right side represents the velocity component in the spreading direction, Vr.

Figure 7

Figure 8.(a–e) Comparisons between theoretical results and MD data from both current and previous studies.

Figure 8

Figure 9.(a) Results of Edis varying with Ek,0 at different Oh calculated by (3.20) and the corresponding linear fitting results. (b) Comparison between the predicted α by (3.20) and the measured α in previous experimental studies, where the dashed line presents a constant value of α in the inertial regime and the solid line shows the predicted α with the method shown in panel (a).

Figure 9

Figure 10.(a) Comparison among the MD data for impacting nanodroplets on solid surfaces with θ = 180°, the scaling law of βmax ~ We1/2 with a prefactor of 0.26 and the proposed model with Cvf = 0.85. (b,c) Comparison between the proposed model with Cvf = 0.85 and the MD data of βmax for nanodroplets on solid surfaces with (b) θ ≥ 125° and (c) θ < 125°. In addition, an experimental data point is also considered for comparison in panel (b).

Figure 10

Figure 11.(a,b) Phase diagrams at gas pressures of (a) 120 and (b) 450 kPa. The comparison of βmax between MD simulation results and theoretical predictions at gas pressures of (c) 120 and (d) 450 kPa.

Figure 11

Figure 12.Front and side views of binary droplet collisions, where the off-centre volume (Voo) and the total volume of one droplet (Vot) are shown schematically.

Figure 12

Figure 13.Numerical results of Po at various B with different numerical mesh sizes (Dm/D0) of 0.004, 0.01 and 0.2, where Dm denotes the side length of each cubic mesh. Using polynomial fitting, the expression of Po is obtained as Po = −0.5B3 + 1.5B.

Supplementary material: File

Wang et al. supplementary material

Wang et al. supplementary material
Download Wang et al. supplementary material(File)
File 1.9 MB