International Communications in Heat and Mass Transfer 88 (2017) 136138

Contents lists available at ScienceDirect

International Communications in Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ichmt

A thermal immiscible multiphase ow simulation by lattice Boltzmann MARK

method
Wei Gonga, Sheng Chena, Yuying Yana,b,
a
Fluids & Thermal Engineering Group, Faculty of Engineering, University of Nottingham, University Park, Nottingham, United Kingdom
b
Center for Fluids & Thermal Engineering, University of Nottingham Ningbo, China

A R T I C L E I N F O A B S T R A C T

Keywords:
Multi-bubble/droplet coexistence
Unphysical mass transfer
Thermal immiscible multiphase ow
Pseudopotential lattice Boltzmann method
The lattice Boltzmann (LB) method, as a mesoscopic approach based on the kinetic theory, has been signicantly
developed and applied in a variety of elds in the recent decades. Among all the LB community members, the
pseudopotential LB plays an increasingly important role in multiphase ow and phase change problems simu-
lation. The thermal immiscible multiphase ow simulation using pseudopotential LB method is studied in this
work. The results show that it is dicult to achieve multi-bubble/droplet coexistence due to the unphysical mass
transfer phenomenon of the big eat the small the small bubbles/droplets disappear and the big ones getting
bigger before a physical coalescence when using an internal energy based temperature equation for single-
component multiphase (SCMP) pseudopotential models. In addition, this unphysical eect can be eectively
impeded by coupling an entropy-based temperature eld, and the inuence on density elds with dierent
energy equations are discussed. The ndings are identied and reported in this paper for the rst time. This work
gives a signicant inspiration for solving the unphysical mass transfer problem, which determines whether the
SCMP LB model can be used for multi-bubble/droplet systems.

1. Introduction where G , w , e and t are the parameter controlling the interaction

As an unconventional numerical method which is based on the
mesoscopic kinetic equation, the lattice Boltzmann (LB) method [14]
has drawn considerable attention and been applied to a wide range of
elds such as phase change [5], wetting phenomena [6,7], porous
media ow [8], electroosmotic driven ow [9] and ow with moving
boundaries [10], etc. The pseudopotential LB method, which was rstly
presented by Shan and Chen [11,12], has become one of the most im-
portant and popular LB community members due to its conceptual
simplicity and computational eciency [4]. In this model, the pseu-
dopotential is introduced to mimic the interaction forces for the
neighbouring particles, which renders an automatic phase separation
without a specic tracking or capturing method for the interface be-
tween dierent phases. Owing to this particular phase separation
characteristic, the pseudopotential LB has been well applied in multi-
phase ow and phase change.
The original interaction force of single-component multiphase
(SCMP) models acting on a particle at the position x from the nearest-
neighbouring particles is given by [11]
F (x , t ) = G (x ) w (x + e t ) e
(1)
strength, the weights, the lattice speed vector and the lattice time step,
respectively. Besides the use of SCMP models, multi-component mul-
tiphase (MCMP) models are also widely applied, and the interaction
force can be written as [12]
F (x , t ) = (x ) G w (x + e t ) e
(2)
where and denote the corresponding components. It has been re-
ported that generally multi-bubble/droplet congurations cannot be
formed when performing a conventional rst order multiphase pseu-
dopotential model [3], which impedes an extended application of LB in
a number of elds such as atomization, spray formation and phase
segregation. Our study nds that the MCMP models do not suer such a
problem, and some multi-bubble/droplet ow studies from other re-
searchers using MCMP models can be found in reference [1318].
However, the multi-component models have not drawn as much at-
tention as the single-component models because of the density ratio and
viscosity issues [3].
Another major nding in our study is that the selection of a proper
energy equation for thermal LB models can be advantageous for multi-
bubble/droplet systems as well. For thermal LB models, without

Corresponding author at: Fluids & Thermal Engineering Group, Faculty of Engineering, University of Nottingham, University Park, Nottingham, United Kingdom.
E-mail address: yuying.yan@nottingham.ac.uk (Y. Yan).

http://dx.doi.org/10.1016/j.icheatmasstransfer.2017.08.019

0735-1933/ 2017 Elsevier Ltd. All rights reserved.

W. Gong et al. International Communications in Heat and Mass Transfer 88 (2017) 136138

considering the viscous heat dissipation and the compression work, the
internal energy equation is given by [4]
t (c v T ) + (c v Tv ) = (T ) (3)
where is the density, cv is the specic heat capacity, T is the tem-
perature, v is the velocity vector, is thermal conductivity and t de-
notes the time. Furthermore, if cv is a constant, the internal energy
equation can be reduced to
1
t T + vT = (T )
c v (4)
Li Q. used another energy equation in their model [5]

1 T pEOS
t T + vT = (T ) v
c v c v T (5)
where pEOS is the non-ideal equation of state, and this energy equation
is derived from the entropy equation neglecting the viscous heat dis-
Fig. 2. The Maxwell construction with Ts = 0.86Tc for Ping-Robinson EOS.
sipation
Ds
T = (T ),
Dt (6)
with the thermodynamic relation
p 1
Tds = c v dT + T EOS d

T (7)
where s represents the entropy. In addition, in their work it has also
been reported that the thermal lattice scheme can produce a signicant
error by the forcing term, thus in this work all the thermal equation are
solved by the classical fourth-order Runge-Kutta scheme [19].
Through studying thermal LB models using the two energy Eqs. (4)
and (5), it is found that it is much more dicult to form a multi-bubble/
droplet conguration when applying Eq. (4). The fact of unachievable
multi-bubble/droplet conguration is due to an unphysical mass
transfer phenomenon which we call the big eat the small. As shown in
Fig. 1, the small bubbles/droplets become smaller and even disappear
over time while the big ones getting bigger without a physical coales- Fig. 3. The small bubble/droplets variations with dierent energy equations.
cence simulated using the basic Shan-Chen SCMP model [12]. It should
be pointed out that an isothermal system is equal to a thermal LB model conditions are set as Zou-He solid boundaries for the top and the bottom
along with the energy Eq. (4), which means the isothermal SCMP and the periodical boundaries for the left and the right to close a
models also suer the same unphysical mass transfer problem. 200 100 computational region. For the thermal simulation, the top
and bottom boundaries are constant temperature Ts and Tb with a small
2. Numerical simulation temperature dierence T = Tb Ts = 0.001 for a simple heat transfer
case. It should be noticed that it is the energy equation selection that
The original Shan-Chen model [12], an improved Bhatnagar-Gross- matters in our study other than the heat transfer cases. First, a single
Krook (BGK) model with the given code in reference [20], the Gong- bubble with diameter 40 is simulated to test the coexistence of liquid/
Cheng model [21], an improved multiple-relaxation-time (MRT) model vapour densities by the Maxwell construction with the saturation
from Li Q [22] and the midrange repulsion model [23], all of which temperature Ts = 0.86Tc, where Tc is the critical temperature. The co-
involve various forcing schemes, non-ideal equation of states and col- existence densities and the Maxwell construction using Ping-Robinson
lision forms, are all tested in our study. The result shows none of the EOS [24] are shown in Fig. 2. From the picture it can be seen the si-
model can get rid of the unphysical mass transfer when coupling the mulation result is in good agreement with the theory thermodynamic
energy Eq. (4). The following numerical simulation and discussion are consistency.
based on the Li Q. improved MRT model and the details of the model Bubble systems and droplet systems are tested with an initial larger
can be found in reference [5]. Following their model, the boundary

Fig. 1. Multi-bubble conguration evolution in a

200 150 computational domain with periodical boundary
condition (a)t = 200t (b) t = 400t (c) t = 800t.

137
W. Gong et al.
Fig. 4. The density variation in the cross section with dierent energy equations.
bubble/droplet with diameter 40 and a small one with diameter 20
using the two dierent energy equations. The radius variation curves of
the small bubble/droplets are shown in Fig. 3. From the gure it can be
seen that the unphysical mass transfer is much severer in bubble sys-
tems than droplet systems, and the mass transfers faster when the
bubble/droplet volume ratio increases. Besides, the mass transfer only
occurs from the smaller bubble/droplets to the larger ones, but not vice
versa. In addition, the problem can be eectively restrained by coupling
the entropy-based energy equation.
Finally in Fig. 4 the density elds for the two dierent energy
equations are compared. The main dierence of the two density elds is
that when conducting Eq. (5), one layer of high density region appears
outside of the bubble/droplet interface, where for Eq. (4) it should be
the same with liquid/vapour density. This spurious density layer caused
the interface nonmonotonic. If the unphysical mass transfer is caused by
the spurious currents, this high-density layer might be the reason that
disturbs the spurious currents and thereby restrains the phenomenon of
the big eat the small. Nevertheless, the real reason and the way to
eliminate the unphysical phenomenon for SCMP models still need to be
investigated in the future work.
3. Conclusions
In the present study, we report our new ndings of a critical pro-
blem of the conventional thermal/isothermal SCMP pseudopotential LB
models. It is the rst time this type of problem is identied and re-
ported. After testing a number of typical models involving various
forcing schemes, non-ideal equation of states and the relaxation
schemes of SCMP LB models with the internal energy based energy
equation, we found the multiple bubbles/droplets could not coexist due
to the unphysical mass transfer phenomenon, or the big eat the small.
Meanwhile, this unphysical problem can be eectively restrained with
an entropy-based energy equation. Although the mass transfer problem
cannot be solved in this work, our study points out a reasonable or-
ientation for the further investigation. Form our study the following
conclusions can be drawn:
(1) The unphysical mass transfer phenomenon exists in most of the
thermal SCMP pseudopotential LB models with the presented in-
ternal energy based energy equation, of which the isothermal si-
tuation is included.
(2) The mass transfer only occurs from the smaller bubble/droplets to
the larger ones, but not vice versa.
(3) The unphysical mass transfer is much severer in bubble systems
than droplet systems, and the mass transfers faster when the
138
International Communications in Heat and Mass Transfer 88 (2017) 136138
bubble/droplet volume ratio increases.
(4) The unphysical mass transfer can be eectively impeded by cou-
pling an entropy-based energy equation.
(5) A high-density layer outside of the bubble/droplet is produced by
coupling the entropy-based energy equation, which might be the
reason to restrain the unphysical phenomenon.
Acknowledgement
The authors would like to acknowledge the nancial support of this
work by the doctoral degree scholarship of China Scholarship Council
(CSC) for Mr Wei GONG and the University of Nottingham, UK, for
Yuying Yan and Wei Gong, and this work is also supported by Ningbo
Science and Technology Bureau Technology Innovation Team project
under Grant No. 2016B10010.
References
[1] S. Chen, G.D. Doolen, Lattice Boltzmann method for uid ows, Annu. Rev. Fluid
Mech. 30 (1998) 329364.
[2] M.R. Swift, W. Osborn, J. Yeomans, Lattice Boltzmann simulation of nonideal uids,
Phys. Rev. Lett. 75 (1995) 830.
[3] L. Chen, Q. Kang, Y. Mu, Y.-L. He, W.-Q. Tao, A critical review of the pseudopo-
tential multiphase lattice Boltzmann model: methods and applications, Int. J. Heat
Mass Transf. 76 (2014) 210236.
[4] Q. Li, K.H. Luo, Q.J. Kang, Y.L. He, Q. Chen, Q. Liu, Lattice Boltzmann methods for
multiphase ow and phase-change heat transfer, Prog. Energy Combust. Sci.
(2015), http://dx.doi.org/10.1016/j.pecs.2015.10.001.
[5] Q. Li, Q.J. Kang, M.M. Francois, Y.L. He, K.H. Luo, Lattice Boltzmann modeling of
boiling heat transfer: the boiling curve and the eects of wettability, Int. J. Heat
Mass Transf. 85 (2015) 787796.
[6] W. Gong, Y. Zu, S. Chen, Y. Yan, Wetting transition energy curves for a droplet on a
square-post patterned surface, Sci. Bull. 62 (2017) 136142.
[7] W. Gong, Y. Yan, S. Chen, D. Giddings, Numerical study of wetting transitions on
biomimetic surfaces using a lattice Boltzmann approach with large density ratio, J.
Bionic Eng. 14 (2017) 486496.
[8] S. Chen, B. Yang, C. Zheng, A lattice Boltzmann model for heat transfer in porous
media, Int. J. Heat Mass Transf. 111 (2017) 10191022.
[9] B. Li, W.N. Zhou, Y.Y. Yan, C. Tian, Evaluation of electro-osmotic pumping eect on
microporous media ow, Appl. Therm. Eng. 60 (2013) 449455.
[10] Y.Y. Yan, Y.Q. Zu, Numerical simulation of heat transfer and uid ow past a ro-
tating isothermal cylinder a LBM approach, Int. J. Heat Mass Transf. 51 (2008)
25192536.
[11] X. Shan, H. Chen, Simulation of nonideal gases and liquid-gas phase transitions by
the lattice Boltzmann equation, Phys. Rev. E 49 (1994) 29412948.
[12] X. Shan, H. Chen, Lattice Boltzmann model for simulating ows with multiple
phases and components, Phys. Rev. E (47) (1993) 18151819.
[13] B. Sehgal, R. Nourgaliev, T. Dinh, Numerical simulation of droplet deformation and
break-up by lattice-Boltzmann method, Prog. Nucl. Energy 34 (1999) 471488.
[14] M. Nekovee, P.V. Coveney, H. Chen, B.M. Boghosian, Lattice-Boltzmann model for
interacting amphiphilic uids, Phys. Rev. E 62 (2000) 8282.
[15] F. Drer, M. Rauscher, J. Koplik, J. Harting, S. Dietrich, Micro- and nanoscale uid
ow on chemical channels, Soft Matter 8 (2012) 9221.
[16] A. Gupta, R. Kumar, Lattice Boltzmann simulation to study multiple bubble dy-
namics, Int. J. Heat Mass Transf. 51 (2008) 51925203.
[17] Z. Yu, H. Yang, L.-S. Fan, Numerical simulation of bubble interactions using an
adaptive lattice Boltzmann method, Chem. Eng. Sci. 66 (2011) 34413451.
[18] Q. Zhang, D. Sun, Y. Zhang, M. Zhu, Lattice Boltzmann modeling of droplet con-
densation on superhydrophobic nanoarrays, Langmuir 30 (2014) 1255912569.
[19] H. Liu, A.J. Valocchi, Y. Zhang, Q. Kang, Phase-eld-based lattice Boltzmann nite-
dierence model for simulating thermocapillary ows, Phys. Rev. E 87 (2013)
013010.
[20] H. Huang, M. Sukop, X. Lu, Multiphase Lattice Boltzmann Methods: Theory and
Application, John Wiley & Sons, 2015.
[21] S. Gong, P. Cheng, Numerical investigation of droplet motion and coalescence by an
improved lattice Boltzmann model for phase transitions and multiphase ows,
Comput. Fluids 53 (2012) 93104.
[22] Q. Li, K.H. Luo, X.J. Li, Lattice Boltzmann modeling of multiphase ows at large
density ratio with an improved pseudopotential model, Phys. Rev. E Stat. Nonlinear
Soft Matter Phys. 87 (2013) 053301.
[23] S. Chibbaro, G. Falcucci, G. Chiatti, H. Chen, X. Shan, S. Succi, Lattice Boltzmann
models for nonideal uids with arrested phase-separation, Phys. Rev. E Stat.
Nonlinear Soft Matter Phys. 77 (2008) 036705.
[24] P. Yuan, L. Schaefer, Equations of state in a lattice Boltzmann model, Phys. Fluids
18 (2006) 042101.