Professional Documents
Culture Documents
To cite this article: Athanasios G. Kanaris, Ioannis A. Stogiannis, Aikaterini A. Mouza & Satish G. Kandlikar (2014): Comparing
the mixing performance of common types of chaotic micromixers: a numerical study, Heat Transfer Engineering, DOI:
10.1080/01457632.2015.987623
Disclaimer: This is a version of an unedited manuscript that has been accepted for publication. As a service
to authors and researchers we are providing this version of the accepted manuscript (AM). Copyediting,
typesetting, and review of the resulting proof will be undertaken on this manuscript before final publication of
the Version of Record (VoR). During production and pre-press, errors may be discovered which could affect the
content, and all legal disclaimers that apply to the journal relate to this version also.
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
ACCEPTED MANUSCRIPT
numerical study
1
Xaar Plc, Cambridge UK
2
Department of Chemical Engineering, Aristotle University of Thessaloniki, Greece
3
Department of Mechanical Engineering, Rochester Institute of Technology, Rochester, NY, USA
Laboratory of Chemical Process and Plant Design, Aristotle University of Thessaloniki, Greece.
E-mail: mouza@auth.gr
Abstract
technique. The purpose of the present study is to compare different configurations of passive
mixers. Six different microchannel designs, commonly used in the majority of mixing
applications, were numerically evaluated in the present study. Both two-dimensional and three-
dimensional serpentine-type microchannel designs were investigated. For each design, the effect
1
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
of geometrical parameters was studied using a commercial computational fluid dynamics code.
Results are presented in terms of mixing efficiency and pressure drop. It is revealed that
increasing the flow path by providing compact fluid passages is not the only way to achieve
better mixing; inducing a helical flow along the flow length has an additional benefit in
improving the mixing performance. The results of this study can be used as a basis for further
improving the design of micromixers, while they can also provide the engineers with guiding
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
principles, not for selecting the optimum set of geometrical parameters for a specific
microchannel design but for screening out the least efficient micromixer configurations.
2
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Introduction
due to the laminar flow that usually prevails in the low Reynolds number flows at microscale [1].
mixing devices. It is however evident that, when mixing is driven solely by diffusion, long
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
residence time is required. To improve a micromixer performance, either active mixing methods
(e.g. acoustic actuation) or passive mixing features, e.g. surface roughness, complex two-
dimensional or even three-dimensional paths, can be used. Active micromixers - which can
devices and have in general a higher implementation cost, thereby making them unsuitable for
disposable microfluidic devices [2, 3]. For the most appropriate application of passive mixing
features, it is important to understand the flow through these complex channels and assess the
effect of the different geometrical parameters on the mixing performance. This understanding
The design concept of micromixers utilizing chaotic advection is based on the simple idea of
the modification of the channel shape for splitting, stretching, folding and breaking of the flow;
in other words, laminating the flow. A lot of work has already been done in this field. Hong et al.
[2] described the design of a passive mixing microreactor using modified Tesla structure, which
is easy to manufacture and of low cost. A comprehensive review published by Hessel et al. [4]
presents a number of different approaches on passive and active mixing. The authors also
suggest that, as the catalogue of different available micromixer designs is quite broad, it is
3
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
advisable for the future designer to focus mainly on the improvement of the existing designs and
the implementation of additional functions in them, such as heating and sensing. Rawool et al.
[5] suggest that the addition of obstructions as part of the channel geometry can be also
beneficial to micromixer efficiency. Lin and Yang [6] analyzed a planar serpentine microchannel
and presented the flow field and the concentration distribution in the two mixing fluids. Their
study, which was performed for Re=160, suggests that the interfacial area between the fluids is
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
continuously distorted due to the changes in the direction, and the flow trajectories are highly
unstable. Park and Kwon [3] studied a similar serpentine design, extending it to 3D. They
identified rotation, lamination and chaotic advection as the main features that improve mixing in
this type of serpentine micromixers. Tsui et al. [7] placed on the walls of the channels either
grooves or block obstacles that disrupt the flow and suggested that the mixing performance can
be enhanced by having the grooves arranged in a staggered manner, by which the transversal
velocity is largely increased. Jeon and Sin [8] studied, both experimentally and numerically, the
mixing performance of passive planar micro-mixers with different geometries and concluded that
valuable design guidelines can be derived by analyzing the volume fraction of mixing fluids.
Wang and Hu [9] investigated the effect of obstacles in T-type microchannels in the mixing of
fluids. The authors concluded that several geometrical features of these channels as well as
various inlet Re ratios between the inlet branches are not yet fully studied and suggested that
such findings can be useful for providing optimal designs of micromixers. In a more recent
review on micromixers, Kumar et al. [10] discussed different designs and noted that when 3D
curved microchannels are used, mixing can be greatly improved even for very low Re (from
0.001 to 5). Kanaris and Mouza [11] investigated the effect of geometrical parameters by altering
4
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
the original design of a prototype μ-reactor and studied them by comparing Resident Time
Distributions (RTDs) using a numerical approach. The results of their study revealed that the
increase in curvature promotes the mixing performance of the device. In a similar study, Kim et
al. [12] have also numerically investigated the effect of geometrical parameters on the
distribution of two liquids along the channel. Finally, Zhang et al. [13] who have performed
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
have reported that the mixing performance greatly depends on the angle of each cross-section
turns.
As already mentioned, there are a number of micromixers available in the literature and a
number of studies have focused on optimizing their performance. Mixing efficiency can be
significantly improved by chaotic advection, which can be either generated by special geometries
in the mixing channel (i.e., passive) or induced by an external force (active). While many types
of different designs of passive micromixers have been proposed and investigated by various
researchers, a more detailed study on the effect of parameters describing similar geometrical
features across micromixers channels of different designs is warranted. This paper aims at
comparing the performance of six highly successful micromixing techniques and presenting a
comparison under the same operating conditions. To do this, a common ‗design rule‘ will be
adopted for all of them, in the form of ‗building blocks‘ of similar features. In an effort to
contribute to the designing of more efficient microdevices, the scope of this study is to review
and compare various designs published in the literature. Thus, the effect of the key geometrical
parameters of common ―building blocks‖ of micromixers on the mixing efficiency and on the
5
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Micromixer designs
flow channels have been extensively used as passive micromixers. However, the typical
―building blocks‖ which form the microchannels are rather standard [4]. The most common
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
designs for simple passive micromixers comprise orthogonal blocks, located either on one or two
planes, while curved microchannels, which enhance mixing efficiency by promoting helical
flow, can be also implemented into the device [14]. This study attempts to combine these
fundamental features into representative generic designs. Thus, six designs are chosen and are
illustrated in Figure 1, whereas the edge, W, of the square duct is 100μm, for all cases. Namely
the designs investigated are: a simple T-type straight duct (T) used as reference, a simple
orthogonal serpentine geometry (SO), two orthogonal serpentine designs extending on two
parallel planes with channels resembling the form of the letter C (C) and the letter L (L), and a
curved serpentine extended either on a single plane (SCSP) or on two parallel planes (SCDP).
The length of the fluid path is determined by the repeated mixing elements increased by a T-
type fluid entrance region of constant geometrical characteristics. Four recurrent ―building
blocks‖ are considered for all cases as they were proved to be sufficient for achieving a
quantifiable degree of mixing, while on the other hand are feasible to simulate with the available
computer facilities. Each mixing element is described by two characteristics dimensions (L1, L2)
as illustrated in Figure 2 as well as the channel edge (W) which is set to be constant throughout
this study.
6
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
An initial study was performed for the Y and C mixers. An interesting conclusion drawn is
that the overall mixing efficiency of the device, at least for the range of operating conditions
tested, was not practically affected by the angle of the Y-entrance region; therefore, this
parameter was screened and not included in the design of simulations. Typical mixing elements
are presented in Figure 2. Based on the above, two dimensionless parameters are defined and
used for the parametric design: a length ratio, LR, and a normalized length, ZR. LR is defined as:
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
LR = L1 / L2 (1)
ZR = W / L1 (2)
Since the channel edge, W, is considered constant for this study (100μm), it is obvious that the
length of the device, Lz , is controlled by ZR, while the total flow path, Lt , is controlled by LR. A
characteristic quantity would be the ratio of Lt over Lz, which is an indicator of the level of
compactness of the device. Compactness of such devices has been reported as one of the most
important advantages of microdevices when are used in mixing [10] or heat transfer applications
[15]. The design parameters used, i.e. LR and ZR, are illustrated in Figure 2, while their range is
given in Table 1. It must be noted that, as the microchannel is assumed to have a square cross
section, when the microdevice is considered to be comprised of two planes, the total height of the
microdevice is equal to 2W, as the two planes are assumed to be perfectly attached to each other.
However, as the actual microdevice includes an overhead due to the connection between the two
planes, this might not hold completely true in reality. For the sake of simplicity and for the
7
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Water and methyl alcohol (CH3OH) were chosen to be the mixing fluids as under normal
conditions they form ideal mixtures and do not react with each other. The thermophysical
properties for both fluids were evaluated at room temperature and considered constant. A similar
In a recent study [11] it is reported that an increase of Re number leads to a better mixing
efficiency. Based on this, in the present study, whose aim is to compare different configurations,
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
the effect of Re has not been taken into consideration. Consequently, for all simulations the flow
rate has the same value that corresponds to laminar flow (i.e., Re has a value of about 2 for both
fluids) to ensure that the prevailing mixing mechanism is the mass diffusion between the liquids.
In the present calculations, the computational fluid dynamics (CFD) code uses the laminar
flow model of ANSYS CFX and the High-resolution Advection Scheme for the discretisation of
the momentum equations. The two fluids, water and methyl alcohol, are considered having
different physical properties and the prevailing mixing mechanism is the mass diffusion between
them. No chemical reaction was included in the model and the fluids are considered
incompressible. Mass-flow boundary conditions are set on the two inlet ports. The flow rate at
each inlet of the Y-entrance was set equal to 10μL/min, which corresponds to Re=1.67 for water
and to Re=1.77 for methyl alcohol, i.e. to laminar flow. An additional flow rate, 50μL/min, was
also used for a group of simulations with the intention to evaluate whether results are
pressure is set on the outlet port, while the convergence criterion is the mass-balance residual
value to be less than 10-9. Based on the Schmidt number and fluid properties, the diffusion
8
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
In order to compare the mixing efficiency of the different geometries studied it is needed to
quantify the quality of mixing. In the majority of studies, either experimental [16-18] or
numerical [19-21], an Index of Mixing Efficiency (IME) is introduced, based on the standard
deviation of mass fraction from the mean concentration over the cross section area:
c c
2
dA
IME 1 A
(3)
c
2
dA
A
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
In this approach the cross section is divided into several units (in our case the grid elements
of each cross section) and concentration is calculated in each unit. The mass flow averaged value
of the mass fraction of water is calculated using the built-in functions of the CFD code
postprocessing software (ANSYS CFD-Post). Mixing is considered to occur when the water
volume fraction equals methyl alcohol volume fraction (i.e., both attain the value of 0.5). As the
densities of water and methyl alcohol are 997 kg/m3 and 791 kg/m3, respectively, for the case of
1
c 0.55 .
791
1
997
the mass flows of the mixing fluids are constant over time, the mean concentration c will also
As the numerical diffusion in the CFD calculations can influence the accuracy of the
calculations, a thorough grid dependence study is performed to ensure that the solution is
independent of the grid density. As it has been already stated, trying to keep the computational
demand of the simulation low, only four mixing elements plus the entrance region were tested
for each configuration. As it is expected, only a percentage of the two fluids are actually mixed
9
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
at the outlet of the fourth element. During the grid dependence study, the results were evaluated
in terms of pressure drop, ΔP, and mixing efficiency, IME. All grids were constructed with
structured hexahedral elements and the final maximum size of the element was chosen to be
equal to 3.5μm.
In this study, two different lengths were considered; the total length of the fluid path, Lt, and
the length of the device, Lz (Figure 2). It is apparent that, from the designer's point of view, a
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
certain micromixer is superior to another when it achieves better mixing per unit of actual device
length. Increasing the flow path on the directions perpendicular to the flow direction can be
beneficial to the mixing efficiency, provided that the inevitable increase in pressure drop is also
In order to estimate the performance of this type of micromixers, two response variables are
considered: ΔP/Lz and IME/Lz. The former is defined as the pressure drop per unit of device
length, as proposed above, while the latter refers to the mixing performance, again per unit of
device length. Both of these response variables, for all the designs considered in this study, are
using standard Response Surface Methodology (RSM) [22]. To reduce the number of simulations
needed for each micromixer configuration for obtaining a full factorial design, a Design of
Experiments (DOE) method is applied. This approach has been successfully used before [11, 12].
In order to acquire confidence for the method used and to validate the resulting model, two
ACCEPTED MANUSCRIPT
10
ACCEPTED MANUSCRIPT
To assess the overall performance of the designs two performance indices that compare each
microchannel type with the reference case design of a T-type square duct are introduced:
Lt
n (5)
Lt ,Y
IME
Lt
nIME (6)
IME
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
Lt ,Y
The nΔΡ index is a friction-related term while nIME is a mixing-related term and both can be
used to assess the interplay between mixing enhancement and pressure drop.
Results
In order to ensure that the selected CFD code can accurately predict fluid flow and mixing
performance in micromixers, the code is validated using experimental data available in the
literature. Liu et al. [23] studied a 3D serpentine microchannel, which consisted of the same
building blocks as the C-type micromixer design model used in this study. In their approach
CCD camera. Light intensity is used to quantify the uniformity of mixing in a plane visible to the
mixing index, identical to the one used in the present study, was calculated for assessing mixing
A C-type model of this work‘s approach is modified in order to meet the geometrical
characteristics of the experimental device [23] and has been then modeled in a series of CFD
11
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
runs using the same model parameters employed throughout this study. Results from the
numerical simulations, shown in Figure 3, are in excellent agreement (deviation less than 3%)
with the experimental results for a range of Re starting from 5 up to 12. Simulations are focused
on the lower part of the experimental Re range due to memory limitations of the available
computational facilities. However, it should be noted that CFD code is validated in a range of Re
numbers significantly higher than the one used thereafter in the study and thus it can be
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
considered that it is valid to predict fluid flow and mixing performance for Re numbers lower
than the range used, where the demand on grid size is also lower.
CFD simulations were performed for the six different geometries described in detail in a
previous section and for the range of design variables reported in Table 1. The IME was
evaluated for all studied geometries at the flow outlet using ANSYS CFX Expression Language
Using RSM methodology a second-order polynomial equation was fitted to the results, whose
coefficients are presented in Table 2, while the value of R-square is calculated to be over 99% for
all cases. Simulation results show the geometric variations have a distinct influence on mixing.
Figure 4 presents the streamline plots for all designs in a representative area of the conduit,
while Figure 5 shows the mixing efficiency and pressure drop against the total flow path length,
Lt. As it is expected, when the actual device length, Lt, is increased or equally ZR is increased,
The characteristic length for all cases is the same (equal to W) and as a result, a linear
increase of pressure drop with unit length can be expected. This trend is indeed shown on Figure
12
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
5a. Based on the above, pressure drop vs. flow path length (ΔP/Lt) is the same regardless of the
microchannel design. This means that, for the low Re values usually encountered in micro-
devices, the number and type of bends does not significantly affect the pressure drop per unit
length. The same trend has been reported by Kumar et al. [10], where for the same Re number in
the laminar regime, friction factor is practically independent of the geometry of the
microchannel.
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
Lt L1 L2 number of elements .
Lz L2 number of elements .
Lt
LR 1 .
Lz
P
Lz
LR 1 .
P
Lt
ads to the prediction of pressure drop vs. the length of the device to be:
P P LR 1 (7)
Lz Lt
As LR is increased for a given value of ZR, and as pressure drop vs. flow path length (ΔP/Lt)
is found to be constant for this study, an increase of pressure drop per unit of device length
13
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
(ΔP/Lz) is also inevitable. To evaluate the effect of LR, additional simulations for the SO design
were run for a wider range of LR values. The dependence of the micromixer performance on the
value of LR , which defines the channel bend ratio and controls the total fluid flow length, is
presented for the SO type device in Figure 6. It is obvious that the rate of increase of IME
(representing mixing performance) is higher than that of the pressure drop. This might lead to the
However, for the final decision one must rely on the relative importance of mixing efficiency
versus the corresponding increase in pumping cost. For the two-layer configurations (i.e., C-
Type, L-Type and SCDL) the interfacial area between the fluids is distorted and enlarged in a
spiral behavior, effectively promoting the fluid mixing. The effect of channel curvature seems to
The L-Type design performs slightly better in terms of mixing compared to the other designs,
as shown in Figure 5b. In terms of mixing performance, it can be seen that L-type stands out as
the best, albeit with a small difference compared to the C-Type and SCDP, which perform very
similar to each other; SO and SCSP perform lower and also similar to each other (Figure 5b).
The similarity between C-Type and SCDP as well as SO and SCSP can be explained by the fact
that the incorporated designs are practically the same, where SCDP & SCSP include curved
channels and C-Type & SO use orthogonal channels. It must be noted that simulations were also
run for a higher flow rate, i.e. 50μL/min, which corresponds to Re=8.35 and 8.85 for water and
methyl alcohol, respectively, and the results are found to be consistent with the above
conclusion.
14
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
It is also shown that the mixing efficiency per device length (Lz) attains a higher value for the
L-Type mixer. In Table 3 the values of pressure drop and the mixing efficiency of the different
configurations are evaluated for a pair of ZR and LR values and they are compared to the base
case (i.e. T-type design). In all cases, increase in pressure is about 40 to 50%, while the increase
in mixing efficiency is always more than 200%. As mentioned above, for this range of Re
numbers, pressure drop is not notably influenced by the number of bends in the microchannel,
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
but by only its actual length, Lt. This statement is verified also by the practically similar pressure
drop calculated for a typical set of the design variables for all design cases (Table 3). Therefore,
choosing a design for the improvement of the overall performance of the microdevice should
The question that rises is why L-Type performs better than C-Type in terms of mixing
efficiency per flow path length, for the same LR; it is not the number of bends (since both have
the same) but the induced rotation of the flow that enhances mixing efficiency. The fluid path in
L-Type actually follows a helix, completing a 360 deg turn due to the type of bends introduced
(Figure 7a). On the other hand, in C-type the fluid follows a 3D zig-zag movement (Figure 7b),
which is less effective. As a result, the performance of the micromixer design cannot be
expressed in terms of number of bends, since it also depends on the actual rotation of the fluid
path. The comparison between C-Type and L-Type, and the effect of inducing rotation in the
fluid flow path are also mentioned in the paper by Park and Kwon [3], and the results of this
15
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Conclusions
This numerical study considered five channel geometries as alternatives to the well-known T-
type straight-duct micro-mixer. A CFD software was used to characterize these different
geometries for their effectiveness in mixing and flow impendence. All the tested geometric
The L-type micromixer tends to have a marginally better performance in terms of mixing
efficiency, since the chaotic behavior of fluid trajectories produces the greatest mixing
intensification. The mixing efficiency for the best case scenario is improved almost 3 times
in comparison to the baseline design. The L-type microdevice design has an additional
advantage over its similar counterpart, the C-type design, as for the same geometrical
design parameters the induced helical flow further improves the performance of the
micromixer.
The pressure drop of the device is only attributed to the increase in total fluid flow length.
Since the flow is laminar, pressure drop increases linearly with flow length, regardless of
the existence of other geometrical features. This observation allows the designer to seek
The results from this study provide guidelines for the engineers in selecting appropriate
type of micromixer for a specific application. The micromixers that can be clearly
excluded are identified and guidelines are provided for selecting among the available
choices.
16
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
It can be deduced that the geometries, which induce helical flow path, offer a number of
improvements over typical designs. Further study of these configurations can provide devices
that are more robust. The results of this study can be used as a basis for further improving the
design of micromixers, while they can also provide the engineers with guidelines, not for
selecting the optimum set of geometrical parameters for a specific microchannel design but for
Nomenclature
A cross-section area, m2
L L-type geometry
17
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Re Reynolds number
SO Serpentine-Orthogonal geometry
W channel width, m
Greek symbol
ΔP Pressure drop, Pa
18
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
References
[1] Gravesen, P., Branebjerg, J., and Jensen, O. S., Microfluidics-a review, Journal of
[2] Hong, C.-C., Choi, J.-W., and Ahn, C. H., A novel in-plane passive microfluidic mixer
with modified Tesla structures, Lab on a Chip, vol. 4, no. 2, pp. 109-113, 2004.
[3] Park, J. M., and Kwon, T. H., Numerical characterization of three-dimensional serpentine
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
[4] Hessel, V., Löwe, H., and Schönfeld, F., Micromixers—a review on passive and active
mixing principles, Chemical Engineering Science, vol. 60, no. 8–9, pp. 2479-2501, 2005.
[5] Rawool, A. S., Mitra, S., and Kandlikar, S. G., Numerical simulation of flow through
microchannels with designed roughness, Microfluidics and Nanofluidics, vol. 2, no. 3, pp.
215-221, 2006.
[6] Lin, K.-W., and Yang, J.-T., Chaotic mixing of fluids in a planar serpentine channel,
International Journal of Heat and Mass Transfer, vol. 50, no. 7–8, pp. 1269-1277, 2007.
[7] Tsui, Y.-Y., Yang, C.-S., and Hsieh, C.-M., Evaluation of the mixing performance of the
micromixers with grooved or obstructed channels, Journal of fluids engineering, vol. 130,
[8] Jeon, W., and Shin, C. B., Design and simulation of passive mixing in microfluidic
systems with geometric variations, Chemical Engineering Journal, vol. 152, no. 2–3, pp.
575-582, 2009.
[9] Wang, C., and Hu, Y., Mixing of liquids using obstacles in Y-type microchannels,
19
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Tamkang Journal of Science and Engineering, vol. 13, no. 4, pp. 385-394, 2010.
[10] Kumar, V., Paraschivoiu, M., and Nigam, K. D. P., Single-phase fluid flow and mixing in
microchannels, Chemical Engineering Science, vol. 66, no. 7, pp. 1329-1373, 2011.
[11] Kanaris, A. G., and Mouza, A. A., Numerical investigation of the effect of geometrical
[12] Kim, B. S., Kwak, B. S., Shin, S., Lee, S., Kim, K. M., Jung, H.-I., and Cho, H. H.,
surface method, International Journal of Heat and Mass Transfer, vol. 54, no. 1–3, pp.
118-125, 2011.
[13] Zhang, Y., Hu, Y., and Wu, H., Design and simulation of passive micromixers based on
capillary, Microfluidics and Nanofluidics, vol. 13, no. 5, pp. 809-818, 2012.
[14] Schönfeld, F., and Hardt, S., Simulation of helical flows in microchannels, AIChE
[15] Kandlikar, S. G., High Flux Heat Removal with Microchannels—A Roadmap of
Challenges and Opportunities, Heat Transfer Engineering, vol. 26, no. 8, pp. 5-14, 2005.
[16] Jayaraj, S., Kang, S. M., and Suh, Y. K., A review on the analysis and experiment of fluid
flow and mixing in micro-channels, Journal of Mechanical Science and Technology, vol.
[17] Bhagat, A. A. S., Peterson, E. T. K., and Papautsky, I., A passive planar micromixer with
ACCEPTED MANUSCRIPT
20
ACCEPTED MANUSCRIPT
[18] Nimafar, M., Viktorov, V., and Martinelli, M., Experimental comparative mixing
[19] Wu, C. Y., and Tsai, R. T., Fluid mixing via multidirectional vortices in converging-
[20] Jeon, W., and Shin, C. B., Design and simulation of passive mixing in microfluidic
systems with geometric variations, Chemical Engineering Journal, vol. 152, no. 2-3, pp.
575-582, 2009.
[21] Alam, A., and Kim, K.-Y., Mixing performance of a planar micromixer with circular
chambers and crossing constriction channels, Sensors and Actuators B: Chemical, vol.
[22] Myers, R. H., Montgomery, D. C., and Anderson-Cook, C. M., Response Surface
2011.
[23] Liu, R. H., Stremler, M. A., Sharp, K. V., Olsen, M. G., Santiago, J. G., Adrian, R. J.,
2000.
ACCEPTED MANUSCRIPT
21
ACCEPTED MANUSCRIPT
ZR=W/L1 0.25-0.50
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
LR=L1/L2 0.30-0.70
22
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
23
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Baseline
C L SO SCSP SCDP
(Straight)
24
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
LIST OF FIGURES
Figure 3. Comparison of CFD code results with experimental ones from Liu et al. [23] using
Figure 5. a) Pressure drop and b) Mixing efficiency vs total flow length. Error bars at -5%.
Figure 6. Dependence of the two performance indices on LR values (SO type, ZR=0.24).
25
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
ACCEPTED MANUSCRIPT
26
ACCEPTED MANUSCRIPT
(SCSP) (SCDP)
ACCEPTED MANUSCRIPT
27
ACCEPTED MANUSCRIPT
Lz L2 Lt
L2
L1
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
L1
(a) (b)
ACCEPTED MANUSCRIPT
28
ACCEPTED MANUSCRIPT
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
error bars at
±3%
Figure 3. Comparison of CFD code results with experimental ones from Liu et al. [23]
29
ACCEPTED MANUSCRIPT
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
(b)
(a)
ACCEPTED MANUSCRIPT
(c)
30
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
(d) (e)
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
SCDP
31
ACCEPTED MANUSCRIPT
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
(b)
(a)
ACCEPTED MANUSCRIPT
32
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
Figure 5. a) Pressure drop and b) Index of Mixing efficiency vs. total flow length.
33
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
Figure 6. Dependence of the two performance indices on LR values (SO type, ZR=0.24).
34
ACCEPTED MANUSCRIPT
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
(a)
ACCEPTED MANUSCRIPT
35
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Downloaded by [University of Western Ontario] at 10:11 03 February 2015
(b)
Figure 7. Partial view of the fluid path focusing on the directional changes for a) L and b) C
mixers.
36
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
PhD on the numerical and experimental study of plate heat exchangers for
France, optimizing the design of fin and tube heat exchangers. He is an Associate Member of the
Institute for Chemical Engineers and he currently serves as a Fluid Systems Design Engineer in
Xaar plc working on the design of complex fluidic paths in the microscale and hydraulic
systems.
University of Manchester. His research interests include heat transfer, micro equipment design
and CFD.
37
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
equipment design, process simulation, CFD, use of nanofluids for heat transfer enhancement and
multiphase flow. Using advanced non-intrusive measuring techniques (μ-PIV, LIF, high speed
RIT. He received his Ph.D. degree from the Indian Institute of Technology in
Bombay in 1975 and has been a faculty there before coming to RIT in 1980.
He has worked extensively in the area of flow boiling heat transfer and CHF
phenomena at microscale, single-phase flow in microchannels, high heat flux chip cooling, and
water management in PEM fuel cells. He has published over 200 journal and conference papers.
He is a Fellow member of ASME and a former Associate Editor of ASME Journal of Heat
Transfer. He has received the RIT’s Eisenhart Outstanding Teaching Award in 1997 and Trustees
Outstanding Scholarship Award in 2006. He has received the 2008 Rochester Engineer of the
Year award from Rochester Engineering Society. He is the recipient of the 2012 ASME Heat
Transfer Memorial Award. Currently he is working on DOE and GM sponsored projects on Fuel
Cell water management under freezing conditions, and an NSF sponsored projects on developing
38
ACCEPTED MANUSCRIPT