Numerical characterization of simple three-dimensional chaotic micromixers
Introduction
The past few decades has witnessed an important role of microfluidic devices in chemical, analytical and biochemical applications. Especially, microfluidic devices with the advantages of high efficiency and precise control provided a facile strategy to chemical synthesis [1], [2]. All the applications need to mix two or more fluid streams to fulfill the task. Therefore, the micromixer is one of the essential units of these applications.
Numerous studies have focused on enhancing the fluid micromixing in both active and passive micromixers. Compared with active micromixers, passive units are stationary and do not require any external energy input apart from the pressure drop needed to drive flow. This enables simpler fabrication and fewer external connections. Passive micromixers can be classified into four types, namely, lamination, injection, droplet and chaotic advection micromixers [3], [4]. The chaotic advection mechanism can be easily generated by special channel geometries that split, stretch, fold and break the flow. Thereby, it was widely investigated and several micromixing structures became classic. For example, a zigzag microchannel proposed by Mengeaud et al. can induce transversal flow at high Reynolds number due to the flow recirculation behind the sharp turns, and thereby improve the mixing efficiency [5]. Similar results were observed in the square-wave microchannels [6]. Another planar chaotic micromixer with curved channels also can produce secondary flows in the manner of vortices [7]. The three types of micromixer have meander channels on a plane, and can be called as two-dimensional serpentine micromixers. To further enhance the mixing, Liu et al. [6] proposed a three-dimensional serpentine design with the repetitive C-shaped units, and experimentally demonstrated its excellent mixing performance compared with the square-wave channel. Another serpentine micromixer using an L-shaped unit was presented by Vijayendran et al. [8]. Both of the two three-dimensional serpentine micromixers are advantageous in terms of an efficient mixing performance by the significant advection effect with a simple design. Thus they have been employed in the further optimal design [9] or novel micromixer development [10].
In addition, the patterned grooves on the channel wall also can be utilized to induce the lateral flows, hence enhancing the mixing process. It was first proposed by Stroock et al. [11] and experimentally demonstrated its chaotic mixing performance with the help of twisting flows. Moreover, the staggered herringbone grooves were placed instead of straight grooves, leading to the counter-rotating flows. More details of the flow and mixing characteristics of the staggered herringbone micromixer (SHM) were further investigated numerically by other researchers [12], [13], [14]. Inspired by this, Kim et al. [15] placed periodic barriers on the channel walls and also observed the rotating transversal flows in the micromixer. A large number of studies illustrated the obstacles or grooves on channel walls can result in rotating transversal flows and further good mixing [16], [17], [18]. Such a simple design was even applied to combine with a two-dimensional serpentine microchannel for developing new efficient micromixers [19], [20], [21].
The present work aims at evaluating the mixing performance of different micromixers in the medium to high Reynolds numbers and elucidating the mechanisms of mixing enhancement in each structure. Within this Re range, the periodic arrangement of the simple channel geometry can lead to an excellent mixing quality without a high pressure drop. To take advantage of it, two representative types of micromixers were applied in this work. One is the 3D serpentine micromixer (TSM), including C-shaped and L-shaped units. The other type is the 2D serpentine (square-wave) micromixer with cubic grooves (GSM). In this type, the grooves are configured in three manners. Numerical analysis of the five micromixers was performed to investigate the flow characteristics and mixing behavior by means of the computational fluid dynamics (CFD). Two miscible fluids, ethanol and water, were used for mixing. The flow field, mixing quality and pressure drop were examined to clarify the mechanisms of the mixing enhancement.
Section snippets
Three-dimensional chaotic micromixer
Two representative types of three-dimensional chaotic micromixers are introduced briefly with corresponding schematics in this section. Fig. 1(a) shows TSMC which consists of C-shaped units, and Fig. 1(b) shows TSML has L-shaped units. Both of them are fundamental designs of the 3D serpentine micromixer using the simple geometry in two layers. The other type is the square-wave micromixer with periodic cubic grooves (GSM). Fig. 1(c) shows GSMMO whose grooves arranged at the midstream positions
Numerical simulation
The performance of five micromixers was evaluated by numerical simulation using ANSYS CFX 12.0. The CFX code solved three-dimensional continuity, as well as Navier–Stokes and convection–diffusion equations by the finite volume method via a coupled solver. These micromixers were modeled and meshed using ANSYS ICEM 12.0. All meshes in the simulations were composed of hexahedral cells.
The flow was defined viscous, isothermal, incompressible, laminar and in steady-state. Ethanol and water were
Results and discussion
A preliminary mesh independency test was conducted to find the optimum number of cells for all the five micromixers. With GSMUT as an example, four different structured cell systems with a node number ranging from 9.26 × 105 to 4.96 × 106 were tested. The distribution of the mixing index along GSMUT was evaluated with a gradual increment in node number. The result is shown in Fig. 2. Beyond the node number of 3.73 × 106, the influence of increasing the node number on the mixing index was negligible.
Conclusion
In the present work, the flow and mixing characteristics of five simple three-dimensional micromixers have been numerically investigated by using two representative types of TSM and GSM. A wide range of the Reynolds number (Re = 8–160) has been utilized. At low Reynolds numbers, the mixing in GSMMO and GSMMT is mainly dominated by diffusion. Hence the raised Re shortens the residence time, leading to the decreased mixing index. With an increase in the Reynolds number, the mixing is enhanced by
Acknowledgements
This study was financially supported by Program for Young Teachers in Shanghai Colleges (ZZyyy12306) and Program for Introduced Talents in Shanghai Institute of Technology (YJ2012-14).
References (21)
- et al.
Synthesis of Fe3O4@ZIF-8 magnetic core–shell microspheres and their potential application in a capillary microreactor
Chem. Eng. J.
(2013) - et al.
Parametric study on mixing of two fluids in a three-dimensional serpentine microchannel
Chem. Eng. J.
(2009) - et al.
Shape optimization of a micromixer with staggered herringbone groove
Chem. Eng. Sci.
(2007) - et al.
Design and characterisation of the staggered herringbone mixer
Chem. Eng. J.
(2008) - et al.
Two-step design protocol for patterned groove micromixers
Chem. Eng. Res. Des.
(2013) - et al.
Mixing enhancement of novel passive microfluidic mixers with cylindrical grooves
Chem. Eng. Sci.
(2012) - et al.
Mixing performance of a planar micromixer with circular obstructions in a curved microchannel
Chem. Eng. Res. Des.
(2014) - et al.
Analysis of mixing in a curved microchannel with rectangular grooves
Chem. Eng. J.
(2012) - et al.
Design and evaluation of an easily fabricated micromixer with three-dimensional periodic perturbation
Chem. Eng. J.
(2011) - et al.
Direct manipulation of particle size and morphology of ordered mesoporous silica by flow synthesis
RSC Adv.
(2015)