4.1 Optimization Strategies for Micromixers
47
investigated the mixing behavior of three different grooved micromixers, viz. slanted
groove micromixer, SHM and barrier embedded micromixer. A ‘colored particle
tracking method’ was developed to study the mixing performance both qualitatively
and quantitatively.
Liu et al. [8] solved a variational optimization problem to obtain different types of
passive micromixers using a layout optimization method. Constraints for the variational problem were Navier–Stokes equations and the convection–diffusion equation,
with the mixing performance as an objective function. Figure 4.2 shows numerical
simulation results for a passive micromixer designed by the layout optimization
method, where measurements reveal a good mixing performance. They showed that
the proposed optimization framework could reduce the dependency on the experience
and intuition of designers. Experiments were conducted to support the effectiveness
of the layout optimization method for conceptual design of the micromixer.
Numerical optimization techniques [9–15] based on CFD analysis have been
proven to be an effective tool for robust and efficient design of passive micromixers.
The objective function(s) for the optimization of a micromixer can be selected among
the performance parameters such as mixing index, pressure loss, and residence time,
etc. In the case of micromixers, the important performance parameters are the mixing
efficiency and pressure loss. The mixing efficiency is the critical performance parameter related to the mixing performance of the device. The pressure loss is directly
related to the pumping power required to drive the fluids through the micromixers.
The optimization can be single-objective (e.g., to maximize the mixing efficiency) or
multi-objective (e.g., to maximize the mixing efficiency and minimize the pressure
loss).
Generally, the optimization procedure requires a large number of evaluations for
the objective function(s) rendering conventional optimization techniques to be very
expensive. To reduce the computational cost, surrogate model(s) is used to generate
functional relationship between inputs and outputs, which is reliable representation
of the simulation model as closely as possible. Queipo et al. [17] and Forrester
and Keane [18] reviewed various surrogate models used in aerospace applications.
Surrogate-based analysis and optimization have been applied to various optimization
problems [19–24].
Fig. 4.2 Series-wound
extension of the bending
cells obtained by the layout
optimization method [8]
47
investigated the mixing behavior of three different grooved micromixers, viz. slanted
groove micromixer, SHM and barrier embedded micromixer. A ‘colored particle
tracking method’ was developed to study the mixing performance both qualitatively
and quantitatively.
Liu et al. [8] solved a variational optimization problem to obtain different types of
passive micromixers using a layout optimization method. Constraints for the variational problem were Navier–Stokes equations and the convection–diffusion equation,
with the mixing performance as an objective function. Figure 4.2 shows numerical
simulation results for a passive micromixer designed by the layout optimization
method, where measurements reveal a good mixing performance. They showed that
the proposed optimization framework could reduce the dependency on the experience
and intuition of designers. Experiments were conducted to support the effectiveness
of the layout optimization method for conceptual design of the micromixer.
Numerical optimization techniques [9–15] based on CFD analysis have been
proven to be an effective tool for robust and efficient design of passive micromixers.
The objective function(s) for the optimization of a micromixer can be selected among
the performance parameters such as mixing index, pressure loss, and residence time,
etc. In the case of micromixers, the important performance parameters are the mixing
efficiency and pressure loss. The mixing efficiency is the critical performance parameter related to the mixing performance of the device. The pressure loss is directly
related to the pumping power required to drive the fluids through the micromixers.
The optimization can be single-objective (e.g., to maximize the mixing efficiency) or
multi-objective (e.g., to maximize the mixing efficiency and minimize the pressure
loss).
Generally, the optimization procedure requires a large number of evaluations for
the objective function(s) rendering conventional optimization techniques to be very
expensive. To reduce the computational cost, surrogate model(s) is used to generate
functional relationship between inputs and outputs, which is reliable representation
of the simulation model as closely as possible. Queipo et al. [17] and Forrester
and Keane [18] reviewed various surrogate models used in aerospace applications.
Surrogate-based analysis and optimization have been applied to various optimization
problems [19–24].
Fig. 4.2 Series-wound
extension of the bending
cells obtained by the layout
optimization method [8]