Mixing in microfluidic devices driven by oscillatory channel flow is studied. Numerical simulation of the unsteady Navier-Stokes equations is used to investigate generation of flow controlled chaotic mixing, in which the channel geometries have stationary boundaries, and chaotic conditions are achieved solely through manipulation of inflow boundary conditions. A two-fold approach is used to break down the generation of chaotic flow by such means. First, purely oscillatory motion is modeled. In this case, the net throughput flow in the system is zero. This allows identification of boundary condition regimes which are viable for producing chaotic mixing and represents the maximum possible effective mixing that can be obtained from a given flow configuration. In the second phase, the effect of imposing a fixed throughput flow on top of the oscillatory motion is examined, as a function of throughput flow rate. Three different representative microfluidic geometries are modeled, a cross cell, an H-cell and a star-cell. For all geometries, calculations for the case of zero throughput flow show that chaotic flow can be achieved in which Lagrangian particles orbit on an attractor whose size and characteristics are determined by a Strouhal number. Approximately 10 to 15 cycles are needed in order to effectively mix different groups of passive tracer particle blobs. For the case of finite throughput, the time of flight on the attractor follows a distribution of residence times, in which particles undergo chaotic orbits in the mixing regime, before being convected downstream. Downstream dispersion patterns are a strong function of the ratio of the throughput to oscillatory velocity. Strategies for controlling downstream dispersion are probably the biggest challenge to the making effective use of these types of mixing devices.
Citation: Physics of Fluids
Pub Type: Journals
chaos, chaotic advection, microfluidics, mixing