| The physics and chemistry governing mixing and cell/particle trapping in microsteady streaming devices have been advanced to the point where computer-aided design and optimization is possible. Experiments show that the flow is 2D at the intermediate plane where particle trapping occurs. Thus, 2D simulations are used throughout to permit high-resolution simulation of the fast oscillating primary flow while also capturing details of slow secondary steady streaming phenomena. An analytic/numeric approach is used to further accelerate numerical solutions. Flow fields computed in this manner are readily integrated into mass transfer, reaction, and particle motion studies to understand the processes that govern particle trapping and chemical treating of trapped cells. Concentric cylinders with the outer cylinder oscillating and the inner cylinder stationary is used initially to understand the oscillating flow, steady streaming eddy, and reagent dosing behavior adjacent to the inner cylinder. Simulations of flow and local chemical environment of steady streaming eddies match experiments well. Since practical devices are built into microchannels, simulation has taken the channel/cylinder geometry into account to better understand the role of a confined geometry and a mean flow on eddies formed near cylindrical obstructions embedded in the microchannels and on wall features. The particle trapping in eddies near the inner cylinder is shown to arise from a balance between the mean Stokes drag on the particle and time average oscillating inertial forces that we call the drift force. The importance of the drift force is emphasized by the fact that dense particle trapping is predicted to be impossible without its inclusion in the particle equation of motion. However, with the addition of this so-called drift force, it is possible to make predictions of the trapping locations, trapping force, and trapping stability, all of which match experiments well. We find that the trapping force can exceed 30 pN, and that it depends quadratically on the flow oscillation amplitude, as does the trap stability. The trapping location scales with the oscillating Stokes layer thickness. These results are the foundation for predicting where cells and particles will be trapped for any given device, particle size, or relative density. |
