Spiral microfluidic-based devices play an important role in particle separation and concentration which is essential for many biomedical, chemical, and environmental purposes. Although many platforms were developed for the separation and purification of various microparticles, there is a lack of practical rules for designing these microfluidic platforms. Moreover, the physics of migration for non-naturally buoyant particles was overlooked in the previous studies. In this vein, here, we used direct numerical simulation (DNS) and particle tracing analysis to examine the physics of particle migration for both naturally buoyant and non-naturally buoyant particles. Later, by combining our simulation results with artificial intelligence (AI), we introduced a software, hereafter called SpiralDesigner, that can easily predict the final position of particles (naturally buoyant and non-naturally buoyant) in the outlet of the spiral microchannel with an accuracy of ∼98 %. SpiralDesigner just needs the particle size and density, length and radius of curvature of the spiral microchannel, and the total flow rate for prediction and after that, it provides the separation index that tells the user whether two particles are separable or not. This software can be very useful for designing a spiral microchannel for particle separation since it is very easy to use and also does not require any computational fluid dynamic simulations or fabrication. Hence, SpiralDesigner can potentially be used for a variety of purposes such as circulating tumor cell separation, stem cell harvesting, powder technology, environmental technology, and food science.
Read full abstract