Computationally efficient concentration-based model for accurate evaluation of T-junction inlet staggered herringbone micromixers

Micromixers are microfluidic components of critical importance in complex lab-on-chip devices. Passive devices in particular have attracted much attention over active ones due to their fabrication and operation simplicity. Staggered herringbone micromixers seem to be dominating the area of passive micromixers, and hence a lot of studies have been published on their optimisation, in terms of mixing efficiency (ME) and pressure drop. However, numerical simulations leveraging concentration-based models have not been convincingly accurate, owing to computational limitations and erroneous assumptions on diffusion coefficients that often are higher than those of large biomolecules in aqueous solutions. Accurate concentration-based models are a prerequisite in biochemical reaction studies. This work quantitatively shows how increasing mesh density increases solution accuracy, applying concentration-based modelling for biochemically realistic diffusivity. A numerical simulation methodology for obtaining less numerically diffusive results without further increasing mesh size is also given, leveraging quadratic discretisation and appropriate consistency stabilisation methods. They demonstrate that with the proposed approach, it is possible to obtain more accurate ME and pressure drop estimations for two different staggered herringbone micromixer variations, while at the same time limiting the usage of computational resources to a practical level.