In micromixer devices, laminar characteristics of the flow domain and small diffusion constants of the fluid samples that are mixed characterize the mixing process. The advection dominant flow and transport processes that develop in these devices not only create significant challenges for numerical solution of the problem, but they are also the source of numerical errors which may lead to confusing performance evaluations that are reported in the literature. In this study, the finite volume method (FVM) and finite element method (FEM) are used to characterize these errors and critical issues in numerical performance evaluations are highlighted. In this study, we used numerical methods to evaluate the mixing characteristics of a typical T-shape passive micromixer for several flow and transport parameters using both FEM and FVM, although the numerical procedures described are also equally applicable to other geometric designs as well. The outcome of the study shows that the type of stabilization technique used in FEM is very important and should be documented and reported. Otherwise, erroneous mixing performance may be reported since the added artificial diffusion may significantly affect the mixing performance in the device. Similarly, when FVM methods are used, numerical diffusion errors may become important for certain unstructured discretization techniques that are used in the idealization of the solution domain. This point needs to be also analyzed and reported when FVM is used in performance evaluation of micromixer devices. The focus of this study is not on improving the mixing performance of micromixers. Instead, we highlight the bench scale characteristics of the solutions and the mixing evaluation procedures used when FVM and FEM are employed.
Keywords: CFD; artificial diffusion; false diffusion; finite element method; finite volume method; grid type; microfluidics; micromixers; numerical diffusion.