The flow of blood is a highly complex phenomenon, exhibiting intricate behaviors such as the Fahraeus–Lindquist effect, migration, and the formation of a cell-free layer. These phenomena, along with the practical aspects of developing innovative blood-based diagnostic devices, have garnered significant research interest. With this in mind, MiDOpt-H aims to study blood flows using a synergistic approach that combines experimental and numerical methodologies to advance the design and optimization of microchannels that manipulate blood flows. The project combination of high-fidelity CFD simulations, low-fidelity surrogate reduced-order models, and numerical optimization techniques will expedite the design process.

MiDOpt-H will pursue innovative research in applying gradient-based optimization techniques, utilizing the continuous adjoint method, and combining both numerical and experimental methodologies to verify the optimized designs. In this endeavor, the CFD software will be implemented within the OpenFOAM library, enabling it to incorporate complex phenomena such as shear-induced migration, typically observed in blood flows confined within channels. The high-fidelity numerical methodology will be supported by surrogate reduced-order models derived from asymptotic analysis and experimental studies aimed at measuring velocity profiles, shear rates, and hematocrit distributions, which will be invaluable for the validation process. The main outcomes of the project include the development of a state-of-the-art 'principled design framework' that will streamline the analysis and optimization of blood-based microfluidic devices. This framework will be beneficial for diagnosing blood-related diseases and creating drug delivery micro-systems, thereby increasing their accessibility. Additionally, the experimentally obtained distributions of blood flow parameters will provide valuable fundamental insights.