Computational Models for Optimizing Particle Separation in Spiral Inertial Microfluidics: A Step Toward Enhanced Biosensing and Cell Sorting

Inertial microfluidics is essential for separating particles and cells, enabling numerous biomedical applications. Despite the simplicity of spiral microchannels, the lack of predictive models hampers real‐world applications, highlighting the need for cost‐effective computational tools. In this study, four novel data fitting models are developed using linear and power regression analyses to investigate how flow conditions influence particle behaviors within spiral microchannels. These models are rigorously tested under two different flow rates, focusing on a smaller particle representing Salmonella Typhimurium and a larger particle representing bacterial aggregates, aiming for effective separation and detection. A critical parameter, the sheath‐to‐sample flow rate ratio, is either interpolated or extrapolated using the microchannel's aspect ratios to predict particle separation. The models show strong agreement with experimental data, underscoring their predictability and efficiency. These insights suggest that further refinement of these models can significantly reduce research and development costs for advanced inertial microfluidic devices in biomedical applications. This work represents a crucial step towards establishing a robust computational framework, advancing inertial microfluidics towards practical biomedical applications. Inertial microfluidics is vital for particle and cell separation, pivotal for biomedical applications. Despite spiral microchannels' simplicity, predictive model gaps hinder real‐world use. This study introduces four novel data fitting models via linear and power regression, tested under varied flow rates to predict particle behavior. Findings reveal strong experimental alignment, promising reduced R&D costs and advanced microfluidic device development.