Numerical and experimental investigations of single and multiphase flows in microchannels

In this thesis, different aspects of microfluidics are deeply studied, employing a wide array of methods including Computational Fluid Dynamics modelling, machine learning algorithms, image processing, Particle Image Velocimetry (PIV), and Particle Tracking Velocimetry (PTV). This approach allows for an extensive analysis of some phenomena in microfluidics, highlighting the field’s interdisciplinary and innovative nature. The first part of this thesis examines microdrop formation within microjunctions, addressing gaps in this long-studied field. An automatic algorithm was developed to analyze droplet formation, complemented by PIV, providing a robust tool for validating numerical simulations. Validation focused on droplet sizes under various flow conditions, interface dynamics during breakup, and velocity fields within droplets. The validated simulations were further used to develop new models that describe the dimensions of the droplets. Moreover, an optimization approach was introduced to provide operational parameters for achieving droplets of the desired sizes. The second part of this thesis focused on thermal phenomena within microchannels. Typically, buoyancy effects are often neglected in microfluidic applications due to the small dimension of the channels. However, it was demonstrated that mixed convection can occur even under relatively low thermal gradients. To characterize this phenomenon, an approach combining numerical and experimental methods was used. Specifically, the flow was analyzed with PIV and PTV techniquesl. Moreover, numerical simulations were performed to study the temperature distribution and mass transfer within these devices. The thesis concludes with a numerical analysis and the development of a simplified model for simulating the hybrid primary heat exchanger in a nuclear reactor. This model achieves comparable results to the full model while reducing computation times, supporting experimental testing to characterize the component. The methodologies presented provide a multi-scale framework for designing, modeling, and validating innovative microchannel heat exchangers in forced and mixed convection regimes.