Characterization of turbulent flows of non-Newtonian fluids using magnetic resonance velocimetry
Time: Mon 2024-11-11 10.15
Location: Kollegiesalen (Nr 4301) Brinellvägen 8
Video link: https://kth-se.zoom.us/j/65208609933
Language: English
Subject area: Engineering Mechanics
Doctoral student: Martin Leskovec , Teknisk mekanik
Opponent: Professor Mark Martinez, University of British Columbia
Supervisor: Professor Fredrik Lundell, Linné Flow Center, FLOW, Wallenberg Wood Science Center, SeRC - Swedish e-Science Research Centre, Teknisk mekanik; Professor Daniel Söderberg, Linné Flow Center, FLOW, Wallenberg Wood Science Center, Teknisk mekanik, Fiberprocesser
QC 241021
Abstract
Fluids consisting of dispersed spherical and non-spherical particles are commonly found in many natural and industrial processes. The intricate interplay between particle-particle and particle-fluid interactions, along with flow dynamics and geometries, leads to complex flow phenomena, which still needs to be better understood. Accurate modelling of such flows are crucial for predicting and optimizing processes occurring in, for example, paper industry and wastewater treatment. Both experimental and numerical approaches have been utilized to obtain the flow information, each offering distinct insights. However, a major challenge lies in developing and validating numerical models due to limited or non-existent experimental data. Capturing and measuring these complex flows experimentally is difficult, due to limited optical access, high solid concentrations and often, the measurement probes can affect the flow. On the other hand, numerical simulations can capture these intricate interactions, however, they rely on assumptions and are constrained by high computational expenses.
For high solid concentrations of particles, magnetic resonance velocimetry (MRV) has emerged as a successful non-invasive experimental technique capable of capturing the flow phenomena. Nonetheless, the existing MRV protocols contains questionable assumptions and is dependent on specific parameters for measurements. This thesis addresses these challenges by investigating and characterizing turbulent flows of non-Newtonian fluids through combined MRV measurements and computational fluid dynamics (CFD) simulations. The work also demonstrates how robust quantitative comparisons between MRV measurements and CFD simulations can be achieved, paving the way for development of accurate, calibrated flow models and highlighting the need for size-dependent rheological models in particle-laden turbulent pipe flows.
The findings of this work contribute to a deeper understanding of complex flows, offering a pathway for improved predictive models that can be applied in various industrial and environmental contexts. The research also highlights the potential of MRV to overcome traditional experimental limitations and underscores the importance of integrating experiments and numerical simulations to advance the study of non-Newtonian and complex fluid flows.