Content area
Abstract
Numerical simulation has become an indispensable tool for investigating physical phenomena that are difficult or impossible to probe experimentally, particularly in biological and physiological systems where in vivo measurements are limited by resolution, cost, or technical constraints. This dissertation advances the application of the lattice Boltzmann method (LBM) to model complex flow phenomena across two broad domains: cardiovascular hemodynamics and biologically inspired propulsion. In addition to application-driven studies, it introduces methodological advancements in pressure boundary condition (BC) treatment for LBM, addressing a persistent challenge in the field.
The first study presents a fast and robust algorithm for estimating parameters of the three-element Windkessel (WK3) model for use as outlet pressure BCs in multi-scale computational fluid dynamics (CFD) simulations of patient-specific aortic flows. By combining a single steady-state 3D simulation with an optimization framework that explicitly accounts for geometric resistance, the method achieves physiologically accurate pressure and flow distributions at minimal computational cost. Validation in normal and pathological geometries demonstrates its reliability and potential for clinical integration using non-invasive measurements.
The second study develops two novel LBM pressure BC schemes – the Density Correction (DC) and Single Node (SN) schemes, designed to improve accuracy, stability, and locality in simulations involving curved or unaligned boundaries. Comparisons with existing methods in canonical and bifurcation flows show that both schemes achieve second-order accuracy and extended stability ranges, offering new approaches for pressure-driven LBM applications.
The third study applies an LBM–immersed boundary method (IBM) framework to investigate the unsteady propulsion of Cuvierina atlantica (commonly known as sea butterfly), a shelled pteropod employing an overlap-and-fling swimming motion. High-fidelity body and wing kinematics derived from experimental video are incorporated to resolve the detailed vortex dynamics, revealing that propulsion is primarily lift-based, driven by leading- and trailing-edge vortex (LEV/TEV) interactions. Parametric analyses show how viscosity, flapping frequency, and Strouhal number influence thrust and efficiency, with the species’ natural frequency aligning with the optimal Strouhal range for swimming performance.
Finally, the work demonstrates the versatility of LBM in tackling diverse and complex flow problems, from accelerating clinically relevant cardiovascular simulations to advancing boundary condition formulations and elucidating biological propulsion mechanisms. The methodological and application-oriented contributions provide a foundation for future research, bridging numerical technique development with real-world fluid dynamic challenges.