Dancing with the flow: Numerical modelling of non-ideal micro-particles in fluids
Year: 2026
Type: dissertation
Abstract: Particles in flows are ubiquitous and are key to various natural and industrial processes, encompassing a diverse range, from asbestos fibres, pollen, volcanic ash and aerosols in the air to aquatic microorganisms, drug delivery particles and micro/nano robots. Their behaviour can vary widely. To date, accurately describing particle motion in fluids remains a major challenge in science and engineering, especially for micron-sized (or smaller) particles where the complexity and cost of experimental techniques and computational models are high. In cases where large particle populations need to be considered, the Lagrangian point-particle method in combination with state-of-the-art CFD simulations is the method of choice. So far, research typically simplifies the complexity of particle dynamics by assuming spherical or ellipsoidal shapes, a homogeneous mass distribution, isolated behaviour (neglecting the interactions between particles and adjacent boundaries) and rigid material properties - idealizations that often fail to capture the real-world particle dynamics. In this work, useful extensions of state-of-the-art Lagrangian point particle tracking are presented that allow efficient modelling of dilute dispersed non-ideal particles with diverse characteristics in flow suspensions. In this context, a novel model for inhomogeneous particles is introduced, where inhomogeneity is exemplified by a spherical mass inclusion embedded in the particle matrix. Furthermore, non-spherical particles are represented by the superellipsoidal surface equation, where a novel surrogate model is implemented to describe the complex barycentric and orientational dynamics. Furthermore, a novel superellipsoid particle contact model based on the Lagrangian multiplier technique in combination with the Newton-Raphson method is proposed. In addition, a novel tangential restitution model (applicable to arbitrary particle shapes) is presented to account for friction during particle-particle and particle-wall interactions. Finally, an utmost efficient and highly versatile model for soft particles in the Lagrangian point-particle framework is developed. The model is based on the pseudo-rigid body theory with affine deformations and does not require surface discretization. Note that all extensions discussed were implemented within the open-source code OpenFOAM. All models successfully reproduce pertinent benchmark cases from the literature and are broadly applicable to arbitrary macroscopic flow fields. These contributions significantly enhance the Lagrangian point-particle method and enable more accurate predictions of realistic particle behaviour. By expanding the range of problems that can be studied, this work can contribute to a better understanding in fields such as environmental science, biomedicine and industrial processing, and paves the way for optimising existing systems.
Author: Wedel, Jana
Language: English
Subfield: Ocean Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7121685191
Cites:
Cited by:
Related to: