Date

2017

Document Type

Dissertation

Degree

Doctor of Philosophy

Department

Mechanical Engineering

First Adviser

Liu, Yaling Liu

Abstract

In this dissertation, we have developed a fluid-structure interaction code specifically designed to simulate soft microparticle deformation in biological flow. We have used this tool for two different applications. First, we study red blood cell deformation under shear flow to evaluate stress distribution on membrane and subsequently pore formation on RBC membrane. Second, we utilized this code to show a proof of concept for an idea where we can separate soft particles based on their biophysical properties. In the following, these applications are discussed in more details.Under high shear rates, pores form on RBC membrane through which hemoglobin leaks out and increases free hemoglobin content of plasma leading to hemolysis. We hypothesize that local flow dynamics such as flow rate and shear stress determines blood cell damage. In this dissertation, a novel model is presented to study red blood cell (RBC) hemolysis at cellular level. The goal of the proposed work is to establish multiscale computational techniques to predict the blood cell dynamics and damage in complex flow conditions, i.e., blood-wetting biomedical devices. The cell membrane damage model will be coupled with local fluid flow to study cell deformation and rupture and a generalized cellular level blood cell damage model will be developed based on these simulations. By coupling Lattice Boltzmann and spring connected network models through immersed boundary method, we estimate hemolysis of a single red blood cell under various shear rates. First, we use adaptive meshing to find local strain distribution and critical sites on RBC membrane, then we apply underlying molecular dynamic simulations to evaluate damage. Our approach is comprised of three sub-models: defining criteria of pore formation, calculating pore size, and measuring Hb diffusive flux out of pores. Our damage model uses information of different scales to predict cellular level hemolysis. Results are compared with experimental studies and other models in literature. The developed cellular damage model can be used as a predictive tool for hydrodynamic and hematologic design optimization of blood-wetting medical devices.Isolating cells of interest from a heterogeneous mixture has been of critical importance in biological studies and clinical applications. In this dissertation, we have proposed to use ciliary system in microfluidic devices to isolate target subpopulation of soft particles based on their biophysical properties. In this model, the bottom of microchannel is covered with an equally spaced cilia array which can be magnetically actuated. A series of simulations are performed to study cilia-particle interaction and isolation dynamic. It is shown that these elastic hair-like filaments can influence particle’s trajectories differently depending on their biophysical properties. This modeling study also uses immersed boundary (IB) method coupled with lattice Boltzmann method. Soft particles are simulated by connected network of nonlinear springs. Moreover, cilia is modeled by point-particle scheme. It is demonstrated that active ciliary system is able to continuously and non-destructively sort cells based on their size, shape and stiffness. Ultimately, a design map for fabrication of a programmable microfluidic device capable of isolating various subpopulation of cells is developed. This biocompatible, label-free design can separate cells/soft microparticles with high throughput which can greatly complement existing separation technologies.

Share

COinS