A novel multiblock immersed boundary method enabling high order large eddy simulation of pathological and medical device hemodynamics

Kameswararao Anupindi, Purdue University

Abstract

Computational fluid dynamics (CFD) simulations are becoming a reliable tool in understanding disease progression, investigating blood flow patterns and evaluating medical device performance such as stent grafts and mechanical heart valves. Previous studies indicate the presence of highly disturbed, transitional and mildly turbulent flow in healthy and pathological arteries. Accurate simulation of the transitional flow requires high order numerics together with a scale resolving turbulence model such as large eddy simulation (LES). This in turn limits one to use a structured fluid flow solver on which complex, branching arterial domains that are typical in the human blood circulatory system could not be handled. To overcome this, a novel multiblock based immersed boundary method (IBM) is developed based on high order discretization schemes that can efficiently simulate blood flow in complex arterial geometries using structured Cartesian fluid flow solvers. The developed solver, WenoHemo, is systematically validated for each of the newly introduced numerics using a variety of numerical and experimental results available in the literature. Three dimensional laminar flow over a sphere, laminar flow in a backward facing step, laminar and transitional flow in an abdominal aortic aneurysm (AAA), transitional flow in a model stenosed artery, and turbulent flow in a mixing layer are used as benchmark cases for validating the solver thoroughly. WenoHemo is then applied to study blood flow patterns in a pathological thoracic aortic aneurysm (TAA) and in a resulting thoracic aorta with a stent graft (TASG) geometry after an endovascular repair (EVAR). Phase averaged velocity profiles, turbulence kinetic energy levels, viscous wall shear stresses and turbulence energy spectra are used to compare the similarities and differences between the blood flow patterns obtained. Presence of well developed turbulence is detected in the case of TAA whereas TASG showed periodic vortex shedding with lower turbulence levels and improved blood flow to the descending aorta. Application of the solver to simulate blood flow patterns obtained in a bi-leaflet mechanical heart valve (BMHV) placed in a model aorta with imposed kinematics of the leaflets is also carried out, which reveals complex blood flow patterns that need to be considered in the design of the same for reliability and to reduce post surgical complications.

Degree

Ph.D.

Advisors

Frankel, Purdue University.

Subject Area

Biomedical engineering|Mechanical engineering

COinS