Investigation of aortic aneurysms by using computational fluid dynamics (CFD) in terms of hemodynamics

Vascular flow conditions play an important role in development of cardiovascular diseases. In this dissertation, aortic aneurysm was investigated by Computational Fluid Dynamics (CFD) and Fluid-Structure Interaction (FSI) methods. It is aimed to determine rupture risks that may occur with increase in the Ascending Aortic Diameter (AAD). Aortic geometries of individuals obtained from computed tomography were constructed by segmentation technique. Patient-specific studies were performed with a normal aorta and a aneurpatient with two different aneurysms. The velocity profile and boundary conditions in aorta were obtained by clinical measurements. Rigid and elastic artery analyzes were performed. Newtonian blood model and five different non-Newtonian viscosity models were used. A machine learning model in Python was developed on rupture risks by using Artificial Neural Networks (ANN) method, which is one of artificial intelligence methods with numerical results. When CFD results were examined, it was observed that pressure on vessel of aorta with aneurysm was 18.56% and 24.7% higher than normal aorta. Shear stresses were lower in aneurysm regions. In viscosity-based analyzes, it was observed that non-Newtonian viscosity models present a difference of 32.5%, 25.7%, 22.4%, 12.3% and 8.7% in shear stresses compared to Newtonian blood model. Elastic artery analyzes showed lower shear stress than rigid vessel analyses, while risk of rupture was found to increase up to three times. Risk of rupture in aorta was estimated by safety factor with R2 performance of 0.986. As a result, it has been shown that CFD and FSI analyzes can support clinical methods with engineering studies.