The aortic valve is a valve made of three thin flexible leaflets which open and close in order to allow the unidirectional passage of blood from the left ventricle to the aorta. Because of this repetitive motion at each cycle, the valve is constantly exposed to large variations of pressure and hemodynamic forces. Deterioration in its peculiar functioning could lead to the arising of valvular diseases such as stenosis or regurgitation. A widely spread procedure to cure severely damaged heart valves is the replacement with artificial heart valves. However, up to now, the natural aortic valve properties and functionalities have never been equalised by any manufactured prototypes in terms of efficiency and durability. This limitation has inspired many engineering researches with the purpose of gaining a complete understanding on the working principles of the valve and the evolution of the flow through it in order to enhance the performances of these medical devices.
In this thesis, numerical simulations of blood passing through the aortic valve have been performed. In particular, this study proposes an alternative investigating approach which consists on prescribing the rigid motion of the leaflets a priori as taken from experimental values and without the employment of the structural solver. To execute this strategy, the Arbitrary Lagrangian-Euler method along with remeshing techniques has been employed in Ansys Fluent. The main objectives concern the study of the resulting hemodynamics in the aortic root and the validation with other investigating techniques such as experimental work and Fluid-Structure Interaction simulations.
First, the methodology was tested in a two-dimensional geometry which embodied a simplified representation of the aortic root. Secondly, a more realistic three-dimensional case was analysed with the employment of a commercial bileaflet mechanical valve. In this case, two different methods were adopted to deal with the transition of the flow to a turbulent regime. Namely, the analysis was conducted both by the resolution of the unsteady Reynolds-Averaged Navier-Stokes equations with the k-ε model and by maintaining a laminar approach.
Validation of the method with available experimental data showed that the two-dimensional case is well reproduced. On the other hand, mainly qualitative agreement is obtained for the three-dimensional case. This suggests that the proposed approach is able to capture the main features of the flow to a certain extent and future investigations are required to obtain more accurate predictions of the flow.