Font Size: a A A

Development and validation of a computational fluid dynamic methodology for pulsatile blood pump design and prediction of thrombus potential

Posted on:2009-10-11Degree:Ph.DType:Dissertation
University:The Pennsylvania State UniversityCandidate:Medvitz, Richard BrianFull Text:PDF
GTID:1442390005956445Subject:Engineering
Abstract/Summary:
A computational fluid dynamics (CFD) methodology is developed to assess the hydrodynamic and hemodynamic performance of a positive displacement left ventricular assist device (LVAD). The pulsatile behavior of the device is developed by implementing valve and piston models to simulate the dynamic effects of valve closure and chamber compression. High fidelity unstructured meshes along with an implicit-LES (Large Eddy Simulation) approach to turbulence are used to model the transitional to low Reynolds number turbulent flows. Grid, time step, and piston modeling studies are performed to evaluate the computational methodology. The thrombosis susceptibility potential (TSP) is developed to gain insight into the device hemodynamic performance and elucidate regions of consistently low wall shear as potential deposition sites. The area- and time-integrated TSP value allowed a quantitative comparison as a means of ranking device performance. A scaling analysis is performed to study the effects of geometric scale on thrombus deposition. Several design variants are analyzed and validated against in vitro PIV data by comparing velocity and wall shear stress measurements. Finally, various mechanical valve types and valve orientations are assessed.;The computational results compare well against in vitro measurements, demonstrating similar mitral jet and chamber rotational patterns within the device. These comparisons give confidence that CFD adequately predicts the device flow field, including aspects which are not easily measured experimentally, in particular the device wall shear. A scaling analysis demonstrated which device geometric and flow field parameters are important for device scaling. The CFD results show that maintaining Reynolds and Strouhal number when geometrically scaling from the 70 cc to the 50 cc device results in a less thrombogenic chamber flow. Design comparisons indicate that the addition of a curved front face to the chamber design increases the TSP of the modified device. Finally, the use of Bjork-Shiley mono-strut valves in the mitral port yields increased wall washing within the device.
Keywords/Search Tags:Computational, Device, Methodology, CFD, Wall, Valve
Related items