Rochester Institute of Technology
Design Strategy for the Optimization Of the Impeller in an Axial Ventricle Assist Device
Abstract
dc:description.abstractThe objectives of this study were to create and implement a design strategy to improve the impeller design of the Left Ventricle Assist Device (LVAD) here at RIT. CFD software paired with parametric design optimization has proved as a useful tool set to improve VAD performance in a timely and cost effective manner. A design search of one design variable was conducted to test the implementation of the design strategy. Performance of a LVAD is split between three individual parameters: 1) Pressure generation, 2) Rate of Hemolysis, and 3) Risk of Thrombosis. CFD simulations provided the data necessary to assess all areas of performance to provide a basis for comparing performances of created designs versus the original impeller. The current version of the LVAD was modeled using CFD simulations and was tested for repeatability and accuracy with bench top test data serving as the basis. Data collected from the simulations was then assessed to calculate rate of hemolysis, thrombosis risk and the pressure performance of design iteration. A weighted matrix was used to create a means to compare overall performance. Iterations were created until the local optimum was found for the design parameter tested. Pressure performance raised in all designs was tested, while there were mixed results in terms of rate of hemolysis and thrombosis risk. The top performing design dropped the risk of thrombosis by 10.8%, reduced rate of hemolysis by 0.8% and increased pressure output by 17.4% when compared to the original impeller performance. Various matrix weighting schemes were applied to verify the design as the local optimal. This also served to find the most appropriate weight balance to the three performance parameters. The design strategy that was tested and implemented has proven successful by creating an optimal impeller design for the design space explored. This study has provided the basis for a more complete design search to be completed over multiple design parameters. Another area for future studies is to complete bench top testing of the optimal design iteration created. The results from the testing can be used to further improve blood damage modeling by raising the accuracy of performance assessment.
Degree
thesis:*- Name thesis:degree_name
- Mechanical Engineering (MS)
- Level thesis:degree_level
- Thesis
- Discipline thesis:degree_discipline
- Mechanical Engineering (KGCOE)
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Peyton, Jonathan
- Contributors dc:contributor
-
- Steven Day
Subjects
dc:subject × 3Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.rit.edu/theses/8601