Cornell University
A forked-accelerated simulation technique (FAST) for modeling blood flow in the cardiovascular system
Abstract
dc:description.abstractEmploying simulation to customize patient treatment requires an affordable and fast computational technique due to the time-critical nature of the clinical intervention, with some being limited to a few hours. To lower this cost and allow for seamless adoption of predictive simulations in patient-specific treatment planning, this thesis proposes a novel strategy called Forked-Accelerated Simulation Technique (FAST). FAST is built on a reduced-order model (ROM) network and orthogonal solution basis (OSB), which are constructed based on a time-spectral Complex-valued Stokes solver (SCVS). FAST presents the following improvements comparing to the traditional computational fluid dynamics (CFD) solver. Firstly, it provides a significant saving in computational cost for boundary conditions smoothly varying in time, as is typically the case of blood flow simulation. It accomplishes this feat by solving only a few modes rather than thousands of time steps by formulating the problem in the time-spectral domain. Secondly, the proposed technique is embarrassingly parallelizable as the modes and solution basis are linearly independent, thus enabling scalable calculations at a much larger number of processors. Thirdly, the ROM and solution basis constructed in FAST are reusable for different boundary conditions, which provides a highly efficient solution procedure for inverse problems such as parameter identification, uncertainty quantification, and optimization. The comparison of FAST and a traditional stabilized finite element CFD solver is performed using five canonical and complex geometries. The results show that the FAST algorithm can provide accurate results at less than 20% CPU hours and 1.6% wall-clock time of the standard CFD solver.
Degree
thesis:*- Name thesis:degree_name
- M.S., Mechanical Engineering
- Level thesis:degree_level
- Master of Science
- Discipline thesis:degree_discipline
- Mechanical Engineering
- Grantor
- Cornell University
- Year dc:date.issued
- 2020
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Meng, Chenwei
- Chair dc:contributor.chair
-
- Moghadam, Mahdi Esmaily
- Committee member dc:contributor.committeemember
-
- Desjardins, Olivier
Rights
dc:rights- Statement dc:rights
-
- Attribution 4.0 International
- Licence dc:rights.uri
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Dc Identifier Bibid
- 13312153
- Dc Identifier Other
-
ProQuest Submission ID: 11130
ProQuest Publication ID: 28261177