Fluid-structure Interaction Simulation of the Cerebrovascular Circulation: Immersed Boundary versus Arbitrary Lagrangian Eulerian Mesh Formulations

Abstract

PurposePatient-specific fluid-structure interaction (FSI) simulations allow for the in silico modeling of vascular pathology. Though existing attempts to model the cerebrovasculature confirm the potential of FSI as a future diagnostic tool, standard simulation methods and modeling parameters remain undefined. The purpose of this investigation was to compare immersed boundary (IB) and arbitrary Lagrangian Eulerian (ALE) formulations to discern whether the increased modeling complexity offered by ALE is necessary for the modeling of smaller-caliber vessels, given increased computational load. MethodsDirect comparisons of Fluent and Mechanical behavior were conducted between IB and ALE methods of FSI simulation. Simulations utilized an internal carotid artery geometry conduit with optimized mesh. Boundary constraints were derived from previous investigation of vascular tissue and fit to a Prony series. Both qualitative profile comparisons and quantitative parametric analyses of variance were conducted to assess differences in simulation output. ResultsIn this study, we report deviations in Fluent and Mechanical output between IB and ALE cases of FSI simulation. More specifically, ALE-method simulations boast higher stress, lower wall shear stress, and lower strain. These differences persist across the vessel geometry and increase with high strain. Additionally, inconsistencies between solving methods are exacerbated in areas of more complex mesh geometry (i.e. vessel bifurcation). ConclusionSubstantial alterations in intraluminal stress, shear stress, and strain suggest that ALE formulation is necessary for modeling blood vessels of the cerebrovasculature. Our findings highlight the importance of accurately modeling the dynamic interactions that occur between the fluid and material domains of simulation.