A protocol for automated a posteriori adaptive meshing with SimVascular: a test case

Akash Gupta,Ethan Kung

doi:10.1186/s13104-020-05057-7

Abstract

ObjectiveOperational details regarding the use of the adaptive meshing (AM) algorithm available in the SimVascular package are scarce despite its application in several studies. Lacking these details, novice users of the AM algorithm may experience undesirable outcomes post-adaptation such as increases in mesh error metrics, unpredictable increases in mesh size, and losses in geometric fidelity. Here we present a test case using our proposed iterative protocol that will help prevent these undesirable outcomes and enhance the utility of the AM algorithm. We present three trials (conservative, moderate, and aggressive settings) applied to a scenario modelling a Fontan junction with a patient-specific geometry and physiologically realistic boundary conditions.ResultsIn all three trials, an overall reduction in mesh error metrics is observed (range 47%–86%). The increase in the number of elements through each adaptation never exceeded the mesh size of the pre-adaptation mesh by one order of magnitude. In all three trials, the protocol resulted in consistent, repeatable improvements in mesh error metrics, no losses of geometric fidelity and steady increments in the number of elements in the mesh. Our proposed protocol prevented the aforementioned undesirable outcomes and can potentially save new users considerable effort and computing resources.

Highlights

ResultsAn overall reduction in mesh error metrics is observed (range 47%–86%)
The finite element (FE) method is a powerful tool for simulating complex haemodynamics observed in cardiovascular flows
The protocol resulted in consistent, repeatable improvements in mesh error metrics, no losses of geometric fidelity and steady increments in the number of elements in the mesh

Summary

Results

The net reduction in the MIE after six iterations of each trial is 86% for Trials 2 and 3 and 47% for Trial 1. Trials 2 and 3 representing the aggressive and moderate settings exhibit monotonous downward trends (Fig. 3a). For Trial 1 an initial uptick is followed by a steady decreasing trend. Trials 2 and 3 are characterized by steady increases in the number of elements while in Trial 1, an initial drop in the mesh size is observed, after which a modest increasing trend is established (Fig. 3b). After the first iteration a b c d. Starting from the same initial mesh (Fig. 2c), all three trials exhibit the formation of a vortex in approximately the same region by the sixth iteration. In all three trials it can be observed that the mesh density increases in and around the regions where the velocity gradient is large (i.e. in and around the vortex)

Introduction

Main text

Discussion

Limitations