In this paper, we develop a geometrically flexible technique for computational fluid-structure interaction (FSI). The motivating application is the simulation of tri-leaflet bioprosthetic heart valve function over the complete cardiac cycle. Due to the complex motion of the heart valve leaflets, the fluid domain undergoes large deformations, including changes of topology. The proposed method directly analyzes a spline-based surface representation of the structure by immersing it into a non-boundary-fitted discretization of the surrounding fluid domain. This places our method within an emerging class of computational techniques that aim to capture geometry on non-boundary-fitted analysis meshes. We introduce the term "immersogeometric analysis" to identify this paradigm.The framework starts with an augmented Lagrangian formulation for FSI that enforces kinematic constraints with a combination of Lagrange multipliers and penalty forces. For immersed volumetric objects, we formally eliminate the multiplier field by substituting a fluid-structure interface traction, arriving at Nitsche's method for enforcing Dirichlet boundary conditions on object surfaces. For immersed thin shell structures modeled geometrically as surfaces, the tractions from opposite sides cancel due to the continuity of the background fluid solution space, leaving a penalty method. Application to a bioprosthetic heart valve, where there is a large pressure jump across the leaflets, reveals shortcomings of the penalty approach. To counteract steep pressure gradients through the structure without the conditioning problems that accompany strong penalty forces, we resurrect the Lagrange multiplier field. Further, since the fluid discretization is not tailored to the structure geometry, there is a significant error in the approximation of pressure discontinuities across the shell. This error becomes especially troublesome in residual-based stabilized methods for incompressible flow, leading to problematic compressibility at practical levels of refinement. We modify existing stabilized methods to improve performance.To evaluate the accuracy of the proposed methods, we test them on benchmark problems and compare the results with those of established boundary-fitted techniques. Finally, we simulate the coupling of the bioprosthetic heart valve and the surrounding blood flow under physiological conditions, demonstrating the effectiveness of the proposed techniques in practical computations.
|Original language||English (US)|
|Number of pages||49|
|Journal||Computer Methods in Applied Mechanics and Engineering|
|State||Published - Feb 1 2015|
Bibliographical noteFunding Information:
Funding for this work was supported by NIH/NHLBI grants R01 HL108330 and HL119297 , and FDA contract HHSF223201111595P . T.J.R. Hughes was supported by grants from the Office of Naval Research ( N00014-08-1-0992 ), the National Science Foundation ( CMMI-01101007 ), and SINTEF ( UTA10-000374 ) with the University of Texas at Austin. M.-C. Hsu and Y. Bazilevs were partially supported by ARO grant No. W911NF-14-1-0296 . D. Kamensky was partially supported by the CSEM Graduate Fellowship. D. Schillinger was partially supported by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) under grants SCHI 1249/1-1 and SCHI 1249/1-2. We thank the Texas Advanced Computing Center (TACC) at the University of Texas at Austin for providing HPC resources that have contributed to the research results reported in this paper. We would also like to thank Dr. Laura De Lorenzis at Technische Universität Braunschweig for helpful discussions on the contact problem and related algorithms.
- Bioprosthetic heart valve
- Fluid-structure interaction
- Immersogeometric analysis
- Isogeometric analysis
- Nitsche's method
- Weakly enforced boundary conditions