Projection-based stabilization of interface Lagrange multipliers in immersogeometric fluid–thin structure interaction analysis, with application to heart valve modeling
Abstract:This paper discusses a method of stabilizing Lagrange multiplier fields used to couple thin immersed shell structures and surrounding fluids. The method retains essential conservation properties by stabilizing only the portion of the constraint orthogonal to a coarse multiplier space. This stabilization can easily be applied within iterative methods or semi-implicit time integrators that avoid directly solving a saddle point problem for the Lagrange multiplier field. Heart valve simulations demonstrate applica… Show more
“…Moreover, the exact solution of the Eulerian pressure is usually discontinuous at the fluid-solid interface, which leads to poor approximation properties of the discrete pressure spaces used in immersed FSI methods. As a result, immersed FSI methods, whether they follow Peskin's idea or not (see [35,36,37,38,39]), often have profound difficulties to accurately impose the incompressibility constraint at both Eulerian and Lagrangian levels [40,41,6,34,38,25,42]. This issue is extremely important since large errors in the incompressibility constraint are able to even alter the qualitative behavior of numerical solutions in challenging FSI applications, such as heart valves [40,38] and cell-scale blood flow [25].…”
Section: Introductionmentioning
confidence: 99%
“…To the best of our knowledge, the only work that has tried pointwise divergence-free Eulerian discretization in the context of immersed FSI methods is the recent paper [46]. In [46], divergence-conforming B-splines [47,48,49,50] were applied to the nonboundary-fitted FSI method developed in [38,39]. The method used in [38,46] defines a fluid subproblem and a Kirchhoff-Love shell subproblem with no-tailored discretizations.…”
Section: Introductionmentioning
confidence: 99%
“…In the last decade, the use of spline functions, such as B-splines, non-uniform rational B-splines (NURBS), analysis-suitable T-splines (ASTS), and hierarchical B-splines, has become widespread in computational mechanics thanks to isogeometric analysis (IGA) [51,54]. In the field of immersed methods for FSI, IGA has already been used to perform NURBS-based and ASTS-based generalizations of the IB method [15,16], develop the immersogeometric method [38,55,39], couple shells with Stokes flows using the boundary integral method [56,57], solve air-blast problems [58], develop a fictitious domain approach [59], and a stabilized cut-cell immersed framework [60].…”
Immersed boundary (IB) methods deal with incompressible visco-elastic solids interacting with incompressible viscous fluids. A long-standing issue of IB methods is the challenge of accurately imposing the incompressibility constraint at the discrete level. We present the divergence-conforming immersed boundary (DCIB) method to tackle this issue. The DCIB method leads to completely negligible incompressibility errors at the Eulerian level and various orders of magnitude of increased accuracy at the Lagrangian level compared to other IB methods. Furthermore, second-order convergence of the incompressibility error at the Lagrangian level is obtained as the discretization is refined. In the DCIB method, the Eulerian velocitypressure pair is discretized using divergence-conforming B-splines, leading to inf-sup stable and pointwise divergence-free Eulerian solutions. The Lagrangian displacement is discretized using non-uniform rational B-splines, which enables to robustly handle large mesh distortions. The data transfer needed between the Eulerian and Lagrangian descriptions is performed at the quadrature level using the same spline basis functions that define the computational meshes. This conduces to a fully variational formulation, sharp treatment of the fluid-solid interface, and a 0.5 increase in the convergence rate of the Eulerian velocity and the Lagrangian displacement measured in L 2 norm in comparison with using discretized Dirac delta functions for the data transfer. By combining the generalized-α method and a block-iterative solution strategy, the DCIB method results in a fully-implicit discretization, which enables to take larger time steps. Various two-and three-dimensional problems are solved to show all the aforementioned properties of the DCIB method along with mesh-independence studies, verification of the numerical method by comparison with the literature, and measurement of convergence rates.
“…Moreover, the exact solution of the Eulerian pressure is usually discontinuous at the fluid-solid interface, which leads to poor approximation properties of the discrete pressure spaces used in immersed FSI methods. As a result, immersed FSI methods, whether they follow Peskin's idea or not (see [35,36,37,38,39]), often have profound difficulties to accurately impose the incompressibility constraint at both Eulerian and Lagrangian levels [40,41,6,34,38,25,42]. This issue is extremely important since large errors in the incompressibility constraint are able to even alter the qualitative behavior of numerical solutions in challenging FSI applications, such as heart valves [40,38] and cell-scale blood flow [25].…”
Section: Introductionmentioning
confidence: 99%
“…To the best of our knowledge, the only work that has tried pointwise divergence-free Eulerian discretization in the context of immersed FSI methods is the recent paper [46]. In [46], divergence-conforming B-splines [47,48,49,50] were applied to the nonboundary-fitted FSI method developed in [38,39]. The method used in [38,46] defines a fluid subproblem and a Kirchhoff-Love shell subproblem with no-tailored discretizations.…”
Section: Introductionmentioning
confidence: 99%
“…In the last decade, the use of spline functions, such as B-splines, non-uniform rational B-splines (NURBS), analysis-suitable T-splines (ASTS), and hierarchical B-splines, has become widespread in computational mechanics thanks to isogeometric analysis (IGA) [51,54]. In the field of immersed methods for FSI, IGA has already been used to perform NURBS-based and ASTS-based generalizations of the IB method [15,16], develop the immersogeometric method [38,55,39], couple shells with Stokes flows using the boundary integral method [56,57], solve air-blast problems [58], develop a fictitious domain approach [59], and a stabilized cut-cell immersed framework [60].…”
Immersed boundary (IB) methods deal with incompressible visco-elastic solids interacting with incompressible viscous fluids. A long-standing issue of IB methods is the challenge of accurately imposing the incompressibility constraint at the discrete level. We present the divergence-conforming immersed boundary (DCIB) method to tackle this issue. The DCIB method leads to completely negligible incompressibility errors at the Eulerian level and various orders of magnitude of increased accuracy at the Lagrangian level compared to other IB methods. Furthermore, second-order convergence of the incompressibility error at the Lagrangian level is obtained as the discretization is refined. In the DCIB method, the Eulerian velocitypressure pair is discretized using divergence-conforming B-splines, leading to inf-sup stable and pointwise divergence-free Eulerian solutions. The Lagrangian displacement is discretized using non-uniform rational B-splines, which enables to robustly handle large mesh distortions. The data transfer needed between the Eulerian and Lagrangian descriptions is performed at the quadrature level using the same spline basis functions that define the computational meshes. This conduces to a fully variational formulation, sharp treatment of the fluid-solid interface, and a 0.5 increase in the convergence rate of the Eulerian velocity and the Lagrangian displacement measured in L 2 norm in comparison with using discretized Dirac delta functions for the data transfer. By combining the generalized-α method and a block-iterative solution strategy, the DCIB method results in a fully-implicit discretization, which enables to take larger time steps. Various two-and three-dimensional problems are solved to show all the aforementioned properties of the DCIB method along with mesh-independence studies, verification of the numerical method by comparison with the literature, and measurement of convergence rates.
“…Recent years have seen great interest in numerical analysis of fluid-structure interaction (FSI) due to its relevance to structural [27,28], biomedical [29], and other engineering applications [30]. In a recent series of articles [31][32][33][34][35][36], we developed a framework for simulating FSI dynamics of thin, flexible shell structures immersed in a viscous, incompressible fluid, where we assume that the thin structure can cut through the fluid meshes and the fluid/structure meshes do not have to match each other on the fluid-structure interface [37][38][39]. The target application was bioprosthetic heart valve [40] analysis.…”
Section: Introductionmentioning
confidence: 99%
“…This paper focuses on the particular numerical method introduced in [31], and refinements of it developed in subsequent work [36]. This numerical method is specialized for problems in which the structure is modeled geometrically as a surface of co-dimension one to the fluid sub-problem domain.…”
In this work, we analyze the convergence of the recent numerical method for enforcing fluid–structure interaction (FSI) kinematic constraints in the immersogeometric framework for cardiovascular FSI. In the immersogeometric framework, the structure is modeled as a thin shell, and its influence on the fluid subproblem is imposed as a forcing term. This force has the interpretation of a Lagrange multiplier field supplemented by penalty forces, in an augmented Lagrangian formulation of the FSI kinematic constraints. Because of the non-matching fluid and structure discretizations used, no discrete inf-sup condition can be assumed. To avoid solving (potentially unstable) discrete saddle point problems, the penalty forces are treated implicitly and the multiplier field is updated explicitly. In the present contribution, we introduce the term dynamic augmented Lagrangian (DAL) to describe this time integration scheme. Moreover, to improve the stability and conservation of the DAL method, in a recently-proposed extension we projected the multiplier onto a coarser space and introduced the projection-based DAL method. In this paper, we formulate this projection-based DAL algorithm for a linearized parabolic model problem in a domain with an immersed Lipschitz surface, analyze the regularity of solutions to this problem, and provide error estimates for the projection-based DAL method in both the [Formula: see text] and [Formula: see text] norms. Numerical experiments indicate that the derived estimates are sharp and that the results of the model problem analysis can be extrapolated to the setting of nonlinear FSI, for which the numerical method was originally proposed.
In continuation of our previous work on nonlinear stability analysis of trimmed isogeometric thin shells, this contribution is an extension to dynamic buckling analyses for predicting reliably complex snap‐through and mode‐jumping behavior. Specifically, a modified generalized‐α$$ \alpha $$ time integration scheme is used and combined with a nonlinear isogeometric Kirchhoff–Love shell element to provide second‐order accuracy while introducing controllable high‐frequency dissipation. In addition, a weak enforcement of essential boundary conditions based on a penalty approach is considered with a particular focus on the inhomogeneous case of imposed prescribed displacements. Moreover, we propose a least‐squares B‐spline surface fitting approach and corresponding error measures to model both eigenmode based and measured geometric imperfections. The imperfect geometries thus obtained can be naturally integrated into the framework of isogeometric nonlinear dynamic shell analysis. Based on this idea, the different modeling methods and the influence of the appropriately considered geometric imperfections on the dynamic buckling behavior can be investigated systematically. Both perfect and geometrically imperfect shell models are considered to assess the performance of the proposed method. We compare our method with established developments in this field and demonstrate superior achievements with regard to solution quality and robustness.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.