Coupling edge-based smoothed finite element method with smoothed particle hydrodynamics for fluid structure interaction problems
Introduction
FSI is a significant field in aerospace, marine engineering, biomechanics and other applications (Bennet and Edwards, 1998; Van Loon and Van de Vosse, 2010; Taylor and Figueroa, 2009). In recent decades, a plenty of numeric methods have been proposed for FSI problems, among which the grid-based method occupies the dominant position in numerical research. The Lagrangian grid-based methods, such as the Finite Element Method (FEM) (Strang et al., 1973), are commonly used for structural modeling. While the Eulerian grid-based methods, such as the Finite Difference Method (FDM) (Brandimarte, 1960), the Finite Volume Method (FVM) (Versteeg and Malalasekera, 1995) and Finite Element Method (FEM) (Strang et al., 1973), are classically used for fluid modeling. There are some ways for coupling Eulerian grid and Lagrangian grid, such as the Immersed Boundary Method (IBM) (Peskin, 2002) and Arbitrary Lagrangian Eulerian (ALE) method (Donea et al., 1982; Hirt et al., 1974). However, these methods are time-consuming for tracking moving boundaries or interfaces (Donea et al., 1982; Hirt et al., 1974; Morinishi and Fukui, 2012; Peskin, 2002). It is very attractive to solve the FSI problem under the unified Lagrangian description, because it can deal with the moving interface naturally. Recently, the FSI problem is solved by coupling the finite element method with mesh free methods (Fourey et al., 2010; Groenenboom and Cartwright, 2009; Hu et al., 2014; Long et al., 2017; Vuyst et al., 2005; Yang et al., 2012; Zhang and Wan, 2018), and the coupling approaches leverage the advantages of convenience of mesh free methods to simulate hydrodynamics and robustness of FEM to solve structural dynamics.
The FEM has become a powerful and effective technique for numeric simulation in science and engineering. However, some shortcomings also exist in the conventional FEM. It is well-known that the ‘‘overly-stiff’’ characteristic of the conventional finite element models often leads to a significant reduction in the accuracy of numerical solutions (Liu and Quek, 2003). Certainly, the FEM-SPH approaches have the ‘‘overly-stiff’’ characteristic because of conventional finite element model. Thus, it is important to develop an effective numerical technique for dealing with this issue. The smoothed finite element method (S-FEM) was proposed by Liu et al. (2007a), which is based on the gradient smoothing technique of mesh-less method (Chen et al., 2001). The gradient smoothing technique provides a good way to reduce the ‘‘over-stiffness’’ (Liu, 2008). A new class of smoothing methods have been developed that can effectively “soften” the numerical model, such as cell based S-FEM (CS-FEM) (Liu et al., 2007b), nodal based S-FEM (NS-FEM) (Liu et al., 2009b), edge based S-FEM (ES-FEM) (Liu et al., 2009a) and face based S-FEM (FS-FEM) (Nguyen-Thoi et al., 2009). In addition, thanks to the gradient smoothing technique, smoothed point interpolation methods (S-PIMs) have been proposed (Gui-rong and Gui-yong, 2013; Liu and Zhang, 2008), and S-PIMs are developed for FSI problems (Wang et al., 2017, 2018; Zhang et al., 2018). Furthermore, S-FEM is also utilized to FSI problems. Zhang et al. proposed Immersed S-FEM (IS-FEM) to solve FSI problems (Zhang et al., 2012). Recently, Jiang et al. proposed a sharp-interface IS-FEM for interactions between large deformation structures and incompressible flows, the method has great application prospects for FSI problems(Jiang et al., 2018). Since the ES-FEM provides more precise solutions than the traditional linear FEM (Liu et al., 2009a). In this paper, the ES-FEM is used to model the structure domain.
The SPH approach is a purely Lagrangian mesh-free approach, which was originally developed by Lucy (1977) and Monaghan and Gingold (Gingold and Monaghan, 1977; Monaghan and Gingold, 1983). It is widely applied in astrophysics, impact dynamics, fluid mechanics and other fields. Recently, Han and Hu used a uniform SPH frame to modeling fluid structure interaction problems (Han and Hu, 2018). Yang et al. proposed an improved SPH-EBG (Element bending group) method for interactions between free surface flows and flexible structures (Yang et al., 2016). Zhang et al. reviewed the latest developments in SPH methods and their typical applications in FSI problems (Zhang et al., 2017). Asai et al. proposed a stabilized Incompressible SPH to simulate the free surface flow problems (Asai et al., 2012). He et al. (2020) used a finite particle method to modeling the solitary wave breaking over a slope. In this article, the SPH approach is applied to solve the fluid domain. The meshless property of SPH approach eliminates the troubles caused by large deformations. Therefore, SPH is quite suitable to solve fluid flows with moving interfaces and free surfaces.
In the FEM-SPH model, the SPH is used to solve the fluid domain and the FEM is applied to solve the solid domain, where the coupling strategy usually utilizes a pure Lagrangian description in both media. The FEM-SPH coupling approach was firstly developed by Attaway et al. (1994) to investigate impact problems. Vuyst et al. (2005) investigated fluid structure impact problems using an FEM-SPH model. The FEM-SPH coupling approach can also be applied to interaction between elastic structures and free surface flow (Fourey et al., 2010; Groenenboom and Cartwright, 2009; Hu et al., 2014; Yang et al., 2012). The master-slave schemes are utilized in most FEM-SPH models. And the contact forces are used to prevent particles from entering the element mesh (Attaway et al., 1994; Groenenboom and Cartwright, 2009; Hu et al., 2014). Each coupling approach has its own specific feature. For the modeling of complex FSI problems, some FEM-SPH coupling approaches without contact algorithm are also proposed (Fourey et al., 2010; Long et al., 2017; Yang et al., 2012). Fourrey et al. (Fourey et al., 2010) utilized the virtual particles coupling strategy to realize the FEM-SPH model. Yang et al. (2012) used Monaghan type repulsive force to prevent the particles penetrate the element meshes. Long et al. (2017) developed a virtual particle coupling scheme for FEM-SPH model with complex geometry interface, in which the virtual particles are produced through discretizing the truncated region of support domain of kernel function. Recently, Fourey et al. (2017) proposed an efficient coupling scheme between SPH and FEM for violent fluid–structure interaction modeling.
In this paper, the coupling method of ES-FEM and SPH is proposed for simulating FSI problems, and the virtual particles coupling scheme is used for the ES-FEM-SPH model. Three-node triangular elements are used to discretize the structure domain, which is simple and flexible. The particles are used to discretize the fluid domain and the FSI in ES-FEM-SPH model is implemented through deploying virtual particles within ES-FEM elements. Since ES-FEM provides a more precise solution than traditional linear FEM using the same constant strain triangular element, it can be expected that ES-FEM-SPH coupling model will be applicable to FSI problems. Comparisons between the ES-FEM-SPH model and the FEM-SPH coupling approach are given for several numerical examples.
The organization of this paper is as follows. Firstly, the ES-FEM to solve the solid domain is concisely introduced. Secondly, the SPH approach to solve hydrodynamics is briefly introduced. Thirdly, the coupled ES-FEM-SPH model is briefly described. Finally, numerical examples are given to illustrate the effectiveness of the ES-FEM-SPH model.
Section snippets
ES-FEM formulations
In this paper, the ES-FEM for solving the deformation and motion structure based on total Lagrangian format is adopted and the structure is discretized by using linear triangular elements (Zhang and Liu, 2014). The smoothed deformation gradient, smoothed Green Lagrange strain tensor, and the smoothed second Piola-Kirchhoff (PK2) stress tensor are derived at the initial configuration. In ES-FEM, the displacement and the current spacial position are approximated by
Governing equations
In this paper, the working fluid is assumed incompressible and Newtonian. The weakly compressible SPH (WCSPH) method is used for modeling incompressible fluid flows. This treatment actually assumes the incompressible fluid to be slightly or weakly compressible. The mass conservation and momentum conservation equations are as followswhere t is the time, is the density of fluid, is the velocity vector, is the kinematic viscosity of the fluid, p is the
Coupling strategy of SPH with ES-FEM
As shown in Fig. 2, The structure domain is simulated by ES-FEM, and the fluid domain is solved by SPH approach. As shown in Fig. 3, the virtual particle coupling strategy proposed by Long et al. (2017) is used in this paper, in which kernel support domain's truncated region is divided into sub-regions relating to interface segments, and the sub-regions are discretized to generate virtual particles. Then, the interaction pairs between the particle and element can be established by using the
Numerical examples with fluid-structure interaction
In this section, six examples of hydrostatic water on a linear elastic plate, a hyper elastic gate under water pressure, an elastic wall in a lid-driven cavity flow, an elastic beam in a fluid tunnel, flow past a cylinder with an elastic beam and elastic structure impacting on free surface are utilized to validate the ES-FEM-SPH model. Comparisons between the ES-FEM-SPH and the FEM-SPH coupling approach are provided, and the same coupling scheme is used for ES-FEM-SPH and FEM-SPH coupling
Conclusions
In this paper, the recently developed ES-FEM is coupled with SPH for modeling FSI problems. In this coupling approach, ES-FEM is used to model the movement and deformation of structures and SPH is used to model fluid flow. Linear triangular elements are adopted to discretize the structure making the ES-FEM-SPH model suitable for FSI problems with complicated geometries. As ES-FEM is associated with the gradient smoothing technique in structure domain, the “overly-stiff” effect in the
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This work has been partially supported by the National Natural Science Foundation of China (Grant Nos. 11902005 and 12002002) and the National Numerical Wind Tunnel Project (Grant No. NNW2019ZT2-B02).
References (64)
- et al.
Numerical simulation of fluid-structure interaction by SPH
Comput. Struct.
(2007) - et al.
Coupling of smooth particle hydrodynamics with the finite element method
Nucl. Eng. Des.
(1994) - et al.
An arbitrary Lagrangian-eulerian finite element method for transient dynamic fluid-structure interactions
Comput. Methods Appl. Mech. Eng.
(1982) - et al.
An efficient FSI coupling strategy between smoothed particle hydrodynamics and finite element methods
Comput. Phys. Commun.
(2017) - et al.
Numerical investigation of the solitary wave breaking over a slope by using the finite particle method
Coast Eng.
(2020) - et al.
An arbitrary Lagrangian-Eulerian computing method for all flow speeds
J. Comput. Phys.
(1974) - et al.
Fluid–structure interaction analysis by coupled FE–SPH model based on a novel searching algorithm
Comput. Methods Appl. Mech. Eng.
(2014) - et al.
A sharp-interface immersed smoothed finite element method for interactions between incompressible flows and large deformation solids
Comput. Methods Appl. Mech. Eng.
(2018) - et al.
Incompressible smoothed particle hydrodynamics for free-surface flows: a generalised diffusion-based algorithm for stability and validations for impulsive flows and propagating waves
J. Comput. Phys.
(2012) - et al.
An edge-based smoothed finite element method (ES-FEM) for static, free and forced vibration analyses of solids
J. Sound Vib.
(2009)
A node-based smoothed finite element method (NS-FEM) for upper bound solutions to solid mechanics problems
Comput. Struct.
An arbitrary boundary with ghost particles incorporated in coupled FEM-SPH model for FSI problems
J. Comput. Phys.
Simulating free surface flows with SPH
J. Comput. Phys.
Shock simulation by the particle method SPH
J. Comput. Phys.
An Eulerian approach for fluid–structure interaction problems
Comput. Fluids
Modeling low Reynolds number incompressible flows using SPH
J. Comput. Phys.
Hydroelastic behaviour of a conical shell impacting on a quiescent-free surface of an incompressible liquid
J. Sound Vib.
An improved SPH method for modeling liquid sloshing dynamics
Comput. Struct.
The δplus-SPH model: simple procedures for a further improvement of the SPH scheme
Comput. Methods Appl. Mech. Eng.
Fluid–structure interaction involving large deformations: 3D simulations and applications to biological systems
J. Comput. Phys.
Accuracy and stability in incompressible SPH (ISPH) based on the projection method and a new approach
J. Comput. Phys.
Free-surface flow interactions with deformable structures using an SPH–FEM model
Ocean Eng.
Numerical modeling of dam-break flow impacting on flexible structures using an improved SPH–EBG method
Coast Eng.
Smoothed particle hydrodynamics and its applications in fluid-structure interactions
J. Hydrodyn.
Coupling immersed method with node-based partly smoothed point interpolation method (NPS-PIM) for large-displacement fluid-structure interaction problems
Ocean Eng.
MPS-FEM coupled method for sloshing flows in an elastic tank
Ocean Eng.
Solution bound and nearly exact solution to nonlinear solid mechanics problems based on the smoothed FEM concept
Eng. Anal. Bound. Elem.
A fixed-mesh method for incompressible flow–structure systems with finite solid deformations
J. Comput. Phys.
A stabilized incompressible SPH method by relaxing the density invariance condition
J. Appl. Math.
An Overview of Recent Developments in Computational Aeroelasticity, the 29th AIAA Fluid Dynamics Conference, Albuquerque, NM
Benchmarking a coupled immersed-boundary-finite-element solver for large-scale flow-induced deformation
AIAA J.
Finite-difference Methods for Partial Differential Equations
Cited by (50)
Numerical modelling and experimental validation of squeezing flows in the automobile production
2024, International Journal of Adhesion and AdhesivesA novel artificial neural network-based interface coupling approach for partitioned fluid–structure interaction problems
2023, Engineering Analysis with Boundary ElementsCoupling edge-based smoothed finite element method with incompressible smoothed particle hydrodynamics for thermal fluid structure interaction problems
2023, Journal of Fluids and StructuresA finite strain model for multi-material, multi-component biomechanical analysis with total Lagrangian smoothed finite element method
2023, International Journal of Mechanical SciencesCitation Excerpt :The core idea is to transform the integral operation from the element into the smoothing domain through gradient smoothing technology [14]. The typical S-FEM contains: the cell-based smoothed finite element (CS-FEM) [15–19], the node-based smoothed finite element (NS-FEM) [20–23], the face-based smoothed finite element (FS-FEM) [24–26], the edge-based smoothed finite element (ES-FEM) [27–29]. Because of the different ways of constructing the integration domain, these methods have different properties, which is completely discussed in the review article [30].
Multi-level adaptive particle refinement method with large refinement scale ratio and new free-surface detection algorithm for complex fluid-structure interaction problems
2023, Journal of Computational PhysicsCitation Excerpt :In this figure, the result using the refinement scale ratio 2 almost coincides with that using the scale ratio 4. The present results agree very well with the semi-analytical solution by Scolan [57] and the results solved by the method coupling SPH with FEM by Fourey et al. [58], and are closer to the FEM-SPH results by Long et al. [59]. Regarding the vertical force applied on the wedge in Fig. 17(b), the results of simulations (i) and (ii) agree with each other quite well, and the present results also have a good agreement with the results by Fourey et al. [58], Long et al. [59] and the semi-analytical solution [57].