Introduction
The fluid-structure interaction (FSI) problem remains one of the most challenging in computational
mechanics and computational fluid dynamics. Researchers have conducted various studies on certain types
of FSI problems (such as the fluid-rigid body interaction [1, 2, 3, 4, 5], fluids with a non-rotational structure
[6, 7, 8, 9, 10, 11], and FSI with a stationary fluid domain [12, 13, 14, 15]). However, there is a dearth
of practical models on FSI problems involving a structure that rotates and deforms, i.e., an elastic rotor.
The development of mathematical models and numerical methodologies is critical in practice for large-scale
advanced FSI simulations involving an elastic rotor, in order to guide the design, evaluation, and prediction
of various applications, such as hydro-turbines, jet engines, and artificial heart pumps. Therefore, it is of
great significance to develop an efficient and accurate mathematical model and numerical method to handle
fluid-structure interactions involving a rotational and deformable structural motion.
However, numerical simulations of fluids interacting with a rotating structure can be challenging. The
difficulties associated with simulating the rotational structure in FSI stem from the fact that the simulations
rely on the coupling of two distinct descriptions: the Lagrangian description for the solid and the Eulerian
coordinate for the fluid. The arbitrary Lagrangian Eulerian (ALE) method [6, 16, 17, 18, 19] copes with this
difficulty by adapting the fluid mesh to accommodate the deformations of the solid on the interface. When
the ALE method is used, the meshes of the fluid and structure conform on the interface if the Lagrangian
structure mesh is moved to the Eulerian one following ALE mapping. This is important as the degrees
of freedom on the interface are naturally shared by both the fluid and the structure, which facilitates the
implementation of the discretization proposed herein. However, the ALE method has a severe drawback; i.e.,
when the structure has a large displacement or deformation, it is very likely that ALE mapping will distort
the fluid mesh. Even the most advanced and most finely tuned ALE-based scheme cannot perform well
without re-meshing. And, if re-meshing is used to produce the fluid mesh, then the number of mesh nodes
and/or elements over different time levels can no longer be guaranteed to be the same. Thus, interpolations
of variables between every two adjacent time steps are unavoidable, resulting in a time-consuming and even
unstable geometrical process, especially in high-dimension cases. In fact, it is difficult to directly apply the
ALE method to fluid-rotating structure interaction problems.
Many approaches have been proposed to deal with rotational structures in FSI problems. Several of these
approaches model the wind turbine rotor [20, 21, 22, 23, 24] by coupling the finite element method (FEM) for
fluid dynamics, isogeometric analysis (IGA) for structural mechanics, and the non-conforming discretization
on the interface between fluid and structure. In order to apply ALE, these approaches introduce an artificial
cylindrical buffer zone to enclose the rotor; let the mesh of the sliding cylindrical interface weakly enforce the
continuity of the solution fields; and introduce extra unknowns and/or penalties to reinforce the continuity
on the interface via, for example, the Lagrange multiplier or the discontinuous Galerkin (DG) method.
The shear-slip method [25, 26, 20] was introduced to locally reconnect the mesh in order to maintain its
the quality when the structure is undergoing translation or rotation. There are also some interesting ALE
remeshing techniques that are able to deal with large structure displacement. For example, the fixed-mesh
ALE method [27] and the universal mesh method [28] locally rebuild the mesh in order to accommodate the
arbitrary motion of the interface.
In our work, we develop a new ALE method in order to produce a body-fitted moving fluid mesh that
conforms with the rotational and deformable structure mesh on the interface. And, to make our ALE
method work for an elastic rotor immersed in fluid, we first derive a linear structure equation that involves
the rotational matrix from the nonlinear structure model based on decomposing the structure displacement
into two components: rotation and deformation. In addition, we define an artificial cylindrical buffer zone
in the fluid domain to enclose the elastic rotor and rotate together on the same axis of rotation with the
same angular velocity. If the fluid channel is non-axisymmetric, we need to determine the relative motion
information between the rotational fluid subdomain (the cylindrical buffer zone) and the stationary fluid
subdomain (the rest of fluid domain) by matching the grid on the sliding interface and defining them in
our new ALE mapping. Finally, we develop a very stable and easily attainable ALE method designed to
generate a rotational and deformable fluid mesh that matches the structure mesh on the interface.
1
Next, with the velocity instead of the displacement as the principle unknown of the structure, we define
a monolithic saddle-point system for the studied FSI problem, and further a monolithic algorithm to solve
the coupled fluid and structure equations. Moreover, we are able to prove the well-posedness of the dis-
crete monolithic linear saddle-point system resulting from mixed finite element discretization and Newton’s
method. Numerical experiments are carried out for a self-defined elastic rotor and a realistic hydro-turbine
to illustrate that our developed structure model and ALE-based monolithic method are efficient and stable
for the FSI problem involving an elastic rotor. A numerical validation is also conducted to demonstrate the
consistency of our rotational structure model with respect to the different structure parameters.
References
[1] J. Sarrate, A. Huerta, J. Donea, Arbitrary Lagrangian-Eulerian formulation for fluid-rigid body inter-
action, Comput. Methods Appl. Mech. Engrg 190 (2001) 3171 3188.
[2] B. Desjardins, On weak solutions for fluidrigid structure interaction: compressible and incompressible
models, Commun. Part. Diff. Eq. 25 (7 8) (2000) 263 285.
[3] S. Capdevielle, Modeling fluid-rigid body interaction using the arbitrary Lagrangian-Eulerian method,
Masters Abstracts International 51.
[4] H. Hu, Direct simulation of flows of solid-liquid mixtures, Int. J. Multiphase Flow 22 (2) (1996) 335
352.
[5] A. Johnson, T. Tezduyar, 3D simulation of fluid-particle interactions with the number of particles
reaching 100, Comput. Methods Appl. Mech. Eng. 145 (3) (1997) 301 321.
[6] C. Hirth, A. Amsden, J. Cook, An arbitrary Lagrangian-Eulerian computing method for all flow speeds,
J. Comput. Phys. 14 (3) (1974) 227 253.
[7] C. Taylor, T. Hughes, C. Zarins, Finite element modeling of blood flow in arteries, Comput. Methods
Appl. Mech. Eng. 158 (1) (1998) 155 196.
[8] T. Belytschko, J. Kennedy, Computer models for subassembly simulation, Nucl. Eng. Design. 49 (1)
(1978) 17 38.
[9] T. Belytschko, J. Kennedy, D. Schoeberle, Quasi-Eulerian finite element formulation for fluid-structure
interaction, J. Press. Vess-T ASME 102 (1) (1980) 62 69.
[10] F. Nobile, Numerical approximation of fluid-structure interaction problems with application of haemo-
dynamics, Ph.D. thesis, Ecole Polytechnique Federale de Lausanne, Switzerland (2001).
[11] J. Xu, K. Yang, Well-posedness and robust preconditioners for discretized fluid-structure interaction
systems, Comput. Methods Appl. Mech. Engrg. 292 (2015) 69 91.
[12] Q. Du, M. Gunzburger, L. Hou, J. Lee, Analysis of a linear fluid-structure interaction problem, Disc.
Conti. Dyn. Sys.-A 9 (3) (2003) 633 650.
[13] L. Zhang, Y. Guo, W. Wang, FEM simulation of turbulent flow in a turbine blade passage with dynam-
ical fluid-structure interaction, Internat. J. Numer. Methods Fluids 61 (12) (2009) 1299 1330.
[14] L. Zhang, Y. Guo, H. Zhang, Fully coupled flow-induced vibration of structures under small deformation
with GMRES method, Appl. Math. Mech. 31 (2010) 87 96.
[15] L. Zhang, Y. Guo, W. Wang, Large eddy simulation of turbulent flow in a true 3D Francis hydroturbine
passage with dynamical fluid-structure interaction, Internat. J. Numer. Methods Fluids 54 (5) (2007)
517 541.
2
[16] T. Hughes, W. Liu, T. Zimmermann, Lagrangian-Eulerian finite element formulation for incompressible
viscous flows, Comput. Methods Appl. Mech. Eng. 29 (3) (1981) 329 349.
[17] A. Huerta, W. Liu, Viscous flow structure interaction, J. Pressure Vessel Technol. 110 (1) (1988) 15
21.
[18] C. Nitikitpaiboon, K. Bathe, An arbitrary Lagrangian-Eulerian velocity potential formulation for fluid-
structure interaction, Comput. Struct. 47 (4) (1993) 871 891.
[19] M. Souli, D. Benson (Eds.), Arbitrary Lagrangian Eulerian and Fluid-Structure Interaction: Numerical
Simulation, John Wiley & Sons, 2010.
[20] Y. Bazilevs, M. Hsu, K. Takizawa, T. Tezduyar, ALE-VMS and ST-VMS methods for computer mod-
eling of wind-turbine rotor aerodynamics and fluid-structure interaction, Math. Models Methods Appl.
Sci. 22 (2) (2012) 1230002.
[21] Y. Bazilevs, M. Hsu, M. Scott, Isogeometric fluid-structure interaction analysis with emphasis on non-
matching discretizations, and with application to wind turbines, Comput. Methods Appl. Mech. Eng.
249 252 (2012) 28 41.
[22] M. Hsu, Y. Bazilevs, Fluid-structure interaction modeling of wind turbines: simulating the full machine,
Comput. Mech. 50 (6) (2012) 821 833.
[23] Y. Bazilevs, M. Hsu, J. Kiendl, R. W¨uchner, K. Bletzinger, 3D simulation of wind turbine rotors at
full scale. Part II : Fluid-structure interaction modeling with composite blades, Internat. J. Numer.
Methods. Fluids. 65 (1 3) (2011) 236 253.
[24] Y. Bazilevs, M. Hsu, 3D simulation of wind turbine rotors at full scale. Part I: geometry modeling and
aerodynamics, Internat. J. Numer. Methods. Fluids. 65 (1 3) (2011) 207 235.
[25] M. Behr, T. Tezduyar, The shear-slip mesh update method, Comput. Methods Appl. Mech. Engrg.
174 (3) (1999) 261 274.
[26] M. Behr, T. Tezduyar, Shear-slip mesh update in 3D computation of complex flow problems with
rotating mechanical components, Comput. Methods Appl. Mech. Engrg. 190 (24) (2001) 3189 3200.
[27] R. Codina, G. Houzeaux, H. Coppola-Owen, J. Baiges, The fixed-mesh ALE approach for the numerical
approximation of flows in moving domains, Journal of Computational Physics 228 (5) (2009) 1591–1611.
[28] E. S. Gawlik, A. J. Lew, High-order finite element methods for moving boundary problems with pre-
scribed boundary evolution, Computer Methods in Applied Mechanics and Engineering 278 (2014)
314–346.
3