Hron-Turek FSI3

FSI3 benchmark (Hron & Turek 2006): laminar channel flow around a rigid cylinder with an attached elastic beam — unsteady periodic solution.

Repository path: tutorials/fluidStructureInteraction/HronTurekFsi3

Reference: Alkafri et al. (2024), multiRegionFoam: A Unified Multiphysics Framework, Sect. 7.4

Overview

The moving-mesh ALE interface tracking method for multiphase flow is extended here to fluid–structure interaction (FSI). The case follows the benchmark defined by Hron and Turek (2006): a laminar incompressible flow in a 2D channel past a rigid cylinder with an attached elastic beam.

The configuration consists of:

  • Channel: $L = 2.5\,\text{m}$, $H = 0.41\,\text{m}$
  • Rigid cylinder: centre $C(0.2,\, 0.2)$, radius $r = 0.05\,\text{m}$
  • Elastic beam: tip point $A(0.6,\, 0.2)$, length $l = 0.35\,\text{m}$, thickness $h = 0.02\,\text{m}$

A parabolic velocity profile is prescribed at the inlet with maximum velocity $1.5\,\bar{U}$. Three test cases are defined in the benchmark suite—FSI1 (steady), FSI2 (unsteady periodic), and FSI3 (unsteady periodic)—by varying the mean inflow velocity. FSI3 with $\bar{U} = 2\,\text{m/s}$ is considered here.

Computational domain

Fig. 18 — Computational domain for the FSI3 case with zoom on the elastic structure. Channel H = 0.41 m, L = 2.5 m; cylinder centre C(0.2, 0.2), r = 0.05 m; beam tip A(0.6, 0.2), l = 0.35 m, h = 0.02 m.

Fig. 18 — Computational domain for the FSI3 case with zoom on the elastic structure. Channel H = 0.41 m, L = 2.5 m; cylinder centre C(0.2, 0.2), r = 0.05 m; beam tip A(0.6, 0.2), l = 0.35 m, h = 0.02 m.

Material properties

Table: Fluid and solid properties for FSI3

PropertySymbolUnitSolidFluid
Kinematic viscosityνm²/s10⁻³
Mean inflow velocityŪm/s2
Densityρkg/m³10³10³
Young’s modulusEkg/ms²5.6 × 10⁶
Poisson ratioν0.4

Boundary conditions

Table: Boundary conditions for FSI3

BoundaryVelocityKinematic PressureTraction
Fluid
interfaceShadowmovingWallVelocityzeroGradient
inletparabolicVelocityzeroGradient
outletzeroGradient0 m²/s²
cylinder, bottom, top(0 0 0)ᵀ m/szeroGradient
front & back planesemptyempty
Solid
InterfaceregionCoupledTraction

Mesh refinement study

Three mesh levels are used. The fluid mesh is generated with blockMesh (coarse) and refined with refineMesh + extrudeMesh:

LevelSolid cellsFluid cellsΔt (s)
Coarse6305 3361.0 × 10⁻³
Medium21 3447.5 × 10⁻⁴
Fine85 3765.0 × 10⁻⁴

Simulations are run until a periodic solution is reached. Mean and amplitude values are computed from the last oscillation period; frequency is extracted via fast Fourier transform.

Results

Fig. 19 — x and y displacement of beam tip point A over time obtained from the finest mesh, showing periodic oscillation.

Fig. 19 — x and y displacement of beam tip point A over time obtained from the finest mesh, showing periodic oscillation.

Fig. 20 — Velocity and pressure fields in the fluid alongside displacement and stress fields in the solid structure at the instant of maximum deflection.

Fig. 20 — Velocity and pressure fields in the fluid alongside displacement and stress fields in the solid structure at the instant of maximum deflection.

Validation results

Table: Displacement of point A — mean ± amplitude [frequency in Hz] (values in mm and Hz, matching benchmark format mean ± amplitude [frequency])

Caseu_y (×10⁻³ m)u_x (×10⁻³ m)
Coarse1.67 ± 28.55 [5.9]−1.99 ± 1.84 [11.5]
Medium1.53 ± 33.41 [5.3]−2.66 ± 2.45 [10.9]
Fine1.47 ± 34.37 [5.3]−2.67 ± 2.53 [10.9]
Benchmark1.48 ± 34.35 [5.3]−2.69 ± 2.53 [10.9]

As the fluid mesh resolution increases the computed displacements converge towards the benchmark values.

Reference

J. Hron, S. Turek (2006). A monolithic FEM/multigrid solver for an ALE formulation of fluid-structure interaction with applications in biomechanics. In: Fluid-Structure Interaction, Lecture Notes in Computational Science and Engineering, vol 53. Springer, Berlin, Heidelberg.

Running the case

cd $FOAM_RUN/../multiPhysicsFoam/tutorials/fluidStructureInteraction/HronTurekFsi3
./Allrun