Proposal For Numerical Benchmarking of Fluid-Struc
Proposal For Numerical Benchmarking of Fluid-Struc
Proposal For Numerical Benchmarking of Fluid-Struc
net/publication/226447172
CITATIONS READS
542 5,439
2 authors:
Some of the authors of this publication are also working on these related projects:
Stabilised finite element methods for problems in fluid mechanics View project
All content following this page was uploaded by Jaroslav Hron on 05 June 2014.
Abstract. We describe new benchmark settings for the rigorous evaluation of dif-
ferent methods for fluid-structure interaction problems. The configurations consist
of laminar incompressible channel flow around an elastic object which results in
self-induced oscillations of the structure. Moreover, characteristic flow quantities
and corresponding plots are provided for a quantitative comparison.
1 Introduction
2 Definitions
∂ vf
f + f (∇v f )v f = div σ f
∂t in Ωtf . (1)
div v f = 0
∂ vs
s + s (∇v s )v s = div(σ s ) + s g in Ωts . (3)
∂t
Written in the more common Lagrangian description, i.e. with respect to
some fixed reference (initial) state Ω s , we have
∂ 2 us
s = div(Jσ s F −T ) + s g in Ω s (4)
∂t2
where F = I + ∇us is the deformation gradient tensor. For further details
see for example [1].
The material is specified by giving the Cauchy stress tensor σ s (the 2nd
Piola-Kirchhoff stress tensor is then given by S s = JF −1 σ s F −T ) by the
248 S. Turek, J. Hron
1
σs = F (λs (tr E)I + 2µs E) F T (5)
J
S s = λs (tr E)I + 2µs E (6)
σf n = σs n
on Γt0 , (9)
vf = vs
where n is a unit normal vector to the interface Γt0 . This implies the no-slip
condition for the flow, and that the forces on the interface are in balance.
H h
l
(0, 0)
L
Fig. 1. Computational domain
C A
B h
r
l
such that the mean inflow velocity is Ū and the maximum of the inflow
velocity profile is 1.5Ū.
– The outflow condition can be chosen by the user, for example stress free
or do nothing conditions. The outflow condition effectively prescribes
some reference value for the pressure variable p. While this value could
be arbitrarily set in the incompressible case, in the case of compressible
structure this will have influence on the stress and consequently the de-
formation of the solid. In this proposal, we set the reference pressure at
the outflow to have zero mean value.
– The no-slip condition is prescribed for the fluid on the other boundary
parts. i.e. top and bottom wall, circle and fluid-structure interface Γt0 .
kg kg 6 kg
material s [ m 3] ν s E [106 ms 2 ] µ [10 ms2 ]
s
polybutadiene polypropylene
parameter
& glycerine & glycerine
kg
s [103 m 3] 0.91 1.1
νs 0.5 0.42
kg
µs [106 ms 2] 0.53 317
3 kg
[10 m3 ]
f
1.26 1.26
2
ν f [10−3 ms ] 1.13 1.13
Table 3. Proposed material combination
1. The y-coordinate y(t) of the end of the beam structure at point A(t) (see
the Figure 2).
2. Forces exerted by the fluid on the whole submerged body, i.e. lift and
drag forces acting on the cylinder and the beam structure together
(FD , FL ) = σndS,
S
= σndS.
S0
S1
S2
S0
The time dependent values are represented by the mean value, amplitude
and frequency. The mean value and amplitude are computed from the last
period of the oscillations by taking the maximum and minimum values, then
the mean value is taken as average of the min/max values, and the amplitude
is the difference of the max/min from the mean:
1
mean = (max + min)
2
1
amplitude = (max − min)
2
The frequency of the oscillations can be computed either from the period
time T as
1
frequency =
T
or by using fourier analysis on the periodic data and taking the lowest signifi-
cant frequency present in the spectrum. Additionally, a plot of the quantities
over the period should be presented.
Proposal for numerical benchmarking of FSI 253
For the validation of the employed fluid and solid solvers, we first describe
partial tests which are performed on different levels of mesh refinement (see
Fig. and Table 4) with various time steps. We provide the results for the
different discretization levels in the following since these sequences of results
indicate that our given ”reference results” are almost grid-independent. All
simulations have been performed with a fully implicit monolithic ALE-FEM
method with a fully coupled multigrid solver as described in [2].
Fig. 4. Example of a coarse mesh and the number of degrees of freedom for refined
levels
completely rigid by considering the flow domain only with fixed boundary
conditions on the flag interface.
0
-100 438
-200 436
-300
-400 434
-500 432
9 9.1 9.2 9.3 9.4 9.5 9.6 9 9.1 9.2 9.3 9.4 9.5 9.6
time time
Proposal for numerical benchmarking of FSI 255
displacement y
-0.005 -0.02
-0.04
-0.01
-0.06
-0.015
-0.08
-0.02 -0.1
-0.025 -0.12
-0.03 -0.14
0 2 4 6 8 10 0 2 4 6 8 10
time time
0.005 0.02
csm3 csm3
0 0
displacement x
displacement y
-0.005 -0.02
-0.04
-0.01
-0.06
-0.015
-0.08
-0.02 -0.1
-0.025 -0.12
-0.03 -0.14
8 8.5 9 9.5 10 8 8.5 9 9.5 10
time time
Proposal for numerical benchmarking of FSI 257
displacement y
-0.01 0.04
0.02
-0.015
0
-0.02 -0.02
-0.04
-0.025
-0.06
-0.03 -0.08
34 34.2 34.4 34.6 34.8 35 34 34.2 34.4 34.6 34.8 35
time time
FSI2: lift and drag force on the cylinder+flag
250 300
200 fsi2 fsi2
280
150 260
100 240
50
220
drag
lift
0
200
-50
-100 180
-150 160
-200 140
-250 120
34 34.2 34.4 34.6 34.8 35 34 34.2 34.4 34.6 34.8 35
time time
displacement y
0.02
-0.002 0.01
-0.003 0
-0.004 -0.01
-0.02
-0.005 -0.03
-0.006 -0.04
19.5 19.6 19.7 19.8 19.9 20 19.5 19.6 19.7 19.8 19.9 20
time time
FSI3: lift and drag force on the cylinder+flag
200 485
fsi3 480 fsi3
150 475
100 470
465
50
drag
460
lift
0 455
450
-50 445
-100 440
435
-150 430
19.5 19.6 19.7 19.8 19.9 20 19.5 19.6 19.7 19.8 19.9 20
time time
5 Summary
The next step will be the specification of how to submit and to collect the
results, and the publication of the test configurations in an international
journal. Moreover, it is planned to prepare a webpage for collecting and pre-
senting the FSI results. As we have learned from [3], a very important aspect
will be the submission of the results on (at least) 3 different meshes and time
steps. Then, based on the collected results, quantitative ratings regarding the
main questions, particularly w.r.t. the coupling mechanisms and monolithic
vs. partitioned approaches, might get possible.
260 S. Turek, J. Hron
References
1. P. G. Ciarlet. Mathematical Elasticity. Volume I, Three-Dimensional Elasticity,
volume 20 of Studies in Mathematics and its Applications. Elsevier Science
Publishers B.V., Amsterdam, 1988.
2. J. Hron and S. Turek. A monolithic FEM/multigrid solver for ALE formulation
of fluid structure interaction with application in biomechanics. In H.-J. Bungartz
and M. Schäfer, editors, Fluid-Structure Interaction: Modelling, Simulation, Op-
timisation, LNCSE. Springer, 2006.
3. S. Turek and M. Schäfer. Benchmark computations of laminar flow around cylin-
der. In E.H. Hirschel, editor, Flow Simulation with High-Performance Comput-
ers II, volume 52 of Notes on Numerical Fluid Mechanics. Vieweg, 1996. co. F.
Durst, E. Krause, R. Rannacher.
4. W. A. Wall and E. Ramm. Fluid-structure interaction based upon a stabilized
(ALE) finite element method. In S. Idelsohn, E. Oñate, and E. Dvorkin, editors,
4th World Congress on Computational Mechanics ? New Trends and Applica-
tions, Barcelona, 1998. CIMNE.