One-way fluid structure interaction of pipe under flow with different boundary conditions.
Dahmane, M. ; Boutchicha, D. ; Adjlout, L. 等
One-way fluid structure interaction of pipe under flow with different boundary conditions.
1. Introduction
Hydraulic induced vibration starts with a continuous flow
disturbance that creates a periodic pressure pulse. At changes in
direction (elbow, tee, bend) or changes in flow cross section (valve,
orifice, reducer) this pressure pulse causes pulsating forces on the
pipe and makes it vibrate [1]. The complexity of the vibro-acoustic
behaviour of flexible, fluid-filled pipe systems is predominantly
determined by two parameters: the frequency and the ratio of the masses
per unit pipe length of fluid and pipe wall. The number of
simultaneously propagating waves in the pipes increases with frequency.
The mass ratio determines whether fluid borne and structure-borne waves
may be treated separately. If this ratio is close to one, fluid
pulsations and mechanical vibrations will be strongly coupled, so that
it becomes necessary to consider their interaction when analyzing the
dynamic behavior [2]. Therefore, it is vitally important that the
evaluation of natural frequency of a pipe be accurately to prevent any
resonance condition. There have been extensive studies on the modeling
and analysis of fluid conveying pipes over the past half-century, as
reported by Paidoussis and al [3, 4], the pipe conveying fluid has
established itself as a generic paradigm of a kaleidoscope of
interesting dynamical behaviour. The effect of internal flow on
transverse vibration of a pipe was studied by [3]. Coriolis acceleration
of the internal fluid was taken into account in [5, 6].
In practice, the geometry of fluid conveying pipe systems is
generally complicated. For these problems, analytical methods are not
sufficient for the vibration analysis of real pipeline systems.
Therefore, it is desirable to utilize numerical or approximately
analytical methods such as finite element method (FEM) [7], (FEM)-state
space approach [8], spectral element method (SEM) [9], transfer matrix
method (TMM) [10], differential transformation method (DTM) [11],
generalized integral transform technique (GITT) [12] and differential
quadrature element method-(DQEM) proposed for obtaining highly accurate
natural frequencies of multiple-stepped beams with an aligned neutral
axis[13] can be mentioned. The problem of nonlinearities can be found in
[14] by considering the dynamical behaviour of a fluid-conveying curved
pipe subjected to motion-limiting constraints and a harmonic excitation.
Based on a Newtonian method, the in-plane equation of motion of this
curved pipe is derived. Then a set of discrete equations in spatial
space obtained by the differential quadrature method (DQM) is solved
numerically. In his study, H.R. OZ [15] considered the non-linear
transverse vibrations of highly tensioned pipes conveying fluid as being
investigated.
Practically, long, cross-country pipelines rest on an elastic
medium such as a soil, and hence, a model of the soil medium must be
included in the analysis. Chellapilla and al [16], studied the effect of
a Pasternak foundation on the critical velocity of a fluid-conveying
pipe. Chellapilla and al [17] studied problem of vibrations of
fluid-conveying pipes resting on a two-parameter foundation model such
as the Pasternak-Winkler model. This work has been extended to the study
of the effect of the Pasternak foundation on the natural frequencies of
the pipeline for the pinned-pinned, clamped-clamped and clamped-pinned
boundary conditions. It is well established from published literature
where there exists a critical velocity of the fluid near which the
natural frequency of the pipeline tends to zero. Dynamic response of
linear systems subjected to dynamic loading, such as shock or seism,
with finite element procedures is of paramount importance in many
engineering applications [18]. Advances in finite element methodology
have made it possible to simulate the dynamic motion of the fluid
coupled with the flexible pipe for arbitrary geometries within the
context of a general purpose finite element program [19].
Numerical enhancement of the finite element code ANSYS and in order
to produce coupled fluid-structure dynamic analysis with pressure-based
formulation has been exposed and validated [20], using modal and
spectral methods. Enhancement of the modeling possibilities within the
ANSYS code is carried out with implementation of fluid-structure
symmetric formulations for elasto-acoustic and hydro-elastic problems.
Using symmetric formulation enables linear dynamic analysis with modal
projection techniques for a fluid-structure coupled system. In the
mentioned study, comparison of numerical results has been performed for
modal analysis for (3D) using modal methods (temporal and spectral
approaches). Taking fluid-structure interaction into account in such
problems is made possible by finite element analysis (FEA), using ANSYS
code [19]. This latter paper presents a state-variable model developed
for the analysis of fluid-induced vibration of composite pipeline
systems. Simply supported, clamped and clamped-simply supported
pipelines are investigated. The influence of fluids Poisson ratio, the
ratio of pipe radius to pipe-wall thickness, the ratio of liquid mass
density to pipe-wall mass density, the fluid velocity, initial tension
and fluid pressure are all considered. The effected numerical
simulations to rest most of the time on a modeling simplified of one of
the two mediums, or appeal specified procedures of coupling which permit
to function together the computer codes specially, is developed for
fluid from on a side and the structure of the other [21].
In this work, the fluid is considered incompressible and the solid
is an elastic body. Therefore, this study will be devided into two
parts. The first part is specified for analytical modeling and formulas
of natural frequencies and shapes modes, by establishing the equations
of motions of the pipe with different boundary conditions. One may
consider a two-mode Galerkin approximation has been employed to get
results for the pinned-pinned beam. Only the effect of fluid flow was
taken into account, and the theoretical eigen-frequencies are obtained,
under different fluid velocities. For the numerical approach, the
interaction enters the fluid and the solid involves the velocity
continuity and the mechanical equilibrium condition to the interface.
This interaction is based on a procedure of instantaneous coupling,
where they obtained pressure in iterations of fluid numerical model is
used as extern solicitation for the solid numerical model.
Unidirectional approach has been used for the three cases;
pinned-pinned, clamped-pinned and clamped-clamped.
2. The analytical model
2.1. Equations of motion and boundary conditions
The pipe is long and straight, thus facilitating the use of
Euler-Bernoulli beam theory; the motions are small so that the system
can be analyzed by the linear theory; and, the effects of internal
pressure and external forces are neglected in the analysis [3]. By
applying Newton's second law of motion for afluid-conveying pipe of
length L with lateral displacement w, four coupled, linear pipe-dynamic
equations have been derived in the previous work [4] as follow:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)
a) Pinned-Pinned Pipe:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)
b) Clamped-Pinned Pipe:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)
c) Clamped-Clamped Pipe:
And those for a clamped-clamped pipe are:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)
The dimensionless form of Eq. (1) can be written as:
[eta]"" +
[u.sup.2][eta]"+2[beta]1/2u[eta]'+[eta] = 0, (5)
where
([eta]) = [partial derivative]([eta]) /[partial derivative] [tau]
and ([eta])' = [partial derivative]([eta]) / [partial derivative]
[xi]; (6)
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (7)
According to Galerkin's method, the solution of Eq. (1) may be
expressed as:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (8)
where, Re denotes the real part, [omega] denotes the mode of
vibration and where [[empty set].sub.r]([xi]) are the dimensionless
eigen-functions of a beam with the same boundary conditions as the pipe
under consideration, [[empty set].sub.r] ([xi]) given by [22]:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], (9)
where the constants [C.sub.r],r = 1 to 4 are determined into the
boundary conditions; the frequency functions [MATHEMATICAL EXPRESSION
NOT REPRODUCIBLE IN ASCII].
In the above equation, [[alpha].sub.r] is the frequency parameter
of the pipe without fluid flow, which is considered as a beam, and
it's values are:
- [[alpha].sub.r]L = n[pi] ; r =1, 2, 3...N for the pinned-pinned
case;
- [[alpha].sub.x]L =3.9266; [[alpha].sub.2]L =7.0685;
[[alpha].sub.3]L =10.2101 for the pinned-clamped case;
- [[alpha].sub.1]L = 4.7300; [[alpha].sub.2]L =7.8532;
[[alpha].sub.3]L =10.9956 for the clamped -clamped case.
The equation (8) is substituted into the left-hand side of Eq. (5),
the result will generally not be zero, but equal to an error function,
which may be denoted by [[member of].sub.N] [[w.sub.N] (x, t)].
Galerkin's method requires that:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (10)
Substituting Eq. (9), we obtain:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (11)
Using the non-dimensional parameters, we obtain:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (12)
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (13)
The dimensionless frequency [OMEGA] = Re([OMEGA]) + iTm([OMEGA]) is
plotted parametrically with the dimensionless flow velocity u [23],
where Re ([OMEGA]) the real part and Tm ([OMEGA]) the imaginary part.
Re([OMEGA])>Tm([OMEGA]). (14)
Multiplying Eq. (8) by [[empty set].sub.S] and using the orthogonal
property of the eigen-functions and integration over the domain [0, 1]
yields:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (15)
where the different matrices are defined as follows:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (16)
The Eq. (15) may be written in matrix form as follows:
First, one may consider a two-mode Galerkin approximation of the
system, namely:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (17)
For pinned ends [b.sub.rr] = 0, [b.sub.sr] = -[b.sub.rs] and
[C.sub.rs] = 0. Eq. (17) may be written as:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (18)
Eq. (16) may be written in matrix form as follows:
[K-[[OMEGA].sup.2.sub.j]MI]{a} = 0, (19)
where K is the stiffness matrix, I is the identity matrix and
[a.sup.T] = {[a.sub.1],[a.sub.2],...[a.sub.n]}.
Setting the determinant ofthe coefficient matrix above equal to
zero and retaining thefirst two terms of the above equation, we get the
frequency equation. For the pinned -pinned case it is:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (20)
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (21)
The critical velocity parameter for the mode 1:
[u.sub.cr] = N [pi].
2.2 Interpretation and analytical results
The frequencies parameters for the first two natural modes are
calculated bending movement of a pipe in different boundary conditions.
The results are presented for the following cases: 1) Fluid conveying
pipe without foundation; 2) No-flow, pipe with parameter elastic
foundation and 3) Fluid conveying pipe with Winkler foundation.
Comparison has been made with available literature wherever possible.
For the pinned-pinned boundary condition, numerical results have been
obtainedconsidering the first two terms of the equation resulting from
using Galerkin approximation.
Case 1 : Fluid conveying pipe without foundation
Results have been obtained for the pipe with fluid (no elastic
foundation). They are compared with available literature. Table 2 show
the two first frequencies parameter for pinned-pinned fluid-conveying
pipe. The two methods employed for comparison are Fourier series and
Galerkin.
The error on the frequencies of the order 0.075 is very
satisfactory. The Fig. 1 watches the effect Coriolis on the vibration of
a pinned-pinned pipe with flow.
Case 2: Pipe with elastic foundation (Winkler model) for no flow
For the pinned-pinned fluid-conveying pipes ([beta] [not equal to]
0), however, the effect of an elastic foundation is stabilizing, as
shown by [4]. Galerkin expansion was employed to calculate numerical
solutions, for the Pasternak-Winkler model [16].
The terms [k.sub.1w] and [k.sub.2][[partial
derivative].sup.2]w/[[partial derivative]x.sup.2] are added to the
equation of motion which becomes as the following:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (22)
Non-dimensional frequency equation for the pinned-pinned pipes with
elastic foundation is:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (23)
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
Results for no-flow have been presented for parameter elastic
foundation (the presence of the Winkler foundation) and compared with
available literature. Table 3 shows the values of the first frequency
parameter for a pinned-pinned pipe for different values of
[[gamma].sub.1]. There is very strongly an increase for [[gamma].sub.1]
in the surrounding by [10.sup.4].
Case 3 : Fluid-conveying pipe with elastic foundation (Winkler
model)
Results for a pipe with fluid flow have been presented for various
values of [[gamma].sub.1] (Winkler foundation), u and [beta] is 0.1.
According to the Table 4 case of a pinned-pinned pipe, one notice that
the first eigen-frequency increases appreciably with the values of the
[[gamma].sub.1], very strongly increases for [[gamma].sub.1] in the
surrounding by [10.sup.3]. According to Table 4, the Winkler foundation
has a stabilizing effect in the pipe and increasing values of
[[gamma].sub.1] tend to increase both the critical flow velocity and the
first frequency parameter pinned-pinned fluid-conveying pipe.
[% Variation = |[[OMEGA].sub.1]present-[[OMEGA].sub.1](Ref.[17])]|/[[[OMEGA].sub.1](Ref.[17])x100]
3. The numerical model
3.1. Fluid-structure interaction problems in the ANSYS code
Design of structures in contact with fluid needs specific
attentions on the static and dynamic analysis and requires more accurate
simulation tools (finite element method) to improve performance of the
system while minimizing the cost of manufacturing. The commercial
package ANSYS was used with Fluent for computational fluid dynamics and
ANSYS mechanical for the structural simulation. The traditional approach
to determine the mechanical stress and deflection of a structure in
contact with fluid consists of a steady state computational fluid
dynamics (CFD) simulation which provides the fluid pressure on the
structure. This can then be applied as a boundary or load condition for
the FEM simulation of the configuration, like in Schmucker and al (Ref.
[24]).
Fluid Model
- The simulation has been realized under the following conditions:
- The fluid is Newtonian and incompressible, Flow 3D is turbulent.
- Mesh with hexahedral elements, (hexahedral) 8 nodes.
Structure Model
The object of this section is the determination of proper
frequencies of the right pipe, using a tridimensional element and the
study of proper from for each case. The conditions:
- The structure geometry supposed uniform along the length;
- Elastic modulus of material component the structure is chose
2.0*[10.sup.11] [N/[m.sup.2]], while the solid density, the
Poisson's ratio are accepted respectively 7850 [Kg/[m.sup.3]],
0.26;
- The used element for the simulation (mesh), from the ANSYS
Mechanical User Guide is the hexahedral (three-dimensional) known under
<<SOLID 45>>, using eight nodes isotropic.
3.2. Fluid-Structure coupling conditions
ANSYS Workbench permits of making a simulation FSI (Fluid Structure
Interaction) in a very easier manner, the schema the charged transfer,
which is in our case the pressure.
The coupling conditions consist of transfer field-fluid operation
on the structure and concern the pressure transfer on solid. The types
of transfer are the following:
[1..] The fluid cod transmit the pressure, interface nodal, on
structure code
2. The structure code calculates the solution and put on again
nodes coordinates.
3. The structure transmits new position of interface nodes. In
(Workbench) the simulation is done with Mechanical code, see Fig. 3.
3.3. Interpretation and numerical results
The results are presented for the following cases: 1)
Fluid-conveying pipe without foundation; 2) No-flow, pipe with parameter
elastic foundation and 3) Fluid conveying pipe with Winkler foundation.
Numerical results have been obtained considering the first two
frequencies. To validate the convergence between the analytical method
and the numerical one, a typical uniform model is used according to the
length of profiles. For this, we considered the profile of the pipe as a
model: pinned-pinned beam, pinned-clamped and clamped-clamped beam of a
uniform circular section. The commercial finite element program ANSYS is
employed for modeling the steel pipeline using eight node isotropic 3D
solid elements (SOLID 45). Fig. 2 shows the profile tube mesh with
hexahedral elements. Fig. 3 shows fluid-structure coupling is only
modeled between the elastic structure and the fluid (the hydrodynamic
load is imported) of clamped-clamped pipe for mode 1. In the ANSYS code,
this is achieved with 45390 elements, (hexahedral) fluid finite elements
(3D element with eight nodes and three degrees of freedom). Fig. 4 shows
the plot of the displacements nodal of clamped-clamped pipe for mode 1.
Case 1 : Fluid conveying pipe without foundation
The first two vibration frequencies as functions of fluid
velocities in the simply-supported, clamped and clamped-simply supported
pipes without foundation were determined and are shown in Table 5 and
Figs. 5-6, respectively. In these table and figures, values of the
velocity [V.sub.0] are varying from zero to the value [V.sub.0] = 103.17
m/s which corresponds of the velocity parameter u = 2 for the mass
ratios parameter [beta] is 0.6. For boundary conditions pinned-pinned,
we notice a good agreement with the results presented analytically for
this case, the variation on the first frequency of the order 0.80 is
very satisfactory.
Case 2: Pipe with elastic foundation (Winkler model) for no flow
The first two vibration frequencies as a function of elastic
foundation (Winkler model) for no flow where [V.sub.0] = 0 m/s (where
the velocity parameter u = 0) and [beta] = 0.5 in the simply-supported
and clamped-simply supported pipes were determined and are shown in
Tables 6 and 7. A similar trend is noticed in these cases, the error on
the first frequency of the order 1.15 is very satisfactor.
Case 3 : Fluid-conveying pipe with elastic foundation (Winkler
model)
The first two vibration frequencies as a function of fluid
velocities in the clamped-clamped pipe with elastic foundation were
determined and are shown in tables. 8 and 10, respectively. For this
case, values of the velocity [V.sub.0] are varying from zero to the
value 103.17 m/s (u = 2) for [beta] =0.6.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
4. Conclusions
The present work is a contribution to the study of pipes vibration,
hydrodynamic structures, through analytical and numerical approaches. A
tool is used to intend the calculation of natural frequencies and
visualize the flow in the pipe system filled on fluid. At first time,
the hydrodynamic problem is treated by analyzing the first two natural
modes of particular cases. Several examples were processed to determine
the influence of the fluid velocity and different geometrical and
physical parameters on the phenomenon of fluid-structure interaction.
The main findings can be summarized as follows:
[1..] The first conclusion which one can draw from this study is
that the frequencies of the system fluid-structure depend on the
geometrical and physical properties.
2. The natural frequencies of the pipe conveying fluid depend on
the velocity of the fluid. At a certain critical velocity the first
natural frequency decreases until vanishes, the natural frequencies of
this case can be obtained analytically.
3. The increase in the [beta] (mass ratios) increased the value of
the added mass to the system and therefore decreases the frequency
parameter.
4. The attempt was made to validate the current formulation of the
problem.
5. The natural frequencies increase significantly with the values
of [[gamma].sub.1], and increases sharply for [[gamma].sub.1] at around
[10.sup.3].
6. The results with Winkler foundation parameter of a pinned-pinned
pipe under flow are obtained analytically.
7. The dimensionless critical velocity of [[gamma].sub.1] values of
pinned-pinned pipe is presented.
8. To validate the numerical results, we carried out the analytical
calculation of turbulent flow with [beta] = 0.6.
9. The results obtained numerically are similar to those obtained
by the analytical approach for the determination of the natural
frequencies. A global error on the first frequency of the order 1.15% is
very satisfactory.
[10.]. After the numerical validation, we can study the one-way
fluid solid interaction of pipe under flow with different different
physical and geometrical parameters.
[11.]. Numerical simulation with ANSYS is therefore placed as a
promising tool for the design of pipes, on the condition of being of a
cost of accessible calculation. To obtain a very good precision, using
mesh with hexahedral elements (SOLID 45). This code also makes it
possible to determine all the modes of vibration of the pipe.
References
[1.] George, A. Antaki 2006. Piping and Pipeline Engineering:
Design, Construction, Maintenance. Integrity, and Repair. Aiken, South
Carolina, U.S.A., 564 p.
[2.] De Jong, C.A.F. 1994. Analysis of pulsations and vibrations in
fluid-filled pipe systems, PhD Thesis, Eindhoven University of
Technology.
[3.] Paidoussis, M.P; Li, G.X. 1993. Pipes conveying fluid: a model
dynamical problema, Journal of Fluids and Structures 8: 137-204.
http://dx.doi.org/10.1006/jfls.1993.1011.
[4.] Paidoussis, M.P. 1998. Fluid--Structure Interactions: Slender
Structures and Axial Flow, vol. 1, Academic Press, San Diego.
http://dx.doi.org/10.1016/S1874-5652(98)80001.
[5.] Anklin, M.; Drahm, W.; Gieder, A. 2006. Coriolis mass flow
meters: overview of the current state of the art and latest research,
Flow Measurement and Instrumentation 17: 317-323.
http://dx.doi.org/10.1016/j.flowmeasinst.2006.07.004.
[6.] Jon Juel Thomsen; Jonas Dahl 2010. Analytical predictions for
vibration phase shifts along fluid-conveying pipes due to Coriolis
forces and imperfections, Journal of Sound and Vibration 329: 3065-3081.
http://dx.doi.org/10.1016/j.jsv.2010.02.010.
[7.] Olson, L.G; Jamison, D. 1997. Application of a general purpose
finite element method to elastic pipes conveying fluid, Journal of
Fluids and Structures 11: 207-222. http://dx.doi.org/10.1006/jfls.
1996.0073.
[8.] Morteza, H. Sadeghi; Mohammad, H. Karimi-Dona 2011. Dynamic
behavior of a fluid conveying pipe subjected to a moving sprung mass-An
FEM-state space approach International Journal of Pressure Vessels and
Piping 88: 123-131. http://dx.doi.org/10.1016/j.ijpvp.2011.02.004.
[9.] Lee, U.; Park, J. 2006. Spectral element modelling and
analysis of a pipeline, Journal of Fluids and Structures 22: 273-292.
http://dx.doi.org/10.1016/j.jfluidstructs.2005.09.003.
[10.] Dai, H.L.; Wang, L.; Qian, Q.; Gan, J. 2012. Vibration
analysis of three-dimensional pipes conveying fluid with consideration
of steady combined force by transfer matrix method, Applied Mathematics
and Computation 219: 2453-2464.
http://dx.doi.org/10.1016/j.amc.2012.08.081.
[11.] Ni, Q.; Zhang, Z.L.; Wang, L. 2011. Application of the
differential transformation method to vibration analysis of pipes
conveying fluid, Applied Mathematics and Computation 217: 7028-7038.
http://dx.doi.org/10.1016/j.amc.2011.01.116.
[12.] Jijun Gu, Chen An, Menglan Duan, Carlos Levi, Jian Su. 2013.
Integral transform solutions of dynamic response of a clamped-clamped
pipe conveying fluid, Nuclear Engineering and Design 254: 237-245.
http://dx.doi.org/10.1016/j.nucengdes.2012.09.018.
[13.] Xinwei Wang, Yongliang Wang 2013. Free vibration analysis of
multiple-stepped beams by the differential quadrature element method,
Applied Mathematics and Computation 219: 5802-5810.
http://dx.doi.org/10.1016/j.amc.2012.12.037.
[14.] Wang, Lin.; Qiao, Ni. 2008. Nonlinear dynamics of a
fluid-conveying curved pipe subjected to motion-limiting constraints and
a harmonic excitation, Journal of Fluids and Structures 24: 96-110.
http://dx.doi.org/10.1016/j.jfluidstructs.2007.07.002.
[15.] Oz, H.R. 2001. Non-linear vibrations and stability analysis
of tensioned pipes conveying fluid with variable velocity, International
Journal of Non-linear Mechanics 36: 1031-1039.
http://dx.doi.org/10.1016/S0020-7462(00)00065-2.
[16.] Chellapilla, K.R.; Simha, H.S. 2007. Critical velocity of
fluid-conveying pipes resting on two-parameter foundation, Journal of
Sound and Vibration 302: 387-397.
http://dx.doi.org/10.1016/j.jsv.2006.11.007.
[17.] Chellapilla, K.R.; Simha, H.S. 2008. Vibrations of
fluid-conveying pipes resting on two-parameter foundation, The Open
Acoustics Journal 1: 24-33. http://dx.doi.org/10.1016/j.jsv.2006.11.007.
[18.] Makerle, J. 2000. Finite element vibration and dynamic
response analysis of engineering structures, Shock Vibration 7: 39-56.
http://dx.doi.org/10.1155/2000/405046.
[19.] Olson, L.G.; Amison, D.J. 1997. Application of a general
purpose finite element method to elastic pipes conveying fluid, Journal
of Fluids and Structures 11: 207-222.
http://dx.doi.org/10.1006/jfls.1996.0073.
[20.] Zou, G.P.; Cheraghi, N.; Taheri, F. 2005. Fluid-induced
vibration of composite natural gas pipelines, International Journal of
Solids and Structures 42: 1253-1268.
http://dx.doi.org/10.1016/j.ijsolstr.2004.07.001.
[21.] Pierre-Antoine Emorine, Alexandre Veillon, Jerome Vinconneau
2010. Modelisation des interactions fluide-structure, LAMIH ENSIAME C2S
2 (a), Rapport de Projet.
[22.] Singiresu, S. Rao 2011. Mechanical Vibrations 5th ed, Pearson
Education, Inc., publishing as Prentice Hall, 1 Lake Street, Upper
Saddle River, NJ 07458.
[23.] Paidoussis, M.P. 2005. Some unresolved issues in
fluid-structure interactions, Journal of Fluids and Structures 20:
871-890. http://dx.doi.org/10.1016/j.jfluidstructs.2005.03.009.
[24.] Schmucker, H.; Flemming, F.; Coulson, S. 2010. Two-way
coupled fluid structure interaction simulation of a propeller turbine,
IOP Conference Series: Earth and Environmental Science 12(1): 012011.
http://iopscience.iop.org/article/10.1088/1755-1315/12/1/012011.
M. Dahmane, D. Boutchicha, L. Adjlout
ONE-WAY FLUID STRUCTURE INTERACTION OF PIPE UNDER FLOW WITH
DIFFERENT BOUNDARY CONDITIONS
Summary
This paper deals with the analytical modelisation and the numerical
simulation of fluid structure interaction (FSI) phenomena of pipeline.
The equations of motions and equilibrium are coupled together taking
into account the influence of the fluid flow inside the pipe. The fluid
circulating in the pipe has the flexional motion as that of structure.
The analytical model is based on the Newtonian approach. The
practicability of the calculation model and the effects of
fluid-structure interaction are illustrated by calculations for some
simple systems, for pipe with boundary condition pinned-pinned.
Independently of the analytical method, a numerical modal analysis is
realized in the fluid structure configuration. This configuration has
been simulated by coupling the two commercial solvers: Fluent code of
fluid mechanical (CFD, Computational Fluid Dynamics), and ANSYS
classical code of structure mechanical (CSD, Computational structure
Dynamics) for pipes with different boundary conditions: pinned-pinned,
clamped-pinned and clamped-clamped. We describe the technique used to
solve the equations associated in the determination of natural
frequencies. For this, we use code ANSYS Workbench where the equations
are discritized with finite element method. The results obtained
numerically are similar to those obtained by the analytical approach.
Keywords: Fluid conveying pipes frequency, fluid structure
interaction, Galerkin, foundation, finite element.
Received September 20, 2015
Accepted November 25, 2016
M. Dahmane (*), D. Boutchicha (*), L. Adjlout (**)
(*) LMA, Mechanical Engineering Department, USTO-MB, BP 1055 El
Menaour, Oran 31000, Algeria, E-mail: dahmanemoul@gmail.com
(**) Mechanical Engineering Faculty, USTO-MB, BP 1055 El Menaour,
Oran 31000, Algeria, E-mail: adjlout@yahoo.fr
Table 1
Integral values [b.sub.sr]. [b.sub.rr]. [C.sub.sr], [C.sub.rr] for
pinned-pinned pipe by [4]
Parameters Pinned-pinned pipes
[b.sub.sr] [2[[alpha].sub.r][[alpha].sub.s]/
([[alpha].sub.r.sup.2]-
[[alpha].sub.r.sup.2])]
[[[(-1).sup.s+r]-1]]
[b.sub.rr] 0
[C.sub.sr] 0
(s [not equal to] r)
[C.sub.rr] -[[alpha].sub.r.sup.2]
Table 2
The two first frequencies parameter pinned-pinned fluid-conveying pipe,
comparison Fourier series and Galerkin method for various values of u
u [[OMEGA].sub.1] [[OMEGA].sub.1] % Var. [[OMEGA].sub.2]
Fourier series Galerkin method Fourier series
0.0 9.869 9.8685 0.005 39.478
0.1 9.864 9.8603 0.037 39.475
0.3 9.820 9.8163 0.037 39.466
0.5 9.731 9.7281 0.029 39.401
1.0 9.310 9.3046 0.054 39.168
1.5 8.574 8.5705 0.035 38.773
2.0 7.455 7.4515 0.035 38.208
2.5 5.779 5.7751 0.039 37.456
3.0 2.784 2.7790 0.050 36.497
3.141 0.0 0.0 0.0 36.183
u [[OMEGA].sub.2]
Galerkin method
0.0 39.4635
0.1 39.4604
0.3 39.4362
0.5 39.3863
1.0 39.1532
1.5 38.7583
2.0 38.1932
2.5 37.4417
3.0 36.4822
3.141 36.1699
Table 3
The first frequency parameter pinned-pinned pipe for various values
[[gamma].sub.1]
[[gamma].sub.1] [[OMEGA].sub.1] [[OMEGA].sub.1]present % Variation
(Ref.[l7])
0.0 9.865 9.869 0.040
[10.sup.2] 14.047 14.049 0.014
[10.sup.4] 100.480 100.491 0.010
Table 4
The first frequency parameter pinned-pinned fluid-conveying pipe for
various values of [[gamma].sub.1], u and [beta] = 0.1
[[gamma].sub.1]
u 0.01 0.5 2.5 10 100
0.0 9.8695 9.8943 9.9949 10.3632 14.0498
1.0 9.3463 9.3730 9.4789 9.8484 13.4649
2.0 7.5792 7.6112 7.7404 8.1175 12.5138
3.1415 0.0988 0.6987 1.5624 2.3882 9.8822
3.1416 0.0 0.2677 1.2134 2.3834 9.8801
3.1495 0.0 0.6331 2.2785 9.8580
3.1953 . . 0.0 2.3834 9.7107
3.2988 . . . 0.0 9.3521
4.4723 . . . . 0.0
Table 5
Two first natural frequencies (in Hz) pinned-pinned fluid-conveying
pipe for various values of u, [beta] = 0.6
u [V.sub.0] Analytical results
[[OMEGA].sub.1] [[OMEGA].sub.2] [f.sub.1], [f.sub.2],
Hz Hz
0.00 0.00 9.8596 39.4780 58.5150 234.0596
0.10 5.15 9.8466 39.4699 58.4379 234.2472
0.20 10.31 9.8453 39.4647 58.4302 234.2163
0.30 15.47 9.8240 39.4383 58.3037 234.0596
0.40 20.63 9.7869 39.4383 58.0812 233.7036
0.50 25.79 9.7402 39.3780 57.8064 233.7018
1.00 51.58 9.2983 39.3119 55.1838 233.3095
1.50 77.37 8.5753 38.7731 55.1838 230.1118
2.00 103.17 7.4449 38.1691 44.1842 226.5272
u Numerical results % Variation [f.sub.1]
[f.sub.1], Hz [f.sub.2], Hz
0.00 58.8430 234.2900 0.56
0.10 58.8410 234.3120 0.68
0.20 58.4210 234.3100 0.18
0.30 58.2470 234.3100 0.56
0.40 58.2330 234.2660 0.26
0.50 57.3400 234.1260 0.80
1.00 55.4210 230.1320 0.42
1.50 55.2835 230.0115 0.18
2.00 44.3412 227.1233 0.35
Table 6
Two first natural frequencies (in Hz) fluid-conveying pipe for various
values of [[gamma].sub.1] where [beta] = 0.5
[[gamma].sub.1] Pinned-pinned pipe
Numerical Analytical
results results
[f.sub.1] [f.sub.2] [f.sub.1] [f.sub.2]
0.01 58.84 234.79 58.57 234.09
0.50 58.85 234.84 58.65 234.14
2.50 59.56 235.01 58.87 234.91
10 62.14 235.31 62.11 235.07
100 78.36 241.51 78.13 241.22
1000 212.56 309.53 211.96 308.93
10000 231.56 577.25 231.12 576.85
[[gamma].sub.1]
%Var.
[f.sub.1]
0.01 0.45
0.50 0.33
2.50 1.15
10 0.04
100 0.29
1000 0.28
10000 0.19
Table 7
Two first natural frequencies (in Hz) clamped-pinned fluid-conveying
pipe for various values of [y.sub.1], [beta] = 0.5
[[gamma].sub.1] Clamped-pinned pipe
Numerical Results Analytical Results %Var.
[f.sub.1] [f.sub.2] [f.sub.1] [f.sub.2] [f.sub.1]
0.01 91.74 295.63 91.50 295.36 0.26
0.50 91.78 295.66 91.54 295.39 0.26
2.50 92.24 295.70 91.84 295.66 0.43
10 93.68 296.34 93.08 296.05 0.64
100 112.02 300.64 111.92 300.22 0.08
1000 224.13 357.64 223.97 357.52 0.07
10000 292.02 651.12 291.88 651.01 0.04
Table 8
First natural frequencies (in Hz) clamped-clamped fluid-conveying pipe
for various values of [[gamma].sub.1] and [V.sub.0], [beta]=0
[[gamma].sub.1]
[V.sub.0], m/s 0.01 0.5 2.5 10 100 1000
0.00 132.78 132.80 132.90 133.90 147.32 243.72
25.79 132.61 132.63 132.85 133.71 147.14 243.62
51.58 129.96 129.99 130.26 131.15 144.82 242.22
77.37 128.06 128.68 128.73 129.66 140.02 237.04
103.17 124.30 124.59 125.80 126.53 146.56 222.38
Table 9
Second natural frequencies (in Hz) clamped-clamped fluid-conveying pipe
for various values of [[gamma].sub.1] and [V.sub.0], [beta] = 0.6
[[gamma].sub.1]
[V.sub.0], m/s 0.01 0.5 2.5 10 100 1000
0.00 362.98 363.01 363.12 363.53 367.65 415.46
25.79 361.11 361.88 361.99 362.41 367.52 415.34
51.58 360.11 360.14 360.25 360.67 365.80 413.82
77.37 353.79 353.82 353.99 354.35 359.58 408.34
103.17 337.12 337.14 337.27 337.72 343.21 394.03
COPYRIGHT 2016 Kauno Technologijos Universitetas
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.