scispace - formally typeset
Search or ask a question
Journal ArticleDOI

Calculating Rotordynamic Coefficients of Seals by Finite-Difference Techniques

01 Jul 1987-Journal of Tribology-transactions of The Asme (American Society of Mechanical Engineers)-Vol. 109, Iss: 3, pp 388-394
TL;DR: In this paper, the Navier-Stokes equations in connection with a turbulence (kappa-epsilon) model are solved by a finite-difference method, where a motion of the shaft round the centered position is assumed.
Abstract: For modelling the turbulent flow in a seal the Navier-Stokes equations in connection with a turbulence (kappa-epsilon) model are solved by a finite-difference method. A motion of the shaft round the centered position is assumed. After calculating the corresponding flow field and the pressure distribution, the rotor-dynamic coefficients of the seal can be determined. These coefficients are compared with results obtained by using the bulk flow theory of Childs and with experimental results.

Summary (1 min read)

FINITE-DIFFERENCE METHOD

  • Because of its general convergence a 'hybrid -difference' method is used, which meansthat the convective terms are calculated by a 'upwind'-or a 'central-difference' method as a function of flow-velocity and grid-distance.
  • To satisfy the entrance condition the authors make use of the iterative character of the finite-difference method.
  • This means that the authors start with a pressure PlBj^at the entrance and after every iteration step they check if the calculated UlB j satisfy condition (21).
  • If not, the pressure _IBj will be corrected.
  • The precession frequencies can be arbitrarily chosen, because the dynamic coefficients are mostly independent of them.

RESULTS FOR A GROOVED SEAL

  • The authors also made some calculations, for the grooved seal, whose geometry and seal data are shown in Fig. 12.
  • The superiority of the theory versus other methods is the simplicity in use for grooved seals by A only neglecting the terms D 1 -D8 in the equations for GI, _1' GI'.

Did you find this useful? Give us your feedback

Content maybe subject to copyright    Report

N87-22204
CALCULATING ROTORDYNAMIC COEFFICIENTS OF SEALS
BY FINITE-DIFFERENCE TECHNIQUES
F.J. Dietzen and R. Nordmann
University of Kalserslautern
Kalserslautern, Federal Republic of Germany
For modelling the turbulent flow in a seal the Navier-Stokes equations in con-
nection with a turbulence model (k-_-model) are solved by a finite-difference
method. A motion of the shaft round the centered position is assumed. After cal-
culating the corresponding flow field and the pressure distribution, the rotor-
dynamic coefficients of the seal can be determined. These coefficients are com-
pared with results obtained by using the bulk flow theory of Childs [i] and with
experimental results.
INTRODUCTION
It is well known that the fluid forces in seals, which are described by equa-
tion (I)
have a strong influence on the dynamic behaviour of rotating turbo-machinery.
While there exist some good theories for calculating the coefficients of
straight seals [I], no satisfactory model is known to describe the effects of
grooved seals. Reference [2] presents a survey and comparison of results of
existing theories. The authors' opinion is that the existing methods are not
at all satisfactory. The main weakness of these theories is the fact, that they
are using so called 'bulk-flow-theories' which connect the wall shear stress with
the mean flow-velocity relative to this wall. Howeve_in the region of a groove
there occur stresses in the fluid which cannot be neglected. Calculating the flow
by using the Navier-Stokes equations in connection with a turbulence model elimi-
nates this disadvantage. Therefore, a finite difference model is presented which
allows the calculation of the coefficients by using these equations.
77 PR CF.DmiGPAGE NOT

Nomenclature:
Fz , Fy
K, k
D, d
M, m
U, V_ W
P
k
s
We' PI' _t
P
t
x, r, 8
q
o k , oS , K
Cp, C1, C2
So
C
0
6
r o
r o
e -
Co
k
Forces on the shaft in z and y direction
direct and cross-coupling stiffness in eq. (1, 24)
direct and cross-coupling damping in eq. (1,24)
direct and cross-coupling inertia in eq. (1, 24)
axial, radial and circumferential velocity
pressure
turbulence energy
energy dissipation
effective, laminar and turbulent viscosity
density
time
axial, radial and circumferential coordinate
radial coordinate after transformation
Constants od the k-s-model
Constants of the k-s-model
general variable standing for u, v, w, p, k or s
general source term
seal clearance by centric shaft position
seal clearance by eccentric shaft position
radius of the precession motion of the shaft
perturbation parameter
rotational frequency of the shaft
precession frequency of the shaft
entrance lost-coefficient
Length of the seal
78

r i
r a
Subscripts
0
1
R
S
radius of the rotor (shaft)
radius of the stator
zeroth order variables
first order variables
rotor
stator
MATHEMATI CAL MODEL
To describe turbulent flow by the Navier-Stokes equations the velocities and the
pressure are separated into mean and fluctuating quantities.
u = u + u' v = v + v'
w = w + w' p = p + p'
Time-averaging of the Navier-Stokes equations leads to terms of the following
form: _-r_-r, _-r_-r, _-r_T.
TO substitute these terms one can use the Boussinesq's eddy-viscosity concept.
For example:
Pt 3u 3v
= - 7 (_ + Tx ) (2)
Pt is the turbulent viscosity, which is not a fluid property but depends strongly
on the state of flow. Summing up the laminar and turbulent viscosity to an effec-
tive viscosity
_e = _I + _t ' (3)
one obtains the following time-averaged Navier-Stokes equations for turbulent
flow. (In the following the overbars are omitted.)
I. axial momentum:
3u 3 3 , 3u, 13 13 3u 13 , wu" la ,1 3u,
p_ + _-_(puu) - _t_eT_; + _(rpvu) - _(rl_e_- _) + _-_tp ; - _-ot_Pe_; =
_@__p a , au, la , 3v, 13 , Bw,
3x + _t_ea-x; + r_e-@x; + raOtPe_-£;
(4)
79

2. radial momentum:
3v
P_ + __x(pUv) 3 , 3v, 13 "r " 13 r 3v, 13 13 ,1 3v,
- _-_,pe_-_) + -_-_ pvv) - r_-{(pe_--_-) + _--6(pwv) - _-6_--_e_-_; :
__p 13 @v 3 , 3u, 13 3 w 2 3w 2 _0 2
3r + r_-r(rPe_ ) + 3-x_Pe_-r ) + r3-c)(rPe_-r(-r)) - -r2"e_ - r2Pe v + r w (5)
3. tangential momentum
3w 3 _ , 3w, 13 13 3w 13 13 ,i 3w,
p_ + _-x(pUw) - _tpe_-_) + _(rpvw) - _-_(rlJe_ ) + r_(pww) - r_-6trlJe_-_)=
_lBp 13 _v 3 ,1 3u, 1 3v
r30 + r_(_e_ ) + _£r_e_-O _ + rZ_e_
4. continuity equation
w 3 13 ,2 v" 13 ,1 3w, Pvw
rz_(r]_ e) + r_-_r]_e ) + r_-E)_Pe_-_).-
(6)
13 'r " 13
x(PU)+ pv; + T (pw): o
(7)
To describe Pt we use the k-_ turbulence model [3, 4]. This model determines Pt
as a function of the kinetic energy k of the turbulent motion and the energy
dissipation _. It is relative simple and often used to calculate the turbulent
flow in seals [12, 13, 14, 15]. Stoff [12], for example, compares his flow meas-
urements in a labyrinth seal with calculations on base of the k-e model. He
observes that both agree well.
k2 (8)
Pt = c p_--
The equations for k and c can be derived in exact form from the Navier-Stokes
equations
5. turbulence energy k
. 3 _Pe3k_ . 13 13 , Pe3k
3_k 3 _puk) .... _ -
_St " 3-x' 3X_OkSX_ _(rpvk_ _-_£r_kk_-_)
G - pe
+ l_(pwk) 13 _IPeSk_
- =
(.9)
80

6. energy dissipation
ae a _(laeae la la , Ueae,
p_ + _(pue) - ax'a a-x) + _(rpve) -_-r_r_--a--6)
la la (l_eae_ =
+ Fge(pwe) - Fa-o'F6-g6'
ez (10)
C2p_-
CI[G -
G = lae{ 2(/av_ 2 tau/2 /law + v 2) av au_2 flav
'_' + W_' + 'rae -_) + (gi + ag' + 'Fa_ +
aw w = aw lau_2}
ar r) + (a-x + raC)'
C = 0.09 CI = 1.44 C2 = 1.92
K
< = 0.4187 o k = 1. o = C_(Cl - Cz)
(II)
To model the flow in the case of a shaft moving on an eccentric orbit, a coordi-
nate-transformation [5, 6] is made. (Fig. 1)
r e-r
n = r a - _ CO (12)
8(O,t) is the seal clearance, varying with angle @ and time t. By this trans-
formation the eccentric moving shaft is reduced to a shaft rotating in the
centre of the seal.
We must note that the following relations of the transformation must be used.
+ (an)O(_-_)r
aq_ aq_ aq_ an
(Y[)r = (_)n + (a-n)t(a-%-lr (131
(_x) r a@
:
81

Citations
More filters
Journal ArticleDOI
TL;DR: In this article, the authors used 3D computational fluid dynamics (CFD) to model the labyrinth seal flow path by solving the Reynolds Averaged Navier Stokes equations.
Abstract: Labyrinth seals are utilized inside turbomachinery to provide noncontacting control of internal leakage. These seals can also play an important role in determining the rotordynamic stability of the machine. Traditional labyrinth seal models are based on bulk-flow assumptions where the fluid is assumed to behave as a rigid body affected by shear stress at the interfaces. To model the labyrinth seal cavity, a single, driven vortex is assumed and relationships for the shear stress and divergence angle of the through flow jet are developed. These models, while efficient to compute, typically show poor prediction for seals with small clearances, high running speed, and high pressure.* In an effort to improve the prediction of these components, this work utilizes three-dimensional computational fluid dynamics (CFD) to model the labyrinth seal flow path by solving the Reynolds Averaged Navier Stokes equations. Unlike bulk-flow techniques, CFD makes no fundamental assumptions on geometry, shear stress at the walls, as well as internal flow structure. The method allows modeling of any arbitrarily shaped domain including stepped and interlocking labyrinths with straight or angled teeth. When only leakage prediction is required, an axisymmetric model is created. To calculate rotordynamic forces, a full 3D, eccentric model is solved. The results demonstrate improved leakage anti rotordynamic prediction over bulk-flow approaches compared to experimental measurements.

105 citations

Journal ArticleDOI
TL;DR: In this paper, the linearized rotordynamic coefficients for a seal with a large aspect ratio were calculated by means of a three-dimensional CFD analysis performed to predict the fluid-induced forces acting on the rotor.
Abstract: Traditional annular seal models are based on bulk flow theory While these methods are computationally efficient and can predict dynamic properties fairly well for short seals, they lack accuracy in cases of seals with complex geometry or with large aspect ratios (above 10) In this paper, the linearized rotordynamic coefficients for a seal with a large aspect ratio are calculated by means of a three-dimensional CFD analysis performed to predict the fluid-induced forces acting on the rotor For comparison, the dynamic coefficients were also calculated using two other codes: one developed on the bulk flow method and one based on finite difference method These two sets of dynamic coefficients were compared with those obtained from CFD Results show a reasonable correlation for the direct stiffness estimates, with largest value predicted by CFD In terms of cross-coupled stiffness, which is known to be directly related to cross-coupled forces that contribute to rotor instability, the CFD also predicts the highest value; however, a much larger discrepancy can be observed for this term (73% higher than the value predicted by the finite difference method and 79% higher than the bulk flow code prediction) One can see similar large differences in predictions in the estimates for damping and direct mass coefficients, where the highest values are predicted by the bulk flow method These large variations in damping and mass coefficients, and most importantly the large difference in the cross-coupled stiffness predictions, may be attributed to the large difference in seal geometry (ie, the large aspect ratio AR >10 of this seal model versus the short seal configuration the bulk flow code is usually calibrated for using an empirical friction factor) [DOI: 101115/14007341]

38 citations

Journal ArticleDOI
TL;DR: In this article, the rotordynamic forces arising from shrouded centrifugal impellers were analyzed using finite element stress analysis and computational fluid dynamics (CFD) techniques, and the results showed good correlation with experiment for both performance and rotord dynamic forces.
Abstract: The demand for higher efficiencies and performance of modern centrifugal turbomachinery requires improved knowledge of critical design factors in strength of materials, aerodynamics, and rotordynamics. While tremendous strides in finite element stress analysis and computational fluid dynamics (CFD) have addressed the first two areas, the lack of accurate prediction tools for centrifugal impellers typically leaves rotordynamics out of the design loop. While several authors have analyzed the rotordynamic forces arising from shrouded centrifugal impellers, there has been no study to couple the secondary shroud passage with the three-dimensional primary flow model. The strong interaction between these domains makes this approach advantageous. The current study utilizes CFD techniques to analyze the full three-dimensional viscous, primary/secondary flow field in a centrifugal pump impeller to determine rotordynamic forces. Multiple quasisteady solutions of an eccentric three-dimensional model at different precessional frequency ratios yield the rotordynamic impedance forces. Performing a second order least-squares analysis generates the skew-symmetric stiffness, damping, and mass matrices. The results show good correlation with experiment for both performance and rotordynamic forces.

35 citations

Journal ArticleDOI
TL;DR: In this paper, the rotordynamic forces are obtained by means of a time-dependent perturbation of the rotor position with respect to the stator, and a sequence of perturbations frequencies is utilized to obtain the frequency dependence of the rotor dynamic force coefficients.
Abstract: This paper deals with modeling of hole-pattern and honeycomb seals. These are frequently used as balance piston seals in high pressure centrifugal compressor applications as they have the potential to facilitate superior rotordynamic damping characteristics while providing good leakage control. On the other hand it is also well-established that the rotordynamic performance of hole-pattern and honeycomb seals is very sensitive to convergence and divergence in the streamwise direction.The ISOTSEAL bulk-flow code has shown difficulties in predicting the rotordynamic coefficients for convergent seal geometries or in cases with negative preswirl. This has lead to increased interest in CFD-based analysis of seal dynamics. CFD-based models generally have less assumptions and are applicable for complex geometries or operating ranges not covered by bulk-flow codes.The CFD-based Instationary Perturbation Model (IPM) is utilized for the analysis of the hole-pattern and honeycomb seals. The rotordynamic forces are obtained by means of a time-dependent perturbation of the rotor position with respect to the stator. A sequence of perturbation frequencies is utilized to obtain the frequency dependence of the rotordynamic seal force coefficients.A strong effort has been put into validating the CFD-based perturbation modeling techniques against published experimental seal test data and the paper describes selected validation cases. A constant-clearance hole-pattern seal and a convergent honeycomb seal are analyzed and the results are compared to experimental results. The frequency dependence of the rotordynamic stiffness and damping characteristics of the seals is very well-captured for both types of seals.Finally the IPM method was applied to a convergent hole-pattern seal to investigate the effects of eccentricity on the rotordynamic coefficients. The results are consistent with available experimental data.Copyright © 2012 by ASME

35 citations

01 Jan 2013
TL;DR: In this article, Wu et al. investigated the dynamics of a TCP system described by a first-order nonlinear delay differential equation and showed that a Hopf bifurcation sequence occurs at the pos- itive equilibrium as the delay passes through a sequence of critical values.
Abstract: This paper investigates the dynamics of a TCP system described by a first- order nonlinear delay differential equation. By analyzing the associated characteristic transcendental equation, it is shown that a Hopf bifurcation sequence occurs at the pos- itive equilibrium as the delay passes through a sequence of critical values. The explicit algorithms for determining the Hopf bifurcation direction and the stability of the bifur- cating periodic solutions are derived with the normal form theory and the center manifold theory. The global existence of periodic solutions is also established with the method of Wu (Wu, J. H. Symmetric functional differential equations and neural networks with mem- ory. Transactions of the American Mathematical Society 350(12), 4799-4838 (1998)).

31 citations

References
More filters
Journal ArticleDOI
TL;DR: In this paper, a finite-length solution procedure is developed for perturbation equations which are based on Hirs' (1973) turbulent lubrication model and apply to small motions about a centered position.
Abstract: A finite-length solution procedure is developed for perturbation equations which are based on Hirs' (1973) turbulent lubrication model. The equations apply to small motions about a centered position and include the influence of swirl and changes in Reynolds number due to perturbations in clearances. Numerical results are presented for a range of L/D ratios, with and without swirl. For zero swirl, changes in the L/D ratios show results which are similar to those obtained by Black and Jenssen (1970), but when L/D = 1, differences of about 15 percent appear. The results including swirl give physically insupportable results at small L/D ratios, such as a negative cross-coupled stiffness coefficient at L/D = 0.2. This result demonstrates that the complete Hirs turbulence model is not suitable for short seals with significant swirling flow.

131 citations

Journal ArticleDOI
TL;DR: A new approach was developed and tested for alleviating the substantial convergence difficulty which results from implementation of the QUICK differencing scheme into a TEACH-type computer code, and the resulting CPU time and number of numerical iterations required to obtain a solution compare favorably with a previously recommended method.
Abstract: A new approach was developed and tested for alleviating the substantial convergence difficulty which results from implementation of the QUICK differencing scheme into a TEACH-type computer code. It is relatively simple, and the resulting CPU time and number of numerical iterations required to obtain a solution compare favorably with a previously recommended method. This approach has been employed in developing a computer code for calculating the pressure drop for a specified incompressible flow leakage rate in a labyrinth seal. The numerical model is widely applicable and does not require an estimate of the kinetic energy carry-over coefficient for example, whose value is often uncertain. Good agreement with measurements is demonstrated for both straight-through and stepped labyrinths. These new detailed results are examined, and several suggestions are offered for the advancement of simple analytical leakage as well as rotordynamic stability models.

54 citations

DOI
01 Jan 1979
TL;DR: These Ecole polytechnique federale de Lausanne EPFL, n° 342 (1979) Reference as mentioned in this paper, Section 5.1, Section 2.2.1.
Abstract: These Ecole polytechnique federale de Lausanne EPFL, n° 342 (1979) Reference doi:10.5075/epfl-thesis-342Print copy in library catalog Record created on 2005-03-16, modified on 2016-08-08

4 citations

01 Oct 1982
TL;DR: In this paper, a finite-length solution procedure is developed for perturbation equations which are based on Hirs' (1973) turbulent lubrication model and apply to small motions about a centered position.
Abstract: A finite-length solution procedure is developed for perturbation equations which are based on Hirs' (1973) turbulent lubrication model. The equations apply to small motions about a centered position and include the influence of swirl and changes in Reynolds number due to perturbations in clearances. Numerical results are presented for a range of L/D ratios, with and without swirl. For zero swirl, changes in the L/D ratios show results which are similar to those obtained by Black and Jenssen (1970), but when L/D = 1, differences of about 15 percent appear. The results including swirl give physically insupportable results at small L/D ratios, such as a negative cross-coupled stiffness coefficient at L/D = 0.2. This result demonstrates that the complete Hirs turbulence model is not suitable for short seals with significant swirling flow.

2 citations