Available online at www.sciencedirect.

com

Procedia Engineering 31 (2012) 156 165

International Conference on Advances in Computational Modeling and Simulation

Numerical simulation of cavitating turbulent flow in a

Department of Engineering Mechanics,Kunming University of Science and Technology,kunming 650500,P.R.China

Abstract
The paper presents numerical analysis of cavitating turbulent flow in a high head Francis turbine at part load
operation. Analysis was performed by OpenFOAM code. The k  Z SST turbulence model is employed in the
Reynolds averaged NavierStokes equations in this study. A mixture assumption and a finite rate mass transfer model
was introduced. The computing domain includes the spiral casing, the guide vanes, the runner, and the draft tube,
which is discretized with a full three-dimensional mesh system of unstructured tetrahedral shapes. The finite volume
method is used to solve the governing equations of the mixture model and the pressure-velocity coupling is handled
via a Pressure Implicit with Splitting of Operators(PISO) procedure. The tendency of the cavitating flow in the runner
and draft tube agrees well with the turbine empirical results.

2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of Kunming
University of Science and Technology Open access under CC BY-NC-ND license.
KeywordsCFD; cavitating turbulent flow; Francis turbine; OpenFOAM

1. Introduction
Cavitation is a three-dimensional, unsteady and discontinuous phenomenon of formation, growth and
rapid collapse of bubbles. It makes the hydraulic machinery present unwanted consequences such as flow
instabilities, excessive vibrations and damage to material surfaces and degradation of machine
performance [1]. A Francis turbine operating at part load will works under off-design condition. In this

* Corresponding author. Tel.+86-18064878136;

1877-7058 2011 Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
doi:10.1016/j.proeng.2012.01.1006

Hongming Zhang and Lixiang Zhang / Procedia Engineering 31 (2012) 156 165

case, cavitation could not be avoided. The main types of the cavitation in Francis turbines are leading
edge cavitation, travelling bubble cavitation, von Karman vortex cavitation and draft tube swirl.
Traditionally, the studies for cavitation rely heavily on experimental model testing, which usually is very
costly, difficult and time consuming. In recent years, due to the rapid development in CFD technology
and computer hardware, numerical cavitating flow simulations have been applied to rotating machinery
such as pumps, marine propeller, and water turbines [2-5].However, since the cavitation is such a
complicated physical phenomenon, it is impossible to treat all factors which may affect the cavitation.
Several physical and numerical models have been developed to investigate cavitating flows. Studies may
be classified into two categories: the interface fitting method and the continuum modelling method [6].
Interface methods assume there is a clear and distinct interface between the liquid and vapour, which can
be determined via an iterative procedure. Its applications are limited to simpler problems .Continuum
methods make no attempt to track the liquid and vapour interface but treats the flow as two phases with a
mixture density, which continuously varies between the liquid and vapour extremes. These two-phase
models are implemented using different approaches: the one-fluid method, the two-fluid method and the
mixture method. The two-fluid approach assumes that both phases co-exist at every point in the flow field
and each phase is governed by its own set of conservation laws. Because the exchanges of mass, the
transfer terms have to be known; such quantities are usually very difficult to obtain. The one-fluid method
treats the cavitating flows as a mixture of two fluids behaving as one. The governing equations are
composed of three conservation laws written for the mixture. These models are based on the assumption
of local kinematic equilibrium between phases and local thermodynamic equilibrium. Moreover,
vaporization or condensation processes are assumed to be instantaneous. But, because of its simplicity, it
is often used for numerical simulations. Mixture models are intermediate models between one-fluid and
two-fluid models, based on four or five equations. The model is obtained by adding a mass equation for
the vapour or liquid density including a cavitation source term. The main difficulty is related to the
formulation of the source term and the tunable parameters involved in the vaporization and condensation
process. This model is very popular in simulating cavitating flows[7-11].
For simulating cavitating flows in a francis turbine, Liu et al.[8] used a mixture model for the
cavity/liquid two-phase flow together with the RNG k - H turbulence model to study the cavitation
performance of a Francis turbine. The unsteady cavitating flow simulation through the Francis turbine is
conducted as well to study the pressure fluctuation characters caused by the vortex rope in the draft tube
at partial load operation. SUSAN-RESIGA et al.[9] introduce a mixture approach for the two-phase
cavitating flows, together with the model employed for evaluating the interphase mass flow rate. This
cavitating flow model is used to investigate the cavitation in the GAMM Francis turbine runner. Results
for cavity shape and extent, as well as for the pressure distribution on the blade with and without cavity
are presented and discussed. Zhou et al. [10] used the 3-D unsteady Reynolds averaged Navier-stokes
equations based on the pseudo-homogeneous flow theory and a vapour fraction transport-equation that
accounts for non-condensable gas to solve cavitating flow in a Francis turbine. The calculation results
agreed with experiment data reasonably. Wu et al.[11] presented the simulation results for the unsteady
cavitating turbulent flow in a Francis turbine using the mixture model for cavityliquid two-phase flows.
In the mixture model, an improved expression for the mass transfer is employed which is based on
evaporation and condensation mechanisms with considering the effects of the nondissolved gas, the
turbulence, the tension of interface at cavity and the effect of phase change rate and so on.
OpenFOAM is an object-oriented C++ library of classes and routines of use for writing CFD codes. It
has a set of basic features similar to any commercial CFD solver, such as turbulence models and
discretization schemes. With OpenFOAM it is easy to add any modification to any part of the
implementation. It is increasingly being used in the numerical simulation of hydraulic machinery.
Nilsson[12] has evaluated the OpenFOAM code for turbulent flow in water turbines. His aim is to

157

158

Hongming Zhang and Lixiang Zhang / Procedia Engineering 31 (2012) 156 165

validate the OpenFOAM results for cases that have also been computed with other CFD codes, as well as
have been experimentally investigated. Comparisons are made both with numerical and experimental
results. Petit et al.[13] presented numerical results, using OpenFOAM, of the flow in the swirl flow
generator. The numerical results are validated against experimental LDV results, and against designed
velocity profiles. The investigation shows that OpenFOAM gives results that are comparable to the
experimental and designed profiles. Bensow et al.[14] considered the cavitating flow as a single fluid,
two-phase mixture (by Kunz et al.[15]) and used an incompressible Large Eddy Simulation technique in
OpenFOAM code to study cavitating turbulent flow on a NACA0015 hydrofoil and the INSEAN E779A
propeller etc.. The cases presented in the paper show that the cavitationg model in OpenFOAM have the
ability to capture several important cavitation mechanisms, necessary for the prediction of cavitation
nuisance such as erosion or noise.
This paper presents using OpenFOAM code to simulate the cavitating turbulent flow in a high head
prototype Francis turbine at part load operation .The k  Z SST turbulence model for the Reynolds
averaged Navier-Stokes equations and a mixture cavitation model for the mass transfer by Kunz et al.[15]
is employed in OpenFOAM code.
2. Governing equations and Cavitation model
To simulate cavitating flows, the two phases need to be represented in the problem, as well as the
phase transition mechanism between the two. The governing equations include the mixture Navier-Stokes
equations and a two-phase cavitation model, which consider the local vapour volume fraction and having
the spatial and temporal variation of the vapour fraction described by a transport equation including
source terms for the mass transfer rate between the phases.
Continuity equation of mixture

w
( U )  <( U u) 0
wt

(1)

Where U is the density of the mixture of cavity phase and liquid phase,

u is the Reynolds averaged

relative velocity.
Continuity equation of cavity phase

w
( UvD v )  <( UvD v u)
wt
where

m

D v is the volume fraction of cavity phase; U v

(2)
is the density of vapour cavity phase; and

m is

the mass transfer between vapour and liquid phase.

w( U u)
 <( U uu)
wt
where,

1
p  [(P  Pt )<u]  <[( P  Pt )u]  U g (3)
3

p is the static pressure; P is the viscosity of mixture; Pt is the eddy viscosity of mixture.The

density of the mixture

and the viscosity of mixture

can be expressed as

U Dv Uv  (1  Dv ) Ul

(4)

159

Hongming Zhang and Lixiang Zhang / Procedia Engineering 31 (2012) 156 165

P Dv Pv  (1  Dv )Pl
Where,

(5)

Pl is the viscosity of water, Pv is the viscosity of vapour.

The expression of mass transfer rate in (2) is based on the ideas of Kunz et al.[15]. It uses different
  , is modled to be proportional to the
strategies for vaporization and condensation. The vaporization, m
amount by which the pressure is below the vapour pressure and the amount of liquid present, while the
  , is based on a third order polynomial function of the vapor volume fraction,
condensation, m

min[0, p  pv ]
1
UlU f
2

m 

A Uv (1  D v )


m

A Uv (1  Dv )Dv 2

(6)

(7)

and the specific mass transfer rate is computed as

 m
 m

m

(8)

pv is the vaporization pressure and A and A is empirical constants determining the mass
transfer rate, U is the far-field velocity.
f
where,

3. Turbulence model
In this paper, The k  Z SST turbulence model is adopted for simulation of the unsteady cavitating
flow through a Francis turbine. It has been implemented in the OpenFOAM libraries. The turbulent
kinetic energy equation, the specific dissipation rate equation, and the turbulent kinematic eddy viscosity
are given by[16]

wk
wk
U j
wt
wx j

Pk  E *kZ 

w
wk
[( P  D k Pt )
]
wx j
wx j

wZ
wZ
w
wZ
U j
GZ  EZ 2 
[(P  D k Pt )
]  (1  F1 )CDkZ
wt
wx j
wx j
wx j
a1k
Pt
max(a1Z , SF2 )

(9)
(10)
(11)

Where

Pk

min(Gk , c1E *kZ )

(12)

Gk

Pt S

2Sij Sij

(14)

Sij

1 wU i wU j
(
)

2 wx j
wxi

(15)

(13)

160

Hongming Zhang and Lixiang Zhang / Procedia Engineering 31 (2012) 156 165

GZ
CDkZ

J
G
Pt k
2DZ 2

(16)

1 wk wZ
Z wxi wxi

(17)

The two blending functions are given by

F1

F2

tanh{{min(min[max(

tanh{[min(max(

500P 4DZ 2 k
),
],10)}4 }

2
2
E Z y y Z CDkZ y
k

(18)

2 k 500P
,
),100)]2}
E *Z y y 2Z

(19)

Where
(20)
max(CDkZ ,1e10 )
and y is the distance to the wall. F1 = 0 in the freestream ( k  H model) and F1 = 1 in the boundary
layer ( k  Z model).
All constants are computed by a blend from the corresponding constants of the k  H and the k  Z
models as D k D k1F1  D k 2 (1  F1 ) . The constants used here are:

CDkZ

E * 0.09,D k1 0.85034,D k 2 1.0,DZ1 0.5,DZ 2

0.85616, J

0.5532, E

0.0750

4. Numerical algorithm
In the OpenFOAM , the spatial discretization is performed using a cell centered co-located finite
volume method for unstructured meshes with arbitrary cell-shapes, and a multi-step scheme is used for
the time derivatives. For the simulations presented in this paper, a second order implicit time scheme is
used combined with second order linear interpolation in space, except for the convective terms. The time
step is equal to 1 degrees of runner revolution.The iterative solvers are considered converged when the
residuals have been reduced by a factor of 10-5. The pressure-velocity coupling is handled via a Pressure
Implicit with Splitting of Operators(PISO) procedure. Before entering the PISO-loop, the vapor volume
fraction transport equation is solved as well as a momentum predictor step. The mass transfer terms are
incorporated into the pressure Poisson equation as a split source term with the part including the pressure
is treated implicitly, whereas the rest is treated explicitly[14].
5. Simulation of unsteady cavitating flow in a high-head prototype Francis turbine
Figure 1 shows the calculation domain of the high-head prototype Francis turbine. The parameters
are listed in Table 1.
The computational grid of the complete turbine is generated using ICEM CFD ,including spiral
casing with stay vanes, guide vane cascade, runner and draft tube with 8,300,000 elements was used.
After importing the meshed geometry in OpenFOAM, it is divided into four domains namely spiral

Hongming Zhang and Lixiang Zhang / Procedia Engineering 31 (2012) 156 165

casing ,vanes, runner and draft tube, in which runner is rotating domain and others are stationary domains.
In the algorithm, The length of the time step was equal to one of the runner revolution. Rotation
speed of the runner was 600 r/min. Thus, the time step was 0.0002778 s. In the average a converged
solution was obtained after five to ten iterations in each time step, depending on the flow rate operating
regime.

Table 1 Parameter of Francis turbine

Fig. 1 Francis turbine

Diameter of runner(cm)
Number of runner blade(long)
Number of runner blade(short)
Number of stay vane
Number of guide vane
Rated head (m)
Rated power (kw)
Rotational speed (rpm)
Rated flow rate (m3/s)

175
15
15
24
24
231
25000
600
13.1

5.1 Analysis of the volume fraction of water vapour and pressure on the runner blade
The volume fraction of water vapour is an important feature for cavitating flow in hydraulic turbine. It
varies between 0 and 1. Figures 2 show the volume fraction of water vapour distribution for runner
blade(long) on the front and back. It is clearly seen that the volume fraction of water vapour at the top
right of the blade on the front as well as at bottom left of the blade on the back is higher than others.
Compared with the pressure distribution of Figure 3, the maximum of volume fraction of water vapour is
not at same location with the maximum of pressure, but at the location away from there for some distance
along the flow direction. Figures 3 and Figures 4 are the volume fraction of water vapour distribution and
the pressure distribution for runner short blade on the front and back. The maximum of volume fraction of
water vapour and the maximum of pressure is not at the same location too. This is the characteristics of
the cavitation model, indicating that cavitation is not instantaneous. It also reflects the mechanism of
cavitating flow.

(a)

(b)

Fig.2 The volume fraction of water vapour distribution for runner blade(long) on the front and back

161

162

Hongming Zhang and Lixiang Zhang / Procedia Engineering 31 (2012) 156 165

(a)

(b)

Fig.3 The pressure distribution for runner blade(long) on the front and back

(a)

(b)

Fig.4 The volume fraction of water vapour distribution for runner blade(short)

(a)

(b)

Fig.5 The pressure distribution for runner blade(short) on the front and back

(a)

(b)

Fig.6(a) The volume fraction of water vapour distribution on the bottom of runner exit(b) The photo of a real runner cavitation
damage

Figures 6(a) show the volume fraction of water vapour distribution on the bottom of runner exit and

Hongming Zhang and Lixiang Zhang / Procedia Engineering 31 (2012) 156 165

Figures 6(b) is the photo of real runner cavitation damage. The location of the maximum of volume
fraction of water vapour in Figures 6(a) is very similar to the location of the most severe cavitation
damage in Figures 6(b). This indicates that the code has some ability to capture the cavitation erosion
process.
5.2 The pressure distribution and vortex rope in the draft tube
Pressure fluctuations are usually a result of a strong vortex created in a centre of a flow at the outlet of a
runner. The draft tube vortex appears at partial load operating regimes in francis turbines. Figures 7(a),
(b), (c), (d) show the pressure distribution at middle plane of draft tube when the runner turned a quarter,
two quarters, three quarters and one revolution. Figures 8(a), (b), (c), (d) show the pressure distribution
on the middle cross-section of diffusion tube at corresponding time. It can be seen that the pressure
pulsation in the draft tube is larger.

(a)

(b)

(c)

(d)

fig.7 Pressure distribution in draft tube (middle plane)

(a)

(c)

fig.8 pressure distribution on the middle cross-section of diffusion

(b)

(d)

163

164

Hongming Zhang and Lixiang Zhang / Procedia Engineering 31 (2012) 156 165

Fig.9 Numerically obtained shape of the vortex

(a)

(b)

Fig.10 The shape of the vortex rope in draft tube , ref.[17]

Figure 9 shows the shape of the vortex rope calculated by us with OpenFOAM. The boundary of the
bubble is settled according to the volume fraction of water vapour is 10%. Fig.10(a) shows the vortex
rope in the experiment and Fig.10(b) shows the vortex rope in numeracal simulation by -27DWDO[17].
Compared with the two results, both in the shape of the vortex rope is similar.
6. Conclusion
In this paper, the simulation of cavitating turbulent flow through a whole passage of a prototype high
head Francis turbine has been carried. It is very effective that The k  Z SST turbulence and the mixture
cavitation model for the mass transfer by Kunz et al.[15] used in numerical simulation of cavitating flows.
Simulation results have shown that the cavitating flow in Francis turbine are well predicted. It shows that
OpenFOAM is able to generate good computational results in an efficient way.
Acknowledgements
This work was financially supported by the National Natural Seience Foundation of china
(No.50839003) and the Natural Science Foundation of Yunnan Province(No.2008Ga027).

Reference:
[1]Pardeep Kumar, R.P. Saini. Study of cavitation in hydro turbinesA review, Renewable and Sustainable Energy Reviews, 14

Hongming Zhang and Lixiang Zhang / Procedia Engineering 31 (2012) 156 165

(2010) ,374383
[2]Lanjin Zhang, Jing Xingying, Zhengwei Wang, Chang Jinshi, Peng Guangjie. Analysis of Francis Pump-turbine Runner
Cavitation Flows in Pump Mode, Proceedings of the ASME 2009 Fluids Engineering Division Summer Meeting, FEDSM2009
August 2-6, 2009, Vail, Colorado USA
[3]Anees Ahmed, J.T. Kshirsagar. NUMERICAL PREDICTION OF CAVITATION FREE ZONE OPERATION FOR
FRANCIS TURBINE, Proceedings of the ASME 2009 Fluids Engineering Division Summer Meeting, FEDSM2009August 2-6,
2009, Vail, Colorado USA
[4]Andrej LIPEJ,Dragica JOST,Zoran MARKOV. NUMERICAL PREDICTION OF CAVITATION IN WATER TURBINES,
Proceedings of the XXIst IAHR Symposium on Hydraulic Machinery and Systems, September 9 - 12, 2002, Lausanne
[5]Bin Ji , Xian-wu Luo , Yu-lin Wu, Shu-hong Liu , Hong-yuan Xu , Akira Oshima. Numerical investigation of unsteady
cavitating turbulent flow around a full scale marine propeller, 9th International Conference on Hydrodynamics,October 11-15, 2010
Shanghai, China
[6] Eric Goncalvs. Numerical study of unsteady turbulent cavitating flows, European Journal of Mechanics
B/Fluids,30,(2011),2640
[7]Shuhong Liu, Yulin Wu, Xianwu Luo. NUMERICAL SIMULATION OF 3D CAVITATING TURBULENT FLOW IN
FRANCIS TURBINE, Proceedings of FEDSM2005, 2005 ASME Fluids Engineering Division Summer Meeting and Exhibition,
June 19-23, 2005, Houston, TX, USA
[8] Shuhong Liu, Liang Zhang, Michihiro Nishi, Yulin Wu. Cavitating Turbulent Flow Simulation in a Francis Turbine Based
on Mixture Model, Journal of Fluids Engineering, MAY 2009, Vol. 131, 051302-1-8
[9]Romeo SUSAN-RESIGA, Sebastian MUNTEAN,Ioan ANTON,Sandor BERNAD. NUMERICAL INVESTIGATION OF
3D CAVITATING FLOW IN FRANCIS TURBINES, &RQIHUHQFH RQ 0RGHOOLQJ )OXLG )ORZ &0)) The 12thInternational
Conference on Fluid Flow Technologies, Budapest, Hungary, September 3-6, 2003
[10]Lingjiu Zhou, Zhengwei Wang, Yongyao Luo, Guangjie Peng. Numerical Simulation of Cavitating Flow in a Francis
Turbine, Proceedings of FEDSM2008, 2008 ASME Fluids Engineering Conference, August 10-14, 2008 Jacksonville, Florida USA
>@<XOLQ :X  6KXKRQJ /LX  +XD-6KX 'RX /LDQJ =KDQJ 6LPXODWLRQV RI XQVWHDG\ FDYLWDWLQJ WXUEXOHQW IORZ LQ D )UDQFLV
turbine using the RANS method and the improved mixture model of two-phase flows, Engineering with Computers, (2011) 27:235
250
[12]Hakan NILSSON. Evaluation of OpenFOAM for CFD of turbulent flow in water turbines,23 rdIAHR
symposium,Yakohama,October 2006
[13] O Petit, A I Bosioc, H Nilsson, S Muntean , R F Susan-Resiga. A swirl generator case study for OpenFOAM.
http://iopscience.iop.org/1755-1315/12/1/012056
[14] Rickard E. Bensow and Goran Bark. SIMULATING CAVITATING FLOWS WITH LES IN OPENFOAM, V European
Conference on Computational Fluid Dynamics, ECCOMAS CFD 2010, J. C. F. Pereira and A. Sequeira (Eds), Lisbon, Portugal,1417 June 2010
[15] Robert F. Kunz, David A. Boger, David R. Stinebring at al.. A preconditioned NavierStokes method for two-phase flows
with application to cavitation prediction, Computers & Fluids, 29(2000), 849-875
[16] Openfoam web site. http://www.openfoam.com
[17] DRAGICA JOsT,ANDREJ LIPEJ. NUMERICAL PREDICTION OF THE VORTEX ROPE IN THE DRAFT TUBE, 3rd
IAHR International Meeting of the Workgroup on Cavitation and Dynamic Problems in Hydraulic Machinery and Systems, October
14-16, 2009, Brno, Czech Republic

165