IOP Conference Series: Materials

Science and Engineering

PAPER • OPEN ACCESS You may also like

- Intermittent cavitating flow field in tandem
Modal analysis of Kaplan turbine in Haditha cascades of Kaplan turbines
R F Susan-Resiga, T Ciocan, A Stuparu et
hydropower plant using ANSYS and SolidWorks al.

- Numerical study on Unstable Hydraulic

Factors of Kaplan turbine with semi-spiral
To cite this article: Abdulmuttalib A. Muhsen et al 2021 IOP Conf. Ser.: Mater. Sci. Eng. 1105 012056 case at large flow rate conditions
Yaping Zhao, Zhihua Li, Mengfan Dang et
al.

- Text Categorization on Hadith Sahih Al-

Bukhari using Random Forest
View the article online for updates and enhancements. Muhammad Fauzan Afianto, Adiwijaya and
Said Al-Faraby

This content was downloaded from IP address 117.198.141.197 on 25/09/2024 at 08:15

PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

Modal analysis of Kaplan turbine in Haditha hydropower

plant using ANSYS and SolidWorks

Abdulmuttalib A. MUHSEN1,2, Ali A. Resan AL-MALIK3, Bashar H. ATTIYA2,

Omar Farhan AL-HARDANEE4 and Karam Aied ABDALAZIZE2
1
Faculty of civil and Transport Engineering, Poznan University of Technology, Poznan, Poland
2
Haditha Hydropower Station, Ministry of Electricity, Haditha, Iraq
3
College of Engineering, Misan University, Misan, Iraq
4
College of Engineering, Ankara Yıldırım Beyazıt University, Ankara, Turkey

Abstract. In this study, numerical analysis is conducted to investigate the failure modes in
Kaplan turbine. All necessary steps for Kaplan turbine failure analysis are presented in this
work using the modal analysis computational approach. The modal behaver analysis is carried
out on a model of an existing Kaplan turbine blade, which is based on the existing turbine used
in Haditha hydropower plant in Iraq. This work investigates the modal behavior of the blade of
interest, which aid in predicting structural damage initiation. The Kaplan turbine blade is
designed using the commercial software ANSYS and SolidWorks. To simulate the blade in
operation, the blade is fixed from one end, and all degrees of freedom are measured. Moreover,
the turbine blade is moved and rotated to simulate multiple operational conditions. Both mode
shapes and natural frequencies are predicted and analyzed using the two aforementioned
commercial software and the numerical formula involving the arrest Lanczos method. It is
clear from the results that the natural frequency of the specified mode shape does not match
with the natural frequency of the runner blade. Hence, there is no failure due to resonance
phenomenon in this specific Kaplan turbine. The future work must investigate other aspect of
the failure modes in such turbine, such as unbalance dynamic loading. The Results obtained
from this study will help study the different possibilities for detecting the failure of the Kaplan
blade by examining the modal behavior of the blade.

Keywords: Kaplan turbine; Modal analysis; Failure analysis; ANSYS; Solid Works.

1 Introduction:
Vibrations induced by transient flow behaviors and its adverse effects are critical problem in
hydropower plants. This kind of adverse fluctuations can affect the structural integrity of the hydro
unit in terms of the components fatigue as well as the unit stability and performance. As soon as, the
operating frequencies match with the unit resonance frequency, the whole unit operational mode
would become unstable, and one example of such operational issues is seen in Haditha hydropower
plant. Haditha plant is the second largest hydropower plant in Iraq. The plant is located at Euphrates
river in the west part of the country, which contains six vertical types Kaplan turbines with a 110
MWe per turbine. Haditha plant operating conditions is highly fluctuated due to the fact that it is the
only fully operated plant in the west region of Iraq. This fact forces the plant to operate at highly
fluctuated and wide operational regimes, which in most cases lies out of the optimum operational

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

zones, to provide continues power supply to the customers. Severe cavitation which results in high
structural vibration levels because of the high dynamic loading, and runner blade erosion damages
have been occurred due to these operating conditions. Moreover, structural cracks have been observed
at the runner blades, and these issues increases with the continues off-design operation of the units.
The aforesaid operational demands made several companies to be forces to operate at low water levels
which is associated with low efficiency, making the turbine suffer from high dynamic loading and
severe cavitation erosion [1].
Operating units at extreme off-design conditions cause blade damages, wear in various turbine parts,
and eventually fatigue damage. Figure 1 shows an example in the structural damages that Haditha
plant turbines suffered from. Kaplan turbines are widely used in low to medium water heads with high
capability to produce power. This is due to the high efficiency in their entire operational range [2,3],
and the reason comes from the ability of such turbine to change the angle of blades to adapt to the
flow and head requirements to maintain high efficiency. Nonetheless, some failure cases have been
reported in Kaplan turbines [4-6]. Several studies [7,8] have shown that the water transient behavior
can causes high dynamic loading which might induce high-stress concentration at the blade's root area,
hence give high potential to cause cracks to occur there. Moreover, if the metal structure has an
accumulated problem in this area, the degradation process can be substantially accelerated.
Frunzaverde et al. [4] reported a situation where a crack showed up at the advanced side origin
location of the turbine blade due to high-stress concentrations.
These large fractures usually stem from micro fractures, which typically are undetected by the present
non-destructive techniques and condition monitoring systems, and they will continually expand.
Otherwise detected in time, sever structural failure in the turbine might occur. In the early studies on
this topic, cracks were typically assumed to be absolutely open, like a narrow vacant space [9,10]
However, it was demonstrated that the natural frequency decrease because of a practical split is always
smaller than the one forecasted via making use of an absolutely open fracture [9,11].
This is brought on by the nonlinear behavior induced by the contact between the two fracture faces
during the resonance cycle. One of the most significant impacts of the nonlinear practices is that the
fractured framework's robust regularities shift substantially towards higher frequency. Bovsunovsky et
al. [9] were among the first writers to acquire an estimated frequency solution of a beam with a crack
considering the nonlinear aspects. According to whether the nonlinear behavior is thought about, the
fracture simulation approaches are typically split right into two categories: direct method and a
nonlinear process. For the immediate process, the crack is assumed to be as a slim void or low
stiffness material. The force acting on the contact pair along with the regular at a discrete time
immediately is calculated for the nonlinear approach.
In this paper, the modal analysis of one blade of a Kaplan turbine ( see figure 2) is conducted using
ANSYS software, and then compared with the modal analysis of the same turbine's blade by using
SolidWorks software. The aim is to predict the behavior of the blade under the influence of different
outside conditions and predict the failure mode. A Mathematical Modal Analysis is carried out for the
blade by using the two software. Numerical outcomes have been contrasted as well as talked about.
Results obtained allow analyzing the possibilities of predicting this type of crack by examining Kaplan
turbine blades' modal behavior.

Figure 1. Damaged runner blades.

2
PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

Figure 2. Runner blade - Kaplan turbine.

2 Numerical Approach:
Every structure can be understood as a superposition of masses, dampers and spring attached in
between them. The matrix form of the structural formula of a structure based on a dynamic load [12] is
the following (equation (1)):

, - ( ) , - ( ) , - ( ) ( ) where * + (1)

( ) ( ) ( ) stands for displacement, velocity, and acceleration in the n points (degrees of

freedom (DOF)) within the time domain. The [M], [C] and [K] are, respectively, the matrices of mass,
dampers, and springs. These parameters represent the relation between the various DOFs within the
predefined structure. F(t) is a vector that stand for the force applied in each DOF in the time domain.
By the implementation of the Fourier Transform (FT) to equation (1), the Frequency Response
Function (FRF) is given by:

* ( )+ , ( )- * ( )+ (2)

here {X (jω)} and {F (jω)} stands for the equivalent vectors x(t) and F(t) within the frequency
domain . As a result, [H (jω)] represents the FRF. This FRF can be also expressed as shown below:

* + * +
, ( )- ∑ (3)
( )

In the above equation, N stands for the number of the different vibration modes, and from the
above equation the modal parameters can be computed. Those parameters include natural frequency
(ωr), the damping factor (θr) which governs the amplitude of the structural vibration in the natural
frequency, the mode shapes ({ϑ}r) which describe the deformation profile and finally, the scaling
factor (Qr) which represents a fixed factor for every mode r. Moreover, the natural frequency (ωr)
depends on both modal stiffness (Kr) and also modal mass (Mr) of every mode-shape r, which is
depicted in equation (4), as shown below:

√ (4)

In the case of Experimental Modal Analysis (EMA), the two vectors {X(jω)} and {F(jω)} are acquired
at the same time with multiple transducers, which is typically accelerometers for the response part, and
impact hammers to take care of the force. Hence, the FRF can be acquired and thus, consequently the

3
PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

modal parameters can be extracted [13]. The experimental part is out of the scope of the current study;
however, it will be included in the future work to expand the outcomes of this work.

3 Numerical simulations (Results and discussion)

Numerical approaches based on finite element method modeling has been used in this study to analyze
the Kaplan runner blade deformation modes. In this part, a description is given for the necessary steps
implemented in the modal analysis simulation. The modal analysis is carried out on the Kaplan turbine
blade (see figure 2). The Kaplan turbine blade was fixed from one end and all degrees of freedom on
this end were taken, meanwhile the runner blade was allowed to move and rotate to simulate actual
runner blade operational environment. Mode shapes and natural frequencies are computed in ANSYS
and SolidWorks with the numerical formulation of the direct solver.
3.1 Calculation the corresponding working frequency:
The presence of the fluid exerted significant effects on the natural frequencies of the blade working
inside the path. Rotating frequency of the blade [14-16] is given by equation (5):

(5)

For Haditha turbine case, the rated speed n is 100 rpm, the number of the blade (Z1) is six, and the
number of the wicket gate (Z2) is twenty four, the frequency of blade in the operating flow path would
be calculated as follows:
Blade frequency: f1= f ×Z1=100/60 ×6=10 Hz Wicket gate frequency: f 2= f × Z2=100/60 × 24=40 Hz
Haditha turbine operates at 100 RPM, so the corresponding working frequency will be 10 Hz. In all
simulations the blade material properties were taken from the original supplier of Haditha plant
turbine. The runner blade is made of EN 1.4313 stainless steel. The mechanical properties of this type
of stainless steel is tabulated in table1. The vanes are manufactured from-cast stainless steel containing
13 percent of chromium and 6 percent of nickel. They may resist abrasion and mechanical and
dynamic straining under any operational conditions.
EN 1.4313 hardened steel is martensitic tempered steel figured for essential shaping into fashioned
items. 1.4313 is the EN numeric assignment for this material. X3CrNiMo13-4 is the EN compound
designation. It has a modestly high base expense among the created martensitic hardened steels.
The synthesis of EN 1.4313 treated steel is striking for containing nearly high nickel (Ni) measures
and chromium (Cr). Nickel is principally used to accomplish a particular microstructure. Furthermore,
it beneficially affects mechanical properties and specific kinds of consumption. Chromium is the
characterizing alloying component of treated steel. Higher chromium content gives extra consumption
obstruction. [17].

Table 1. Mechanical properties

Elastic (Young's, Tensile) Modulus 200 GPa, 28 x 106 psi
Fatigue Strength 340 to 510 MPa, 50 to 73 x 103 psi
Poisson's Ratio 0.28
Shear Strength 460 to 600 MPa, 66 to 87 x 103 psi
Tensile Strength: Yield (Proof) 580 to 910 MPa, 84 to 130 x 103 psi
Elongation at Break 12 to 17 %
Impact Strength: V-Notched Charpy 55 to 70 J, 41 to 51 ft-lb
Shear Modulus 76 GPa, 11 x 106 psi
Tensile Strength: Ultimate (UTS) 750 to 1000 MPa, 110 to 150 x 103 psi

3.2 Modal analysis results with ANSYS:

The modal analysis for the hydro turbine blade is executed by ANSYS Workbench (see figure 3).
Modal analysis is a technique to study the dynamic characteristics of a structure under vibrational

4
PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

excitation. Natural frequencies, mode shapes, and mode vectors of a system can be determined using
modal analysis. A graphical variation of the number of modes vs. the frequency can also be obtained
from ANSYS Workbench. The mode shapes are observed, which provide a comprehensive picture of
deformations occurring. The material properties are assigned to the blade and the boundary conditions
are defined (see figure 4), the blade is fixed from one end using a fixed support (blue part in figure 4).
This condition prevents the movement of the surface in space. The mesh structure for the is generated
automatically using ANSYS built-in meshing package, with a maximum size of the element of 5 mm.
The mesh density of 524806 elements was used for this study due to computational limitation. The
mesh structure is depicted in figure 5.

Figure 3. Graphical environment of ANSYS Workbench.

Figure 4. Boundary conditions. Figure 5. Mesh structural details

Figure 6 and 2able 2 shows the simulation results that were predicted using ANSYS software. The
figure depicts the six modal shapes that corresponds to six natural frequencies. The six natural
frequencies of the runner blade are detected between 189.1 and 584.1Hz (see table 2). The mode-
shapes associated to every of those natural frequencies are shown in figure 6. In general, the results
show a trend of increasing deformation with higher frequencies and a higher distribution in the zone of
the tip-lip of the runner blade. The results indicate higher deformation at the fourth mode (figure 6d)
and the sixth mode (figure 6 f). However, the results indicate that mode-shapes does not change
considerably for the first three modes. It is worth to mention that the influence of the computational
time of the analysis is crucial when a range of frequencies or number of mode shapes is specified.
Also, the type of the solver will affect the predicted results.

5
PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

Figure 6. Runner blade mode shape at natural frequency: (a) 189.1 Hz, (b) 248.06 Hz , (c) 332.37 Hz ,
(d) 434.81 Hz, (e) 472.87 Hz, (f) 584.1 Hz. ANSYS data.

Table 2.The first six natural frequencies (ANSYS).

6
PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

3.3 Modal analysis results with SolidWorks:

The modal analysis for the Haditha turbine blade is furthermore performed by SolidWorks (see figure
7). The aim here is to compare the two software results and what kind of discrepancies will be
detected. Solidworks simulation is a multi-discipline computer-Aided Engineering (CAE) tool that
enables users to simulate the physical behavior and allows users to improve the design. Solidworks
simulation can predict how a system will behave in the real world by calculating stresses, deflections,
frequencies, heat transfer paths, etc. The material properties are assigned to the blade and boundary
conditions are defined.
The fixed end boundary condition is shown in blue (see table 3). This condition prevents the
movement of the surface in space. The meshing process is done by the SolidWorks and the mesh on
the blade surface is generated automatically. It is noted that the mesh size in this case is lower than the
case of ANSYS software (see table 4).

Figure 7. Graphical environment of SolidWorks.

Table.3. Boundary condition.

7
PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

Table 4. SolidWorks mesh parameters.

Figure 8. SolidWorks mesh.

Figure 9 and table 5 shows the simulation results that were predicted using SolidWorks software.
Again, the modal analysis is carried out by SolidWorks and mode shapes and natural frequencies are
also predicted similar to what conducted in ANSYS. The first six mode shapes are shown with the
corresponding six natural frequencies as shown in the table 5. The results show similar trends as the
ones shown in the ANSYS results (see figure 6), and generally the deformation mode shows higher
intensity with larger distribution on the blade surface as the natural frequencies of the modes
increases. Higher modal intensity shapes are associated with the higher frequencies. Results obtained
from SolidWorks modal analysis shows a higher deformation mode than the one predicted by ANSYS
( for example, see figures 9 d and f) and figures 6 d andf. This might be due to the fact that the used
mesh within Solidworks is very coarse mesh when it compared to the mesh used in ANSYS. This will
have its effect on the accuracy of the predicted modal shapes.
The modal analysis shows no resonance in any of the six mode shapes. The turbine operates at 100
rpm, so the corresponding working frequency is 10 Hz. Moreover, the natural frequency of all mode
shape does not match with the natural frequency of the runner blade. Thus, no resonance was detected
in the data from the modal analysis.

8
PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

Figure 9. Runner blade mode shape at natural frequency: (a) 190.48 Hz, (b) 251.69 Hz,
(c) 339.62 Hz, (d) 445.45 Hz, (e) 487.23 Hz, (f) 600.71 Hz. SolidWorks data.

Table 5. The first six natural frequencies (SolidWorks).

Mode shape 1 2 3 4 5 6

Frequency (Hz) 190.48 251.69 339.62 445.45 487.23 600.71

9
PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

4 Conclusion:
In this study, the modal behavior of Haditha hydropower plant Kaplan turbine has been
investigated. To conduct this investigation , two different commercial software were used (ANSYS &
SolidWorks) to conduct the numerical simulations with a blade model mimicking the original runner
blade at Haditha plant in Iraq. The aim was to give an insight and understand the unusual failure
modes that occurred in Haditha plant Kaplan turbine. Usually, Kaplan turbine blades are prone to have
fatigue issues due to high hydraulic loads that is the results of the wide operational modes associated
with this type of hydro turbines. The blade root zone is normally associated with high-stress levels and
it is easy for the cracks to initiated from it. In the case of Haditha plant , a big fracture occurred on one
blade starting from the leading-edge side root hole. It is not clearly understood why this type of failure
occurred in the plant, and this study tries to shed some light on this subject by investigating the modal
behavior to see if any resonance might happen during operation of the unit.
In both programs, the modal analysis of the hydro turbine blade was executed, and the first six
mode shapes and its corresponding natural frequencies were computed. Mode shapes behavior for
both software results is very close in the general behavior. Higher natural frequencies were associated
with higher deformation shapes which is related with the modal shapes at each frequency. Natural
frequencies are very close in the case of the two numerical results for both programs. The deviation in
the results between the two software might be related to the nature of the mesh used in each one of
them. Higher mesh density was used in the ANSYS simulation compared to lower one in the case of
SolidWorks analysis. The numerical investigations show no sign of resonance in any of the six mode
shapes and the associated frequencies. The natural frequency of all mode shape does not match with
the natural frequency of the Kaplan runner blade. Hence, we conclude that failure mode due to such
phenomena is not possible according to the available data. Moreover, the blade acts as a fixed blade
during the modal analysis, which in some cases were associated with high deformation at the blade tip.
However, the results shown that it was within safe limits at the edges of the runner blade for all mode
shapes.
In the future work, the authors will conduct an experimental and numerical analysis to investigate the
effect of an existing crack on the natural frequencies and mode-shapes between a damaged blade and
undamaged one. The blade with the fatigue crack is currently in Haditha plant and can be used to
conduct the experimental part. The modal shapes and natural frequencies will be quantified and
analyzed to evaluate the differences between the two different blades.
References:
[1] Muhsen, A. A., Szymański, G. M., Mankhi, T. A., & Attiya, B. (2018). Selecting the most
efficient maintenance approach using AHP multiple criteria decisions making at Haditha
hydropower plant. Zeszyty Naukowe Politechniki Poznańskiej. Organizacja i Zarządzanie.
[2] B.G. Mulu, P.P. Jonsson, M.J. Cervantes, 2012, Experimental investigation of a Kaplan
draft tube–Part I: best efficiency point, Appl. Energy. 93 (2012) 695–706.
[3] P.P. Jonsson, B.G. Mulu, M.J. Cervantes, Experimental investigation of a Kaplan draft
tube–Part II: off-design conditions, Appl. Energy 94 71–83.
[4] D. Frunzaverde, V. Campian, D. Nedelcu, G.-R. Gillich, G. Marginean, 2010, Failure
analysis of a Kaplan turbine runner blade by metallographic and numerical methods, in:
Proc. 7th WSEAS Int. Conf. FLUID Mech. (FLUIDS’10), Univ. Cambridge, UK, , pp. 60–67.
[5] Y. Luo, Z. Wang, J. Zeng, J. Lin, 2010, Fatigue of piston rod caused by unsteady,
unbalanced, unsynchronized blade torques in a Kaplan turbine, Eng. Fail. Anal. 17, 192–199.
[6] G. Urquiza, J.C. Garcia, J.G. Gonzalez, L. Castro, J.A. Rodriguez, M.A. 2014, Basurto-
Pensado, O.F. Mendoza, Failure analysis of a hydraulic Kaplan turbine shaft, Eng. Fail. Anal.
41, 108–117.
[7] L. Zhou, Z. Wang, R. Xiao, Y. Luo, Analysis of dynamic stresses in Kaplan turbine
blades, Eng. Comput. 24 (2007) 753–762.

10
PEC 2020 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1105 (2021) 012056 doi:10.1088/1757-899X/1105/1/012056

[8] Z.W. Wang, Y.Y. Luo, L.J. Zhou, R.F. Xiao, G.J. Peng, 2008, Computation of dynamic
stresses in piston rods caused by unsteady hydraulic loads, Eng. Fail. Anal. 15, 28–37.
[9] G.-L. Qian, S.-N. Gu, J.-S. Jiang, 1990, The dynamic behaviour and crack detection of a
beam with a crack, J. Sound Vib. 138 233–243, https://doi.org/10.
1016/0022-460X(90)90540-G.
[10] W.M. Ostachowicz, M. Krawczuk, 1991, Analysis of the effect of cracks on the natural
frequencies of a cantilever beam, J. Sound Vib. 150 191–201.
[11] A.P. Bovsunovsky, V.V. Matveev, 2000, Analytical approach to the determination of
dynamic characteristics of a beam with a closing crack, J. Sound Vib. 235, 415–434.
[12] R.D. Blevins, 2001, Formulas for Natural Frequency and Mode Shape, Krieger,
Publishing Company,.
[13] Zhang, M., Valentin, D., Valero, C., Presas, A., Egusquiza, M., & Egusquiza, E., 2020.
Experimental and numerical investigation on the influence of a large crack on the modal
behaviour of a Kaplan turbine blade. Engineering Failure Analysis, 109, 104389.
[14] Abubakar, M., Badshah, S., Ahmad, T., & Rahman, N. 2014. Modelling and Analysis of
a very Low Head Kaplan Turbine Runner Blades for Rural Area of Punjab. International
Journal of Scientific & Engineering Research, 5(7), 905-912.
[15] Lixia, Z., Wei, Z., & Zhaohong, Y. 2010, Modal analysis of Francis turbine blade based
on fluid-structure interaction theorem. International Journal of Modelling, Identification and
Control, 10(1-2), 101-105.
[16] Lengvarský, P., Bocko, J., & Hagara, M. 2013, Modal analysis of titan cantilever beam
using ANSYS and SolidWorks. American Journal of Mechanical Engineering, 1(7), 271-275.
[17] INGRA Co., 1985, Operation & Maintenance Instructions Manual Book 8..

11