fluid structure interaction study in francis turbine
DESCRIPTION
Numerical Simulation of Fluid Structure Interaction in Hydro Turbines by Srinivasa Rao, Project Head, QuEST and Katsumasa Shimmei, Deputy General Manager Hitachi Works HITACHI, Ltd., Power Systems.TRANSCRIPT
Fluid Structure Interaction study inFluid Structure Interaction study in Francis Turbine
Introduction
About QuEST GlobalQuEST Global Engineering is a diversified Engineering Services company, employing 2000 professionals across 12 global delivery centers The company helps customers in Aeroprofessionals across 12 global delivery centers. The company helps customers in Aero engine, Aerospace & Defense, Power Generation, Oil & Gas, and other verticals to cut product development and project costs, shorten lead times, extend capacity and maximize engineering resources availability by providing support across the complete product life cycle from design and modeling through analysis prototyping automation data documentationfrom design and modeling through analysis, prototyping, automation, data documentation, instrumentation and controls, embedded systems development, manufacturing support, vendor management, and in-house precision machining. We leverage our local presence and global reach to support engineering cost reduction initiatives for our customers.
About HitachiHitachi, Ltd., (NYSE: HIT / TSE: 6501), headquartered in Tokyo, Japan, is a leading global diversified product & services company with approximately 400,000 employees worldwide. Th ff id f t d t d i i k t tThe company offers a wide range of systems, products and services in market sectors including information systems, electronic devices, power and industrial systems, consumer products, materials, logistics and financial services.
QuEST Global – Hitachi Ltd. Confidential
Typical Hydropower Plant Setup
Francis Turbine
QuEST Global – Hitachi Ltd. Confidential 3
Basic Concepts
High head turbines are subjected to dynamic forces.
Some significant sources are1. Interaction between runner and guide2. Vortex shedding at the trailing edge of vane3. Cavitation at the trailing edge of the runner vanesg g
Testing of hydraulic turbine are expensive
h d h
http://en.wikipedia.org/wiki/File:Turbine_Francis_Worn.JPG
QuEST Global – Hitachi Ltd. Confidential 4
Is there any way to reduce the testing cause…?
YES, through CFD…
Basic Concepts
Physical Problems
Analytical Solutions Experimental
CFD
What is CFD…?It is the science (art) to replace the partial differential equations (which in this case describe the flow of fluids) with numbers and simple algebraic expressions, and to solve them through iteration in time and/or space to obtain a final numerical description of the total flow field under consideration.
CFD works by dividing the region of interest into a large number of cells or control volumes (the mesh or grid).
It solves the following equations for any problem
QuEST Global – Hitachi Ltd. Confidential 5
Conservation of MassConservation of MomentumConservation of Energy
Basic Concepts
Advantages1. Accuracy2 E f U2. Ease-of-Use3. Speed 4. Powerful Visualizations
Disadvantages or LimitationsDisadvantages or Limitations1. CFD solutions and accuracy rely upon physical models (e.g. turbulence,
compressibility, chemistry, multiphase flow, etc.).2. Solving equations on a computer invariably introduces numerical errors.
• Round-off error • Truncation error
CFD can be used in:1. Conceptual studies of new designs2. Detailed product development3. Troubleshooting4. Redesign
QuEST Global – Hitachi Ltd. Confidential 6
CFD analysis complements testing and experimentation. It reduces the total effort required in the laboratory.
Overview
Hydro turbines are desired to work over a wide range of operating conditions due to load variations.
Fatigue:Fatigue:During dynamic loading, structures fails below its yield strength.
In some cases, when fluid interacts with a solid structure, exerting pressure that may cause deformation in the structure and, thus, alter the flow of the fluid itself.
When the flow induces significant stresses in the solid; however, since the resulting deformation of the solid is small, the flow field is not greatly affected
QuEST Global – Hitachi Ltd. Confidential 7
Background of the study
1. In the present study, Experimental testing of a Francis Turbine had shown that there is a crack developed at the root of the runner blade after some hours of operation, and this was the probably caused due to the runner blade interaction with the guide vanesprobably caused due to the runner blade interaction with the guide vanes.
When looked at this problem, the flow induces stress in the turbine and the deformation is very small and will not alter flow.
2. In an other study, Experimental testing of a Francis Turbine stay vane had shown that there was2. In an other study, Experimental testing of a Francis Turbine stay vane had shown that there was vortex shedding from the vane during operation, and this was the probable cause for the observed structural failure in the vane
In this case, the vortex shedding frequency when match the resonance of the vane, the vane vibrates violently and will alter the flow.
These two problems has to solved using computational methods with different approaches . Such problems can be solved using Fluid Structure Interaction (FSI) approach.
FSI simulations can be broadly categorized as one-way or two-way coupled.
One way FSI analysis (Interaction of runner blades with guide vanes)
Two ways FSI analysis (Vortex shedding)
QuEST Global – Hitachi Ltd. Confidential 8
Case study
1. In Vortex shedding was one of the causes proposed for the failure of the original Tacoma Narrows Bridge (Galloping Gertie) in 1940, but was rejected because the frequency of the vortex shedding did not match that of the bridge The bridge actually failed by aero elastic fluttershedding did not match that of the bridge. The bridge actually failed by aero elastic flutter.
2. A thrill ride "Vertigo" at Cedar Point in Sandusky, Ohio Suffered the fate of vortex shedding during the winter of 2001, one of the three towers collapsed. The ride was closed for the winter at the time.
7th November1940; 11:00 AM
QuEST Global – Hitachi Ltd. Confidential 9
http://www.mae.cornell.edu/images/etc/fluids/structure/tacoma.gif
http://www.cigital.com/justiceleague/wp-content/uploads/2007/07/bridge.gif
1940; 11:00 AM
Contents
1. Executive Summarya Objectivea. Objective
b. Approach
2 Model and Assumptions2. Model and Assumptions
3. Boundary Conditions
4 R lt d Di i4. Results and Discussions
5. Conclusions
QuEST Global – Hitachi Ltd. Confidential 10
Executive Summary
Objective of the study
One Way FSI AnalysisTo develop the numerical methodology of Fluid Structure Interaction to find the intensity of stress levels at root of the Francis Runner at operation condition
Two Way FSI AnalysisTo develop the numerical methodology of Fluid Structure Interaction to evaluate the ability of an unsteady numerical simulation to accurately reproduce the vortex shedding frequency on the
t l f f th tnatural frequency of the stay vane.
QuEST Global – Hitachi Ltd. Confidential 11
Executive Summary
ApproachOne Way FSI AnalysisOne Way FSI Analysis1. Perform the unsteady CFD simulation of Francis Turbine and save the result file
2. Map the results (Pressure) on the structure and perform the FEA simulation
3 Check for the maximum stress in the runner3. Check for the maximum stress in the runner
Two Way FSI Analysis1. 2D Flow field Analysis to Determine Required Mesh Density2. 3D Modal Analysis of the Vane to Determine Natural Frequencies3. 3D Flow field Analysis of the Vane to Determine Span wise Variation in Shedding Frequency4. 2D Flow field Analysis of the Vane to Determine Shedding Frequency at Various Flow rates5 3D FSI Analysis at a velocity where the shedding frequency coincides the natural frequency of5. 3D FSI Analysis at a velocity where the shedding frequency coincides the natural frequency of
the stay vane
QuEST Global – Hitachi Ltd. Confidential 12
Model and Assumptions
One Way FSIA design geometry was used for creating a CFD and FEA modelIt should be noted that the design model does not include the fillets Cavitation effect are neglected do to higher computational time
Draft tube
RUNNERFluidDomain
SolidD i
RUNNER
Domain
QuEST Global – Hitachi Ltd. Confidential 13Geometry Mesh
Guide vaneSolid region
Model and Assumptions
Two Way FSIA design geometry was used for creating a CFD and FEA modelIt should be noted that the design model does not include the fillets Cavitation effect are neglected do to higher computational time
Fl idFluidDomain
FluidMesh
QuEST Global – Hitachi Ltd. Confidential 14
SolidDomain
SolidMesh
Boundary Conditions
Working Fluid : Water
Guide
Runner
D f
UpperCavity
InletGuideVanes
DraftTube
LowerCavity
Outlet
One Way FSI
QuEST Global – Hitachi Ltd. Confidential 15
Turbine Material : Structural Steel
Boundary Conditions
OutletFluid Structure
Interaction
Blade
Frictionless Support(Symmetry)Inlet
QuEST Global – Hitachi Ltd. Confidential 16
Two Way FSI
Results and Discussions
ObservationsOne Way FSI Analysisy y1. Form this analysis, it was seen that the stress were maximum at the root of the runner leading
edge.
2. Form the stress harmonic levels is seen at the root that the cause due the interaction of Guide Vane and Runner bladeVane and Runner blade.
3. This analysis doesn't conclude the failure of the runner at root. Fatigue analysis has to be carried our to confirm.
Variation of Nomalized Von Mises Stress 1 02
30
Hub0.98
1.00
1.02M
ax
Probe at which the stress
30 mm
0.92
0.94
0.96
σ/σ
M
QuEST Global – Hitachi Ltd. Confidential 17
Probe at which the stress values were monitored 0.90
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035
Time (s)
Results and Discussions
One Way FSIThe Peak stresses in the runner were
Normalized Pressure Fluctuation at Trailing Edge of Guide Vane
1 02
1.04
predicted to occur at the LE root near the Hub region.
These unsteady pressure effects are quite considerable (~ 4.5% variation) during each 0.96
0.98
1.00
1.02
Ps/P
s (M
ax)
Normalized Von-Mises Stress
( ) gvane passing. This results in huge variations in stress levels.
0.92
0.94
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08Ti ( )
0.8
1.0
1.2
ess
Limitations
Node to node connectivity of fluid and structural mesh will increases
Time (s)
0.4
0.6
Nor
mili
zed
Stre and structural mesh will increases
the accuracy of the stress prediction.
QuEST Global – Hitachi Ltd. Confidential 18
0.0
0.2
Hub Leading Edge Hub Trailing Edge Shroud Leading Edge Shroud Trailing Edge
Results and Discussions
ObservationsTwo Way FSI Analysisy y1. Analysis was carried our with 3 Mesh configurations 13,000 cells, 16,000 cells and 3,50,000
cells. It was seen that 16,000 and 3,50,000 cell had a close match of results and for further analysis 16,000 cells mesh were used.
Cases Mesh SizeFirst NaturalFrequency
Second NaturalFrequency
1 13,000 91.55 (Hz) 183.1 (Hz)
2 16,000 85.45 (Hz) 170.9 (Hz)
3 3,50,000 85.45 (Hz) 164.8 (Hz), , ( ) ( )
QuEST Global – Hitachi Ltd. Confidential 19
2. It was seen from 3D Modal analysis, the major frequencies (Bending: 73 Hz & Torsional: 155 Hz)
Results and Discussions
ObservationsTwo Way FSI Analysisy y3. From CFD results, FFT calculations for different points were done and when plotted, it was seen
that the frequency had very little effect over the span. This concludes this analysis can be done with 2D to determine the vortex shedding frequency.
Frequency Distribution behind the Trailing Edge of Hydrofoil @ different Span
12
34
56
Edge of Hydrofoil @ different Span
1 5
2.0
2.5
de
Point 1 Point 2 Point 3
Point 4 Point 5 Point 67
89
Symmetry0.5
1.0
1.5
Am
plitu
d
Point 7 Point 8 Point 9
QuEST Global – Hitachi Ltd. Confidential 20
Points just behind the hydrofoil at different spans
0.00 200 400 600 800 1000 1200 1400
Frequency (Hz)
Results and Discussions
ObservationsTwo Way FSI Analysisy y4. For 2D CFD analysis with different inlet velocities,
a. From the FFT analysis it was seen that at inlet velocity of 12m/s, the vortex shedding frequency is 73 Hz
b. This concludes that for inlet velocity of 12m/s, the vortex shedding frequency matches the natural frequency of the model
c. So, FSI analysis can be done with a inlet velocity of 12m/s
Velocity Maximum Frequency
(m/s) (Hz)
11 65
Frequency plot @ R1P1
2
2.5
73
11 65
12 73
13 78
0 5
1
1.5
Am
plitu
de
QuEST Global – Hitachi Ltd. Confidential 21
0
0.5
0 100 200 300 400 500 600 700 800 900
Frequency (Hz)
Results and Discussions
ObservationsTwo Way FSI Analysisy y5. The Stress were monitored at the root of the blade. The values were around 10 times less then
the yield stress of the material.
Von Mises Stress (MPa)Displacement of the Hydrofoil at Trailing Edge in Mid Span
4
5
0
1
2
3
cem
ent (
mm
)
-4
-3
-2
-1
Dis
plac
QuEST Global – Hitachi Ltd. Confidential 22
-50.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10
Time (s)
Conclusions
One Way FSIFrom this analysis, it was seen that the stresses were developed at the root of the blade which developed as a crack of the turbine due to runner interaction with guide vane.
Improvement in the results would have been achieved if the there would have been node to node connectivity and using same geometry (Fluid and Surface in common)
Two Way FSIForm this analysis, it was seen that the vortex shedding frequency derived from flow induced vibrations match the resonance (bending mode) of the vane. For such lockup conditions, Karman vortices exhibit a strong instability and substantial increase in vibration level.
The Stress at the root of the blade were around 10 times less then the yield stress of the material. So, cycle calculation would help in determining the fatigue characteristics and number of cycles the vane can sustain.
Coupled analysis did not show the resonant interaction between fluid and solid that was expected. This is due the oscillations have not reached a constant phase and recommended to run from some more time
QuEST Global – Hitachi Ltd. Confidential 23
THANK YOUTHANK YOU