Professional Documents
Culture Documents
Search
Collections
Journals
About
Contact us
My IOPscience
Parametric design of a Francis turbine runner by means of a three-dimensional inverse design method
This article has been downloaded from IOPscience. Please scroll down to see the full text article. 2010 IOP Conf. Ser.: Earth Environ. Sci. 12 012058 (http://iopscience.iop.org/1755-1315/12/1/012058) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 115.113.144.102 The article was downloaded on 30/06/2012 at 09:57
25th IAHR Symposium on Hydraulic Machinery and Systems IOP Conf. Series: Earth and Environmental Science 12 (2010) 012058
Parametric design of a Francis turbine runner by means of a three-dimensional inverse design method
K Daneshkah1 and M Zangeneh2
1 2
Advanced Design Technology, London WC1E 7JN, UK Department of Mechanical Engineering, University College London, London WC1E 7JE, UK E-mail: k.daneshkhah@adtechnology.co.uk
Abstract. The present paper describes the parametric design of a Francis turbine runner. The runner geometry is parameterized by means of a 3D inverse design method, while CFD analyses were performed to assess the hydrodymanic and suction performance of different design configurations that were investigated. An initial runner design was first generated and used as baseline for parametric study. The effects of several design parameter, namely stacking condition and blade loading was then investigated in order to determine their effect on the suction performance. The use of blade parameterization using the inverse method lead to a major advantage for design of Francis turbine runners, as the three-dimensional blade shape is describe by parameters that closely related to the flow field namely blade loading and stacking condition that have a direct impact on the hydrodynamics of the flow field. On the basis of this study, an optimum configuration was designed which results in a cavitation free flow in the runner, while maintaining a high level of hydraulic efficiency. The paper highlights design guidelines for application of inverse design method to Francis turbine runners. The design guidelines have a general validity and can be used for similar design applications since they are based on flow field analyses and on hydrodynamic design parameters.
1. Introduction
The hydraulic design of Francis turbine runners requires accomplishment of several targets and constraints. A high level of efficiency and a cavitation-free flow in the runner is usually desirable. The flow in Francis turbine runners is highly rotational and three-dimensional and therefore only three-dimensional methods will provide effective solution for a Francis runner. A considerable improvement in the design of Francis turbines have been obtained by the use of Computational Fluid Dynamics (CFD). CFD results provide a better understanding of the flow physics and they are now commonly used in industry, ref [1-4]. Although these methods are very useful for analysis in different design configurations, they cannot be directly used as a design tool as they do not provide any direct information on how to change the runner shape. So the designer needs to rely on trial and error to improve the runner geometry. Such an approach, with its reliance on empiricism, may restrict the part of design space that is being used in the design as the designer tends to stay within the bounds of successful previous designs. A major improvement in the design of Francis runners can be achieved by the application of 3D inverse design method for the design of the runner shapes. Unlike conventional direct design methods, where the blade geometry is described by geometrical parameters, inverse design uses hydrodynamic parameters like the blade loading, to compute the blade shape, offering a major advantage in the design process. Such an approach allows designers to directly relate their understanding of flow physics in the design process and hence access a larger part of the design space. The application of 3D inverse design method has already resulted in important design breakthroughs such as suppression of secondary flows in radial and mixed flow impeller impellers [5-6], improvement of suction performance and efficiency of water jet pumps [7], suppression of corner separation in pump diffusers [8] and improvement of cavitation in a Francis turbine runner [9]. In this present paper, a parametric design study of a Francis turbine runner is carried out where an inverse design method is used to parametrically describe the runner geometry and CFD analyses are performed to evaluate the hydrodynamic and suction performance of different configurations. First, a baseline design was created using the basic design specifications of the Francis turbine runner. Next, the impact of stacking condition on the runner
25th IAHR Symposium on Hydraulic Machinery and Systems IOP Conf. Series: Earth and Environmental Science 12 (2010) 012058
performance was assessed. The aim of this study was to understand the effect of stacking condition of on the runner efficiency and its suction performance. Then, the effect of blade loading was studied for an optimum stacking configuration obtained in the previous step so that a cavitation-free flow in the runner is achieved, while maintaining high level of hydraulic efficiency.
p + p = (2 / B ) Wm
( rV ) m
(1)
Where p+ and p correspond to the static pressure on pressure and suction side of the blade, B is the blade number, is the density and Wm is the pitch-wise averaged meridional velocity. The input design parameters required by the program are as follows: Meridional channel shape in terms of crown, band, leading and trailing edge contours. Normal thickness distribution at two or more spanwise sections. Fluid properties and design specifications. Number of blades. Inlet flow conditions in terms of spanwise distributions of total pressure and velocity components. Inlet and exit rV spanwise distribution. By controlling its value, the runner head is controlled Blade loading distribution (drV/dm) at two or more spanwise sections. The code then automatically interpolates the blade loading in spanwise direction to obtain two-dimensional distribution of the loading over the whole meridional channel. Stacking condition. The stacking condition must be imposed at a chord-wise location between leading and trailing edge. Everywhere else the blade is free to adjust itself according to the loading specifications.
One unique feature of TURBOdesign1 is that it allows designers to vary one parameter (e.g stacking or blade loading) while fixing the other parameters. The program then automatically arrives at the blade shape that satisfies the necessary specific work at the correct flow rate and specified blade loading or stacking. It is this feature of the code that is used in this paper for parametric study. In order to verify the different configurations that were designed, CFD calculations were performed using the commercial software ANSYS CFX 12.1. The computational domain was discretized by means of a hybrid H-C-O type structured mesh with approximately 375K nodes per blade passage. The Reynolds Averaged Navier-Stokes equations were solved using a finite-volume approach and k- model with standard wall function implementation was used for the turbulence closure. The average value of total pressure, which occurs at the runner inlet was imposed as a boundary condition at the inlet of the computational domain. For cavitation analysis, a two phase Rayleigh-Plesset model is used. The interphase transfer is governed by a mixture model where the interface length scale is 1 mm. Flow is assumed to be homogeneous and isothermal at 293.15 K. The saturation pressure is 3619 Pa and the mean nucleation site diameter is 2m.
1/ 2 (2 gH ) 3/ 4
25th IAHR Symposium on Hydraulic Machinery and Systems IOP Conf. Series: Earth and Environmental Science 12 (2010) 012058
The runner meridional geometry is presented in Fig.1. The runner maximum diameter is 157.5 mm and its axial length is 140 mm. The runner meridional shape is usually fixed by design constraints and therefore it was not changed during the design process. The runner has 13 blades with a maximum profile thickness of 7 mm at the crown and 4 mm maximum thickness at the band. The runner operating conditions are listed in Table1. Before proceeding with the parametric study, a baseline design was created using TURBOdesign-1. The design specifications and inlet condition were imposed according to their values at the operating condition. A free-vortex flow distribution (uniform spanwise rV ) was assumed at the runner inlet. The value of rV was chosen to produce the available head at runner inlet. A zero stacking was imposed at runner LE. Table 1. Francis Runner Design Specifications Rotational speed 1350 Runner Head 42 m Design flow rate 0.45 m^3 min-1 Inlet total pressure 415 kPa Guide vane opening 73 deg Required Shaft Power 165 kW
0 Crown Band -1
Blade Loading
-2
-3
-4
0.2
0.4
0.6
0.8
Streamwise Distance
Figures 2 represents the normalized loading distribution of the baseline runner design. The loading is defined at two sections (band and crown) and it is then interpolated over the meridional channel. Each loading distribution is plotted against the normalized streamwise distance from leading edge (streamwise distance=0) to trailing edge (streamwise distance=1). Both sections are mid-loaded with a constant loading from 25% to 75% of blade chord. The value of blade loading at the leading edge controls the flow incidence at design point (see equation [1]). The baseline design runner geometry obtained by the inverse code is presented in Fig.3.
25th IAHR Symposium on Hydraulic Machinery and Systems IOP Conf. Series: Earth and Environmental Science 12 (2010) 012058
K u = U / 2 gH
The hydraulic efficiency is given by:
(3)
T gQH
(4)
Figure 6 shows the velocity vectors on the suction and pressure surfaces on the runner. The flow is roughly aligned with the streamwise direction on the suction side of the blade, whereas near the pressure side inside the boundary layer the flow is forced towards the band, which indicates its strong three-dimensional character and the distinct secondary flows in Francis runner. Figure 7 shows the runner pressure distribution at three spanwise sections, i.e., crown , midspan and band. The low pressure region on the band suctions side indicates that this area is prone to severe cavitation. This is further confirmed by a two-phase flow cavitation analysis, as it can be seen by contours of water vapor volume fraction in Fig.8, confirming strong cavitation on the shroud near the trailing edge region.
25th IAHR Symposium on Hydraulic Machinery and Systems IOP Conf. Series: Earth and Environmental Science 12 (2010) 012058
42
185 180
40
0.98
Head [m]
38
36
34
0.96
0.94
32 0.5
0.6
0.7 Ku
0.8
0.8
0.9
0.92 0.5
0.6
0.7 Ku
0.8
0.9
(a)
(b)
(c)
Figure 5 Runner performance characteristics at design flow rate, Runner Head (a), Shaft Power (b), Runner Efficiency (c)
(a)
(b)
Figure 6 Baseline design: Velocity vector on the blade suction surface (a) and pressure surface (b) at design point
300 250 200 150 100 50 0 -50 -100 Crown Midspan Band
0.2
0.4
0.6
0.8
Streamwise Distance
Figure 8 Baseline design: contours of water vapour volume fraction at design point
25th IAHR Symposium on Hydraulic Machinery and Systems IOP Conf. Series: Earth and Environmental Science 12 (2010) 012058
(a)
(b)
(c)
Figure 9 3D blade geometries at -15 deg (a), -30 deg (b) and -45 deg (c) stacking
300 250 200 150 100 50 0 -50 -100 Crown Midspan Band
300 250 200 150 100 50 0 -50 -100 Crown Midspan Band
300 250 200 150 100 50 0 -50 -100 Crown Midspan Band
0.2
0.4
0.6
0.8
0.2
0.4
0.6
0.8
0.2
0.4
0.6
0.8
Streamwise Distance
Streamwise Distance
Streamwise Distance
(c) (a) (b) Fig. 10 Blade pressure distributions for -15 deg (a), -30 deg (b) and -45 deg (c) stacking design configuration
25th IAHR Symposium on Hydraulic Machinery and Systems IOP Conf. Series: Earth and Environmental Science 12 (2010) 012058
(a)
(b)
(c)
Figure 11 Contours of water vapour volume fraction at -15 deg (a), -30 deg (b) and -45 deg (c) stacking design configuration
-2
-3
-4
Blade Loading
-1 0 0
0 .2
0 .4
0 .6
0 .8
25th IAHR Symposium on Hydraulic Machinery and Systems IOP Conf. Series: Earth and Environmental Science 12 (2010) 012058
(a)
(b)
Figure 15 Design S30_MF: Velocity vector on the blade suction surface (a) and pressure surface (b)
Z
Baseline DesignS30_MF
Baseline DesignS30_MF
Baseline DesignS30_MF
(a)
(b)
(c)
Figure 16 Comparison of baseline and DesignS30_MF blade section geometries at crown (a), midspan (b) and band (c)
25th IAHR Symposium on Hydraulic Machinery and Systems IOP Conf. Series: Earth and Environmental Science 12 (2010) 012058
42
185 180
0.99
40
175 170
0.985
Head [m]
38
Power [kW]
0.98
165 160
0.975
36
155
34
Baseline DesignS30_MF
150 145
Baseline DesignS30_MF
0.97
Baseline D esignS30_M F
32 0.5
0.6
0.7 Ku
0.8
0.9
140 0.5
0.6
0.7 Ku
0.8
0.9
0.965 0.5
0.6
0.7 Ku
0.8
0.9
(a)
(b)
(c)
Figure 17 Comparison of the baseline and DesignS30_MF runner performance characteristics at design flow rate, Runner Head (a), Shaft Power (b), Runner Efficiency (c)
7. Conclusion
In this paper, a 3D inverse design method was applied to a Francis turbine design. Effect of inverse design parameter (stacking condition and blade loading) on the flow field inside the runner was studied in a parametric way. The aim of design was to obtain a cavitation free runner with high hydraulic efficiency. The flow field and suction performance obtained by CFD with single-phase and two-phase flow models were compared between different designs. The effects of stacking condition on the spanwise work distribution and the associated pressure field was studied in details. By a combination of stacking condition and blade loading parameters, the static pressure field inside the runner was optimized so that the low pressure region on the blade suction side was eliminated and a cavitation free runner was realized. It was shown that parameterization of blade geometry using the inverse design flow related parameters can provide the designer with control over the pressure field inside the runner, which can be used effectively to suppress cavitation phenomena without deteriorating the hydraulic efficiency. The design guidelines presented in this paper can be applied easily to the optimization of other Francis turbine runners. The 3D inverse method is an extremely powerful and practical design tool for designing hydraulic turbine runners.
Nomenclature
B H LE Ku m P Q r T Number of blades Runner head [m] Leading edge Non-dimensional blade velocity Merdional distance Static pressure [Pa] Flow rate [m3/s] Radius [m] Torque [N.m] TE U V vs W Trailing edge Blade velocity[m/s] Absolute velocity[m/s] Specific Speed Relative velocity[m/s] Circumferential direction Density [ kg/m3] Rotational Speed [rad/s]
References
[1] [2] [3] [4] Drinta P, Sallaberger M 1999 Hydraulic Turbines- Basic Principle and State-of-the-Art Computation Fluid Dynamics Application Proc. Institute of Mechanical Eng.vol 213 (Part C) pp 85-102 Sallaberger M 1996 Quasi-Three-Dimensional and Three-Dimensional Flow Calculation in a Francis Turbine IGTI (Birmingham) p 96-GT-38 Keck H, Goede E and Pestalozzi J 1990 Experience with 3D Euler Flow Analysis as a Practical Design Tool In Proc. of 16th IAHR Symp.(Sao Paolo, Brazil) Nagafuji T, Uchida K, Tezuka K and Sugama K 1999 Navier Sokes Prediction on Performance of a
25th IAHR Symposium on Hydraulic Machinery and Systems IOP Conf. Series: Earth and Environmental Science 12 (2010) 012058
Francis Turbine with High Specific Speed ASME Fluids Eng. (FEDSM99-7815) Zangeneh M, Goto A and Takemura T 1996 Suppression of Secondary Flows in a Mixed Flow Pump Impeller by Application of Three-Dimensional Inverse Method ASME J. of Turbomachinery 118 536561 [6] Zangeneh M, Goto A and Harada H 1998 On the Design Criteria for Suppression of Secondary Flows in Centrifugal and Mixed Flow Impellers ASME J. of Turbomachinery 120 723-35 [7] Bonaiuti D, Zangeneh M, Aartojarvi R and Eriksson J 2010 A Parametric Design of a Waterjet Pump by Means of Inverse Design, CFD Calculations and Experimental Analyses ASME J. of Fluids Eng.132 031104 [8] Goto A, Zangeneh M 2002 Hydrodynamic Design of Pump Diffuser Using Invese Design Method and CFD ASME J. of Fluids Eng. 124 319- 328 [9] Okomoto H, Goto A 2002 Suppression of Cavitation in a Francis Turbine Runner by Application of 3D Inverse Design Method ASME Fluids Eng. (FEDSM2002-31192) [10] Zangeneh M 1991 A 3D Design Method for Radial and Mixed Flow Turbomachinery Blades Int. J. of Numerical Methods in Fluids 13 599-624 [5]
10