A NOVEL DESIGN OF COMPOSITE WATER TURBINE USING CFD*

WANG Ji-feng, PIECHNA Janusz, MüLLER Norbert

Turbomachinery Laboratory, Department of Mechanical Engineering, Michigan State University, East Lansing, USA, E-mail: jwang94@illinois.edu

A NOVEL DESIGN OF COMPOSITE WATER TURBINE USING CFD*

WANG Ji-feng, PIECHNA Janusz, MüLLER Norbert

Turbomachinery Laboratory, Department of Mechanical Engineering, Michigan State University, East Lansing, USA, E-mail: jwang94@illinois.edu

(Received May 22, 2011, Revised September 15, 2011)

This paper presents computational investigation of a novel design of composite material axial water turbine using Computational Fluid Dynamics (CFD). Based on three-dimensional numerical flow analysis, the flow characteristics through the water turbine with nozzle, wheel and diffuser are predicted. The extract power and torque of a composite water turbine at different rotating speeds were calculated and analyzed for a specific flow speed. The simulation results show that using nozzle and diffuser can increase the pressure drop across the turbine and extract more power from available water energy. These results provide a fundamental understanding of the composite water turbine, and this design and analysis method is used in the design process.

water turbine, composite material, Computational Fluid Dynamics (CFD), extracted power

Introduction

Hydropower, the energy from moving water, is one of the oldest renewable energy sources and the total global electric power capacity of hydropower, including large hydropower, small hydropower, and ocean power, was approximately 820 GW in 2005, which accounted for almost 20% of the renewable energies[1]. In 2003, the world- first commercial-scale marine current turbine with a 300 kW rated power was successfully installed by Marine Current Turbines (MCT) Ltd and IT-Power[2]. Myers and Bahaj[3]tested a 0.4 m diamater horizontal axis marine current turbine in a circulating water channel and measured the power output over a range of flow speeds, and the results were comparable to previous studies. Canadian National Research Council Hydraulics, Energy, Mines and Resources designed and tested a straight bladed darrieus rotor set with an installed capacity of 5 kW[4]. However, sub-marine structures have to withstand the notoriously aggressive marine environment with its corrosive salt water, fouling growth and abrasive suspended particles. Traditionally, steel has been used to produce marine rotors with the stiffness required to combat yielding, however, it is very expensive to achieve the necessary compound-curved profile in steel. In addition, steel is heavy, prone to fatigue and susceptible to corrosion induced by salt water. These disadvantages prompted a decision to adopt composites instead[5]. Advanced composite materials are broadly used in water turbine application because of their high strength to weight ratios and high corrosion resistance, which are expected to be the key to the success for these devices to operate in the harsh marine environment. A novel manufacturing approach similar to filament winding is able to produce the automated composite wheels in different possible patterns in Turbomachinery Lab at Michigan State University (MSU) (Fig.1)[6].

Fig.1 Different patterns of wound wheels

The advantage of using filament winding method to manufacture high performance and light-weight composite wheels is that the production can be rapid, inexpensive and utilize commercially available winding machines[7]. Another important advantage of using this novel pattern of wheel is that by using integrated magnetic rotor it will achieve a minimal impact on underwater life forms. Most marine creatures will be able to pass through the center of the wheel avoiding the blades. To extract more water power, we designed a ducted turbine and performed the simulation. There are some major advantages by using diffuser-augmented marine current turbines[8]: (1) A diffuser placed behind the wheel can reduce the downstream pressure and draw in more flow for a given sized turbine. (2) A large inflow area containing a large amount of energy is concentrated into a smaller area so that a smaller turbine can be used for a given power output. (3) The duct eliminates tip losses on axial flow turbine blades and improves its efficiency.

To understand the hydrodynamic performance of this totally untypical wheel as a water turbine, as well as the influence of the wheel’s rotating speed on the power output, numerical approach was performed using Computational Fluid Dynamics (CFD). The extracted power and torque were calculated at different rotating speeds in a specific flow speed. Static pressure distribution on computational domain and turbine was given to evaluate the turbine’s characteristics.

Fig.2 A multiple axis winding machine and a schematic representation of fiber wetting and wounding

1. Novel composite material water turbine

1.1 Novel concept of the turbine

The invention at MSU can be realized by a lowcost flexible and fully automatic manufacturing method using commercially available CNC machines shown in Fig.2 for filament winding of the turbo-compressor wheels.

These machines can be used for rapid-prototyping and mass-production. They integrate conveniently into CAD/CAM systems. During or after the winding process, the support structure can either be removed or remain in the wheel as a structural element, especially if the support is of a magnetic material and used as an electromagnetic element of an integrated motor or bearing. In this and other ways, motor and bearing elements can be integrated during the winding -all in one production step. Depending on the size and sophistication, one wheel may cost even less than ten dollars.

Fig.3 Free-stream flow direction

1.2 Working condition of the turbine

In this work the turbine was simulated in a free stream velocity with 5 m/s which was chosen according to the near future test condition. The turbine will be tested by mounting it in a moving carriage and driving it at a steady speed, in still water. This is equivalent to mounting the turbine under a fixed pontoon in moving water. The rotating speed of the turbine can be changed corresponding to the different extracted torques and powers, and the optimal rotating speed will be chosen based on the maximum power generation. The free-stream flow direction is from nozzle to diffuser as shown in Fig.3.

Table 1 Main parameters of the modeled turbine

1.3 Turbine design

The turbine wheel parameters were constrained by the dimensions of the water tow tank where the testing will be preformed in the near future. The parameters of modeled turbine in this study are shown in Table 1.

Fig.4 Composite material wheel

1.3.1 Influence of the blade number

During the performance analysis of a water turbine, blade number is one of important design parameters. If the blade number is too great, the crowding out effect phenomenon at the turbine is serious and the increase of interface between free stream and blade will cause the increment of hydraulic loss, while if the blade number is too small, the diffuser loss will increase with the growth of diffuse extent of flow passage[9]. Of the many different patterns shown in Fig.1 designed by Eyler[10], one pattern with the greatest potential of structure stability and fluid dynamic performance has been selected for the marine current turbine shown in Fig.4[11]. To achieve a minimal impact on underwater life forms, without using an external shaft, this wheel can be actuated from a magnetic force between poles in the outer shroud and coil poles on a stationary device surrounding the wheel and housing[7]. The selected wheel pattern with 8 blades shown in Fig.4 gives most uniform distribution of free space for marine creatures to go through its central area. In our design process, we want to take into account not only the power generation, but also the environmental influence and the totally untypical wheel pattern with 8 blades gives us the most promising solution.

1.3.2 Influence of the blade angles

The analysis of influence of blade angles on extracted power has been made in a wide range of angles (from 25oto 60oas shown in Fig.5) to find the relation between the power generation and blade angles. Taking into account of safety factors, we have limited the wheel’s tip speed to 5 m/s. Because of this, we have chosen the rotating speed 20 RPM as the closest speed to this limit. Since we want to connect the water turbine generator to electric net which requires a constant rotational speed of turbine wheel, we concentrated our investigation on the water stream velocity as 5 m/s, which is achieved on the testing frame in the near future water tow tank, and rotational speed of turbine wheel as 20 RPM (electric output frequency is 60 Hz which is achieved by the gear between the generator and turbine wheel).

Fig.5 Normalized extracted power variation with rotating speed for different blade angles

Fig.6 Modeling of wound turbine wheel

2. Numerical simulation of the turbine flow

2.1 Geometry modeling and grid preprocessing

In this study, the preprocess including model and non-structure mesh generation was completed in NX and GAMBIT. The modeling of a wound turbine wheel is shown in Fig.6.

In order to accurately simulate the flow in a turbine passage, further mesh refinement of the turbine is required. The meshing sizes in the turbine are strictly controlled and particular refinements have been made. In all CFD simulations, a mesh dependence test is important in order to check the convergence of the computationl results with respect to spatial resolution[12]. The mesh dependence test is performed by refining the mesh to its final configuration shown in Fig.7 that has been selected for the analysis.

2.2 Governing equations

CFD is fundamentally based on the governing equations of fluid dynamics. They represent mathematical statements of the conservation laws of physics,where the following physical laws are adopted:

(1) Mass is conserved for the incompressible fluid at steady state as

(2) For the fluid analysis of the entire turbine at steady state, the Navier-Stokers equations are used in the following form

Fig.7 Meshing of computational domain

2.3 Boundary conditions and solving procedures

Since the working fluid is ocean water, the CFD simulations were calculated based on assumptions as follows: (1) incompressible flow, (2) steady state flow, (3) smooth walls with no-slip boundary condition. Boundary conditions were set as follows: velocity at inlet and pressure at outlet, interfaces between nozzle, wheel and diffuser, and symmetry boundary conditions for the side walls[13]. In Fluent, no defining surfaces are considered walls. Since the Navier-Stokes equations are solved inside the domain, no-slip boundary condition is applied to all walls in the domain[14]. Therefore, at all surfaces the flow velocities in three directions u = v = w = 0 .

Modeling the wheel rotation is complex as the relative motion between the rotating wheel and the stationary fluid zone causes a cyclic variation of the solution domain. Sliding Mesh (SM) is one of commonly used modeling methods. By using the SM model, the wheel region can slide relative to the other regions in discrete time steps and using implicit/explicit interpolation of data at successive time-steps[15,16]. In this paper, the SM model was used to obtain more accurate results of the actual phenomena of the wheel rotation.

Reliable results can be given only by a well converged, posed and grid independent simulation. Convergence is determined by the order of magnitude of the residues. It should be noted that convergence criteria must assure that the results do not change as the iterations proceed. In this study, the mass flow rate and momentum change for the convergence tolerance 10–4were monitored, and when they stay at a certain number and do not change as the iterations continue, it can be stated that the solution has converged.

Fig.8 Static pressure distribution on turbine

Fig.9 Static pressure distribution in the computational domain

3. Results and discussion

3.1 Static pressure and velocity distribution

To extract energy from the water, a thrust force T directed upstream must be generated. On this axial marine current turbine, the thrust is obtained by rotating blades, which create a pressure drop across the turbine. The ideal extracted power, which is not what would actually be recovered at the shaft, is the product of the water flow through the turbine and the pressure drop across the turbine[12]. Figure 8 shows the static pressure distribution on the turbine, in which it is seen that the pressure is higher on the turbine front.

Figure 9 shows the static pressure distribution on the whole computational domain, in which it is observed that the pressure drop is increased due to the nozzle and diffuser.

Fig.10 Velocity contours in the computational domain

Fig.11 Velocity vectors in the computational domain (upper) and on turbine (bottom)

Figure 10 shows the velocity contour in the computational domain, in which it is seen that the water velocity is increased just before entering into the turbine due to the nozzle’s effect. Figure 11 shows the velocity vectors in the computational domain and turbine.

Fig.12 Variation of extracted torque with rotating speed of turbine at inlet flow speed of 5 m/s

3.2 Influence of the rotating speed

Figure 12 shows the extracted torque variation in a free stream of water with various hydrodynamic flow conditions. The maximum extracted torque is 190 kNm. For a specific water flow speed, different rotating speeds of the turbine are simulated. From this figure, it can be observed that the torque decreases with rotating speed in a quasi-linear fashion for a specific water flow speed. Torque also decreases with flow speed for a specific rotating speed.

Fig.13 Extracted power variation with rotating speed of turbine at inlet flow speed of 5 m/s

Figure 13 shows the variation of extracted power in a free stream of water with various hydrodynamic flow conditions. The maximum extracted power is 220 kW. This figure indicates that for a specific water flow speed, the extracted power increases with rotating speed until obtaining a maximum power, at which further increases in rotating speed serve to reduce the amount of power extraction. To extract the possible maximum power, the water turbine should be working at a suitable rotating speed. For a specific rotating condition, the power is dependent on the water flowspeed, e.g., it increases with the water flow speed.

4. Conclusion

A novel manufacturing approach similar to filament winding has been developed and is able to produce the composite material water turbine, which have significant advantages over traditional designs. In order to use this totally untypical wheel to succeed in extracting ocean current energy, a CFD simulation using Fluent has been performed. The simulation results show that using nozzle and diffuser can increase the pressure drop and extract more power from available water energy. The torque decreases with rotating speed in a quasi-linear fashion for a specific water flow speed, the extracted power increases with rotating speed until obtaining a maximum power, so the water turbine should be working at a suitable rotating speed for the maximum power output. These results provide a fundamental understanding of the composite water turbine, and this design and analysis method is used to determine the turbine’s performance. The future work is testing the turbine by mounting it in a moving carriage and driving it at a steady speed in still water to simulate the free stream and validate the numerical results.

[1] YUN S., SEUNG H. and KIM J. Optimization of cycloidal water turbine and the performance improvement by individual blade control[J]. Applied Energy, 2009, 86(9): 1532-1540.

[2] PONTA F., JACOVKIS P. Marine-current power generation by diffuser-augmented floating hydro-turbines[J]. Renewable Energy, 2008, 33(4): 665-673.

[3] MYERS L., BAHAJ A. Power output performance characteristics of a horizontal axis marine current turbine[J]. Renewable Energy, 2006, 31(2): 197-208.

[4] PONTA F., DUTT G. An improved vertical-axis water current turbine incorporating a channeling device[J]. Renewable Energy, 2000, 20(2): 223-241.

[5] MARSH G. Tidal turbine harnesses the power of the sea[J]. Renewable Energy, 2004, 48(5): 44-47.

[6] WANG J., LI Q. B. and MüLLER N. Mechanical and optimization analyses for novel wound composite axial impeller[C]. Proc. ASME 2009 International Mechanical Engineering Congress and Exposition. Lake Buena Vista, Florida, USA, 2009, IMECE 2009-12938.

[7] WANG J., MüLLER N. Numerical investigation on composite material marine current turbine using CFD[J]. Central European Journal of Engineering, 2011, 1(4): 334-340.

[8] KIRKE B., LAZAUSKAS L. Variable pitch darrieus water turbines[J]. Journal of Fluid Science and Technology, 2008, 3(3): 430-438.

[9] LIU Houlin, WANG Yong and YUAN Shouqi et al. Effects of blade number on characteristics of centrifugal pumps[J]. Chinese Journal of Mechanical Engineering, 2010, 23(6): 742-747(in Chinese).

[10] EYLER A., MüLLER N. Simulation and production of wound impellers[C]. Proc. ASME 2008 International Mechanical Engineering Congress and Exposition. Boston, USA, 2008, IMECE 2008-51310.

[11] WANG J., VAGANI M. and MüLLER N. Design of composite water turbine in free stream using CFD[C]. Proc. ASME 2010 International Mechanical Engineering Congress and Exposition. Vancouver, Canada, 2010, IMECE 2010-39763.

[12] SAEED R., GALYBIN A. and POPOV V. Modelling of flow-induced stresses in a Francis turbine runner[J]. Advances in Engineering Softwares, 2010, 41(12): 1245-1255.

[13] HONG Fang-wen, Dong Shi-tang. Numerical analysis for circulation distribution of propeller blade[J]. Journal of Hydrodynamics, 2010, 22(4): 488-493.

[14] HONG Fang-wen, Dong Shi-tang. Numerical simulation of the structure of propeller’stip vortex and wake[J]. Journal of Hydrodynamics, 2010, 22 (5 Suppl.): 457-461.

[15] DEGLON D., MEYER C. CFD modeling of stirred tanks: Numerial considerations[J]. Minerals Engineering, 2006, 19(10): 1059-1068.

[16] ZHANG De-sheng, SHI Wei-dong and CHEN Bin et al. Unsteady flow analysis and experimental investigation of axial-flow pump[J]. Journal of Hydrodynamics, 2010, 22(1): 35-43.

10.1016/S1001-6058(11)60213-8

* Biography: WANG Ji-feng (1979-), Male, Ph. D. Candidate

MüLLER Norbert,

E-mail: Mueller@egr.msu.edu

2012,24(1):11-16