2009

Vertical Wind Energy Engineering Ian Duffett – 009723628 Jeff Perry – 200211837 Blaine Stockwood – 009224597 Jeremy Wiseman - 200336428

DESIGN AND EVALUATION OF TWISTED SAVONIUS WIND TURBINE

Table of Contents 1 PROBLEM DEFINITION ........................................................................................................................... 1

2 BACKGROUND ....................................................................................................................................... 1

2.1 Wind Energy and Wind Power ...................................................................................................... 1

2.2 Vertical Axis Wind Turbines .......................................................................................................... 1

2.3 The Twisted Savonius .................................................................................................................... 2

3 DESIGN .................................................................................................................................................. 2

3.1 Modelling ...................................................................................................................................... 2

3.2 Computational Fluid Dynamics ..................................................................................................... 3

3.2.1 CFD Parameters..................................................................................................................... 3

3.2.2 CFD Results ............................................................................................................................ 3

3.3 Prototype Foil Design .................................................................................................................... 6

4 PROTOTYPE FABRICATION & TESTING .................................................................................................. 7

4.1 Blade Fabrication .......................................................................................................................... 7

4.2 Prototype Setup and Testing ........................................................................................................ 8

4.2.1 Setup ..................................................................................................................................... 8

4.2.2 Testing ................................................................................................................................... 9

1

1 PROBLEM DEFINITION

The twisted savonius concept combines the advantages of a savonius wind turbine with the twisted design of a helical darrieus. While the proposed concept is not experimentally proven, it is theorized that such a turbine will be self-starting and its helical shape will create an even torque distribution throughout a complete revolution. On suggestion of the project supervisor, this project will focus on optimizing this concept by modeling and evaluating the effect of foil design, pitch, rotation and shape of the foil face on performance. The project proposal can be found in Appendix A. Once optimized, a scaled prototype will be fabricated and tested under various wind tunnel conditions to develop a performance matrix.

2 BACKGROUND

2.1 Wind Energy and Wind Power The conversion of wind energy into various other useful forms such as electricity is known as wind power. Large scale wind farms are typically connected to the local electric power transmission network with smaller turbines being used to provide electricity to isolated locations.

Wind energy is an ample and renewable source of green energy. The wide spread distribution suitable wind patterns and the declining cost of producing wind energy makes it a viable energy alternative. The main drawback to wind generated power is that wind is erratic and somewhat unpredictable causing intermittent wind energy input. Wind energy is highly variable making it difficult to manage within commercial electricity generation. Producing irregular wind energy can require energy storage solutions, increasing cost.

Wind energy, as a power source, is favoured by many environmentalists as an alternative to fossil fuels as it is plentiful, renewable, widely distributed, clean, and produces lower greenhouse gas emissions. Although the construction of wind farms is not universally welcomed due to the negative visual impact and the effect on wildlife, it remains one of the largest forms of green energy used in the world today.

2.2 Vertical Axis Wind Turbines

A wind turbine is a rotating machine which converts the kinetic energy of wind into mechanical energy which in turn can be converted into electricity. The main rotor shaft of vertical axis wind turbines are arranged vertically giving them the key advantage of not having to be aligned with the wind. This type of arrangement is highly advantageous on sites where the wind direction is highly variable as VAWTs can utilize wind from varying directions. The generator and gearbox of a VAWT can be placed near or at ground level, eliminating the need to be supported by a tower. This also makes them more accessible for maintenance.

Major concerns of VAWTs are a pulsating torque created by some models and drag forces experienced as the blades rotate into the wind. This pulsating torque is infamous for its negative effect on wildlife and as such, efforts are being made to eliminate the pulsating torque and to develop more wildlife-friendly designs.

2

2.3 The Twisted Savonius Darrieus turbines have good efficiency but produce large torque ripple and cyclic stress on the tower contributing to poor reliability. The blades of a Darrieus turbine can be canted into a helix. This allows the wind to pull each blade around on both the windward and leeward sides of the turbine. As this feature spreads the torque evenly over the entire revolution, it prevents destructive pulsations. Savonius turbines have the advantage of being self starting and are considered more reliable; however they are low efficiency power turbines.

The Twisted Savonius combines the advantages of the savonius turbine with the twisted design of the helical darrieus. The blades, used for converting the power of the wind into torque on a rotating shaft, are uniquely designed to catch the wind from all directions, while the skewed leading edges reduced resistance to rotation. While this design has not been experimentally proven, theoretically it will have the self starting features of the savonius while spreading torque evenly throughout the revolution like the helical darrieus. Analysis into foil design will determine twist angle and pitch for optimal performance.

3 DESIGN The twisted savonius wind turbine design is based on complex fluid-structure interaction. It is in this regard that a computational fluid dynamic (CFD) analysis was undertaken in order to model and simulate the fluid interaction with varying design parameters.

3.1 Modelling Through modelling varying savonius blade designs within Solidworks 2008, Vertical Wave Energy Engineering has developed an optimal prototype concept. The concept has been designed based on changing two independent variables:

• Angle of Twist, α (Figure 5)

• Blade Radius (Figure 6)

Figure 1 - Twisted Savonius

3

Figure 2 - Angle of Twist

Figure 3 - Blade Radius

The importance of foil size is an obvious consideration in developing the prototype. However, the foil size is restricted by the limitations of the rapid prototype machine and the wind tunnel. The prototype can easily be scaled based on swept area to determine the expected energy output of the full size device.

3.2 Computational Fluid Dynamics

3.2.1 CFD Parameters CFD analysis was utilized with the SolidWorks 2008 FloWorks software package. The following parameters were held constant for all prototype design variations:

• Fluid: Air

• X direction velocity: 15m/s

• Y and Z direction velocity: 0m/s

3.2.2 CFD Results CFD analysis of variations of both angle of twist and long and short radius yielded the following results:

Top Plane

Bottom Plane

α

Short Radius, r

Long Radius, R

4

3.2.2.1 Circular Foil Design Varying the angle of twist and measuring the torque generated on a static foil yielded the results shown in Table 1. Torque versus angle of twist is plotted in Graph 1.

circle radius

angle of twist (deg)

torque (N-m) value max avg min

47.8 45 -0.1595 -0.15987 -0.1591 -0.15631 47.8 90 -0.23381 -0.23397 -0.23224 -0.22986 47.8 180 -0.37821 -0.37821 -0.37535 -0.37123 47.8 270 -0.45554 -0.45562 -0.4435 -0.42907 47.8 315 -0.3479 -0.34802 -0.34576 -0.34209 47.8 360 -0.4708 -0.47348 -0.4716 -0.47006 47.8 405 -0.30022 -0.30242 -0.30144 -0.30022 47.8 540 -0.3131 -0.31657 -0.314 -0.3131 47.8 720 -0.2379 -0.23908 -0.23834 -0.23759

Table 1

Graph 1

From this analysis it is clear that maximum torque is achieved at a twist angle of 360°.

00.05

0.10.15

0.20.25

0.30.35

0.40.45

0.5

0 90 180 270 360 450 540 630 720

Torq

ue (N

·m)

Angle of Twist ( )

Circular Foil Design

5

3.2.2.2 Elliptical Foil Design

Holding angle of twist constant at 360°, based on results of Circular Foil CFD, and varying the long radius resulted in the static torque measurements shown in Table 2. Torque versus long radius is plotted in Graph 2.

long radius (mm)

short radius (mm)

angle of twist (deg)

torque (N-m) value max avg min

20 47.8 360 -0.27497 -0.27632 -0.27545 -0.27485 40 47.8 360 -0.40699 -0.40931 -0.40733 -0.4058 80 47.8 360 -0.41232 -0.4168 -0.41354 -0.41199 60 47.8 360 -0.49076 -0.49076 -0.48759 -0.48541

100 47.8 360 -0.51999 -0.5219 -0.518 -0.515 108.3 47.8 360 -0.56037 -0.56549 -0.56242 -0.55956 160 47.8 360 -0.55805 -0.56369 -0.56176 -0.55805

Table 2

Graph 2

From this analysis it is clear that maximum torque is achieved at a long radius of 108.3 mm.

To verify that a twist angle of 360° is optimum for various elliptical designs, the twist angle was also varied at various long radii (refer to Table 3).

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 40 60 80 100 120 140 160

Torq

ue (N

·m)

Long Radius (mm)

Elliptical Foil Design

6

long radius (mm)

short radius (mm)

angle of twist (deg)

torque (N-m) value max avg min

80 47.8 90 -0.04497 -0.07151 -0.04697 -0.03456 80 47.8 180 -0.29442 -0.30212 -0.29507 -0.28883 80 47.8 360 -0.41232 -0.4168 -0.41354 -0.41199 80 47.8 405 -0.20715 -0.21397 -0.20818 -0.20599

108.3 47.8 0 0.425486 0.395606 0.417385 0.451704 108.3 47.8 45 0.3183 0.304751 0.335832 0.35658 108.3 47.8 90 0.208451 0.163101 0.198852 0.232821 108.3 47.8 180 -0.21527 -0.21826 -0.21599 -0.21457 108.3 47.8 270 -0.42199 -0.42884 -0.42261 -0.42068 108.3 47.8 360 -0.56037 -0.56549 -0.56242 -0.55956 108.3 47.8 450 -0.42772 -0.42781 -0.42342 -0.4182

Table 3

The results of Table 3 indicate that a twist angle of 360° results in the greatest static torque generated by the foil.

3.2.2.3 Summary of Results The following conclusions can be drawn based on the CFD analysis conducted:

• A 360° angle of twist results in the greatest static torque for both circular and elliptical foil designs.

• A long radius of 108.3mm at a twist angle of 360°, results in a larger static torque than that generated for a circular design at a 360° twist angle (0.56037 N·m and 0.4708 N·m respectively).

3.3 Prototype Foil Design The prototype design was determined from the findings of a CFD analysis. The prototype foil is currently being fabricated by Memorial University’s rapid prototyping machine. Due to the restrictions of the rapid prototype machine and the wind tunnel, a 22.86cm (9”) by 68.58cm (27”) foil is being fabricated. The rapid prototype machine can build an object of maximum dimensions of 10” by 10”. Therefore, three 9” by 9” foils are being fabricated and once all three foils are completed they will be assembled into a complete unit using ABS solvent cement.

7

4 PROTOTYPE FABRICATION & TESTING

Prototype design and construction is focused on turbine performance, not on electrical generators or support structure. This will allow for improved analysis of the turbine performance without introducing unknown variables such as generator efficiency or line losses.

4.1 Blade Fabrication The twisted savonius blade will be fabricated using the Fused Deposition Modeling (FDM) rapid prototyping machine available at Memorial University of Newfoundland as seen in Figure 4. The blade fabricated will consist of 3 sections: Bottom, Middle, and Top, as shown in Figure 5. This is due to the manufacturing limitations of the rapid prototyping machine Stratasys FDM1650. The rapid prototyper uses FDM to turn computer-aided design (CAD) geometry into solid physical objects, the blade sections of the turbine. The rapid prototyping machine has a build size of 254mm x 241mm x 254 mm (9.5" x 9.5" x 10") with an achievable accuracy of ±.127 mm (±.005 in). Each section has been designed to these build restraints, therefore each section has been designed with a

diameter of 217mm (8.55”) and a height of 217mm (8.55”), once joined together the blade will be 651mm (25.65”).

The PC-ABS thermoplastic material is heated by the FDM head so it comes out in a semi-liquid state. Figure 6, Figure 7, and Figure 8 show the FDM model tray, injection head and PC-ABS material. The successive layers fuse together and solidify to build up an accurate, three-dimensional model of the blade design.

Figure 5- Blade Sections

Figure 4 - Stratasys FDM1650

8

Figure 6 – Model Tray Figure 7 – Injection Head Figure 8 – PC-ABS Material

The Figure 9, Figure 10, and Figure 11 below show the completed bottom section of the blade.

Figure 9 – Blade with Support Figure 10 – Blade Profile Figure 11 – Blade Isometric Material

4.2 Prototype Setup and Testing

4.2.1 Setup The blade section will be tested using Memorial University’s wind tunnel. The wind tunnel is a horizontal open-circuit facility with a rectangular 20.0 x 0.93 x 1.04 meter test section. The turbine will be installed centered both horizontally and vertically within the wind tunnel with both ends of the shaft extending through the bottom and top of the tunnel. The shaft shall have low resistance bearings connected at both ends to a support base. Two assemblies/setups will be tested, and are shown in Figure 12.

Assembly #1 - Eddy Current Brake Setup

The eddy current brake setup will try to slow the spinning flywheel down by creating eddy currents which create resistance on the spinning flywheel. This resistance will impose a load on the eddy current brake which, through analysis, will give the torque on the blade shaft. The rotational speed of the shaft will be captured using a tachometer.

9

#2 – DC Motor Setup

A permanent magnet dc motor will be mounted on the top end of the shaft which will produce a DC voltage while the shaft rotates. As in the setup #1, the rotational speed of the shaft will be determined using a tachometer.

Figure 12 –Blade Setup Assembly

In both setups, a data acquisition system will be utilized to capture and record instrumentation values during testing. These values will be analyzed upon testing and compared with the estimated power output for the blades design.

4.2.2 Testing The test matrix for the wind turbine tests shall be broken up into 2 sections, Static and Dynamic testing. All tests runs shall have data recorded for a minimum of 2 minutes once a steady-state condition is achieved.

Static testing - The wind turbine shall be locked at particular angles starting at 0 degrees relative to wind flow and the torque produced by the blade will be recorded through a load cell. This shall be repeated at 45 degree segments at nominal tunnel speeds of 2.5, 5, 7.5 and 10 m/s.

10

Dynamic testing - The blade shall be free to rotate under the load provided by the wind travelling through the tunnel at 2.5, 5, 7.5 and 10 m/s. When dynamic testing begins, the blade shall be rotated by the wind motor at a very low, but constant, rotational speed to obtain bearing friction.

Both static and dynamic testing will be performed with a flat plate closing in the blade section. This date will be compared with the unclosed blade section testing to see which design performs better.

Test Test ClosingNumber [km/h] [m/s] Condition Plate

S1 2.5 0.7 Static NoS2 5 1.4 Static NoS3 7.5 2.1 Static NoS4 10 2.8 Static No

S1C 2.5 0.7 Static YesS2C 5 1.4 Static YesS3C 7.5 2.1 Static YesS4C 10 2.8 Static YesD1 2.5 0.7 Dynamic NoD2 5 1.4 Dynamic NoD3 7.5 2.1 Dynamic NoD4 10 2.8 Dynamic No

D1C 10 2.8 Dynamic YesD2C 20 5.6 Dynamic YesD3C 30 8.3 Dynamic YesD4C 40 11.1 Dynamic Yes

Wind Tunnel

Helical Savonius Wind Turbine Test Matrix

5 Economic Analysis As the wind turbine design has been completed, it is now important obtain a cost estimate for both the prototype and full-scale concept. The prototype cost can be estimated using the hourly rate for use of the rapid prototype and the labour necessary to fabricate the final prototype. Memorial University Technical Services will be contacted to obtain an estimate to fabricate a full-scale model of this turbine using various different materials. This information will then be used to develop a cost per kilowatt-hour for this particular VAWT model.

11

6 APPENDIX A: PROJECT PROPOSAL

12

13

14

7 APPENDIX B: PROJECT SCHEDULE

ID Task Name Duration Start Finish

1 Wind Energy Research 8 days Mon 1/12/09 Wed 1/21/092 Concept Selection 1 day Thu 1/22/09 Thu 1/22/093 Preliminary Design 3 days? Fri 1/23/09 Tue 1/27/094 1st Meeting With Supervisor To Review Concept

Selection1 day? Tue 1/27/09 Tue 1/27/09

5 Finalize Concept Selection 2 days? Tue 1/27/09 Wed 1/28/096 Design & Simulations 18 days? Wed 1/28/09 Fri 2/20/097 Foil Designs 18 days? Wed 1/28/09 Fri 2/20/098 Foil Simulations 18 days? Wed 1/28/09 Fri 2/20/099 Foil Selection Based On Simulations 1 day? Fri 2/20/09 Fri 2/20/09

10 Develop SDL Files On Selected Foil Design 1 day? Fri 2/20/09 Fri 2/20/0911 Fabrication 20 days? Mon 2/23/09 Fri 3/20/0912 Foil 15 days? Mon 2/23/09 Fri 3/13/0913 Support Structure For Testing 10 days? Mon 3/9/09 Fri 3/20/0914 Testing 16 days? Fri 3/13/09 Fri 4/3/0915 Foil Outfitting 6 days? Fri 3/13/09 Fri 3/20/0916 Testing Instrumentation Setup 5 days? Mon 3/23/09 Fri 3/27/0917 Wind Tunnel Setup 5 days? Mon 3/23/09 Fri 3/27/0918 Foil Testing Matrix 3 days? Wed 3/25/09 Fri 3/27/0919 Analysis Of Testing Data For Reporting 5 days? Mon 3/30/09 Fri 4/3/0920 Reports & Presentations 46 days? Mon 2/2/09 Mon 4/6/0921 Report 1 - Project Update & Status 1 day? Mon 2/2/09 Mon 2/2/0922 Interim Presentation 1 day? Wed 2/18/09 Wed 2/18/0923 Report 2 1 day? Mon 3/2/09 Mon 3/2/0924 Final Presentation 1 day? Mon 4/6/09 Mon 4/6/0925 Final Report & Submission Of Logbooks 1 day? Mon 4/6/09 Mon 4/6/09

Jan 4 Jan 11 Jan 18 Jan 25 Feb 1 Feb 8 Feb 15 Feb 22 Mar 1 Mar 8 Mar 15 Mar 22 Mar 29 Apr 5 Apr 12ry February March April

Task

Progress

Milestone

Summary

Rolled Up Task

Rolled Up Milestone

Rolled Up Progress

Split

External Tasks

Project Summary

Group By Summary

Deadline

Page 1

Project: Wind Schedule 2Date: Wed 3/4/09