Windmill Design

The goal of this project was to optimize a windmill. To do so, we first edited the design we were given then focused on the electrical components. The goals for our electrical components were to first enhance the power provided by the windmill by optimizing the power output on the DC motor we were given. Second, through filtering the signal sent from our motor, we needed to obtain a constant voltage. Finally, we needed to research techniques to optimize our windmill design to harness more power and compare all of our prototypes to decide which is the most effective. 

Windmill Baseline

At the beginning of this project, we were given this windmill blade design and this windmill blade hub design to 3D print, test, and then optimize. I sent these files to Fusion360 to combine them and then I saved it as an STL. I then sent it to Prusa Slicer, added supports, then saved the file as GCode. I printed this file on the 3D printers provided by the Charlotte Latin Fablab and then used super glue to attach the blades to the blade hub. Using a small drill, I carved a hole into the center of the blade hub, which now had the blades attached. With this hole, I inserted the rod of the motor into my windmill. 

WaveForm

First, we wanted to get an understanding of how the motor would work during testing so without any windmills attached, we took our motor and connected it to an Analog Discovery. An Analog Discovery is used to generate the input/output of voltage. We connected the orange wires, representing channel one positive and negative respectively, of the Analog Discovery with alligator clips to the red and black wires, representing positive and negative respectively, of the motor. We connected the USB and the Analog Discovery to the computer and opened up WaveForm. We used wave form to track the changes in voltage of the AC signal from our motor. I spun the rod of our motor manually and watched the reading on WaveForm. The spikes were uneven, due to my arbitrary spinning, and there were tiny uneven spikes on top of those large spikes, due to the fact that we were measuring AC instead of DC voltage. This unsteady voltage that we saw would need to be filtered and transformed into DC voltage to get a steady flow. The purpose of using WaveForm was to obtain an understanding of why we would need to convert the AC voltage to DC voltage and filter the signal. 

Filtering Signals

To filter the voltage signals while testing, we were given a small green PCB board, for surfacing mount soldering, four different capacitors of varying strengths, and two different resistors also of different strengths. We soldered these items onto the PCB board. For resistors who would determine our DC voltage, we soldered a 5 ohm resistor to the r1 place indicated on our board and a 1 mega ohm resistor to r2. For our capacitors who filtered the voltage and would help to produce a more constant voltage, we started off with a low capacitance at 22pF for C1 as indicated by our board. We placed the capacitors in parallel and next added a high capacitance of 10,000pF on C2 to filter the low signals while the low capacitance of C1 would filter the high signals. For C3 we used 100pF and C4 we used 1,000pF, following a similar pattern with the low for high and high for low. Because we added the capacitors in parallel, we could test the different capacitors at different times without shutting off or disturbing the rest of the circuit. We connected the positive and negative wires of the motor to the positive and negative hols in the PCB board. As demonstrated by the comparison of signals below, filtering the signals means decreasing the ripples and arbitrary spikes in voltage and providing a more steady voltage flow. We included a range of capacitors to account for all different sizes of spikes to filter the entire output of signal from the windmill. For reference, the range in voltage for the baseline windmill unfiltered was 73.52mV and filtered it was 36.84mV which is around half of the unfiltered test. 

Windmill Baseline Testing

Our setup for windmill testing was simple. We set a fan on top of a bucket and set a ruler from the edge of the bucket to the tip of a stand. We placed the motor attached to a windmill inside of the stand and tightened the screws to make sure the motor didn't fall out. We started the fan on high and made sure to use the high setting for testing all of our windmills. We connected our motor to the Oscilloscope, which would read the signal sent by the motor. We connected them by attaching the end of two alligator clips to the ground and power wires of our motor and then we touched the measuring tips of the Oscilloscope to the capacitors on our PCB board. We tested different angles of the Oscilloscope (0°, 45°, 90°, 135°, 180°) to compare the average voltage. 

Windmill Optimization #1

After using multiple sources to determine the next step in optimizing our windmill, we decided to simply modify the blades of our windmill. Research found here led us to round the blades in an attempt to have wind more easily roll off the blades and make them longer to touch more wind. We used the same blade hub as the provided baseline windmill. We 3D printed the blades and assembled our windmill and replaced the given windmill with the new windmill on the motor. We used the same testing process as the baseline windmill.

Windmill Optimization #2

For our next optimization, we made a windmill which was already connected to its hub. We also edited the blades by getting rid of the ridges and making them a little bit thinner. We 3D printed this windmill and used the same testing process as the other windmills. Unfortunately, the windmill's blades were so thin, it could not spin the motor at any angle. Therefore, we have no data for this windmill.

Windmill Optimization #3

For our next optimization, we decided to make our blades shorter and thicker, again with a built-in hub which this time was round. We 3D printed this windmill and used the same testing process as the other windmills. 

Windmill Optimization #4

For our last optimization, we decided to use the given base hub again but make a fully crescent shaped wing which was long and had a hole in the middle. Our research led us to believe a longer wing and a scooped wing would harness more wind and a hole in the middle would allow for wind to pass through with less resistance therefore making the windmill spin faster. We 3D printed this windmill and used the same testing process as the other windmills. 

Conclusion

The most successful windmill was windmill optimization #4 because it generated the highest average voltage. The most consistent windmill was windmill optimization #3 with its average voltages being around the same at varying angles. In the future, I would prioritize length and thickness when optimizing. Our thinnest windmill did not spin at any angle and our shortest was consistent for many angles but at an overall low voltage.