My entry for the regional and Canada-wide science fair was my electric skateboard, but it was more of an engineering project than a science fair experiment. Since the judges were tasked with analyzing the scientific thought/method of the contestants, I decided to run experimental trials on my latest prototype in order to determine how the speed, load, and gear ratio effects power consumption. From that, I found the optimal speed for maximizing the range of the skateboard, the optimal gear ratio, and I also provided proof that my particular ESC had substantial regenerative braking capabilities.
In total, I did an absurd amount of trials (absurd = 48). I had two different groups; a group that accelerated uniformly, and a group that rode however they wanted. I would log the power consumption, location and speed of the electric skateboard during the trial with my Bluetooth data logger, which would conveniently place an .xml file in a folder on my phone.
I also had two different groups with two different gear ratios, as I wanted to see which one would perform better. All of these groups had four different rider loads. Here are all of the trials and the data I have collected for my experiment.
In order to log the data for analysis, I had to build a piece of hardware similar to an RC logger. Above is the schematic I came up with, which I etched using my dad’s CNC machine. I put a stronger emphasis on accuracy, so I included a tunable voltage reference for the microcontroller. The Bluetooth logging system would log the voltage of the batteries, current, temperature, throttle, location, speed, and position, and send them over Bluetooth to an Android device. The Android device would be running BlueTerm during a trial, and would be set to record the data.
The most interesting data that I collected from this experiment was the effect of gear ratios on power consumption. I tested two very different gear ratios. I hypothesized that the larger gear ratio would consume less power, since it exerted a lower torque load on the motor, and that lined up with a higher efficiency on a BLDC efficiency vs torque-load graph. My results were the opposite, and they were statistically significant enough to prove the opposite. After a bit of communication with my mentor (who provided an unbelievable amount of help over email), I realized that the max load had not been achieved on the motor, and thus, I was sampling the wrong part of the graph.
From a similar data set as the data above, I was able to estimate the speed at which the skateboard should get the farthest range. I derived a formula which worked off the best fit line for the previous chart, and turned it into a range vs. speed graph. The result showed that the absolute maximum range is about 19km, and that could be achieved by a 140lbs rider at a speed of 8.9km/h. This has yet to been tested.
Another thing was proven as a result of my experiment was the regenerative braking of the ESC. A lot of people doubted that the ESC I was using could put power back into the battery pack, but no-one could explain where the energy would go, other than the battery. I was able to prove that regenerative braking exists, and provides a significant amount of power into the battery. I was also able to determine the peak current, average current, preferred riding speed and more. For more information and data regarding my experiments on the electric skateboard, check out this link.