October 2020 - May 2021
Overview:
While in my senior year at Boston University I did research studying the aerodynamics of a rotor under Dr. Sheryl Grace.
Background
Urban Air Taxis are a hot topic in aerospace right now with many startup companies racing to make reliable vehicles and build up infrastructure to make flying taxis a part of everyday life. Many of the routes taken by these vehicles would be through Urban landscape densely packed with tall building. While much research has been conducted on the aerodynamics of helicopters, These are predominantly single or double rotor aircraft. Most air taxi designs have 4 or more rotors which significantly change the fluid dynamics and require further study. Additionally, outside of urban air mobility, multirotor drones have become popular vehicles for other like search & rescue, and delivery both of which require flying in confined spaces close to buildings and walls.
As you move towards a wall, the airflow around a propeller will change, possible changing thrust and/or producing a torque. For example if you are close to the ground you will experience extra thrust. The purpose of this research is to characterize the aerodynamic performance of rotors near walls and other obstacles. With this information control systems could be improved to perform better in spaces which are more confined.
The first stage of research was to study a single rotor by running it at different speeds and measuring the thrust coefficient. This would be done with walls varying distance away and in different directions (above, besides, below, near a step, etc). Then later the experiments would be repeated for a quadrotor.
I was not able to finish this project unfortunately. I spent majority of my time trying to get the Jr3 sensor to give accurate measurements. I then built the test rig and took preliminary measurements but this was as far as I got.
Experimental Setup
We used a mt2212-11 motor which ran at 11.1 volts and was attached to a 8 in propeller. The motor was attached to a Jr3 6-axis load cell which could measure all three directions of force and moments. A tachometer was attached to measure the rpm of the motor. This whole setup was mounted on some 80-20 bars to raise it off of the ground. The tochometer measurements were read by an arduino, while the motor speed was controlled by a separate controller. The Jr3 sensor was read through its included DAQ which plugged directly into the computer.
Now majority of the time was spent trying to get accurate measurements from the Jr3 force/torque sensor. This sensor did not come with MATLAB software so we had to figure out how to read from the dll via Python. Luckily Dr. Grace had a neighbor who was a fantastic programmer who made an object in Python to interact with the Jr3 sensor DAQ system. Through this I was able to figure out how to get values from the DAQ. The python code can be found on my github. The next issue was that the values were not correct. I calibrated the sensor by placing known weights on it and measuring the voltage. There was a setting in the DAQ that had to be changed to make the voltage match up with the weight. Unfortunately however, this setting would change back to its factory setting every once and a while and I could not figure out why is was changing. Since it created a nonlinear change to the voltage output, I also could not just ignore it.
To read rpm I first made my own tachometer using an LED with a photodiode, which can be seen in the left image above. The motor is black with a white stripe. When the white stripe passes the tachometer the photodiode reading spikes, which tells the computer on rotation has passed. The homemade sensor was not able to ready very fast, perhaps due to the cheap photodiode. However, I later just bought a photodiode which worked in the same way, but much better.
By the time I graduated and left BU I had set up the testing rig and figured out how to read from the Jr3 sensor. I had taken some initial measurements, however, I had not done in depth analysis of it.
While I did not get far in the data collection and analysis and theory, from this project I gained a lot of experience as an experimentalist. I was able to set up sensors and data acquisition systems, write code in Python and Matlab, and build an experiment.