Variable Sweep on Flying Wing Aircraft can Reduce Drag

Kevin Shen

I’ve been fascinated with flight for almost as long as I remember. There is something about such advanced and precise things operating completely untethered from the ground that was simply awe inspiring to me.

I first began experimenting with Radio Controlled (RC) model airplanes, first watching online videos, to flying myself, to designing and building from tutorials. After I became more comfortable with off-the-shelf electronics for model airplanes, I learned about the wonders of Arduino and self-programmable electronics systems. It all seemed so overwhelming to me at the time, how the electronics worked, but one google search after another eventually taught me the things I needed to know. I thought it would be really neat to be able to control one of my planes with my own code, but I had very limited success due to many issues, some of which I continue to work on to this day. However, the experience and ground testing that I had done set me up to be able to work on something larger.

As I entered the eighth grade, we were assigned a school-year long project that would simply have to learn something from. It was incredibly open ended, so I chose to participate in a local science fair, (PLU Discovery Science and Engineering Fair, formerly PLU South Sound Regional Science and Engineering Fair), and try to “clean up” my random and scattered knowledge and skills of aeronautical engineering and make something new.

I was interested in the concept of flying wings, and read that sweep was one of the important factors for flying wings to work well. However, it created inefficiencies that offset the lack of tail, and therefore usually isn’t used in aviation. I decided that having a varying sweep wing on a flying wing would be an interesting design concept and have unique advantages (and disadvantages) over other designs and should be attempted. Some quick Googles showed that the concept hadn’t been tested before, at least not officially. I found oblique flying wings, and variable sweep on fighter jets to increase supersonic performance, but nothing of the type I was imagining. So, I signed up to the fair as that project and would figure out the details later.

Figuring out the details was much more difficult than I expected. The basic concept was relatively simple, but deciding where CG movement, trim changes, how to compare flight performance, and more, took me quite a while and was the majority of the time spent on the project.

In the meantime, as part of the eighth grade project requirements, I was to find a mentor. After sending a few dozen emails I got a response from an aeronautics professor at the Embry Riddle Aeronautical University; Prof. Richard Prazenica. He specialized in controls, which seemed very applicable to part of my project. We exchanged emails and met, and he told me about different control methods, more than just the simple PID loops I’d learned before. However, as the first full lockdown-ed year went on, he became very busy with his profession and it became difficult to communicate. A month later, one of the professors I had emailed at the University of Washington referenced me to the Design, Build, Fly team, whose leaders, Ethan and Daniel, answered quite a few of my questions about some design decisions, but being university students were also quite busy.

At this point I was trying out different ways to make this project work, and over the course of a few months struggled to find a fully foolproof way to test the concept. After an initial burst of progress designing, optimization with MATLAB, and flying a control design to compare with the variable sweep design I spent far too long perfecting a design, only for it to not work well. My goal was to make the comparison as fair and with as few variables as possible, even matching the structural integrity of both designs. However, this sacrificed some aerodynamic qualities (of both designs equally), which I only found out about a week before the project deadline when I began the last round of flight testing.

With limited time and nothing working, I thought to just make something, anything work and collect some basic data, or I wouldn’t have anything to show. In a few short days I redesigned the whole variable sweep UAV system just to fly (although still keeping the basic dimensions of the control UAV) and in the gaps between rain would do my data collecting flights. I was quite disappointed with myself as the tests weren’t very reliable the data was short and uneven (only about 2 minutes of flight time, due to crashing into a tree compared to the 6 minute flights of the control), the data did not show any improvements, and the presentation put together quickly and submitted midnight the night the project was due.

However, despite this I placed first for the eighth grade category at the PLU fair and was nominated to Broadcom MASTERS. This gave me a little bit of time to think about my project, and the data I collected, and realized that although it wasn’t very good, it showed the concept was viable and that drag was reduced- although the extra weight of the variable sweep design likely was what kept its absolute efficiency lower. I submitted my application and was placed in the top 300.

Finally, after a few months had passed in the summer of 2021, after I technically entered high school, I advanced to the Washington State Academy of Sciences which nominated me for AJAS and led to being inducted as an AJAS fellow.

Includes:

Research Project Paper

Short video demonstrating UAV flight