The Mars Electric Reusable Flyer is a three-year project—with at least another year and a half left—so my final post obviously can’t overview the entire project. This summer was dedicated to building and testing a prototype capable of consistent, stable flight. Furthermore, the prototype needed to be able to house the Piccolo autopilot (which is about the size of an iPhone). Sometime this fall or next spring, the rest of the team (those employed by NASA) will conduct a high-altitude balloon drop of the Flyer from 100,000 feet. In anticipation of this test, my research revolved around building and testing multiple Flyers with a team of other students under the guidance of a NASA engineer.
As I mentioned, in my previous blog post, the CTOL prototype was not performing as we had hoped. We therefore decided to switch to a VTOL model. You may remember that the end goal of the project is to have a VTOL Flyer for use on Mars. So you may, then, be wondering why we even bothered with a single-engine, CTOL prototype. Well, this summer’s prototypes are only being used for testing the airframe in anticipation of a glider drop from the balloon. Translated, this means that it really doesn’t matter how we got the Flyer to fly this summer, we just needed a model that flew. In theory, CTOL is more simplistic and in turn more reliable than VTOL. Therefore, we initially chose the launch/power system that is easier to set up and use. However, after a series of problems with the CTOL prototype we went ahead and made the decision to switch to VTOL.
The conversion of a single-motor CTOL model to a twin-motor VTOL is rather complicated and time-consuming. To make things more difficult, the model that we overhauled was 3D-printed out of polycarbonate (a relatively fragile material) and covered with fiberglass and carbon fiber (very strong materials). Point being, we needed to be careful not to crush the model while drilling/cutting through the stiff outer layers while retrofitting the model. From here on I will refer to the model as MERF-1 (short for Mars Electric Reusable Flyer).
One of the biggest challenges was designing and fitting the electronics architecture inside the nose of MERF-1. Packed into the nose are two batteries, a kk board, a multiplexor, two electronic speed controllers (ESC), and wiring. I helped design a plywood mount that all of the electronics attached to; the mount would keep them organized and secure during flight. We used computer-aided design (CAD) to design the mount and perfectly fit it in the nose of MERF-1. Instead of fabricating the mount by hand, we sent the CAD file to a water jet cutter, which can precisely cut a multitude of materials. Using the water jet sped up the process and enabled us to rapidly make spare parts. In addition to designing the architecture, I also soldered the connections and installed the electronics. The building process was obviously a lot longer than this one paragraph but the rest of the details get a bit dry so I’ll move on.
The last few days of my research involved flight testing MERF-1. Given that it had never flown before and has two propellers (a potentially dangerous scenario), we initially tested it in an enclosed cage with a tether. The first test we did was hover testing—essentially you power up the aircraft, hover a foot or so above the ground, and then land. This type of test is like the ‘test flight’ of the test flights; it is a test that checks all the systems before more aggressive flights. The test was quite successful—a much welcomed change. The following tethered tests were also successful which gave us the go-ahead for open air flight tests.
The first full-blown flight test was a bit nerve-racking. Just as in the cage, we started with a simple hover test to make sure all the systems were working well. Following that, the pilot did a ‘high-altitude’ hover at about 30 feet. Again, MERF-1 performed extraordinary well. Now came time for the first horizontal flight. The challenge of this flight was transitioning from vertical flight to horizontal flight, a transition that occurs over four seconds as the motors tilt. This part of the flight was nominal but problems soon arose in horizontal flight. MERF-1 was in a shallow turn when (we believe) it lost power to one of the motors. Without power on one side, it was impossible to pull out of the turn completely. Instead, the bank angle shallowed and the turn radius increased. After about two revolutions the turn radius had increased enough to put MERF-1 on a collision course with the tree line. Needless to say, we lost MERF-1. We hypothesize that MERF-1, a 13.5-pound aircraft, collided with a tree at approximately 30 miles per hour. Despite hunting around the woods for upwards of an hour, we were unable to recover MERF-1. I was disappointed in the failure, but my time at NASA taught me that failure is part of the development process. Although that was my last day at NASA, the project continues. After I left, another model, presumably MERF-2, was built and testing has most likely begun. My boss invited me to come back any time, so I hope to spend winter break working on the project.
Working at NASA this summer was a dream come true. The experiences and knowledge that I gained during my tenure there are unmatched in my life. I am very grateful for the opportunity that I had. Before I wrap up, I’d like to thank the people that made my summer research possible: Lisa Grimes and the Charles Center; my advisor, Professor Wouter Deconinck; my boss at NASA, Dave North; and the other students I had the privilege of working with, Ashwin Krishnan, Arjun Krishnan, Erin Clifton, and Henry Kwan.