What I have learned building an airplane

This is an excert of the final report for an Aerospace Engineering class called Senior Design. I took this chair over 1 year while I was studying at University of Tennessee, Knoxville as a exchange student of the Brazilian Scientific Mobility Program

Helping building an entire model airplane was a really tough challenge for this semester. At the beginning of the year, we barely have started the report we headed to AIAA DBF. And 3 months later, we were flying the machine. It was a growing, unique and very gratifying experience for me. Our goal was to build two airplanes. A big and a smaller one. The smaller one (aka Panda/Production Aircraft/PA) was going to carry a full 32oz Gatorade bottle. And the payload of the big one (Dumbo/Manufacturing Support Aircraft/MSA) was both the bottle and the small airplane. A real flying crib for the smaller airplane. Generally, cribs are not as aerodynamically shaped as airplanes, but we managed to develop a very beautiful design of a flying wing, which was the first challenge. And here I had my first lesson: to choose a design based on the ambient conditions of the competition. We notice that most teams chose the standard airplane shape for their projects. But the windy Wichita didn’t forgive them. Despite the poor piloting skills (such pilot provided by the competition), our plane cut through the wind gracefully. A standard approach would be a way easier solution for the design, but we would have probably suffered a lot with the high speed winds from Kansas.

My main role before manufacturing was to develop the CAD of the project. And that was the most painful lesson lessons: leave the CAD for who knows CAD. If you don’t, learn it as fast as possible. Although I had classes on that, I didn’t have enough expertise to develop an entire airplane on SolidWorks. Of course, YouTube saved me. In less than one day, I watched about 95 video-classes. Here one more lesson: Google is like the Hand of the King. And you are like the King. The King messes, the Hand cleans. Then I was ready. Started drawing the flying wing. My roommate worked with SolidWords and Catia as an intern at Embraer. And he told me that sometimes, the CAD guy captures engineering mistakes. And it happened here. The main spar tape was too thick to fit inside the 4th rib (yes, the spars are like columns and ribs are like floors. But on the horizontal). And then, I also noticed that the rear spar was going to suffer of the same problem. That occurred because until the 3rd rib, the profile was a NACA4412 and from the 4th and on, it was a NACA23021. Then, the 4th rib was reengineered to be a transition between two different airfoil profiles. Problem solved. Lesson: everything is beautiful and perfect until it comes to the dirty reality of CAD. Unfortunately, I ended up having trouble with timing and the drawing, which led the CAD to be delayed and the deadline not reached. On the very last night, before delivering the report for the last review, my teammates drew the CAD in my place. It was shameful for me. And laborious for them, who had to stay up all the night. I deeply regret my mistakes that led to this.

After sending the the report to be evaluated by AIAA, we started the building process. Actually, we started a bit before, building a prototype with foam. This is a good one: build a prototype for everything you need to invest a lot of time and effort. I will save you time and effort. We could collect some data and get ready for the building process, which was started by Spring Break. All the material was already bought, and we had no money for anything more, but from our own pockets. Therefore, we began by building the ribs of the MSA. We had not planned our workflow for the building process. So we had just figured out that, although we had the ribs’ profiles measures calculated, we did not know how we were going to pass those measurements to the balsa to cut it. We could have chosen a couple of alternatives, such as printing from CAD, or using the CAD model to laser-cut the balsa. But we thought that we were not going to have enough time to schedule the laser cutter (it was Spring Break) and cut everything. So we did it by the simplest and laborious way: using a Matlab script and the calculated measures, we could get the points of profile’s curve and drew it by hand on the balsa. Then, cut wood on the vertical saw. Definitely, it was not the smartest option. We had a lot of measurements errors, parallaxes, and even the pen’s ball created some errors. Plus, the sawing process also introduced errors, since the blade was not thin enough and cut irregularly. The result: asymmetry and imprecision.

Messy measurements. From decimal to fractional numbers. Double pain.

Here, a very important lesson: plan and predict the manufacturing workflow beforehand. It is going to save a lot of time and avert imprecisions and errors. And reserve some money for unexpected problems.

After a couple weeks, we moved to a bigger workshop that could nest both airplanes manufacturing (yes, we had to build two different airplanes for the competition, with two different groups inside our team, in which a smaller airplane was going to fit inside the bigger one). Everybody spent a lot of time on that workshop. At least, almost everybody. We had a lot work to be accomplished, but sometimes, there was too much people there, most of those doing nothing, because they needed to wait the current part of the process to be done to start theirs. I spent hours just watching people working. That was a very bad practice and should be avoided. What we should have done is the next lesson: in a process that involves team members, each task should have a specific schedule. It avoids workforce, time and stamina to be wasted. If we had scheduled each task and specified who was responsible for that, no one was going to waste their time by just watching other peers working. Besides, we would not have spent entire nights doing work, and waiting for some peers to show up.

When we got to the day before the competition, we still had a lot to do. A couple of peers and I spent the entire two nights previous the depart day on the workshop. Then, we finally departed. Fourteen hours of road after, we arrived at Wichita, Kansas, at almost 4am, directly to the meeting room to finish the airplanes. One thing that I noticed was that all the tools were not properly organized. Lesson: spending some time organizing the tools would have saved a lot of precious time during the competition. Besides, bringing tables and chairs would also have spared our backs.

On the first day of competition, I arrived by 12pm on the site, and surprise: the propeller crib was broken. And the ease it was broken sort of terrified me. Some of the rubber bands we used added to am accidental hit broke it. When it came back to our spot, we realized that the crib was not even screwed to the base, which was held only by epoxy glue. We took about 40 minutes to fix it, screwing and zip tieing straightly onto the base.

Such a beautiful Dumbo. Such a good team.

We had to use a lot of rubber bands to hold the airplane’s cap on the right place. And also a lot of taper. Mainly on the third day, when it was a really bad windy rain. We had 3 successful take offs out of 4 tries. From those 3 tries, only one landed successfully. Not our fault. The pilot, sourced by AIAA, took a lot of amateur decisions that led the airplane to crash twice. On the first crash, the nose landing gear broke. We took about 2 hours to fix that. But we had a very good problem-solving workflow and team work. The second time, the carbon fiber rear landing gear structure was broken. Two to three hours to fix. We replaced the carbon fiber one for a glass fiber. It made the Center of Gravity to go forward (that's a bad thing) because it weighted almost tree times more. Again, a good workflow and team work. On the last flight, we could, finally, take off and land safe and successfully, thanks to the pilot, which was replaced. Hardly ever should you trust on a pilot outside the team. They seemed that didn’t care about you and your machine. If something crashes, keep calm and fix it, and never forget to have a good team work and workflow.

In the end, I have learned a way more about engineering building this airplane than taking classes. Of course, classes are very important to build the basis of thought and engineering. But the on-hands work generates a better learning curve. And there are some things that are not on books that you can learn only by working with it. It was one of the most interesting, resourceful and astonishing experiences on my student life (and résumé).

Mission accomplished!
One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.