Spring 2018

The spring semester is already ending, and I’ll do my best to cover all that I can. Between the OSU Rocketry Team, my senior capstone project, side projects, following SpaceX, and finishing undergrad work, it has been the best semester of college yet.

This final semester of my undergraduate coursework consisted of only two classes: my senior design (capstone) course and an aerospace laboratory that has some parallels to the capstone course. Only 7 credit hours has been bliss, and while it is just as busy as taking 15 credit hours’ worth of professional school classes, 90% of my time and work is towards rockets, and I couldn’t be more excited about it.

My senior design project was with 9 others, some mechanical engineering majors, some aerospace engineering majors. Most of us were active on the OSU Rocketry Team, and we worked towards winning the 2018 Argonia Cup. We called ourselves the OSU Rocket Squad, a nod towards the quad(copter) we developed to fly aboard our rocket.

The objective of the competition was to take a golf ball payload to 8000 feet and recover it as closely to the launch site as possible. The first 3 weeks of the semester involved choosing the best means to recover the golf ball, whether it be a deployable glider, deployable quadcopter, integrated glider, integrated quadcopter, or just basic dual deployment recovery via parachute. The integrated options posed a lot of problems about returning a 20lb+ vehicle from the sky, so those were eliminated early. We chose a deployable quadcopter for its active stabilization, powered flight, and prior team member experience. The quadcopter would be the rocket’s nosecone and contain spring-loaded arms, folding propeller blades, GPS, its own flight computers, a backup recovery system, remote control override, and of course, a golf ball. The nosecone and quadcopter structure were made from 3D printed PLA plastics, and the nose was only 5 inches long. While the nose was short, it contained internal shelves that housed the electronics, making every cubic centimeter count.

Our team was split into 3 groups: Structures, Avionics, and Integration. Structures (my team) focused on building our own fiberglass tubes that would become our airframe, couplers, and motor mount tubes, and our own fiberglass sheets that were CNC cut to yield fins, centering rings, bulkplates, and quadcopter motor mount supports. By using fiberglass cloth, fiberglass resin, and a casting mandrel, we built countless tubes of varying lengths and diameters that would have cost over $1000, had we purchased all the parts commercially. Even with R&D and some failed parts, we estimate that we spent less than $500 on materials.

Avionics focused on the design, programming, and electronics needed to make a quad from scratch. Integration designed spring-loaded arms and worked closely with Avionics to build the functional quad.

In February, our Structures team flew a small rocket made entirely of our own fiberglass parts on an I357T as a test to ensure it could withstand all stages of flight including launch, ejection, and landing. It performed beautifully, and because the tacky fiberglass resin got stuck to Chad’s garage floor, tabletop, our hands, forearms, arm hair, and shoe soles, the rocket was appropriately named Everything Is Sticky.

Chad Kenkel, Gerald McCullers, and myself (Structures Team)

On the structures side of things, our tubes were then verified to withstand flight stresses. We worked with a 5” steel pipe that was the only suitable 5” OD tube on the entire internet long enough and could be purchased in a single quantity. There were plenty of cardboard mailing tubes that would have worked, but we would have had to purchase a dozen of them for over $150. This steel pipe ended up being $50 and worked great for us.

We eventually nailed down a system called the “Pour and Pet” method for wrapping fiberglass cloth around the mandrel, pouring fiberglass resin between the layers, and petting it with nitrile gloves on to evenly spread it. To make tubes from start to finish we needed the mandrel, fiberglass cloth, fiberglass resin and hardener, red Solo cups for scientific resin measurement, popsicle stir sticks, butcher paper for wrapping around the mandrel, scissors to cut parts to size, nitrile gloves, petroleum jelly, and vegetable oil for lubrication. For making sheets, the process was similar but with 8-12 squares of cloth precut and put down with resin between each by means of the “Slather and Scrape” method. Patent pending, of course. 

Our senior capstone counterparts doing Speedfest laughed at our methods because they have vacuum sealed components, entire labs, and unlimited resources at their disposal. Not to mention we did all of this in our team member’s garage because we had no place on campus to do this. It shows just how accommodating our department is to the success of our team representing OSU on the national level against 9 other universities in a collegiate rocketry competition. But that’s a whole other topic I could rant about.

We all diligently worked in our respective subteams, overcoming one challenge and moving on to the next. In March, we flew again with the Kloudbusters in Argonia, KS, this time a full-scale configuration using the same motor we expected to use in competition. On the structural side of things, we made good use of lubricants to aid in removing the tubes from the casting mandrels, and this rocket was affectionately named Everything Is Slippery. It became our backup rocket in case the final iteration was damaged prior to flight in any way, and this flight verified that we would reach the necessary altitude. This launch also tested a basic version of our quad: the 3D printed structure and a backup recovery system. This backup system contained a small parachute that would catch the quad in the event that all systems went offline.

The launch also featured the worst possible wind conditions, so if the flight went well, we knew that we could fly in the worst-case weather conditions. Most of the day was spent resolving issues with the quad, and the generosity of the Kloudbusters and their electrical power supply helped. We finally launched with 15 minutes remaining in the day’s waiver, and it took off without a hitch. The up part was great, but a tighter-than expected fit of the quad inside the airframe prevented the quad from separating at apogee, and therefore, the drogue also didn’t come out. A ballistic descent wasn’t good, but fortunately the main parachute caught the rocket and threw the quad away from the rocket. We recovered the rocket to find no damage whatsoever which was a huge step regarding the validity of making our own fiberglass components. The quad took a while to find (it got thrown far away), and it didn’t fare well. The SLA-printed nosecone and body were smashed to smithereens, and there really wasn’t much else to see.

Awaiting the altimeter to reach flight-ready status

Caleb Ritchie, Chad Kenkel, me, and Nick Foster

What was left of the quad

It was a day with mixed results, but the fact that the rocket held up to the extreme jerk of a main-only deployment event was a big step forward. It was one less thing to worry about, and since that part is what I was responsible for, I felt relieved, as did my fellow structures team members.

Spring break followed shortly after this launch and that consisted of ejection charge tests to get the piston working and attempting to fly the quad. On the quad side of things, 3D printed parts kept breaking, it couldn’t stabilize, rotor motors malfunctioned and burned out; there were all kinds of problems. By the end of that week, we couldn’t fly the quad anymore if we tried because we had to wait on more components to arrive in the mail.

Following spring break, our structures team got back to work building the tubes and sheets we needed to assemble a second iteration of the competition rocket. For some unknown reason, our lubrication methods were not working, and we went to extreme measures to attempt to pull off the tubes. One of these measures included drilling small holes in the fiberglass tube to tie a harness of paracord through and then attach those loops to the hitch of an F350 diesel truck. The pipe was tied to a telephone pole, and then the truck slowly drove forward. The truck engine began to whine and then the paracord ripped clear through 3 inches of fiberglass. To remove these failed tubes, we heated the fiberglass with a heat gun and used a screwdriver to chisel all the way through it and along the tube’s length to cut it off the mandrel. I began talking to the Speedfest teams to see what options there may be using expensive wax and release methods. Before we went through with that, we tried just adding a second wrap of butcher paper, and while the tubes weren’t as clean (seams and small bulges were present), these could be removed from the mandrel with a gentle pull. It was frustrating to have wasted over a week on three failed tubes (including $50 worth of cloth and resin), but this method is now fool-proof. We even perfected the butcher paper wraps to minimize the seams and overlaps. This method of an extra wrap of wax paper and petroleum jelly worked even better than a $50+ wax and release method would have that the Speedfest people would have done.

Now that we had new tubes again, I began building the competition version of the rocket, identical in dimensions to the previous, but more soundly built because I had additional time. With this new rocket, we also used new fiberglass cloth that we purchased (rather than use that from the sailing club), that made the tubes much heavier but even more stout. The 40” aft section that weighed 33 ounces on the last rocket now weighed 54 ounces with the same number of resin and wraps. Fortunately, we were so far above our required altitude of 8,000 feet, that the additional weight wasn’t of concern to us.

Throughout this entire semester, my roommate’s woodshop tools came in very handy. We were able to use his table saw to cut the fuzzy and ragged edges off our tubes and sheets, and his new CNC machine was put to good use. Prior to April, he did all the cuts for us, but during spring break, he showed me how to operate everything, so I could do all the cuts myself. When cutting our fins, I added some flair by engraving the OSU logo into the fiberglass. It was a clean touch that cut into our altitude but looked neat. One of our team members has a friend who makes vinyl wraps for automobiles and such, so he got this rocket wrapped for us. It was cleaner and more durable than any paint job would be.

The rocket was named The Other Things, inspired by John F. Kennedy’s “We choose to go to the moon” speech when he says, “We choose to go to the Moon! We choose to go to the Moon in this decade and do the other things, not because they are easy, but because they are hard.” It pairs nicely with our quad which was named The Eagle because it would return to us to complete its mission. The Other Things was completed a week before launch and allowed for ejection charge testing.

The day before launch, we prepared to ejection charge test all of our rockets flying over the weekend. This included our capstone rocket, the rocket built by the rocketry team, and our 12’ tall rocket for the Spaceport America Cup (more on this later). Delays kept us from doing these tests any earlier, but Friday morning on April 6th, our team congregated at my house for testing in the backyard. Delays kept us from testing anything until the late afternoon, and after a few attempts, we got the club’s rocket and the SA Cup rocket to separate properly. The capstone rocket took so long because new carbon fiber arms arrived that morning, and it look four hours to get them cut and fitted properly. Once we finally ejection charge tested it, the brand-new carbon fiber arms broke on impact with the ground. Turns out, the cheap PVC works better, so we re-configured the quad with those. We still had issues with our piston pushing out both the quad and drogue parachute, and we ran out of ideas on how to fix it. This was baffling because we got it to work over spring break, but now the night before the launch, we were unsuccessful.

At this point it is 9:00 in the evening, and my roommate wanted to have the house back (I’d had the team over all afternoon and evening), so we regrouped at another team member’s apartment to do some final testing and get it right. A nasty cold front moved in throughout the day, so we were working outside in freezing temperatures with 20+ mile per hour winds. At midnight, we finally achieved a partial success, but on the final test, several rivets holding parts of the quad together sheared through their 3D-printed holes because the force of ejection was so great. This posed a bigger problem because we had to secure these parts of the quad back together. Fortunately, some sheet metal and machine screws did the trick. Amidst all of this, an idea came to me to get both the quad and drogue to come out properly and that was by means of an apogee delay on our secondary altimeter. The primary altimeter would fire at apogee to separate the quad, then the secondary altimeter would first 3 seconds after apogee and release the drogue parachute. I felt like a genius, except I wish that idea had come to me weeks ago.

We get everything packed and ready to go at the lab at 2:30 in the morning the day of the launch. I went to Wal-Mart to buy snacks and hand warmers for the weekend. I managed to find handwarmers only at the self-checkout isle, the very last place in the store I happened to be. I bought all 5 packs there.

Once my personal gear was all packed away at home, I got to bed at 4:00 to sleep for 90 minutes. Funny enough, that was the exact same amount of sleep I got last year the night before the Argonia Cup. We left Stillwater just before 7:00 in brutally cold conditions, but with clear skies.

Saturday was relatively uneventful. I assembled the L1500T motor for our capstone and packed the recovery for that rocket and the SA Cup test flight. Delays with the quad kept us from flying our capstone rocket that day. The club flew their entry for the Argonia Cup but because the team’s recovery lead attached the main parachute too far down the length of the shock cord, the parachute never exited the tube and the rocket came down hard. Two fins fractured and could be repaired, but it would take an evening’s worth of work in Stillwater to fix and fly again. The entire day there hadn’t been a single qualifying flight by any other team, so as long as we could get up to altitude and land safely, we had a good shot of winning.

With that, we called it a day and hoped for better success on Sunday. Many of us wanted to camp out, but only four of us ended up staying the night on the Rocket Pasture. I may have been able to help out with the evening’s work, but the 11 hours of sleep I got in my warm sleeping bad was bliss.  

Sunday was a bit warmer, but also windier. We worked first to get the SA Cup rocket in the air. After a lot of delays with recovery and loose coupler sections, Results May Vary took off at 12:30 on an M1500G. Other than an altimeter being programmed to deploy main at 1200m instead of 1200ft, the flight was flawless. It did also go 1500ft lower than predicted, and we are still investigating that. In the grand scheme of things, this was the most important flight of the weekend. It validated our own fiberglass parts and that joining many sections together wouldn’t be an issue. It was the first successful M flight for the team, and without a doubt the largest rocket flown in school history. At the Spaceport America Cup this summer, it will weigh over 60 pounds, carry 360° video, utilize a PID control-looped airbrake, and reach 10,000ft.

Cowboy Rocketworks with Results May Vary on the pad

Under M1500G boost

Once Results May Vary was recovered, we could go on get the two Argonia Cup entries in the air. Additional delays persisted, and the weather worsened a bit. Our capstone rocket, The Other Things took off at 2:30 in the afternoon and the rocket part went without a hitch, from launch to touchdown. The quad wasn’t big enough to be spotted for a while, but a minute after launch, we start to see a tiny black dot far above our heads. Our quad created a buzz among the other teams and spectators, and I think they were as excited as we were to see whether this would work or not. As this dot grows larger, we start to see a quad-like shape to it, and everyone starts losing their minds. Bear in mind, no one has ever accomplished a rocket-launch autonomous drone before that we are aware of. What was concerning, however, was that our ground control station that received information from the quad had no kind of telemetry or data whatsoever. This meant that our 7lb drone had no remote control and with its long battery life could go rogue, and nothing we could do could stop it. Once the drone gets within a few hundred feet, we see it is all white, has no spring-loaded arms, and is carrying a GoPro. That’s when it hits all of us: that isn’t our quad. That was the most screwed-over I’ve ever felt in my entire life. Apparently, some guy who spent the whole weekend in his car with an FPV headset on was the pilot of that drone, and he claimed to have no control over it when it landed in front of everyone. I call bullshit, and I can only laugh now that someone actually did that to us when he knew we were all expecting a quad to land there after that rocket launched.

The OSU Rocket Squad prior to launch

The Other Things under an L1500T boost

Anyway, I go out with a couple others to retrieve the rocket which was completely intact and flew to 8,500ft. The rest of our team found what was left of the quad after that, and to this day, we’re still trying to figure out what happened. Not that we were surprised it failed, but it didn’t look as though it completely fell from 8,500ft either.

The remains of The Eagle

While we were out recovering, the rocketry team flew again on a second L1000W. It too came down hard but was in good enough condition to be a qualified flight. They landed closer than the only other qualifying flight that day, so OSU Cowboy RocketWorks became back to back Argonia Cup champions!

Our capstone team did have one other flight. A 54mm minimum-diameter rocket called If You’re Reading This, It’s Over that was purely for a backup. On a J90W, it chuffed and spiraled upwards into the clouds in the last 5 minutes of the weekend’s launch window. The radio tracker suddenly went silent, and I figured it lawn-darted. An hour later we found it thanks to a feeble signal from the tracker, and it was buried up to its fins in soil. Even if it had landed safely, I don’t think it would have reached the necessary altitude to qualify.

As the rocketry team gathered with the trophy for the picture, I stood with most of my capstone team, and we talked about the semester’s work, discoveries, failures, and takeaways. It was hard to not have anything fruitful to show for our efforts, but that’s sometimes how it goes. Our Avionics team talked about how we truly pushed the limitations of software and control algorithms to make the quad turn on and fly.

The rocketry team received $1000 cash, another $400 TeleMega altimeter, and two EasyMini altimeters (each team won one upon registration, and they are keeping our capstone team’s). I was grateful for that, and it gives us even more cushion for our expenses at the SA Cup this summer.

As per tradition, we ate at the Taco Bell in Wellington, but we took our time before leaving. All of us were in deep conversation either about this semester’s work, what the team could do next year, and how to proceed with the rest of 2018. Back in Stillwater, we cleaned up the DML and I was home around midnight. It was a long weekend and not the outcome I was hoping for, but it was unbelievable how much we learned and were able to take away from the last 3 months of work. Our discoveries will help the capstone team next year and our structures work made huge strides for the rocketry team, as this greatly reduces our reliance on commercial composites.

Since the competition ended, it has been a completely different couple of weeks. I’ve had abundant free time, time to go to the gym again, be outside, sleep plenty, and recap our efforts through this senior design project. We met with our advisors 2 days after the competition, and we are making a display poster, writing an internal paper to document everything we did and why, we’ll do a simple rocket launch demonstration flight at Speedfest (on Everything Is Sticky, our first test rocket), and after Speedfest, well lift the quad and forward rocket body via drone, drop the assembly, then eject the quad out and attempt to fly it down from there. There are fewer variables at play, and we hope we can get that to work. But as far as rocket-launching the drone, that part is done.

That is about it in terms of the Argonia Cup, and our senior capstone project. Because we are still writing about our work, and we have some minor tasks to still complete, we don’t have a truly “finished” feeling, but that is alright. This part comes easily because we’ve already done the work; now we just have to write it down. I’ve been told there are already 7+ juniors who are interested in doing this capstone project next year, and design ideas are already being tossed around. A fear of mine was that if we failed this year, it wouldn’t have been worth the financial investment from our professors, and they’d discontinue it next year because of our lack of success. There are certainly other capstone projects that involve just writing a paper which is significantly cheaper, but nowhere near as exciting. I hope to see OSU has 2 or more teams registered for next year.

Outside of the capstone project, there has been lots else going on with rocketry, both OSU-wide and with me personally. I’ll talk about those in a separate post. I’ll share that one first, so this post appears first. Thank you for reading, and I look forward to writing here more often throughout the summer.