Last year while working at Syntheon LLC, a Miami company that designs and develops hightech surgical instruments, Derek Deville caught wind of the Carmack 100kft Micro Prize during the long Fourth of July weekend. The competition offered $5,000 to the first team to build a rocket that flew to over 100,000 ft.
Deville thought he was just the right person to take on such a challenge. After all, he was a lifelong rocketry hobbyist, a mechanical engineering grad from Purdue University, and he had worked professionally in rocket design with Darpa, NASA, and the Air Force Research Lab, not to mention building rocket engines for Burt Rutan’s SpaceShipOne. He’d already built many rockets; this one just seemed like it would take a somewhat scaledup effort.
Planning and red tape
Deville carefully but quickly drew up plans for the rocket, which he named Qu8k in tribute to the man sponsoring the prize, John Carmack, developer of the popular Doom and Quake video games. The FAA examined the plans and by the end of July granted Deville a waiver for launch. Deville planned to launch his rocket in October at BALLS, an annual event held at Black Rock Desert, a dry lake bed in Nevada just north of Reno. The event draws rocketry enthusiasts from all over the world who launch large rockets, some with complex staging, others with multiple motors, as well as experimental and homebuilt designs. The event is sponsored by the Tripoli Rocket Association, a group that helped the Qu8k project in several ways.
For example, Tripoli cut through a lot of red tape by getting an FAA waiver that covered the entire BALLS event, clearing it for rockets to fly up to 150,000 ft. Deville and other hobbyists were allowed to piggyback on this waiver. “Tripoli also assisted with the highimpulse (Class 3) flight package that had to show the FAA an analysis of my rocket’s flight profile, charts of possible crash sites, what the dispersion pattern of wreckage should be if there was an accident, and prove that the risk of injury and death were sufficiently low,” say Deville.
Tripoli also made rocket hobbyists’ life easier when working with one of the most-common rocket fuels, ammonium perchlorate composite propellant (APCP). “APCP had long been classified as an explosive by the Bureau of Alcohol, Tobacco, Firearms, and Explosives, making it difficult to work with. For instance, you needed a license to handle APCP, an approved storage magazine to house it, and the magazine had to be inspected,” says Deville.
But Tripoli spent 10 years suing the ATF, trying to get APCP reclassified. About two years ago, it won that legal battle and ATF declassified the propellant so that a license and permit are not required to use it.
Keeping it simple
Deville designed Qu8k with a single rocket motor and nozzle to keep things simple. With a cluster flight, or rocket with several engines on a single stage, it becomes challenging to get all the engines to ignite and shut down at the same time. If they don’t, the resulting asymmetric thrust veers the rocket off its straight-line course. “With those risks, I shy away from cluster rockets, though I’ve built and launched many in the past,” says Deville.
Because the goal was to reach the highest altitude possible, Deville wanted his rocket to fly straight up. One way to ensure stable, straight flight is to have the rocket spin or roll. This evens out aerodynamic loads and tends to make the rocket fly straighter.
“But high roll rates give you horrible videos because of the camera spinning around with the rocket,” notes Deville. “And one of my objectives was to get good video of the flight. So I designed Qu8k as a low-roll rocket, one that would spin about four to five times in 90 seconds.”
To do this, he and his team carefully aligned the fins when they were welded in place. In fact, they used CNC to construct a jig from medium-density fiberboard to hold the fins square to each other while being welded onto a thin-walled cylinder that fit over the motor tube.
The rocket had several electrical components, including dual timers for parachute ejection, three video cameras, and four independent GPS units. To keep the layout simple, Deville had separate battery packs for each. The GPS devices, for example, drew power from lithium power packs, which Deville was unsure would work at altitude.
“We tested the pack in a vacuum chamber at our shop to make sure the lower pressure, nearly a vacuum at 100,000 ft, wouldn’t hurt their performance,” says Deville. “I expected them to swell and possibly open circuit or drop in voltage. And although they did swell somewhat, they did not lose charge and continued working.”
Other electrical devices relied on N-style disposable batteries.
Taking a mechanical approach
With most amateur-built rockets, the friction-fit nose cone pops off from the pressure generated by a timed black-powder charge. The charge goes off based on calculations of when the rocket will reach its apogee or highest point. But most amateur rockets don’t soar to 100,000 ft.
Qu8k couldn’t use this method for several reasons. First, the rocket would run out of thrust before it got to its targeted altitude, so it would, in effect, “glide” there. But when the thrust abruptly stopped and acceleration slowed, the nose cone’s momentum might pull it off the rest of the rocket before the rocket reached 100k ft. And second, if the nose cone stays on, any air trapped behind it would remain at atmospheric pressure as the rocket climbed. Meanwhile, outside pressure would drop and no longer push down on the nose cone. So the trapped air would be like an inflating balloon pushing on the nose cone.
“Assuming air pressure at 100,000 ft is negligible, the 15-psi air trapped inside would push on the 50-in.2 bottom surface of the nose cone with 750 lb of force,” says Deville. “So I had to vent the small area behind the nose cone. This meant I could not trap the pressure generated by a black-powder charge to release the nose cone.”
Deville fell back on his mechanical background and devised an approach never used before, a piston-cylinder actuator. Two timers (one for redundancy) were each set to activate an electrical match, a simple device that uses an electric current to ignite an explosive charge. The matches were set to light a black-powder charge inside the sealed 2-in.-diameter piston. At the right time, the matches lit the charge, and the resulting 200 psi of pressure extended the piston and pushed off the nose cone.
“We were able to ground-test this subsystems and it worked perfectly,” says Deville.
Deville also took a novel approach to building the nozzle, a critical component that must withstand 1,000-psi gases flowing at Mach one and 3,000°F.
“I used a piece of graphite for the nozzle, which is traditional, but put it in a machined phenolic carrier,” says Deville.
Deville and his team machined a nozzle with thin walls, which reduces thermal expansion. The thin cross section also means it heats up more evenly, so there is little differential expansion, which creates high internal stresses and can lead to cracks. The phenolic carrier mechanically supports and isolates the nozzle from the motor casing, shielding it from heat.
Payloads and problems
Qu8k also carried a couple of payloads into the atmosphere. A GPS unit broadcasting the rocket’s position on the 70-cm ham-radio band let Deville and his team monitor rocket position, though the GPS lost positional lock as the rocket climbed. But data resumed when the rocket began riding down under the parachute. “The GPS let us drive right to the touchdown point,” notes Deville.
The other payload was a cosmic-ray detector (Geiger counter), which was part of the Symbiosis Foundation’s Ergo project. The project’s goal is to equip at least a thousand classrooms across the globe with similar detectors to record and analyze high-energy cosmic rays coming from space. “The rocket carried one of their data packages, which consisted of a 5-in.-square circuit board and some daughterboards, and it was a challenge getting it to fit in the nose-cone area,” says Deville. “But apparently it detected cosmic rays with eight times the strength of readings taken at ground level.”
Qu8k carried other GPS units that also failed, a sticking point when it comes to claiming the $5,000 Carmack Prize. It stipulated that proof the rocket climbed to at least 100,000 ft would come from GPS, but all three units lost positional lock during the climb. And there are two theories as to why this happened.
The first postulates that the crystal oscillators used in the units to measure the time between signals received from GPS satellites was disturbed by the 15-g takeoff. This altered the timing, confusing the software, and making the GPS unit stop working.
The other theory, one Deville believes is more likely, is that the high speed of the rocket and the resulting Doppler shift confused the GPS circuitry doing periodic calculations. It couldn’t accept the high speed or vast difference in position between readings. “This could probably be remedied with the right software, but it would likely be expensive,” says Deville.
Still, Deville had other proof that his rocket met the Carmack Challenge. Experts could tell it climbed to at least 100,000 ft based on acceleration data recorded onboard the rocket. And flight times matched simulations of it traveling to that height.
The Qu8k team
Jorge Pinos and Angel Fernandez: Endless machining
Guy Kress: Launch tower
Greg and Rowan Mayback: Financial, moral, and physical support
Bret Ranc: Launch support
Korey Kline: Inspiration and design review
Carlos Rivera: Road tripping to Pitt Tripoli Pittsburgh: Motor transport
Al Bychek: BRB, tracking, and launch support
Chuck Rogers: Simulation and load calculation
Miguel Hernandez: Late-night support and heavy lifting
Jim Harper: Logistics Marc Devits: Electronics support
Ky “The-Rocketman” Michaelson: Parachute
Michael and Danah Kirk and Ed Ampuero: Propellant casting
Syntheon LLC: Machining and material support