Embedded Files
BSME Senior Capstone | Portland State University | September 2024 to June 2025
As part of a 4-person senior capstone team, we designed and manufactured a new 240 N LOX/IPA rocket engine after tracing the legacy engine's combustion instability back to a single missed setting in the previous team's design process. My own piece was the thermal and propulsion analysis behind the new engine: heat-transfer modeling, nozzle design, and CFD validation.
The Problem
The Problem
Portland State Aerospace Society's Minimal Test Stand, MinTS, is meant to be the club's low-barrier, reusable platform for testing liquid-fuel propulsion, so getting it right would affect every team that used it after us, not just ours. Previous capstone teams had left the stand incomplete and unfireable, with an engine already mounted that was known to chuff, sputtering instead of burning cleanly, for reasons nobody understood. Project Lazarus took over the engine and combustion side of the rebuild, partnered with a second capstone team, Chicken Tendies, who handled the stand's piping and sensors. Our first task was to get the stand fireable so we could actually test the engine and figure out why it was chuffing.
The Goal
The Goal
Diagnose the chuffing, then design, build, and hot-fire test a new 240 N engine and injector module that doesn't have the same problem.
Approach
Approach
Before designing anything new, we recreated the existing engine in RPA using its original design parameters and got a nozzle sized for roughly 25 times its rated thrust, about the size of a dry-erase marker. Tracing that back, we found a single missed setting in the prior team's design process that told the software the combustion gas had infinite length to react in, meaning the nozzle effectively had no combustion chamber. That was the likely root cause of the chuffing all along.
With the bug identified, I took on the thermal and propulsion analysis for the new 240 N engine and injector module. I modeled all three modes of heat transfer through the combustion chamber wall in a MATLAB script, and analyzed the nozzle contour's exhaust flow in ANSYS Fluent. The injector and chamber geometry were designed to accept either a regeneratively cooled nozzle or a simpler heat-sink nozzle, so the team could manufacture and test the cheaper heat-sink version first to verify the geometry before committing to the more expensive DMLS-printed regenerative part.
Results
Results
Our CFD analysis validated the new nozzle design, matching predicted pressure and thrust performance, with only a small (1 to 3 percent) specific-impulse loss from visible Mach diamonds in the exhaust.
The heat-transfer analysis delivered the bigger finding: the bare, uncooled wall would melt in about 1.1 seconds, and regenerative cooling only extended that to 1.3 seconds, since at this scale the coolant can't carry heat away fast enough to matter. Regenerative cooling isn't viable for an engine this size, so we built the steel heat-sink nozzle instead. Our capstone wrapped up before the new engine's gaskets arrived; in April, once they came in, the next MinTS team test-fired it and it ran cleanly with no chuffing.
Rocket Engine Test FireRocket Engine Test Fire EnhancedChalleges
Challeges
Diagnosing the chuffing meant repeated failed hot-fires with no clear answer for a while: the first test produced four ignitions, and three of them just kept chuffing for their entire duration no matter what we tried, with only the fourth calming into a stable flame after we guessed the mixture was running fuel-rich and cut the regulator pressure.
Regenerative cooling looked promising on paper but the channels a 240 N regenerative engine would need are only about 0.5 mm wide. This is much tighter than the roughly ±0.2 mm tolerance that DMLS printing would allow, and PSAS had already watched a similar print fail once before.
LOX itself got harder to work with on hot test days: warmer weather sped up boil-off in the tank, stretching out fill times, and on at least one fire we ran short of propellant mid-test because we'd misread bubbling off-gassing as a sign the tank was already full, when it wasn't.
Skills
Skills
Rocket propulsion, heat-transfer analysis (Bartz equation), nozzle and combustion-chamber design, RPA and NASA CEA, ANSYS Fluent CFD, root-cause diagnosis, hot-fire test procedure, cryogenic propellant (LOX) handling, MATLAB.
Page updated
Google Sites
Report abuse