Post-Launch assessment Review

Transcription

1 Post-Launch assessment Review Iowa State University Cyclone Student Launch Initiative April 27, 2018

2 Iowa State University Post-Launch Assessment Review TEAM SUMMARY School: Organization: Iowa State University Cyclone Student Launch Initiative 2 LAUNCH VEHICLE SUMMARY Name: Size: Mass: Launch Motor: Payload: Altitude: Day This Side Up 117 in long, 6-in diameter lbs. Aerotech L2200 Target detection system composed of five downwardfacing Pi cameras 5680 ft 3 VEHICLE DIMENSIONS Total Size: Nose Cone: Parachute Bay 1: Parachute Bay 2: Avionics Bay: 117 in long, 6-in diameter throughout 33 in long 28.3 in long 12 in long 21 in long Flight Bay: Computer 12 in long Motor Mount: Recovery System: 34 in long 18 in diameter drogue chute (deploys at apogee) 120 in diameter main chute (deploys at 800 feet above ground level) 4 VEHICLE SUMMARY Our rocket utilized a modified form of our design from the previous year and retained the same basic layout. The main sections of the rocket are: the nose cone, parachute

3 Iowa State University Post-Launch Assessment Review bay 1, avionics bay, parachute bay 2, flight computer bay, and motor mount. The nose cone was selected to minimize drag and aid the rocket in achieving its target altitude. Parachute bay 1 is located below the nose cone and contains the main parachute. The avionics bay is in the fuselage between the parachute bays and houses the rocket s altimeters as well as the experimental equipment. Parachute bay 2 is aft of the avionics bay and contains the drogue parachute, which deploys at apogee. The flight computer bay contains the rocket s flight computers which govern the air braking system and is located aft of parachute bay 2. Lastly, the motor mount is located at the very aft end of the rocket. The rocket diagram is shown in Figure 4-1. Figure 4-1 This Side Up Diagram 4.1 NOSE CONE The nose cone is a 5.5:1 Von Karman design. We selected this shape after considering several options including the 4:1 and 5:1 Ogive shapes used in previous designs. The Von Karman nose cone is the optimal shape for the velocity and flow regime we expect to be operating at. It is 33 inches long with a base diameter of 6 inches. This diameter extends through the rest of the rocket as there are no variations in the diameter of the airframe. 4.2 PARACHUTE BAYS Parachute bay 1 is located between the nose cone and the avionics bay. It is 28.3 inches long and is large enough to contain the main parachute as well as the coupler shoulders from the nose cone and avionics bay. Parachute bay 2 is located between the avionics bay and flight computer bays and contains the rocket s drogue parachute. It measures 21 inches to provide sufficient space for the parachute as well as the coupler shoulders from the two bays on either end. 4.3 AVIONICS BAY The avionics bay is located between parachute bays 1 and 2. It measures 12 inches long and contains the altimeters and on-board experiment. The altimeters, batteries, and other avionics electronic components are mounted to wooden plates secured by

4 Iowa State University Post-Launch Assessment Review threaded steel rods affixed to the bulkplates on either side of the bay. The avionics bay also houses the experiment s Raspberry Pi s and cameras mounted to the exterior of the bay through camera ducts. 4.4 FLIGHT COMPUTER BAY The flight computer is placed between parachute bay 2 and the motor mount section. It measures 12 inches and contains the rocket s flight computer, GPS system, an accelerometer, and a servo and pulleys that control the rocket s air brakes. Like the avionics bay, the electronic components are mounted on a wooden sled attached to a pair of metal rods, which are themselves connected to the bulkheads at either end of the bay. 4.5 MOTOR MOUNT The final section of the rocket is the motor mount. The section is 34 inches long to accommodate the fin can. The motor mount features eight fiberglass primary fins, four carbon-fiber airbrakes, and four secondary fins printed of ABS plastic which house the pulleys which actuate the airbrakes. The chosen motor for CySLI s competition rocket was the Aerotech L2200. While the team had initially considered both the Aerotech L2200 and the Cesaroni L1115, the final decision to use the L2200 was made due to the mass requirements. 5 VEHICLE DATA & RESULTS Overall, our rocket s launch and recovery systems worked properly. The airbrakes did not function during flight because a small error in the code caused altitude and acceleration readings to be half of their actual values (Figure 5-1). This broke all our fail safes put in place for flight ceiling, as well as the apogee prediction function. On launch day we did limited testing of the systems outside of power cycling them to make sure they initialized successfully. More diagnostics should be run before a launch to confirm sensor connections and readings are nominal. In the future, we will be doing more ground tests with all airbrake- and payload-related programming and embedded systems. With data gathered from launches this year we can feed data into our flight computer and see how it reacts as it would during a launch. Having multiple members reviewing the code running in onboard systems would also be beneficial in catching any last-minute errors in any codebase changes, especially those made close to launches. Standardized tests should be made to evaluate the braking systems after software or hardware changes.

5 Iowa State University Post-Launch Assessment Review Figure Printout from flight computer (Note Filtered Altitude is half of Height) Our recovery system worked perfectly, and both the main and drogue parachutes deployed as expected. This was an improvement over our test launch, in which our main parachute did not inflate. The successful competition descent confirmed that the earlier issue was most likely due to tangled lines, which was corrected with good packing technique. We estimate that the rocket drifted 1320 to 2640 ft from the launch pad, although it dragged a significant distance after landing. The estimated distance is within our mph calculated range (shown in Table 1-1), and competition wind speed was about 10 mph. Total Drift Wind Speed (mph) Drift (feet) Table 1-1 Estimated Drift 6 PAYLOAD SUMMARY The experiment we selected for our launch was the Target Detection System. The payload bay of our rocket, which was a 12-inch coupler section, contained all the hardware for our target detection system. This included five Raspberry Pi 3 boards and accompanying battery packs. Each Pi board was connected via ribbon cable to a separate Pi camera. Five statically mounted Pi camera V2s were spaced to encompass the entire field of view. The combination of Raspberry Pi and Pi camera V2 allowed us

6 Iowa State University Post-Launch Assessment Review to differentiate and process the images captured in-flight. In addition, the upper bulkhead plate contained two different altimeters wired to the ejection charges. The target system was programmed to save video with images of targets marked in green. In addition, the upper bulkhead plate contained two different altimeters wired to the ejection charges. The target system was programmed to save video with images of targets marked in green. 7 PAYLOAD DATA & RESULTS Our payload was the target detection system. Before the competition, a few improvements were made to bump the frames per second (fps) from 15 to 30 fps. This new code was not put through the same tests as the old and as a result, the camera exited with an error at an hour and a half, just before the launch. The camera exited because the OpenCV function to write video can only write up to two gigabytes on a 32 bit system. This did not come up in the original testing because we ran it for two hours at 15 fps which did not result in a file bigger than two gigabytes. We believed that the video writer would keep writing video until the microsd card was used up which was about six gigabytes. 8 SCIENTIFIC VALUE This year we created a color detection and differentiation in Python on a Raspberry Pi using a Pi Camera that runs at 30 frames per second. Figure 8-1 is a screenshot of a video from a Pi that is using this algorithm. This figure clearly shows color detection and differentiation. The algorithm works very well, and we are confident that it would have worked at launch if the memory of the SD card had not been filled. Figure 8-1: Color detection and differentiation Also, our airbrake system works as demonstrated during our test launch. Mechanically, the rocket is sound and only code failures prevented all systems from working properly. Our recovery system worked as expected, and we feel that our rocket design and predictions have been validated over the course of the year.

7 Iowa State University Post-Launch Assessment Review CySLI updated our airbrake code and made significant advances towards the creation and implementation of a visual recognition system. This constitutes the bulk of the scientific value introduced by the team over the school year. 9 VISUAL DATA OBSERVED This Side Up launched beautifully (Figures ). The drogue chute deployed at apogee and the main parachute appeared to deploy at 800 ft, as expected. This is clearly visible in Figure 9-4. On landing, the rocket was dirty but undamaged, and it was ready for relaunch after cleaning (Figure 9-5). Figure This Side Up on launch rail Figure Rocket on launch

8 Iowa State University Post-Launch Assessment Review Figure Rocket in flight Figure Recovery configuration at apogee and around 800 feet

9 Iowa State University Post-Launch Assessment Review Figure The rocket was undamaged on landing 10 LESSONS LEARNED During the competition rocket build, we had a small hiccup where we were temporarily unable to connect to the Raspberry Pi's. We reset the connection and everything continued with the build. This put us back minutes and as a result we were rushed to get the rocket to the RSO tent. We should have had a backup connection method ready in case of failures. Just because we have never run into an issue does not mean it will not happen.

10 Iowa State University Post-Launch Assessment Review Our failures in avionics can also be traced to not retesting after changing code. Next year, we plan to make a list of tests and expected outcomes that need to be run after any change before launch. Throughout the year, the entire team learned about rocket design, software and hardware, CAD modeling, safety, and working with teams. All members of CySLI gained valuable experience in engineering, communication, and problem-solving. The launch team also enjoyed meeting other teams during Launch week, interacting with students at the Rocket Fair, and touring NASA. 11 PROJECT SUMMARY Our initial plan was to construct a vehicle that would successfully detect three colored tarps during its ascent. The Pi cameras onboard were functioning perfectly, however due to issues detailed above, we only managed to record the process of setting up the rocket and placing it onto the launch rail. Another function that did not perform as expected was the airbrake system as described above. However, launch was excellent and the recovery system worked properly. We followed safety protocols and had a safe launch. Even with the numerous setbacks experienced during launch, the team felt that the rocket was a resounding success. We learned many aspects of rocket design, avionics and computer control systems, and the importance of preparation and safety. We look forward to applying this knowledge to next year s project. 12 EDUCATIONAL OUTREACH SUMMARY 12.1 FIRST LEGO LEAGUE CySLI participated in the outreach activities at the FIRST Lego League Regional Competition in Webster City, Iowa on December 2 nd. The team brought the rocket from the competition Cardinal Heavy to the FLL competition and spoke to the students about high-powered rocketry and Student Launch Initiative. The team engaged with some 50 students over the course of the day IOWA STATE UNIVERSITY 4U PROMISE CySLI participated with ISU s 4U promise at King Elementary School in Des Moines, Iowa on February 21st. The team brought the rocket from the competition Cardinal Heavy to the event along with a small wind tunnel to demonstrate the effects of drag on different shapes. Small Styrofoam airplane assembly kits were also distributed and assembled with the students on a crafts table for those interested. The team engaged with about 160 students on college night educating on basic aerodynamic

11 Iowa State University Post-Launch Assessment Review principles, airplane arts-and-crafts and basic high-powered rocketry through our previous years designs, videos, and models. 13 BUDGET CySLI is funded by the joint contribution of the Iowa Space Grant Consortium and Iowa State Universities Make-to-Innovate program. Through funding applications to both ISGC and the M2I program, CySLI was granted a total budget of $10,800 for the NASA SLI competition period. Progress and milestone reports are presented to M2I faculty and filed to the ISGC department office throughout the competition period on major milestones and achievements. Any funding that remains unspent by the end of the term will be returned and refunded back to funding organization(s) FUNDING Iowa State University Make to Innovate $5, Iowa Space Grant Consortium $5, Total: $10, BUDGET SUMMARY Rocket Team $3, Avionics Team $ Experimental Team $ Travel $4, Total: $9,227.33