A Great Team in front of the "Rattlesnake" launch vehicle and its payloads
NSL 2019 Launch
Launch Close-up
-
NASA Student Launch 2019 challenged university students to launch a rocket, recover it safely, then deploy a UAV payload
-
3 test launches validated the subsystems as they were developed and manufactured
-
8 months of extracurricular work culminated with the competition in Huntsville, Alabama
-
My club, the Northwestern University Space Technology And Rocketry Society (NUSTARS) placed 15th out of the 45 teams which competed
-
I led 7 engineers on the Payload Mechanical Systems Team
Render of the final payload CAD assembly (Siemens NX)
Render in context of the rocket payload bay and nose cone
REQUIREMENTS
-
Launch a rocket to 4500 feet and safely recover it
-
Deploy a UAV from the rocket and fly it to a designated zone to deliver a beacon
-
The beacon has to occupy a volume of 1 cubic inch
-
UAV flight can be RC or autonomous
-
-
Launch vehicle preparations at the launch site must take fewer than 2 hours
-
Payload weighs below 5 pounds
Beacon Design
Inspired by the logo for the Nintendo 64, I designed a beacon with the "N" geometry used by Northwestern sports teams. The beacon was 3D printed with PLA. Extras were printed for display at the competition and made great keepsakes for team members!
The payload was divided into 7 main mechanical components. The descriptions below are color coded with the diagram at the right. Note the UAV itself was built from a quadcopter kit which was handled by my peers in the separate payload electronic systems team.
UAV orientation system
-
IMU senses the orientation of the UAV
-
Stepper motor rotates the UAV skyward once the rocket landed
Pneumatic deployment system
-
2 Pneumatic pistons separate the UAV and nose-cone from the payload bay
-
16 gram CO2 cannister provides pressure to the system
-
Solenoid valve triggers piston extension once the nose cone unlocks and UAV is oriented
Aluminum support bulkhead
-
Secures the payload inside the launch vehicle
-
Reacts loads from the orientation and deployment systems
3D printed electronics sled
-
Provides mounts for the electronics which control the pneumatic and orientation subsystems
-
Minimized payload assembly steps required on launch day
UAV release mechanism
-
Tray and cushion protect the UAV from heavy vibrations during the rocket's flight
-
Torsional spring opens the cushion once pistons extend the UAV outside the payload bay
Nose cone locking mechanism
-
Connects the nose cone to the payload bay during flight
-
Unlocks once rocket lands
-
Actuated by a servo motor which has an independent power supply
-
First mechanism activated in payload deployment sequence
Beacon deployment mechanism (not noticeable in diagram at the right)
-
Unique Northwestern "N" beacon tied to a nichrome burn wire on the UAV
-
Once the UAV is positioned over the delivery zone, a current passes through the wire, burns the thread, and drops the beacon
The payload on display prior to competition
Section view of the payload bay and nose cone, as they would appear during the rocket's flight
Section view of the same systems upon UAV deployment
Brainstorming:
Development began with a few team brainstorming sessions. A few key questions guided the payload design process:
-
How is the UAV released from the rocket?
-
What keeps the UAV safely inside the rocket during flight?
-
How do we ensure the UAV flight path is unobstructed, regardless of the rocket's orientation upon landing?
-
-
Which payload parts can be left assembled prior to launch day?
-
How can batteries and other components be installed quickly at the launch site?
-
What must be inspected before launch & what spares/ redundancies are possible?
-
-
How are electronic connections packaged to avoid wire tangling?
-
What modifications to the launch vehicle are necessary for payload integration?
-
What materials & manufacturing processes are needed to execute each design feature?
-
Who [amongst the team] has experience with design and fabrication of similar mechanisms?
-
What is each team member passionate about developing further?
-
-
How might failure of a given component propagate to other parts?
In the previous year, NUSTARS had worked on a similar payload challenge, except with a ground-based rover in place of the UAV. Myself and three other teammates applied some lessons learned that year as we narrowed ideas down. I turned the leading idea into powerpoint art for the proposal we submitted to NASA which described our initial plans for the competition. The payload bay was at the top of the rocket. The nose cone would be detachable from the bay during the deployment process. A captive leadscrew would turn through a rear bulkhead to push the UAV outside the bay and orient it skyward.
Powerpoint art depicting the conceptual elements of the payload
First Test Flight / Nose Cone Lock:
The first flight test focused on the nose cone locking mechanism. It was the first system manufactured and tested; if it failed to retain the nose cone, then other valuable payload parts could hazardously jettison from the vehicle.
A servo motor rotates a grooved plate. The grooves in the plate drove pins inserted into spokes, which would extend through slots in the nose cone. While the spokes are extended, the nose cone cannot be separated from the rest of the launch vehicle. The videos below demonstrate this mechanism in motion
Nose Cone Lock prior to launch
Nose Cone
Payload Bay
View of the spokes outside the rocket
FEA of a spoke under load
Nose Cone Lock technical details:
Mechanism made from a stack of 0.25" thick plates of UHMW plastic
-
UHMW selected for high impact strength and low friction coefficient
-
Impact strength critical to handling load from the ejection charges which separate the stages and deploy the parachutes
-
200 lb-f was the expected value of this deployment force
-
Safety factor of 3.3 estimated for a spoke under max load
-
-
Design evolved from a similar lock used for the 2018 competition
-
Spokes were prone to slipping out and finicky to reposition
-
~6 hours to machine the lock in 2018
-
-
~2 hours to make everything for 2019 redesign
-
Water-jet cut plates
-
Spacers for the first flight test were turned from scrap PVC; this proved time consuming & spares were desired, so:
-
3d printed PLA spacers in the final competition version
-
-
Bulkhead is 0.5" thick UHMW to accommodate the 1/4-20 fasteners used to interface with the launch vehicle
Comparison of relevant properties for various plastics (source)
A similar lock used in 2018 had 3D geometry that required time consuming milling set-ups (left)
The 2019 version used stacked 2D plates instead so everything could be rapidly made with a water jet cutter (right)
Fastener Consolidation:
The 2018 payload required too many different fastener sizes and lengths. Fumbling through small boxes of 4-40, 4-48, 6-32, 10-32, 1/4-20, M2, and M3 screws, plus matching nuts, was frankly a huge pain in the assembly. Thus I made sure each teammate stuck to imperial fasteners and held a review to consolidate similar sizes.
To fasten electronics and other elements which would not see high loads, I bought a 100 pack of 1" long nylon 4-40 screws. Lighter than metal screws, they were also so soft I could cut them to shorter lengths with my swiss army knife. This saved considerable cost and effort of managing multiple screw sizes. When these tiny screws were occasionally misplaced, it was super easy to cut another to a required length and move on.
Exploded view of the Nose Cone Lock
Learnings from frigid first flight:
The flight test happened on February 9, 2019 in Three Oaks, Michigan, where the high temperature that day was 27 degrees Fahrenheit. The nose cone locking mechanism worked normally when initially assembled and tested indoors at room temperature. However, I had not accounted for the thermal contraction which would occur in the UHMW plate under the frigid temperatures which exist in this part of the country! UHMW has a relatively high thermal expansion coefficient of 1.1*10^-4 in/in/°F . The tolerance on the spokes were too tight (specifically parts in the blue box above). Fortunately, I always carry a swiss army knife and the UHMW plastic was relatively soft, so once this problem was discovered, I was able to disassemble the mechanism and shave a few thousandths of an inch off each spoke until the mechanism worked again. After this roadbump, the test was successful!
The locking mechanism and my swiss army knife
Flight test team bundled up for the frigid temperatures
A funny final detail: extra ballast was needed for this test flight to compensate for the missing weight of other payload components. We used corn found on the field of our test flight. In case you were wondering why corn cobs appear in various photos (like in my hand above and in the first photo on the page), that is why.
Second Test Flight / Active Drag System (ADS):
Altitude targets are often used to score collegiate rocketry competitions, so the club wanted to develop control over the rocket's apogee. It was easy to imagine tweaking the mechanism behind the nose cone lock to extend surfaces which induce additional drag during flight. Thus, NUSTARS felt adding the active drag system (ADS), would be a valuable addition to our competition entry with minimal additional effort. ADS functionality was the focus of the second test flight.
Render of the ADS electronics
ADS Mechanism test video
ADS Design, successful Ctrl + C, Ctrl + V:
The ADS was developed by a separate team from my payload group; however, its mechanical elements were similar to my payload team's work, so I also led ADS mechanical design and manufacture. The systems behave so similarly, that I just cloned the Nose Cone Lock subassembly in NX to make one for the ADS. The ADS was simpler because it did not need to interface with the UAV payload, nor did it need to be screwed into the rocket frame. Most dimensions were driven by an expression table and easy to change. The deleted plates left a few mates to reattach, but it definitely beat modeling everything from scratch! I <3 NX : )
The ADS was also made from the same 0.25" thick UHMW as the Nose Cone Lock. The ADS parts could have been thinner, but the slight mass penalty was not worth the cost of getting other material in different thicknesses (in the end the ADS weighed only 1.5 lbs, less than 2.7% of the total vehicle mass). The UHMW and aluminum parts were also cut with a water jet. A belt sander and files were used to remove rough edges to facilitate smooth motion in the mechanism. The battery bracket was 3D printed in PLA.
Finally I will mention that the groove geometry was different between the ADS and Nose Cone Lock. For control purposes it was convenient to maintain a linear relationship between the rotation angle of the groove and its resulting fin extension length. While the nose cone lock used a simple arc for its on/off behavior, the ADS used a spiral (specifically the Spiral of Archimedes) for its graduated extension behavior.
ADS drag:
For simplicity, the ADS would not deploy until the rocket motor burnt-out. At burnout, the onboard IMU would sense change in acceleration and activate the ADS (in case you were wondering why I was shaking the mechanism in the video above, that was an easy way to trick the IMU into triggering the ADS control). With altimeter data, the ADS microcontroller (an Arduino Teensy) would run a PID control scheme and extend the ADS fins proportional to any additional drag required to hit the final altitude target.
ADS Fin load
The fins were 2.75" wide and could extend 1" beyond the rocket body. The maximum speed at motor burnout was 530 ft/s, with simple drag analysis (F_drag = 0.5 * Drag Coefficient * air density * area * velocity ^2), this could result in a maximum drag load of ~250 lb per fin. That drag could induce a bending stress around 8.5 ksi, so I changed the ADS fin material to aluminum instead of UHMW (UHMW yield strength is only 4.5 ksi vs 35 ksi for 6061 aluminum). The sum of 3 drag fins could increase drag by ~140%.
Learnings from the second frigid flight test:
The ADS must align with slots in the launch vehicle's avionics bay. A sled holding altimeters that track the rocket's altitude during flight is mounted in this bay on a set of threaded rods. During the second test flight test, the ADS was positioned against a set of doubled nuts on those threaded rods (left image below). It quickly became annoying to turn and adjust those nuts, so the final version in competition used 3d printed spacers to position the ADS against the avionics bay (right).
Test Launch ADS positioned with nuts
Final ADS positioned with custom spacers (2 blue sleeves between the ADS and Black Sled)
The challenge of adjusting the ADS position cascaded into a series of unfortunate events. A wire needs to pass through the ADS so altimeters on the avionics sled can trigger ejection charges located at the bulkheads outside the bay. This wire was inadvertently pulled through the ribbon cable connecting the 2 ADS PCBs and ripped the connection.
Fortunately the electronics lead left header pins on the board to aid debugging, so connection without the ribbon cable was possible. Unfortunately, the electronics lead was not available for that flight day and thus no spare female-female jumper wires were packed with the team's launch equipment.
Having to launch in remote farmland, a 3hr drive from campus and over an hour from any store, this problem had to be solved with available materials. Some scrap wire and female-male header pins were left in a box of spare materials we brought, so in the 20 degree Fahrenheit weather, I got my gloves off and jerry-rigged the two boards together like shown at the right. A dirty solution, but it worked!
Last resort PCB connections
By the time I reconnected the boards, the battery died for the rocket's radio tracker. It was a cloudy day, so a functional tracker was essential to finding the rocket after landing. Thus the launch vehicle leads took apart the rocket and replaced the relevant battery. Then we got the rocket on the pad and found the avionics sled was on the wrong threaded rod, so it could not be armed through the existing hole (highlighted in green above). We came prepared with a hand drill, so a new one was made with the rocket on the launch pad. By the time this happened, the Nose Cone Lock batteries died (the cold temperature was a real enemy that day). So, we took the rocket off the pad, took the nose cone lock out, replaced the battery, put it back together, and were finally ready for launch! The launch was successful despite our hardships. The multiple cycles of assembly and dissassembly informed MANY new flight prep checklist items, which became useful for the competition.
Final Test Flight / Other Payload Systems:
Secure in the launch vehicle's functionality, the remaining payload elements could be tested. The final flight test was less eventful than the first two. Ground testing was effective for catching most issues in the efficacy of the payload mechanisms.
Orientation System:
The rocket lands in an unpredictable orientation. The UAV needs to be upright to fly. Thus a stepper motor and an IMU are used to sense the UAV's orientation and make corrections.
Because orientation is not time critical, a motor with a 1:100 gearbox was used to maximize torque. It was a bit overkill, but motor masses only varied by a few grams in this product category and this one happened to be the cheapest.
The rotor is held stationary, while the motor's housing and the other payload systems rotate. This prevents wires from tangling during the orientation process.
The golden rectangle outlines the motor used for orientation
All parts in the red rectangle rotate together while the disks on the sides are bolted into the body of the rocket
Pneumatic Deployment:
Once the IMU detects the UAV is oriented upward, a solenoid valve triggers, allowing pressure from a small CO2 canister to move a set of pneumatic pistons. This pushes the nose cone away from the payload bay and exposes the UAV to the air.
Pneumatic Deployment Test
A test of the pneumatic deployment system done on a bed to simulate the uneven terrain of the landing zone
Erroneously, the IMU was installed upside down, so the system oriented upside down too. The fix was simple but that test video had awful camera angles.
UAV Release:
During flight, the UAV needs to be secure to avoid potential damage from rattling around the payload bay. After landing, the UAV must be clear for takeoff. To accomplish this, a spring loaded hinge was attached to the tray holding the UAV. This hinge is actively pushing against the interior wall of the payload bay. Once the tray is deployed by the pneumatic pistons, the hinge swings upward. A memory foam cushion underneath the hinge protects the UAV propellers. This system also prevents unwanted rotation of the payload during flight.
Competition Payload Deployment
The payload deployment after the competition flight
The payload systems were designed for 2 hours of idle time between assembly of the rocket and landing, as specified by competition guidelines. Unfortunately the competition had several delays, so we were unable to launch for about 3 hours. In that time (and in the Alabama heat) the batteries drained and the pneumatic system lost pressure. Thus our payload was ultimately unsuccessful. This video starts after the pneumatic system stalled, so we pulled the nose cone off manually to demonstrate successful orientation and the UAV's survival. The UAV did not have enough battery charge to lift off.
ARTURAS'S RESPONSIBILITIES
Mechanical Engineer :
-
Owned design of the UAV orientation system
-
Owned design of the UAV release system
-
Created each red 3d printed part visible here
-
Spacers for aligning various plates in the payload bulkheads
-
Tray to hold electronic components for the pneumatic system
-
UAV Release hinge
-
-
Designed the 3D "N" beacon
-
Reviewed teammates designs for manufacturability and assemblability concerns
-
Managed fastener consolidation
-
Operated water jet cutter to cut the UHMW and Aluminum plates used in the nose cone lock and bulkheads
-
Cut slots in the launch vehicle for the bulkhead mounting bolts and nose cone lock
-
Assembled payload parts in the launch vehicle prior to launches
Team Lead:
-
Coordinate with electronics and launch vehicle teams to
-
Teach team members relevant CAD and machining skills
-
Set deadlines for payload deliverables and tracked them in a GANTT chart shared with the team
-
Owned top level NX assembly
-
Resolved mating issues and part discrepancies from parts made by teammates
-
Implemented file naming and assembly saving conventions to streamline collaboration between 7 users (our university software lacked a PDM system)
-
Produced the sweet cross-section views and renders seen on this page and in the project's numerous reports
-
-
Managed the payload's bill of materials
-
Organized documentation on the payload for reports required by the NASA Student Launch administrators
-
Project Proposal Report
-
Preliminary Design Review Report and Presentation
-
Critical Design Review Report and Presentation
-
Flight Readiness Review and Presentation
-
Note: This webpage primarily covers the mechanical design of the rocket's payloads, as they were my primary contribution to the team. The payloads were only possible due to the coordinated effort of my teammates, and co-leads. The team's success in competition is also due to the hard work of the Launch Vehicle and Electronics subteams, the NUSTARS co-presidents, safety officer, project manager, and chief engineer.
NUSTARS has been one of the premier engineering groups on Northwestern for several years, supported by the hard work of many talented individuals. Information about other NUSTARS teams, and the current club activities is available at http://www.nu-stars.org/