First flight of a minimum-diameter & supersonic rocket
Post Launch Assessment Review Slides: https://docs.google.com/presentation/d/1Rhy48ysssNaJ00C0Y2lKBGgRjSm6Qu-QznzRDb1azWU/edit?usp=sharing
Systems Architecture Overview 3
Mission Concept of Operations Overview 7
Conclusion and Lessons Learned 8
System weights and measurements 10
Mindi is STAR’s first minimum diameter rocket, designed to be a one-semester project to optimize for apogee while practicing the necessary engineering techniques for the team’s future minimum-diameter projects, such as the lower stage of a future multi-stage rocket. The target apogee is 15,000 feet, and to this end the rocket is flying a K1103 Propellant X motor. Onboard electronics include an Altus Metrum TeleGPS for tracking purposes, 2 Eggtimer Apogee Deployment Controllers to deploy the accompanying Jolly Logic Chute Release, and 2 Mini WiFi Switches to arm the avionics from outside the rocket. The airframe consists of 54mm thin-walled fiberglass tubing, a 54TAC precision CNC-machined tail cone and an accompanying rail guide from the Wildman Mach 2 kit. The fins were cut from ⅛’’ fiberglass stock, secured to the rocket through a combination of JB Weld Steel and a robust fin glassing procedure, and manually airfoiled. Our electronics were mounted into a 3D-printed avionics bay, which was screwed onto the forward end of the motor retainer. This motor retainer, into which the motor’s forward closure was screwed, was epoxied into the airframe. An ⅛’’ aluminum annulus was epoxied into the airframe above the avionics bay, through which the assembly was inserted and screwed in, and an ⅛’’ aluminum bulkhead was screwed onto the annulus to close off the avionics section of our rocket. Mindi was designed to separate at apogee and fall freely until a height of 600ft, at which point the Jolly Logic Chute Release would deploy a 30’’ parachute. This deployment system was designed to expedite the physical recovery process for the rocket, as deploying a parachute at such a height would cause great drift. The duplicate electronics were intended for redundancy to ensure safe recovery.
Space Technologies and Rocketry (STAR) is affiliated with the University of California, Berkeley, and supported by the Associated Students of the University of California (ASUC), Boeing, Tameson, Procter & Gamble, Bay Area Circuits, Dassault Systems, Northrop Grumman, Technetics Group, Seamless Tanks, and Graphite Store. This project was a joint effort between members of our airframe and recovery subteams.
Mindi was initially envisioned as an experimental airframe project. The airframe challenges in question are primarily twofold: 1) mounting and reinforcing the fins to support the rocket through its high-velocity flight, and 2) retaining the motor. A third challenge which became clear during the manufacturing process was inserting and securing components within the 54mm airframe.
To address the first point, we sought a way to ensure a rigorous mounting procedure for Mindi’s fins given the impossibility of traditional through-wall mounting. For phase one, we applied JB Weld Steel to the fin root chords and pressed them to the outer airframe. A 3D-printed jig was designed to ensure perpendicularity with the airframe, proper alignment and equal spacing. For phase two, we applied a 2-layer fiberglass cloth tip-to-tip layup with West Systems 207 as the resin and hardener of choice.
Motor retention proved easier than expected. We purchased a 54mm minimum-diameter motor retainer which screwed onto the motor’s forward closure, and then was installed by inserting the assembly into the airframe with JB Weld Steel applied in a ring at the desired location.
For the third challenge, a variety of jigs and special tools were designed and purchased to help us reach long distances down Mindi’s body tube. This included a pair of tongs, a specialty screwdriver, and various long, thin sticks.
The recovery system on Mindi is fairly straightforward. However, our team has never worked with the Jolly Logic Chute Release before, so this was tested by our recovery subteam and redundancy was incorporated into our avionics system to ensure the intended operation. As the rocket was designed to separate at apogee and fall freely for around 14,000 feet, the parachute also had to be selected to be able to deploy at such speed and remain undamaged – the ramifications of this will be discussed in a later section. Black powder charges were used to separate the rocket at apogee, and controlled by the Eggtimer Apogee Deployment controllers.
Since we knew our rocket would be going transonic during flight, we could not rely on our usual simulation software of choice (OpenRocket) and turned instead to engines which are particularly suited for simulating trans- and super-sonic flights. RasAERO II was the software we used to model our general airframe setup. The only involved parameters were the total weight of the rocket and motor, all external geometry (body tube, fins, and tail cone), basic recovery deployment altitudes, and the K1103X thrust curve. From this information we were able to generate altitude, drag, and stability plots, to name just a few. We mainly used this software for its apogee estimate, as well as to demonstrate the importance of surface finish on the rocket, as varying the “smoothness” parameter to increase roughness lead to an apogee loss of up to 3,000 feet. We addressed this by developing a robust procedure for our paint job. We also discovered that including the tail cone increases the apogee by up to 1,000 feet.
As Mindi was intended as an experimental project, the diameter of the body tube was arbitrarily determined; choosing such a small diameter later posed some interesting engineering challenges. Due to the high forces experienced during flight, we knew that body tube materials such as blue tube would not be suitable for the vehicle, and turned instead to fiberglass. Although we originally planned to purchase fiberglass components from Apogee Rocketry, a contact of our team offered to sell us parts from his Mach 2 kit, and after conducting a cost comparison, we eagerly accepted, gaining in the process a filament-wound, thin-wall, 35-inch 54mm body tube. The main factor in the improved cost-benefit was that we were able to use the RMS 54/1706 motor casing on loan (granted no damage) rather than purchasing it for around $200.
We wanted our shear pins to be as small as possible, so that they would better conform to the rocket’s outer surface and not add tremendously to drag. We selected 2-56 nylon shear pins, since by using their reported maximum and minimum shear strengths in PSI we calculated a shear force range of 93-137 lbf for later use in black powder calculations. The PSI range for our shear pins fell well within the fiberglass tensile strength range of 45,000 - 50,000 PSI so we were confident that we could use 2-56 shear pins without damaging the rocket’s airframe. Three corresponding holes were drilled into the top of the body tube, halfway down the nose cone coupler, to secure these.
Two vent holes were drilled into our avionics section (discussed in Avionics, below) to allow the vehicle’s internal pressure to equalize with the external atmospheric pressure. Considering an atmospheric pressure of about 101.325 kPa at ground level and about 57.2 kPa at our target apogee of 15,000 feet, and combining this data with our body tube diameter of 54mm, or about 2 inches, we obtain an upwards force of about 80 lbf on our nose cone with no vent holes. This is not enough to shear the shear pins, but the primary purpose of our vent holes was to provide our altimeters with accurate enough static air pressure to effectively report altitude.
The rocket was recovered with no tube delamination or failure, which validated our design choices.
Our 13’’ fiberglass nose cone also came with the Mach 2 kit, and it featured an aluminum tip with an eye-bolt which we epoxied in with JB Weld Steel, and to which we then attached the shock cord.
Due to the nose cone and body tube being a part of Wildman’s Mach 2 kit, we felt confident making assumptions that these rocket components could handle Mach 2 speeds and the accompanying forces, so we focused more attention on the vehicle components which we were designing and manufacturing, which is what follows.
Our fins were water-jetted from ⅛’’ fiberglass. We determined the dimensions, such as root chord and tip chord length, based on what yielded a favorable stability (one-tenth of the length-to-diameter ratio, about 1.8 calibers) on the OpenRocket simulation software.
Fin dimensions
For the first step of securing the fins to the airframe, we designed a jig to support the fins throughout their root chord length, to ensure that they were perpendicular not only to the outside of the airframe tube, but also to the lower edge of the airframe itself. Additionally, this 3D-printed jig helped us ensure that each of the 3 fins was 120° from the next. After applying JB Weld Steel to the fin root chords and attaching them to the airframe under jig, we were able to reach into the jig’s slots to make fillets at the fin bases. In retrospect, the height of the jig should have been increased to allow our fillets to run the length of the fin root chords, but in order to stay on our timeline we did our best with the jig which was printed. CAD is available in the Appendix.
Fin jig on the rocket, pre-epoxy
For the second step, we glassed the fins. To this end, we sized 3 cuts of cloth for each pair of fins – two to cover the fillet areas, and one piece which extended from one tip chord to the other, running along the area of each fin and the section of body tube separating them. We lay down the corresponding sizes of fiberglass cloth with 45° translation between layers, where each layer was also separated by a thin layer of West Systems 207 Resin and Hardener, brushed on with a paintbrush. This was our first time performing this procedure on a rocket and we practiced extensively before glassing Mindi. Considering that the fins only showed some paint damage post-launch, we consider this effort successful.
After glassing, the excess cloth was removed with a dremel. We then manually airfoiled the fins with an orbital sander, ⅜’’ from the fin edges, leaving 1/16’’ of fin thickness to ensure stability. This proved to be a difficult task, since our fins are so small, and we used a dremel in conjunction with sandpaper to ensure a smooth finish. They are far from professional airfoils but we decided to include them to further increase apogee and reduce drag on the fin section of the vehicle.
Post-airfoiling with West Systems applied above fins to patch glassing
Additionally, we used the fin simulation software AeroFinSim to ensure that our fin design was robust enough to handle the forces which it would be subjected to during flight.
AeroFinSim results
We aimed for our simulated velocity not to exceed either the divergence or flutter velocities, (3098.59 ft/s and 4215.16 ft/s, respectively) both of which would indicate fin instability. From our RasAERO II simulations, the maximum in-flight velocity was 2200 ft/s, which is comfortably under the two critical velocities.
We used the RMS 54/1706 casing in conjunction with our K1103X motor. Since we did not use the motor’s ejection charge to deploy a parachute, we had to purchase a plugged forward closure, which featured a threaded hole compatible with our motor retainer. On the aft end, our tail cone acted as an aft closure. Additionally, we initially planned to use the aluminum seal disk included with the motor casing, but upon research we discovered that the high burn temperature of the K1103X motor would damage this disk, and we sought a stainless steel alternative instead.
To assemble the motor, we first cleaned the inside of the casing with acetone and water. We then sanded the casing interior and the outsides of the 4 grains to create more surface area for our glue to bond to. As per recommendation from Aerotech, we used Elmer’s Glue-All Max. We then applied lube to the O-rings for the tail cone and forward closure and let the grains set.
A cleaning and inspection post-launch confirmed that the casing and hardware incurred no damage.
Our recovery team selected a 30’’ octagon-shaped nylon parachute with a shroud line length of 30’’. In retrospect, after analyzing flight data from our TeleGPS, Mindi was falling at roughly 100 ft/s at the time of chute deployment at 600 feet from the ground, and this parachute was not rated to handle such forces. In fact, selecting a parachute to withstand such a high snatch force is near impossible and we should have re-evaluated our drogueless design in light of this. This will be discussed more in the Lessons Learned section.
Mindi flew one Altus Metrum TeleGPS, 2 Eggtimer Apogee Deployment Controllers, and 2 Mini WiFi Switches. These were selected mainly due to their size, which made them the simplest recovery electronics available. Our avionics bay design took several iterations to perfect, and we settled on essentially mounting the boards and the batteries required to power them to a long block with a 10-24 threaded rod on the bottom to screw into the motor retainer, and a handle at the top to aid in the insertion process. See Appendix for the CAD model.
The paint job for this rocket proved non-trivial. Ensuring a smooth finish after all of our hard work was essential, so we developed a schedule to apply primer, paint, and then gloss to the rocket in several coats, with the appropriate amounts of time between each to allow for complete curing.
Design & Manufacturing
Due to the small diameter of the rocket, the airframe tube also serves as the motor mount tube. This eliminates the need for centering rings.
Attaching fins will be different from other rockets built in the past
Will need to be attached to the tube since there will not be slots for them
JB Weld on the root, then make fillets with 30 minute epoxy reinforced with carbon fiber threads. Fin glassing will be done for maximum support.
Pre-launch Preparation and Arming:
Vehicle is prepared for flight on or before the day of launch. Preparations end with the arming of energetic systems while the rocket is on the pad.
Ground Test
Recovery systems are tested for functionality. Minor modifications for testing are allowed.
Flight Preparation
Recovery systems are prepared in a flight configuration.
Arming
Vehicle is loaded onto the rail and armed for flight.
Ignition and Boost:
The motor is ignited and the vehicle exits the launch rail with an appropriate velocity. It later enters a powered, stable flight.
Coast:
The motor burns out after 1.6 seconds and the rocket continues an unpowered, stable flight.
Recovery:
Apogee
Vehicle separates.
Descent and Chute Deployment
Jolly Logic Chute Release deploys parachute at 600ft.
Landing
The vehicle descends while drifting an acceptable range for the launch site. It lands with an acceptable amount of kinetic energy. An additional parachute may be deployed to meet drift and kinetic energy requirements.
Drift radius must be less than 2500 ft in 20 mph wind.
Each component must not land with greater than 75 ft-lbs of kinetic energy.
Tracking
Vehicle is located and recovered
Designing, building and launching Mindi proved to be a wonderful learning experience for our team.
As our design process happened over the summer, there were some choices that were made and then never revisited during the school year; we simply operated as if these were absolutes. For example, selecting a 54mm diameter was arbitrary at first but not questioned during our design reviews. In the end, this proved useful because our members, particularly new members, were able to confront and think their way around the challenges of securing components into such a thin tube, but we could have avoided this at least in part by selecting, for example, a 98mm body tube. Additionally, after we chose to use the Jolly Logic Chute Release, we should have done the necessary calculations to discover that the parachute would end up deploying at an unsuitably high speed. Instead, we mistakenly used our numbers for apogee deployment. The reasons for never revisiting this are more complex, though, and will be discussed in the Project management lessons section below.
Many of these issues could have been resolved had we had a more lenient timeline, but since we aimed to launch this rocket within a semester, we were often forced to work with what we had rather than iterating. For example, our fin jig for the first step of attaching fins was a bit too short, meaning that our fillets did not fully extend to the tops of the fin root chords. This was somewhat remedied by post-processing the layups, but having that thick, smooth fillet as a base would have helped us lay down fiberglass cloth and be more secure that bubbles would not arise during the glassing process.
On the day of launch, we discovered that our simulated center of pressure and measured center of gravity were about two times as close as desired (2 inches, which, divided by the rocket diameter, yields a starting stability of 1.0 calibers – too low). To mitigate this, we were forced to create a last-minute ballast of 160 grams, consisting of sand in a Ziploc bag, and secure this to the parachute shock cord. We calculated the appropriate weight using center-of-mass equations. This increased our measured on-the-rail stability to 1.75 and proved effective, but in the future the center of pressure and gravity should be marked during integration.
Left: ballast, Right: center of mass calculations on day of launch
Additionally, we extended our glassing cloth too far outside the fin leading edges, causing unfavorable geometry above the fins which had to then be sanded down and patched up with epoxy. This was due to a miscommunication and we will adjust our procedures accordingly.
Excess fin glassing cloth
The main lesson we learned from Mindi was the importance of communication. The cooperation of the recovery and airframe subteams on this project in particular was imperative and fell short at some times, as the project was headed by two airframe members. In the future, having one airframe lead and one recovery lead co-manage the project would be more beneficial. Additionally, having joint planning meetings throughout the course of the project and sharing a timeline, which we only did in the last month of the project, would be valuable from the beginning.
In general, though, all of our team members contributed their ideas, skills, and enthusiasm to the project. We are proud of what we were able to accomplish and that so many new members were exposed to the design, manufacturing, and launch process, and we look forward to what we can accomplish together next.
Airframe length: 45.5 inches
Airframe diameter: 54mm, ~2.126 inches
Fin span: 6.236 inches
Mass with motor: 7.2 pounds
Number of stages: 1
Propulsion type: Solid
Off-the-rail velocity: 80 ft/s
Target apogee: 15,000 ft
Projected apogee: 13,975ft (OpenRocket), 18,726ft (RASAero II)
Stability: 1.75 cal
Maximum velocity: 2,200 ft/s (RASAero II)
Preliminary FMEA
CAD
Avionics bay
Fin jig
