Summary

From Spring 2023 to Spring 2024, onboard camera housings were designed and fabricated for projects Medusa and Prometheus. As a full redesign of its predecessor, the Prometheus aeroshell 

Background 

The need for new housings was made apparent after the launch of Project Phoenix in the IAP prior.

These were the Phoenix aeroshells (images from 2022): 

And images from Phoenix Launch, after recovery (2023):


Two designs were made in parallel, one to hold the Runcam Split flight cameras and another to hold 808 Keychain Cameras. These aeroshells were 3D printed from SLA resin and bolted onto the mission package tube. After launch, multiple issues were identified in this design:

  • Many of the cameras turned off on the pad, so they did not record footage. Overheating was identified to be the most likely cause. 
    • The cameras were turned on and the rocket left on the pad for an unexpected length of time (~1 hour).
    • Since launch happened in January, the temperature outside was still chilly. The camera generates heat while operating, so the overheating most likely happened because its own heat was trapped inside the shell.
    • They may have also absorbed heat from the warm sunlight.
  • After recovery, some of the aeroshells were noticed to have a gummy or sticky surface texture. 
    • After launch, avionics placed another set of aeroshells outside in the sun, with cameras inside them. The experiment recreated the camera shutoff and sticky texture from launch.
    • The stickiness could be due to improper curing of the resin print.
    • Could also have been softening from in-flight heating (air friction, stagnation).
  • The aeroshells required access to the inside of the tube to mount, as well as space for nuts on the interior. This can obstruct the AV bay from being inserted into the rocket, and requires the aeroshells to be integrated before anything else gets put inside.
  • The shape is bulky, with many corners, flat faces, pockets, and protrusions that disrupt airflow. Phoenix had a significant (~10,000 ft) discrepancy between predicted apogee and actual apogee. The shape of these aeroshells likely exacerbated coning and increased drag on the rocket in a way that was not predicted in simulations.
  • The aeroshell cracked on impact with the ground. This is fine, we don't need to design things to survive hitting the ground.

With the new design, we hope to fix the issues of the previous while improving the aerodynamic performance and thermal robustness.

Preliminary Designs and Research

In Spring 2023, Summer Hoss '23 led aeroshell redesign sessions. The goals were to conduct research and establish a preliminary design. 

  • Research was conducted to narrow down a set of materials that provided sufficient thermal insulation while being reasonable to buy and fabricate with. These materials ended up being:
    • Nylon 66 3D print - somewhat more heat-resistant than PLA or SLA, though still runs the risk of softening. 
    • carbon fiber - very heat-resistant. Pure carbon fiber may soften from heat, depending on the thickness. May be difficult to work with, given the size of the aeroshells and the team's level of experience with composite materials.
    • Other composite materials - same as carbon fiber, but the team has limited experience with it.
  • Some design ideas were generated:
    • Incorporating a mirror so that the camera can be oriented differently from the direction we want a recording of.
      • In-flight vibrations were found to not interfere with this type of design.
    • Incorporating air pockets for increased insulation
      • Using aerogel
        • It's expensive
      • Using honeycomb structures such as Nomex
        • Adds a lot of thickness
    • Relocating the onboard cameras to inside the rocket.
      • People really wanted them on the exterior though.
    • Using the shape of the existing but unused aeroshell CAD model made by Chad Meier '23.

In Fall 2023, the design of the aeroshells became a collaborative effort between the Medusa Aerodynamics and Prometheus Structures subteams. Interested members cooperated to produce rough designs for the Medusa conceptual design review (CoDR) and preliminary design review (PDR), slides linked. These were designed for the 808 cameras, to be made with T300 standard modulus twill weave carbon fiber and withstand a maximum in-flight temperature of 1300°F. A negative mold would be made for them using foam and a CNC router. After preparing the mold, the carbon fiber would be laid up and vacuum bagged. Ideally, the part can be left to cure in a composites oven so that it attains a higher operating temperature. We obtained access to Prof. Zachary Cordero's composites oven in his lab for this purpose. Post processing would be done to create holes for mounting these to the mission package tube.


For the Medusa conceptual design review (CDR), the development of the aeroshells shifted to be more the responsibility of the aerodynamics subteam. Design changes were made so that the shells could be more easily fabricated and design details were further developed. Thermal simulations were attempted, but unsuccessful. The slides are attached here.

This design involves two plies of carbon fiber over a 3D printed frame, with heat-resistant paint applied to the exterior as an ablative coating.

Design Development

After the cancellation of Project Medusa, the remaining members of the Medusa Aerodynamics subteam (Vealy Lai '26, Conrad Casebolt '26, Ethan Wong '26) were absorbed into Project Prometheus and turned their primary focus to developing the aeroshells. With renewed efforts, a clearer design process was defined for this component.


Goals [TBD, KEEP WRITING HERE - VL 5/5/24 

The aeroshells, of which there are four (4), shall be mounted to the outside of the mission package tube on each stage.  Each aeroshell must contain and protect a camera which records the flight of the rocket.  The upward facing aeroshells shall capture the deployments of the recovery systems.  The downward facing aeroshells shall capture the ignitions and staging.

The aeroshell design must be streamlined in shape to minimize drag.

The upward-facing and downward-facing aeroshells must be aerodynamically similar to the greatest extent possible.

The aeroshell shall thermally insulate the camera from both the external thermal loads of in-flight drag and the stagnation temperature experienced by the rocket.

The camera must not overheat inside the aeroshell when left on (includes pad time and flight time).


Aerodynamic Design

The aeroshell is streamlined on both ends.  Its diameter is only minimally larger than the Foxeer Camera which it houses.



(Previous Iteration of Aeroshell IAP 2024)

Aerodynamic Testing

CFD analysis will be performed in the future.

  • Update 5/5/24: No aerodynamic or thermal simulations have been conducted on the aeroshells


Manufacturing (Nov. 2023)

The aeroshell mold is 3D printed using PLA.  A piece of heat resistant glass is cut to the required shape.  Three layers of carbon fiber are attached to the exterior of the mold with an epoxy layup using a vacuum bagging technique.

  • We performed this procedure for two first-iteration aeroshells, but the carbon fiber did not bind well to the outside of the aeroshell.  It is difficult to map the planar sheet of carbon fiber onto the steeply curved exterior of the aeroshell.  As a result, a negative mold may be used in the next iteration.

Thermal Testing

In order to recreate the thermal loads experienced by the aeroshell, the aeroshell will be exposed to the stagnation temperature of 350°F using a heat gun for one minute.  It will be mounted to a piece of fiberglass to emulate the exterior of the mission package tube.  

  • Thermal testing concluded that fiber glass is a better choice of composite for the Aeroshell compared to carbon fiber.  An identical fiber glass aeroshell maintained a lower internal temperature compared to a carbon fiber counterpart.

Current Design (April 2024)

Overview

The current design features two separate parts: The Shell and the Skeleton.  There is also a glass panel attached to the end of the viewing aperture and camera.


Shell


Skeleton


Manufacturing

Due to the non-planar shape of the shell, the only viable method of manufacturing is to use a negative mold with composite-epoxy layups sealed in a vacuum bag.  The general overview of the process is as follows:

  • Print a negative mold of the shell using PLA.
  • Perform three consecutive epoxy layups in the negative mold using three layers of fiber glass composite, abiding by the general procedures of epoxy layups.
  • Vacuum bag the negative mold and hold near/complete vacuum for 24 hours using a vacuum pump.


Detailed Manufacturing Instructions

Materials/Equipment Master List

    • 3D Printer
    • Sandpaper
    • Acetone
    • Wax
    • Gloves (and optionally a mask)
    • PVA release film
    • Vacuum bag plastic
    • Vacuum tape (double sided)
    • 1x1" squares of heat resistant glass 
    • Glass scorer (Metropolis has some)
    • Glass-breaking pliers (Metropolis has some)
    • Thin-woven fiber glass sheet (Gelb Lab)
    • Woven fiber glass sheet (Gelb Lab)
    • 2-part epoxy resin
    • Peel ply (Gelb Lab)
    • Breather
    • Vacuum pump
    • Four (4) M2 screws
    • Four (4) M3 or 1/8" screws and corresponding nuts
    • Dremel with cutting and sanding wheels

Obtaining and 3D Printing the Negative Mold

Materials: 3D Printer

    • Obtain CAD file of a negative mold of the exterior shell (YouTube tutorial for creating a negative mold linked here; CAD for this iteration of the Aeroshell has already been developed).
    • Convert the CAD to a .STL file. Print the negative mold on a 3D printer using low infill (10-20%) and material PLA.

Mold Preparation Steps

Materials: Sandpaper, acetone, wax, gloves, mask (optional), PVA release film

    • Sand down the inside of the mold to get rid of any burrs and smooth out surfaces. A few minutes of sanding will suffice.
    • Remove all the dust from the inside by washing and drying or wiping with acetone.
    • Rub down the inside and top planar surface of the mold with wax until glossy and smooth. A pea-sized portion of wax should be sufficient.
      • The wax is toxic if ingested, so wash your hands after use.
    • Put gloves on (and optionally a mask) and move to a well-ventilated area, preferably outside. Apply a coat of PVA release film by gently rubbing it in a layer over the wax. A bottle cap of PVA should suffice. 
      • Recommended to do this outside and stand upwind because plastic fumes are not good to breathe in. 
      • Visually, the layer should appear shiny.
      • Optionally: If you want multiple coats, wait until the first coat dries (drying time listed on the bottle).

Vacuum Bag Preparation Steps **@Vealy can you describe this process?**

Materials: Vacuum bag plastic, vacuum bag double-sided tape *Vealy what are these actually called?* 

    • Vacuum Bag Prep (with pleats)
    • Vacuum Pump Tube Prep
    • Leaving one side open


Epoxy & Fiber Glass Layups

Materials: Vacuum Pump, thin-woven fiber glass sheet, thick-woven fiber glass sheet, 2-part Epoxy Resin, peel ply, breather, gloves, mask (optional)

    • Cut one rectangle of thin-woven and two rectangles of regular-woven fiber glass (per Aeroshell) in a sufficient size to fit inside the negative mold cavity.
    • Cut a layer of peel ply slightly larger than the fiber glass rectangles.
    • Cut a rectangular layer of breather roughly equal in size to the upper planar face of the negative mold.
    • Put on gloves.
    • Thoroughly mix __grams of epoxy and resin in the correct ratio (per Aeroshell).
      • Once they are mixed, there is a limited time window in which this process can be successfully completed. We have found that after about half of the listed epoxy working time, it becomes stickier and difficult to work with.
    • Coat fully the inside of the negative mold with a layer of epoxy, using about 1/3 of the total epoxy allotted for one Aeroshell.
    • Lay the thin-woven layer of fiber glass into the negative mold cavity. Gently press/move the layer until it contacts the surface of the cavity and nearly every point.
      • This process is difficult and requires practice. Tips:
        • Start from the deepest point in the mold.
        • Provide excess material at first and flatten it out later.
        • Don't overstretch the material.
        • Use nails to lightly fit the sheet into sharp corners.
    • Repeat the above process two more times but using the regular-woven sheets of fiber glass cut out earlier and 1/3 of the remaining epoxy each time.
    • Press the precut sheet of peel ply into the mold covering all of the fiber glass.
    • Place the breather on top of everything else.
    • When this is done proceed quickly to the following section.

Holding a Vacuum 

    • Put the mold and layups into the open end of the previously prepared vacuum bag.
    • Close the vacuum bag.
    • Turn on the pump. The bag should quickly pull a vacuum and the shape of the negative mold will become visible.
      • Not recommended but sometimes necessary: The vacuum can be tested by turning off the pump briefly and observing how quickly air is let back in. Sometimes the pump has trouble turning back on if this is attempted.
        • If the bag is not holding a vacuum well at all, the manufacturing process may be compromised.
    • Leave the pump on for 24 hours.
    • When the process is complete, leave the pump on and cut a large hole in the bag near the pump tube. Allow the pump to run freely (not pulling vacuum) for a few minutes before shutoff.  

Removing the Aeroshells from the Mold

    • This process can be a little bit tricky, so try not to damage the Aeroshells in the process.
    • Separate the shells from the mold. There is no real procedure, so use the following tips:
      •  Try to separate by pulling up on the corners of the peel ply or the fiber glass itself. 
      • Use a blunt and thin tool, like a skinny screwdriver, to carefully disconnect the edges of the shell from the mold.
      • Support the sharp corners at the edges of the cavity of the shell as you pull the shells out. 
      • You will visibly see the areas which are free of the mold. Use these to continue to make progress on getting the shells out – they won't come out in one piece.

3D Printing the Skeleton

Materials: 3D Printer

    • Convert the CAD file into a .STL file.  Material as PLA and 0.2mm thickness are sufficient.
    • When slicing the file, change the following settings:
      • The side which touches the MPT should be normal to the print bed (even though it is not flat). 
      • Infill should be >70%
      • Supports set to ON EVERYWHERE.
      • Brim set to ON.
    • Make sure to remove all the excess material from the screw holes and supports after printing.

Cutting the Aeroshell to the Proper Size *Conrad return to this 5/16/2024

Materials: Skeleton, four (4) M2 screws, Dremel with cutting and sanding wheels

Glass Pane

Materials: 1x1" square of heat resistant glass, marker, glass scorer, glass-breaking pliers, sandpaper

    • Break the 1x1" square in half into two pieces.
      • Use a marker to draw a line divided the glass into two equal rectangles. Score along the midline on both sides of the glass. 
      • Carefully break the glass in two using the glass breaking pliers. A useful technique is to hang the excess glass corner over an edge and repeatedly hit (rather softly) the overhanging piece using the glass-breaking pliers.
      • In theory, each glass panel should yield two useful glass pieces.
    • Cut the glass into the correct shape to fit into the aeroshell.
      • Use a marker and ruler to trace the correct five-sided shape on the 0.5x1" glass panel
      • Score the glass along the lines using a glass scorer on both sides.
      • Carefully break off the excess glass using the glass-breaking pliers.
    • Sand down the edges until the desired dimensions are achieved.
      • Caution: Don't scratch up the glass surface through which the camera is viewing.

Attaching the Glass to the Shell *Conrad return to this 5/16/2024


Results

These Aeroshells flew on the Prometheus test flight in April 2024.  All four remained intact through subsonic flight speeds and an estimated apogee of 5,000 feet (**double check this CC 5/16/2024).  Only one Aeroshell carried a live camera, and the footage captured was deemed sufficient documentation of the booster recovery system deployment.  The camera did not overheat during flight or on the pad.


Integration

The two-piece approach necessitates the following integration steps:

  1. Mount the skeleton to the mission package tube using four (4) M3 or 1/8" screws and corresponding nuts, with the nuts inside **or outside? CC 5/16/2024** the MPT.
  2. Attach the camera to the skeleton and connect the camera to the avionics bay.
  3. Mount the shell to the skeleton using four (4) M2 screws.


Future Design Improvement Ideas (as of May 2024)

  • The exterior screws should be angled to keep the screw heads normal to the face of the exterior shell.
  • The exterior screws are currently threaded directly into the PLA of the skeleton.  A future design should add a threaded metal component on the skeleton in order to better secure the assembly together.
  • The aerodynamics of the up- and down-facing aeroshells are similar but not identical.  Perhaps a new glass shape could be designed, or non-planar glass could be used.
  • The cross-sectional profile could be further reduced by using a parabolic shape instead of circular, although a redesigned shell must retain adequate room for the structural supports of the skeleton.
  • The field of view could be increased. 

Future Manufacturing Process Improvements (as of May 2024)

  • The skeleton could be printed in two parts to get rid of the roughness on the face bordering the MPT.






  • No labels