Composite Filament Winding
Composite Filament Winding
Embedded Files
As Advanced R&D Intern at bluShift Aerospace, I worked alongside another intern, Sam Bonauto, to expand the companies manufacturing capabilities by developing a composite filament winding program to produce rocket components such as fuselages and fuel tanks.
As Advanced R&D Intern at bluShift Aerospace, I worked alongside another intern, Sam Bonauto, to expand the companies manufacturing capabilities by developing a composite filament winding program to produce rocket components such as fuselages and fuel tanks.
Phase One: Assembly
Phase One: Assembly
The X-Winder desktop filament winder had been disassembled for storaged when it failed to produce useful parts in 2020. We reassembled it following the procedures provided by X-Winder inc.
Phase Two: 3" Tubes
Phase Two: 3" Tubes
We first tried to wind 3" diameter tubes to develop an understanding of the software and hardware.
Our first mandrel, created out of a cardboard poster tube, PVC endcaps, and $5 of hardware
The first test dry wind. The filament is slipping on the mandrel, ruining the pattern
The first successful wind with epoxy.
Phase Three: 12" Tubes
Phase Three: 12" Tubes
After building confidence winding 3" tubes we moved onto 12" diameter tubes.
12" tubes required more robust endcaps. We made these out of plywood. The blocks that extend beyond the edge keeps the endcap flush in the end of the tube
When winding 3" tubes we discover it was very difficult to remove the mandrel without damaging the composite tube. Here, Sam is applying packing tape on top of a layer of receipt paper. This will make it easier to remove the tube later.
Video of winding 12" fiberglass tube
Phase Four: Filament and Epoxy Delivery Troubleshooting
Phase Four: Filament and Epoxy Delivery Troubleshooting
We had issues with an excess of epoxy ending up on the final part. This photo shows a double wiper system we created to squeeze more epoxy out of the filament
This photo shows an incomplete part. The wind went perfect until the software said the part was done when the part itself was 1" from completion. This was fixed by trial and error to determine the issue with the software
We had issues with twists in the filament ending up on the tube. We needed to create a longer run of filament to allow the twist to "cancel" each other out. This was one attempt to lengthen the run of filament by have a fixed roller far from the winder itself. We ended up being able to achieve a long enough filament run on the X-Winder itself by mounted filament on a "tower" (see videos below).
Phase Five: Expanding the X-Winder Capabilities
Phase Five: Expanding the X-Winder Capabilities
Through a complete electro mechanical overhaul, we expanded the X-Winders capabilities from a maximum of 12" diameter tubes to 24" diameter tubes
Disassembly
Reassembled with wider cross bars and taller mandrel posts to accommodate the 24" mandrel shown in this photo
We added a second mandrel motor to provide the necessary torque to pull filament using the 24" mandrel
Phase Six: Winding 24" Tubes
Phase Six: Winding 24" Tubes
To allow for easier mandrel removal, we cut down the length of the mandrel and made it whole again using wood blocks. This allowed us to remove the blocks after the wind and compress the mandrel slightly.
Finishing Process. After winding we applied heat shrink tape to compact the weave, squeeze out excess resin and prevent air bubbles. Heat lamps and heat guns shrink the tape and maintain proper temperature for optimal tube strength
Video of winding our final piece of the summer. 5-foot long, 24" diameter carbon fiber tube.
Removing the tube from the mandrel still required large amounts of pulling force despite our release mechanism improvements. Part of the issue was that out mandrel surface was not perfectly smooth create friction between the tube and mandrel after ther tube cured.
Phase Seven: Transfer of Knowledge
Phase Seven: Transfer of Knowledge
At the end of the summer, I led a transfer to BluShift engineers and technicians to allow them to continue to use the X-Winder to transition bluShift to composite rocket bodies.
Page updated
Google Sites
Report abuse