Projects‎ > ‎

Filament Winder

posted Dec 22, 2013, 5:39 PM by Sean M. Messenger   [ updated Apr 6, 2014, 4:04 PM ]
In the summer of 2013, fresh out of our custom rocket manufacturing experience for E80, Taylor suggested we design and construct a filament winder for general use by MARC (Mudd Amateur Rocketry Club) members. This would allow construction of composite tubing for various projects. For example, our rocket was a vulcanized cellulose tube wrapped in fiberglass, but could have been much lighter and stronger had we been able to do the wrapping ourselves instead of just epoxying a sheet of fiberglass fibers to the outside of the tube.

We did initial research and stumbled upon many designs that had already been implemented. It had already been determined that ones available on the market would cost on the order of thousands of dollars to buy. Most of what we ran across were semi-industry level projects with a few senior design projects. In short, this was not something that there are tutorials readily available for. 

The basic design consists of a rotating mandrel and a linear guide head. The rotating motor winds a filament around the mandrel as the guide head produces a weave pattern along the length. By varying the rotation and feed speed, different patterns can be obtained. The tension in the fiber as it is wound around the mandrel, density of the fiber toes, pattern, and number of layers ultimately determine the strength of the final product. The fibers are set in place with epoxy and dry as a single tube that can be slid off the mandrel for immediate use.


1. Closed-Loop Motor Controller (Sept. 2013 - Dec. 2013)

The first step towards the final project is implementing a way to finely control the rotation speed of the mandrel. Stepper motors are typically the ideal choice when accurate position (or steady state velocity) are desired, but tend to be much more expensive and provide less torque than conventional brushed motors. We decided to go with the cheaper solution, as the accuracy from the stepper motor could easily be reproduced using an encoder on the motor shaft. 

This portion of the project was completed as a final project in the course ENGR 155: Microprocessor Systems - Design & Applications. This was our first experience integrating two pieces of hardware plus an input in an embedded microprocessor, so we anticipated it taking a bit of time and having a learning curve. In the end of this portion, we wanted the motor controller to stand alone, drive a brushed motor at a given set speed, report that speed, and stand entirely on its own. That is, no computer controlling it -- everything was to be programmed on the microprocessor and the only input was a wall plug for power.

Please read these three status reports:

2. Concluded

Unfortunately, Taylor and I decided we would not be pursuing the filament winder further. It was a fun project where we learned a lot in the first phase, but as a whole have other interests that we would prefer to place our time in. This would have been a very large project that required a significant amount of time. 

Started: August 2013
Finished: NOT