When I went to the high power launch that renewed my interest in rocketry, Tom Kaye mentioned that he and Dave Zupan had been planning to put a GPS receiver in the rockets that would feed serial data back by radio link.  Not only would this give a way of recording the entire flight path, but also of finding the rocket more easily after it lands.  When you're spending a couple hundred dollars in fuel to launch a rocket that costs a few hundred dollars, you don't want to lose it because the wind carried it too far away. 

The problems they faced came with feeding the data over radio.   They were trying to feed RS232 data which works best bi-directionally, rather than in a single one way transmission, and may have been at a data rate higher than the radio bandwidth could handle. 

A solution I suggested was using a compact microcomputer to transcode the GPS data into DTMF (touch-tone) and another computer on the other end to decode the DTMF signals.  I had recently gotten into microprocessor technology for use in paintball airguns, and this seemed another natural application. 

That got me to thinking, why not use the GPS data to plot a return course to the launch pad, and a steerable recovery system controlled by the computer to guide it there?  Project Icarus was born.

In high school I had experimented with limp wing parachutes after reading about the designs of Francis Rogallo which had been considered  for use on the Gemini manned space flight program.  Rogallo's limp wings were basically delta wings that deployed like parachutes and were steerable.  Early models used metal frames, and you can see one being tested HERE.  I'd built a few limp wings, and even used them successfully with model rockets.

The first stage of project Icarus is to test different limp wing designs for ease of deployment, and effectiveness in steering.  Initial tests will be performed with wings dropped from a second story balcony with payload, and deployed from an Estes Mongoose rocket (the yellow rocket in the picture at right).  Once I've determined effective controls by flights with set control positions (for example with the proper lines shortened to produce a yaw to the right), I will move on to the second phase.

Phase II will involve computer control.  Either an R/C Servo, or muscle wire cylinders (they are elegant, compact and simple, but I suspect they are too slow to be effective) under control of a microprocessor will steer the wing and rocket in preset patterns such as squares and figure eights.  Phase II testing is planned for flights in an Estes Super Nova Payloader (the blue rocket in photo - note top half of clear payload section and nosecone are cut off in photo).  Optionally, Phase II may culminate in the use of light sensors to steer the wing into the wind minimizing the drift away from the launch point, using the position of the sun as a geographic reference.

Ultimately phase III will combine a GPS receiver with the steering technology from Phase II, and include the development of software that will return the rocket to its point of origin automatically, regardless of prevailing winds.  The software will ideally detect launch by the increase in elevation, and and begin steering once the wing has been deployed, keeping the wing and expended rocket on course to return to the launch point.  Alternately the rocket may be carried to a landing zone, that location recorded into memory, and then launched to land on target in target landing rocket competitions.  I had expected the the testing of the Phase III system would take place with ballistic launches from an aircannon, as the equipment will probably be too bulky for model rockets, and I do not plan on licensing and funding high power launches. Once through initial trials a system would be sent to Zupe and Tom to put through high power testing.  Through recent web-searches, I have learned of medium power rockets, and it is now most likely that the main test vehicle for Icarus Phase III will be built from a Launch Pad kit.