Steerable Parachute Research

As many people who have chased balloons have discovered, finding exactly where a payload comes back to the earth is not a simple task. GPS telemetry has certainly revolutionized tracking and recovery, but it is still a pain if your payload ends up in the middle of a forest, lake, swamp, or other difficult-to-access area. Directing the descent of a payload to a predesignated landing area would greatly simplify things.

The use of steerable parachutes on both manned and unmanned payloads is nothing new. However, commercial systems that accomplish spot landings of payloads are rather expensive, and also much heavier than our balloons can lift (they are usually designed to land heavy cargo dropped from aircraft). We therefore decided to research the idea of using a small steerable parachute system for directing our payloads' descent.

A wide number of design considerations must be made for whatever system is chosen. For the first testing prototype, I decided that rather than use the ram-air square parachute design favored by sky divers and most automated designs, I would try a hexagonal 'parasheet' similar to what is used in model rocketry.

I chose the parasheet design because of its simplicity and low cost to build. Also, with fewer shroud lines, the theory is that there is less probability of entanglement during deployment. The ability to flare the 'chute and a long glide path are benefits in sky diving, but are not needed in our balloon work (and in fact they may cause problems with drift during a long descent). The 'sheet I built for this test was about 40" in diameter, cut from a plastic garbage bag and attached to the gondola with lengths of 20 pound monofiliment fishing line.

To give the parasheet a prefered direction of flight, I lengthened two of the shroud lines by about 2" as compared to the others. The length of these two longer lines were adjustable via a radio controlled servo in the test gondola. Thus, by varying the lengths of the two lines in flight from a radio transmitter, I should be able to steer the 'sheet during descent.

On March 21, we conducted drop tests of the first prototype from a three story metal tower next to the parking lot at the UND Aerospace complex. This tower used to have a radar antenna mounted on it, and was very useful for the tests.

Here are two shots of the payload during construction. The frame is a 1" diamater pine dowel about 8" long. Screweyes at each end of the dowel serve as attach points for the fixed shroud lines and guides for the two adjustable ones. On the near side of the dowel at the middle are glued the radio receiver and power switch. The servo is glued on the far side opposite the receiver. The lever attached to the servo wheel is made from several layers of corregated cardboard. The eyelet on the end of the lever is a bent paperclip. All attachements were made with hotglue. The battery pack (4 AA cells, made in two groups of 2 cells) had not been added when these photos were taken.

The completed gondola, mounted in its protective styrofoam cradle, is ready for drop testing.

Here the test rig is being hung out over the edge of the tower with the aid of the "drop stick". The drop stick is a length of 1"x2" pine with a clothespin glued to one end. A length of string runs from the clothespin down the stick via screweyes. This allowed the test rig to be hung out away from the tower and dropped without anyone having to risk their safety by leaning over the railing. Since we only had about three stories to test the droped rig, we wanted to have the 'sheet deployed at the moment of drop (rather than rolling it up and throwing the rig off the tower).

We conducted several drops during the first test. Due to payload weight and parasheet size, we only had about 3 seconds of flight time per drop to test the system. No fancy maneuvers were possible during these brief flights, but the parasheet did respond by turning the direction we expected it to when the appropriate shroud line was shortened by the servo. Pulling the right rear line made the rig spin right, the left line to the left, and keeping the controls centered the rig went fairly straight.

These first positive results are encouraging. More testing needs to be done by dropping the system from a higher altitude (we plan to use a tethered balloon and a radio controlled release). The relativly small parasheet will be replaced with a larger one for future tests as well. We also need to determine what control inputs are needed to steer the system without having it spin wildly.

The end goal will be to have the flight computer on a payload determine headding and direction via GPS and compass readings, and tell a servo to adjust the direction of descent by controlling the parachute. A designated landing site will be programmed into the computer before the flight, and the system will try to land the gondola on that site. If the system can be made to work, it will simplify recovery operations considerably.

More updates to follow

Return to the UND HABP webpage.