Before I began building amateur rockets, I built and flew model rockets, starting in the summer of 1971. Looking for a greater challenge, the following year I decided to begin building my own rockets, entirely from scratch. Although my first rocket was a simple one, not equipped with a parachute, I was soon working on my second, and larger rocket, which I wanted to recover by parachute. It seemed natural, therefore, to base the parachute recovery system for this rocket on a concept that was similar to that used for model rockets, specifically, the concept of using an "ejection charge" with a time delay, to deploy the parachute. Unlike model rockets, however, I used an ejection charge that was entirely separate from the engine, a self-contained unit containing a small charge that was electrically ignited. Instead of a time delay, I utilized a pendulum switch, similar to the method I read about in the book "The Amateur Scientist", which was based on the naive notion that, as the rocket would "turn over" at the peak of the trajectory, the pendulum would fall over, closing the switch contacts. Needless to say, if the parachute did successfully eject during these early flights, it did so almost immediately after engine burnout. Thinking that the pendulum was simply too sensitive to angle, I tried a mercury switch, which needless to say, led to the same futile results. It wasn't until I began to understand the concept of free-body motion and the consequences of aerodynamic drag, that I realized that such a method is doomed to failure. Any such inertial type of switch will close immediately after burnout, as the rocket begins to decelerate at a rate greater than that due to gravity alone, as a consequence of drag. The pendulum, mercury, or any other inertial mass, however, decelerates at a rate dictated by gravity alone, thus moving forward relative to the rocket body to which the switch is attached, thereby closing the switch contacts.
By this point in time, I had developed a reasonably effective method for ejecting , or deploying, the parachute from the rocket (even at high speed, a benefit from the trials and tribulations experienced !), and it was simply a matter of ejecting it at a slow enough speed such that it was not torn off immediately!
The solution seemed to be an air-speed switch, that is, a switch with a spring-loaded external flap or vane that would deflect due to aerodynamic drag. As the rocket would slow down near the peak, the spring would return the vane to the undeflected position, closing an electrical switch. This would then trigger the parachute ejection charge. This method proved to be, after several design iterations, quite successful. The eventual design incorporated a backup timer, in case the rocket veered significantly from vertical during ascent, and would consequently not slow sufficiently to close the air-speed switch (due to horizontal component of velocity). As well, to help overcome this possible problem, the air-speed switch was put in series with a short delay timer (2-3 seconds), which would allow for a higher speed setting of the air-speed switch. The short delay would allow the rocket to slow further before triggering parachute ejection.
The goal of developing a highly reliable parachute recovery system was never fully achieved. Although the method developed for ejecting the parachute from the rocket was very reliable, as was the air-speed switch for sensing the point in trajectory when the parachute should be released, the problem with reliability laid in designing an electrical circuit for linking these two actions. In hindsight, it does not seem like a difficult problem, and any difficulties were likely a result of trying to be too fancy in the circuit design. And besides, electronics never were my greatest expertise !
After engine burnout, a rocket continually slows down as it ascends skyward, succumbing to the effects of gravity and air resistance. Whether the rocket is launched vertically, or at some initial angle, it reaches its minimum velocity at the peak of its trajectory. It is at this point that ejection of the parachute is normally desired. Under certain circumstances, it may be desirable to have the parachute eject sooner, or later during the flight. Such would require a parachute and rigging system designed for higher speed deployment. For my system, the goal was to have the parachute deploy at the peak of the trajectory. An air-speed switch (Figure 2), mounted inside the rocket, is fitted with a flap that extends out of an opening in the fuselage. When the rocket is at rest, the flap extends outward nearly horizontally, under spring action. At high speed, aerodynamic force causes the flap to pivot and fold flush against the fuselage to provide minimal drag. As the rocket slows near the peak altitude, the spring force overcomes the drag force, extending the flap to an angle where a microswitch closes, triggering a short-delay timer. The precise speed at which this occurs is determined by the adjustable spring tension setting.
Prior to launch, the rocket is, of course, at rest, and therefore the air-speed switch is closed. To overcome this predicament, the air-speed switch is put in series with a mercury switch (Figure 3). This serves as an inertial type switch which maintains an open circuit while the rocket is at rest. Once the rocket begins to decelerate at engine burnout, the inertial mass (mercury) moves forward, relative to the rocket body, and closes the switch contacts. Since the rocket continues to decelerate at a rate greater than that due to gravity (as a result of drag), the mercury switch remains closed. With such an arrangement, the only time that both switches are closed at the same time is when the rocket slows sufficiently as it approaches the peak.
The air-speed switch is typically set to close at a speed of about 80 mph (130 km/hr). Since this is too high a speed to safely deploy the parachute, the short-delay timer provides for a 3 second delay to allow the rocket to slow further before triggering parachute ejection. After this delay, the electronic circuit causes a micro-relay to energize, which closes the circuit which provides electrical power to the ignition plug.
The Ignition Plug in screwed into the bottom of the Ejection Cylinder, and serves the purpose of igniting the ejection charge. The ignition plug has two (for redundancy) nichrome wire filaments which glow red hot when sufficient electrical current is passed through. Surrounding the filaments is a small primer charge of black powder, contained in a thin plastic sleeve which fits snugly at the top of the plug. Black powder is used since it ignites readily. Once ignited, the primer then ignites the main ejection charge, consisting of granulated propellant contained in the ejection cylinder. Granulated propellant is prepared by coarsely grinding, using mortor & pestle, freshly melted propellant (left over from casting the grain). Typical particle size is 1 mm. Granulated propellant is used as the ejection charge for three main reasons:
1. Very rapid burning
2. Re-usability. Relatively low combustion temperature allows use of aluminum components, and cleanup of burnt residue is simple using warm water
3. Large volume of white smoke is produced, forming a highly visible cloud when the parachute charge fires (excellent for tracking)
The burning charge generates high pressure in the cylinder, forcing the piston with attached nosecone, cradle and parachute, out of the rocket. Note that the cylinder bracket pushes (reacts) against the forward perimeter edge of the fuselage while the cylinder is pressurized. Once the piston is fully out of the cylinder, the pressure is immediately exhausted, and the cylinder assembly is pushed aside by the cradled parachute as it is pulled out of the rocket.
In order to allow initial pressure to build up in the cylinder to produce rapid burning of the charge, and to eject the nosecone with sudden force, I found it was important to secure the nosecone to the fuselage with strips of masking tape. I found that without the tape, the nosecone would simply move out of the cylinder as soon as the charge began to burn, failing to fully eject. Ground testing demonstrated that three strips of 3/4" (1.91 cm) masking tape worked best, with each strip having an average tensile strength of 13 lbs (58 N.). This allows the cylinder pressure to build up to a pressure greater than 100 psi (7 atm.), facilitating rapid burning.
Immediately after the parachute cradle is pulled from the fuselage, the springs on the cradle disc rod force the cradle arms apart, allowing the disc to separate from the cradle, completely freeing the folded parachute. Airstream motion then causes the parachute to unfold and blossom. A high strength braided steel cable tethers the parachute to the rocket fuselage. The nosecone, ejection cylinder assembly, and cradle disc each have a light gauge steel cable tethering to the parachute, such that all components are recovered for reuse.
The purpose of the Ejection Cylinder (Figure 6) and Piston / Rod assemblies (Figure 7) is to provide a mechanism for ejecting the nosecone and attached parachute cradle. The hollow aluminum nosecone, besides acting as an aerodynamic fairing, serves as a compartment for the ejection cylinder assembly. Screwed into the inside top of the nosecone is the piston and connecting rod assembly. When assembled, the ejection cylinder slides into to nosecone, mating with the piston. The ejection cylinder bracket engages the slots cut into the side of the nosecone flange.
The nosecone is machined from a single piece of aluminum bar stock. The cylinder is made from a length of aluminum tubing, with the end plug (threaded to mate with the ignition plug) machined from brass bar stock. The end plug is retained by four 1/8 in. brass pins soldered in place.
The ejection triggering circuitry is shown in a schematic in Figure 4 . The purpose of the circuitry is to supply electrical power to the ignition plug at the precise moment determined by either of the two sensors-- the air speed switch (primary system) or the mercury inertial switch (backup system). Note that the circuit incorporates two mercury switches-- one in series with the air speed switch, to maintain an open circuit on the ground, and the other to trigger the backup timer immediately after burnout. The timer module consist of one 556 dual timer chip or two 555 (or 7555 cmos) single timer chips. The logic control module contains a series of gates and flip-flop chips which interpret the output from the timers and supply a signal to the relay driver at the proper moment.
Problems I had with the circuit related to the detail design, not the basic concept as shown. Problems included:
1. Relay driver. Since the relay was a subminiature type, the coil current was relatively high, requiring a driver. I tried an SCR then MOSFET for the driver . The SCR seemed to be too sensitive, triggering for no apparent reason. The MOSFET seemed to work better.
2. False triggering. The relay seemed to be causing the problem. I tried putting diodes in parallel & series with the relay to dissipate the stored coil energy, with limited success.
Perhap an opto-coupler would have been a solution to these problems (to isolate the relay circuit from the rest of the circuit), but I never got around to trying this.