The Construciton of the All-PVC Amateur Hybrid Rocket Engine

Evan Daniel

Last Updated March 12, 2006

Back to main page

Construction

For simple operation of the engine, the chamber needed to be detachable, but firmly sealed to the aft end of the tank (which would form the injector bulkhead). For PVC pipe, the easiest way to do this was to use NPT threaded fittings. Due to the incomplete engagement of the threads, the female threads will get some exposure to hot combustion gases. So, the female threads go on the combustion chamber, and the male threads on the tank.

The hardest construction problem to solve is the injector mounting. A compression fitting is needed for the UC valve, which means that the tank needs an NPT fitting to connect to the compression fitting. What would be ideal would be a PVC fitting with a socket weld fitting on one end, 1-1/4" NPT male threads on the outside of the other end, and 1/2" (or 3/4" or 1") female threads inside those. Unfortunately such does not exist.

The solution is to use stacked sections of pipe. 1" pipe fits inside 1-1/4" with a small clearance; fitting 1/2" inside 3/4" and 3/4" inside 1" requires some preparation of the ends and the gentle use of a sledge hammer or press. As long as the joint is made long enough (L/D > 2 seems sufficient), the gap between 1" and 1-1/4" pipe can be filled simply by using excess PVC cement.

Click to enlarge	Full size image

The final assembly for the injector then consists of a piece of 1/2" pipe with a female adapter on the end, inside a piece of 3/4" pipe, inside a piece of 1" pipe. This whole assembly is glued into the aft end of the tank, and then the 1-1/4" male adapter is glued on over it, so that the ends of the two threaded fittings are flush with each other. After all the glue has dried, a 3/8" compression x 1/2" NPT brass fitting is installed as the injector. The gap between the two PVC fittings is filled with silicone caulk or a plastic epoxy to prevent hot gas from circulating and damaging the tank fittings.

Click to enlarge	Full size image

Click to enlarge	Full size image

The fore end of the tank is much simpler -- a 1-1/4" socket weld end cap is sufficient. The UC-valve fill system requires a vent at the top of the tank, which is most easily crafted by simply drilling a small hole in the side of the tank, about halfway up the end cap. I use a 1/16" drill bit for about 3/4 of the depth (drilled on a drill press, though a hand drill would suffice), and then finish the hole with a #72 bit (0.025"), drilling by twisting the bit between thumb and forefinger. A drill press would be preferable, but the press I have access to has enough wobble that it breaks drill bits that small. (Note that if you plan to hydro test your tanks, which I recommend, you will want to drill the hole after testing.)

The last piece is the chamber. The forward end consists of a 1-1/4" female adapter, a short length of 1-1/4" schedule 40 pipe, and a 1-1/4" coupler. The pipe should be sized so that the adapter and coupler are flush against each other. This section is to provide a space for the ignition grain to reside. The bulk of the chamber is a length of 1-1/4" schedule 80 PVC, which provides both the chamber pressure wall and the fuel grain.

The nozzle throat is an SAE steel washer, epoxied onto the inside of a 1-1/4" x 3/4" socket weld bushing. This is then glued into a coupler and glued onto the back of the chamber. The expansion section could then be formed into a smooth cone or bell with the use of epoxy or concrete. Early testing has been conducted without this present to simplify construction, with the hope that recirculating exhaust gases will form a usable expansion cone in a manner analogous to an aerospike nozzle turned inside out. Performance differences between the nozzles will be investigated in the near future. Obviously none of the nozzle materials are capable of withstanding the temperatures of the combustion gases; however, with short burn times, they survive well enough to withstand a single burn, and that is all that is required.

The nozzle, after firing:

Click to enlarge	Full size image

The aft end of the tank and the chamber (after firing; the cleanly cut fill tube / injector can be seen here):

Click to enlarge	Full size image

Hydro testing

For safety, hydro pressure testing is recommended. The plumbing required to do this is relatively straightforward. The main item is a hydraulic cylinder (Surplus Center sells a nice simple 3/4" bore single acting cylinder for $20; you'll need an SAE x NPT fitting for easy plumbing as well), which will be compressed with a bar clamp, jack, or similar tool to create the pressure. Next, a reservoir of hydraulic fluid and a manual valve is required. For the reservoir, I use a 6" length of 1-1/2" PVC pipe with a screw cap on one end and an end cap on the other, with the end cap drilled and tapped to take a 1/4" NPT to compression fitting adapter. One could of course use couplers and bushings instead, but since I had the tap this was cheaper and easier. The valve is a surplus needle valve rated to 3000 psi. Connected to all this is a pressure gauge and an output port which goes to the vessel under test.

I fill the test vessel with water (cheaper and more environmentally friendly than hydraulic fluid), and install a check valve between it and the hydro testing apparatus. McMaster-Carr carries a wide variety of such valves; I used a compact ball check valve rated to 3000psi. The check valve is important for keeping the hydraulic fluid dry, and also allows for a simple manually operated pump if there is too much air in your system or the vessel under test expands too much -- just release the bar clamp, open the valve to the resevoir, pull the cylinder out to refill it with fluid, close the valve, and clamp it again to pump more fluid into the vessel.

You'll want to bleed most of the air out of the system before use for simplicity and safety (compressed air stores energy; compressed liquids generally don't). This can be accomplished by working the cylinder back and forth and tapping the lines to get bubbles into the reservoir. Having everything plumbed using translucent high pressure nylon tubing makes this fairly straightforward.

Finally, the system is pressurized to the desired pressure with the bar clamp. I tested my tank to 900 psi without incident.

Ignition grain

The UC-valve ignition system uses a solid propellant grain to both preheat the combustion chamber and also to cut the fill tube and form the injector. Any small solid propellant grain and igniter combination should work just fine; this merely the one I chose for cost and simplicity.

For simplicity, cost, and safety reasons, a solid propellant formula of 80% Potassium Perchlorate (KClO₄, purchased from Firefox) and 20% Epoxy (Mr. Fiberglass medium cure rate epoxy) was chosen. Potassium Perchlorate propellants have a reputation as having highly pressure dependent burn rates, making them difficult to use for solid rocket motors. Here, however, that is not as much of an issue -- the ignition grain does not have enough burning area to pressurize the chamber significantly, and the feedback cycle of increasing burn rate causing increasing pressure causing increasing burn rate is highly damped in the hybrid motor, where most of the propellant mass flux comes from other sources.

I decided against HTPB or PBAN binders because of toxicity (HTPB uses isocyanate curatives) and simplicity (PBAN requires an elevated cure temperature). Potassium Perchlorate won out over other oxidizers because it burns at a higher temperature than Potassium Nitrate, and ammonia based oxidizers (Ammonium Perchlorate / Ammonium Nitrate) react with the polyamine hardeners in the epoxy. Chlorates and other pyrotechnic oxidizers were avoided for reactivity and other reasons.

The ignition grains are hand mixed, and cast into 3/4" PVC end caps. These make a convenient casting mold and inhibitor, and fit nicely into 1-1/4" PVC pipe. While the mix is still wet, a pair of resistors are pushed into the surface, near the edge (far enough apart to leave room for a 1/2" hole for the fill tube to fit through). I use 1/8 W, 10 Ohm resistors from Mouser.com. These will fail and catch fire in < 1s when connected to a 12 volt supply.

After curing, the grains are drilled to fit around the fill tube. A 1/2" hole for a 3/8" tube works well.

Click to enlarge	Full size image

An ignition grain, before and after use.

Lastly, the surface of the grains should be painted with a high temperature ignition paste to ensure even ignition (the resistors don't reliably light the grain otherwise). The mixture I use is 6 parts Potassium Perchlorate, 1 part Aluminum powder, and 1 part Red Iron Oxide (Fe₂O₃).

Safety note: mixtures containing powdered metals and oxidizers are known as flash powders, and can be highly sensitive to ignition, burning very rapidly or exploding if confined. Proper safety precautions for flash powders should be observed.

The powder is then mixed with Weldwood Contact Cement to form a paste, and then painted on the flat face of the grains.

The cavity to hold the ignition grains is too large, so a support will be needed. A small ring of 1" PVC pipe works well. It should be sized to hold the top of the grain against the injector (about 0.6").

Fill tube

The fill tube is a 20" length of 3/8" OD high pressure nylon tubing, available from US Plastics and elsewhere.

The tube needs to be notched to reduce the thickness where it will be cut. The notch should be just aft of the face of the ignition grain, and should remove about 50-60% of the wall thickness of the tube. The easiest way to notch the tube appears to be to put it on a lathe (or in a drill press chuck, or a hand drill rigidly mounted), and spin the tube while removing material with a triangular file.

Final Assembly

The final step is preparing the motor for flight.

First, drop the PVC spacer ring into the chamber to hold up the ignition grain.

Next, the fill tube and ignition wires are run through the chamber. Next, install the ignition grain (resistors / ignition paste facing aft) around the fill tube and attach the wires (wiring the resistors in parallel). Don't solder the resistors; just twist on the wires (soldering near highly flammable rocket propellants is asking for a Darwin Award). Then slide the grain into the combustion chamber, as far as it will go against the spacer ring.

Then, the fill tube is inserted into the compression fitting and the compression fitting tightened. Use care not to over-tighten the fitting, as this will constrict the injector reducing flow and also weaken the tubing. I have been unable to find specs on how exactly it should be tightened, but one full rotation after the nut is hand-tight seems to be approximately correct (though less may suffice).

Finally, apply thread sealant to the tank threads and thread the chamber onto the tank. I've been using high temperature silicone grease for this. PTFE (teflon) sealant pastes or tape could be used here, but they will probably release toxic fumes thanks to the fluorine content, and so should be avoided. A light coating of grease on the brass fitting will help protect it from combustion gases and reduce wear.

At this point the motor is ready for flight; it just needs to be connected to a nitrous fill system and ignition system. Lots of other people have described these, so I won't do that here (for now). I will recommend the use of quick-disconnect fittings on the nitrous line (the Whitakers launch site uses McMaster-Carr PN 6537K53 male plugs (high-flow style, 1/4" size)). This allows all compression fitting plumbing to be done at the flight line or before arriving at the launch, and then only plumbing QD fittings at the pad.