Hybrid Rocket Motor

This page details my experiments in building a hybrid rocket motor. Please do not attempt to emulate what I am doing here, at least not until the design has been thoroughly tested. I will continue to update this page as the design, construction and testing progress.

The first question has to be, why a hybrid? From information I have gleaned from the web, books and other hybrid builders/users. There a certain advantages.

Basic Design:

Most hybrid use thin wall light weight tanks for oxidizer flight tanks. The advantage there is weight savings, but the draw back is that thin walled uncertified tanks need an always open vent to the outside. This vent bleeds off N2O and head pressure. While those losses may be small, they are still losses. It also means oxidizer fueling must be done at the launch pad, requiring more launch support equipment.

I have opted to go with a sealed, high pressure aluminum tank. While it is heavier, I can fill the tank in my shop before a flight, and know with certainty exactly how much N2O is in the tank. My launch support equipment is also the same as for any solid launch. I have decided to use a 20 oz. CO2 tank for the oxidizer flight tank. Rated at 1,800 psi with an actual burst pressure of over 7,200 psi, the tank is more than adequate for the 600 to 900 psi nitrous oxide. I am also leaving the burst disk on the tank valve for added safety.

For ease of construction, I decided to use the valve that came on the tank, and modify it to function as fill adapter and injector, as well as the connection to the combustion chamber. I started by removing the valve from the tank. Then I removed plug, seat and spring from the valve opening. I also removed the burst disk assembly while I worked on the valve.

Below are some photos as I worked on the valve.

This is the top of the valve, the part that's sticks up out of the tank. It was threaded for the plug and the seat and spring I removed earlier. I drill it out to 3/8". This is where the injector will go.

I am now tapping the hole with a 1/8" NPT tap. This will hold a 1/8" NPT plug that will be drilled out to form the injector.

I have drilled a small pilot hole into the valve side, this will be where the fill valve goes.

Here is the enlarged hole to accept the 1/8" NPT tap. I tapped this hole as before.

Here is the valve on the tank, with a small needle valve and the injector plug installed.

That's where the project is at this point. Here are a few more design thoughts.

As you noticed, I have used a small needle valve as the fill valve. While I believe it will handle the pressure, I'm not entirely sure. I do plan on using a plug on the end of the valve after it is filled to prevent any bleed by of N2O past the needle valve after filling.

I plan on sealing the injector orifice with an epoxy based pyrogen to initiate the flow of N2O. With an injector hole of .1", the actual force acting on the epoxy is only about 6 pounds. I need a gas tight seal there, so I may try a drop of RTV, or a small plastic disk under the epoxy to make a better seal. It would be nice to be able to change the injector plug with the epoxy pyrogen on it. That would make for rapid turn around time on refueling and flying, but would also mean a longer injector assembly. I'll have to think on that for a while...

My design will allow for the oxidizer tank to be simply screwed on to the combustion chamber at the launch site.

The 20 oz. CO2 tank should hold just over 3 lbs. of N2O when filled, that's allowing a 15% gas area and 85% liquid N2O.

3.5 pounds total propellant weight when fully fueled. At an Isp of 185 = 647.5 pounds of thrust or 2,881 newtons, making it a low "L" class motor. A ten second burn would be an average thrust of 64.7 pounds, total rocket weight should come in around 11 pounds.

Just from memory here, but I think I had the throat diameter at about .6" and the chamber pressure at 287 psi.

The next step was construction of the adapter which connects the tank valve to the combustion chamber. I cut about a 3/4" length of steel bar from some 1 7/8" diameter stock. I then turned it down to 1.6" to fit inside my 1.5" EMT. I know it's weird, but 1.5" EMT has an internal diameter of 1.6". I left about a 1/4" lip the full diameter to act as a stop when the casing slides on. I then center drilled the stock, and drilled all the way through using progressively larger bits. The last bit was 3/4", slightly undersized for the hole needed. I then removed some material from the inside to lighten the piece as much as possible.

I had initially planned on just using a plate as a bulkhead and a nut to hold it in place. But, I decided threading the bulkhead plate would be a cleaner design. The problem was, the thread required was a 13/16" at 13 tpi (threads per inch) and I didn't have a tap that large. I also didn't have an inside threading tool for my lathe. But I did have some tools blanks to grind. So I proceeded to see if I could just grind a blank into a thread cutting tool. After some time on the grinder, I had a tool that looked like it should work.

Keep in mind I have never cut inside threads on my lathe before, so this was going to be interesting! I started the cutting process and made 10 or 12 cuts, and wasn't very happy with the way it was cutting. I took a close look at the tool next to the work and decided the tool was high on the works' center, so I took the tool out and ground it down a bit to get it on the centerline of the work. The cutting went smoother now, and I also used some cutting fluid to aid in clean cutting. I checked the work with my calipers at first, then I used the valve itself to check the threads as I got close. While it's not the greatest threads I've ever seen, they certainly work and I am pleased with the results.

Below are some photos of the work.

Above is the tool I ground for the inside threads. It's the top left tool on the tool post.

Here is the adapter from the top. You can clearly see my first inside threads.

Here is the adapter from the bottom, you can see a groove cut to capture an o-ring to seal the combustion chamber to the valve. To assemble the motor, you simply screw the casing onto the N2O tank valve.

Here is a picture of the adapter screwed onto the valve.

Here is a picture with an old casing stuck on the adapter, just to give you an idea of what it will look like.

The next step is to prepare the injector for testing. So I drilled the orifice to a diameter of .075". I slightly rounded the entrance area where the liquid N2O flows into the injector, just to try improving the flow through the orifice. I then installed the injector and the fill valve, using teflon tape to seal the threads.

Here is a photo of the valve, with the injector orifice drilled.

Here is another photo viewed from the tank side.

The next step is to see what is going to be required to seal the orifice. I used a hack saw and cut 8 very small grooves in the outside of the injector, I'm hoping they will give the epoxy pyrogen something to grab onto. I mixed a small batch of KNO3 72%, Epoxy 26%, and Iron Oxide 2%. I started with slightly less epoxy, but increased it to 26% to get a mix that would flow better. I then wrapped masking tape around the injector to form a mold, into which the epoxy pyrogen was poured. I used a toothpick to help the epoxy flow around the injector.

Here is a photo of the injector covered with epoxy pyrogen.

The next step will be to fill the oxidizer tank with air, and do a leak test. If everything holds fine, I will do a pyrogen burn test, with air pressure in the tank to see if the pyrogen will burn cleanly from the injector orifice.

If all goes well I will repeat the above tests with increasing air pressure, until I reach about 900 psi.

Here is the pyro valve ready for testing.

Here is the valve with a fill adapter so I can fill the tank with air.

I performed the first tests today. I pressurized the tank to 350 psi with air. There is a small leak around the fill valve fitting, it will have to be addressed but it wasn't enough of a leak to prevent a pyro valve test. My greatest concern was the tiny orifice plugging with the epoxy pyrogen. And that concern was justified. I ignited the pyro valve with a small motor igniter, it lit quickly and easily. The epoxy pyrogen burned for about 9 seconds, without ever releasing the 250+ psi air in the tank.

Above is a picture of the pyro valve burning. It did not release the pressurized air in the tank.

8.6.03 I have recast the pyro valve again. This time using a small plastic disk over the orifice. I think the plastic disk will keep the epoxy out of the orifice, and still blow away when the epoxy burns. Of course, the plastic may just melt into the orifice too. I'll find out tomorrow. There are other ways to make a pyro valve work. But they would require a little more hardware and more combustion chamber length. Both of which I am trying to keep at a minimum. Or course, my pressure is only 1/3 of what the N2O will be, so the higher pressure should help, but I figure if I can get a clean flow at 250 psi, I won't have any problems in the future.

Perhaps I should have used a more heat resistant material to cover the orifice. Even a tiny tin disk should be fine, as it is much smaller in diameter than the nozzle throat will be.

8.07.03 I test fired the pyro valve again today, this time with the small plastic disk covering the orifice. The tank was filled with 350 psi air. This time the pyro valve opened cleanly when the disk blew off after the epoxy burned away.

Next, I need to fix the small leak in the fill valve. Then I will do some higher pressure tests.

Above you can see the air flow in the grass to the right of the tank as the pyro valve opens.

Update: Sept. 6, 2003

In the past several weeks I have been procuring what I need to get the hybrid project off the ground.(pun intended) I finally got my nitrous tank, it's a 15 lb. tank. I also ordered a fitting to fit the CGA 660 valve of the tank. The fitting has an out fitting of an AN
4. As luck would have it, I had an adapter to convert the AN 4 fitting to 1/4" NPT, and I also have a high pressure hose with 1/4" NPT fittings on each end. So I have the fill line ready to go.

One little problem has been troubling me, I need to leave a gas space in the flight N2O tank, and I'm not sure how I want to go about doing that. Most hybrids use an always open vent, just below the top of the tank to leave a gas head space. I don't\can't have an open vent. The next best option would be to drill and tap a hole into the top of my flight tank, and install another valve with a dip tube extending into the tank, the valve would be cracked open to vent gas as the flight tank is filled, when the liquid nitrous level reaches the dip tube indicated by the white plume, both fill and vent valves would be closed. My flight tank is rated for pressures way over that of N2O, plus I have the burst disk should the tank overpressure. So, I don't think I would have a problem. But, the valve and dip tube should eliminate any questions in that area.

This is a picture of the completed nozzle. It has a 60 degree convergent angle, a 30 degree divergent angle and a .56" diameter throat. Behind the o-ring groove are 8 holes drilled and tapped for #10-32x3/8" screws. I left the metal thicker than in most of my nozzles, I wanted the extra thermal capacity the heavier metal would create. If the throat erodes due to heat, I may have to make a graphite insert for the nozzle.

In this picture you are looking into the divergent cone.

Here is the nozzle next to the combustion chamber.

Here is the R 45 mixed and ready to cast as a fuel grain.

I decided to just cast the R 45 into the combustion chamber. I made a guide from a 1/2" wood dowel, then placed a rubber washer, then a metal washer on the dowel and used a wood screw to hold them on the dowel. The dowel assembly is inserted through the threaded top of the chamber to keep the R 45 from draining out the hole.

A section of PVC pipe was then lubed with silicone grease, and slipped over the dowel. The PVC pipe is the actual coring tool. The wood dowel just holds the PVC in place. I then poured the mixed R 45 into the chamber to form the fuel grain. The R 45 was cured with 8% Isonate 143-L.

Update: Sept. 7, 2003

I attempted my first test fire of the hybrid today. I hooked up the fill line to the flight tank, slowly opened the supply tank valve, once the line was pressurized I slowly open the needle valve on the flight tank. The tank quickly equalized with the supply tank. I had the engine with attached flight tank on a digital scale, so I could monitor the weight of the nitrous in the flight tank. I only got .4 lbs. into the flight tank. I raised the supply tank about 3 feet over the flight tank, hoping the gas space in the flight tank would bubble up through the fill line and back into the supply tank. The only problem was, the supply tank has a dip tube going to the bottom of the tank, making it difficult for any gas to go up the fill line then down the dip tube, so I laid the supply tank on it's side, hoping it would help, it didn't. After 20 minutes no more liquid nitrous went into the flight tank. I guess I will have to install that dip tube/valve I mentioned earlier into the flight tank.

I decided to do a leak test on the flight tank and engine. Soapy water indicated no leaks, and the needle valve was working fine. At least I that much was working well. I decided to do a test firing even with the minimal amount of nitrous in the tank. I hose clamped the hybrid to a metal tower in my yard and hooked up an igniter to my launch system. 5,4,3,2,1 ignition...the igniter popped loudly. Then, nothing. I kept the video camera running for about 10 seconds, no smoke at all. My pyro valve had failed to light. So I turned off the camera and went for a new igniter, as I was walking past the engine the pyro valve cracked open and the nitrous vented out. The tank emptied in several seconds.

Upon disassembly of the engine, it was evident the pyro valve never burned. I'm sure the igniter popped and stressed the pyro valve, causing it to break.

While my first static attempt was not a rousing success, I did learn a lot. I need the vent valve in the top of the flight tank, the flight tank and hardware handle the nitrous pressure well, and I may need less vigorous igniter that burns rather than pops. I'll try again tomorrow.

Update: Sept. 8, 2003

I installed a top vent valve in the flight tank yesterday, I also recast the pyro valve. This time I didn't use a pyrogen igniter, I just slightly embedded a resistance wire on the outside surface of the epoxy pyro valve. The idea is to get the pyrogen burning, not to have a mini explosion when the igniter lights.

I refilled the flight tank, using the new flight tank top vent I was able to determine when the tank was full of liquid nitrous. I once again used hose clamps to attach the engine to my metal tower.

I assembled my launch controller, 5,4,3,2,1 ignition... I heard a slight pop at ignition, then several seconds of smoke, I knew the pyro valve was burning. Then engine ignition! The sound was terrific, much louder than any of the solids I have built in the past.

Here is the moment of ignition.

Here is the engine at full thrust.

Click Here for a video of the static test. 6.5 mb It's just to bad my new video camera has auto sound level control, the sound of engine firing was very impressive.

Now for a few notes on the test. Of course, I didn't get any thrust measurements so it's hard to tell how the engine performed. There was about 1.5 lbs. of nitrous in the flight tank. Going by the video, I would guess there was about 6 seconds of good thrust, followed by four of 5 seconds of blow down. That's a little longer burn than I wanted. The burn was a little fuel rich also, the motor produced smoke and a slightly yellow flame. Taking those into consideration, I can see I need a larger orifice diameter.

Post burn inspection of the engine revealed no erosion of the throat, no blow by at any of the seals and no heat damage. Which is good, but had the engine burned at the o/f ratio I wanted, I would have expected some heat damage to the throat.

Update: Sept. 9, 2003

I decided to try another test of the hybrid today. Frankly, I should have stayed in bed! But, here it goes. The first test went ok, but it was a far too long of a burn. So the first thing I did last night was drill the orifice from .075" to .1 inch. I then cleaned the flight tank valve and injector. I soaked the parts in vinegar, and brushed them clean with a brass brush. I then recast the pyro valve. The combustion chamber was pretty clean, I just scrubbed the nozzle and rinsed it out. The fuel grain was still very thick, so I decided to do a second burn on the fuel grain.

This morning I made sure everything was dry and made one last inspection of all the parts. The engine was screwed together and ready to be filled. I decided to throw together a quick test stand, just to get an idea of what kind of thrust the motor would put out. The design is flawed however, the way the springs are attached they never add much pull after the initial tension is overcome. So, the quickie test stand is not of much real value.

After the test stand was made and attached to my tower. It was time to fill the engine with N2O. A little history of my nitrous tank is in order. I got the tank from an OEM supplier, plain label so to speak. I ordered it with a CGA 660 valve, the 660 valve is supposed to be the hot rod valve, and since that was where I would refill, that's the type of valve I ordered. Hot rod tanks have a dip tube in them, so the tank must stand upright, or lay on it's side to draw liquid nitrous. As it would while mounted in a car. So I assumed with a 660 valve, I would have a dip tube. Never, never assume! When I went to refill the motor, I could never get any liquid nitrous in, so I kept trying and trying. In the end, I gave up, and went outside to drain the tank and pull the valve out, just to see what kind of dip tube was in there. I had the tank with the valve up, and cracked the tank valve open, just vapor came out. So, leaving the valve open I laid it on it's side, still vapor. Then I turned the tank upside down, ahhh, liquid nitrous! So there was no dip tube. By now the tank was almost empty. But I decided to try to get one last fill out of it. So I hooked it up to the hybrid engine, turned the tank upside down and started filling. I got about one pound of nitrous in the flight tank, not full, but enough to do another burn.

Click Here for a video of the test.

A lot of things went wrong with this test. If you watch the video, you notice a couple of seconds into the burn the engine has a burst of power, I think the orifice must not have fully opened at first, then all at once it opens up and takes off.

Here is a shot shortly after ignition.

Here is a shot of the burst of thrust.

Here's the engine at full thrust.

I'm not sure why the pyro valve didn't open all at once. I did cast the pyro valve a little thicker, I suppose if the pyro valve ignited more to one side, it could take an extra second or two to burn all the way through. I also ran out of fuel! The R45 was completely gone, so I guess I should have recast the fuel grain as well. As a result of the fuel grain running out, there was no insulation for the casing and there was some burning of the zinc coating on the outside of the combustion chamber. If you watch the video, you will see the outside of the combustion chamber casing burning. That was not a burn through of the case, it was some R45 that had dribbled down the outside when the fuel grain was cast. It burned because of the lack of insulation on the casing when the fuel ran out.

Next, I need to get that *%@! nitrous tank filled. Then I'll give some thought to other fuels. And perhaps a new test stand for hybrids.

Update: Sept. 16, 2003

Several days ago I refilled my nitrous tank. So two days ago I recast my pyro valve and fuel grain. The flight tank was filled, it held 1.75 lbs. of nitrous. The engine was set in the test stand, and the igniter wires connected. Just as I was moving to the launch controller, when the pyro valve cracked and the nitrous vented out. I was a little worried about the pyro valve, I had set the igniter wire rather deeply into the epoxy pyrogen. I believe the soldered wire weakened the pyrogen, causing it to crack.

At any rate, it was a problem that needed solving. So, for the next pyro valve I decided to chop up some fiberglass strands, and mix them into the pyrogen when cast. I added about .5 grams to a 38 gram batch of the pyrogen. I also used a very small piece of masking tape to cover the orifice, rather than the plastic disk I had used in the past. As an after thought, I considered the possibility that my plastic disk may dislodge when the pyrogen is cast on the valve. That may have caused my irregular burn on the last test. I also made sure the igniter wire was just barely into the surface of the pyrogen.

The new pyro valve was ready, the fuel grain was ready. But I wanted some real data, or another test burn was not going to give me any more information than I already had. So I decided I needed a test stand to better measure thrust. That brings us to today, I decided to use the hydraulic load cell I had made for my solid motor test stand, and modify it to be used on both stands. While I really wanted the engine to sit on a rail system of some sort, I also wanted it now! (I know, I know) So I ended up just using a plastic sleeve as a guide. I will try to come up with a rail system in the future.

Above is a picture of the hybrid test stand. You can see the temporary plastic sleeve to guide the engine and hold it vertical.

In the above picture you can see more detail of the load cell. The load cell mount is adjustable by means of a slotted bracket, allowing engines of various diameters to be tested. I think I should have better luck with attaining reliable numbers from the load cell, than I did testing solid motors. The hybrid burns much longer, with lower levels of thrust, which should make data easier and more accurate.

Here is my 15 pound nitrous tank, with the fill adapter. The pressure gauge was already there so I just left it on. I use a digital shipping scale to see how full the tank is, and to measure the nitrous in my flight tank.

Update: Sept. 19, 2003

A couple of days ago I attempted another test of the hybrid in the new test stand. Unfortunately, when the pyro valve ignited, the N2O was released but the engine failed to ignite. I have a feeling I used too much epoxy and not enough KnO3 in the pyrogen mixture. The flight tank was a little short of full anyway. I only got about one pound of nitrous in the tank. It is still a bit of a mystery why I can't always get the flight tank full. For the next pyro valve, I will pay closer attention to the percents in the mixture, and also cast it as large as I can, more burning area and a hotter mix should aid in rapid ignition.

Today I decided to give the engine another try. The pyro valve has been recast, it is larger and contains a higher percent of KnO3. The fuel grain for this test is a thin wall PVC pipe with R45 cast into it. The PVC was a bit of a lose fit. I considered making a cardboard tube for casting into, or even a fiberglass tube, but I had the PVC on hand and decided to give it a try.

I filled the flight tank with nitrous oxide, I made sure the fill tank stayed completely inverted, and cracked open the vent valve on the flight tank to just a slight hiss. The flight tank quickly filled with over 1.6 lbs. of nitrous. That's a good fill, as full as I've had it I think.

5,4,3,2,1 ignition. You could see the pyro valve burning well, more smoke than in the past. Then the engine lit and came to full thrust almost instantly.

Click Here for a video. 2.92 mb

I had two video cameras running, one on the engine and the other zoomed in on the pressure gauge. The sound still surprises me, I was holding the camera zoomed in on the gauge, and when it lit, I jumped, and pulled the camera off the gauge for a second. But I still got good data from the video. This test was almost perfect, the burn was just over 5 seconds, just where I wanted it. The only bad thing, and it's interesting to see on the video, at the end of the burn the fuel grain is expelled from the engine. The nozzle is plugged for just a moment, you can see a flash of flame and smoke come from the top of the combustion chamber, that was the burst disk blowing out. After closer examination, the injector orifice and the burst disk were damaged. I always wondered what would happen if the chamber pressure went higher than the nitrous pressure. I guess that answers my question, the flame front obviously traveled into the oxidizer flight tank, causing combustion of the nitrous in the tank. The injector orifice was burned out to several times its initial size. If I had a lightweight thinwall oxidizer tank, no doubt it would have failed. I can see the PVC got way too soft from heat during the burn. I'll either just cast into the combustion chamber in the future, or use a fiberglass casting tube.

Above is a photo showing a new injector on the left, the burned out injector in the middle and the blown burst disk on the right.

Above is a picture of the ejected fuel grain.

Above is a photo of the engine just before burnout.

Here is the moment of nozzle blockage, notice the small flame at the top. That is coming from the burst disk.

The thrust data was interesting, but not complete. Thrust started at 10.417 seconds and ended at 15.492 seconds. Initial thrust was about 30 lbs. and quickly went up to the top scale on the gauge, maximum thrust is not known. The gauge maximum is 43 pounds. The total fuel and oxidizer consumed was 1.8 lbs.

Update October 5, 2003:

I decided to try a new injector on the hybrid. I felt that by directing the flow of N2O onto the fuel grain itself, engine efficiency may improve. I also reduced the fuel grain length by one inch to 7.5 inches. The result was a 1" area behind the fuel grain for a post combustion area. The new injector has four .06" diameter holes drilled at a 30 degree angle from perpendicular. While the total surface area is greater than the .1" diameter single injector, the resistance to flow is undoubtedly greater, I'm hoping the flow rate of N2O is similar to the original .1" injector.

Above is a photo of the quad injector.

I wanted better data from my tests. So I used both of my video cameras, with my high resolution camera on a tripod zoomed in on the test stand pressure gauge. After carefully zooming in and focusing on the pressure gauge, I fired the engine without hitting the record button. Oh well, I'll get it next time. I did manage to record the test with the other camera, but it wasn't close enough to get a reading from the pressure gauge.

Above is the engine at full thrust.

Here are some of the numbers from the test:

N2O tank empty: 940 grams

Complete engine weight less N2O: 4.2-4.4 lbs. (sorry, too heavy for my lab scale)

Combustion chamber starting weight:1024.1 grams

Combustion chamber ending weight: 943.5 grams

Fuel consumed: 80.6 grams

N2O consumed: 590.2 grams (approx.)

N2O to R45 ratio: 7.32/1

Start of burn: 9.937 sec.

End of burn: 16.429 sec.

Total burn time: 6.492 sec.


I've decided I need to redesign the pyro valve, or do away with it completely. I'm leaning towards making the engine easier to scale up. I think I will use a 1/4 turn ball valve, spring activated by means of a burn through retaining line, maybe a small nylon cord that can be cut with a resistance wire. It will be a few extra ounces, but should be a cleaner start, and could incorporate a throttle of some sort.

More to come....

Home Contact Me