The Junkyard Turbojet Engine
A Real Working Jet Engine Built From Junkyard Parts!
How to build a primitive gas turbine engine.
I have always been fascinated by all things mechanical, and all things powerful. Gas turbine engines (jet engines) fall into both categories. During my endless hours surfing the net I've come across many pages about home built turbine engines, and have considered building one myself. However, I was worried that I didn't have the necessary tools, skills, and money to pull it off. I am primarily a wood worker. I don't own a metal cutting lathe or a milling machine, and I can't weld. I am also chronically short of cash. These factors prevented me from seriously persuing the idea at first.
As I found more and more examples of home built turbines on the net, I realized that there was a great variation in the level of sophistication. Some were built by master machinists, and looked it. Others seem to have been banged together with only the aid of a hacksaw and a hammer. Whether finely crafted, or crudely thrown together, they all worked. I began to believe that I could also build one, and started to make plans.
I decided I would try to build an engine only from parts that could be obtained easily and cheaply. They would come primarily from auto scrap yards and hardware stores. The engine must be designed so that I can build it with my limited skills, tools and funds. I also decided that the engine should do something useful, and not just make lots of noise and heat while burning up lots of fuel. My ultimate goal is to add a second turbine, and extract energy to run a generator or do something else useful. But before I could get fancy, I first had to actually build a working engine. This is a chronicle of my efforts.
My further developments page chronicles the ongoing modifications and experimentation with the completed engine.
The photos are large, numerous, and will take a while to load.
I kept my eyes open for something to use as an oil pump while I was at the scrap yard. I may try a power steering pump. I can get one very cheap and easy. I can also get 12 volt fan motors to drive it. I'm working on ideas for the combustion chamber and flame holder.
This is my $50 turbo charger. The car it came off of had 114,000 miles
on it, and the bearings are in pretty bad shape. I'm not going to worry
about them for now. This is my "learner turbo", once I really know
what I'm doing I'll have it rebuilt or buy a better one.
I decided that the engine should ultimately run on liquid fuel, probably alcohol. However, I will probably do the initial combustor tests with propane, because it's just so much easier.
The conduit is harder to work than I expected, but it is slowly coming along.
The process of flaring the flame holder tube left 4 triangular slots at the turbine end. I had planned to completely close these slots with metal shims. Now I think I will make the shims adjustable so that excess air can be dumped behind the flame tube. No-doubt it will take a lot of experimentation to determine the optimum placement of the shims for best operation.
This is the flame tube. It is made from 1.25 in. steel electrical conduit.
The fuel injector passes through the hole on the right. The turbine end is
flared out to the inside diameter of the air case. The four slots left
from flaring act as the last set of dilution holes.
The combustion chamber, flame holder, and fuel injector are all essentially completed. All that's left to be done before I can test it is to settle on an ignition system. I haven't decided just what to do about that yet. I go back and forth between using a traditional spark plug, and a somewhat untraditional brainstorm I had about striking an arc directly to the fuel injector. It's too hot for more work today, so I'll just think about it for a while.
I bought some high pressure hose and fittings today so I can get propane from the tank to the fuel injector.
Progress will be slow until cooler weather arrives.
Today I bought a small motorcycle spark plug. With some difficulty I managed to drill and tap a hole for it in the heavy pipe of the combustion chamber. I also drilled a passage hole for it in the flame holder. I have an old LASER power supply that will make a nice hot arc across the plug gap.
I need to build some sort of test stand so I can test the combustor soon. The blower from my shopvac can provide air, and I have propane tanks, high pressure hose, and fittings. All I need to do is put it all together.
This is the completed combustor. The fuel inlet is on the right. The air inlet
is on the lower right. The spark plug is at the top. The exhaust is through the
flange on the left. The combustor is built entirely from standard pipe fittings.
Four new holes have been drilled in the flange to mate with the bolt holes on the turbo.
I made a makeshift test stand for the combustor from a small stepladder, some scrap wood, and lots of clamps. I mounted the combustor, electric blower, and HV power supply all on the stand. It's a little rickety and strange looking, but I think it will work fine for testing purposes. The old LASER HV supply makes a nice hot spark across the plug gap. I'll run it from batteries for now. Eventually I'll probably build a plug-in power supply for it.
I spent a good while plumbing together the fuel line and fittings. I also had my spare propane tank refilled. I'm almost ready to test. Just a few more odds and ends need to be completed. I have a very busy weekend ahead of me, but perhaps I can find the time to do a test run or two.
It's going to take guts to fire this thing up. It's mean looking, and real loud just with the blower running. I suspect there will be a very loud bang when it ignites, and it will probably really roar when it's burning. I wonder how my neighbors will react?
This is my makeshift test stand for the combustor. The injector end of
the combustor is facing the camera, and the air blower is on the left.
I have formed several theories about what problems could be causing this, and how to fix them.
Success! Ignition was achieved easily. There was no loud bang as I had feared. Ignition is like turning on a fire hose. It burns with a very loud low frequency rumble that brought the neighbor lady running to see what was happening. The rumble comes from instability in the combustion. Combustion will not self-sustain, and the flame blows out immediately if the ignition is switched off. I suspect that there is not enough turbulence to produce adequate mixing of the propane and air. There is also only a very narrow region of fuel pressure where it will burn. A little above or below, and the flame blows out. My next modification will probably be to try to induce swirl inside the flame holder.
looking into the open end of the combustor while it's burning requires standing nearly 20 feet away, but even so the air blast is amazingly hot and powerful. I think I have created a monster. The flame is pale blue, and seems to be entirely contained in the combustor.
I wobbled out some of the holes in the primary zone in such a way as to hopefully induce some swirl. I did another test run in the early evening. There was no noticeable difference in the performance. Under the darker conditions I did notice that the flame extends several inches beyond the end of the combustor. I am beginning to suspect that the injector is a big part of the problem.
This is a test firing of the completed combustor after dark. This photo is a several
second exposure taken to capture the flame. The flash was fired at the end of the exposure
to illuminate the rest of the scene.
I may also experiment with closing down the triangular slots on the flame tube. (I've been running with them wide open.) I want to force as much air as possible through the dilution holes to cool the combustion gasses before they leave the combustor.
The original injector is on the left. The new one is on the right
The new injector is a piece of 1/8 in. copper tubing smashed nearly
flat. The opening is a very thin slit that is nearly invisible.
When the combustor was cool enough to touch I tore it down. The last few inches of the flame holder are fairly discolored, but there is no sign of melting as I had feared could happen. Discoloration so close to the open end leads me to believe that the flame front is too far down the tube. This is probably why the flame extends past the end of the combustor. I decided to close down the triangular slots in the flame tube to see if that would cause the flame front to retreat toward the injector. After putting it all back together, it would not ignite. I seem to have the same condition that existed a couple of days ago. I assumed it was running too lean to ignite, so I put the old injector back in. I know that injector is providing a lot more propane, but it still won't ignite.
It's possible that the problem is with the HV power supply. It did really badly overheat, enough to burn my finger. It could be damaged. The spark plug is probably a much lower impedance load than it was designed for. I guess I will have to build a more traditional HV power source. I have several old auto ignition coils laying around. I'll build a driver circuit for one of them and try again in a few days.
This is the new high voltage power supply for the spark plug.
You can see it arcing away at the upper left. Don't laugh at
the construction technique, it works.
Today I spent most of the afternoon trying to get back to where I was 09/27/99 before I started tinkering with the combustor. I hooked the new HV supply up to the combustor. No joy, it still won't ignite. I then put the second injector back in, still no ignition. Then I partially opened the slots in the flame tube, and was rewarded with some half-hearted chugging and puffing. I opened the slots some more, and it got better. I then removed the shims and opened the slots fully. Now it runs nearly as well as it did before. Obviously there is too much air going into the flame holder with the slots closed.
The combustor is still not ruining as smoothly as it was. I think my new HV supply is not producing an arc as hot as the old one was. The air blast is really doing a number on the arc. This would not be a problem if the flame would self-sustain, because it would only have to ignite once. The combustion got smoother as I opened the slots until they were all the way open. I suspect that there is still too much air going into the flame tube even with the slots all the way open. I can't open the slots any more, but I can close down holes in the flame tube. that will be my next project.
First I added a welding gas type regulator to the propane tank in an effort to get better control of the fuel pressure. I have noticed that as the propane tank gets cold during use the pressure drops. This makes it very hard to keep a constant pressure going to the injector. The pressure in the propane tank is so low that the high pressure gage never even moves off it's peg. Otherwise it works great. I wish I had thought of it earlier.
Next I tried a run with ALL the holes in the primary zone closed off. No good, could not achieve ignition, only some pathetic chugging. Next I opened 4 - 1/16 in. holes just beside the injector. The combustor ignited, but burned really roughly. I noticed that when I turned off the propane at the tank valve, the combustion smoothed out for a few seconds before the flame blew out.
This is when I made my big breakthrough. I started it up again then slowly reduced the pressure with the regulator. The combustion got really rough as I lowered the pressure. Suddenly, at about 38 psi the flame moved to the back of the flame tube and really smoothed out. There was no more of the rumble I was so used to. It was burning much smoother. It now roars like a high pressure air leak. I turned off the ignition, and sure enough it self-sustained. No wonder, when I stood well back and looked up the exhaust, I saw that the spark plug and all the little metal burrs around the holes in the flame tube were glowing bright orange. Even without the spark there were so many ignition sources that the flame couldn't possibly have blown out. I shut it down because I was afraid I would melt the spark plug. If I had not installed the regulator I never would have found the region of stable burning. Controlling the fuel pressure with the tank valve was just too touchy.
As it stands now, the fuel pressure must be brought up to about 60 psi before ignition is achieved. At 60 psi the combustion is rough and has the loud rumble of instability. After it's been burning at 60 psi for a few seconds, I can reduce the pressure to the stable area around 38 psi. The flame front retreats up the tube, smoothes out, and gets real hot. The flame seems to be burning well up the flame tube, and not extending past the end of the tube. I need to do another run in the dark to make sure.
I will probably enlarge the passage hole for the spark plug in the flame tube in an effort to blow more cool air on the plug. I suspect this will make the combustor more difficult to ignite. The air/fuel mix near the plug is already too lean to ignite unless the fuel pressure is really cranked up. The spark plug is probably too far away from the fuel injector. Unfortunately, moving the plug closer to the injector end would require a major redesign of the whole combustor.
In the early evening I did some more testing. I found that the combustor is very sensitive to injector orientation. With the slit in the injector in the same plane as the spark plug it will ignite. Turn it 90 degrees, and it won't ignite no matter what. In between it's iffy. I also found that the flame is indeed fully contained inside the flame tube at 38 psi. I decided not to worry about the spark plug, and let it run a while. After a few seconds I noticed the whole flame tube was glowing dull red. The plug and all the little metal burrs were again glowing bright orange. I tried throttling the combustor and found it burns very smoothly over an amazing range of fuel pressures once the flame tube is hot. It will burn smoothly and self-sustain from only a few psi up to just over 60 psi. Over 60 psi the flame becomes unstable and won't self-sustain. At low pressure it just whispers, but at 60 psi it really roars. The flame starts to peek out of the end of the combustor at about 45 psi. At 60 psi it roars like a space shuttle trying to take off, and has a nice tongue of pale blue flame extending several inches beyond the end of the combustor. I ran it at high power for several minutes just to see what would happen. No problems were observed. When I shut it down I could see that the entire length of the flame tube was glowing orange.
I pulled out the spark plug to see if it was damaged. Other than a slight discoloration of the metal, it looks just fine. I think I'll stop worrying about it, and leave modifications for the MK II engine.
The propane fuel tank with the new regulator that allows fine
control of the fuel pressure. The yellow handle is part of a
quick shut-off valve. It's come in handy a couple of times.
When the weather gets a little cooler I'll return to the scrap yard to find the pieces I still need to finish the engine. I need an oil pump of some sort. An engine oil pump or perhaps a power steering pump will do. I'll need a motor to drive the pump. Perhaps a 12V radiator fan motor. I suspect I will need an oil cooler of some sort. I may even pick up an inter cooler and experiment with it if I can find one cheap and easy. It's been cool enough to fool around with the combustor on the shady side of my house, but not cool enough to bake under the sun in the scrap yard. Perhaps in a couple more weeks.
I'll need to build a framework to support all the bits and pieces. The test stand I'm using now won't cut it for testing the full blown engine. I priced some heavy duty metal shelving components a few weeks ago, but decided they were too expensive. I'll probably bang together something from 2X4s and plywood. I'll also probably put wheels on it to save my back.
I've also been thinking about how to extract power from the engine once I get it all together and running. Someone on the turbo turbine mailing list is using a homemade air motor running on compressor bleed air to power a generator. This is brilliant! No need to worry about high temp exhaust gasses, high speed reduction gearboxes, or precision balanced machinery. I'm not yet sure my engine will produce enough excess air to drive an air motor, but I could always use exhaust from the first turbo to drive a second turbo charger and use all the air from it's compressor. I'm going to be keeping my eyes open for a cheap air motor.
I have also made a tap in the side of the combustion chamber so I can read the pressure with a gage. This is just the first bit of instrumentation that I will have to build. Eventually I want to also monitor RPM, EGT, and fuel flow rate.
A machinist I work with has agreed to make an adapter plate to go between the flange on the combustor and the turbo. I decided that the flange is not really a very good fit on the inlet of the turbo, and some sort of adaptor is needed. It should be done sometime this week.
I also spent some time poking around the scrap yard today. I found some potentially useful bits, but didn't have the proper tools with me to remove them. I'll go back another time and try again.
Here is the "wheel barrow" type test stand for the engine. It will allow me
to wheel it in and out of my garage. At the moment the only component
mounted on it is the turbo charger.
First I came across a power steering pump from a Toyota Corolla that someone else had already removed to get at some other bit. It was just hanging by the high pressure hose. Couldn't have been easier to remove. It cleaned up nicely, and now looks like new.
Next I saw an easy to remove transmission oil cooler and grabbed it. It's nothing fancy, just a mini radiator. I don't remember what type of car it came off of. I almost chucked it when I found the next bit, but decided to keep it anyway, just in case.
Lastly I removed a combination oil cooler and oil filter assembly from some sort of Mazda. It's a lovely little unit that kills two birds with one stone. I had been struggling with the problem of how to mount an oil filter. It bolted directly on to the top of the engine, so unfortunately there are no fittings for getting the oil in and out, only passage holes. I'll have to manufacture some sort of plate with fittings that I can bolt onto it. I kept the other cooler in case the job turns out to be too difficult.
Total cost, a little over $30 and a few cuts and bruises from slipping wrenches and sharp edges. I still need some sort of reservoir for the oil and some kind of motor to run the pump. I also need to do a lot of plumbing.
10/17/99 - 10/23/99
I still need to mount some components and do a lot of plumbing before I can do a test run.
Here you can see the three main components of the engine mated together.
The turbo charger on the right, combustor with it's spark plug, and PVC air pipe.
10/24/99 - 10/27/99
There is only one major component left to mount, the oil filter/cooler assembly. It is a very awkward shape. After much brainstorming, I had the idea of making some mounting brackets out of aluminium angle that I can bolt onto it. I'll save that job for another day.
I am very close to being ready to test.
The first attempt failed, but was a learning experience. Ignition was easy, and very smooth. The obstruction of the turbine must slow the airflow through the combustor enough to make ignition easier. As I throttled up the engine, it was responding well and spooling up. At about 3 or 4 psi of boost, the PVC air pipe blew off the combustor, and brought the test to a premature end.
I made some modifications to to the pipe and remounted it. It stayed on during the second test. The engine spooled up nicely, but then topped out at what seemed like a fairly low speed, and would not self-sustain without the blower. I don't have a tachometer yet, so I am only guessing at the speed. I lowered the oil pressure to the turbo and found that the engine sped up. I then lowered the pressure as far as my crude regulator would allow, and was rewarded with a further increase in speed, but it still would not self-sustain. Obviously the oil pressure was still to high and creating too much drag on the turbo. The oil pump will supply a little over 40 psi. My regulator will cut that down to just below 20 psi, but that is still way to much. Obviously it still needs some work. As a temporary measure, I simply kinked the oil line to the turbo further reducing the flow. This resulted in a dramatic increase in speed, and sure enough it self-sustained. I attempted to explore the performance of the engine, but it was proving too difficult to maintain a constant oil pressure. I will rework the regulator tomorrow and try again.
I was surprised at how quiet the engine is when running. I had expected terrible noise, but it just whispers. The blower I use to start the engine is orders of magnitude more noisy than the engine itself. Once the engine is self-sustaining and the blower is off, the noise is not really that bad at all.
I also noticed that the turbo leaks oil through the bearings, although the excessive oil pressure may be exaggerating the problem. The engine blows a fine mist of oil while running. The oil can puddle in the turbo housing if the pump is left on while the engine is not running. This resulted in an oil fire in the hot turbo once when I shut off the fuel and the blower, but left the oil pump running. I simply turned the blower back on and blew the burning oil out of the turbo.
A side view of the completed engine and test stand.
With the lower oil pressure I no longer have the misting and puddling problem from the turbo. However, there is a lot of gas blowing past the bearings and through the oil return line from the turbo. A jet of oil mist and smoke blows out of the breather hole on the cap of the oil reservoir. I may run a hose from the breather hole to the inlet of the compressor to recycle the waste products. The exhaust from the turbo itself is smoke free.
I noticed that the turbine was running very hot on one run this morning. The turbine blades were glowing brightly. It doesn't normally run hot enough that the turbine can be seen to glow in daylight, so I looked around for a problem and found that the PVC air pipe was trying to blow off again, and was leaking air badly. It's interesting that the leak increased the EGT, but really didn't reduce the RPMs much. I shut down and reworked the pipe again. Now it is much stronger and sealed air tight. It could probably withstand several pounds more boost without further problems. Eventually a boost level will be reached where the air temperature (from the compression) is high enough to soften the PVC. Then I will have problems again. At 6-7 psi the pipe gets fairly warm to the touch, but not nearly hot enough to soften.
I also found that there is an air leak from the joint between the adaptor plate and the turbine inlet. Man o man is that some hot air coming out that leak. It is scorching the wooden frame of the test stand and has partly melted the water return hose from the turbo. I found the leak accidentally while running my hand along the air supply pipe looking for leaks. I jerked my hand back quickly and fortunately was not badly burned. After I demonstrate the engine to some of my friends, I'll tear it down for inspection, and when I reassemble it I'll use some high temp gasket material on those mating surfaces.
I am still amazed at how quiet the engine is. At 2-3 psi of boost it really runs very quiet. The sound it makes is like rushing air. Very similar to what you would hear driving at highway speed with the windows down in your car. The whine of the turbo is really only noticeable during acceleration or deceleration when it is changing pitch. Above 5 psi of boost the rushing noise gets loud enough to be irritating without ear protection if you are within a few feet of the engine. However, the noise is not loud enough to bother my neighbors further away. They didn't even know I had it up and running. The engine as a whole is quieter than the combustor alone was.
I also spent considerable time exploring the performance envelope of the engine. With a delicate touch on the fuel valve it can be made to run smoothly from just over 1 psi of boost up to about 5 psi. The response of the turbo is far from instantaneous, and it can't be throttled up or down too quickly. At 5 psi of boost and above I am seeing problems. I must have been lucky yesterday because today the rotor tends to start wobbling in its worn bearings above about 5 psi. This results in a rapid decrease in speed accompanied by loud and scary noises that I can't even begin to describe. When this happens the only thing to do is shut it down quickly. Because of this I have decided not to run it much above 4 psi anymore.
On the whole I am very happy with the engine. It starts easily, runs smoothly, and is much quieter than I expected. I will continue tweaking, improving, and adding instrumentation to it. I may also try adapting it to liquid fuel. However this little engine has only whetted my appetite for more power. Now that cooler weather has arrived I'll be haunting the scrap yards looking for turbo chargers that are larger and in better condition.
The goal of this web page was to describe the design, construction and testing of a primitive turbine engine. Now that the engine is up and running, I think that goal has been achieved. In the process, this page has gotten quite long and cumbersome. Making it any longer with reports of further developments on the engine is probably not a good idea. All further developments related to the engine can be found on the further developments page.
|Home||The Engine||Further Developments||A New Beginning||Turbine FAQ||Turbine Links||Contact Me|
[Mike's telescope workshop]
[Mike's gold prospecting page]
[Mike's home-built wind turbine page] [Mike's Home-Built Solar Panel page]
Copyright © 1999-2011 Michael Davis.
All rights reserved.