Turbo-charger based Gas Turbine Engine
Running Gas Turbine
|I have built a simple gas turbine engine from a Holset diesel engine turbo-charger by adding a
home made combustion chamber to it. I was given the turbo some years ago and for a long time wondered if it could be made into a gas turbine. More recently I began a series of experiments with the turbo, initially these were a bit half hearted and produced nothing more
than lots of flames but in 1993 a very hot turbo belching flames began to drive itself. I have refined this design and it now runs quite well. The fundamental problems to be overcome were to provide enough air to start it and to construct an efficient combustion chamber. A turbo-charger is quite similar to a gas turbine in many ways, the original engine built by Frank Whittle partly consisted of some super charger components and it was my inspiration for this project.
This was built from a steel tube cut from a satellite dish ground stand, the tube is clamped between two plates to form the ends. The bottom plate is bolted to the turbo turbine inlet volute and the top plate originally accepted compressor air through a tube but now air passes into the combustion chamber at the side near the top.
|Air is delivered to the combustion chamber through plastic drain pipe, this tends to blow off if the unit is allowed to go too fast. The flame tube or combustion chamber liner was made from a camping gaz tin and extended using steel sheet. The gaz tin gives the right sort of domed top to the liner. Holes are drilled into the liner to allow air into the combustion zone. The size and positions of the holes were guessed from looking at various diagrams of commercial engines, no calculations were done. The engine is fueled by propane gas, the gas enters the combustion chamber via a burner ring made from copper pipe drilled with 1mm holes.
A motorcycle spark plug is inserted into the combustion chamber to "light up" the engine. I have tried several different sources
of ignition the best being a HT Ignitor unit from an early jet aircraft. I have also used a motorcycle ignition coil driven from
a home built transistorised inverter. Once ignition has taken place, the combustion chamber seems to hold the flame well,
the throttle can be backed right off and the flame does not go out.
Oil is circulated into the turbo-charger plain bearing by a car engine oil pump driven by an induction motor originally from
a photo copier. The pump is of a type fitted to ford crossflow engines and is easily modified as it is an external type with
integral oil filter. A metal tank beneath the turbo collects oil from it ready to be recirculated by the pump. When the oil is
cold it is quite hard work for the motor, during starting the oil pump is stopped to lower the drag on the turbo-charger
rotor and then switched on when the engine self sustains. The oil used is conventional Mobil 1 formula which is what the
turbo is designed for, turbine oil should not be used as it is designed for ball races not plain bearings. During running the oil
gets hot, a future modification might be to add an oil cooler.
A complete compressor assembly from another similar turbo-charger is driven by a spin-drier motor running off the mains. The compressor forms a blower which is coupled into the front of the engine and acts as a "Windmill Starter". A dimmer switch connected to the motor controls the amount of air admitted to the engine, to achieve ignition only a gentle breeze is required otherwise the engine lights with a loud bang. To start the engine the blower works at full power and is removed when the engine self sustains at about 35,000 rpm. Interestingly the blower with a cold engine barely turns the rotor yet the airflow with combustion is enough to wind it up as the oil warms.
I have used an optical method for measuring the speed of the gas turbine. An optical fiber illuminates a small portion
of the rear face of the compressor wheel, the surface of the wheel is alternately shiny aluminum and matt black, a second
optical fiber receives reflected light from the wheel and conveys it to an electronic sensor. As the wheel rotates the reflected light pulses on and off. The sensor converts the light to an electrical signal which operates a homemade rev-counter calibrated
0-100,000 rpm. I have found this system to work but the reflected light is quite dim requiring a sensitive amplifier, I have used
a He-Ne Laser to provide the light as it couples to the optical fiber efficiently. Another problem is that the optical fibers are actually polymer ones which can melt due to heat soak into the compressor section of the turbo when the unit is shut
down. After shutting down the turbo I blow air through it to cool it, during this operation the turbine is locked using
a spanner to prevent it from turning as the lubrication system is switched off.
Exhaust temperature is measured using a standard inconel K type probe feeding into an AD595 thermo couple amplifier integrated circuit and then on to an analogue meter calibrated 0-1000 degrees C. I much prefer analogue meters, they are easier to watch as the engine parameters change during acceleration and deceleration. The AD595 IC makes temperature
measurement easy as it converts the mV output from the thermocouple to an output of 0-10V. The 0-10V output
then corresponds to a temperature range of 0-1000 degrees C.
I have fitted a pressure gauge to measure the compressor delivery pressure. The indicated pressure seems to fluctuate
so I have inserted a restriction in the gauge feed pipe to dampen oscillation.
The engine is fueled by propane gas delivered from a portable caravan type cylinder. The regulator is removed and the
valve mounted on the cylinder used as the throttle control. The engine has a very healthy appetite for fuel and only lasts about
10 - 15 minutes on a 3.9 Kg cylinder. Due to the fast fuel delivery the cylinder sits in a bowl of warm water to aid evaporation
of the liquid propane into gas.
I have tried liquid fuel using a car Bosch "K" type fuel injector nozzle, this almost worked but the single nozzle would not cope
with the required fuel flow. The nozzle using kerosene at low flow rates produced a near ideal spray pattern, but this
deteriorated as the flow was increased. Ignition was more tricky to achieve with liquid fuel, if light up did not occur quickly after
turning the fuel on, the engine would rapidly flood with fuel threatening a very "wet start" when ignition finally occurred.
An aircraft fuel primer pump was used as a fuel pump, this would supply fuel at up to 60 PSI, a needle valve was used to
spill fuel from the pump outlet back to it's inlet and so act as a throttle. The injector opens at about 15 PSI but as the
pressure was increased (needle valve closed) the device would choke and not atomize the fuel properly.
This is where the fun starts, to start this home built gas turbine the starter is coupled directly to the turbo inlet and the air
turned on gently. The ignition is switched on and the fuel valve opened again gently until the engine lights with a "fut".
After the engine has lit up, the air is turned on fully and the throttle opened, initially the rotor spins slowly but as the oil thins
and heats up the engine begins to accelerate and at about 35,000 rpm the air supply to the engine is quickly pulled off
to allow it to suck in more air and accelerate to a comfortable speed of 50,000 rpm. During starting the oil supply
is switched off and only momentarily pulsed to provide some lubrication without causing too much drag, when self sustaining
speed is reached the oil is turned on permanently. Once the engine has completed a run and is hot it is much easier
to restart, the rotor spins up much faster.
In operation the engine is quite noisy, although with ear defenders the unit sounds pretty good emitting a delightful "Whistle"
from the compressor and a rumble from the combustion process. Listening with eardefenders helps hear the compressor
speed more clearly which aids throttling the engine which can be tricky. If you close your eyes you can imagine that you
are at the controls of a real jet, I was stood listening to the Vulcan XH558 the other day and the similarity in the sound
to my engine was uncanny.
So far the gas turbine has achieved about 70,000 rpm and at 50,000rpm the exhaust gas temperature is only 500 degrees C
not bad for a home built engine. The limit to rpm at the moment is the compressor delivery pipe, it seems to blow off if
the engine runs too fast, flames shoot out of it and the compressor shrieks as it runs rapidly down. Some of my early
attempts suffered from the compressor pipe blowing off, the original engine would hardly self-sustain before the pressure
build up was too much for it.
As time allows I hope to develop this demonstration gas turbine further, it can never be used as a propulsion engine
as it is far to heavy but with a more secure compressor pipe I think it will spin faster. It exhibits all the characteristics
of any other gas turbine and was built at a fraction off the cost of a commercial unit or even a model aircraft turbo-jet.
The cost of the project is only £100 or so as the turbo bits were scrap surplus units.
I have tried a perspex top to the combustion chamber to perhaps see into it during operation, this seems to work and does not
get hot. A blue glow can be seen through the air holes in the top of the flame tube but the holes are not large enough to give
an idea of flame distribution.
I would like to revisit liquid fuel as some point, I may try fitting a burner/atomizer from a "Solent" gas turbine starter
unit but this is cheating as it means that I am fitting components which originate from highly specialized aircraft
systems and not "Junk yard" automotive parts.
The home built engine works well, but is not very elegant and needs all manner of services to get it working, what I really
wanted is a commercial small gas turbine engine which is electric start and runs on Kerosene. I find small aircraft gas
turbines most interesting and satisfying to operate.
Gas Turbine Rear View
Starter Coupled to Engine
30/12/1997 The engine is now running very well. I have replaced the compressor delivery pipe with a new stainless
steel item and the joints are now fabricated using special turbo-charger hose purchased from a motor sport shop. A
colleague of mine has very kindly built me a new oil pump connection block. The oil pump now bolts into this aluminum block which ducts the oil in and out and provides mountings for the oil pipe fittings. A new shaft seal is fitted to the pump and the unit
is very oil tight.
The turbo has now run up to just over 80,000 rpm, at this speed it produces about 0.9 bar of boost pressure. At this speed
the pressure rise increases with compressor speed very rapidly. I believe that the engine will go even faster, the exhaust
temperature at 70-80,000 rpm is quite low at about 450 degrees C lower than at slower speeds. The steady exhaust temperature suggests that the unit is functioning efficiently at high speed. I will find out what the limits are for this type of
turbo, it's a fairly old-fashioned unit so I guess that I'm not far off the turbo's limits.
The turbo is getting pretty loud at high speeds and is fast approaching my beloved Garrett GTP30 in terms of noise levels.
The limit to the running time appears to be the oil temperature. The oil capacity is fairly low (About 1litre) and so gets hot
quickly as it circulates through the hot bearing assembly. A future improvement will be to fit an oil cooler with electric
fans. I will also have to fit an oil temperature indicator fed from a thermocouple fitted inside the oil tank.
Developments to Date
In 1999 my brother built me a high energy ignition system to my good friend Roger Marmion's design. The unit uses a surface discharge plug taken from a racing car engine, light up tests have shown that this arrangement is superior to the high tension systems previously adopted. The ignitor works by using an inverter to charge up a 2uf capacitor, a special trigger circuit provides a low energy spark which ionizes the air and causes a high curent flash-over at the spark plug tip.
|Home Built High Energy Ignitor In Action!|
This home built engine has more or less been abandoned in terms of development in favor of more APU type units. It represents the early years of my gas turbine development and is a personal historic artifact. The home built engine is also a great mechanical engineering achievement and the for-fillment of a long held dream to construct such a unit.
Oil Pump Mounting Block
Stainless Steel Compressor Delivery Pipe
Links to other Turbo Based Gas Turbine Projects
Roger Marmion's Turbo
Wooshmaster 2000 Rotary Blowtorch
John C Williamson
Kevin Wolfes Smart Looking Turbo-Turbine
Larry's Home Made Jet
Nick Haddock Turbocharger Fun!
New Jan 2005 PowerLabs
DIY Turbo Homepage