Bits made so far thrown together for a general view of the shape of things..

Flame tube alongside combustor.

Flame tube made from 316 stainless steel, swg22 (0.7mm). 5" in diameter, 345mm long. The calculated tube length given the size of turbo is only 6 times the inlet diameter of 44mm i.e. 264mm, the extra length is to accomodate the point at which the tangetal airflow from the compressor will enter in order to minimise disturbance of the flame.

 There are three general rows of holes, the first ( bottom 2 rows ) is the primary zone, delivering main air for combustion. The second row delivers air ensuring that the combustion process completes. The third row is the dilution ( Tertiary ) zone, ideally all fuel is burnt by this point and the air is used purely to mix in with the hot combustion gases so that it can be heated and reduce the temperature of the exhaust gases.

The holes were positioned by drawing the layout on a piece of paper and then wrapping this around the tube. The centres of the holes were then marked with a centre punch and drilled out. The total area of the holes combined, is equivalent to the area of the turbo inlet. The distribution of the given area into the three zones is 30% Primary, 20% Secondary and 50% Tertiary. Hole size diameter is roughly in the ratio1 to1.5 to 2.5 for Primary, Secondary and Tertiary respectively. Generally, small holes allow air to penetrate a relatively short distance and are used for cooling the tube walls, larger holes allow the air to penetrate more deeply for mixing purposes.

 This will be the source of power for the oil/fuel system instead of using batteries and electric motors. It is a Russian made model aero engine of size 40 ( 0.4cu/in or 6.5cc ) which are relatively cheap and easily available. Engines of this size have a maximum power output of about 1hp which is a lot for it's small size. I will only need about half this power to drive the oil and one fuel pump. A larger engine may be needed when I come to drive a second fuel pump for the afterburner.  The engine will also have to be cooled with a separate fan probably from a computer as the cooling is normally provided by a  propeller.

Two 10mm thick pieces of mild steel bar sandwiched together for shaping of the bearing mounts. I had to make these by hand because I couldn't find pillow blocks for the size of bearing ( 22mm ) I am using. The bearings will carry an 8mm diameter shaft fashioned from steel drain rod ( used for unblocking drains ), the power for the fuel and oil pumps will be taken off this.

These are some of the power transmission components bought from a single supplier. They consist of small ( 10mm wide ) timing belts and pulleys and universal joints for coupling the shaft directly to the fuel pumps. The shaft is made from 8mm diameter drain rod cut to length and reduced in diameter by a few thousandths of an inch. I found that these small timing belts and pulleys these were the best way to transmit the power to the various components in terms of their flexibility for changing ratios and compactness.

Delta VU Universal fuel pump from an oil fired domestic boiler. Rated at 18bar ( 260psi ) max pressure @ 2800rpm, power consumption 80Watts, although in practice I was able to push the pump to nearly 400psi with a corresponding increase in flow. I disabled the fuel cut-off solenoid and altered the pressure regulator to act as a relief valve  to cut fuel flow when max rpm is reached.

I discarded the original filter housing found at a scrap yard because it was too bulky and had no provision for attaching oil lines for which I had to make a separate face plate. This new one has 1/2 BSP threads with two bosses for tapping out to take pressure and temperature gauges.

Home made regulating valve made from two 1/4 BSP to 10mm tube couplings. The two BSP threads are screwed into the barrel nut sandwiching the spring an ball bearing between the couplings. The spring pushes the ball bearing against the conical face of one of the couplings providing a relatively oil tight seal. Oil enters from the right. Relief pressure is adjusted by inserting or removing washers between the couplings and barrel nut.

 Complete re-design of the Oil/fuel system using a model aero engine. Model engines are generally noisy but this one is surprisingly quiet as it is driving a flywheel as opposed to a propeller. The flywheel is made from a spare motorcycle clutch spacer which  had an ideal weight for starting as well as threads going through for connecting to the back mounting plate and small oil holes  in which to insert a shoelace for starting! The whole unit shouldn't weigh more than about 5kg, about the same as the turbo itself. The steel mounting plate is 6mm thick and has yet to be finished off.

 The main shaft ( not shown ) will be driven by the pulley on the engine with a larger pulley wheel on the shaft giving an engine/shaft ratio of 2:1. With the engine running at about 6000-8000rpm this will give a shaft speed of about 3000- 4000rpm the maximum rating for the fuel pumps that will be driven directly from it using universal couplings. The belt for the oil pump will be driven by a small pulley off the main shaft with a ratio of about 8:1 giving about  500rpm pump speed which is more than enough to provide the required oil flow.

Views showing orientation of the oil pump ( KYB power steering, positive displacement vane type ). The pulley wheel on the oil pump is the standard wheel but had to be modified slightly to take the narrow ( 10mm ) timing pulley. I did this by working my way around the wheel with a vice, squeezing the V-groove together so that it matched the width of the timing belt. I then back-filled the groove with a metallic epoxy resin so that the timing belt had something to sit on. The result grips the belt very nicely with virtually no slippage.

Further work on the oil/fuel system. The above pictures show the fuel pump ( oil fired burner pump ), pillow blocks for drive shaft and jaw coupling to connect the shaft to the fuel pump.

The mounting plate has extra space to accomodate another fuel pump on the opposite side which will be driven by the same shaft. This extra pump could be used to fuel an afterburner or combined in parallel with the first pump for greater flow capability. This particular fuel pump can be driven in both clockwise and anti-clockwise directions.

 Almost finished Oil/fuel system mounted in frame with turbo positioned roughly where it will be when the test rig is complete.

The aero engine needs cooling so I made a shroud from 60mm diameter tubing and used a 12V CPU cooling fan to provide the air throughput.

Clutch Mechanism

Click for here animated version of above sketch

This part was to be the aero engine mounting for a direct drive starting device ( see sketch above ). Made from 5mm steel plate and 3x10mm M10 pieces of studding. One of the plates would connect to the mouth of the compressor housing via a short piece of pipe. The other plate would have supported the starting motor. The 3 pieces of studding allow the motor to be oriented squarely to the nut on the compressor wheel. I have now shelved this idea for a less complicated indirect method ( see next panel ).

Experimental onboard starting device similar in concept to my previous unsuccessful attempt to make a blower ( see Phase 1 ), but employs a much more powerful motor and larger compressor. The engine is the same as the one used for the fuel/oil system ( 1bhp ). The compressor inlet diameter is 60mm ( 2.35" ) which theoretically will flow as much air as the two compressors on the smaller turbo's combined. Whether this will provide enough flow/pressure to start one of the turbo's remains to be seen..

My own design for a way to vary a jet nozzle opening. Finding the right nozzle size is a bit of a trial and error process, so rather than having to make up lots of different sized nozzles or using a 'turkey feather' arrangement, I thought this may be a more convenient way of going about things. The smaller inner cone is designed to give a slightly undersized jet nozzle and the larger outer cone giving an oversized nozzle when fully extended. The ideal nozzle size will be somewhere in between which would be found through trial and error. It may be possible to use this design for the nozzle of an afterburner where the smaller cone would be sized to the ordinary jet nozzle which would be used when running "dry", i.e. without the afterburner switched on. When the afterburner is turned on ( running "wet" ), the nozzle size has to increase to accomodate increases in jet mass flow and velocity. This could be achieved manually by sliding the outer cone accordingly or have some sort of automated mechanism ( spring loaded? ) which would open when the pressure in the afterburner can increases as the extra fuel is burned so acting like a pressure regulating valve.