Oops!

After getting some advice, I decided to remake the Combustion Chamber.

The old one wasn't working properly, the gas mixture appeared too rich, with yellow flames visible at times.

So I made this one, with much smaller air-holes.

I think the theory is that the air is forced to move faster through the smaller holes, giving more turbulence and better mixing.

Anyway, it seems to work.

The flames are much bluer-whiter and the turbine seemed to have more shaft power.
I say 'seemed' because I got carried away with the throttle setting in an attempt to reach maximum speed and before I could measure anything the back end of the engine suddenly vomited a shower of flame and molten metal fragments!

Here's what's left of the NGV assembly. It appears that several of the stainless steel blades melted and exited through the turbine wheel.
Fortunately, they melted so completely that the turbine itself was undamaged (after I scraped the gobs of molten metal of it!)

That's despite the fact that it stopped so suddenly that the drive shaft (which continued to rotate) unscrewed itself completely and continued to spin.

It's going to be a few days before I finish making a new one.

The new NGV under construction

A carbon fibre compressor impeller

Click the picture for a larger view.

Just to let my readers know that I haven't given up and gone away ;-) I'm posting this composite picture of the stages in making a CF impeller.

The reason it's been a while since I posted is that I'm remaking the entire engine compressor section and some of the parts take a while to make. The impeller in particular. It doesn't take long in terms of time spent, but the bonding/glueing processes take 24 hours at each step.

The process starts (frame 1) with an aluminium boss shaped on the lathe with a nice smooth 'conical' curve to help the airflow. A similarly shaped laminar plate for the coverplate is made in CF (two layers). In the shot you can see the two parts of the mould I used for this job, turned out of a bar of nylon.

Frame 2 - The boss and cover are slotted to take the curved blades. The slots are started using a 'Dremel' tool router bit (0.8mm). I roughed them out using a piercing saw then finished them with the router bit, this time using a pillar drill at high speed.

Frame 3 - The blades are cut from a sheet of 2 ply CF. I make this sheet by laying up two layers then clamping them between boards. (I've found the use of peel-ply to be very helpful in all of this work but it produces a rough texture that I try to avoid on the 'airside' surfaces by using polythene sheeting. - one side of the sheet is rough(peel ply), the other smooth (poly).)

After cutting, I glue the blades into the boss and leave to set.

Frame 4 - The fiddly bit! The cover plate has to be persuaded to fit over the blades and eased down to be parallel with the base boss. A spacer is inserted between the two parts and helps a lot with setting the plate correctly.

I could probably get more accuracy using jigs here but the end result is acceptable for the prototype.

Frame 5 - After glueing and hardening, the blades are trimmed down to the cover plate (using the Dremel again and finishing with a spin in the lathe using glasspaper) and then (Frame 5) a final layer of CF is bonded over the cover plate.

Frame 6 - The impeller is mounted in the lathe and the inner orifice is roughed out allowing the spacer to be removed.

All that remains is another layer of CF to be bonded over the bottom of the boss for added strength and the whole thing can then be trimmed to size and polished up.

Finally! - some help

I posted a query on the DIYGasTurbines group at Yahoo and got some helpful advice.
(Thanks guys)
It seems I need to really 'push' this thing - as much starter speed and fuel as I can - push it right up to the highest speed I can get.

Before anyone get's excited - I didn't get it going, but I think I got a lot closer and identified a couple of problems.

Taking it up to maximum speed on the big blower and opening the throttle (carefully) to the max, I got over a foot of water pressure in the casing - 0.03 bar. I had taken the tacho off in order to get a better airflow into the intake so speed is unknown - although that pressure corresponds to about 30,000 rpm.

When I tried to remove the blower, I got a huge roaring, drumming sound from the front that I could feel in the air around the air intake. I believe this is compressor surge. In which case, the impeller is unable to handle the back-pressure from the engine, leading to a rhythmic stall-and-recover from the blades.

In a way it's good news. It means the compressor is close to the working region.

I'm going to rebuild the front of the engine, impeller and diffuser, to a slightly different design to give a smoother airflow.