A period of turmoil in my life is slowly resolving itself. After buying a house and moving in May last year, I was made redundant only three months later. Bummer. I spend a while in Canada with my family, then was and still am very busy with designing a very cute but incredibly complicated bit of electronics for Empeg Ltd., a company started by a good friend of mine. As a result, it's been a long time since I have been in a turbine mood, but today I had some time to play
I have been acquiring more turbochargers slowly over the last few months, whenever I have some free time to poke around one of the local scrapyards. Recently I had a lucky find, a virtually new Holset WH2D, after a chance meeting of a mechanic from a local truck repair company. Apparently the thing was retrieved at great expense from a truck shipped to Italy, after allegedly failing. When they had flown someone out to replace it, brought the original back, and tested it, they found nothing wrong. The truck ECU was faulty!
Bad luck for them, but great for me. They couldn't reuse it, since it was now second-hand, and so it sat on the floor of the workshop for two years until I came along, and offered 20UKP for it. The turbo was filthy, but the bearings seemed, unsurprisingly, to be in perfect condition. I got it home, stripped it completely, and sandblasted the compressor and turbine housings to remove the crud. They came up looking like new, and after cleaning out the soot and dead spiders, and oiling the innards, I reassembled it with new stainless steel bolts. Result: One practically new turbocharger, for practically nothing.
It was sitting on my kitchen counter for a couple of weeks, but today I decided to try temporarily attaching the combustion chamber from my original engine, to see it it would run the new turbo. It's considerably larger than the old KKK unit I have been using, and the ratio of compressor to turbine is noticeably different, so I was unsure whether the combustor would be man enough for the job. Nevertheless, I tried fitting it, and was pleasantly surprised to find that it fitted nearly perfectly. The compressor inlet tube was eerily bang on to the outlet from the turbo, and the output of the combustor could be persuaded to match up to the turbine inlet with one bolt from the original, and a couple of clamps.
These images show how it all went together. The original engine is there so I could use the ignition system, without removing it from the engine frame. The wires were just long enough to reach, luckily.
As you can see from the photos, this turbo has an integral wastegate. I left the actuator fitted, because I have some ideas for using it. A little duct tape around the compressor joint, to seal any air leaks, and I was ready to fire it up. I fitted a temporary thermocouple, and set up my camcorder looking straight into the turbine from a safe distance, both to record the sound and so I could see the turbine behaviour.
I was using compressed air from a hand-held air jet to spin up the turbo, in the same way as my original engine experiments. I was recently loaned an old dental compressor for an extended period, from a friend who had two of them but only needed or had space for one. It's brilliant for this sort of thing, because it has a much higher airflow than my other compressor, and is very quiet.
I filled the oil input pipe with machine oil and put some old cloths underneath the turbo to catch the output oil, and connected up the propane. The first couple of attempts at starting were only partially successful. I hunted down and sealed up a couple of leaks, and tried again. This time, it worked. The compressor is only just able to produce enough air to spool up the turbo to the sustain speed, which is considerably higher that that of the KKK-based engine. I'd estimate it was around 25kRPM or so. I had to increase the fuel pressure quite a lot, to get it to run, but run it did.
The exhaust temperature was high compared to my first engine, at around 650 deg. C while running at a constant speed, and nearly 920 deg. C when accelerating. Looking at the camcorder tape, this would appear to be from unburnt propane igniting in the output, since the turbine only glows dull red at the blade-tips, rather that bright red all over which is what I would have expected from such a high temperature. There is a visible blue flame coming from the exhaust at high revs, which would support this theory.
Obviously, I will need to design a new combustion chamber more suited to this turbo. However, I proved it works, and in doing so, found that it is much quieter than the KKK-based turbojet, even at higher revs. I had it up to an estimated 45kRPM, and the noise was only a moderately loud whooshing sound, rather than the loud roar the other engine makes at the same speed. I'm not sure what causes the difference, but it may well be down to much smoother airflow at both input and output. It's quite a pleasing sound, in fact, and the turbine whine is rather more obvious than on the other engine, although not intrusive.
I have sampled the camcorder soundtrack during a start sequence, and you can download the Mpeg3 file here (431k). One thing that did show up is how much better the good bearings are than the old KKK turbo's ones. The heavier turbine and compressor wheel couple with good bearings means the shaft spins for nearly a minute after the fuel is turned off, which is quite impressive.
So, what's next? Well, I have been talking to a local company about making a custom epicyclic gearbox that could be fitted to a large turbocharger turbine shaft, and gear it down from around 69kRPM to a more reasonable 3000 to 4000 RPM. I should be able to extract 20 or 30 BHP from such a unit, which would be enough to run a go-kart or small kitcar very nicely. I have also been playing around with designing an ECU for turbine engines, based around a PIC16C77 microcontroller, with Windows control software written in Delphi. I have already written the first test version of the control software, and sent it to a few friends with turbines, for comments and suggestions. I hope to have some hardware and a more finished version of the software in a couple of months, and eventually to put the circuit diagrams, PCB layouts, and software on my web page.
Larger versions of the above pictures can, as usual, be downloaded below. I must apologise for the rather crappy images, but the light was fading and my digital camera doesn't really produce stunning results at the best of times. I'll have to get a better one eventually.
Back to the beginning of turbines. Back
