I haven't had a lot of time to play with turbines for the last few weeks, due to pressure of work. However, over the last weekend I added an exhaust nozzle and a prototype afterburner to the engine. The exhaust nozzle is a 14cm piece of 76mm OD 3mm wall carbon steel tube, with about 1cm of one end turned down to a diameter of approximately 74mm, which makes it a tight friction fit into the exhaust port of the turbocharger. Although the turbine casing expands significantly when the engine is hot, the pipe does as well, so the fit stays tight.
The exhaust nozzle has a set of flow straightener vanes inside it, at the turbine end. The are made from 1.6mm stainless steel sheet, the same stuff I used to make the combustion chamber ends. I cut three rectangular pieces 50mm wide and 70mm long, and bent them into a V shape across the middle of the longer dimension. This gave three 60 degree folded pieces, each 50mm wide and 35mm long (the internal radius of the exhaust tubing) on the arms. I then clamped them together equally spaced out, meeting along the folds. A small tack weld was then made at either end where all three parts met. The end result was an assembly with six equally spaced fins 50mm wide, of 70mm overall diameter, which was a friction fit into the nozzle. I then bent the end of each of the fins that was furthest away from the turbine to about 25 degrees in the opposite direction of the turbine rotation. The whole assembly was then ground down to be an exact fin into the nozzle, and tack-welded inside, leaving about a 2cm gap between the turbine and the vanes.
The results of the completed nozzle are quite good. The exhaust flow is much tighter, giving a coherent jet for about 1.5m past the end of the nozzle. The exhaust gas had very little residual rotation, which can be seen easily in the smoke trail produced by the afterburner. The only problem is that the steel pipe will probably eventually oxidise past the point of usefulness, or simply melt. I will have to replace it with a fully stainless steel version at some point.
The above system can be seen in the pictures below. The one on the right show a view up inside the exhaust nozzle, and the flow straightener vanes can clearly be seen. In addition, as you can see I have put PCL type quick release fittings on both the starting air inlet and the propane inlet. This allows me to set up the engine very quickly and easily.
The afterburner prototype is very simple. It consists of about 2m of 3mm ID soft copper tubing (car brake line) which is wrapped several times around the turbine end of the exhaust nozzle. One end of the tubing is the inlet, which is about 30cm long to dissipate the heat so it won't melt the plastic tubing connecting it to the fuel pump. This is fixed to the engine frame. The other end of the copper tube is bent up 90 degrees, and then forms a loop pointing down towards the exhaust nozzle about 4cm above it. This has an inline compression coupler (again this is a replacement car part) attached to it, to make a point at which the system can be separated from the injector. This assembly is a simple heat exchanger, which is heated to about 400 deg. C by the exhaust gas flow.
The injector is a short piece of the same copper tubing, which has one end crimped shut. The tubing is bent into a ring about 30mm in diameter, with a stem that goes through a hole in top of the nozzle and attached to the other end of the compression coupling. The overall shape of the injector is much like a question mark, but with a complete loop rather than half of one. The loop has a number of 1mm holes drilled around its periphery, and it is mounted concentrically within the exhaust nozzle, just in front of the flow vanes. The end result has the small holes pointing at right angles to the exhaust gas flow, which gives good mixing of the fuel vapour and hot air. The coil can be seen in the picture above on the left, and the injector can just be made out inside the nozzle in the picture above on the right. As in the case of the nozzle itself, the injector at least will definitely need to be made of stainless steel of a similar high-melting-point alloy, to stop it melting in operation. However, for ease of testing, the copper tubing is ideal, and I also conveniently had a coil of it lying around.
The theory of operation is that diesel fuel or kerosine is pumped through the heat exchanger coil, where it vaporises near the output end. The high pressure vapour squirts out into the exhaust gas, where it mixes with air and hopefully ignites.
My first tests used a 30cc syringe to pump the fuel through the heat exchanger, both because this gives an easy way to measure the fuel consumption and also because it limits the amount of fuel available in the event of a cock-up. This will eventually be replaced with an electric fuel pump. I used straight diesel fuel for these tests, mainly because I had a gallon of it in the spare fuel can for my car.
To test the afterburner, I started the turbine and ran it up to about 25000 rpm. I let id run for a couple of minutes, to warm up, and then gingerly depressed the syringe plunger. After a few seconds, I got a small amount of fuel into the hot bit of the heat exchanger, and it behaved exactly as I had hoped, in that I got a large cloud of smoke out of the back of the engine. This showed that the fuel was vaporizing correctly, since there was no liquid coming out. I then squirted fuel a bit more vigorously, and was rewarded with a much larger smoke cloud, and a short burst of hissing noises from the injector. A bit harder, and I got a really enormous smoke cloud and a sustained hiss, but still no flame. I then threw caution to the winds, and put about 15cc of fuel through in about a second. That did it. You can see the results below. This picture is a video grab from the camcorder I was taping the tests with, and is not the best possible quality image. The other thing to note is that it was a bright sunny day, and the camcorder still stopped down considerably due to the brightness of the flame!
The blast of flame out the back of the engine was accompanied by a bass 'whoomp' sort of noise, and the entire engine jumped about an inch forward. It only lasted for about half a second, which works out to a really silly fuel consumption of about 1.8 litres per minute, or 108 l/hour (half an oil drum full), which is a bit worrying since the turbine was effectively at minimum revs. I repeated this test a few times, and found that ignition of the diesel fuel was actually quite difficult. The EGT of this turbine is only just high enough to ignite the fuel mix, and at minimum revs the engine tends to flame out when the afterburner lights. I found that if I opened the propane valve quickly just as the fuel was being injected into the afterburner, I could make it ignite quite reliably from the short burst of flame through the turbine, and the flameouts were not as common. I suspect it would work much better at higher revs, which will have to wait until I finish making the liquid fuel system for the main combustor. I simply haven't had time to do this yet.
The next step is to fit the afterburner fuel pump, which will allow me to run the thing in sustained bursts, and probably melt the exhaust system. However, these tests have proven that the design is functional, which was the main goal.
Larger versions of the above pictures can be downloaded below.
In addition, the file jet2.mp3 (114k) is a better sound sample of the engine starting, running, and stopping. This was sampled from the sound track of the camcorder tape, which was taken from the front of the engine at a distance of about 2m. Unfortunately, I haven't yet managed to get a good sample of the afterburner ignition, again due to the microphone overloading from the loudness of the noise.
Back to the beginning of turbines. Back
