Building a Disk Turbine
The parts of a simple model turbine. Rotor and backplate on left, stator to rear, and cover plate in front.
|April 13, 2003 Cont'd. --
In some of these experiments there was barely enough capacity to spin a 10 inch rotor up to 1000 RPM before the reserve in the air tank was completely depleted. I have seen one thesis presentation where a diesel powered industrial compressor was borrowed for a day to do test trials. Even then, the test turbine still only reached 9000 RPM. Normal operating speed for a 10 inch rotor in a working engine might be 20,000 to 25,000 RPM, with idle speed at 8000 rpm. Clearly tests at 1000 RPM can provide almost none of the information that would be applicable to a functioning turbine engine, other than the fact that a disk rotor turns when a gas is directed at it. The efficiency and aerodynamics of turbine disks and gas flow are not representative at such a small percentage of working RPMs.
A second factor was the cost involved. I have read of several projects in that size range costing from one to four thousand dollars, generally utilizing stainless steel laser cut rotors. Many are based on a popular experimental design featured by one of the first Tesla turbine building societies.
A third factor was the successful development of model airplane gas turbine engines using a (non-Tesla) radial compressor stage and an axial hot turbine stage. These are complete functioning miniature gas turbines. They commonly run at 70,000 RPM to 105,000 RPM and feature turbine and compressor rotors in the 2-1/2" to 2/3/4" size range.
Kurt Schreckling's book available from Traplet Publications, in the U.S. (800) 695-0208
Unlike the Tesla engine, a lot of verified hard data is available for these small engines, and several are in actual production.
Taking these three considerations into account, it seemed to me that a reasonable size to experiment with was a disk rotor of about 2-3/4" in diameter. A turbine of that size could be expected to spin up to a useful speed with a normal shop air compressor, and the data obtained compared with the same size conventional radial and axial turbines used in model engines. In fact, it might prove possible to combine model engine components with a successful disk turbine to yield a hybrid engine for testing. Incidentally, a model turbine engine of this size is capable of developing nearly 2 horsepower, which is a useful amount for creating a small portable generator. It is by no means a trivial amount of power.
(c) Copyright 2003, Stephen Redmond, all rights reserved
|April 14, 2003
Having decided on the disk size, I then thought about how to go about building it. Conventional turbine engines have three main stages, a turbine compressor in the front, a combustion chamber, and a power turbine at the rear. I was mainly interested in using the Tesla disks in the compressor stage. I had a suspicion that a conventional axial flow wheel was better suited as a power turbine, while the Tesla disk rotor might be more appropriate for the compressor section.
Because a compressor doesn't require the heat resistance of a power turbine, I decided to build my first model using cast aluminum for the housing and disks. These would be unsuitable as a fueled power turbine because of aluminum's 1100 F. melting point, but would be feasible for compressor mode tests, and might even be useful for steam powered tests as well. I estimated that sufficient steam at temperatures in the 300-400 F range might be sufficient for an engine of these dimensions.
I had by now a year's experience in casting aluminum, and if I used that method, the cost to build would be low. I had managed to scrounge 150 lbs. of aluminum pistons from an engine rebuilder for $10. This has lasted me through the entire lathe building project including construction of a center rest, chuck, faceplates, angle plates, core plates, fly cutting heads, and other accessories, the milling attachment, and now the disk turbine. I probably still haven't used half of the pistons yet.
I've found that the alloy used for pistons pours and machines very well, and the resulting castings are hard and strong. This seems reasonable considering their original application. I would caution other experimenters not to use mixed scrap aluminum for similar engine experiments, but to try to locate an engine shop for scrap pistons. Soft aluminum is definitely unsuitable for this application.
I would say that my cost to build the turbine has run to under $20 so far -- the main expense being a bag of charcoal briquettes. Also needed was 6" piece of 1/2" drill rod for the turbine shaft, a couple of bronze sleeve bearings, six bolts, and a half inch shaft collar. That's all there was to it.
I decided to depart from Tesla's standard engine form, which features bearings and ports on on either side of the main rotor. A conventional turbine engine has the compressor forward of the first main bearing, with one inlet at the front. That was the form I chose, since it is also the configuration of model airplane gas turbines. By using the same layout, details that have already been worked out for these miniature powerplants might then be applied to a Tesla version.
I designed a rear housing with a long shaft tunnel with two bronze sleeve bearings that I happened to have on hand. I expect to switch to ball bearings in the future, and sized the tunnel to accommodate them. The model airplane turbines use precision metric ball bearings. The Cairns turbine uses bronze sleeve bearings, and reportedly turned up to 50,000 RPM, so it seemed feasible to get initial results with these. I didn't expect high efficiency in initial tests. Like most home engine builders, I first just wanted to see my engine spin in the lower speed ranges to start with. Refinement would follow.