Combustor Improvements, Nozzle Construction & More Test Results

November 27, 2001

Here we are towards the end of November -- almost a full year of action-packed Turbine Builders Club activities.

Last week we were busy entertaining family visitors, Yooper style, before sending the Trolls merrily on their way. For those of you who are unfamiliar with Michigan's Upper Peninsula, Yooper is a name for Upper Peninsula residents. A Troll is a down-stater who lives "under the bridge" (Mackinaw Bridge connecting two peninsulas), and Yooper-style entertainment involves beer, snowmobiles and shooting guns.

By now most of you should have some semblance of a turbine put together, ready for trials. This month we will cover combustor improvements, construction updates, and more test results.

Combustor Improvements

Let's start with combustor improvements. Photo (a) shows our basic combustor with a new threaded exhaust tube and a couple of reducers. This allows us to step the exhaust down from a 2-inch tube to a 1.25-inch tube. The gas vaporizer was rerouted through the side of the 2-inch tube, making it a lot easier to connect the combustor to the turbine. Also shown in the photo to the right are the spark plug and vaporized fuel delivery tubes.

Photo (b) shows a close-up of our fuel and spark controller/sequencer. (Click on image to view full size.) The potentiometer on the far right controls the frequency of the combustion cycle from about 1 cycle per second to about 100 cycles per second.

The sequencer sends a control power pulse to an electronic gas valve immediately followed by a burst mode ignition pulse packet to the spark coil. We've tested the combustor in continuous and pulse modes. 

Continuous combustion delivers a massive amount of heat with very low velocity and kinetic energy. While this may be beneficial for steam generation, it is not ideal for kinetic energy machines like the Tesla turbine.

Pulse combustion, on the other hand, delivers less heat volume, requires less fuel, and produces a very energetic, high velocity shock wave. While these shock waves will destroy piston and conventional turbine engines, the more robust Tesla design easily withstands and seems to work very well with this type of combustion.

Photos (c) and (d) show our turbine assembled and fed with a compressed air line.

Photo (e) shows me with one of my helpers running a spin test on about 150 psi of compressed air.

Nozzle Construction

You may have noticed from Photo (c) that we are using 1-inch square tubing for our inlet nozzle. This allows us to use nozzle inserts with a horizontal slot profile, distributing high velocity gas equally across the width of the rotor or disk pack.

Presently all of our initial tests use compressed air as the fluid. Refinements to our combustor technology will allow us to eventually move to hot gas.

To shape a nozzle insert, we started with a 3-inch piece of 0.75" x 0.75" square steel stock. Using a small electric hand grinder, the square stock was carefully ground to approximate the cross section of an airplane wing. (See Figure 1). A (0.25-20) thread was tapped through the insert to attach and hold it in the 1-inch nozzle.

In our first test the insert was oriented to configure the nozzle as a convergent type. (See Figure 2)

In the second test the insert was flipped 180 degrees to configure the nozzle as a convergent-divergent supersonic nozzle. (See Figure 3)

Nozzle Test Results

The air compressor we are using is relatively small. It takes about 10-12 minutes to pump up the 30-gallon tank to 150 psi. Even though the nozzle slot is only 0.125-inch by 1-inch wide, the air tank is exhausted in less than 30 seconds. Since the momentum and energy are both related to fluid mass times velocity, we had to use a 0.5-inch feed line to the turbine nozzle to get enough air mass delivery.

Using the nozzle insert in its convergent or subsonic mode, the turbine spooled up with no problem using an initial 150 psi tank pressure. It continued to spin until the air pressure dropped to 20 psi.

Again, reconfiguring the nozzle with the insert oriented for convergent-divergent supersonic mode, the turbine spooled up easily on 150 psi. This time it continued to spin even when the air pressure dropped below 20 psi -- indicating a higher energy transference efficiency when using supersonic nozzles.

The conclusion is that the gas or energetic fluid must reach supersonic speed before entering the disk pack for highest overall efficiencies -- the turbine chamber and disk pack cannot be used as the divergent section of a supersonic nozzle.

Next month we'll review the year's accomplishments, make a few projections for progress in 2002, and make a few suggestions for design improvements -- maybe even have more test results.

We'd also like to hear from others of you who have working turbines up and running, and what results you are getting. We'll only make progress by sharing what we know and discover.