View Full Version : Tesla Turbine

01-13-2002, 11:10 PM
Does any one know what test fluid, pressure, & temperature the author of this article was using to get his test results?

Tim Clarke
01-14-2002, 06:55 PM
I met the builder, and watched the prototype
run, at the P.R.I.M.E. show last fall. He was
running it on compressed air, that was plumbed throughout the display area. I assume that is was line pressure, I don't remember seeing any pressure regulator. I doubt the line was running much over 125lbs.

He had a generator coupled to it, and was using 3 flourescent lights for a load

Tim in Oregon

Jeff Maier
01-14-2002, 08:20 PM
I used air at 120psi to get my published results. I could get 130V with about a 30W load.
The pressure at PRIME was a bit less than 50 psi. That was enough to light a compact fluorescent bulb at around 90 to 100V.

--Jeff Maier

01-14-2002, 10:06 PM
Jeff: I was curious because I wondered what was limiting your RPMs. With air at essentially room temperature you might be suffering from critical flow restrictions in your elbows and the nozzle. I don't know the I.D. of the elbows but looking at the photos it strikes me as a possibility. At room temperature in air, it appears to me that your rim velocity is very close to the critial V in air at those conditions.

Jeff Maier
01-14-2002, 11:42 PM
I agree. The optimal speed at the rim of a Tesla turbine runner is about 750 fps; for my turbine that's about 24,000 RPM. My nozzle is a very simple, and certainly not optimum, design. I'd also bet that straightening out the inlet path (no elbows) would make a difference as well. I've been reading up some on nozzle design and am planning some experiments with re-designing the one I have. Hopefully I can get the speed I should get out of it.

I also found that removing the cover plate significantly reduced RPM; this tells me that air is blowing around the outside edges of the runner, and the cover plate keeps it in place and it adds to the force on the runner. Ideally, all the air should go into the inside discs and not around the outside, so a cover plate on or off should make very little difference. Thus, my nozzle does not direct all of the air flow where is should be. I'll also bet that it doesn't turn all of the air pressure on the inlet side into velocity. If I can do better at both direction and velocity, I'm sure that will translate into more RPM.

01-14-2002, 11:53 PM
I was wondering if you had tried putting the plates closer together and if this affected output. Pictures I have seen of commercial units (pumps) had close spacing and many more plates. If I remember correctly the rims of the plates were in close proximity to the housing as well. Are the rims of your discs square or rounded and does this have any effect on output? Neat project nicely done!


01-15-2002, 09:00 AM
Jeff: I meant that your rim speed is very close to the speed of sound in air at the test conditions you were using. You will need a sophisticated nozzle design to do any better. It may not be the turn in the elbow
but the I.D. that would cause a problem. You may be trying to exceed the speed of sound in the elbows. I haven't the time right now to look up the formula, but the pressure drop is high in that situation.

Jeff Maier
01-15-2002, 08:56 PM
According to the formula I used for torque (derived from Tesla Engine Builder's Assoc. literature), the torque varies with the reciprocal of the distance between the discs. In theory, the closer they are spaced the more power you'd get. I'm sure there's a limit. If I understand correctly, the working fluid adheres to a smooth surface in a boundary layer. The discs need to be closer than the thickness of this layer so that torque is exerted on both surfaces. I'm sure the ideal spacing differs depending on the fluid; air vs. steam for instance. The ends of my discs are square; I don't know if rounding makes a difference, but all the pictures I have seen appear to have squared off edges.


I had never thought of the air going supersonic. If the rim speed is 750 ft/s, then the air velocity could well need to be supersonic. That would complicate things (is it even possible?)


01-15-2002, 11:34 PM
Yes, it is possible, but a very sophisticated nozzle is required. I think your airflow is being limited somewhere in your supply sysrem, the valve, elbows, or unintended restrictions. As far as air bypassing the runners into the exhaust, a long time ago I read an article about the Tesla Turbine and it was that author's opinion that the inefficiency caused by this was the reason the Tesla is not the turbine of choice for large power plants. It is obvious that it would be if iot coild be made highly effeicient, since it is so much cheaper to manufacture than bladed turbines.

01-16-2002, 12:37 AM
Thanks for the reply. Have you tried your turbine hooked up to a high pressure water supply instead of air? It should work better with water than air because of viscosity. A venturi (rocket nozzle) would greatly accelerate the air and could be worth investigating. At 750FPS you are only at 511.36MPH well below the speed of sound.

There seems to be an ommision in the formulae on pg.21 - what does the greek letter (mu) just after Pi represent?


Jeff Maier
01-17-2002, 09:36 AM

I got the mu from the TEBA literature; it is given as the "dynamic viscosity" of the fluid with a value of 1.79 X 10-5 Ns/m2 for air.

I haven't tried running the turbine with high pressure water. Other than my own household plumbing, I don't have a source for it.

Interesting what you read about air leaking around the sides of the runner; any air that did this would certainly be lost, and you could end up wasting a significant portion of your energy source in this manner.


[This message has been edited by Jeff Maier (edited 01-17-2002).]

01-18-2002, 02:10 AM

Thanks for the info!


John McClain
01-18-2002, 09:49 AM
As I understand it, the nozzles need to be as close to the runners as possible, the nozzle orfice needs to be about the same diameter as the space between runners, with an orfice for each space, and the cross section of the inlet line needs to be larger than the total cross sectional area of the nozzles added up. I would use hypodermic tubing inset into a block to space the orfices to match the runners. If the nozzles are close enough, and aimed tangent to a circle just inside the O.D. of the runners then there will not be a problem with pressure running around the outside of the rotor, and you will need the space between the rotor and the inside of the case to be greater than twice the separation between runners to minimize adhesive reaction between the ambient air adhering to the case and the peripheral air which will inevitably linger around the outside of the rotor inherent in the design due to frictional losses. The interaction between these two air masses will be frictional in nature and will be greatest when the distance between elements(static and dynamic elements) is less than or equal to double the theoretical film thickness of the fluid, in this case the air which powers the engine. Theoretically, excluding friction, the adhesive qualities of fluids allow us to direct a jet of fluid between the runners, with the adhesion converting the potential energy of pressure into kinetic energy of rotation, as the rotor obsorbs energy from the stream, it loses velocity which means it also loses centrifugal force said force being motive for the fluid to remain at the periphery, therefore the fluid will seek the center of the rotating mass as it transfers energy. As long as you exclude friction as an element of the equation, all of the fluid will exit out of the center and theorectically be at 0 psi, it is the friction losses which cause a small portion of the fluid to remain at the periphery and thus impede the rotation of the rotor. Basic thermodynamics says that the higher the energy potential difference the greater the effiency possible, this is particularly important in engines as they have to deal with heat losses directly through conduction, convection and radiation although radiation probably doesn't affect us very much in this instance. Compressed air has a relatively small amount of energy compared to steam or water due to the effect of mass as well as the differences in temperature. This engine is far more efficient driven by steam or water than just air.

02-13-2002, 08:42 PM
<font face="Verdana, Arial" size="2">Originally posted by Jeff Maier:
The ends of my discs are square; I don't know if rounding makes a difference, but all the pictures I have seen appear to have squared off edges.


If you'll allow a newbie (both to the board and machining) to stick a few words in here, I'd say that the ends of the runners need to be square, since that seems to be where the initial impetious comes from. Rounding the edges would make it easier for the airflow to slip around and escape through the exhaust without doing any work. There was an article TEBA News a couple of years ago about a Tesla Turbine that used serrated edges (thing looked like a bunch of radial saw blades) the idea being that the edges would increase starting torque for the turbine. The problem with that, is that the edges also increased the drag, thus cutting down on the efficiency of the turbine.

In theory, one could serrate the edges of the runners like those on a US quarter, which would improve the starting torque and still keep the drag caused by them relatively low. The biggest problem with Tesla Turbines is that no one with lots of money has bothered to do any research on them.

From my research (and it was the Tesla Turbine that got me interested in machining), it looks like that the turbine works based on a combination of the Bernoulli principle/Coanda effect/or something else. There was also an article published in Popular Science back in the '80s which talked about a cylindrical sail. This was basically a vertical cylinder hooked up to a generator which revolved due to a low pressure area that formed behind the windward side, the Tesla Turbine might work based on a similar principle.

In order for us to find out what's really going on with these things, one of us "armchair experimenters" is going to have to get up and make enough of the things that the folks with the money and equipment to do a serious study decide to pay attention.

02-24-2002, 08:28 AM
Jeff: Just got the latest issue of HSM. With the dimensions of your nozzles I will do some air flow calculations to see if I can tell what is limiting your RPM. If anything interesting shows up, I will post the results here.


02-24-2002, 09:31 PM
Stupid question:

Why would there not be a nozzle exit for each slot between wheels? Article has two nozzles, but there's more slots, what's up?

02-25-2002, 12:08 AM
Just easier to do one than many (KISS).


02-26-2002, 10:40 AM
Jeff: The following contains some symplifying assumptions. I get a total nozzle area of 0.0186 sq in and the annulus area between the runners and the casing of .05624 in sq. This gives a ratio of 3 for the annulus to nozzle areas. Using the conservation of mass, the V(nozzle)X A(nozzle)= v(annulus)X a(annulus) then V X 0.0186 = v X (.05625) or v = V X .33.
Assuming the air temperature going to the nozzles was about 72 Deg. F the the velocity of sound at the nozzle would be about 1140 ft/sec. The air velocity in the annulus would then be about 377 ft/sec. The rim velocity at 12000 RPM is 366 ft/sec for a slippage of about 3%. This sounds reasonable. The slippage at 6000 RPM is about 51%. I calculate that if you doubled your nozzle area you would get 24,000 RPM no load, and tripled, you would get about 36,000 RPM. Fred