Solving the Tesla Turbine Bearing Temperature Problem

[JoeEngineer]

I know getting funding is probably going to rub some of the guys funny in this forum-

But as I have stated before- while basic turbine operation is reasonably straight forward- competitive optimization is highly complex. This is not going to get anyone rich quick- if it were that simple someone would have done it long ago.

I feel this is a very promising technology- but not one that is going to get developed by someone selfishly trying to keep profits to themselves. Success will ultimately be a group effort (just like most technology start ups) and eventual profits will be distributed by many. I know inventors hate the idea of others profiting on their ideas, but- a wise person once told me- "its better to have part of something than all of nothing."

 

[dr dodge]

Wanted to hear an update on the Ronl saw turbine
that was my idea too

dr

 

[NateD2]

Just wanted to post an update to let everyone know I'm still working on things. Right now I'm working on some programming for the data collection.

 

[dr dodge]

If you need instrumentation consultation, let me know.
I have some resources available

dr

 

[Scurvy]

I found the following article...

Flat Disks: Now here is the true real killer of Tesla turbine efficiency. Imagine a simple Tesla turbine, a stack of 11 flat disks each 10 inches in diameter with a .1 gap between them and an exhaust hole in the center 1 inch in diameter. The surface area for the driving fluid entering the Tesla turbine is the (Circumference (or Diameter times Pi) * the width of the input area) = (10*Pi)*(10*.1) = 10*Pi square inches or about 31.4 square inches. But the exit hole is only one tenth the size. So any fluid entering a Tesla turbine gets slowed down while inputting energy into the Tesla turbine and at the same time must exit from the smaller area of the exhaust hole. But notice it gets worse, the area of disk gaps doing the exhausting is (1*Pi)*(10*.1) = Pi or about 3.14 square inches but the area of the hole is .5*.5*Pi or about .79 square inches.

Exhaust Hole Size: To solve the two above problems we have to do two things, both affect the efficiency of the turbine. First we have to make the Exhaust hole closer in size to the surface area of outer surface. So we get sqrt(10Pi/Pi) = sqrt(10) = about 3.16 in radius if we exhaust from one side only. This of-course does not take care of the surface area of the exhausting disk gaps which are still too small at 6.32*Pi*10*.1 about 19.87 square inches. To make up for this we need to taper the thickness of the disks so as to increase the gap size between the disk's surfaces as they near the center.

I was thinking of how to solve this problem and may have come up with a solution.

Take the disk pack and rotate it into the third dimension. Instead of a pack of disks, think of a series of cylinders inside each other. Taking a slice from the center out will result in a similar cross section as looking at the edge of a disk pack. You can use bolts and washers similar to how they are used in the disks. This will ensure an output area equal to the input area without needing tapering or losing a large amount of work area.

For injection, use a nozzle (or series of nozzles) around the end of the cylinders angled such that the flow strikes the end of the cylinder similar to the angle the flow strikes the edge of the disks. The flow should have enough tangential momentum to spiral around the length of the pack.

The downside to this is that the cylinders will have progressively less surface area and volume the closer to the center, but the difference should be less than increasing the central hole size.

 

[chayced]

Flat Disks: Now here is the true real killer of Tesla turbine efficiency. Imagine a simple Tesla turbine, a stack of 11 flat disks each 10 inches in diameter with a .1 gap between them and an exhaust hole in the center 1 inch in diameter. The surface area for the driving fluid entering the Tesla turbine is the (Circumference (or Diameter times Pi) * the width of the input area) = (10*Pi)*(10*.1) = 10*Pi square inches or about 31.4 square inches. But the exit hole is only one tenth the size. So any fluid entering a Tesla turbine gets slowed down while inputting energy into the Tesla turbine and at the same time must exit from the smaller area of the exhaust hole. But notice it gets worse, the area of disk gaps doing the exhausting is (1*Pi)*(10*.1) = Pi or about 3.14 square inches but the area of the hole is .5*.5*Pi or about .79 square inches.

Exhaust Hole Size: To solve the two above problems we have to do two things, both affect the efficiency of the turbine. First we have to make the Exhaust hole closer in size to the surface area of outer surface. So we get sqrt(10Pi/Pi) = sqrt(10) = about 3.16 in radius if we exhaust from one side only. This of-course does not take care of the surface area of the exhausting disk gaps which are still too small at 6.32*Pi*10*.1 about 19.87 square inches. To make up for this we need to taper the thickness of the disks so as to increase the gap size between the disk's surfaces as they near the center.

The original article is very flawed in its asumptions. There are no tesla turbines which emit steam around the whole outside circumference. If this asumption were true then impeller driven pumps would not work since they take a liquid from a small opening in the center and then discharge it from a much larger opening at the outside. Steam is injected through nozzles tangentaly to the disks. The size of the exhaust hole could come into play, but comparing the inside surface area to the outside surface area is nonsense. Also tapering the disks is equal nonsense.

 

[dr dodge]

hollow shaft would get the most "linear distance" from the disk OD to the exhaust. Steam is going to be changing volumn as it cools requiring less exhaust dia. I am amazed how many mis-conceptions there are to a tesla turbine on many web sites. I have seen impeller and flat vane turbines called tesla's. Good luck

dr

 

[Wonderous Mtn]

I am building a disc rotor in the next few days. Hope to post some hard data. I have several ideas for improving efficiency, but want to calculate my actual losses before proceeding with a new design.

If anyone has legitimate information that might be useful, or a pet theory they would like tried; I will accomidate all suggestions that aren't too difficult to perform.

I have many resources available, including aluminum casting, vertical mill, lathes, welding equipment, compressors, and several tools for measuring results. If there is someone that can take high quality measurements, I would be glad to bring it anywhere in the continental United States.

Cheers,

Wonderous Mtn

 

[linuxguy]

I feel the most important thing to remember is back around 1906 -1913 when Tesla invented the turbine there were different materials and machining techniques. A lot of people try to start off by modifying his original plans thinking they will get more efficient output from mathematical equations. The reality is that historical documentation shows that it works.

Let's start with his original model and research the material and machining techniques of the time. Then, you have some comparisons to work with. They were also using quaterion mathematics.

I myself have studied this Turbine for years and proceeded to have a machinist to build such a unit. But, to my surprise even though this machinist had the latest equipment and produced massive parts for DOD could not produce this simple model. A lost investment.

We have to think in terms of starting with the original first then do modifications. Set the standard first.

For a turbine which is supposed to be simple to manufacture today's machinist is totally lost.

It is not a question whether the turbine is efficient of the time, but what we can do to produce an original working model.

The Tesla Museum in Belgrade has an original..

linuxguy

 

[NateD2]

Linuxguy,

I highly doubt there is any problem creating the turbine today. All the parts, materials and processes exist to make the unit. The problem I have run into is getting a machinist who is creative enough to make the parts. But if you really know how to model and draw the parts that is a non issue. A modern shop with the right tools can easily make a model to the original specs.Additionally I totally disagree with your assessment on mathematics. Applying the proper mathematics to the Tesla turbine is not hard, but it is not easy either. But it can be done.

Altering the design is not a problem as Tesla's understanding of flow losses I do not think was as good as modern day. Additionally Tesla did not have the ability to simulate flow like can be done today.

On a last note the efficiency of the Tesla turbine has somewhat changed relatively. When Tesla made his turbine it may have been more efficient then competing reaction and impulse turbines, however today's turbines are very efficient compared to then. I've heard numbers in the 95% range. But who knows what Tesla's stuff can actually do no one seems to have released the data to figure out just how efficient it is or is not.

 

[NateD2]

Dr. Dodge,

In conventional turbines the cross sectional area of the fluid path increases as the working fluid is expanding to extract energy. The same is true of the Tesla. The exhaust ports need to be sized correctly for the specific volume of the working fluid in the expanded condition.

 

[linuxguy]

Please see the following:

This article goes into some detailed data related to efficiency and other related material.

Further Investigations into Tesla Turbomachinery Author:Peter Harwood, SID:3046768, Supervisor:Professor Mark Jones November 19, 2008.

http://www.scribd.com/doc/27375893/tesla-pumps-and-compressors

 

[NateD2]

I Skimmed the article. There are a few major problems I can see with it.

The first one is whether or not they use a flow solver that handles viscous effects (most flow solvers that I have seen can not). In fact most solvers that solve the viscus equations have to remove some other flow characteristic in order to be solved.

Simply put a rotational, moving mesh, fully compressible, viscous flow solver is one of the hardest problems in CFD (to my knowledge).

Additionally their results are for a blower not a turbine. There is always a HUGE difference in the two.

The other question I have is about their efficiency rating. Many people have rated efficiency as (enthalpy input - enthalpy output)/enthalpy input.

From my point of view this ONLY gives the thermodynamic efficiency and not the real world efficiency which would be written this way: (enthalpy input - work out) / enthalpy input.


94% efficiency seems good but unrealistic.


I'll have to spend some time digesting the paper, but sooner or later I will have all my test results on hand to create my own curves.

 

[linuxguy]

Please be sure to check out Tesla's Valvular Conduit as a source of input.

Hope this helps,

Linxguy

 

[andrewr]

Tesla's design requires quite a bit of rigidity on the blades. A neighbor of mine patented a design for a simple blade using dimples in sheet stainless steel to solve that problem. I don't have the patent number, but look up Guy Letourneau (sp?) on the patent database.

He was using steam injection on his particular run, and I recall watching the video and discussing the water injection after the heat source to produce the steam. I don't think he was trying for maximum efficiency -- but just raw power conversion at reasonable conversion rates. he did succeed at that -- there was significant horsepower output from a small coffee can sized turbine -- enough to burn the brakes out on a differential set of tires he was using as a load.

If the efficiency isn't mentioned in his patent -- I could ask him about what he got (if he knows), and I would expect that to be a lower value turbine. He definitely knows the amount of power it generated, and he might know how much fuel he burned in the test.

 

[fsjer]

Is it possible to run a tesla turbine by using household water suppy with only 10.3m residual head?

 

[potatoecannon]

Hi NateD,
How is your data collection rig comming? I'm looking forward to seeing the curves from your first round of tests.

 

[Sixdeuce062]

new to the forum thing I have always been interested in many of teslas works ( he's sort of my hero ) and thanks to A very great teacher i had in high school introduced me to the tesla coil ( which is awsome in its own right) but what caught my eye was the tesla turbine, black box ( most likely more myth than fact ) and wireless transmission. This is the about the turbine though i have built two prototypes and currently working on a finished model. So i don't sound like a scam artist trying to tell the world that i have made the magic tubine than makes more power than it uses ( about as true as unicorns fairies) I have Have five basic componets: the block ( housing, shell, etc), the rotor, ( outlets or exhust holes steped down, stainles steel 18Guage 6in dia, that's right no metric here becuse this is america), inlet nozzles two types, one of tesla spec and one of my own( brass), compressor with resivour ( air @ 50-100 psi ), and an oil cirulating system ( due to high speed went with an open bearing style in gear oil to keep heat down ). Now that u have a rough idea what i am working with I am about 2 months into build ( you don't rush any thing that spins over 1000 rpm or cost a lot in materials) and i am looking for some tips on test equipment for determining the test Eff. rating. i already have heat temp., tourqe, and pressure. looking for rpm, and yes i know there's formulas out there and some test equipment but we are talking thousands of rpm here and i am old school i like to have tools in my hands more than the pen. I am also going to have a generator hooked up to it ( geared down of course) to see if if can push it and test turbine under load would be very interested in what u got nate

 

[nunyabuizness]

Any updates on the thread? I made an account here just to get involved as I, too, want to build a TT when I get the chance. Physics undergrad and willing to help anyone still working on it.

 

[Kjunjoe]

Everyone I asking for Data on Efficiencies, Over and OVER. The funny thing is NO ONE HAS ANY. This is not a mistake or overlooking of data. This is because the TT is dynamic in it construction and efficiencies. The truth be told. Everyone is wrong in some way when it comes to TT, Maybe even Tesla himself. Let’s get into the nuts and bolts as well as why there is no data:
1) The size of the housing? Let’s assume you are using a standard size disc of 4” in length. Should it be 4 ¼” or 4 ½” or 5”. You get the picture. The answer is Boundary Layer (BL). The meniscus of the fluid x2 would me a better definition. One of the key things that Tesla said outside the Patent was that the fluid should have as little resistance as possible in the engine, in other words, flow of motion.
2) The size of the Disc? Any size would be the answer but it changes the speed and Torque of the turbine.
3) The size of the discharge holes? Large enough not to offer resistance of the outgoing fluid.
4) Bearings? As little resistance as possible but consistency is important.
5) Pressure applied? ??
Unlike any turbine or engine you have dealt with in the past, This turbine changes due to all of the above reasons. When one changes, they all change. To many times I have seen people on the internet putting disc so close as to cause restriction……Wrong. They Increase the input without accounting for the output…….Wrong. Due to all the variables and ways to damage the outcome, even scientist and universities can’t agree on the basic numbers. 25%-60% are the ranges that we have gotten but Possible to 95% is always the end of the statements. Why? Unlike a normal engine where you plug in numbers into a model and get a result, you have to first build it changing only one variable at a time then go back to each previous variable and modify them till you get the proper efficiency. The Truth is that there are efficiencies out there. We just need to take the time to find the truth.

 

[kland1]

Is it possible to run a tesla turbine by using household water suppy with only 10.3m residual head?

Hello fsjer,

Yes it is possible to build a Tesla turbine with the water pressure you have, the key will be making a nozzle that will take the pressure much higher and faster then what you have now.

The inlet nozzle is considered the most important part of a working Tesla turbine so designing that part will be the most work. Take a look at this article http://www.seabirdadventure.com/tes...chives/three-keys-to-tesla-turbine-efficiency where the discussion on the three parts; the inlet nozzle, disk geometry and outlet nozzle are shown with design ideas.

You will be facing another problem which I faced when building my first Tesla turbine with a power steam washer. In order for the turbine to work the inlet nozzle gas or in your case water stream will be at the outer diameter of the disk pack and then spiral towards the center exhaust port of the disk pack. And the disk pack will need to be enclosed so that the stream does not simple escape, this however creates an area where the water will gather in the bottom and create a pool that the disk pack will have to spin in causing water drag against the spin that you are creating.

In my case the steam worked at the beginning but because the heat was not enough the steam condensed in the disk chamber and created the same pool which ended up slowing the pack down to almost a stop.

So it will be fun to try but you'll have to get to something more like high heat steam or air pressure to do the job.

Cheers, Kris

 

[kland1]

new to the forum and i am looking for some tips on test equipment for determining the test Eff. rating.

Hello Sixdeuce062,

Sounds like you have a great setup already and I would love to hear more on your project. I just finished testing a third generation Tesla turbine prototype end of last year and am now also in the efficiency testing phase to squeak out as much performance as I possibly can. To see the last test being run in March of 2010 take a look at http://www.seabirdadventure.com/tesla-turbine/tesla-turbines-are-very-different this has a picture of the turbine not running and also one of it running.

In this case I've got a Tesla turbine running on 90% H2O2 through a catalytic engine at roughly 165HP which is gear reduced using a S-10 transmission of 4.03 to 1 which is in turn running a 125KW generator.

Now to answer your question, if you tie your Tesla turbine to a generator and then in turn tie the generator to an electric load you can measure the wattage costumed by the load to calculate energy produced. This is also a great writeup on efficiency calculations http://www.ehow.com/how_6539894_calculate-steam-turbine-efficiencies.html

Cheers Kris

 

[kland1]

Any updates on the thread? I made an account here just to get involved as I, too, want to build a TT when I get the chance. Physics undergrad and willing to help anyone still working on it.

Hello nunyabuizness,

I just started posting here and would love to see these discussions start up again. I have been working on Tesla Turbines for a few years now and am slowly making progress on getting the system to be more and more efficient. I think this is a great article to read up on to get started http://www.seabirdadventure.com/tesla-turbine/tesla-turbine-archives/tesla-turbine as it discusses the three main types of turbines manufactured and the basics of how they work.

The work involved in the seal between the moving plates and the housing is my current problem so I am planning on making a set of interlocking graphite rings that will mount on the inner housing plates and the rotating disk plates to keep some of the high pressure gases from escaping around the disk plates directly to the exhaust manifold.

Cheers, Kris

 

[Vespa71]

Tesla did more than just invent stuff. He told riddles. Some of them might not have good answers. A machine that relies on friction, or adhesiveness, to transfer power, can not reach good efficiency. Sorry, believers. The main feature of the Tesla-turbine design, is its simplicity, which discourage scientific research, and attracts amateurs.

I will not loose sleep waiting for the effective Tesla-turbine, but the problemsolver in me will suggest a couple of principles to be sought to reduce the flaws of design:

-2-step, allows the exhaust to expand towards sufficient low pressure. One of Tesla's own prototypes featured such design. A single turbine suffer exhaust back-pressure, since the port(s) must have a sufficient area to accommodate high volume and low speed of the expanded gas, and therefore ports can not be at absolute center. Two machines in series increace resistance but relieves the problem described.

-Low speed-differance between incoming gas and wheel. Allthough adhesiveness is necessary to achieve power, this friction also looses energy. Lessing the braking of gas speed in the peripheral contact, reduce frictional loss. The Tesla-induction-generator (similar design-principle) has a moderate possible drag, an so should the turbine, to achieve max efficiency. Low speed difference and small distance between discs (High adhesiveness) would impersonate the generators action.


The Pelton Hydro-power-turbine works effectively at 2-1 speed, because the water is not braked, but turned 180 deg. at paddle and therefore reach 0-speed relative to surroundings. That is allmost without friction. Why would friction be as effective?

Sorry if I cursed in somebody elses church. Amen.

 

[TMT]

I'm interested with TT can be utilized as a part of HE to EE conversion. Since it is simple & cost effective relative to other turbines. But efficiency is my primary concern also.

E.Enginer computer control graduated and worked very long time on programming.

As a part of programming analysis & break down the problem to objects and build solver as assembly of those objects gave me skill to look problems in different view angle.

Let me explain how TT work as my approach.

The main component of system is working fluid. It is actually build from molecules. Therefore we have to focus on molecules behaviors in different stages.

An escape hole (exit) on presurized container let some molecule found no other molecule to collide between openings and leave container. (Now molecule has a velocity vector.)

Molecule meets with disk molecules. Some disk molecules elevated in front of our traveling molecule and cause a collision (this is friction effect). Others are ordered parallel to velocity vector of traveling molecule. Interaction of moving molecule and disk surface molecule cause an attraction force. Force vector has same direction on disk molecule and reverse direction on moving molecules with velocity vector (drag, boundary layer effect).

Hence disk surface is very smooth, disk molecules very mostly subject for boundary layer effect.

In any case moving molecule very close to disk surface will slow down (loosing its KE) while disk molecule forced in velocity vector direction (causing disk spin). But conservation of momentum principle let moving molecules keeping velocity vector direction even reduce in magnitute. Now moving molecule is not on same radius circle which is tangential before and be on circle greater radius one and not tangential with it. But we know that molecule is going on smaller radius and eventually leaving from central exit. Some documents explain this case as losing energy of moving molecule. This is not convincing me and I try to explain as different hypothesis.

Assume some moving molecule is out of disk perimeters. They will go ahead and collide to TT case and build a pressure rise in that area. This pressure create a radial force toward center direction is making change in moving molecule direction.

As I observed from some experiment video recording, TT without closing case either not starts to spin or very low rpm. This observation is somehow support my hypothesis.

Conclusion: The pressure build on area between TT case and disks pack may be one of the key parameter on efficiency.

NateD

if you need any C coding support I may help.

 

[TMT]

My second hypothesis is about a vortex created in case that leading spin of disk pack.

If we remove the disk pack we (most probably) will observe fluid first follow case chamber wall and complete turn. Now we have a spinning mass that have moment of inertia. The kinetic energy of fluid is converted to inertial energy.

Pumping more fluid will help pressure build up near to chamber wall. This pressure will yield a radial force on spinning mass. Spinning mass contract towards center, than leave case from central exit.

If so, we are actually deal with a vortex structure in case and a standing disk pack at just the middle of this vortex.

Eventually, The rotating fluid Inertia is transferred to disks by means of boundary effect.

May be a larger gap between disk pack and enclosing case inner chamber will lead a larger Inertia build up. (than we may focus on inertia transfer via B.L. effect). In that case we will not set flow direction tangential to disk, but set direction tangent to case inner chamber.

If this hypothesis is valid, thickness of disks is irrelevant.

Otherwise, we can not utilize KE of the portion of fluid that hitting to cylinder formed due to thickness. In that case I suggest a very sharp or airfoil formed edge for disks may contribute to efficiency.

 

[sagsec]

Many people are modeling TTs and steam engines to run by compressed air, not with the real steam. and trying to show they are working with the steam in the same manner. But it is having a considerable difference when it is running with the steam, rather than Com. Air. When I failed to find the real figures about the TT running with steam I myself designed and fabricated a TT about 7 years ago.
Here I used Circular saw blades ( I used 9 of 14" saw blades.) after removing all the tooth in a lathe machine. To maintain the gap I used thin Aluminum washers between blades. The gap I kept was 0.4 mm, as I could remember. It works fine, had a considerable torque even in the beginning. I never expected a that much of a torque ( I am a mechanical Engineer), but it was. It could start under generator load and worked fine and developed about 3 kW. Problem to run continuously was the temperature of bearings. Those days I did not have a good solution for it but now I have. Now I have a even better design and hope to start again. I will put all the pictures taken when I was building the turbine showing internal structure and all.

 

[bayleebug13]

i just want to know if its renewable, nonrenewable or inexhaustible

 

[billy_joule]

i just want to know if its renewable, nonrenewable or inexhaustible

A TT is not a source of energy, it is a method of conversion. They could be used with any heat source ie coal, gas,solar, nuclear, geothermal etc
http://en.wikipedia.org/wiki/Thermal_power_station
Conservation of energy tells us no energy source is inexhaustible.

 

[Gary K]

Many people are modeling TTs and steam engines to run by compressed air, not with the real steam. and trying to show they are working with the steam in the same manner. But it is having a considerable difference when it is running with the steam, rather than Com. Air. When I failed to find the real figures about the TT running with steam I myself designed and fabricated a TT about 7 years ago.
Here I used Circular saw blades ( I used 9 of 14" saw blades.) after removing all the tooth in a lathe machine. To maintain the gap I used thin Aluminum washers between blades. The gap I kept was 0.4 mm, as I could remember. It works fine, had a considerable torque even in the beginning. I never expected a that much of a torque ( I am a mechanical Engineer), but it was. It could start under generator load and worked fine and developed about 3 kW. Problem to run continuously was the temperature of bearings. Those days I did not have a good solution for it but now I have. Now I have a even better design and hope to start again. I will put all the pictures taken when I was building the turbine showing internal structure and all.

I'm curious what you determined to be a solution for the bearing temperature problem. Thanks!