DIY Tesla Impulse Turbine

[NapierDeltic]

What about turbine's rpm?

 

[Steamchick]

Turbine "speed" (that is a misnomer): 3 parts of dead crawl... I guess around 300~400rpm? It speeded-up a tad when the gears and shafts were given a drop of oil.
But I was amazed that it worked so well and so slowly! (No power of course). Ran for at least 30 mins while I familiarised myself with pumping water (almost constant!) versus using steam... It was throttled to a level the burner could maintain).
Watch the videos and try not to fall asleep? - Watch the drips from the drain hole to understand the amount of "shed" condensation from the discs...

K2:

 

[Steamchick]

Perhaps I'll remove the secondary gear next time?
K2

 

[NapierDeltic]

Hi K2. Can you please specify the measuring units and the reading of manometer on attached video? I am used to think in bars and image is not focussed.
With all consideration and ignoring condensation, it is obvious that the turbine is (almost) completely out of its "comfort zone". The fact that you have achieved 20 000rpm with 7bar/100psi compressed air is more than explicite.
I have re-read some posts and attachments of this thread. It is not obviously for me that 100 000rpm should be the target. I would guess it is rather an escape goat. Something like: see, you have not obtained 100 000rpm we claim - it is obviously you have failed! Your source mentions 2000rpm for 50 cm OD rotor which is far from 100 000. Yes, I'm aware about downscaling -also mentioned! And I have in mind your underline of efficiency increase from 70 000 to 100 000 rpm.
As I see, you have at least a 20cm disk OD turbine. You could have a something below 40% efficiency when working with compressed air (according some of your sources). It is not much but not too low. Gasoline engines have lower than that (overall) and they don't seem to bother.
From what I have read, a 10 times drop of pressure in turbine is considered (yes, CO2 cycle), and for a reasonable power output it should be at a higher max cycle pressure. Conclusion -input parameters in turbine should be raised.
Please don't take this as a negative critic! I am entirely on your side and hooked on your quest.

P.S. I have had numberless technical (strictly technical) failures both in my personal interests and in my job, so I don't speak from an insensible (presumed) point of view! And I don't pretend I see the right way...

 

[Steamchick]

Hi Deltic. No problems. Pressure gauge FED 100psi. Red line 30psi. I am delighted it runs! The guy that made it died from cancer before runningnit, so I only know rumours of what he built, without dismantling it to see. Rotor is about 65mm diameter. Without getting too technical. Bearing are second-hand so turbine gets noisy at speed on compressed air. So it needs a service. But should I buy replacements steel or high-speed ceramics? No idea till I get the boiler at its best and see what can be done with the turbine. Final burner design may be about 4 kW instead of the temporary one at 3kW. But the boiler design cannot be certified over 30psi anyway. I may be able to add a decent superheater , but that would be next winter. Next show day end of month.
Ok?
K2

 

[Nerd1000]

Hi Deltic. No problems. Pressure gauge FED 100psi. Red line 30psi. I am delighted it runs! The guy that made it died from cancer before runningnit, so I only know rumours of what he built, without dismantling it to see. Rotor is about 65mm diameter. Without getting too technical. Bearing are second-hand so turbine gets noisy at speed on compressed air. So it needs a service. But should I buy replacements steel or high-speed ceramics? No idea till I get the boiler at its best and see what can be done with the turbine. Final burner design may be about 4 kW instead of the temporary one at 3kW. But the boiler design cannot be certified over 30psi anyway. I may be able to add a decent superheater , but that would be next winter. Next show day end of month.
Ok?
K2

My spreadsheet says that for your design parameters (pressure, rotor diameter, etc) and pure impulse operation your optimal speed will be about 70,000 rpm.

Also in the absence of a condenser or feed water preheating your 3 kW burner should be able to provide enough steam for about 25W at the turbine shaft, assuming a turbine efficiency of 75%. Preheating your feedwater to 100 degrees C using the exhaust steam would raise that to 30W, for an overall thermal efficiency of 1%.

It's the latent heat of vaporisation that is the issue. Takes a lot of energy to turn water to steam, and you don't get it back without more elaborate economising methods I suppose.

 

[NapierDeltic]

Thank you for the answer and Good Luck, K2!
Considering also what Nerd1000 above said, ceramic bearings are the answer. A lot of friction in standard steel, greased, encapsulated bearings.

 

[Toymaker]

<snip> But should I buy replacements steel or high-speed ceramics? <snip>
K2

I've had pretty good luck buying and using cheap, all-ceramic bearings, on eBay; the size I often use is 608 (8 ID x 22 OD x 7 W mm). This size is often advertised as an upgrade replacement for skate boards, which makes them easy to find and cheap. Most of these bearings I've received have been a bit loose, but the looseness goes away when lightly loaded axially. Tesla turbines have no axial loading, so most any bearing will work.

BTW, thanks for all the videos !! Nice to see a running Tesla turbine

The drain hole you installed seems to work well for your turbine; I'm now wondering if I'll encounter the same issue with condensate, or will the six nozzle arrays, equally spaced around the circumference, be sufficient to "blast" the condensate into small enough droplets such that they are more easily pushed out the axial exhaust? Time & testing will eventually provide the answer, but for now, my focus is on building my tiny 9-piston feed pump.

 

[Nerd1000]

I've had pretty good luck buying and using cheap, all-ceramic bearings, on eBay; the size I often use is 608 (8 ID x 22 OD x 7 W mm). This size is often advertised as an upgrade replacement for skate boards, which makes them easy to find and cheap. Most of these bearings I've received have been a bit loose, but the looseness goes away when lightly loaded axially. Tesla turbines have no axial loading, so most any bearing will work.

BTW, thanks for all the videos !! Nice to see a running Tesla turbine

The drain hole you installed seems to work well for your turbine; I'm now wondering if I'll encounter the same issue with condensate, or will the six nozzle arrays, equally spaced around the circumference, be sufficient to "blast" the condensate into small enough droplets such that they are more easily pushed out the axial exhaust? Time & testing will eventually provide the answer, but for now, my focus is on building my tiny 9-piston feed pump.

Apparently full sized turbines have drains, so I'd just add one.

I think you would ideally have enough superheat that the steam does not condense significantly within the turbine, once it is up to temperature. Apparently impact damage from water droplets hitting the blades is a significant concern in full sized plants, so they try not to have their steam condense, often the HP turbine exhaust is sent back to a second superheater for reheating before going to the LP turbine in part to prevent this.

I'd guess that a Tesla turbine is more or less immune to the impact damage issue, but too much condensation is still undesirable regardless.

 

[Steamchick]

Thanks for comments. I am very interested in the Tesla design.
On condensate:
I was very surprised when my 2mm drain hole performed so well! Just enough internal pressure to overcome the surface tension means that the steam really is spiralling in to the centre vent holes! And although the video doesn't really show it there are some good clouds of water vapour from the exhaust. Carefully I felt it was HOT (Hands don't like anything over 60deg.C!) - 6 inches away from the exhaust. I didn't feel any closer! But not really blowing out with any pressure so perhaps it really was using a lot of Steam energy to drive the "noisy" gear chain and rumbling bearings.
On pre-heating water: I plan a pre-heater in the smoke-box end of the boiler and/or flue. 120C gases passing up there can be cooled by pump water through a coil, before the water feeds the boiler.
On superheating: There is a single pass tube through the fire, then up through the boiler to conveniently place the steam valve on top of the boiler. - REALLY just a drier! The superheater passing back through the boiler water and out will probably exit little above boiler steam temperature. But if this is fed a few times back through the fire then a lot of heat/temperature will be gained and improve efficiency. I have a vertical boiler with a superheating coil in the firebox and the steam /pipe is hot enough it chars cotton lagging, and tarnishes the copper pipe instantly with "colours" like tempering steel as it oxidises the copper tube... I'd like to try the turbine with this steam soon? - Watch this space!
On Bearings: THANKS Toymaker for the advice! - I think those bearings may fit, but I'll have to strip the turbine first to find out what will suit. n water droplets damaging turbine blades: Yes, a real problem for High pressure high superheat Parsons turbines as they carry a lot of momentum, but then destroy the blade/cavity shape as soon as the droplets get between the blades. - as well as mechanical damage of the blades. Gas turbine engines on commercial aircraft do not like rain either! - But are made to cope with Seagull ingestion (which is quite common) - My brother worked with Rolls Royce on the design of the first test rigs for the first RB211 engines. (mid-1970s). They ran-up an engine on the rig, then threw frozen chickens up into the intake draught to be ingested, after which the engine was stripped to study the blade damage. They also had a fire hose they sprayed at the air intake to simulate a heavy Monsoon/hurricane thunder shower. I don't know what they did to simulate large hailstones... but I think compressor icing is a serious problem for those engines... Freezing fog can develop an ice coating on the "vacuum" side of compressor blades changing the aerodynamics of blades? - But our Tesla turbines won't suffer from frozen Chickens and freezing fog... I hope? (My Surname is Chicken! - hence a pun.).
On max rpm.. Thanks Nerd for the 70,000rpm limit. Better than my guess! - I had figured "sonic steam" may relate to a tangential speed around 85,000rpm.... but it was not an accurate "sonic speed" for wet steam at 30psi! - Hence the "Max. 100,000rpm" guess. (I am sure it will never get there without huge modification and dynamic balancing).
I will have to build a huge boiler - flash boiler maybe? - to feed a Tesla turbine at that speed. Until I strip the turbine and see what nozzles are fitted I cannot judge "how-much?" steam it can take, but the current teeny-weeny steam pipe won't give it much! A part of the reason I am surprised it runs on 15pi wet steam at all! I also suspect it is a LOT less efficient than the "25W" shaft-horses (mice?) forecast you predict. I suspect water (condensate) drag on the outer rim of the turbine, heat loss from the un-lagged turbine casing and end plates, bearing friction and gear friction are consuming >20 of those Watts?
My ideas to improve the turbine:
A bigger feed pipe and valve.
SHORTER fed pipe to reduce those losses.
Lag the outer casing.
Improve the drain, as the drain hole isn't right at the edge of the cavity. Perhaps a drain hole on the other side as well will help, as the turbine annular pressure (just outside the disc outer diameter) is probably a function of steam speed after the nozzles, so at very high steam velocity may be negative pressure in the middle of the disc-pack, but just above atmospheric at the outsides of the outer discs, as otherwise it would not drip, but suck -in air. - Nerd, what is your thinking (relating to internal steam velocity and pressure) for the best location for the condensate drain?
Finally: An apology to Toymaker on hijacking the thread, but perhaps this experience can help you plan how to optimise your set-up?
Thanks for the space in your thread.
K2

 

[Steamchick]

Just a couple of questions re: to help my understanding of what is going on inside the turbine:
N1000 - I think your spreadsheet may be interesting - if you can share a screenshot - or something?
1: Can it manage Toymakers combined Impulse and Tesla design?
2: Latent heat during conversion of pressure steam to high velocity steam jet, and loss of heat due to momentum transfer to rotors: I tried to imaging what is happening "inside" the steam.. both thermodynamically, and where the energy goes... And I don't understand what is going on: I.E. The steam with Temperature and pressure and no velocity, has "enthalpy" (potential energy within the gas) - (PLEASE correct me where I am wrong. I failed this at university!):
So when it passes through the nozzle, the enthalpy converts some energy to kinetic energy of the jet stream... I.E. pressure and temperature drop.
As the jet stream slows - because some kinetic energy transfers to the rotor - the steam can start to condense to water vapour, releasing the latent heat of vaporisation in the process (?): the mass flow stays constant , except for maybe in the boundary layer water vapor that gets flung off radially when droplets adhere to discs? - So that is lost kinetic energy? But is the latent heat transferring energy from the steam to kinetic energy because that kinetic energy is being extracted by momentum exchange to the inner zone of the rotor - to maintain the wet steam gas dynamics in the turbine? - IF this is the case, then I can see the latent heat of condensation being partly used by momentum exchange to discs, and partly lost as condensate is shed from the discs.

Can the spreadsheet add anything to this hypothesis? - Or is it all just junk in my head..? (there is a lot in there!).
That s why I see superheating as an efficiency improver - there is less condensate lost through being flung off the discs. But then is the latent heat being lost?
I can see how a condenser creating vacuum helps by increasing the pressure difference across the turbine - hence more kinetic energy can be extracted - and, if a condenser is fitted, the condensate drains from the rotor chamber should be connected to the condenser pressure chamber.
A condenser also enables hot water to be re-cycled into the boiler.... - which, I guess, is Toymaker's plan anyway.
Thanks,
K2

 

[Nerd1000]

Just a couple of questions re: to help my understanding of what is going on inside the turbine:
N1000 - I think your spreadsheet may be interesting - if you can share a screenshot - or something?
1: Can it manage Toymakers combined Impulse and Tesla design?
2: Latent heat during conversion of pressure steam to high velocity steam jet, and loss of heat due to momentum transfer to rotors: I tried to imaging what is happening "inside" the steam.. both thermodynamically, and where the energy goes... And I don't understand what is going on: I.E. The steam with Temperature and pressure and no velocity, has "enthalpy" (potential energy within the gas) - (PLEASE correct me where I am wrong. I failed this at university!):
So when it passes through the nozzle, the enthalpy converts some energy to kinetic energy of the jet stream... I.E. pressure and temperature drop.
As the jet stream slows - because some kinetic energy transfers to the rotor - the steam can start to condense to water vapour, releasing the latent heat of vaporisation in the process (?): the mass flow stays constant , except for maybe in the boundary layer water vapor that gets flung off radially when droplets adhere to discs? - So that is lost kinetic energy? But is the latent heat transferring energy from the steam to kinetic energy because that kinetic energy is being extracted by momentum exchange to the inner zone of the rotor - to maintain the wet steam gas dynamics in the turbine? - IF this is the case, then I can see the latent heat of condensation being partly used by momentum exchange to discs, and partly lost as condensate is shed from the discs.

Can the spreadsheet add anything to this hypothesis? - Or is it all just junk in my head..? (there is a lot in there!).
That s why I see superheating as an efficiency improver - there is less condensate lost through being flung off the discs. But then is the latent heat being lost?
I can see how a condenser creating vacuum helps by increasing the pressure difference across the turbine - hence more kinetic energy can be extracted - and, if a condenser is fitted, the condensate drains from the rotor chamber should be connected to the condenser pressure chamber.
A condenser also enables hot water to be re-cycled into the boiler.... - which, I guess, is Toymaker's plan anyway.
Thanks,
K2

My spreadsheet assumes that the steam doesn't condense, it just expands isentropically through the nozzle, staying at the saturation point. Kind of an idealised scenario, but not far off the mark practically because turbines usually cannot tolerate too much condensation without damage. It also assumes that the turbine is adiabatic (no heat losses to the environment) which is approximately true with appropriate lagging.

The turbine is assumed to be an axial flow pure impulse machine. A stage without velocity compounding runs at 1/2 the steam jet velocity, so I just doubled the number to get an approximate value for a Tesla turbine.

Superheating doesn't just prevent condensation, it adds enthalpy to the steam which can be used in the turbine. An easy way to appreciate this is to consider what happens if we make the pressure and mass flow constant and add a superheater: the superheated steam has lower density at a given pressure, so to keep a constant mass flow rate it must exit the nozzles at higher velocity resulting in more work done on the turbine wheel. In fact I think superheating is critical to getting decent efficiency from a steam turbine, with saturated steam most of the enthalpy is in the latent heat of vaporisation, which is hard to use for anything. The lower your boiler temperature the worse it gets, hence the rather dismal overall efficiency I calculated.

I haven't accounted for superheating in the spreadsheet because it involves pulling data from a different steam table for each pressure, and I haven't figured out how to do that yet.

 

[Toymaker]

There's another rather important principle governing fluid flow within a Tesla turbine which has yet to be discussed, that being what causes the fluid (steam, air, any gas) to flow, without turbulence, in an inward spiral pattern.
Here's how I see it: Once the steam leaves the nozzle it's forced into a circular path due to the metal walls of the casing. As the steam completes the first circle, it encounters more steam from the nozzle which is at higher pressure and velocity, which pushes the second circle steam inward. This steam stream would love go straight, but it can't because the first circle steam continues to push it into an inward spiral. Turbulence cannot occur on either side of the circle due to the close proximity of the disc walls, and turbulence cannot occur upwards because the second circle of steam is at a lower pressure and velocity than the first circle steam, and is prevented from flowing in an outward direction.
Even though the first circle steam is at higher pressure and velocity, it cannot produce turbulence in an inward direction because of Newton's first law of motion. The spiraling steam produces a continuous pressure drop from the outer circumference of the disks, radially, until it reaches the axial exits.

In other words, the centrifugal forces of the spiraling steam, which really, really wants to flow in a straight line, act as soft barriers which control the direction and expansion angle and rate of the steam stream.

 

[Steamchick]

Good stuff! What does it predict for Toymaker's components impulse and Tesla arrangement? I have tried to thinNk-it-out thus.
The impulse bit runs at half velocity of the jet. The remaining steam running at half original velocity then runs through the Tesla discs turning at the speed dictated by the impulse vanes, slowing the steam stream and extracting more energy, thus speeding-up the turbine....
So the impulse vanes then take less energy from the initial steam, the Tesla discs take more from that as a result, and the turbine speeds-up to find an equilibrium speed. Or does it? Is the Tesla part more efficient at this lower speed than the reaction bars? If less efficient, does the turbine slow down?. Can your model predict a speed and efficiency of this turbine? How does that compare to the two individual parts of the turbine?
I am curious, that's all.
K2

 

[Steamchick]

Hi Toymaker, I think you are right.
What I read (years ago, so probably imprecise now!) was that the gas follows Newton's laws... without considering pressure, etc.
Fluid flow tangential to the slots (the discs are just there to make the slots as the fluid sees it) tries to go in a straight line, but "feeling" the push of the fluid following due to "pressure behind is greater than pressure in front" - I.E. an accelerating force, (which is not strictly correct, as fluid cannot control compression as a vector, then it is purely momentum keeping it going in a straight line) so it is trying to accelerate towards the exit - which is at right angles to the way it is going. So it starts to form a curved path. IF the pressure of the outer part is higher than the pressure at the exhaust in the middle of the turbine. (But the high velocity of the fluid at an outer circle is a lower pressure than the gas at the same radial speed but at an inner radius? - or not?), The analogy was planets orbiting the sun, etc.... which is Totally different, as they react to a real force of gravity. Once the curve has formed at the start of introduction of the fluid jet-stream, then drag on the discs starts to rotate the discs and that drags the curve into a helix.... And I saw a utube video of a lab demonstration to show the helix form from a curve to "prove" the idea...
In my head, I can't figure out how the "Low-pressure - high velocity" jet stream after the fluid has been accelerated in the nozzle can see the "Low pressure zone" of the exhaust near the disc centre. Perhaps it takes a very small time to fill the whole turbine chamber and develop the flow into a bent stream to make it work? -
Or is it a dynamic reaction from the movement of the discs - boundary layer - extracting some energy so the jet stream is slower that causes it to bend inwards? - The boundary layer must move in a curve as it is attached to both the disc and the jet-stream, so it drags the jet stream in a curved path until it takes so much kinetic energy that as it slows it has to find a slower disc to adhere to... which is at a smaller radius..
I think I may have seen the whole process demonstrated with a smoke jet on a disc without an outer casing - hence removing the effect of the casing "pushing" the fluid inwards... to a central exhaust? Initially the smoke went straight on past the disc, until the disc speed began to build, when it curled inwards to the exit, before the helix was formed, but as the disc accelerated it formed the helix. More study required I think?

The Tesla turbine works for all fluids, liquids and gases, and mixtures of fluid and "particles" - which is why the pumps are used for things like blood, mixed (crude? Blended?) oils, sewage (?), etc. as the working fluid can carry "less fluid" stuff (passenger material?) in it. - With steam in a turbine, the steam carries condensate in it as passenger material. But in the case of a steam turbine, the passenger material has done 2 things - inter-related. Some steam condenses and in doing so gives up latent heat to the surrounding steam to enable it to carry-on giving energy to the discs. But in condensing, the steam volume is reducing considerably, thus creating a pressure drop - added to the pressure drop developed as the enthalpy of the steam is used. High velocity steam against the disc uses the boundary layer to give-up the kinetic energy (Boundary layer frictional force?) but condensate droplets attaching themselves to discs give-up their kinetic energy by momentum exchange - I think?
I try and resolve things mathematically - e.g. Steam (Mass Mi) at velocity v1 develops a boundary layer frictional force F that causes the disc Mass Md and local velocity vl1 to accelerate. Plus Droplets of condensate (mass Mc) at v1 in the steam transfer their mass to the disc at a lower velocity as they attach themselves to the disc at local velocity vl1 ad the law of conservation of momentum then rules the disc must accelerate to accommodate the lost velocity of the mass of water that has just slowed from gas velocity to disc velocity. - I think? Anyway, the fluid stream slows as it gives up kinetic energy to the disc.
Confused,? I am...
K2

 

[The Ignoble Troll]

There's another rather important principle governing fluid flow within a Tesla turbine which has yet to be discussed, that being what causes the fluid (steam, air, any gas) to flow, without turbulence, in an inward spiral pattern.
Here's how I see it: Once the steam leaves the nozzle it's forced into a circular path due to the metal walls of the casing. As the steam completes the first circle, it encounters more steam from the nozzle which is at higher pressure and velocity, which pushes the second circle steam inward. This steam stream would love go straight, but it can't because the first circle steam continues to push it into an inward spiral. Turbulence cannot occur on either side of the circle due to the close proximity of the disc walls, and turbulence cannot occur upwards because the second circle of steam is at a lower pressure and velocity than the first circle steam, and is prevented from flowing in an outward direction.
Even though the first circle steam is at higher pressure and velocity, it cannot produce turbulence in an inward direction because of Newton's first law of motion. The spiraling steam produces a continuous pressure drop from the outer circumference of the disks, radially, until it reaches the axial exits.

In other words, the centrifugal forces of the spiraling steam, which really, really wants to flow in a straight line, act as soft barriers which control the direction and expansion angle and rate of the steam stream.

I haven't found any papers on it, yet, but this is what prompted me to wonder about discs that taper inwards, so that the resulting drop in Enthalpy might convert more steam to water and by having the spacing gradually change from idealy space for steam to ideally spaced for water, may increase efficiency. After all the turbines work with water at a different blade spacing.

This might act as an all in one turbine/economizer

 

[Steamchick]

Clever idea Troll, but I can see 2 issues to study.
1. Steam volume is a couple of hundred times water volume, so water gap is 1/200th steam gap of 1mm for the job we are looking at.
2. Most of the steam and water exits the exhaust, but some wat er leaves from the water drain at the perimeter of the casing.
Hard to know what to do ...
K2

 

[NapierDeltic]

I have seen the mentioned efficiency increase jump in this video:

Around min. 3.30 . It really is amazing!

 

[Steamchick]

Heroes don't come in shining armour, but are usually modestly dressed... They are not about show, just results.
Personally, I think his work on Alternating current is the most important engineering development since Faraday "tamed" electricity, Maxwell wrote the Maths, and without which we would not appreciate how to "make the lights work" and everything up to the gamma radiation of the Nuclear process, X-Ray telescope, MRI machines, X-ray scanners, mobile phone carrier signals, Radar, microwaves, induction heating, induction chargers (for phones, toothbrushes, etc.), PV cells for power generation (Convert the AC power of E-M radiation into DC for our use), etc. - All radiation is "AC". - It moves... (DC is just AC at zero frequency!).
We are moving away from the "heat engine" era (Burning fossil fuel), to the "Electric era" where all our power comes instantly from the big nuclear furnace above our heads. Thanks Nicola! You took Maxwell's maths and made it run our society. Remove "AC" from "today" and we have only 19th century (Coal, steam and DC) technology. Arguably the single most important person of the 20th century. IMHO.
K2

 

[Steamchick]

This is the video that explains it for me.

The centripital force bit explains helix formation.
K2