DIY Tesla Impulse Turbine

[Steamchick]

Electrically, he also designed a generator - that looks a lot like the turbine -
Put a couple of magnets either side of a disc, spin it FAST and the outer edge develops a high DC voltage compared to the shaft. The difficulty is gathering the CURRENT from a brush on the outer rim of the disc.... it overheats with friction. But if you put a pair of these disc generators together, just touching at the rim so they are contra-rotating at the same speed, and have the same direction of magnetic flux, because the contact is stationary it does not overheat from friction, yet the electrical potential of one disc is reversed in the contra-rotating disc, hence the shaft-to-shaft potential is double the across-disc potential. Fancy making one? Maybe I should? - It was designed to use the high speed of his turbine... but materials and bearings failed due to materials and manufacturing accuracy at the time of Tesla.

Then there was a guy called Wurtz - designed an odd pump.

Fun and rubbish off the web? (The rubbish is mine).
K2

 

[peterl95124]

can anyone point out any uses of Tesla Turbine in the real world, a real product, not just a toy or a demonstration model ?

 

[NapierDeltic]

can anyone point out any uses of Tesla Turbine in the real world, a real product, not just a toy or a demonstration model ?

Hi Peter, there might be military applications of Tesla Turbine; off course we won't be aware of them most probably. Off course, it is not main stream; otherwise we would find it somehow.
For civil applications, I'm aware only of prototypes.

 

[peterl95124]

Hi Peter, there might be military applications of Tesla Turbine; off course we won't be aware of them most probably. Off course, it is not main stream; otherwise we would find it somehow.
For civil applications, I'm aware only of prototypes.

OK, so my point is this, the Tesla Turbine isn't used anywhere for anything for a very good reason, its inefficient, and there's no getting around that fact (hence not even the military would use it), so it never ceases to amaze me that people keep on building demonstration models of something that isn't practical, its a jee-wiz parlor trick engine with no applications, and a really lousy way of getting into actual turbines. its not like building a flame-licker engine that has no applications, in that case you've at least accomplished building most of the parts needed for an IC engine, but with the Tesla Turbine you haven't made a single blade, and that's what turbines are all about, gas or fluid flow around blades; aerodynamics, fluid dynamics, and thermodynamics, which are all extremely interesting and fascinating topics, but the Tesla Turbine doesn't get you there. my "rant cast" for the day, sorry, over-and-out...

 

[NapierDeltic]

People are free to believe in what they want. If you feel you don't belong here (in this topic), you shouldn't be here! Over_and_out...

 

[Nerd1000]

OK, so my point is this, the Tesla Turbine isn't used anywhere for anything for a very good reason, its inefficient, and there's no getting around that fact (hence not even the military would use it), so it never ceases to amaze me that people keep on building demonstration models of something that isn't practical, its a jee-wiz parlor trick engine with no applications, and a really lousy way of getting into actual turbines. its not like building a flame-licker engine that has no applications, in that case you've at least accomplished building most of the parts needed for an IC engine, but with the Tesla Turbine you haven't made a single blade, and that's what turbines are all about, gas or fluid flow around blades; aerodynamics, fluid dynamics, and thermodynamics, which are all extremely interesting and fascinating topics, but the Tesla Turbine doesn't get you there. my "rant cast" for the day, sorry, over-and-out...

They aren't used as turbines but do find some applications as pumps, for situations when tolerating entrained solids or minimising shear forces on the fluid are more important than efficiency.

I've been enjoying this thread mostly because it prompted me to start learning the thermodynamics of turbomachines in more depth than I have previously.

 

[peterl95124]

People are free to believe in what they want. If you feel you don't belong here (in this topic), you shouldn't be here! Over_and_out...

my point is that people participating in this thread are trying to "engineer", in the sense that they hope to apply known formulas to the situation, in the case of turbo machinery (that uses blades) this is fully understood and there are thousands of text books on the subject, in the case of the Tesla Turbine there are no books and no formulas hence no way to "engineer" it.

 

[Steamchick]

Peter, We accept your comments as we all have different views on life and the toys we play with. Most of what I see on this and other threads is about people challenging themselves to make something, usually a something that works. Most make small steam or IC engines, as Stirling engines require a bit more skill getting things to work (extremely low friction required?), but very few actually apply those engines to something. A minority make boats, generators and locomotives, and often people who make turbines make aeroplanes so they can fly But I am sure from the loads of models I have seen that most just make them because they can, and they are made to work. You mention "flame lickers" - as a practical engine they lasted just a few decades - using flue gases from furnaces making iron and steel to power compressors for blowing furnaces, etc. But those were superceded by more efficient engines (Today we use electric motors for those blowers/compressors!).
Just because something is obsolete by today's standards has no influence on whether someone chooses to challenge themselves by making a model. And a part of the challenge is understanding the good and bad points, and improving some of the deficiencies in a particular model.
I am glad you can understand and enjoy the aerodynamics of turbine blades. Not something I have ever studied. - But my step-son worked at Rolls Royce making turbine blades. Otherwise they do not interest me, as I have other "Ducks to shoot". It would also be wrong of me to discuss bladed turbines when Toymaker's thread is about his combined Impulse and Tesla turbine, not the aerodynamic bladed Parsons Turbine design. Especially as I know nothing about them!
I introduced my Tesla Turbine comments as Toymaker is making a different design, combining an impulse turbine with a Tesla boundary layer disc turbine. It doesn't matter how impractical some of us think any if this is, we only get involved in the "pseudo-technical" discussion as there is a lot we don't understand and are learning from the discussions. I feel my "Tesla turbine" experience may help Toymaker, but that is for him to decide. When he says "shut up" I shall. He is not making a "Bladed" turbine, and that is his choice. He will take whatever he wants from my input and discard the rest, as he is a clever guy and can sort the odd useful bit from my experiences.
Sorry if I am a bore, just don't read my input if you don't enjoy it. But I get the idea that some find some of my preachings of interest.
Do what makes you happy, as we do. No hard feelings, just alternative ideas...
K2

 

[Steamchick]

Peter, just checked your comment "no formulas hence no way to "engineer" it." - I disagree.
Napier Deltic has done some valid computations I believe. (Earlier posts).
I have read some Technical papers on Boundary layer and drag effects - which can be calculated. (I cannot say I understood or remember the content!). The De Laval nozzle, steam expansion and work calculations are all reasonable well understood and covered by numerous technical papers. And while I like "easy" calculations that I comprehend, I have never studied/understood Mathematics well enough to approach design of anything aerodynamic, like a turbine blade. So I'll leave that to you, with the respect of a lesser engineer for someone cleverer and more experienced.
Have a nice day..
K2

 

[NapierDeltic]

For the sake of truth, I haven't made what could be called deep dive calculations; as I remember Toymaker or Nerd1000 looked through his charted pool of data, diagrams ... and gave some estimations for certain values related to Tesla Turbine's parameters, and there were also the others sharing parts of their maths.
But for me seeing is believing. I have seen that sudden acceleration of turbine from 50.000 to 70.000rpm that makes me believe that when stars are aligned, efficiency of energy transfer from fluid to rotor increases almost with an order of magnitude (almost 10 x ) and for sure I would be interested in the mathematical model that predicts it. It is a rational man's position.
If scientists struggle to obtain today positive output fusion reactor, it doesn't make it a scam.
For the passion of people on this forum, I completely agree with you, K2!

 

[Nerd1000]

Attached is my new and improved steam turbine spreadsheet, now compatible (well, actually only compatible) with superheating. The sheet is prefilled with a solution for a 3 stage Rataux turbine that is matched to Steamchick's boiler, should he fit it with a superheater. Worth messing around with if you're at all interested in how steam turbines actually work.

A few general notes:

The spreadsheet can model a Rataux or Brown-Curtis impulse turbine with up to three pressure stages, it should be easy for people with good spreadsheet skills to alter it to accommodate more pressure stages if desired. The pressure stages can consist of one or two velocity stages.

The parameters you can change are highlighted yellow, don't mess with the other stuff or you'll break the spreadsheet. Solutions will only be approximate because much of the sheet works using lookup tables (steam tables). The results are a bit 'notchy' because of the gaps between data points in the tables. Important output parameters are highlighted in green.

To get a sensible solution, you need to do a few things: first set the boiler temperature, the condenser pressure, the superheater temperature and the rated power. 101 KPa is one atmosphere, and should be used as the 'condenser pressure' for non-condensing designs. Then you need to set your turbine diameter, speed and number of velocity stages. The spreadsheet will automatically calculate a trial number of pressure stages (which is not necessarily an integer) from these parameters, you need to adjust the turbine parameters so that the trial number of stages is close to the number of stages you want. If you want fewer stages: larger diameter, higher RPM, velocity staging or lower temperatures will get you there, vice-versa for increasing the number of pressure stages. Then set the parameter for pressure stages to the number you want.

Now you need to check the outlet pressure for the last pressure stage; ignore ones later than the number you chose in the parameters. E.g. if you set pressure stages to 2, look at the nozzle exit pressure for the stage 2 nozzles. Don't look at the stage 3 nozzles results as they will be nonsense in that case. A valid solution will have nozzle outlet pressure close to (slightly above) the condenser pressure. If the outlet pressure is too high try increasing the temperature of the superheater; if it's too low try the opposite.

While I've got a handle on superheating now, I still don't really know how to deal with partial condensation through the turbine stages. As a result the sheet will give strange and obviously incorrect results (such as 'steam' with a density of 2kg/L) if the steam reaches a condensation point prior to the condenser. I don't consider this a major issue because it should become obvious from playing with the parameters that efficiency is generally optimised when you superheat your steam as much as possible without causing the final stage pressure to go below the condenser pressure. This usually results in superheated steam exiting the turbine at the condenser pressure.

Once everything is set, check the output results! Especially my super snazzy Temperature-Entropy diagram that maps out the working cycle of the engine, and the thermal power and efficiency calculations. If you want to build the turbine you designed, the spreadsheet will also give you the blade angles and nozzle outlet sizes. The sizes should be correct even for supersonic nozzles, though the sheet doesn't know how to calculate the throat size so that's up to you if your steam going faster than Mach 1.

Obviously this isn't directly relevant to tesla turbines. However, you can estimate your tesla turbine's peripheral speed by designing a single stage turbine (one pressure, one velocity) that can handle your steam conditions, then doubling the RPM. It should also be possible to estimate the efficiency, just set the 'leakage + fanning losses' parameter to 0.7 and you'll get a rough estimate of the performance of a tesla turbine on these steam conditions at double the listed RPM...

 

[NapierDeltic]

Thank you very much Nerd1000 for taking time to put everything in a useful parametric format and for sharing it!
I think it will help many in their calculations; either estimative or detailed.

 

[Toymaker]

The main reason I see for why Tesla Turbines (TT) have not been put to work in some commercial way is their high rpm output isn't very compatible with existing devices. So their outputs are coupled through a gear reduction transmissions before extracting power from it; Steamchick's TT is a good example.

However, there's a simple way to generate AC electrical power using the full RPMs of the TT.

The rods in the sketch below are super magnets that are radially polarized, as shown. The right hand sketch shows a single copper wire partially wrapped around the magnet, and is shown only to help explain how the coil is made. The left hand sketch is my rather poor attempt to show many loops of a single copper wire looped around the magnet, following the same geometry as the single wire. Use a little epoxy glue to hold the coil wires in place.

Machine a coupler from a non-magnetic material, aluminum, nylon, brass, etc., to attach the magnet to the end of the TT's output shaft, covering as little of the magnet rod as possible. Getting the balance right will likely be the hardest part.

Spinning the magnet as shown by the arrow in the right had sketch will cause current to flow in the coils. Direction of rotation is not important, can be CW or CCW.

The output frequency of the alternator will be equal to the RPM of the TT. So, if your TT spins at 50,000 rpm, the AC voltage from the alternator will be 50,000 Hz (50 kHz). Connecting the AC output to a rectifier (a single diode will work, but a full wave rectifier is better) will give you DC voltage. Be careful with your alternator; depending on how many loops of wire in your coil, and the RPM, voltages can easily exceed several hundred.

 

[Nerd1000]

The main reason I see for why Tesla Turbines (TT) have not been put to work in some commercial way is their high rpm output isn't very compatible with existing devices. So their outputs are coupled through a gear reduction transmissions before extracting power from it; Steamchick's TT is a good example.

However, there's a simple way to generate AC electrical power using the full RPMs of the TT.

The rods in the sketch below are super magnets that are radially polarized, as shown. The right hand sketch shows a single copper wire partially wrapped around the magnet, and is shown only to help explain how the coil is made. The left hand sketch is my rather poor attempt to show many loops of a single copper wire looped around the magnet, following the same geometry as the single wire. Use a little epoxy glue to hold the coil wires in place.

Machine a coupler from a non-magnetic material, aluminum, nylon, brass, etc., to attach the magnet to the end of the TT's output shaft, covering as little of the magnet rod as possible. Getting the balance right will likely be the hardest part.

Spinning the magnet as shown by the arrow in the right had sketch will cause current to flow in the coils. Direction of rotation is not important, can be CW or CCW.

The output frequency of the alternator will be equal to the RPM of the TT. So, if your TT spins at 50,000 rpm, the AC voltage from the alternator will be 50,000 Hz (50 kHz). Connecting the AC output to a rectifier (a single diode will work, but a full wave rectifier is better) will give you DC voltage. Be careful with your alternator; depending on how many loops of wire in your coil, and the RPM, voltages can easily exceed several hundred.

View attachment 156943

It's more than just the output rpms being too high. Practical tesla turbines usually have isentropic efficiency of about 30%. A large multi-stage reaction turbine can exceed 90% isentropic efficiency, so between the two it is no contest. Then you have the problem that no practical turbine (bladed or otherwise) can handle the full enthalpy drop available from even a rather modest pressure and temperature steam supply in a single stage without spinning at speeds high enough to make the disc explode. So for a practical steam plant you have no choice but to use multiple stages. This state of affairs very much favours axial flow turbines because they can be set up in series on the same shaft with each stage discharging directly to the next, whereas a radial flow design would need ducting taking the steam from the middle of one stage to the outside of the next stage.

All this makes the Tesla turbine more of an interesting curiosity than a practical machine. I like interesting curiosities.

 

[Toymaker]

It's more than just the output rpms being too high. Practical tesla turbines usually have isentropic efficiency of about 30%. A large multi-stage reaction turbine can exceed 90% isentropic efficiency, so between the two it is no contest. Then you have the problem that no practical turbine (bladed or otherwise) can handle the full enthalpy drop available from even a rather modest pressure and temperature steam supply in a single stage without spinning at speeds high enough to make the disc explode. So for a practical steam plant you have no choice but to use multiple stages. This state of affairs very much favours axial flow turbines because they can be set up in series on the same shaft with each stage discharging directly to the next, whereas a radial flow design would need ducting taking the steam from the middle of one stage to the outside of the next stage.

All this makes the Tesla turbine more of an interesting curiosity than a practical machine. I like interesting curiosities.

As has been pointed out, there are no practical TTs available for efficiency tests. They're mostly (perhaps all) one-offs and hobbyists creations.

Comparing power station, or ocean-going ship sized steam turbines to all the desk-top TTs is a bit of apples to oranges comparison. Study a desk-top sized power station turbine and I'm quite certain you'll find the efficiency drops considerably. For several fluid flow reasons, axial steam turbines don't scale down very well.

Several studies indicate TT rotor efficiencies can exceed 90% Wiki Tesla Page. Seems reasonable to believe that TTs have substantial growth potential. Because of their low manufacturing cost, I can see where TTs might be a good choice for residential power production.

 

[NapierDeltic]

It is not a direct proof of Tesla's turbine potential, but as long as internal combustion engines were the kings, most of the money for development went in their improvement. There should be no wonder why progress in electrical power chain was so slow and that it looked unpromising. Now, as due to certain circumstances electrical option is almost on pair, their advantages start to show and much more, we read daily of new research directions of their improvements, technically closer or not to current main configuration.
It is obviously that the only thing I try to point out is that development goes mostly in actual main branch applications because here losses are translated to real money waste. I do not want to make any technical comparison between each of 3 mentioned devices/power solutions.

 

[Nerd1000]

As has been pointed out, there are no practical TTs available for efficiency tests. They're mostly (perhaps all) one-offs and hobbyists creations.

Comparing power station, or ocean-going ship sized steam turbines to all the desk-top TTs is a bit of apples to oranges comparison. Study a desk-top sized power station turbine and I'm quite certain you'll find the efficiency drops considerably. For several fluid flow reasons, axial steam turbines don't scale down very well.

Several studies indicate TT rotor efficiencies can exceed 90% Wiki Tesla Page. Seems reasonable to believe that TTs have substantial growth potential. Because of their low manufacturing cost, I can see where TTs might be a good choice for residential power production.

It's true that conventional turbines perform a lot worse at small scales.

Here's some commentary from Prof. Warren Rice (Arizona State University) on the topic:
https://www.gyroscope.com/images/teslaturbine/TeslaTurboMachinery.pdf

Of note, he states that tesla turbine rotor efficiency can exceed 95%. That's really good. However the process of getting fluid into or out of the rotor winds up being very inefficient, leading to poor overall efficiency of the stage. Also the disc spacing has to be vanishingly small to obtain such good rotor efficiency, which results in the rotor being physically large for the flow rate.

 

[Nerd1000]

It is not a direct proof of Tesla's turbine potential, but as long as internal combustion engines were the kings, most of the money for development went in their improvement. There should be no wonder why progress in electrical power chain was so slow and that it looked unpromising. Now, as due to certain circumstances electrical option is almost on pair, their advantages start to show and much more, we read daily of new research directions of their improvements, technically closer or not to current main configuration.
It is obviously that the only thing I try to point out is that development goes mostly in actual main branch applications because here losses are translated to real money waste. I do not want to make any technical comparison between each of 3 mentioned devices/power solutions.

Internal combustion has a big unfair advantage over steam, it gets to just use the hot gas at the combustion temperature rather than a temperature that the boiler can withstand. As per Carnot the temperature drop across the engine directly correlates with efficiency, so it's very hard for a steam plant to surpass an internal combustion plant on the same fuel.

This is of course why steam largely got displaced to burning solid fuels like coal, as liquid or gaseous fuels can usually be used more efficiently in an internal combustion engine.

 

[NapierDeltic]

It's so good that it affords to evacuate waste gasses at a temperature higher than usual steam's top T; and still stands tall.
Of course, I try not to mix apples with pears, as steam cycles are still the highest efficient thermodynamic and overall energy generation path.

 

[Nerd1000]

It's so good that it affords to evacuate waste gasses at a temperature higher than usual steam's top T; and still stands tall.
Of course, I try not to mix apples with pears, as steam cycles are still the highest efficient thermodynamic and overall energy generation path.

The best diesels just surpass 50% thermal efficiency. Supercritical steam slightly less than this, up to 48%. Of course one can do better than either by using an ICE's exhaust to heat the boiler of a steam engine, like many gas fired plants. Overall efficiency of more than 60% is possible that way.

Anyway, it's a bit moot because as small scale model builders we can't go anywhere near the state of the art temperatures and pressures. I think I would be happy to achieve 10% thermal efficiency in a model steam engine, that would already be beating full sized steam locomotives. Playing with my spreadsheet I managed to get a design with moderate temperatures and pressures up to 16%, which would be at least competitive with an inefficient ICE like a simple 2-stroke. But at model scale the real performance would probably be quite a bit worse than the sheet predicts.