DIY Tesla Impulse Turbine

[NapierDeltic]

I guess we all can see particular facets of phenomenon and the more we underline them the better.
Can't stop but thinking of these days Chinese tests with hypersonic missile electric/rail gun. The test was not 100% successful and according predictions; and I'm sure hundreds of thousands were spent on simulations to guarantee success. I am sure also corrections will be applied soon...
Science goes hand in hand with experiment in order to succeed. It is not only a matter of what happens but also how much of each.

 

[The Ignoble Troll]

This is annoying. I can't find the paper.

It should be in my downloads but it's full of pdf's with names like te6uhfft6.pdf and it's probably one of those.

I did find this
https://www.google.com/url?sa=t&sou...ChAWegQIBRAB&usg=AOvVaw0mqTszPtBkblqJqmmMb4Nb

Which has lots of other variables and their results.

Sorry K2


My gut would say to treat the edges in accordance with the working fluids expected impingement speed.

 

[Nerd1000]

ADHD struck (damn me and my malfunctioning neurons) and I've spent a bunch of time I could have been doing stuff I'm supposed to do (like work) making a steam turbine design spreadsheet.

Something that came up from this is that the speed of sound in steam is quite a bit faster than air: even at 100 C it is over 400 m/s. Because the Tesla turbine quickly loses torque when there isn't a good match between the jet velocity and the disc velocity, I think convergent-divergent nozzles are counter-productive, your peripheral speeds are going to be impossibly high if the steam is supersonic.

 

[The Ignoble Troll]

ADHD struck (damn me and my malfunctioning neurons) and I've spent a bunch of time I could have been doing stuff I'm supposed to do (like work) making a steam turbine design spreadsheet.

Something that came up from this is that the speed of sound in steam is quite a bit faster than air: even at 100 C it is over 400 m/s. Because the Tesla turbine quickly loses torque when there isn't a good match between the jet velocity and the disc velocity, I think convergent-divergent nozzles are counter-productive, your peripheral speeds are going to be impossibly high if the steam is supersonic.

With appropriate blade geometry, why do you think it would be an issue? Planes, like the F16's weird intake, go through a lot of effort to deal with boundary layer flow?


Wouldn't a higher speed be good because you are giving the blades a higher top end?


Genuinely, as always, asking.

 

[The Ignoble Troll]

I found this

https://www.internationalsteam.co.uk/trains/newsteam/modern26.htm

 

[Toymaker]

With appropriate blade geometry, why do you think it would be an issue? Planes, like the F16's weird intake, go through a lot of effort to deal with boundary layer flow?


Wouldn't a higher speed be good because you are giving the blades a higher top end?


Genuinely, as always, asking.

No matter what shape you make the edge of a Tesla Turbine disc, rounded, flat, or knife-edge sharp, if steam flow is above Mach 1, shock waves will form on both sides of the disc and act to reduce steam flow between the discs. However, it's my understanding that gases exiting a divergent nozzle quickly lose velocity and so quite likely slow to under Mach 1 by the time they impact the discs.

 

[The Ignoble Troll]

I think that a de laval could be positioned a little ways from the blades to tune where the shock wave forms.

The cooling tempurature inversion of a de laval may allow the blades to operate at higher power levels as the gas will be cooler but moving much faster?

This should help with blade stretch, I would think?

 

[The Ignoble Troll]

Funny, most detailed studies omit disc edge experiments.

Regardless, this one has some useful info

https://www.google.com/url?sa=t&sou...0QFnoECDoQAQ&usg=AOvVaw0L85bI0clSPeBAM6YTP8N8

 

[Steamchick]

I must spend more time studying all this. Just been busy with life stuff. On holiday next week.
I understand speed of sound for steam to be more like 550m/s at 100psi... so over 85000rpm for the 3 in turbine?
Needs ceramic bearings! And balancing!
But if below that speed, efficiency is way down. So almost impossible to use them as a driver of anything useful.
Just trying is a but of fun!
Still working on a burner for the boiler. ..
I don't expect to ever get the turbine working - properly. Would it stay in one piece at 100, 000 rpm?..
K2

 

[Nerd1000]

No matter what shape you make the edge of a Tesla Turbine disc, rounded, flat, or knife-edge sharp, if steam flow is above Mach 1, shock waves will form on both sides of the disc and act to reduce steam flow between the discs. However, it's my understanding that gases exiting a divergent nozzle quickly lose velocity and so quite likely slow to under Mach 1 by the time they impact the discs.

There's nothing to be gained using C-D nozzles if you don't want supersonic gas velocity. Note that the turbine is spinning, it sees only the difference in velocity between itself and the gas, so there shouldn't be shockwaves on the rotor even when the impinging gas is travelling at supersonic speed.

The Euler turbomachine equation basically says that (in an impulse turbine) the energy you can extract from the gas is a function of the change in tangential velocity of the gas across the turbine. This is tied directly to your stage pressure ratio, higher tangential velocity means higher pressure ratio of the stage. So if you wish to avoid supersonic conditions at the turbine entry, you automatically restrict yourself to a fairly low stage pressure ratio and thus will need compounding.

Messing about with my spreadsheet, with a 50mm pitch diameter DeLaval turbine at 50,000 rpm I still needed 2 stages to fully expand 200 degree C saturated steam. The same diameter of Tesla would need 100,000 rpm and 2 stages.

 

[The Ignoble Troll]

I must spend more time studying all this. Just been busy with life stuff. On holiday next week.
I understand speed of sound for steam to be more like 550m/s at 100psi... so over 85000rpm for the 3 in turbine?
Needs ceramic bearings! And balancing!
But if below that speed, efficiency is way down. So almost impossible to use them as a driver of anything useful.
Just trying is a but of fun!
Still working on a burner for the boiler. ..
I don't expect to ever get the turbine working - properly. Would it stay in one piece at 100, 000 rpm?..
K2

From what little I've read on failure modes, they tend to implode not explode like a bladed turbine. So at least if you fail it might be gentle.

 

[HMEL]

With appropriate blade geometry, why do you think it would be an issue? Planes, like the F16's weird intake, go through a lot of effort to deal with boundary layer flow?


Wouldn't a higher speed be good because you are giving the blades a higher top end?


Genuinely, as always, asking.

One of the reasons a Tesla turbine is hard to make work is due to its physics result in high rotational speeds. This results in very high root stresses at the axle area. For most materials these loads are destructive. This is one of the technical difficulties in making tesla turbines.

 

[Toymaker]

<snip> Note that the turbine is spinning, it sees only the difference in velocity between itself and the gas, so there shouldn't be shockwaves on the rotor even when the impinging gas is travelling at supersonic speed.

<snip>

Interesting theory,...I've not read any data on this condition. Examining the relative motion between spinning disc and super sonic gas flow; yes, the disc is spinning away from the gas flow, but the disc is still in a fixed position relative to the gas flow. There's an argument to made that spinning the disc, which is still in a fixed position relative to gas flow, will have no impact on shock wave formation.

I don't know the answer,...I'm just theorizing.

 

[The Ignoble Troll]

A quick Google turns up references.

For example

https://www.sciencedirect.com/science/article/abs/pii/S0894177714003100

 

[Nerd1000]

Interesting theory,...I've not read any data on this condition. Examining the relative motion between spinning disc and super sonic gas flow; yes, the disc is spinning away from the gas flow, but the disc is still in a fixed position relative to the gas flow. There's an argument to made that spinning the disc, which is still in a fixed position relative to gas flow, will have no impact on shock wave formation.

I don't know the answer,...I'm just theorizing.

It's certainly true for a bladed turbine.

I think the point is probably moot anyway because Tesla turbines with supersonic tip speeds in steam are probably beyond the capabilities of most materials, the wheel will just explode. Instead use more reasonable subsonic speeds (which also.will better match most loads) and multiple stages.

 

[Steamchick]

To help you appreciate what I am playing with, here are a couple of pictures of the Tesla Turbine made by the local Club's later Chairman. One of the last engines he made.
K2

 

[NapierDeltic]

Was it tested? From the notes on it might seem so. Do you know with what results?
Is it still in working order and can be run?

And a different comment - I have seen a Tesla turbine after a test -run with all disks from package twisted in a horrible way. Care should be taken for heat expansion of disks and mostly not to keep the rim cool (I am thinking of condensate) while the core is heated. I know the temperature difference is smaller for the steam application - if I remember right, that specific turbine was tested with hot gases.

 

[Steamchick]

Running on steam it made a paddle wheel sound at only a couple of hundred rpm, as it filled with condensate. As it became hotter with a few attempts with steam, the speed increased.. a bit - Not a lot! but it was trying to boil/blow-out the water from inside the turbine and we didn't manage to get it hot enough (clear of condensate) so it ran properly.
I have since added a small drain so it can clear the condensate... I hope - at the next run! The boiler is only good for a limited volume of steam at 30psi... this turbine takes a lot more! So I really need to find/make a bigger boiler.

Most use flash boilers, but the club is not insured for flash boilers and testing. Maybe I can get a tap off a small traction engine or railway loco boiler though? I need to find what is available in other club members' cupboards... We did have a 6in diameter 12inch high coal fired boiler - until it evaporated...
I later ran in on my home compressor, at 100psi (Air) it ran above 20,000rpm. (- when the reflective tape flew off and the meter stopped counting and recorded zero rpm!).
The bearings are a bit "second-hand" and feel knackered when you spin-it by fingers, then rest it on your palm. you can feel some horrible roughness. - It NEEDS new (ceramic) bearings capable of over 100,000rpm... as it SHOULD get towards 100000rpm with enough input, when balanced properly, if it comes to true "Tesla" mode.... But it may take a lot of steam to get it above 70000rpm for that to happen...
Also when running on air there was a lot of vibration (It needs balancing! - When stripped for new bearings)...
Most turbine makers make their own balancing rigs... so maybe I need to find one on the web and copy it? - sometime...
K2

 

[Steamchick]

Hi Nerd,
I wonder when you suggest using multiple stages of the Tesla turbine, if one of us doesn't understand it properly? (I.E. Me?).
An impulse turbine can develop about 50% of tractive energy - per rotor - I think? (something due to geometry when the gas stream turns 90degrees from input to exhaust direction against each blade? ). But maybe it was "half-speed" (1/4 kinetic energy remaining?). So multiple rotors each extracting 50% of energy (or is that velocity??) leave a lot to be collected by a later stage (I think?), as the exhaust from the first stage can give 50% to the second, then 50% to the third... etc. so the shaft accumulates 50% + 25% + 12.5% etc. thus achieving high efficiencies?
But the whole principle I think Tesla achieved was a single stage obtaining ~80~90% of the energy, so the exhaust contained too little to make a further stage useful? (back then!).
Also, in the Tesla mode the helix is very fine (e.g. there are smoke demos to show this on Utube?). It starts as a coarse helix at lower speeds, but when the turbine achieves true Tesla mode the helix manages to make thousands of rotations from the outer high velocity zone to the centre "near rotor speed at exhaust ports" - which was maybe 1/10th of the initial speed. And energy is proportional to Vsquared... so necessarily Tesla was right in the possible efficiency he could predict - and achieved. All in a single stage. Or maybe I missed something and got it all wrong? (I often do!!). It is all due to the progressive slowing of the gas stream per revolution that the gas stream makes, as it slows it naturally finds a slower surface - which is at a smaller radius - thus generating the helix, naturally. So the exhaust outflow from the Tesla turbine is very slow (and low pressure, and cool, having lost the energy). You can't get enough velocity, pressure or temperature from the Tesla turbine to power another stage. - I think? - but I really am unsure now, with your suggestion.
An impulse turbine does a Momentum exchange, by turning the gas stream through 90degrees, it needs a force to turn the gas stream, reacted on the blade (like any aerofoil works?), so the reaction force turns the turbine wheel. But the gas stream still contains a lot of energy. Hence the Parson's turbine multi-stage approach?
The idea that Toymaker is using seems interesting: Useing an impulse turbine to extract "half" the energy at a lowish turbine speed, then let the remaining gas stream pass through a Tesla type array of disc-slots, to extract more energy... But I fear the impulse momentum exchange will direct the gas stream "backwards" (and inwards?) from the tips, so rotating in the wrong direction, and thus effectively extracting some energy from the discs... lowering the efficiency? Or maybe simply going in the right direction, but so slowly so as to make very little difference? I.E> after the impulse outer ring, the gas simply transits directly to the exhaust holes in the middle, without any "Tesla" style input to the rotor. But I don't know! - SO I am watching the thread to find out and learn..
Cheers!
K2

 

[NapierDeltic]

Drain, I think, is very good.
Definitely you need good bearings.
You might try to use a reversible electrical motor to bring it first to a reasonable speed and only after work as generator, but the problem of condensate generation remains.
Clearly it needs more steam volume and if possible overheated.
But it is a nice assembly and I hope you will improve it.

Later edit: Would you consider a simple setup like a copper serpentine 6-9mm OD (for split heaters/coolers) and a propane blow torch to be too risky for superheating steam released from small boiler? No big engineering behind; just for test-run.