Ideally the tangential speed of the gas should be zero at the exhaust, it should be flowing entirely radially or axially depending on the turbine layout. Take a look at a turbocharger turbine and you'll note that the blades are angled away from the direction of rotation at the exhaust- that's intended to get rid of the last dregs of whirl from the gas as it leaves the outlet.
DIY Tesla Impulse Turbine
- Thread starter Toymaker
- Start date
Help Support Home Model Engine Machinist Forum:
NapierDeltic
Well-Known Member
Thank you, Steamchick, for sharing your experiments and the theoretical info, and I would love to hear more about results of your trials. Some pictures would be more than welcome!
OK. Next Time I am at the club (and rememeber).
K2
K2
NapierDeltic
Well-Known Member
Little by little.... 
How much pressure will the boiler give you, and do you think it will be able to keep up with the demands of the turbine ?
That boiler with current burner cannot run the Tesla engine.
At the pressure the boiler can take, without much superheating, it may never be capable of running the Tesla engine satisfactorily.
But I am making a larger burner (It can take about a 4.5kW gas burner in my calculations with un-forced exhaust up the chimney) - of around 3.5kW, possibly more if my design is good enough?
The significant problem is the water (condensate) that turns the turbine into a water wheel! It ran the other day when I did a mock-up with a 3kW burner, but slowly (maybe 100~200rpm) and sounded like a paddle wheel sloshing around. But even so it demonstrated the Tesla surface friction turbine action, when it simply started turning as soon as the steam was introduced. It runs (just!) as low as 10psi (without water drag).
The whole project is to teach a new guy about steam, so a project with problems (that is robust and relatively low pressure!) is teaching him very well what can - and does - go wrong, and he has remarked to me how he is learning more by watching the engine "not work" then understanding what is wrong, than by just seeing something that "works out of the box". It is not about making a Tesla turbine do real work. The steel bearings are pretty rough (you can feel it when you hold the engine and spin it by hand). With 100psi air from my compressor it only manages 20,000rpm, yet the maths say it needs to be at around 100,000rpm for the true "Tesla effect" to work. Bearings are only made for 30,000rpm. so ceramic bearings and fine balancing will be needed... if we are bothered...
It is just a bench demo project for open-day running for education really.
In re-furbishing the boiler, I have added longitudinal stays, after re-setting the blown end plates, re-set the pressure relief valve to match the capability per calculations, hydraulically and steam tested and certified the boiler with the inspector, and the "apprentice" has been involved all the way. Next he has to add cleading/insulation, based on a radiant heat double barrier. (2 layers of shiny aluminium foil separated by an insulating material).
https://www.hammockforums.net/forum/archive/index.php/t-7215.html
Finally, I shall make a superheater (when a big burner shows what it can do) so we can experience improvements from superheat.
Does this help?
K2
At the pressure the boiler can take, without much superheating, it may never be capable of running the Tesla engine satisfactorily.
But I am making a larger burner (It can take about a 4.5kW gas burner in my calculations with un-forced exhaust up the chimney) - of around 3.5kW, possibly more if my design is good enough?
The significant problem is the water (condensate) that turns the turbine into a water wheel! It ran the other day when I did a mock-up with a 3kW burner, but slowly (maybe 100~200rpm) and sounded like a paddle wheel sloshing around. But even so it demonstrated the Tesla surface friction turbine action, when it simply started turning as soon as the steam was introduced. It runs (just!) as low as 10psi (without water drag).
The whole project is to teach a new guy about steam, so a project with problems (that is robust and relatively low pressure!) is teaching him very well what can - and does - go wrong, and he has remarked to me how he is learning more by watching the engine "not work" then understanding what is wrong, than by just seeing something that "works out of the box". It is not about making a Tesla turbine do real work. The steel bearings are pretty rough (you can feel it when you hold the engine and spin it by hand). With 100psi air from my compressor it only manages 20,000rpm, yet the maths say it needs to be at around 100,000rpm for the true "Tesla effect" to work. Bearings are only made for 30,000rpm. so ceramic bearings and fine balancing will be needed... if we are bothered...
It is just a bench demo project for open-day running for education really.
In re-furbishing the boiler, I have added longitudinal stays, after re-setting the blown end plates, re-set the pressure relief valve to match the capability per calculations, hydraulically and steam tested and certified the boiler with the inspector, and the "apprentice" has been involved all the way. Next he has to add cleading/insulation, based on a radiant heat double barrier. (2 layers of shiny aluminium foil separated by an insulating material).
https://www.hammockforums.net/forum/archive/index.php/t-7215.html
Finally, I shall make a superheater (when a big burner shows what it can do) so we can experience improvements from superheat.
Does this help?
K2
A bit of a tangent, but has anyone tried a De Laval turbine? Seems like a better match to normal RPM ranges and retains the simplicity of a pure impulse turbine. Though you would have to make a multitude of tiny blades.
Seems you're an excellent teacher K2.
Funny you should mention De Laval. When the center drills I ordered arrive, I will use them to drill out the current straight drilled holes, reshaping them into a more divergent (De Laval) nozzle, as this drawing shows.
The conical nozzles formed by the center drill isn't ideal, but it's a lot closer than the current straight holes shown in the first post on this thread.
Of course, after the steam from the nozzle impacts the blades, which are simple flat plates, the steam is forced inwards, between the discs, which isn't how a De Laval turbine is designed. Still, those tiny, flat blades have already proven pretty good at extracting energy from compressed air. It will interesting finding out how much power this little Impulse-Tesla turbine can generate.
Do you have the convergent part of the nozzle too? It's not in the drawing.
Unfortunately, no; the nozzles are not Convergent-Divergent,...only Divergent. Each nozzle hole was made by drilling a hole into a brass tube (see post #1). I have access to the outside of the tube to machine the divergent cone, but there's no easy way to get access to the holes from inside the tube. The entire cross section of the brass tube looks like this:
Does that big hole (6mm) count as a convergent nozzle section? I'm thinking, No.
Last edited:
Depends on the nozzle diameters that inject into the hole? Maybe?
Are you familiar with air knives?
Not sure what holes you refer to?? Drilled nozzle holes are 0.026" (0.66mm) diameter.
if it helps, that drawing is to scale.
I know a few uses for air knives, but I've never studied how they're constructed.
I'm not sure how long the throat of a de laval can be but there is room in venturi's for stretch... so depending on your geometry you might have some pressure/velocity changes that are de laval like.
Just wondering if the geometry of an air knife orifice might help. On the surface they would seem ideal
The ideal divergence in a venturi is a 9 degree included angle.... the gas will do that whatever cone you drill that is 9 degrees or larger. And I think I read somewhere that over 15 degrees included angle there is no benefit...? Any better advice please?
K2
K2
I know at some point that you get nozzzle flop potential with a divergent nozzle but I couldn't find that info in my brief search.
I have heard of "nozzle flop" that Old Men discuss. - not there yet so can't help...
(Sorry Ladies - It's "a man thing").
Moderator. Please delete as appropriate.
K2
(Sorry Ladies - It's "a man thing").
Moderator. Please delete as appropriate.
K2
I've been reading that the efficiency of a De Laval turbine is roughly 85 to 90%; of course, that's design efficiency which is seldom, if ever, achieved. Still, being able to extract even 80% of the energy contained in steam in only a single turbine stage is quite an achievement.
