Steam turbine, which of the design is effective??

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

Dear all folks, i am planning to make own steam turbine with reduction gear, all will be ball beared. Which of the turbine wheel is effective when we are talking about the same turbine wheel diameter with difference construction and use same steam pressure: Reaction-, pulse- or Tesla turbine?

 

I saw the Jensen and Wilesco steam turbine has difference construction from other such as Saito steam turbine and Proteus turbine. Will theJensen- and Wilesco turbine be useful for modelboat?

 

Come with the advice from you about steam turbine

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

I can help wth turbine design but I would need to know a bit more about what you are trying to achieve first .

 

Here are a few notes just to get started :

 

(1) All steam turbines are either impulse or reaction or sometimes both together . The terms impulse and reaction only describe how the turbine transfers energy from the steam – there are always many practical alternative turbine designs to actually utilise this energy transfer .

 

(2) For simple one stage turbines the DeLaval design is usually the most efficient but it can be a bit awkward to make – so easier to machine designs have evolved using the basic DeLaval concept but simplified shapes .

 

(3) Descriptions of model turbines usually concentrate on the design of the rotor itself

but the whole success or failure of a small turbine depends far more on the nozzle design .

 

[Keith Wardill 1Participant @keithwardill1]

Hi, Jens,

 

There may be some useful information to help you in the following thread on this forum

 

http://www.model-engineer.co.uk/forums/postings.asp?th=47249

 

Good luck

Keith

[Richard ParsonsParticipant @richardparsons61721]

Jens – Do not be put off by the thread above. Have a go but be aware

1. Small turbines are greedy for steam. Successful units tend to use waste steam.

2. Small turbines are FAST. I had one go up to 2 million RPM before it disintegrated. The aircraft Gyros run at about 14-16,000 rpm. They run in a partial vacuum because the air is sucked into them, but NO power is taken from them (except overcoming friction).

3. Their efficiency is abysmal. I was after power output.

4. Both Stumphs and DeLavals are not so difficult to make if you make the correct tooling.

5. You will have 3 or 4 BIG problems

a. Bearings, you will ‘fry/curry’ them by the dozen, But things have moved on in the past 20 years,

b. Reduction gears – I often got the dreaded ‘Grrrrr-wheee-poof’. As the gear stripped, the turbine ran away and disintegrated!.

c. The physical strength of the turbine wheel. Calculate the internal stresses on your wheel at 2,000 rpm, 20,000 rpm, 200,000 rpm and finally at 2,000,000 rpm. It is horrendous.

d. Nozzle design. In these sizes remember you are well down into the boundary layer.

6. Beware there are many claims and conspiracy theories about Tessla’s turbines and the rest of his works BUT that is enough of that.

Go for reaction turbines, I did try an expansion turbine, a lot of work and I could not get it to spin. I did but the clearances were too large.

I am afraid my note books vanished some years ago

If you succeed please write it up. Turbines are such simple machines and well worth working on. I will watch this space for signs of success.
Good luck

Dick

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

The two main reasons why almost all model turbines are seemingly so inefficient are that :

 

(1) A large part of the energy in the incoming steam is lost in the nozzle before it ever comes into contact with the turbine . In a simple turbine with plain hole nozzle operating at modest pressures and temperatures over 90% of the energy is lost .

 

(2) The angle of attack of the nozzle and the entry and leaving angles of the turbine blades have to have a specific and fairly exact relationship to one another for higher efficiencies and this relationship is not the intuitively obvious one .

 

To understand this without doing any maths do two things :

 

(i) Imagine sitting on the leading edge of the turbine blade and looking back at the

steam jet approaching you from the nozzle and try to visualise the angle which it

seems to be coming at you from .

 

(ii) Draw out simple DeLaval turbine with a nozzle about 60 deg off axis (30 deg off

the turbine disk) and blade entry and leaving angles of 45 deg and try to imagine

why these angles are a better bet for efficiency than the more obvious 45 deg

nozzle and 45 deg blade angles .

 

Anyone familiar with maths can draw out the velocity diagrams but this is not nescessary for basic understanding of what is going on .

 

More :

 

Turbines can be scaled up or down according to a simple set of rules – so get one to work and you can develop a whole family .

 

Turbines venting to atmosphere have to have a totally unrestricted exhaust path . This usually means venting direct to a big area annulus and then into a big area pipe out to fresh air .

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

Hi, also not simple to get enough power out of the steam turbine to run the model boat?

 

The vanes must be curved to make “jet stream” in other side of the turbine wheel?

 

I am planning to make the turbine wheel at 1-1/4″ diametre with 25-30 vanes.. but not sure.. The ball bearings are at 4 mm d x 9 mm D x 4 mm W.

 

I saw at the Karstein turbine wheel, the vanes are mounted at side instead top of the turbine wheel, how are the effect compared to a turbine wheel with vanes mounted in top of the turbine wheel.

 

Better to machine the vanes in the turbine wheel (with milling machine and rotary table) or mount the loose vanes on the turbine wheel and solder the parts of turbine wheel togheter?

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Hi Jens .

 

Your requirements for a steam turbine seem to be very similar to those of a gentleman that wrote in a few weeks ago about a turbo-generator for a traction engine .Though there would be some difference in power output needed in each case the geometry of the turbines would be the same .

 

A turbine wheel to generate useful power really has to be cut from solid so the design has to consider the practicality of manufacture .

 

If you want to pursue this matter see first what means you or any other contributor can devise for cutting true DeLaval blades with a symmetric shape and as near as possible 90 deg deflection angle – ideally 45 deg in 45 deg out .

 

Some methods used before are by copy cutting using a ‘wobbler’ type set up and by hollow milling cutters . Nowadays it would be very easy indeed to make turbine wheels by CNC .

 

Design details that make a difference are :

 

(1) smooth surfaces anywhere that steam touches .

(2) Fine edges to the blades especially on the entry side .

(3) Lots of clearance between the blades so that a secondary nozzle can’t form .

 

If you would like to come up with a practical turbine disc first and then send me the details I will do some sums on it , design a proper turbine nozzle to suit and work out the output power , running speed and steam consumption . If nescessary we can go round the loop more than once until the best overall design is arrived at .

 

Disc stress and bearing loading are unlikely to be a problem but I will check these as well if you wish .

 

MW .

 

PS: In previous posting I ommitted to mention another common reason why model turbines don’t seem to work as well as they should – the turbine nozzle has to be the controlling nozzle in the whole steam line . If there is for instance a partly closed needle valve in the steam line this could be the controlling nozzle and all the pressure drop will occur here leaving very little at the nozzle . It sounds bizarre but you could have 100 psi in the boiler but end up with only 2 psi at the nozzle if you get the design completely wrong .

Edited By MICHAEL WILLIAMS on 29/04/2011 10:15:55

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

Thanks for the answer from you..

 

I searched the info about the model steam turbine ,and i found the book in PDF file. It is a book “Model steam turbines” by H. H. Harrison. I downloaded the book from Library of the university of California. The book is well written about the model steam turbine and how to calculate and make the steam turbine.

 

 

 

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

Hi folks!

 

I has maked the steam turbine after the drawning of Elmer Verburg steamturbine with ball bearing.

 

Then i took the test with difference oils in the bearings.

 

Here are the tests of oils and measured revolutions per minute:

Work pressure= Mobil ESP 5W-30 / Air tool oil /  Diesel oil (fuel for diesel engine as lubricant)

 

1 bar= 4700 / 13230 / 17600

1.5 bar= 6200 / 17900 / 25700

2 bar= 11400 / 21660 / 31500

2.5 bar= 15500/26340 / 36700

3 bar= 18600 / 28800 / 41800

 

As i can see the ball bearing get less friction with diesel oil as lucricant, it is sensitive for jet stream from nozzle. With high revolution above 15000, i can feel there are enough torque when trying to stop the turbine shaft.

 

I can think: With less friction means less waste of the steam and fuel for heating the boiler.

 

What is your meaning about the oil/rpm to a “fixed” working pressure?

 

Edited By Jens Eirik Skogstad on 04/05/2011 19:50:30

Edited By Jens Eirik Skogstad on 04/05/2011 19:57:56

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Hi Jens ,

 

(1) The energy loss which occurs in lubricated ball bearings running at high speed is almost all due to pumping loss – not simple friction as such . The oil has to be continually squeezed or ‘pumped’ through the passages of the bearings as it rotates and there is a large amount of viscous drag .

 

(2) Your results make complete sense – the higher the viscosity of oil , the higher the pumping loss and the lower the running speed .

 

(3) Designing bearings for high speed machines is much more difficult than for low speed ones . If you would like to send me a copy of the part of your drawings showing the bearings I’ll see if I can suggest some improvements .

 

(4) You don’t mention whether you were running your turbine on air or steam for these tests . Lubricated ball bearings running on hot steam have different problems to ones running on cold air .

 

(5) As general priciple most turbine designers use bearings which are much to big for the purpose . There is almost no load on the bearings of a well designed and well balanced turbine wheel so very small bearings will serve and have less pumping loss problems . For your turbine of about 40 mm diameter 4 mm bore bearings would be adequate or even possibly 3mm experimentally .

 

(6) You don’t have much choice in tiny bearings but generally the more ‘open’ the design of bearing and especially of the ball cage are the less pumping drag will occur .

 

Best wishes ,

 

Michael Williams .

[Richard ParsonsParticipant @richardparsons61721]

Jens your results are very interesting. When I was trying to build the things one of my problems (amongst many others was) fried bearings. Some of which were from a very expensive supplier and two sets few given to me by a company which made ‘air driven’ aircraft instruments which ran at about 15,000 RPM.

The best which lasted about 35 minutes total running time were made of the carbon rods from old dry batteries. The shafts were piano wire used on a Spinco ultra centrifuges. They survived three stripped primary drive gears.

One of my problems is that my turbines were small (max rotor size was 50mm) so to get power which was what I was after they had to spin very fast.
Best Regards

 

Dick

Edited By Richard Parsons on 05/05/2011 12:49:28

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

Today i tested with steam with good results No problem with ball bearing with diesel oil, no mixed with water. The turbine ran above 35000 rpm with good torque. When i removed the heat of the boiler, the steam turbine was still running up to 2 minutes before the pressure was dropped of.. there was enough torque.

 

I am uploading the movie of my turbine in Youtube.. ready to see later..

 

Here is the plan of the turbine: http://www.john-tom.com/ElmersEngines/12_turbine.pdf

 

I added the ball bearing 4mm x 9mm x 4mm instead bronze bearing. The turbine wheel in brass, 5/16″ thick and the wheel is near close to the turbine house, about 0,5 mm clearance. The vanes is milled with 2 mm saw cut wheel. The shaft was silverbrazed to turbine wheel before complete machining the whole turbine wheel. There is no out of balance.

 

 

 

 

Edited By Jens Eirik Skogstad on 05/05/2011 23:06:11

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

Here is the movie: http://www.youtube.com/watch?v=OVlh7KDxcdg

Edited By Jens Eirik Skogstad on 05/05/2011 22:54:22

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Hi ,

 

Jens :

 

The mechanical design of your turbine with the 4mm bearings etc is quite ok and for a simple turbine I would not change it .

 

The gas flow design of the turbine is not so good and efficiency will be much lower than it need be . It depends now on whether you want to stay with this simple design and just use it directly for your original purpose of running a model boat or whether you want to get into the more general topic of designing high performance turbines .

 

Richard :

 

A 50 mm turbine with only moderate steam pressure and temperature could in principle develop several horsepower outut . The main problem with small steam turbines is that it is very easy to design one where the steam goes clean through the turbine and out the other side without transfering any energy . Success is all down to creating a clean , stable very high speed jet from the nozzle and setting up the correct interaction geometry between steam jet and turbine blades . I personally would be very interested to hear about some of your experiments and perhaps all people interested in this subject would like to discuss the whole of this topic in more depth .

 

MW .

[Werner JeggliParticipant @wernerjeggli14222]

Posted by MICHAEL WILLIAMS on 08/05/2011 10:41:39:

Hi ,

 

Jens :

 

The mechanical design of your turbine with the 4mm bearings etc is quite ok and for a simple turbine I would not change it .

 

The gas flow design of the turbine is not so good and efficiency will be much lower than it need be . It depends now on whether you want to stay with this simple design and just use it directly for your original purpose of running a model boat or whether you want to get into the more general topic of designing high performance turbines .

 

Richard :

 

A 50 mm turbine with only moderate steam pressure and temperature could in principle develop several horsepower outut . The main problem with small steam turbines is that it is very easy to design one where the steam goes clean through the turbine and out the other side without transfering any energy . Success is all down to creating a clean , stable very high speed jet from the nozzle and setting up the correct interaction geometry between steam jet and turbine blades . I personally would be very interested to hear about some of your experiments and perhaps all people interested in this subject would like to discuss the whole of this topic in more depth .

 

MW .

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

Hi, i has maked the reduction gear to steam turbine, the gear ratio is 10.2:1. But not tested before the sear set is completed and ready to run. I am planning to test the horsepower with miscellaneoud air screws and tachometer + the program “Thrust hp”, then i can learn out how much effective is the rpm range and steam pressure.

[adyParticipant @ady]

Only a guess but maybe there is a connection with inefficient small wind turbines.

 

A chap I know had one with a single blade size of about 1.5 meters for about two years.

When he took it down I asked about it, thinking the powers that be had objected but it turns out it hardly made any power at all for him, and we’re in a relatively exposed windy area.

He reckoned that he would need a blade of around 10 meters to make it worthwhile.

 

His solar unit was fine, and generated a reasonable amount of energy.

 

Soooo.

 

If you can figure out how to make a reasonably efficient blade system for a small wind turbine perhaps this solution could be carried over to helping with small steam turbines.

 

2centrinos

 

Edit:

I watched the video.

Inefficient or not, that’s a darned kewl little bit of kit.

Edited By ady on 09/05/2011 23:46:58

[Richard ParsonsParticipant @richardparsons61721]

Michael – The work I did started in the early 70s with the Stumpf type turbines. Because of limitation in machining capacity (a little Unimat SL) I was limited in sizes and materials. It sort of worked making a ‘wee-ing noise’ when running on low pressure air, but it had no power. It would stop if you looked at it cross eyed.

I read everything I could find in the County Libraries –not much- I admit. But I came across the DeLaval turbine and nozzle. My early DeLaval wheels were simply slots sawn at 45° across the disk and nozzles were made using clock maker’s broaches. These disks were easier to make than the cutters for Stumps type. More by luck than judgement the blades on my second DeLaval over lapped. It looked promising especially when it disintegrated (on air). It was then I saw an illustration of a DeLaval which had its nozzle at 70° to the axel (20° to the plane of the disk). This I discovered had been done to reduce the effective velocity of the air/steam flow. I steepened the blade angle which were ‘slab sided’. As things were so small, I did not think that aerodynamic shapes would have any effect. About 1973 I had saver up enough cash for a Myford ML10 which being more rigid allowed me more scope. I also met the late Professor Chaddock and as a result I started to make curved blade rotors. I did nothing between 1974 and 1977 moving house and work etc. But in 1978 I was given an old Oscilloscope and by attaching a magnet to the turbine shaft I could measure the rotational speed of the disk. It was getting frightening, but I noticed that when I connected up the circuit I could hear the turbine slow down. I tried making annular nozzles (sort of 2 dimensional venturi with vanes in to direct the flow). I even built a 4 stage Curtis unit (never again). I was still getting less than 2-5% of the theoretical output. Or ‘louse power’ as I described it in a letter to Professor Chaddock.

It all came to a stop when I came to the ‘notice’ of the ‘local authorities’. I owned a dog and my new neighbours were Moslems. They did not care but the ‘Authorities’ did and they wanted the dog to go. So for the next year or so I was involved in various legal battles with the ‘authorities’. All of which I won, but they seldom paid my costs. Their eternal excuse was that these ‘expenses’ were not in this year’s ‘budget’. These sorts of niggles rumbled on including my shed being opened (in my absence) and sealed for 2 months by ‘Elfin Safety’. This went on until I just ‘vanished’ early in this centenary. You cannot ‘buck’ city hall.

I am afraid all my work notebooks and drawings from the 70s and 80’s were lost/misplaced/removed etc, but if I can get the metal I might restart work.

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Werner Jeggli : Your posting copies mine but any comment from you seems to have been lost off the bottom – Please try again ??

 

Richard Parsons : Thanks for information – very interesting so far .

 

Ady : If you can get any sketch or photo of the bits I’ll have a look and see if anything simple can be done to get better performance . The concept of matching the ‘jet’ angle and turbine blade angles matters just the same in large installations as little ones and surprisingly small adjustments can make big differences .

 

Oddly the most efficient simple DIY turbine shape for small wind generators is the Sevonius rotor – two halves of an oil drum mounted in antiphase on a pole axle ! It just so happens that the entry and exit angles come about right for the rather moderate winds we have in this country and being vertical axis it doesn’t have a prefered wind direction .

 

MW

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

Hi folks!

 

The turbine is near complete and there is propeller shaft added. It is some parts to make ready to the turbine and add the propeller. I has tested the turbine with reduction gear, there is really good torque, difficult to stop with the fingers at the propeller shaft while the turbine is running, but it is a question about start torque from zero revolution to working revolution. The reduction gear house has oil inside to keep the gear lubricated. I will rig the test stand to test the turbine with boiler connected to the turbine and the propeller in watertank.

 

I early post i wrote: I am planning to test the horsepower with miscellaneoud air screws and tachometer + the program “Thrust hp”, then i can learn out how much effective is the rpm range and steam pressure.

 

I will drop this test since there is enormous torque in the turbine to test a lot of propeller in watertank.

 

[Ian S CParticipant @iansc]

Jens, you could drive a generator from the shaft, and load that to give the electrical power out put, ie., amps and volts. Some where I saw a test rig for a hot air engine using this system, and the generator was mounted so that the body of it was able to rotate, this was to enable the torque to measured. the little generator will be only about 50% efficient, but not as messy as propellers in water. Ian S C

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

Posted by Ian S C on 02/06/2011 13:41:50:

Jens, you could drive a generator from the shaft, and load that to give the electrical power out put, ie., amps and volts. Some where I saw a test rig for a hot air engine using this system, and the generator was mounted so that the body of it was able to rotate, this was to enable the torque to measured. the little generator will be only about 50% efficient, but not as messy as propellers in water. Ian S C

Ian and folks there, how can i calculate the effect by generator and volt/amper if the rpm is known?

 

Edited By Jens Eirik Skogstad on 02/06/2011 14:42:48

[Ian S CParticipant @iansc]

If you know the revs, and can measure the torque, you can calculate the power in watts, or horse power; Watts = RPM x torque(inch oz) divided by 1352. Or Watts = grms x cm RPM x .00001026.

With a generator with a known load measure the amps and volts, multply these and you have watts. If you do both, you will have an idea of the efficiency of the generator, I’v spent / waisted hours with my hot air engines making graphs.

I use a home made rev counter as per James G. Rizzo’s books, and an own design of brake to measure torque. Ian S C

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Hi ,

 

A simple friction brake system is far more effective for measuring relatively small torques than a generator . The friction brake system is (nearly) an absolute value system depending for accuracy only on the dimensions of the parts and the accuracy of the weights used .

 

 

Edited By MICHAEL WILLIAMS on 04/06/2011 09:13:13

[Author]

Posts