Model Turbines

I started this thread to create a place where we could share information about model turbines. To get the ball rolling, I'll share the information I have on my model steam turbine and a more advanced design I may build. The photos of the turbine assembled and the major parts are in my album. The drawings for these turbines are also shown in the gallery. I call my existing turbine a tangential turbine since I have not found a common name given for the open pocket tangential flow turbines. The more advanced design is for a Terry turbine rotor that can replace the tangential turbine rotor and use all the remaining parts. This is a very small turbine with a 7/8" rotor that has been run with a very small boiler and a small airbrush compressor. The testing of this turbine is given in my thread for testing model engines. I hope this information can be useful to someone looking for a very easy turbine to build.

Thanks for starting this, TG ... I'm sure that Turbines are under-used, at least by non-dentists.

I'm already pondering the possibilities of a very high speed milling/drilling spindle for the BCA.

... Keep it coming.

MichaelG.

Thanks for the turbine info. My thought is to make a bigger one for powering a generator for my traction engine. One of those projects on the roundtuit list.

Paul.

There was a plan for a Turbo Generator in ME, it was back in the 1950s/60s somewhere (I think). I also read in ME that someone had built a tool and cutter grinder using an air turbine to power the grinder. I don't know if he built his own turbine, or bought one, you can get quite cheap air powered drill/ die grinders.

Ian S C

The article on a turbo generator that Ian refers to was by a R.E. Rowbottom of South Africa and commenced in the 11 Nov. 1954 issue of ME and ran for a couple of installments.

   There was also a very similar one in 3/4" scale described in the American publication " The Miniature Locomotive" starting in the Feb 1952 issue.

John

Edited By John Purdy on 15/01/2019 18:19:16

Hi All!

I am working on a design for a scale model Curtis type turbine - it will have a 3" diameter 2-row wheel with reversing sector. The screen capture below is of the wheel design using Alibre Atom3D. The rotor body is a dish design, machined out of a single piece of metal. The buckets will be individually formed, silver soldered into slots in the rotor, then the shroud soldered to the outside. Once assembled it will be trued to final dimensions on the lathe.

reversing sector

casing cover

Edited By Tim Taylor 2 on 15/01/2019 19:46:21

Edited By Tim Taylor 2 on 15/01/2019 19:49:26

Hi all,

Thanks for your responses. Based on my modest experience with building steam engines. My simple turbine is much easier to make than a steam engine. The more efficient Terry turbine design is still much less difficult to make. My only concern with my Terry turbine design is the 0.005" minimum thickness of the blades. If the cutter can mill out the opening without bending the blades there shouldn't be any problem making the rotor. The 3/32" diameter, .015" wide, keyseat cutter is very expensive but available. My Terry turbine rotor design is a scaled down version of the Terry turbine rotor with the best performance in Dr. Balje's study of high energy level low power output turbines. I used Dr. Balje's guidelines to make a 12" diameter 8 HP steam turbine that was tested in a lab and had almost the exact performance he predicted.

Tim is taking on a much bigger project with potentially much better results. The velocity compounded impulse turbine is ideal for running at lower speeds. I hope he shares with us pictures of the turbine parts as he finishes them.

Tony Jefree described an air powered high speed router spindle in MEW a while back IIRC. I think it was powered by the vacuum also removing the chips.

The Rateaux turbine in the latest ME is interesting too.

TG,

Pic below is what the machined rotor will look like prior to adding the buckets & shroud. I will likely model most of the major parts on my 3D printer to make sure I have things correct before I machine anything.

I am having a similar issue to yours in that the buckets are small and tolerances are tight. What I'm currently playing with is forming the blade from 0.025" flat brass stock in a press using a die. It would be roughly 0.050" wider than the final dimension, with the excess width machined off in the lathe after assembly, this would also provide the desired sharper edge on the blade. I need 80 of the blades for the wheel and another 8 for the sector.

I tried to find enough information on one of the Coppus velocity staged turbines to calculate it's efficiency. The most information I could find was for a Coppus RL-12L that had an ouput of 6 HP at a speed of 1,200 RPM. The inlet pressure was 230 PSIG, the exhaust pressure was 15 PSIG, and the inlet temperature was 500 F. The isentropic change in enthalpy for for the given pressures and temperature is 170 btu/lbm. The theoretical nozzle spouting velocity is 2,918 ft/sec. For a 12 in. diameter rotor turning at 1,200 rpm the tip speed is approximately 63 ft/sec. This gives a ratio of rotor tip speed to spouting velocity of approximately 0.22. There was no mass flow given for the turbine so I couldn't calculate it's actual efficiency. The following chart gives estimated efficiencies for turbines. Using this chart, the efficiency would be approximately 11% to 12%..

For comparison, my turbine with a 7/8 in. diameter rotor turning at 17,012 rpm has almost the same rotor tip speed (65 ft/sec vs 63 ft/sec). This speed was the maximum I could turn a 2.5 in. diameter propeller I used in the test of my turbine (see the testing models thread). I was able to get an output power of 1.92 watts with an input energy of approximately 18 watts, so the overall efficiency was approximately 10.7%. The nozzle spouting velocity of the air was approximately 1,324 ft/sec, so the ratio of rotor tip speed to spouting velocity was approximately 0.05. From the chart in the preceding post, a high-work single stage turbine would get a efficiency of about 15%. The high work turbine would have enough nozzles to get a high blade efficiency and a axial impulse turbine blade would be much more efficient that the open pocket design of my simple turbine with a single nozzle. Running my turbine as fast as I could with air of a lot lower energy content, is watt allowed me to get close to the same overall efficiency.

Sorry to rain on various parades but 63/2918 = 0.021, not 0.22. The 1200 rpm is nowhere near fast enough for a 12" diameter 2 stage Curtis. 12000 might be nearer the mark.

The passage through the blades should be parallel for an impulse turbine, one can't achieve this with blades pressed from thin sheet, and they should be close pitch

Edited By duncan webster on 16/02/2019 20:17:30

Duncan,

Thanks for your feedback. I apparently didn't check what I typed in the post well enough. The ratio should have been 0.022. The value of the efficiency I gave was based on 0.022 You are also correct that a ratio of 0.22 would put the turbine at it's peak hydraulic efficiency. If the the turbine rotor was running near a vacuum in a condensing turbine you could reach around the 75% efficiency shown in the diagram. The turbine speed I quoted was what was shown for the Coppus turbine I used in the example. Like you, I thought this speed was not correct until I saw several other examples of Coppus turbines that ran at this low of a speed. I picked the Coppus turbine I used in the example because it had enough information to get a ball park estimation of the efficiency. Apparently eliminating a speed reducer was worth the low efficiency to Coppus. Even with the relatively high 15 psig back pressure and wet steam, the windage losses would not become the limiting factor until a much higher speed. Thanks for spotting the error.

Actually Duncan, not to rain on your parade, but the 1200 rpm is very real.

Coppus manufactured radially split 2 row turbines in wheel diameters from 9" to 23", specifically designed for hp ranges from fractional to 1000bhp, at speeds anywhere from 600 to 6600 rpm. As well as most typical speeds in the 3600 rpm range, low speed direct drive applications were quite common, such as lube oil pumps, boiler draft fans and feed pumps. While the lower speed applications were not very thermally efficient, the exhaust steam was typically used to feed a deaerator/feed water heater instead of using a PRV, thereby recovering the residual waste heat. I know this for fact, as I worked with literally hundreds of them over the course of 40+ years (as well as turbines manufactured by Elliot, Worthington, ABB & others).

I am aware that a formed bucket is not an ideal design - if you read my earlier posts, you will note that I stated that in the real world they are broached out of a steel alloy. Using a formed bucket is a compromise, as broaching buckets in the very small size for my scale model is not realistic. That said, I can very accurately produce a nozzle block with multiple parallel convergent/divergent steam nozzles whose flow is parallel and feeds the wheel at the correct angle. As designed, the blades will be deliberately wider than the final dimension, with post assembly machining to final dimension providing a knife edge.

Tim

Edited By Tim Taylor 2 on 17/02/2019 01:30:05

TG,

The most efficient use of the Coppus type single stage turbine was/is in applications where there exists a need for low pressure process steam - the 15psig exhaust in your example was likely used to feed a deaerator, process heater, low pressure steam header or similar application. The power produced by the turbine is derived from the isentropic enthalpy drop. Obviously it varies somewhat with inlet/exhaust conditions, but the rule of thumb is that using a properly sized turbine increases the needed flow for the low pressure application by only about 8% over what would be supplied via a pressure reducing valve.

They did make some versions with integral gear reducers (TFR/RLR), and also could be used with an external gear reducer to improve overall efficiency.

Tim

Tim,

Thanks for showing that top efficiency is not always the top priority. Having a reliable source of power that can utilize some of the available energy is also important. You also make a very good point of capturing some of the energy lost in throttling. I tried to find an example of a Terry turbine since they are used for similar situations. I was surprised how hard it was to get information on these turbines. The following is the only example I found that had enough information to calculate the efficiency. The Terry GLT-360 had a power output of 10 hp at 3600 rpm with an inlet pressure of 80 psig, a inlet temperature of 324 F, an exhaust pressure of 9.7 psig, and a mass flow of 501 lb/hr. No rotor diameter was given so I couldn't calculate the rotor tip speed to spouting velocity ratio. The enthalply drop for the steam conditions given is approximately 105 btu/lb and the exhaust would have about 10% moisture. For an enthalpy drop of 105 btu/lb and a mass flow of 501 lbm/hr the input energy is approximately 20.67 hp and the efficiency is approximately 48%. The Terry turbine used a 4 stage velocity compounding as shown below. The lower blade efficiency of the Terry turbine blades is partially offset by keeping the blades full of steam with their reversing chamber. Filling the empty blades is a major loss in single nozzle impulse turbines.

I should clarify when I talked about a single nozzle in the last post, I meant the combination of the 1 nozzle and 1 reversing reversing chamber as shown. This Turbine might have several off these sets depending on the pressures, temperatures, and mass flows available. The following picture was made from the PDF file that had the information on the Terry GLT-360 turbine I used. I noticed in the picture that the vertical distance for diameter of the inlet flange was reasonably close to the distance from the center of the rotor to its top edge. A 2" sch. 160 flange is 6.5 in in diameter so the rotor is probably 12 in. in diameter. Assuming that it is a 12" diameter rotor, the tip speed at 3600 rpm is 251 ft/sec. The spouting velocity of the steam with a enthalpy drop of 105 btu/lb is approximately 2,293 ft/sec. and the U/Co ratio is approximately 0.11. A 3 velocity stage axial impulse turbine could reach an efficiency of around 58%. A 2 velocity stage axial impulse turbine like the Coppus turbine with this same ratio could reach an efficiency of around 43%. The point I wanted to make was the importance of the blade efficiency. The lower cost of manufacturing is probably what kept the Terry turbines competitive.

I realized I added the wrong data to the last post. The data added was based on the assumption of a 16" rotor. Scaling the photo indicated a 12" rotor which would have the following results. The tip speed at 3600 rpm is 188 ft/sec. The spouting velocity of the steam with a enthalpy drop of 105 btu/lb is approximately 2,293 ft/sec. and the U/Co ratio is approximately 0.08. A 3 velocity stage axial impulse turbine could reach an efficiency of around 50% with the same ratio. A 2 velocity stage axial impulse turbine like the Coppus turbine with this same ratio could reach an efficiency of around 33%. The 4 velocity stages of the Terry turbine yielded an a efficiency of 48% so the Terry turbine was much more competitive than I suggested in the last post if the actual rotor diameter is 12". I think the cost advantage would also be with the Terry turbine,

After testing my existing turbine (see ‘Testing Models’ thread), I looked for ways to improve the performance. The major causes of the relatively low performance of my existing turbine are the open pockets instead of blades, a small arc of admission, and attaching to a propeller that requires too much torque for the rotor to reach its optimum speed. My turbine was built with the tools and materials available to me at the time of construction and has many compromises. I’ve found better sources for tools and materials and have more machining experience now so I’m looking at various designs to improve the performance. The following drawing shows my existing turbine with dimensions given a code letter in parenthesis that are used in my analysis. Dr. Balje’s study of high energy level low power output turbines used the same code letters in his report. The conclusions for Dr. Balje’s study indicated as many blades as possible should be used. My turbine has 24 pockets which is about the lowest number of blades used in his study. The length of admission (a) should be as long as possible and as close to equal increments of the pitch (t) as possible. Ideally the admission length (a) should be 1, 2, 3, and etc. times the pitch (t). My turbine has an admission length (a) of 0.102 in. and a pitch length (t) of 0.114 in. so the ratio of a/t is approximately 0.89. The blade width (e) should be approximately 3.25 times the nozzle diameter (h). My turbine has a pocket width (e) of 0.125 in. that is approximately 4.46 times the nozzle diameter (h) of 0.028 in. The average rotor velocity coefficient found from testing my turbine is approximately 0.34 and the maximum power was 1.92 watts. In Dr. Balje’s report, the maximum rotor velocity coefficient for widely spaced blades was approximately 0.90 and for closely spaced blades approximately 0.74. The average rotor velocity coefficient for partial admission was found to be approximately (1-t/2a) times the maximum rotor velocity coefficient. In the next posts I will show some new designs where I have tried to correct the non-optimum values.