Prototype three rotor steam impulse turbine

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
B

Bogstandard

Guest
This is going to be a quick and dirty exercise, design and make on the run (but I will make sketches as I go).

I walked into this one with my eyes closed. It was on one of my model boats sites, where an American chappie wanted to build a 10ft long warship, but have a pair of steam power turbines in it.

The criteria being, steam driven, as clean running as possible, easy to operate plus easy to build and service, with no special materials required.

So I came up with this

3Roturbinitialsketch.jpg


Steam in one side, using a control valve to give fwd/rev and throttle. Exhaust on the opposite side.

The control valve will be a modified version of the one I designed for my piston valve engine. The outer casing will be ali, rotor of brass and a shaft of ground stainless, running in stainless shielded bearings. The expected size of this proto will be around 3" long (not including shaft length) and 2" square.

The expected RPM of the full sized version (about 2x the size of this one) when running thru a 100 to 1 gearbox will hopefully be between 35K & 40K, giving an output speed of 400 rpm at the 3" props.

This will be, if anyone wants to follow suit, a good exercise in precision turning, boring, milling and rotary table work. Just for the colonials, I will try to remember to do it in imperial, or at least give imperial sizes to use instead, as most of my cutters are metric.

Unlike the other one I built last year, this one will be a lot safer to run, with the bearings being buried a bit deeper.

Now for a search in the stash box for the materials.

John

 
If anyone else wants to sing along with this tune, I went thru my stash this evening and found almost everything that is needed for the main turbine, the control valve is made of scrap bits and bobs of brass, so no need to worry about that yet. Fastenings are what I have in stock and will use whatever is necessary, there will be no requirements for anything special (as far as I know at this moment).
All my materials that I use will be metric, but I will be stating the nearest imperial sizes so you shouldn't have any troubles.

Here is a pic of the recycled materials that I will be using.

3Roturbscrapbits.jpg



#1 - 2 x 3/16" bore (or to suit) stainless bearing with stainless shields (to fit shaft #2)
#2 - 3/16" diameter s.s. shaft (can go up in size a bit, but not smaller)
#3 - 1 1/2" diameter brass rod to make rotor (you will need about 1 3/4" long, but leave on billet if possible (for holding in lathe)
#4 - Approx 2" cube of ali for main chamber
#5 - A piece of 1/2" thick ali plate, 2" x 2" to make the exhaust chamber
#6 - A bit more plate to make end plates for rotor chamber and to mount the bearings into

What is not shown is the material for making the nozzles, that will be about 3/16" to 1/4" diameter brass and about 6" will be required.

If you are only ever going to run this on air, you can make almost everything out of whatever you want, but I would suggest drill rod for the shaft and something heavy for the rotor. I have only used this variety of materials because it will have to be able to resist the corroding effects of live steam, with no lubrication added.

Tomorrow I will be making the rotor and shaft.

John


 
Interesting John I am sure lots of us will be watching this one. Glad to see you back at it and take it easy for awhile with the shoulder.
 
John,

I'm looking forward to seeing how this one goes together. Your single rotor turbine was a very nice looking (and working) project. This one should be even more enjoyable to watch you build.

Chuck
 
You got the idea, now for the hard bit, putting it into practice.

Just a bit of a warning before I start. I am expecting to lose my internet connection fairly soon, but hopefully not for very long, so don't worry if I don't post for a bit, I will still be taking piccies and I will play catchup when everything is stable again.

A slight change to my build materials, the #3 was a nice piece of brass, but a paying job came up, and it was used for that. So I have put my masochist coat on and I am going to make the rotor out of PB1.


Now to start in ernest.

The first grotty sketch is for the rotor, no details about the pockets yet, that will be sorted later. The rotor will be like an extra wide flywheel, with a recess at either end and holes drilled thru to lighten it and get the mass on the outside to retain its momentum.

3Roturbrotorsketch.jpg


This next pic shows the recesses in (using my flywheel recess tool) and a hole wacked thru the centre. I am not too worried about the OD to within a few thou because the rotor case is going to be bored to the OD of the finished rotor.
The hole wasn't reamed or bored, it would be rather difficult boring this length being such a small diameter. The shaft is a sliding fit in the hole, but needs to be made to an interference fit, so what I am going to do is knurl up three raised portions on the shaft and secure it together with force fit and loctite. If I made the hole 1 thou under to begin with, I would expect the shaft to bend as it was being forced in, by doing it the way I am suggesting, I expect to get a good solid fit with no damage to the shaft. Also I will be making the basic rotor first, then fitting it to the shaft, followed by final turning to size, this will ensure I will be fairly close to balance and won't have any unexpected wobbles.
The marks on the side are for where I want to do a bit of rough turning so I don't have to do any heavy stuff when it is mounted onto the shaft.

3RotorTurbine1.jpg


Here it is in the chuck, roughing out to 50 thou deep with my parting off tool, this will save having to have any heavy cutting pressure onto the assembled rotor, and maybe deforming the shaft.

3RotorTurbine2.jpg


Rough cut and ready to go onto the RT to have the lightening holes drilled.

3RotorTurbine3.jpg


Is everything understandable so far?


John
 
HI John,
this looks like a very interesting build, cant say I will get to build, it but I will certainly be following yours...Hope you are recovering well...
All the best for now,
John.
 
This is very interesting!

I've never seen a model steam turbine. I love the use of brass, but I've gotten shy about it of late due to costs. I guess I should just suck it up and soldier on!

I hope someone will build a Tesla turbine sometime too.

Best,

BW
 
BobWarfield said:
...
I hope someone will build a Tesla turbine sometime too.

...

O/T I think Wes (powderkeg) has been designing and casting some parts for a Tesla turbine. I believe he has the intentions of selling said castings...

Eric
 
Just done a bit more, trying to get most of it done before I have to go offline.

I must apologise for a few of thes pics being out of focus, but I don't get to see them until the end, and by that time it is too late to go back and reshoot.

This is the first out of focus one, but what it is supposed to show is the drilling of the lightening holes thru the rotor. I put 8 x 1/4" into it, at 45 deg apart, and they worked out perfectly.


3RotorTurbine4.jpg


This is the drilled rotor, thank goodness that is over, drilling phos bronze is not my idea of fun. Had a few sticky moments with these, but brute force got me thru.


3RotorTurbine5.jpg


Now the next stage, the shaft. I made this one 4 1/2" long, it will be trimmed down at the very end to whatever length is required. What I have done is put the shaft in approx half and half then marked the outside face positions on the shaft.


3RotorTurbine6.jpg


Take the shaft out and use the marked positions to find where I want to put the knurling to form the gripping area.


3RotorTurbine7.jpg


Here is my clamp type straight knurl in operation. It takes very little indeed to raise 2 thou for the grip, in fact the knurl wasn't even properly formed, but more than enough for this job.


3RotorTurbine8.jpg


Position of the raised knurls in relation to the rotor.


3RotorTurbine9.jpg


Another out of focus one, looks like my cam doesn't like the miller.
Anyway, what I have done is centred the drill chuck over the shaft and closed it up. Between the chuck and the shaft is a bit of ali scrap.
I put a ring of loctite around the shaft where it enters the hole, so when the shaft is pushed in, it will pick up the loctite and take it into the hole with it.
It takes very little force using this knurled method, if it was a lot I would have used my press to do it.


3RotorTurbine10.jpg


A lovely assembly job if I must say so myself. It spins up on the bearings really nice, and is only a couple of thou out of true. The next step, after it has dried out overnight is to get it into the collet chuck and true it all up, plus do the final forming of the recesses.
Please don't criticise the condition of my 'precision' tooling. These are the ones I use for my silver soldering, and are not normally seen.


3RotorTurbine11.jpg


For the next thrilling installment, tune in tomorrow, same time same place.

John
 
Had an hour or so, so jumped to the shop again and finished off another stage.

This first one shows the rotor skimmed up and trued, with the correct recesses machined in. If you are going to try one of these and don't have a collet chuck, it can be done in a 4 jaw, but remember that the shaft is the datum to set to, not the rotor.


3RotorTurbine12.jpg


Ready for the milling stage. I have just checked out the free running on the bearings and it looks very promising, about 1 minute run down from around 10k. I have also noticed that the bearings are so free, that I should be able to do the balancing on the setup as shown.


3RotorTurbine13.jpg


Definitely no more machining tonight.

John
 
Today, I should not have bothered.

Went to the shop nice and early, coughing away with the local sparrows. Caught the cat and give it a going over just for luck.

It didn't work.

Turned on the mill and my Y axis readout wasn't working. I was using a squirty bottle for coolant yesterday for drilling the holes, and some had found its way into the slide.

Took it off, give it a good strip down and clean, works like new. But one thing I didn't realise was how badly these things are made. The machining inside looked like someone had done it with a dremel and a carbide burr. But anyway it now works.

So sets up the RT for horizontal work and duly started machining the first round of pockets, after struggling a while with the PB, I got onto the last pocket and duly fluffed it.
In fact to a state where it was not 'fixable'. Caught the cat again, and did a blame recovery exercise on the poor devil, but he now knows how to keep out of my way when something goes wrong.
So to save wasting a total morning, I decided to use this duff one as an experiment.

So went to the next area and cut 12 pockets instead of the 18 I did on the first one. See pic to see what it looks like.

3RotorTurbine14.jpg


This was duly mounted up into bearings and given a dose of air into each set of pockets in turn. My first turbine was made with pockets like the ones in the middle, and it works fine. Thinking I was going to get the same sort of result, deep pocket, good power and runup, shallow, less power and a long runup time. The results were totally the opposite, the shallow pocket run up as though it wanted to fly, at least twice as fast as the deep ones, and the power was still there.

So this morning wasn't a total failure after all (sorry cat).

Using my failure for experimentation has really pushed me up the learning curve on this type of motor.

I reluctantly raided my outside stash and hacked off another lump of 2" brass, and I will proceed to make another rotor.

So one days delay.

John
 
John,

That's an interesting result. I would have agreed with you and thought the deeper pocket would give more power and faster run up. I wonder if the result you got was because the air jet was hitting more pockets at the same time on the 18 pocket side?

Chuck
 
Chuck,

I honestly think it is something to do with the flat area at the back of the deep pockets, causing like a dead area, whereas the shallow ones actually run into each other, so rather than having an interupted pulse of air, it is allowing a smooth transition from pocket to pocket. When I put air onto it, you could hear it 'popping' between pockets on the deep ones.

I will know better when I get the correct nozzles made for it.

John
 
Now back on track, knocked up another one, in a slightly different machining order (just in case it happened again).


What this sketch shows is the depth and position of the pocket in relation to the rotor.
If the middle set is done first (fwd drive), then the outside set (2nd fwd drive) with the 10 degree offset between the two. When they are both completed, just swap over end to end and carry on machining in the same direction, you should find that the reverse pockets will be machined facing in the correct direction and it will be in the right position on the rotor.
Sorry about the shape of the circle but I was trying to use my right hand.

3Roturbrotorsketch2.jpg


What it should look like when machining the pockets.

3RotorTurbine15.jpg


The good and the bad together.

3RotorTurbine16.jpg


The finished rotor. Get the hard bit over and done with, and the rest is plain sailing.

3RotorTurbine17.jpg


Lots of things to do tomorrow, so don't know if anything will get done in the shop.

John

Forgot to mention, I used a 3/8" end mill for the pockets.
 
Wes,

Maybe by the weekend I will have it running, but not under controlled power. It will take nearly a week to make the controller. A lot of very careful measuring and planning in that, I have got to relearn my geometry to get the port in exactly the right position. That is the problem with a few of us as we get older, new stuff pushes the old out of our ears, and it just gets lost.

Just hope the damned thing runs as planned. Mind you with only one moving part (excluding bearings) there is not much that can go wrong.

John
 
I have been hurting my brain a bit more, thinking how to make this drag out a bit rather than wham, bam, that's it.

There are so few components in this engine I can do a lot of piccies to cover everything.

So I have decided to turn it into a little bit of an instructional build, showing newbies, in not too technical terms, how I achieve what I do.

But please bear in mind that this is the way I do it, someone else might do it an easier and better way, it is what I am happy with.
--------------------------------------------------------------------------------------

The rotor is done and dusted.

The next major part is the outer casing. No drawings for this because it is rather easy to remember, there are only two dimensions for the block. Each side is 1 3/4" wide by 2" long.

Here is the raw block of ali, rough cut on the bandsaw out of a much bigger lump. It is about 1/4" oversize in all directions. What I am going to be doing is turning this rough block into a precisely sized block with all faces square and parallel to each other. This technique can also be used for getting square out of round with the same results.

turblock1.jpg


When starting work like this ALWAYS check your fixed vice jaw for running parallel to the spindle cut.
I do it by putting a good solid parallel into the vice, set up a DTI to run against the parallel (you could use a normal dial indicator with a plunger if you can get it in). Tweak up the vice holding bolts, enough to stop it moving about, but with a slight tap the vice can be moved.
Position the DTI in the position as shown and wind the Y axis (back and forwards handle) until the DTI starts to register a reading. Reset the rotating dial to the zero position and then wind the X axis (side to side handle) so that the DTI is run along the face of the parallel. If the pointer starts to go anticlockwise, tap the left hand side of the vice near the tightening handle with a soft hammer until the pointer moves to the zero position. Return the DTI to the start position, re zero the dial and try again. If the pointer on the dial goes clockwise as you move right to left, the right side of the vice needs a tap.
Just keep doing this until you get no or very little movement on the needle while you track the full length of the parallel. Tighten up on the hold down bolts and recheck. If it still runs with no deflection, your vice is set up to perfection.
I always do this check whenever I start a job. No need to slacken of, just pop a para into the vice and do a quick check with a DTI. Tweak if needed.

turblock2.jpg


The above didn't really need to be done with this job, as it is all flycutting top face work, but it is always nice to see how it is done.

Anyway, back to making something.
Pop the hunk of metal into the vice with one of the side faces upwards (the ends are done later), and rest it on a couple of matched para's. Tighten up. Nothing spectacular, just so that we can face the first side.
This pic shows what is needed. A flycutter (not essential, but for big faces like this it is an advantage). A nice dead blow soft hammer for tapping the job down onto para's, and a bit of lube for helping the cutting along. Ali can be a bit of a dog to get a good mirror like finish, so razor sharp tooling and lubricant are the right things to use. For ali, paraffin is the recommended cutting fluid, WD40 is now starting to be recommended as well. Because it is so expensive to use willy nilly, I always make up my own cutting fluid. It is white spirits with about 5% 3in1 oil, and I find it works really well. Dab it on the job when needed, whilst keeping well away from the spinning cutter.


turblock3.jpg


This is what the profile of my flycutter is like, it was shown to me a long time ago and I use this shape all the time, it works in the same way as a round nosed turning tool to produce mirror like finishes. On ali, and a nice rigid machine, 0.100" (100 thou or 2.5mm) can be hacked off in one go, not recommended for beginners to do, but it is there if needed. On a job like this I tend to stick with 50 thou and below. Most people use a more standard straight edged bit in their flycutter. As I said before, I do it the way I am used to.
Back to the job in hand, take a facing cut across the top until all the old cutting marks are gone, then put a very fine cut of say 1 or 2 thou, with some cutting fluid in the mix and take a nice slow feed, fast cutter swipe across the face. You have cut your first datum face.

turblock4.jpg


The first thing to do when it comes out of the vice after every operation, is fully deburr the part, and clean down around the vice jaws. One bit of swarf can knock all your faces out of square.
I am a great believer in marking things up, just so that I don't get confused between operations.
The face I have just machined is marked up D1, standing for datum face 1. I also mark on the end of the block where the datum face is.
Shown also is the most vital part of getting the bits square, a bit of soft round bar. I usually keep this in the channels on my miller table so it is always there, ready for use.
turblock5.jpg


Now we set up for cutting the second side, to be square to the first. Datum face 1 goes towards the fixed vice jaw, the bit of round is put against the job and the vice is tightened up. Just do a rough tap down onto the para's (there is only rough faces against them still). Now do your cleanup and finishing cuts like was done on D1.


turblock6.jpg


Datum 2 is marked up the same way as D1. These two faces should be totally square to each other.


turblock7.jpg


Now mount the block into the vice as shown. D1 goes to the fixed jaw, D2 goes down onto the para's and rod inserted and vice tightened. Now is the time to gently tap D2 down onto the paras until both paras are just gripped.
Now you can machine the top face down to give the correct width of 1.750" between now D3 and D2. Take it down gently when getting close and do a finishing cut of a couple of thou. If you haven't got a mic to measure while still in the vice, the block can be taken out and measured any way you wish to get a reading. As long as the block is put back in the same position, and tapped down onto the paras you won't be far out.


turblock8.jpg


If you have done it right, the faces D2 and D3 should be parallel to each other at the correct measurement and the corners between the datum faces should be totally square.


turblock9.jpg


The bit of round bar is now no longer needed, there are three datum faces that are worked with.
The last datum to be cut goes against the fixed jaw, the first down onto the parallels. Tighten up and tap down. Take the last face down to the same setting as the previous facing cut (to 1.750" thickness). Fully deburr all around and check your sizes and squareness. You should be pleased with yourself if all sides are parallel and square.

turblock10.jpg


I have set this shot up just to show how to get the ends square to the sides. This is the bit where the previous setting up with a DTI is needed, if the vice jaw doesn't run parallel to the cutting stroke you end up with off square ends.
Set the block down onto para's with a small amount protruding out of the end of the jaws (make sure the para's stay within the jaw area, otherwise you will end up cutting steel as well). You can then face across the ends to bring them to squareness and length (if you have a cutter long enough to reach from the top of the face to the bottom). I don't like this method because it usually gives a real crappy finish. There are other ways, like standing the ends up in the vice and flycutting the end faces, but the blocks have to be got perfectly vertical first. That is why I go over to the lathe for end finishing.


turblock11.jpg


What I do is mount the job into my four jaw, it doesn't need to run true, near enough is good enough. Pack the jaws with the same type or softer material than the job. I normally use Pepsi Max cans cut up with scissors (If I tell Pepsi about this free advertising, do you think I will get some freebies). In this case I didn't have any to hand, so I used brass shimstock instead.
Face the end, then turn it A over T in the chuck and face and bring to length.



turblock12.jpg


Job done, with one obvious mistake. Because I used brass instead of ali for protection pieces, I have a bruise mark on a nice clean face. It will easily be eradicated with a swipe across some very fine W & D, but it shouldn't have happened in the first place. Typical, always making extra work for myself, just by being lazy.


turblock13.jpg


The next instructional bit will be how I bore the big hole in the end.

John
 
All fantastic informative stuff, thanks.

Some questions questions come to mind in a sort of devil's advocate kind of way from my experience of operating full size turbines (up to 30,000 SHP a long time ago). It's entirely possible that these points have been considered or are not a problem due to the relative small size of the project.

What sort of steam it the builder of the boat intending? Superheated will obviously be hotter and you may get thermal expansion issues not just radially but axially. Saturated may cause water droplet impingement issues or a puddle in the casing.

Are two ahead stages going to be sequentially valved to admit steam to one then the other or steam will pass from one stage and then to the next?

How will the astern stage be sealed from the ahead steam and how will the ends of the casing be sealed to prevent steam leagage?

Is the steam going to be total loss or will there be a condenser to return it to feed water, if total loss how will the steam exit the ahead and astern stages?

Control between ahead and astern and the firing of the boiler may need careful 'weighting', if the rotor is spinning ahead fast, to go astern or stop in the water, first the rotor must be braked to a stop before changing direction. When at 'stop' full size turbines need to be rolled ahead under steam to maintain even heat and prevent hogging. If ahead and astern valves are both shut the firing of the boiler will need to be reduced to prevent lifting of safeties.

None of the above are meant as any sort of criticism, just some quid pro quo from my erstwhile experience, please continue to fascinate us with your progress.

Al
 
Back
Top