A Simple Uniflow Engine

[Jasonb]

But I bet Garry has his heart set on a flash steam setup like this

[gary.a.ayres]

I do  indeed, Jason.

But I also have my heart set on far too many other things. Funnily enough, stamp collecting hasn't featured among these for more decades than I care to remember. And flower pressing has never been on my list.

But yes - I would love to build a kickass flash steam plant. Equally, I would like to build a big, slow, lazy mill engine. And a beam engine. Maybe with a bit of a contemporary feel to the styling. And perhaps one or two bigger boilers.That's already quite a list alongside a bunch of other, non-steam related pipedreams.

But right now my heart is set on some ice cold cider and some burning BBQ charcoal, so to do MJM's post justice I shall read and respond to it later.

Cheers Guys, and thanks for your support.

 

[gary.a.ayres]

Hi Gary, while you are thinking about silver soldering, I will make a few comments on the overall project.

It is important to separate the ideas of flash steam and high speed engines.  Of course, the two are often associated in a project, and for quite valid reasons, but it is not necessarily so.

A flash steam boiler, sometimes called a mono tube boiler, is a boiler configuration that provides a quite  large heat transfer area for steam raising, with quite low weight for the steam production achieved.  But at the end of the day, the amount of steam raised is determined mostly by the burner, and how much fuel can be burned, so long as you have enough heat transfer area.  I was tempted to include the word efficiently in that sentence, but not all projects are interested in efficiency.  If someone wants to break a record, they want maximum power, and if burning more fuel, even less efficiently, achieves more power, then they go for it.

Part of the low weight is achieved at the expense of water capacity.  There is effectively no water storage in a flash boiler, so they can’t be operated like my little boilers by filling up with water then lighting the burner and start steam production.  Well perhaps they could, but it would be a quite short run.  They really need a feed pump, preferably driven by the engine, but again it is not the only way.  The preference is because both steam consumption and water requirement, the same thing really, are dependent on engine speed, so while it is still not easy to balance the two, it is easier than if a separate pump is used.

Of course, the high power to weight ratio is very desirable for high speed vehicles, whether land based or water, and when combined with a very big burner, can produce awesome power to drive the engine.

But it does not have to be so.  If you use a quite moderate burner then you produce only the proportionate amount of steam.  You could build a quite moderate steam plant for a model boat for very sedate performance.  This approach has its own chapter in Benson and Rayman’s excellent book, even though the rest of the book is a fascinating look into the world of high speed hydroplanes.  I think I remember you buying it, so you will be aware of the valuable information there.

In addition, it is not necessary for there to be no water storage.  A moderate plant can include a pressurised water tank, with an air cushion over the water, between the water pump and the steam coil.  It will smooth out the flow and give a bit of a safety factor against water failures, which might allow you to bring a model back to extinguish the burner.  A separate tank like this can allow design of the plant with a much lower centre of gravity than a more conventional boiler design, which is good for stability of a model boat.  Of course the fire tube marine boiler also has a good low centre of gravity.

The engine is a quite separate issue.  An oscillating engine as you have already built can operate quite slowly when required, particularly a twin cylinder double acting design which does not have the dead spots that the flywheel must supply the energy to carry through.  Similarly a slide valve engine, or any of the other designs we usually see on this site.  The upper speeds are mostly limited by the quality of balancing achieved, not easy with heavy reciprocating parts.  My mill engine achieves around 2000 rpm unloaded with quite low steam pressure, with limited balancing, but I suspect that even at this speed, wear might become an issue if I pushed it that fast for long.  So hardened pins and more attention to bearings might be required.

Slow engines are generally characterised by long stroke compared with the bore, though not necessarily so.  More significantly the steam is admitted mostly after the top dead centre, even when a little is admitted before.  Your engine with the inlet valve operated by a peg on top of the cylinder has the steam admitted equally each side of top dead centre. (I wonder how a double acting engine would be implemented).  More likely a twin single acting, or better still a three or more cylinder configuration with the cranks spaced equally around the shaft.  Because of this necessarily early admission, the engine needs to get enough of a kick every steam admission to accelerate the flywheel to store the energy necessary to push the piston over the next top dead centre position.  Otherwise the engine might just oscillate back and forth through about 340 - 350 degrees.  The engine will be happier going faster for this reason, but it still does not require the extreme speed sought by the racing community, it just won’t run slowly like your little oscillator.  But a quite interesting design, well worth trying.

So while roaring blowlamps, extreme pressure and temperature steam conditions, and high speed racing hydroplanes, are typically based of a flash boiler, uniflow engines, and very highly developed material selection, and manufacturing techniques, there is no reason why you cannot run your simple flash plant with a more moderate burner and your current engine, which might run quite happily around 1500 - 2500 rpm (at a wild guess) but definitely not slow like your oscillator.  Mind you, an oscillator is also is no slouch if you give it enough steam, as the oscillating engine design involves a much smaller oscillating mass than a typical mill engine, in both reciprocating and rotary motions.  Mine go at 2000 rpm measured with a digital tacho when unrestrained and with only the little meths burner, and are really not fussed at this speed.  Much less vibration than the mill engine.

It is one of those things that has to be tried, but if you can tackle a little engine driven pump, perhaps with bought gears, there is no reason that you could not try your flash plant with this new engine.  Keep the burner relatively gentle until you get the feel of it.  But I really would not recommend the flash plant with a hand pump only, unless you have plenty of water capacity.  It all comes back to heat balance.

By the way, stamp collecting is only allowed during the current lockdown situation, and there is another current thread where that can be included.  In the mean time you have an an engine and flash boiler to complete!

MJM460

Thank you for this, MJM.

Although I was aware that the uniflow engine could be appropriately paired with a flash boiler (and have indeed considered that option), I didn't choose to build this engine with that specifically in mind. Another - quite separate - boiler development is on the cards. More of that later.   
However, irrespective of that, it would make sense for me to try running the engine on the flash boiler once they are both complete, but with the following caveats:

First - I may be wrong, but I have a feeling that the engine may require a bigger coil in the flash boiler than the current one as the cylinder bore will be either 30 or 35 mm. Fairly chunky. That said, I understand that coils are considered to be a consumable, and it wouldn't cost much in terms of time or money to twist up a bigger one.

Secondly (and this is a somewhat bigger issue) - I have gathered from the literature that a displacement lubricator won't work with flash steam as superheated steam is too hot to condense within the lubricator and will go straight through to work its mischief on an unlubricated piston and cylinder.  The thing is that I'm not really sure that I want to put my energy into building a mechanical lubricator, at least at this point. One could then legitimately ask why I decided to start building a flash steam boiler if  this is the case. The answer would be that the flash boiler is really a kind of side project for me, something that is motivated more by in-the-moment curiosity than by a clear sense of purpose, built as cheaply as possible and to fill gaps between other projects over a longer period of time. One of those, you know...

I have no plans to build model boats or vehicles - it's stationary engines and boilers that most interest me (though to build an engine that would drive a real small boat would be exciting!). However, I do like the idea of using the pressurized reservoir with air cushion, as mentioned by you  above and as we discussed at the start of my flash boiler thread. That said, unless there is a way round the oil pump vs displacement lubricator issue (and if anyone knows of such it will be yourself!) then the likelihood is that I won't be running anything on flash steam any time soon and that another boiler will be used to drive the uniflow engine on non-superheated steam in the first instance.

I suppose I could always hire out the flash boiler (in its semi-complete state) to philatelists looking for more efficient ways of steaming stamps off envelopes. Or if not more efficient, certainly more powerful ...

Thanks as always for your very knowledgeable input.

gary

[MJM460]

Hi Gary, I would not want to discourage you from your intended course, and I look forward to seeing the new boiler develop.  These are after all, affairs of the heart, and no one else can dictate where we should go. 

However, in the interests of increasing understanding, I hope you will bear with me while I explore some of the issues in your caveats.

The first one introduces a whole bagful of interesting issues.

First, how much steam will your uniflow engine actually need?  The valve principle is actually very different from oscillating or slide valve engines we all normally think about.  The valve is only open for the short time the piston crown is touching that peg which protrudes into the cylinder.  And there is a whole chain of dimensions determining that time, each with its own tolerance, which can accumulate to make the actual time difficult to determine, however, that is not the aspect I mean.

When the usual valve first opens there is an initial inrush of steam to equalise the pressures, then the continuing downward travel of the piston under the influence of the steam pressure, does the work that turns the engine.  But consider the uniflow!  When the valve opens, the piston is travelling up!  Still might be that initial inrush, or might not if the steam remaining in the cylinder after the exhaust closes off is compressed up to pressure by the time the inlet valve opens, but then the engine does work on the steam (it comes from the flywheel slowing) for fully half the time the valve is open.  Think of that for a moment, for fully half the valve opening time, the volume in the cylinder is reducing and only the pressure balances with that in the boiler.  Then steam enters through the now fully open valve and while the pistons goes down until the valve closes, I assume at about 1/4 of the stroke, quite early cut off for any engine, and then takes advantage of the expansion of the steam in the rest of the stroke until the exhaust opening lets enough steam out to lower the pressure again for the return stroke.  Because this engine really does utilise the steam expansion, so I expect it will be quite miserly in its steam consumption.  It is that requirement for the flywheel to do the work before top dead centre that is why the engine will probably need to run faster than typical on this site.  The flywheel needs plenty of stored energy for the non-power parts of the cycle.  Similarly you might need a bit higher steam pressure than is required to run your oscillator.

The next issue is how much coil length to generate the required amount of steam?  Again, first it is important to recognise that the steam produced by a boiler is fundamentally determined by the heat released by the burner.  The efficiency of the boiler is a secondary, but still important consideration.  That efficiency is again determined primarily by the heat transfer area, with other arrangement factors secondary.  My little boilers seem to run at around 60 to 70% for example.  The amount lost goes mostly up the stack in the heat of the flue gas, and is unavoidable.  A much smaller amount is lost from the casing, and this can be minimised by insulation.  There is a lot of variation in the acceptable heat transfer area, but if you calculate it for your vertical boiler, it will give you an idea.  If your coils has less heat transfer area, more heat will be lost up the chimney in the form of higher flue gas temperature, but if more area, then less loss, but it is diminishing returns.  As the steam gets hotter so the temperature difference from the flue gas gets less, you need increasingly more area to get the next little bit of improvement.

As for the coil being disposable, I think that is probably more in the context of those extreme hydroplanes, where they melt copper tubing so have to go to stainless steel, and quite likely push the limits of that.  A more moderate firing rate should allow copper to last as well as your vertical boiler.  But you are also right in assessing that a new coil costs a lot less in material and effort than a conventional boiler, and that is part of the attraction of the concept for moderate applications, while the light weight relative to the steam produced is a major factor for chasing speed.  If you put a little meths burner under your coil, it will not melt the copper, and the displacement lubricator will work just as on my pot boiler.  There is a lot of space between those extremes.

So quite a lot in that first caveat

The second point, about lubrication, is also important and has to be taken into account also for air running if long running hours are required.  It is true that if the displacement lubricator gets too hot, the steam does not condense, and no oil is displaced into the steam pipe.  However, while the jump to an oil pump is necessary for the chasers of extreme, and I am told that the little oil pumps are not very difficult, though I have not yet tried one, there are also a few tricks you can do with a displacement lubricator in a moderate plant if the degree of superheat prevents the lubricator from working.  If you use a steel washer in the connection between the lubricator and the steam pipe, it conducts a little less heat than brass in contact.  And a fibre washer even better.  You can also add some cooling fins to the lubricator body to increase the heat loss.  These require a bit of trial and error with only short runs until you are getting an amount of water accumulating in the lubricator.  Certainly we overlook lubrication at our peril, and as the burner gets bigger, the pump solution becomes preferred.

Regarding your last comments, who among us has not at some time felt “so many projects, so little time”, and those high speed hydroplanes look like wild beasts.  But I suspect we both need to learn a lot more about more moderate plants before we chase that dream too far.  But one day.....

So don’t let me deflect your plans, but it is always useful to have a few more ideas in the tool box for the day they might come in handy, they aren’t very heavy.

MJM460

[gary.a.ayres]

MJM -

Indeed, ideas are not too heavy to carry around. What you have said about ways to make a displacement lubricator work with a monotube boiler (I won't use the term 'flash' here because I understand that the coil can deliver steam at lower temperature) is encouraging, and I will plooter away with the flash/monotube boiler over time alongside other developments. I have no major objection to the idea of building an oil pump other than time and how I want to spend it, but it's nice to know that there are other options too. I had a feeling you would come up with something!

If I understand you correctly, it would seem that the uniflow engine may not demand much steam but that the steam it does use will need to be at high pressure. How that would translate into performance on the coil I have already made will be (for me at least) a matter of trial and error. I take your point about these coils only being a consumable when pushed to the extreme - that makes sense. However, if I found that I needed a bigger coil it wouldn't be a big job to make one. Which I suppose could in turn require a more powerful lamp... we shall see.

All this said, another boiler solution is waiting in the wings. It may seem unnecessarily coy but I'd rather not explain until it's a reality. It will certainly make a difference to he options for running this and other engines though.  Which doesn't mean that the flash boiler will be cast aside unfinished, but it does mean that I'll be able to take my time with it.

 

[gary.a.ayres]

Started work on the two main bearings between the crankshaft and the sides of the frame. Still very much a beginner but now learning to up my game by using my new 4-jaw independent chuck clocked with a DTI to minimise runout instead of the 3-jaw and crossed fingers:



Hopefully this will result in a better engine at the end of the day.

I have also just accquired two of these for parting off - one 2mm and one 3mm:



The 2mm tool sliced through the bronze with very little effort and minimal anxiety on my part! Very nice to use, but counterintuitive how that insert stays in place I reckon.

gary

[gary.a.ayres]

Unfortunately I had a dig-in and mangled the first of the two main bearings, so I put the stock for the bearings aside and ordered some tooling to make the job easier. I switched my attention to making two bronze collars (yay! for ER-32 collet blocks!) ...





... and two cast iron thrust collars:



Here they are, temorarily rigged up on the crankshaft-to-be for demo purposes:



Outboard of each flywheel the cast iron thrust collar will run against the bronze face of the main bearing. The bronze collars are situated between the pulleys and the crank webs, mainly just for the look. I somehow managed to screw up the crank webs by making holes that are out of parallel, so the ones you can see in the picture are destined for the scrap bin and new ones will be made.

Pausing there while waiting for more steel, I began work on the cylinder, which will be of cast iron. I started by squaring a block of the stuff with a facemill and a large endmill:





The cylinder block was then chucked in my new self-centring 4-jaw and clocked in as closely as I could get it (with the aid of a rubber mallet):



This was my first ever experience with a self-centring 4-jaw and I was impressed by how solidly it holds a square workiece.

The cylinder bore was then started with a centre drill and a succession of twist drills...



... and then bored to size (35 mm):



It remains chucked in the lathe as I type. The bore still needs a finishing cut, after which it will be honed. The outside of the block will be given some profiling and the surface finish will need more work.

All in all, some progress, despite a couple of frustrations...

[gary.a.ayres]

The basic bore of the cylinder is now complete. 35 mm (approximately):



The picture below shows the bore, with the cylinder silhouetted against the evening sky. There are obvious machining marks. I don't know what you guys will think, but I feel quite pleased with the internal finish given that I used a toolpost-mounted boring bar, albeit a fairly rigid one. The marks are not deep and the pattern is regular. I have ordered a cylinder hone and some honing oil and I'm hoping that will take care of the internal surface:



I'm hoping that the hone arrives tomorrow. The external surface of the block needs more work.

So, in the absence of a hone, I then started on the cylinder end cap. This will also be square-ish in form and will be bolted to the cylinder but will be visually continuous with it. It will be bolted in situ when I profile the outside of the cylinder block so that the two parts can be milled to shape simultaneously. The cap will have a raised boss on the inside to locate it in the end of the cylinder. It will be drilled and tapped in the centre to accept the steam inlet valve. In this photo I have made a start on turning it down but I thought I'd better stop for the night as the neighbour's kids are probably trying to sleep.



I enjoy machining cast iron but it's a messy old game. Major cleanup required...

Cheers,

gary

[gary.a.ayres]

While waiting for the cylinder hone, I made a start on making some new crank webs. The ones I made before were no good as the holes were out of parallel. I will have to find out where the origin of that problem lies. Meanwhile, though, I have decided to try making a new pair on the mill. I used superglue to fix two pieces of the steel bar together so that they can both be drilled and reamed at the same time:



In the picture it looks like they are not fixed together, but they are. The effect is just because the ends hadn't been squared off when I took the photo. Both ends of the temporary assembly have now been milled square ready for drilling and reaming in the mill. Once that is done I'll use heat to break the superglue bond.

However, these arrived today...



... so it's back to the cylinder. I have never set eyes on this type of hone before, and the stones on the hone are much smaller than I thought they would be, but no problem. I made a start with the honing, the hone being held in a hand drill and the cylinder in the vice:



The bore is coming up with a nice satin finish, but some machining marks are still visible and there is a bit of surface unevenness about halfway down the bore that I can actually feel with my fingertip, albeit slightly. I don't know how much of an issue this is as I'm hoping it will hone out, but I decided to pause with the honing and make a dummy piston out of a piece of scrap brass to test the overall straightness and regularity of the bore. There's no point in further obsessing with the hone until I have checked that the bore is basically true and parallel, and I think the dummy piston will be a good way to gauge that.

[gary.a.ayres]

The next step was to finish the basic form of the cap. It has a spigot which fits the bore fairly snugly. The cap will be bolted to the end of the cylinder and the edges will be milled in situ to create visual continuity between the cylinder block and cap:



In his design, Stan Bray suggests a shop made ball-bearing type check valve as the main steam inlet valve to the cylinder.  I'm not such a purist that I feel compelled to make fittings of this kind when they can be bought from a supplier. I'm using a ready-made check valve sized for 1/4" OD pipe. The cap will be drilled and tapped for this valve, but the photo below shows a mockup of the arrangement just to give an idea:



It was then time to test the cylinder. I turned down a piece of brass on a superglue arbour (yay, Clickspring!) to make a dummy piston:



Then, with my heart in my mouth I anointed it with steam oil and pushed it into the bore:



Amazing! My fears of a crooked bore and jamming were unfounded. The dummy piston behaved in the cylinder just like a real piston should. Smooth action (well, just a couple of very minor tight spots but these will disappear). When I put my hand over the open end where the cap will be, the piston pulled a formidable vaccuum. 

Well pleased with that and tbh I don't feel inclined to obssess much more about the finish of the bore. Once I have drilled the steam exhaust ports I'll give the bore another quick go with the hone and try to achieve the desired cross-hatch pattern, but I'm pretty sure the thing would run fine as it is. In fact I think  that the real piston can afford to be a tiny bit bigger in diameter for even better performance.

The exterior of the cylinder requires more work, mainly of a cosmetic nature. More of that later...

Now, a question: can any of you advise me on my choice of material for the piston? I have round bar of sufficient diameter in cast iron and free machining stainless . I'd be happy to buy something else if needs be, but I suspect I don't need to. I'm aware that cast iron is an exception to the rule about using dissimilar metals together, and I also understand that stainless works fine with cast iron too. Which would you guys recommend?

Also, I'll be quite happy not to use ring(s) of any kind on the piston, though wouldn't mind cutting an oil groove or two in it. However, I'm happy to change my mind in response to advice. Any thoughts?

gary

[john mills]

you have a cast-iron cylinder then the material to use is cast iron .

[Johnmcc69]

you have a cast-iron cylinder then the material to use is cast iron .

But wouldn't SS be better for corrosion resistance?

 My edit: Nice work Gary!

 John

[john mills]

the cylinder is already cast iron so the same care for the cylinder already if you don't want corrosion then cylinder out of bronze  piston bronze too soft packing will serve well .with cast iron then cast iron piston rings rusting is can be a problem.   

[gary.a.ayres]

Thank you both.

I think I'll probably go with a cast iron piston. As john mills notes the cylinder is cast iron so it will be vulnerable to corrosion anyway. I'll just have to keep it dry and oiled when not in use.

I'd rather get away without piston rings if possible. Would just having a plain cast iron piston with maybe oil grooves but no rings be viable or would the absence of rings have a very detrimental effect? If any rings are necessary I'd prefer to go with silicone or viton ones (whichever it is), but would prefer not to use them if possible. I'd welcome any further thoughts on this. The cylinder bore is about 36 mm.

Meanwhile, this evening - a fair bit of thinking done but only modest progress in practical terms. I drilled two exhaust ports through the cylinder walls into the bore, then gave the bore a light honing to remove any burrs, after which I marked out (by hand) and punched for a bolt circle to which the cap will be fixed in due course:



 This is a x 3.5 scale-up from Stan Bray's plans, but his small original version has one exhaust port of the same size as these. His advice is to keep the same size of exhaust port in scaled up versions but to make more of them, so I stuck to his original size but drilled two of them, one on each side of the cylinder. If these turn out to be insufficient I can always drill a third one on the top of the cylinder. The idea is to eventually make a manifold of some kind to connect the exhaust ports together.

Regarding the bolt circle, the bottom right punch mark in the photo is a bit out of alignment. I fixed that after the photo was taken.

[Jasonb]

If you are not going to put the engine to much work then you can run it without rings, easy enough to enlarge one of the oil grooves for a Viton ring if you find you want one at a later date.

Not enlarging the holes is probably more to do with timing so you could put them side by side which may help reduce the number of pipes to couple them up