A Simple Uniflow Engine

[propforward]

Great to hear from you Gary - Happy New Year to you! Good to see you posting again. The new bench addition looks great - a massive improvement! You must be well happy with that.

[gary.a.ayres]

Cheers, Stuart!

Yes, the new bench raises the tone somewhat.   

[MJM460]

Hi Gary, a new chest of drawers with a nice bench top never seems a very exciting purchase when there are so many tools to buy, but it probably makes a bigger difference to all your projects than just about anything else, and gets used more than just about anything else.  Makes everything so much easier to find, so saves frustration all around. 

I went in to buy a small set a year or so ago, and the hardware store was clearing out the bigger industrial ones to concentrate on the smaller ones.  I bought the big one at a very reasonable price, and it has made a huge difference.  I went back recently and bought another one which would fit in another space I was not using well.

I am sure you will find that nice bench top great as well.

And good to see the progress towards a test of the main engine components.  Not far off now.

I hope the clearance where that pin goes through the cylinder head is not too tight, it might mushroom a bit with the hammering it will get.  When you have the length sorted, it might help to go for a tougher material, that can be hardened with appropriate tempering.  All steps to ponder for the future.


MJM460

[gary.a.ayres]

Hi MJM.

Thanks for looking in once again.

Yes, good storage is absolutely invaluable, and tidiness in the shop can make the difference between enjoying the experience of being in it and that of just tolerating it!

I have indeed been thinking about that pin. The ball in the check valve would be easier to replace than the pin so there's no reason not to make the pin hard (unless the ball is swapped from the stainless steel one to a nitrile one). Even though there is enough clearance, mushrooming of the pin would change the length of it so it needs to be durable.

Food for thought, but I'll stay with that nasty galvanised one at this point for test purposes.

I remain very dubious about whether or not the engine will run without enlarging these flywheels, but it's definitely worth a try. Am hoping to get out to the shops for airline connectors tomorrow, but may not be able to find time to test the engine until next weekend. Work...!

Cheers,

gary

[gary.a.ayres]

Hmmmmmmm.....

Got the connections for the air line sorted out - just had to drill and tap one part to take the ME thread on the valve:



I then tried to run the engine in test assembly form on air, not expecting it to work. Sure enough it didn't.

I slackened some nuts and screws to loosen the assembly up as it was running a bit tight. For the same reason I attached a hand drill to the end of the crankshaft and spun it for a while. Back on the air and still nothing, even at 100 psi from my small compressor.

To make a start in testing out the 'flywheels too small' hypothesis, I fixed a lathe chuck to the end of the crankshaft to create more inertia:



Still very, very little. You can feel it trying to turn over, but it's still light years away from running.

I have a small bunch of theories on this, some of them incompatible, though more than one may be correct:

• Too little inertia without attached lathe chuck/too heavy with attached lathe chuck
• Too little inertia even with attached lathe chuck
• Valve lift too small (pin too short)
• Valve lift too great (pin too long)
• Piston too heavy
• Running too tight (needs running in)
• Problems with 'scaling nature', i.e. design not meant to be scaled up x 3.5
• Not a good runner on air - needs expansive nature of steam to drive the piston once the valve has closed
• Exhaust ports too big
• Exhaust ports too small
• Valve bore too big
• Valve bore too small

At this point in time all I can do is fiddle about and test these parameters as far as possible.

Needless to say, if anyone has any comments on the above theories or other suggestions these will be most welcome at this stage...

[crueby]

If it was a normal d-valve engine, I'd say loosen the screws on the cylinder end to see if the compressed-air and exhaust timing was right, and not coming in from both sides at once or not letting the exhaust out. Are you hearing any escaping air at all, anything going through the engine? Air passages large enough, not blocked by a gasket, timing, are all the usual suspects. Never done a uniflow so I dont have a good mental image of how they work. That chuck on the crankshaft should be plenty heavy if all else is okay.

[gary.a.ayres]

Actually Chris I have no gasket between the cylinder and the cylinder head - it's just screwed tightly on. I didn't notice any air escaping but I will check for that at my next shop session. My hunch is that won't fix it but it's definitely worth checking.

The timing will be determined by the amount of valve lift (i.e. length of the pin). I'll be playing around with that too.

Could be multifactorial...

[MJM460]

Hi Gary, disappointing when we get to connect the air line and it doesn’t run, but who among us haven’t tried, just to see, before the engine is quite ready?  I am quite sure that it will run, we just have to see how well.

That’s an intimidating list of potential problems.  Fortunately a number of them are quite low probability, so let’s start with the more likely.

First the obvious errors as already mentioned by Chris, and no doubt others by the time I write this.  It is also worth trying to deal with whether it is just too tight before looking for anything more obscure.  These little engines don’t produce a lot of power, and friction can easily absorb all they make and more.  If it’s tight, apply a little more oil, then take out the wrist pin so you can check separately whether the crank shaft main and big end bearings are running freely, and if not, why are they binding?

Then try moving the piston to see if it is moving freely.

I would then check the air passages.  With the piston at bottom dead centre, the valve should be shut so no air flow. 

Then with the piston at the top, valve should be open, but again no air flow because the exhaust ports are closed.  Don’t need much pressure, 2 -5 psi should be enough to tell, but not rip the flywheel out of your fingers.  Exhaust ports are obviously closed off by the piston, but much leakage around the piston means too loose.

If you then allow the flywheel to turn slowly still with very low pressure, you might feel the valve closure point, and then a pulse from the exhaust when the exhaust ports open.

Then, can you be a little more descriptive of what you mean by “trying to turn over”?  I am guessing you mean that when you pass over dead centre, it turns about 3/4 of a turn, then bounces back when the valve opens, rather than continuing over top dead centre for another revolution.  Probably the same effect no matter which way you allow it to start.  I believe that it should run either direction as the timing is quite symmetrical.

At this point, you have three variables, the air pressure, pin length and that inertia you are worried about.  Much more encouraging than 12, but I will explain more about why I consider the others less likely later.

For the engine to run successfully, when you flick the engine over top dead centre, it should take off for about half a rev to where the exhaust ports open.  They need to be large enough to reduce the pressure in the cylinder to nearly atmospheric pressure.

If the ports are too high in the cylinder, the expansion part of the stroke will be short, and not produce enough power, and if too small, they will not open enough to exhaust sufficient air, and the compression of this air after the ports close will absorb too much energy.

 During the power stroke, the flywheel has to pick up enough speed for the energy in the flywheel to compress that low pressure air remaining in the cylinder to the point where the pin touches the valve, and continue against the incoming air until top dead centre is reached, beyond which point, the air pressure and piston motion are in the same direction, so producing work which will accelerate the flywheel again. 

The principle is that the positive work produced by the air entering the cylinder after top dead centre, plus the work done during the subsequent expansion with the valve closed, must be more than the negative work required to compress the air remaining in the cylinder when the exhaust port closes, plus the work done against the incoming air once the pin starts opening the valve.

Air pressure is the easiest to experiment with.  I am not sure whether high or low is better, but more likely some intermediate pressure.

I am thinking the pin length is the most critical factor.  If the pin is too long, the negative work, before the piston reaches top dead centre might be too much.  You might expect that you will get it all back after top dead centre, but thermodynamics being what it is, dictates that you will get less back.  The positive work by the expansion of the air after the valve seats is likely to be more  than the compression work on the return stroke before the valve opens, as the pressure will be higher throughout the expansion process.  Hence, I suspect that a shorter pin, which opens the valve later might be the constructive direction to try.

The flywheel inertia might be troubling you here.  If it is too low, it would not store enough energy to drive that negative work part of the cycle.  However, a low inertia flywheel will accelerate quickly, and as the energy stored is proportional to speed squared, more speed easily can make up for less inertia, particularly if you start the engine with a really energetic flick.   If it is too low, it will take off for that first half a rev, then stop and bounce back when the pin hits the valve, or perhaps after a very small pulse of incoming air.

On the other hand a very heavy flywheel will not accelerate quickly with the low torque available, and the extra mass will introduce more bearing friction.  With a good flick to start, it might run a few revs then stop due to loosing more than it produces each revolution.  I think you will get an idea whether the flywheel is very lively due to being too light, or very sluggish, due to being too heavy.  But in reality, there is a very wide range of flywheel inertia that will operate quite satisfactorily.  You have covered a good range between your original flywheels and the lathe chuck.

So I am suggesting that the starting point is to try varying the pin length.  Either make up two or three, or remove the piston and adjust the pin, depending on the construction.

I am suggesting that the effect of the pin is all about timing of the opening, and the valve lift in itself is less significant.  I think I would try first a very short pin, that just allows a quick puff of air in each stroke.  Then try the effect of increasing the length of the pin, but that is getting ahead of myself.  The short puff of air means less work to get the piston to top dead centre, but not much effect on the work done during the expansion part of the stroke with the valve closed.

More than enough words for one post.  I hope that it is a help towards understanding how the engine works, and even better if you already have the engine running by the time you see this.  We all want to see the engine run.

MJM460

[gary.a.ayres]

Hi MJM.

Thanks for this fabulous exposition, which I will copy, paste, print off and take out to the shop as a guide to my next attempt to get the engine running. You really 'get' this engine.

I have been busy working all day and into the evening, so no shop time unfortunately and it could well be the same tomorrow. However, last night (before you made your post) I took another look at the chapter in Stan Bray's book about building this engine and saw something that I had forgotten. He says if you make a scaled up version, don't scale up the valve lift. I take your point that pin length is not just about valve lift but also about timing, but SB's point above already had me thinking that I should shorten the pin. Your post this morning also steers me in that direction so it's the first thing I'll do.

If I recall correctly (without going and looking at the book again) he specifies 1/32" as the valve lift. Now, I have no idea what the valve lift on my engine currently is (very difficult to measure accurately) but some trial and error will be instructive I think. Fortunately I made a temporary pin out of threaded bar to enable me to adjust it by winding in and out if needs be. So tinkering with the pin will be the first thing I do, starting with minimal length and gradually increasing...

Also very interesting is your point that maximum pressure is not necessarily optimum. I had rather short-sightedly assumed that 'the more pressure the better', but I now understand from your description of positive and negative work that this may not be the case.

And then there's inertia. That lathe chuck... I suspect Chris is right in saying that ought to be plenty. I could take my three-jaw off the lathe and put it on the other end of the shaft, but no doubt that would be way over the top!

As for the exhaust ports, SB advises against making them bigger, but instead to drill two or more on bigger versions of the engine. Currently I have two, one on each side of the cylinder, positioned relative to the cylinder ends as per plans x 3.5.

You have given me lots to think about and a sense of how to proceed with tweaking the parameters. Much appreciated!

I'll get back to it as soon as I can...

[Jasonb]

I'd time it so the exhaust has opened  a little before the piston reaches the end of its stroke and onlt just gets to closed as the piston comes back to TDC. Air should  not be entering until exhaust has closed at  or just after TDC and have it shut off at 3/4 of travel at the most.

Pressure really only needs to be enough to overcome ant friction as you are not making it work at this stage, use something to control the flow once compressor regulator set to desired pressure

[Zephyrin]

as the piston is also the valve, most probably timing is wrong...I don't see lot of possibilities !
air intake never opens, I would bet.
a piston extension pushes the intake valve, but if the air pressure is too high, or piston extension too short, it never happens.

moreover, the piston pin that pushes the valve also plugs the port, being of the same dia, no chance that it will work without some changes !

[gary.a.ayres]

Thanks Jason and Zephyrin.

It all points to one thing, at least initially - experiment with the length of the pin/piston extension!

I'll start with the air pressure about 30 or 40 psi and fiddle with the length of the pin.

 

[john mills]

is  this design the type whern the pin opens the valve ,the inlet is only open for a short time.from when the pin hits the valve until  piston goes over top dead centre and to the position when the pin is no longer holding the vale open so the time is short and allows for expansion the port for the exhaust is only open when uncovered by the piston  so the ports need to be as much area as possible most of the remaining pressure hats get out while the ports are uncovered .the piston has to get to the valve to open it for the next cycle .it is meant to suit high pressure and speed  if it can get going .if the exhaust ports are to small and are not uncovered for long enough
for enough pressure to get out it will be in trouble .it should try to go as it comes over top dead centre while the
valve is open  after that it will depend on how much pressure got into the cylinder how  much it expands to get to the exhaust ports and how much gets out how much the pressure drops to see if it can get back to open the valve
again.     thats  what i think the disign is .      Is that what it is i have not seen your drawings.

interesting to see how you get on .    John

[gary.a.ayres]

John - yes, that's exactly the kind of engine it is.

Indeed, the area of the exhaust ports did entail a bit of guesswork for me as I increased the scale from the drawings. If adjusting the pin length and the pressure don't get it running, thinking about the size of the exhaust ports could be a next step I guess.

Thanks for your input.

gary

[MJM460]

Hi Gary, you are most welcome.  It does my old brain good to take on the challenge of understanding the physics that drives these engines, particularly ones like this one that are a little different.

It certainly looks like the consensus is to first try shortening the pin.  I think the screwed pin to allow adjustment is a good idea, but you might need a method of locking it in position.  I don’t know if you have room for a lock nut, but the pin gets quite a hammering and this will loosen it quickly.  You don’t want to repeat Rogers experiment with shortening a grub screw.

There are two separate factors involved with the valve and pin.  First, the length of the pin determines the timing.  But then the size of the hole in the head gives the flow area for the incoming charge of air.  Obviously the pin blocks part of the opening.  So a larger diameter pin will need a larger diameter hole.  Once the ball starts lifting off the seat, the flow area opened increases rapidly, though the passage that retains the ball is also important.  I don’t know what shape you have.  It may be something to come back to, but first try and get the pin length for timing.

The exhaust ports are similar.  Timing and flow area have to be considered separately.  You don’t want the exhaust open to early, as that will limit the expansion which is providing the power, but as John says, you do want them to open enough to let out as much air as possible, as once the port is closed, you have to compress any remaining air as the piston returns to where it opens the valves.  Extra holes as you say, is the way to go, but it makes collecting the exhaust a challenge.  No problem for air, but I believe that you want to run on steam eventually.

MJM460

PS I did make a small edit to yesterday’s post.  The energy stored by the flywheel is proportional to speed squared.  A smaller flywheel can store as much energy as a larger one by rotating a bit faster.  The torque available from the engine during that power stroke will accelerate a small flywheel to a higher speed than a larger one.