Greene Automatic Cutoff Steam Engine

[Dennis]

Glad to see you making progress again.

Dennis

[kvom]

After a large number of valve adjustments that resulted in the engine making 1-3 turns at best, I located one problem that may be the key.  The input valve rods are connected to a yoke with a slot in which lives a small bronze slide. The slide has a .125" hole for a pin that is supposed to be threaded 3-48 on each end.  The pin is rotated by a clevis on the actuator rod.

As it happens, I was apparently fooled by the SW model showing the 3-48 nut that I used a 3-48 pin.     This results in significant lost motion meaning the input valve moves only a fraction of the distance it should.  Apparently I did it corrctly on the head side valve.  Short term, I'll use a 5-40 screw that "ought" to fix it, and hopefully I'll someday get a run video.

I found a local woodworker who's committed to making me a base from some brown stuff, and I hope to get it in the next week or so.

[Admiral_dk]

It will be great if it turns out to be the solution 

[kvom]

Alas it wasn't.  But that valve's motion is now much smoother.

[Dennis]

Kirk, good to see you back on line, I was beginning to worry.

It sounds like you have checked and rechecked the valve mechanism many times,  have you looked at the eccentrics?  I remember early on a  builder of this model found his eccentrics were off by 180 degrees.  Also, the two eccentrics are out of phase by 90 degrees because of the motion to open and close the exhaust. 

I know you are very meticulous in tracing the valve movements so if I am stating the obvious, I apologize.  The Green Engine valve mechanism is fussy to get adjusted correctly, however, I haven't heard of any problems from other Green engine builders getting their models to run.

Dennis

[kvom]

Dennis,

The reverse eccentric is either at the top or bottom of its lift when the input eccentric is at the end of the stroke, but I don't see it matters which way.  Since the exhaust clevises have one above the bracket and the other below, the reverse valve motions are opposite.

I've found that the angle of the clevis with the bracket has a big impact on the valve travel, and with the angle set pretty close to 30 degrees, I needed to adjust the length of the valve rod as well. 

I'm only going into the shop one day a week as we have a lot of home renovation in the works.

[kvom]

Getting close as I have half an engine.  That is, half the revolution is powered (crank side valves) but not head side.  Most obvious issue is that the dash pot piston is not closing the input valve completely.  It sticks halfway down into the pot.  I might need to take a thou or so off the piston if nothing else obvious is visible.

[crueby]

Getting close, sure you will get the rest going.   

[Dennis]

Gook news Kirk,  looking forward to seeing that running video.
Dennis

[Craig DeShong]

We're all pulling for you and waiting to see this model run.  It's a tough build but you're getting there; just another issue or two to resolve.

 

[kvom]

Another day of frustration and no evident progress other than perhaps backwards.  Plenty of tweaking of valve motion yielding nothing.  So it occurs that maybe I'm doing the timing completely wrong.  I based my process on how the slide valve works in that at the center of the stroke (piston centered in the cylinder and conrod halfway between TDC and BDC), one input valve is completely open and the other completely closed.  The exhaust valve opposite the open input is full open and the other closed.  At TDC and BDC both input valves are closed.  Regardless of my tweaks, with air applied there is a strong flow through the exhaust hole in the bottom of the cylinder casting, and I'm not getting any resistance when turning the flywheel in the opposite of the correct rotation. 

Since here the input and exhaust are not directly tied as they are with slide valves, perhaps someone who's built a Corliss can chime in, since there too the valves aren't directly tied.

On the bright side, the base that I contracted a local woodworker to build to my plan is finished, and I'll pick it up tomorrow.

[crueby]

If you are getting free flow of air through to the exhaust, that suggests either:
- the piston is not sealing, so the air on one side is free flowing past it and out the exhaust. If you close all the valves, and turn the engine over by hand, you should feel resistance since you are pressuring one side and creating a vacuum in the other.

- the exhaust valve on the side being powered is not sealing, or is not closed when pressure is applied to that side. Perhaps the valves appear close by the linkage, but internally the valve is not rotated to the proper position? When I made my MEM Corliss, that was an issue I had.

Is the air coming out the exhaust regardless of where the crank is in its rotation?

[kvom]

Yes, constant exhaust.  I had previously "verified" with just the cylinder block that with each pair of valves set manually that the piston would move properly.  The piston seals very well with only a thou or two of clearance.

Other than that it is impossible to see exhaust valves directly.  I scribed a line in the fixed exhaust valve showing the open position when aligned with the end of the sliding valve.  And since both exhaust into the same core chamber and out the same hole, with the engine assembled there's no easy way to see which or both are open to air flow.  Unfortunately I didn't mark the closed position, but the slide needs to move only .10" to close.

[crueby]

Can you disconnect the valve linkages to the exhaust valves and move them by hand? If so, can check where the closed positions are that way and mark the movement. Or, if it never closes, then at least one of them is not sealing at all. If air is coming out at all positions, one or both of them is not closing or is being prevented from sealing. My guess is both, since it is coming out at all positions. It would also seem like the input valves are not sealing either, since they should prevent air going through the cylinder at TDC and BDC when the valves are not supposed to be open yet.
One thing easy (I hope) to do is to take off the cylinder end cap, and run pressure to the engine - as you turn it over by hand, the air to that end of the cylinder should stop and start. Could be all the valves are leaking.

[Dennis]

Hi Kirk,
I think Crueby is on the right track.  It sounds like the exhaust timing is off since your tests with the cylinder block show the inlet valves working correctly.

The angle of the links between the rotating exhaust actuating rod and the exhaust valves is very sensitive.  If you reverse the angle of the connecting links between the exhaust valve and the rotating rod, it will change the timing 180 degrees.  One exhaust valve opens with the link perpendicular to the engine then closes as the link rotates down, the other valve opens with the link perpendicular to the engine and closes as the link rotates up.  By changing the length of the link or adjusting the length of the threaded rods between the valve and valve crosshead , you can reverse the direction the valve travels to open and close so the valve is closed when the link is perpendicular to the engine.  The think the point here his that there are a lot of places and options to adjusting the exhaust valve action.

Years ago, I wrote a paper about setting up and adjusting the valves but have not been able to locate a copy and don't remember very much of what was in the paper.  I believe I started with the inlet valve just starting to open at top dead center while the exhaust valve opposite the opening valve completely open.  On a real engine you would probably start to open the inlet valve a few degrees before TDC so the incoming steam could cushion the change in direction for the piston but on a model we look for smooth operation with no load so opening the inlet valve after TDC is going to make the engine run smoother at slow speeds.

Sorry for writing so much, and you have probably already figured out most of this, but hopefully there is something here that will give you an idea where to look next.

Good luck,
Dennis