Relationship between airpressure and flow as it relates to model steam/air engi

[Brian Rupnow]

Yes, a very interesting thread title. When I want to control the speed at which one of my engines run, I turn the air pressure up or down with the regulator on my airline. By changing the pressure, I am actually regulating the flow of air, because the engine is not running under load. So, consequently, I am controlling the speed of the engine, not the power output. I could accomplish the same degree of control over the speed by running a high pressure, but putting a butterfly valve in the line, to regulate the flow of air. When I wanted the engine to run slowly, I could close the butterfly valve untill only very little of the air got past it to operate the engine, even though the air was under high pressure. Now, lets imagine that I am driving something with my engine. As soon as I put much load on the engine, it will stall out, because there just isn't enough air pressure to drive the piston down. Regardless of how much air flow I have, even with the butterfly valve wide open, if the pressure is low, the engine won't develop enough power to turn the crankshaft. The force with which a piston pushes on the connecting rod is a function of square inches of piston surface area times the pounds per square inch (PSI) of the air acting on that area. If I turn the pressure up high enough, the force acting on the piston will become greater than the resistance to turning exhibited by whatever the engine is trying to drive, and the piston will start to move, and cause the crankshaft to turn. If I am going to run a peice of machinery with one of my engines, and control the speed with a flyball governor, then I should have my regulator set for high pressure, and any butterfly valve in the feed line wide open. The governor will control a butterfly valve of its own. Lets assume that to start with, the pressure regulator is set on high (35 to 50 PSI, the main feedline butterfly is closed, and the governor will be "balls in" which means that the butterfly valve it operates will be wide open. Since the engine is a "self starting" design, as soon as I open the main feed valve, air will rush into the engine at full pressure, and this surge of full pressure air will overcome any static inertia in the engine and whatever it drives, and things will start to rotate. Things will rapidly speed up, heading for a "run away engine" condition. However, as the engine speeds up, the governor will rotate faster and faster, untill centrifugal force causes the balls to fly out away from the main shaft which the balls are attached to. This action of the balls flying into a "balls out" condition causes the butterfly valve operated by the governor to close. (When governor is in a full "balls out" condition, the butterfly valve will be completely closed).--This is a progressive event. The more the balls fly out away from the stem post, the more the butterfly valve closes untill it is completely closed at full "balls out". As soon as the engine starts to starve for air flow because of the closed butterfly valve controlled by the governor, it will start to slow down, however the air will still be at full pressure.--It just fills the cylinder more slowly. And--as the engine slows down, the force of gravity and the spring in the governor (Yes, the spring that Tel likes so much) will cause the balls to start to return to the "balls in" condition, and consequently open the governor controlled butterfly valve, to allow more air flow and speed the engine back up again. Depending on the sensitivity of the governor, and the strength of the spring, the engine and governor should reach a state of equilibrium, where the engine runs at a constant speed, neither over revving, nor slowing down to the point of stalling. This "optimum speed" can be "tuned" by dialing in the right amount of force on the governor return spring.---Whew!!! Heady stuff, isn't it!!!

 

[dsquire]

Brian :bow:

After reading that through, it seems to me like it would work. Initially it will probably be a bit touchy until you get everything set at a happy medium. Lot of luck with sorting it out, I'll be watching.

Cheers

Don

 

[Kludge]

I'm not sure but as I visualize this, I see it oscillating around a steady state condition but not actually hitting it. The oscillations would become decreasingly small but would continue and the process would start over again if, for some reason, the load changed.

This is not unlike how constant speed aircraft propellors work, though with some extra gizmachies so the pilot can control engine speed and there's a default state to low pitch if the system fails. Memory serving (and I'm going back a ways), there was a method of damping the system some so it wouldn't be driven nutzoid in unfriendly weather conditions - gusty winds and the like.

I'm sure there's a way to dampen this that would allow controlling the engine speed by adjusting the governor while not letting it oscillate around the desired steady state. The first thing that comes to mind is something like a leather (or plastic) disk attached to the fly weight mechanism that would provide a touch of resistance to its movement. Let me mosey over to the thread on the governor and see what sort of chaos I can cause.

BEst regards,

Kludge

 

[Stan]

I went crosseyed trying to read that long paragraph without any breaks in it, but from the part I read, you seem to misunderstand the concept of air pressure and flow through a valve.

A pressure regulator is just a valve being controlled with a spring instead of you adjusting it by hand. You can use a needle valve or gate valve to give give good pressure control. A ball valve or butterfly is more touchy to adjust but will also control the pressure. In all cases the orifice in the valve must be large enough to pass the maximum amount of air required. The flow at any given pressure is determined by the restriction on the end of the line (an engine?).

If you want to control your engine over varying load, then you have to have a large enough line with sufficient pressure to obtain the maximum speed at maximum load. For anything less than max speed or max load you just restrict the flow by closing a valve.

Probably the best type of valve to control air flow with a governor is one with a long stem through a seat. The stem has two or more triangular grooves that get progressively deeper down the stem. This type of valve takes little energy to move it and gives very fine flow control. HTH Stan

 

[ronm]

Nothing I can see wrong with your theory, Brian. That's the way centrifugal governors work. the lack of equilibrium you describe is called "hunting"...Just listen to a big Diesel when it first starts up & the oil is thick in the fuel rack mechanism. The gov. will "overcorrect" & the speed will "roll" until things warm up a little. 2-stroke Detroits are notorious for hunting, esp. when cold. They have what's called a buffer screw, which is an auxiliary spring that acts to dampen out hunting. Most Diesel governors have something similar, not always called a buffer. Diesels are a little bit of a special case as far as governing-A Diesel HAS to have a governor, as it's theoretically impossible for a Diesel to idle. If you hold the rack in a fixed position, the engine will either die or run away...at any idle condition, the governor is constantly correcting the speed. If it's set up & working right, you can't hear the speed change. A carbureted or steam engine can be made to idle w/ a fixed stop on the throttle. The only problem I can see with your "air engine" is the possible effect on the butterfly from air flow. It may overcome the gov. & go wide open. The type of valve Stan described should eliminate that problem. Big engines, Cat D9 for example, have the governor assisted by oil pressure, because the rack & pump elements are too big for the governor weights to control. As a Diesel mechanic, I've studied governors for 40 years, & some people just can't grasp the concept...you seem to have nailed it pretty well...
Ron in CO...

 

[Brian Rupnow]

Stan---Thanks a lot for your help with the English language. I've only been using it for sixty years or so now, so I really appreciate any help I can get.---Brian

 

[mklotz]

I'm not sure but as I visualize this, I see it oscillating around a steady state condition but not actually hitting it. The oscillations would become decreasingly small but would continue and the process would start over again if, for some reason, the load changed.

<snip>

I'm sure there's a way to dampen this that would allow controlling the engine speed by adjusting the governor while not letting it oscillate around the desired steady state. The first thing that comes to mind is something like a leather (or plastic) disk attached to the fly weight mechanism that would provide a touch of resistance to its movement.

I think Kludge has nailed it here.

If you look at the mathematical analysis of a governor (link posted in governor thread) I posted, you'll see that, for a frictional Lagrangian analysis, it predicts the oscillation Kludge is describing. In fact, almost any undamped (i.e., frictionless) negative feedback system will hunt (i.e., oscillate) around its equilibrium point.

A fabricated, real world governor may have enough native friction to self-damp these oscillations or at least reduce them to the point of unobservability. If you build it too perfectly, you may need to add some extra friction to remove the oscillation.

There's a whole branch of applied mathematics, control system theory, devoted to the details. However, as with so many things, the effort required to accurately model the system far exceeds the effort required to simply build the thing and tweak it.

 

[Stan]

Brian: I wasn't trying to be mean, but with poor eyesight it is hard to read a solid block of print with 40 or 50 lines without any paragraphs. If I offended you, I apologize.