Page 3 of 8

Re: digital steam engine indicator

Posted: Tue May 31, 2016 6:02 pm
by johngriffiths
"It is a noteworthy fact that no railway authority in Great Britain and Ireland ever believed in these engines; Mr. Webb, and Mr. Webb only, had faith in them. Precisely on what evidence that faith was based we have never been able to discover"

from Webb's obit in The Engineer, 20 June 1906

My dwg shows both outside cylinders are HP whilst the single LP lurks between the frames. So if this were near/on top or bottom dead centre, the other cranks not well disposed, considerable back pressure, and a heavy train, you can see why a wise station master would have a banking engine near by. The troublesome ones were in the main either 2-2-2-0 or 2-2-2-2.
His coupled compounds had quite a long working life.

Some shallow draft twin screw river boats had the HP one side and the LP the other. Got a dwg of one of these somewhere.

I know, off topic.

Re: digital steam engine indicator

Posted: Wed Jun 01, 2016 2:59 am
by Lopez Mike
I had no idea that compounding was ever used on paddle boats. Something new.

Back to indicators . . .

All of the indicator cards I have looked at were taken at a fairly significant speed. High enough that the area within the loop was probably affected by port friction. Would a card taken at very low speeds as when a locomotive was starting a train look quite different?

Re: digital steam engine indicator

Posted: Wed Jun 01, 2016 3:24 am
by barts
When the admission valve opens, the piston is moving very slowly. If the pressure never drops until cut-off, the inlet valve area is sufficient. The exhaust port opening is more critical - there's more volume and less pressure to work with... again, how quickly does the cylinder pressure drop when the exhaust valve opens?

= Bart

Re: digital steam engine indicator

Posted: Wed Jun 01, 2016 2:28 pm
by Lopez Mike
I would think that if the piston is moving very slowly that when the exhaust valve opens that the pressure would drop almost instantaneously. I mean, think of what would be happening in the extreme case of the piston taking minutes to make a stroke. The size of the ports and the rate of the valve openings would seem to be unimportant. I would think that it would produce a very odd looking card.

This situation does occur regularly when starting a locomotive under load. Fairly open throttle settings and very low speeds. The condensation would be almost total. Actually, my experience has been that the cylinder drain cocks are usually left open for this reason. At least until the speed is perceptible.

I suppose there has never been any reason to take indicator readings in such a circumstance. The power produced is minimal. All torque and no go. One would want admission for the full stroke. Sort of like a non-flywheel feed water pump. Grossly inefficient.

Re: digital steam engine indicator

Posted: Wed Jun 01, 2016 2:59 pm
by fredrosse
"Another issue is that all of the historical data is based on piston travel and connecting rod angularity was, I believed, quietly ignored."

Yes that is the case, and connecting rod angularity was ignored, because the intent is to record piston stroke, which is exactly the same as crosshead stroke, and the indicators of old times did attach directly to the crosshead.

"If we start recording pressure v.s. crank angle we will do well to be aware of this. It is easy to correct for this in software if we know the connecting rod length, but which data do we want and, as importantly, does it really make any practical difference?"

We have always intended to make an indicator plot of cylinder volume vs. cylinder pressure, so calculations are needed to convert crank angles to piston position. I do not know of any other way to do this if your only measure of piston position is crank angle readings. For engines with long connecting rods, the relationship approaches a simple sine wave, but still a non-linear mathematical manipulation. For engines with relatively short length connecting rods the mathematics are more complex, with very different relationships for the head end cylinder stroke vs the crank end cylinder stroke.

Re: digital steam engine indicator

Posted: Thu Oct 06, 2016 1:59 am
by troye_welch
I've been thinking about this for awhile now (despite the fact that I own half a dozen of the old-style mechanical indicators). I think a good solution would be to fabricate a linear potentiometer out of nichrome resistance wire. Apply a voltage (5,10,12,24 volts- whatever is electrically convenient) and attach a lightweight wiper to the crosshead. The wiper would then have a potential directly related to piston position, which could be fed into an A/D converter, or just one of the axes on an oscilloscope. Some day when I get a minute, I'll rig one of these up with a microcontroller and a program to grab a few cycles on command (once the operating speed and load are attained) average the data (or not), and then send tab-delimited data via a serial port to a PC for graphing in (e.g.) MS Excel. Such a device (circuit board) could be built for under $10, though it's such a niche market. We're probably the only 4 or 5 people on planet earth discussing such things... Lol! -Troye

Re: digital steam engine indicator

Posted: Thu Oct 06, 2016 3:39 pm
by Lopez Mike
I guess it depends on what we want from an indicator card, whether it be on paper or recorded electronically.

In my experience most of the time I was trying to set the valve gear. The shape of the curve was what I was looking at. I never got out the planimeter and measured the area. Actually, when I want to integrate the area on a paper record I cut out the curve with scissors and measure the weight of the hunk of paper with a sensitive scale!

If what one is trying to measure is the indicated power then things like whether the horizontal axis of the card is piston travel or crank rotation would seem to make a difference. Pressure on the piston when it isn't moving at the end of the stoke produces no power whatsoever. Is there a logical fallacy here in my thinking?

I don't mess with my valve settings from one year to the next. I am a bit curious about the ratio of IHP to BHP. In other worlds, the mechanical efficiency of my little engine. It is received wisdom that piston valves are relatively friction free compared to slide valves especially on the high pressure cylinder of a compound engine. The wear and tear on the valve gear I have observed would tend to reinforce this notion. But how much power loss are we talking about?

So the question I'm asking is, "Does the number we get by integrating the area of the curve on our card have a direct relationship with the number we get from a brake?" Minus the losses, of course.

To put this another way, if we hooked up our indicator to a simple reciprocating steam driven pump, there would be no connecting rod angularity or any sort of geometrical conversion to confuse the issue. If we did the same thing with an engine using a scotch yoke, there would be a conversion from linear motion to rotation with a simple sinusoidal relationship.

How do these conversions affect the validity of the number we call Indicated Horsepower and its relationship to Brake Horsepower which is usually taken at a rotational point?


Re: digital steam engine indicator

Posted: Thu Oct 06, 2016 4:34 pm
by troye_welch
The indicator card is the most direct, reliable way of telling what's going on in an engine during different working conditions. I have easily a dozen engines, all from different sources. I've seen many engines (not necessarily my own) that were timed badly. Sometimes it's a simple offset/shifting of the valve on the valve rod and sometimes it requires a phasing adjustment via adjusting the eccentric (assuming it's not keyed to a shaft). Some engines are timed to run better one way versus another and this can be intentional. The small engine in my boat seems to be timed too early, so that steam is admitted to the cylinder way before TDC. This can be ok if done in small amounts and if the engine is run fast, but in my case, it can sometimes cause the engine to lock up and make it difficult/impossible to turn by hand to get it going.

An indicator card would show me exactly how bad the timing was and answer the question of whether it's bad enough to warrant a fix. They can show the additional power gained by dropping the bottom line through the use of a condenser (increased area under the curve). They can, obviously be used to show actual power, and yes, there is a direct relationship between IHP and BHP (BHP=IHP-friction) Also, it will show wiredrawing though the piping and throttling governors and valves and valve events such as recompression, release, and expansion. Mainly though, my main use would be to check valve events, and to make sure the timing is symmetrical, assuming that's what I wanted. When you see indicator diagrams of engines with problems, and how you can identify those problems just at a glance from the card, the power of this tool becomes very clear.

Re: digital steam engine indicator

Posted: Thu Oct 06, 2016 7:27 pm
by Lopez Mike
Well, all very fine and good. I know all of that. And I don't think there is any question whether the values of IHP and BHP rise and fall together.

What I'm wondering is this. The numbers that we get from a brake are not much in doubt. They can be taken from an internal combustion engine, an electric motor, a wind generator, a horse or a mouse. 550 ft./lbs./second is one h.p. Period.

What I'm questioning is whether our measuring of the card area without correcting for the sort of sinusoidal relationship between piston travel and crank angle is, indeed, giving us the power developed within the cylinder.

For the better part of 200 years we have been recording the relationship between piston travel and pressure and calculating a number and calling it power. And naming it IHP. Maybe this is right. But a brake measures the relationship between crank angle and torque. Are they the same thing? That is after subtracting out friction and thermal losses.

This is not a question relating only to reciprocating steam engines. I.C. people are doing the same thing and taking the same measurements. Surely this has been fussed over before!

I may be all confused in my thinking and suspecting an issue that doesn't exist. Certainly not the first time.

Re: digital steam engine indicator

Posted: Thu Oct 06, 2016 8:01 pm
by barts
If I remember correctly, the horizontal axis on the indicator is supposed to represent the piston travel (e.g. volume), not the crankshaft rotation in degrees: ... ag-001.jpg

Attempts to conflate the two will produce incorrect results; a correction factor using the rod length and stroke should be applied to correct the numbers.

Alternatively, a linear encoder on the crosshead would work well.

- Bart