Page 4 of 4

Re: Using a steam indicator

Posted: Fri Nov 16, 2018 12:37 am
by k4kfh
So I checked out a book (which appears to be about 130 years old, funny enough) on steam indicators from the University library. I have been learning quite a lot about how one might use them on a full size engine and what "ideal" looks like, but the book said one thing that made me curious. It said not to connect the indicator to holes on the bottom of the cylinder on a horizontal engine, because they will fill with water. Could this just be because water is bad for the mechanical parts of the indicator compared to steam? Or will water actually screw up the readings? I thought this was odd, seems to me like a "buffer" of water would be better all around.

This might be because you would have to make a U shaped section of pipe to ensure the indicator drum was right-side-up, and the lower portion of the pipe would fill with water while the higher portion up near the indicator would be filled with gradually-cooling steam that's trapped behind the water. Maybe that's ridiculous, but that's the only explanation I could think of (and it would have no effect on my idea, because I was going to put the sensors under the wooden base anyway, so there would be no funky U shaped section)

Re: Using a steam indicator

Posted: Fri Nov 16, 2018 7:25 am
by Bill Shields
the water filled pipes restrict the pressure changes which come VERY QUICKLY....and give very funky readings as a result.

You have the same problem with very small pipe connections to / from a 'model' indicator.

Re: Using a steam indicator

Posted: Sat Nov 24, 2018 9:31 am
by Cabbagestack
FWIW

Aside from the hardware aspects to the digital steam indicator (pressure transducer, crank position encoder, Digital Analog Converter, assorted data acquisition & signal processing gear) -- I forget now some of the hardware stuff I had worked out -- I would think that you can solve for the area of the plot by coding (for me Java , any language will do) something like Simpsons Rule or Reimann sum to arrive at Mean Effective Pressure. I would store the incoming values of the position vs. pressure signals in say a 2x360 array for every degree of rotation. I had been working on just such a project for quite some time, a year or so ago, (if only at the theoretical level) but have put it aside because I must CNC my milling machine and then put together a foundry, both of which taking considerable time and resources - and will be ongoing for quite some time to come, I'm afraid.
But it was my intention to transmit the data via Bluetooth, write an app for an Android mobile device or tablet to compute the data and show plots in real time, and show Indicated HP and speed.
I was having trouble finding a reasonably priced pressure transducer capable of temps up to 500­­°F. I wanted to compare plots on an engine with saturated vs. superheated (actually just very dry would be more accurate, i have somewhat doubts if I could get the steam into the superheated regime)

Re: Using a steam indicator

Posted: Wed Nov 28, 2018 1:46 pm
by k4kfh
I had actually planned a similar thing, minus the Bluetooth because I am not that brave and this is a stationary engine anyway, not a locomotive. I've had some calculus already so I figured a trapezoidal Reimann sum would be accurate enough for kicks and giggles. I think if I played my cards very carefully I could do all this on an Arduino, and display it on a small OLED screen along with the graph.

As for the pressure sensor, I am a little stuck here too, although I may try using one of the little 1/8 NPT sensors that is available on eBay. I believe those were made for oil pressure in cars, so I would guess they could tolerate some heat. For $8 it may be worth the gamble. I have saturated steam now but my engine is a little bronze/brass thing and my boiler is fairly primitive so I see no reason to hassle with superheated steam.

My thinking was that a set of short 1/8" OD tubing could be run underneath the base to the two sensors, with the microcontroller concealed under there, and the indicator card displayed on a little OLED screen in real time. Before I do this, I'd like to pick up one of the little Graham industries engines since they have Stephenson valve gear and thus variable cutoff, which makes this whole thing a lot more interesting.

My only concern is condensation. If any condensation in the tubing will screw up the readings, then this project warrants a pretty complex system of drain cocks to purge water from the cylinder and tubing before enabling the indicator, and this may warrant a larger engine because of the minimum size of drain cocks and pipe threads.

Re: Using a steam indicator

Posted: Wed Nov 28, 2018 11:02 pm
by John Hasler
Tubing that completely full of liquid right up into the transducer will transmit pressure changes very accurately and quickly. Trouble comes when there are bubbles in the tubing. Then all the fluid up to the location of the bubble has to move as the bubble expands and compresses with pressure changes.

Re: Using a steam indicator

Posted: Sat Dec 22, 2018 5:53 pm
by Cabbagestack
John Hasler wrote: Wed Nov 28, 2018 11:02 pm Tubing that completely full of liquid right up into the transducer will transmit pressure changes very accurately and quickly. Trouble comes when there are bubbles in the tubing. Then all the fluid up to the location of the bubble has to move as the bubble expands and compresses with pressure changes.
What plan of action would you suggest?

Re: Using a steam indicator

Posted: Sat Dec 22, 2018 9:11 pm
by John Hasler
Fill the tube with oil. This will work well with a solid-state pressure transducer because the displacement is so small that it can be treated as a sound wave. I don't think it would work with a mechanical indicator because of the relatively large displacement.