The Steamboating Forum

I made a few test runs of Nyitra tonight logging the pressure.

On the second run I left at around my MAWP of 200 psi.

There was a lot going on and whatever reason the ASME Pressure Relief valve that usually goes off at exactly 200psi didn't, I think I saw 205 PSI on the gauge before I really opened up the throttle on the engine and the pressure started decreasing.

At an rate unless steam can somehow compress I don't understand a max of anywhere near the 240psi on the graph. ?

-CB

Zoomed in.

20 strokes that much, almost 50 over observed gauge boiler pressure?

Could this be back pressure? have a very large 1" steam main that is 10ft+.

-CB

P.S. Wondering if there was slug flow from carry over?

P.S. I think I was confusing the two runs yesterday. The safety did relieve somewhere around 202 to 205, and there was carry over going through the safety. I just don't understand the recorded peak, the main boiler pressure doesn't increase that quickly on the pressure gauge and it wasn't up near 240 psi.

Transducer calibration?
RussN

RNoe wrote: ↑

Fri Jul 23, 2021 8:21 pm

Transducer calibration?
RussN

It could be. I just am confused how it could be so far off.

It makes me want to put a sensor right on the boiler so I could correlate boiler pressure available to cylinder pressure.

-CB

Can you add the time scale units to your graphs?

My current guess is say above 150psi I think the sensor is getting damaging or causing anomaly's(Non linear response, and this time out of the specified voltage range for the sensor).

I just got yesterdays data which was near 200psi, but not above, the safety didn't blow.

The data below shows beyond the range of the sensor. That means the voltage from the sensor was outputting a voltage above the specification. (Above 4.5V)


Video that is a small segment of the actual run of the data above:

Unfortunately you can't see the engine because of the rear railing.

https://www.youtube.com/watch?v=N0IlYO71AgQ

I haven't been putting the time scale because I haven't had access to my computer in Chautauqua and I need to improve the arduino firmware, the current time stamps it is recording aren't that useful.

-CB

P.S. If you just look at the left side vs the right side the right side of the graph the quality of the data is more spurious. I think that the sensor was temporarily or permanently damaged. Some combination of the high pressure and temperature doesn't work very well with this sensor. If your boiler only went up to 150psi it might be fine.

I feel like maybe the heat is confusing the sensor.

This graph is zoomed in on the data, just around the time when the sensor was reporting above 200psi.

In this graph If I am understanding correctly each rise and fall is two strokes of this piston.

-CB

P.S. Well it was only a $20 sensor. I think I'm going to order one these for $30 that says its rated for steam and comes with its own thermal management solution:
https://www.amazon.com/Akozon-Pressure- ... s9dHJ1ZQ==

Yes, and you now see the problem I was having with the cheap pressure sensors. If you zoom in on one single stroke, the result is not going to be of any high quality UNLESS you're running the engine very slowly. And those cheap sensors won't hold up for long in these kind of conditions, unless you're working with less than 75psi wet steam. They're just not made to meat high industrial standards.
You CAN trick around by using my arduino setup and use fewer samples per rotation, but my long-time goal is to get a clear and professional pv-diagram. I've been working with 100 samples per rotation and it secretely doesn't fully satisfy me already. I think 200 to 1000 samples would be perfect. And the adc of the arduino is so-so. You can definitely use it to see results, but there are better and faster options. It's all about reducing response time in the signal chain and if just one link is slow, it throws off all data.

But I am positively surprised that someone takes the time and opportunity to accompany me on the search to our goal to a digital indicator. There are so many challenges, but it's really fun and involving to tinker around for a solution.

Thanks, Barts for the link. (edit: I just checked it. Those are the pressure transducers, which I think CAN work under very relaxed conditions, but I'd consider their lag of 2ms already uselessly slow for any serious attempt of drawing a pv-diagram UNLESS the engine is running very slowly. The issue is, that 2ms is not the number you can work with. You would need to have a lag, that is just a fraction of the sample rate you're using to prevent it from having an impact on the pv-diagram. So, a 2ms lag might mean you could only sample every 20ms - with 100 samples each rotation, this would translate to only 30rpm for a very clear pv-diagram with those pressure sensors. The closer you get to the lag of 2ms, the more the pv-diagram is going to be garbled by the sensor's physical restraints. It's more so an issue for me, since I secretly want to make it work into the thousands of rpms)

Last year they were running P vs Crank Angle diagrams on a big Diesel 8 Cylinder engine running at 400RPM, with very good resolution of the compression and power strokes. Around 800 PSI peak pressures, and far hotter than our steam. Those transducers were retailing for about $550 each.

For comparison, 5 degrees of crank angle equals 2/1000 of a second, and it appears that these transducers are far faster than that.

I made a few graphs without averaging every 10 samples into one.

It doesn't really show any more detail, you just see the errors more clearly.

The detail level is ok, but the anomalys above 150psi make the data kind of useless.

I was trying to get some baseline data that might help me recognize when the piston valves have worn out.

Still interesting to see the data it did record.

I wrongly I guess thought that the pressure in the cylinder would get near 0psi at each stroke, but clearly above 100psi it doesn't do this.

I guess you want a cushioning effect a little?

I would have thought you would want the maximum pressure on one side of the piston and almost no pressure on the other side of the piston. Maybe that would be too rough on the engine...

-CB