Perfect control of monotube boilers in our lifetimes!

[flashtwo]

HI Keith,

PM sent

Ian.

[87gn@tahoe]

Ian,

My boat (Mr. Grosjean's is more like it, I'm just the long-term caretaker until I sell or die) has a bypass and a de-superheater controlled by a quartz rod. It was Mr. Grosjean's goal to make the plant completely mechanically controlled. He was a mechanical engineer by trade and he told me he started researching and designing the boat in the '60's.

I can not fathom the time he invested in R&D as well as tuning the system. I have pages and pages of notes from his first outing in the late '80's 'till the early 2000's, noting improvements to the system, etc. If one attempts to start the system in a manner that does not follow his instructions, it pretty much needs to be shut down and started all over again. When done correctly, I have timed from the lighting of the fire to leaving the dock at 6 minutes and the system is EXTREMELY reliable, as can be attested by Keith (boatbum) and others on this board.

I have the utmost respect for the time and money he put into this project, and that is why I cannot claim the boat truly as my own.

I apologize Keith for my slacking and procrastination pertaining to the drawings and info on my plant. I finally dug out, I THINK, all of the drawings and notes and will begin the arduous task of making copies of them all. I think I'll scan them all at the same time, so as to make them more available to others that may ask.

[boatbum]

No worries Wes, the delay actually spurned my own research and uncovered a treasure trove of additional information on monotube designs and temperments. Though, I do look forward to those drawings

Always impressed by the entirely mechanical controls on "Wayward Bell" but something more dynamic is needed for automotive demands in a steam car and the ever changing demands on the monotube boiler an automobile creates. Yes, call me crazy, I want a tire shreading steam powered car and roostertail making boat!!

Keith

[flashtwo]

Hi,

Here is the monotube boiler controller schematic.

Note: the ESCs are the bulk standard pulse width modulation (PWM) to electric motor interface - the name is a misnomer, since, by themselves, they don't actually control the motor speed and they require an additional shaft sensor and controller to control the pumps strokes per minutes and hence feed flow.

Ian.

[barts]

Many larger monotube control systems run afoul of the substantial feed-forward delay caused by trying to measure boiler exit temperatures; a common solution is to measure temps about 2/3 of the way along the coil.

I continue to think that a Lamont type boiler is simpler, with the possible exception of the need for a circulation pump.

- Bart

[flashtwo]

Hi Bart,

This feed forward delay was noticed very early on in the project even with my small boiler.

The current control system detects changes in feed water demand by looking at the incoming signal from the transmitter.

With an increase in feed demand, rather than wait for the outlet steam temperature to drop, it “heads them off at the pass” by opening up the gas valve in anticipation.

The amount of opening and its duration depend on the change in feed flow demand and a manual offset setting of desired steam temperature.

Likewise, if there is a reduction in feed demand, then the gas valve closes in anticipation of a rise in steam temperature.

This method of feed forward control is tunable by manually amending the sensitivity of the gas valve control to feed demand changes.

The analogy is when you drive a car and see a hill approaching – you don’t wait until the car slows down before applying the throttle, you apply it in anticipation; the same applies in going down a hill.

The big advantage of the software based approach was that different control ideas could tried without modifying the electronic or mechanical components. Different control versions can be downloaded in a few minutes.


All the best

Ian

[Lopez Mike]

For what it's worth, I spend all too much time tuning servo loops of this sort in astronomical telescope mounts. It's called a PID controller (Google it if you have the stomach for it!) for Proportional Integral Derivative. Whole generations of mechanical engineers have gone around the bend dealing with this issue beginning with big naval gun mounts back before WWII and continuing through anti aircraft guns controllers and on and on.

These suckers are available in 'black box' form. Myself, I think I'll stay with my crude old donkey boiler and a hot well float. I'm into steam boating to get AWAY from that high tech stuff.

Enjoy yourselves though.

[flashtwo]

Hi Mike,

I appreciate very much your feelings towards the old mechanical control systems having started my apprenticeship in a power station that only had pneumatic and hydraulic controls – not a piece of electronics anywhere not even electronic valves (tubes) let alone transistors!

The software approach has enabled many different modes of control to be explored more easily than constructing various mechanical systems to do the same job.

I suppose I’ve transferred the complexity and hard work in making a conventional boiler with simple controls, to making an extremely simple boiler with complicated controls - I’m not very good at making conventional boilers!

Now that a viable control scheme has been developed, be it in software, it would be an interesting exercise to replicate it in a mechanical form, especially for the purists.

With regard to the PID – I tend not to use the Derivative bit that might measure the rate of change of temperature, because it also amplifies “noise” in the signal and causes instability. As mentioned above, I use feed forward control (not to be confused with the feed water) , which is the opposite of feedback you can get with microphones and load speakers. The feed forward control control measures the variable at the start of the process, which will eventually cause changes further down the process – in our case the change in feed flow will affect the steam outlet temperature.

That’s sounds a very interesting job you have there, tuning telescope mounts. I suppose the difficult part is changing the direction relatively quickly and optimising the time it takes to settle down at its new position without it oscillating – bet you get it in the neck from the astronomers with the fuzzy pictures!

Please find attached a photograph of my “black box”.

All the best.

Ian.

[flashtwo]

p.s. I should have said the astronomers with the Lissajous pictures!

Ian

[barts]

That looks like a neat project... but my experience with wire wrapped boards in high vibration environments has not been good; the wire fatigues after a while. For a boat carrying people, I'd want to make sure things kept working. Perhaps it's our roads, but the vibration and shocks of trailering tends to take its toll. Also. the marine environment is a hard taskmaster. My brother sailed in the South Pacific for eight years, and carried two GPS units and two sextants - and found himself using the sextant more than once.

- Bart