Flash Steam or Monotube boilers

[doubletop]

To continue this thread a bit further I had been in touch with the guy who did "Vital Byte" that has a flash steam plant controlled by pic micro controllers that can be found on the Model Boat Mayhem site. I had asked if I could re-post some of his pics here but no reply so you'll have to go over there. However, I did this diagram.

Basically there are two control loops one with an R/C input of speed demand which controls an electric pump pumping water through the flash boiler and the other with which measures steam temp as an input which controls gas flow and hence burner rate. Request for more speed results in more water being pumped, steam temp drops, burner gas flow increases steam temp is maintained and more steam produced, so engine runs faster.

So the only input is speed demand and the rest is controlled by the micro controllers. The guy has a lot of other instrumentation built in as part of his evolutionarily development but he is in the process of rationalizing that down the minimum necessary.

I should say this is not a 'screamer' as shown in the video in the first post but a means of taming this sort of boiler for general use.

These flash boilers have intrigued me and I had my eye on doing one with a figure of 8 of tube as I'd seen an example in the KN Harris book. I also see that fsheslop's boiler was figure of 8. In fact I'd already done the drawing and worked out the surface area when the Vital Byte guy confirmed that my guesstimates matched his design.

And by way of making it real, I've already made the water pump and had it running under PIC micro control

This is all going to happen in very slow time as I have the other boiler to finish first. And a conundrum 'is it a bird, or is it a plane.........?" "Is this a boiler or is it software and programming thread?"



Pete

 

[fcheslop]

Just a few questions, do you use a PID prog in the pic not been very well up on programming them is there a basic prog available.What temp do you expect at the boiler, are you using a thermo couple or a thermistor for temp measurement?and Hall sensors for engine and pump rpm this is starting to get very interesting not from the point of building a high speed plant but as a simple monotube plant .The pump design could you use a scotch crank to simplify.Just ideas no criticism intended just starting to think out aloud .

 

[doubletop]

Sorry I'm not sure what you mean by a PID. The PIC micro is by Microchip http://www.microchip.com/ there are numerous devices in their catalogue the key thing is choosing the right one for the job. Which is primarily based on the type of I/O you want to do, simple digital I/O, serial ports, A to D converters, pulse width modulation, counters, timers, etc etc the options are nearly endless.

Programming can be done with C compilers, Basic or Assembler. There are probably more but I've always used assembler, that's probably a bit geeky but you have absolute control of the code you produce. Microchip provide a full range of development tools and demo kits. The programming is done on a PC but you need to be able to get the code from the PC to the device. I've build a number of programmers but recently purchased a basic development kit http://www.microchipdirect.com/productsearch.aspx?Keywords=DV164120. It's relatively cheap and has a full tutorial, examples etc. Its certainly the fastest way to get going but there's a few principles that need to be understood but they aren't really a subject for this forum. The PICkit2 has a 16f690 on board and apart from memory capacity I believe would have all the I/O channels required to build this controller. I may have over simplified that but if it isn't right the code is portable to bigger chips with more ports and more memory so it will just be a case of moving up to a different device.

I've no idea how this will go but I'm expecting a boiler pressure of 50psi so a temp of about 140degC I believe. I was considering using a thermistor as I've found a bead type that goes to 300degC but the Vital Byte guy advises using thermocouples. We'll see.

Yes; Hall sensors for rev counters for the engine and pump. Pump motor speed and control of the RC servo (for gas control. valve) will be PWM. Serial port used for the display (basic TTY screen using hyperterminal). All these functions exist on the 16f690

The pump was a bit of a throw together to get me going and see if the motor could do the job. However it does have a torque of 55Kg/cm. My crank throw is 1cm and piston is 1/4" so I'd assume it should be able to pump to 616psi so more than enough. as the stroke is 2cm so 1.26 cc/stroke and the max revs are 140rpm that means it should theoretically pump 1.7litres/min. Which again is probably more than enough water.

So that's it for now really, once I get going I'll look for 8 meters of 1/4" copper tube, which should give me 145sq/ins of heating surface. If that enough or not I don't know but is close to the Vital Byte heating surface

Pete

 

[fcheslop]

Hi Pete, PID is Proportional Band Integral Device it is used for simple examples in a car cruise control or in a domestic heating circuit to enable a constant to be reached .Very simply it looks for a gain say in temp in a given time then uses feed back to adjust itself so that it stays within the set point and stops the temp wildly oscillating around the set point.There is software available for pid control by pic chips.Have a look on WWW.CURIOUSINVENTOR.COM/PROJECTS Or try googling pid control for pic chips theres a load of sites.Apologise offered for over simplifying PID but I didn't want to go into it to deeply.

 

[doubletop]

Interesting; I wasn't aware of PID but thank you for bringing it to my attention. http://en.wikipedia.org/wiki/PID_controller .

I see Microchip have an application note on the subject and source code available. http://ww1.microchip.com/downloads/en/AppNotes/00964A.pdf

It certainly is worth considering. The last time I was involved in this sort of stuff was the inertial navigation systems for the Phantom and Harrier.

(At this point how many on this forum are wondering how far off topic we could really go with this?)

Pete

 

[doubletop]

I've been looking more into this PID stuff and come to the conclusion I've been living under a rock. But its certainly what is required in the two control loops.

 

[flashtwo]

Hi,

I've just found this forum and I'm the "Vital Byte" guy.

The horizontal monotube boiler (a better description than a "flash boiler") has been developed over the last two years to give a very satisfactory performance.

This year the original monotube coil has been fitted into a vertical boiler and, with a new burner arrangement, has improved the boats performance.

I intend to build an Edwardian steam launch, the design of which specifies a vertical rather than an horizontal boiler.

An 7000 word article has been accepted by the British magazine "Engineering in Miniature" and may be published latter this year. The article details the control philosophy with particular reference to the PID mentioned above and the control loops used.

I've been using a Stuart D10 engine and the boiler now has about 240 sq.inches surface area. The engine ticks over at about 5psi and gives continuous performance in excess of 500RPM with a 5 inch, four bladed prop at about 22 psi. The boat can be stopped in the middle of the lake and restarted in reverse if necessary.

Ian.

 

[picclock]

Hi Pete

>>My crank throw is 1cm and piston is 1/4" so I'd assume it should be able to pump to
>>616psi so more than enough. as the stroke is 2cm so 1.26 cc/stroke and the max revs are
>>140rpm that means it should theoretically pump 1.7litres/min. Which again is probably
>> more than enough water.

I'm not sure the numbers compute ???. If the piston is 1/4" diameter (6.25mm) thats an area of 30 square mm, or around 1/3 of a cm2. with a stroke of 2cm's that's 0.6cc per rev, which is only 84cc or 0.8 litres per minute.

Don't know if this helps or not - please feel free to ignore or curse me ;D

Best Regards

picclock

 

[doubletop]

Ian

Welcome aboard. As you can see from this thread, after an initial 'flash' of activity my work around this stopped. Actually I got involved in getting my locomotive running and that has taken all my time until recently. It would be good if you could post a reference to the article when it gets published so those of us on distant shores can source a copy.

For those of you who haven't seen it take a look a Ian's stuff over here on Model Boat Mayhem

http://www.modelboatmayhem.co.uk/forum/index.php?topic=15817.0

picclock

It looks like my numbers got doubled up somewhere along the line. But it didn't get much further than building the pump I'm afraid. Maybe I'll get some inspiration and back to this once Ian's article is published.

Pete

 

[picclock]

Hi Doubletop

What sort of valves were you using in your pump ? Are they homemade or a purchased item ?

One of the reasons I picked up on the size is that one of my projects uses a pump for 1 litre per minute against 120psi. For some reason I find it difficult to design valves that have a good flow rate when open, and yet are small in size. Any details much appreciated.

Best Regards

picclock

 

[doubletop]

Apologies for the delay in replying, I didn't get a notification for some reason.

I scratch built the pump based on a hand pump and I'm not sure what drawings I used. I think it was one of Firebirds but I've just spent 20 mins looking for the thread and can't find it. It was nothing special and I haven't proven the flow capability, this thread is about as far as I got.

Pete

 

[firebird]

Hi picclock

If it was based on my design you can find it here.

http://www.homemodelenginemachinist.com/index.php?topic=2092.120

I built it to test the small boiler I built. It has proved to be very reliable and has been used several times now. Hope this helps.

Cheers

Rich

 

[arnoldb]

I'll second Rich's pump design; works a treat for testing boilers :bow:

Picclock:

I'm not sure the numbers compute. If the piston is 1/4" diameter (6.25mm) thats an area of 30 square mm, or around 1/3 of a cm2. with a stroke of 2cm's that's 0.6cc per rev, which is only 84cc or 0.8 litres per minute.

I agree with the approximate calculations up to 0.6cc per rev (based on the parameters from the earlier thread), but 1cc = 1ml, so 84cc is 0.084 litre.

One of the reasons I picked up on the size is that one of my projects uses a pump for 1 litre per minute against 120psi. For some reason I find it difficult to design valves that have a good flow rate when open, and yet are small in size. Any details much appreciated.

You'll need a bigger pump and higher revs - or a multi-piston pump for that. Have you considered flap valves ?

Regards, Arnold

 

[picclock]

@ arnoldb

My bad - I think my typing fingers are developing Alzheimers ???. Odd thing was I looked at it and thought it must be smaller than that - truly a brainfade moment.

@ Firebird
The pump I'm currently using has a cylinder housing a squarish plate with an oring for the seal. The plate fits into a cylinder but its shape is square with rounded corners. as the valve lifts the fluid flows past the oring and the gap between the square plate sides and the cylinder. It also has a stainless steel spring which acts to keep the valve closed when there is no pressure differential.

I don't think this would be classed as a flap valve as the whole body of the plate is free to move. If I remember correctly, I think the problem of having a hinge mechanism inside the pump body made this too difficult to engineer. However it may be possible to rethink the construction to overcome this limitation on the Mk2.

btw in testing I got the pump up to 900 psi before I chickened out (outlet blocked so no flow).

Thanks for the ideas and comments.

Best Regards

picclock