Model Boat Mayhem

Please login or register.

Login with username, password and session length.
Pages: 1 2 3 4 5 6 7 8 9 [10] 11 12   Go Down

Author Topic: Flash steam plant control.  (Read 95772 times)

Xtian29

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 511
Re: Flash steam plant control.
« Reply #225 on: February 28, 2012, 07:30:38 PM »

The second and final part of the article is in the March issue of Engineering in miniature, and is on the shelves now.....Don't miss it!
Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #226 on: March 02, 2012, 09:02:32 PM »

Hi Mike,

Thanks for sharing the gas valve photos. I've just replaced my gas valve servo (£7 type) with a more expensive all metal gear and spline type (£!), because the plastic spline of the original servo was getting chewed up with the clamping grub screws.

I used to set the closed position mechanically by loosening the grub screws and rotating the gas valve shaft. Now I can set it accurately by setting  values on the display screen, which nominally set the fully open position to 1000us (PWM) and fully shut to 2000us. If the gas valve hasn’t fully shut then I can increase the 2000us value to say 2100us and the servo travels a bit more to shut the valve – the controller remembers this position even when powered off.

I’ve been carrying out some pump flow tests today, since I thought the “A” pump wasn’t performing well. I replaced the steam pipe between the boiler outlet manifold and the engine with one containing a flow control valve with the outlet going to a laboratory 500ml measuring cylinder.

I measured the pump flow over a three minute period with and without constrained flow  - the constrained flow made the pump work against a pressure of about 50 to 200kPa (7 to 30psi).

I found that the displayed flows (“A” &”B” pumps) were reading about 7% lower than that recorded by the measuring cylinder. This can be corrected by amending the pump capacity in the controller.

I’ve recently rebuilt the timing linkages of the D10 engine, since the wear on the original quadrants was making accurate timing impossible. New quadrant bronze castings were obtained  from Stuarts. This time I made a jig to fit on the milling machine’s rotary table, which enabled me to mill the quadrants to the exact radius without resorting to tiresome filing. I also found that I had limited the travel of one of the quadrants when I had originally built the engine 4 years ago – I repinned the linkage to the correct position.

I also tightened up the eccentric straps to reduced the slack.

With both the above overhauls completed, the valve timing setting was greatly improved.

With the above sorted out, I gave Vital Byte a run in the test pool with very good results.

From a low last July of 380RPM (max) with flames leaping out of the stack and very hot boiler casing  -  to a performance today of 590 - 620RPM (peaking at 710RPM) with no flames leaping out the stack – in fact I could place my hand over the stack, on the stack and on the casing without it being unduly hot, indicating a very efficient boiler.

The flaming problem, as discussed in earlier posts, was caused by the mesh in the burner bowl corroding up and limiting the airflow into the boiler.

Ian.
Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #227 on: March 10, 2012, 01:27:39 PM »

Hi Mayhemers,

Here’s a series of photographs show the latest development of “Vital Byte”.

The first couple show the jig made for milling the timing quadrants (octants?) more accurately and quicker than by filing. The dimensions ensure that the exact radius is kept around the piece and a 4mm end miller cuts the correct dimensions.

In the other photographs the new compact water filled gas heat exchanger is shown with the reconstructed burner and the vertical boiler.

From the burner photograph you can see that the liquid fuel enters a the union furthest away, does 3 ½ turns around the burner in 1/8th copper pipe and gas exits at the tee-piece where some is tapped off back to the three pilot lights spread around the burner, which also heats the liquid fuel evaporation coils.

Above the burner is some 1.8mm pitch stainless mesh, which is surviving better than the scroll of mesh originally in the burner bowl. The mesh ensures that the flame doesn’t propagate to the nozzle.

In the previous setup, the exit gas pipe lay in an open trough of water to cool the gas before it entered the servo gas valve, which had an “o”-ring seal and a lead soft seat, both of which I didn’t want to overheat.

The gas exiting the gas valve, when heavily throttled, can come out freezing, so that pipe was put through the same trough to warm up before going to the burner nozzle.

The trough had to be filled with water before running the boat, so a replacement design was installed, which was connected to the incoming feed water from the lake. This design would automatically fill when the boat was floated with the vertical silicon tube indicating that it was full.

With this arrangement, the problems of cold weather operation are eliminated with no gas evaporation taking place in the upside-down cylinder and also all the fuel is used from the cylinder. One has to be aware of any fuel residues blocking the gas valve, so I allow a full flow blast of gas through the valve before reshutting the valve, and allowing the gas to disperse before igniting the pilot lights.

Looking forward to Wickers, and special thanks to the Wickstead Model Boat Club and Martin for all their efforts.

Ian.
Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #228 on: March 23, 2012, 12:21:16 PM »

Hi Mayhemers,

Subsequent to the feed pump tests, I've made some additions to the Display menu (see photograph) and software.

I can now update the pumps' capacity in mm^3 to give a more accurate display, and can also swop over which is the Master Pump without having to swop the ESC and shaft pickup connectors on the interface board.

From the menu I can choose whether the "A" or "B" pump will the Master with the other acting as Stand-by.

"Vital Byte" is getting ready for Wickers - hopefully with a fresh coat of paint!

Ian

Logged

essex2visuvesi

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 6,134
  • Location: Finland, England, Finland!
Re: Flash steam plant control.
« Reply #229 on: March 23, 2012, 12:45:24 PM »

We must all bow down before the almighty B pump!



It is written that it shall be master






Sorry slow afternoon in the office  :embarrassed:
Logged
One By One The Penguins Steal My Sanity
Proud member of the OAM  (Order of the Armchair Modeller)
Junior member of the OGG  (Order of the Grumpy Git)

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #230 on: March 29, 2012, 09:48:00 AM »

Well the "Almighty B pump" wasn't to be bowed down to!

The non-return valves needed to be re-seated, since they were letting by (see "simple feed flow indicator" thread).

Over the last year I have had problems with the gas control valve blocking up causing a much reduced performance of the boiler. I had been lead to believe that using the gas cylinders upside down would allow waxy deposits from the cylinder to pass through the system and eventually blocking the valve. Well, I've dismantled the valve several times and poked a fine wire through to open up the valve and, although I couldn't see any waxy deposit, the valve did clear.

After only one or two runs the valve would block again. This time I opened up an empty gas cylinder to find the "fabled" waxy deposit only to discover the inside of the cylinder was totally clean with not a hint of the "wax".

The actual valve seat is made of lead with a small needle size hole. This soft valve seat, I suspected, had been deformed, over time, by the valve plug and being repeatably partially closed again after I had cleared it with a fine wire. This time I drilled out the hole to 1mm dia. and so far the blockage has not re-occurred and the performance is the best for ages. I think the recently fitted high torque servo may have contributed  to this deformation of the seat.

Another reduction in performance was caused by a sliver of silver solder that had travelled down the gas pipe to the nozzle causing throttling and frosting up at the nozzle. If ever you see frosting up it seems to indicate a partial blockage.

The gas cooler arrangement has been changed with the feed flow being drawn through the cooler as the gas was getting too hot at the valve. I think I will have to reduce the length of the evaporator pipe around the burner as it is heating the gas too much.

Ian
Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #231 on: April 14, 2012, 10:55:12 PM »

Hi Mayhemers,

Having solved a few things that were restricting the power output (engine timing and fuel gas obstructions), I've been experimenting with some of the control system settings.

From previous posts and the magazine article, you may know that the steam temperature desired value is offset by the feed demand, i.e. the higher the feed demand the higher the target steam temperature.

Well today I've run the plant with both pumps running at full speed (170cc/min indicated), with a steam temperature at 145degC (controlled!) and achieved a remarkable 720RPM running (not peak) and steam pressure at 300kPa (45psi). You can imagine that, with the 5 inch prop, the water was quite turbulent especially slamming into astern then ahead.

Normally I've been running with a single pump and getting 530 or so RPM with a gas consumption of about 4.5grams/min, which would be more than adequate for the model boat. I think 720RPM would be showing off! Perhaps the boiler is now oversized at 250sq.inches for the engine.

By the way, are there any questions from those who have read the magazine article (its gone rather quiet)?

Ian
Logged

HS93 (RIP)

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 2,949
  • I cannot spell , tough
  • Location: Rainhill UK
Re: Flash steam plant control.
« Reply #232 on: April 15, 2012, 12:47:37 AM »

the Post has been going on for some time so is there any chance of a run down of what the system now does in the finished state as I think from day one it has changed so much and it would be nice to know what is possable with steam in the modern world, thanks forthe informative Post.

Peter
Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #233 on: April 15, 2012, 02:26:03 PM »

Here is a breakdown of the current situation:-

Engine: Stuart D10 double cylinder, double acting, ¾” bore by ¾” stroke, driving the
             Propshop’s 5 inch prop designed for the D10.

Boiler: Vertical Monotube, 50foot of 3/16 inch copper brake pipe wound in three
             layers.
Boiler dimensions: diameter 108mm, Height 270mm plus 190mm stack.

Electric pumps (2 off): Feed rate 85cc/min each. Filtered Pond water is used.

Fuel: Butane/Propane in liquid form with evaporator on burner.
         The heated gas is attemperated by a cooler fed with incoming feed water.

Burner: Jet 0.7mm dia., mixing tube ¾” dia., 100mm long, Stainless bowl 1¼ “ dia. with 1.2mm pitch mesh to prevent flashback to the jet..

Fuel consumption 4.5g/min at 530RPM.

Currently achieving 720RPM at 170cc/min feed and 45psi, 145degC steam conditions.

The exhaust steam is directed to an oil separator / feed heater (dia.50mm, height 225mm plus 108mm high vent pipe). The oil free water is dumped overboard.

Control system:
PIC based microcontrollers monitoring the feed pumps’ strokes per minute, steam temperature and pressure and the gas cylinder pressure.

The demanded feed flow is set at the radio transmitter and the controller, via ESCs, ensures that the pumps maintain the flow independent of the boiler pressure.

The steam temperature has a manual setting that is amended by the demanded feed flow, such that, as the feed is increased, the desired temperature is raised, which in turn causes an increased firing rate via a servo gas valve, raising more steam and an increase in speed
–   put simply more water in, more steam out and the faster the boat goes.

The controller has a start up mode such that the boiler is warmed up at a optimum rate to the boiling point without over firing and causing excessively high temperatures. Being a monotube boiler, it takes only a few minutes to raise pressure.

There are a few safety features: high temperature shutdown of the gas valve, high pressure shutdown of the gas valve and pumps, pump failure detection and standby pump cut-in. The boat automatically restarts itself if any of the above occurs.

Current problem: the stainless steel burner mesh crumbles away after about 5 hours running and is to be replaced by a ni-chrome mesh once I’ve woven it!

Was it worth it? The initial objective was to build a simple cheap boiler that didn’t require insurance or annual testing. Well the boiler is certainly cheap at £60 (£20 copper brake pipe, £20 toilet brush holder boiler casing, £10 brass stack, £10 sundries) perhaps a tenth of the price of a conventional boiler. Certainly it is very simple to make with only two silver soldered connections, plus one joining the two 25 ft lengths of copper. Not being a pressure vessel it is exempt from the test and insurance requirements.

At the project start, the best method of controlling the monotube boiler was a matter of debate and using a programmable system enabled many ideas to be explored without massive amounts of rebuilding mechanical and/or electronic controls. One thing you don’t have to worry about is the boiler water level!

So we have a boiler that is cheap, quick to start and runs as long as the fuel lasts (100minutes or more).

The controller hardware bits came to about £70 and took around a week to solder together, although it is transferable between models or even a full size boat.

The software is a totally a different kettle of fish, with thousands of lines of code, but, since completed, can be downloaded to the raw controller in minutes.

It took 2 years of experimenting (including pump and boat construction) to arrive at a practical solution with “Vital Byte” having it maiden voyage at Wicksteed in May 2009.

Where to now? I’ve got Selway Fisher plans of a 30ft steam launch and I’m just scaling them for the 5 inch prop; that determines that I build a 75inch launch to install the existing system in.

Now that I’ve broken the back of the problem, it would nice to know that other enthusiasts will take up the challenge and build a similar (perhaps better) monotube boiler system.


Ian

 


Logged

dreadnought72

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 1,898
  • Wood butcher with ten thumbs
  • Location: Airdrie, Scotland
Re: Flash steam plant control.
« Reply #234 on: April 15, 2012, 02:55:10 PM »

Ian, I'm very impressed with the set-up - but could you explain more about this line (maybe it lost something in your post):

Quote
There are a few safety features: high temperature shutdown of the gas valve, high pressure shutdown of the gas valve and pumps, pump failure detection and standby pump cut-in. The boat automatically restarts itself if any of the above occurs.

Is that the right thing to do? If a fault is detected, do you want it to restart with the fault not fixed?

Andy
Logged
Enjoying every minute sailing W9465 Mertensia

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #235 on: April 16, 2012, 05:34:20 PM »

Sorry for the confusion Andy,

During development and use, I adjust the response of the control loops so that they respond either sluggishly or lively to changes in feed flow command. If the controls are too lively then, on occasions, the steam temperature can get out of control and the system shuts down to protect the engine. The boat then sits on the pond and waits for things to cool down before it restarts itself and I can bring it back to shore. If it keeps happening, I then "detune" the temperature control loop to ensure only gradual temperature changes.

Actually discussing this has usefully reminded me (and duly entered in the boat's log) to put back the function of setting both pumps to full speed in case the gas valve jams in the open position; I commented it out of the software last week when I was doing some experimenting - a fateful mistake!

It was very amusing, when "Vital Byte" was steaming down at the Dover Model Boat Club's pond and it did shut down due to high temperature. The lads there were very keen to send out their electric tugs on a rescue mission, but I told them that the boat was in "auto-recovery" mode and would return (fingers crossed), which it did to their amazement (and mine!).

The high steam pressure shut down (which has never occurred) is there in case the prop fouls up with debris and stops turning.

I suppose the priority is to return to shore.

Thanks for your compliments and interest.

Ian.

Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #236 on: April 17, 2012, 04:40:58 PM »

Here are couple of photographs showing the crumbling stainless steel burner mesh and the weaving loom for making a ni-chrome mesh.

The weaving loom has two brass guides each with 60x 0.5mm grooves at 1.2mm pitch, mounted in a steel frame. I did consider drilling 120 0.5mm holes in a brass plate, but then thought the better of it.

The warp is wound around the guides. The two "tangs" are for mounting firmly in a vice as the weft is woven through the warp.

Once the ni-chrome is woven, it will be placed on a former and bent into a slightly domed shape in order to slide into the burner cup.

I can see this is going to try my patience!

Ian.
Logged

boneash

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 187
  • In an actual war zone there are 2 types of ships,
  • Location: Kent UK
Re: Flash steam plant control.
« Reply #237 on: April 17, 2012, 05:08:05 PM »

Quote
I did consider drilling 120 0.5mm holes in a brass plate, but then thought the better of it.

Oh come on now a milling machine, you have, and a Pic, 3 motors and a CAD package......

or  a xyz package off Ebay.....
Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #238 on: April 17, 2012, 06:40:13 PM »

ouch!

I don't want to wear out the milling machine, you might want to use it one day!
Logged

Mad_Mike

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 2,597
  • Lincolnshire
  • Location: Lincolnshire
Logged

HS93 (RIP)

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 2,949
  • I cannot spell , tough
  • Location: Rainhill UK
Re: Flash steam plant control.
« Reply #240 on: April 18, 2012, 07:31:50 AM »

how about  Bruce steam models they do some stainless steel for spark arestors will that work

Peter

Peter
Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #241 on: April 18, 2012, 01:57:20 PM »

Hi,

Thanks Mike for that suggestion - never knew they existed. My burner bowl is about 32mm diameter, so I'll have to make my own. I'm now thinking of a use for the pipe bowl filters.

Peter -  the stainless steel ones work fine until they oxidise due to the burner flame sitting just above it, especially at low gas flows. The ni-chrome wire, which is normally used for electric fires, should not oxidise and last much longer - I'll keep you posted with its performance.

In the mean time, I've found a new method of weaving called the "Scotty" method that enables you to finish before you've started - many an enterprise use it apparently, even up to warp speed 9! (yawn!)

Ian
Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #242 on: April 22, 2012, 06:16:14 PM »

Hi Mayhemers,

Now that I'm using both pumps to achieve "ramming" speeds (700RPM), the previous control philosophy to having one pump purely as the standby has changed somewhat.

I've made some software changes so that, if either pump stops reciprocating or there is a high steam temperature trip (possibly caused by a feed pump problem), the Feed demand is automatically halved. This ensures that the boiler feed requirement can be handled by one pump.

Another change is to now allow 10 seconds for the pumps to re-establish flow, after a trip, before the gas valve starts to open.

The steam temperature target, which is a function of the feed demand, has been further refined by allowing the equation parameters (offset and slope) to be set at the display and command unit whilst the boiler is running. (For the mathematical it is the old "y = mx + c" equation and I can now adjust m and c whilst running). This will allow fine tuning to take place.

....and not a soldering iron in sight!


Ian
Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #243 on: April 24, 2012, 07:56:43 PM »

Just to give Mayhemers a flavour of what is going on, here is a small part of the code that has just been modified for the new method of running with two pumps (without the use of the soldering iron!...)


TEST_A_RUNNING   
      BTFSC   C_FLAGS_6,1   ; (BTFSC) Times up! Is the "A" Pump running?
      GOTO   ENB_GAS_VALVE   ; Yes, the "A" Pump is running, therefore enable the gas valve to open.

      BSF   C_FLAGS_1,7   ; The "A" Pump is not running, therefore set the "B" Pump as Control Master.
      BSF   C_FLAGS_1,3   ; Set because one of the pumps is unavailable and the target T4_DV must be lowered.
      BSF   C_FLAGS_3,0   ; Set the "A" pump failed flag.

      MOVLW   D'100'      ; Set the pump established counter to 10seconds (100d) to allow the pump to start turning.
      MOVWF   PUMP_COUNT
      GOTO   DIS_GAS_VV              ; Disable gas valve from opening.

Ian
Logged

steamboatmodel

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 333
  • Location: Toronto, Canada
Re: Flash steam plant control.
« Reply #244 on: April 27, 2012, 01:32:12 AM »

Ian 30 years ago I mite have understood it better, but I can sort of follow it through. When I my round tuit I am going to have to sit down and do more than glance at embedded C.
Regards,
Gerald.
Logged
Be wary of strong drink. It can make you shoot at tax collectors--and miss. Lazarus Long

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #245 on: May 04, 2012, 05:02:08 PM »

Hi Mayhemers,

Since the gas control valve was drilled out to prevent it getting blocked, the valve characteristic improved such that the performance increased dramatically as previously reported.

With the increase gas valve response came a problem of overfiring when starting up resulting in high steam temperature trips. Apparently the rate of heating the boiler on start-up had less to do with my software controlling the heat input and more to do with the gas valve itself previously limiting the heat rate. The original code would set a target temperature value, wait for the actual steam temperature to catch up and then increase the target temperature to a new value.

To get around this problem a proper piece of software code was installed that enables the target temperature to be ramped up at a set rate. Although the code waits for the temperature to catch up, it won’t increase the target until a timer has expired. The start-up time, from ambient to boiling, can now be set  from anytime between 1 minute and 16 minutes.

Having now sorted out these remaining problems, “Vital Byte” has just had a pre-Wickers run down at Herne Bay with SWMBO taking a few photographs; she apologises for the stern shot since the boat was going “too fast”! Quite a strong hull wave was present.

“Vital Byte” performed without fault throughout her power range (100 to 600RPM) and could be stopped and restarted in the middle of the pond with ease. Going astern – no problem. Gas consumption 5g/min.

She (“Vital Byte” not SWMBO) is now going into dry dock for a repaint and final preparation before Wickers.

I suppose I’ll have to build the proper boat now that the controls are fully functioning!

Ian
Logged

steamboatmodel

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 333
  • Location: Toronto, Canada
Re: Flash steam plant control.
« Reply #246 on: May 06, 2012, 12:56:36 AM »

Looks great.
"Having now sorted out these remaining problems, “Vital Byte” has just had a pre-Wickers run down at Herne Bay with SWMBO taking a few photographs; she apologises for the stern shot since the boat was going “too fast”! Quite a strong hull wave was present."
It looks like a classic case of here she comes there she goes and by the time the brain gets the message to the hand she is gone.
Regards,
Gerald.

Logged
Be wary of strong drink. It can make you shoot at tax collectors--and miss. Lazarus Long

Panoramix

  • Shipmate
  • *
  • Offline Offline
  • Posts: 3
Re: Flash steam plant control.
« Reply #247 on: May 17, 2012, 02:34:08 PM »

Thanks Phil, that's much better.

It looks like one trough behind the bow followed by one crest before the prop, i.e. one wave length.

For those not familiar with the boat, there is a large overhang above the water after the prop (see earlier posts at last year's Guildford Show), which gives the appearance of a much longer hull.

Now, is the hull speed indicated by one wave length or one half wave length?

Ian

I was in admiration when I've read this and seen all the work and couldn't resist registering!

 I don't think that anybody has answered the question and although it is an old one you might be interested in the graph below:


So basically to be at hull speed you need to have the crest of the first wave following the bow wave lined up with the stern (4th sketch from the left for Fn=0.4). I think that in your case there is still some speed to be gained before reaching hull speed. The distance between crests is proportional to the square of the boat speed, so roughly you are looking at 40% extra speed. If there is enough torque on the shaft a propeller with 40-50% more pitch would yield this.
Logged

flashtwo

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 521
  • Location: Sevenoaks, Kent, England
Re: Flash steam plant control.
« Reply #248 on: August 13, 2012, 07:34:06 PM »

Thanks Panoramix for the diagram - "Vital Byte" still appears to be somewhat short of achieving "hull speed", although its square hull section may be offering a lot of resistance.

I renewed the piston 'o'-rings last week (since the performance wasn’t up to scratch) before steaming her at yesterday at the Model Boat Association Dover annual meeting at Kearsney Abbey.

The weather was hot (first outing this year when it hasn't poured with rain!) and the gas cylinder pressure was up at 5bar (75psi) due to boat inboard temperature being at 40degC. On a cooler day the pressure is about 3bar.

The high gas pressure caused a significant amount of steam to be raised purely on the pilot flames causing the engine to tick along at 300RPM. What with the high gas pressure and new 'o'-rings, the boat's performance was the best ever with an average of 480 RPM for the first half-hour run and an average of 532 RPM for the second half-hour run. Previously the averages were in the region of 380 RPM. Maximum RPMs were off-scale >800RPM (RPM is scaled 0-800 in the software).

The hull wave still consisted of the bow, mid and stern components, but greatly enhanced with a very deep dip just before the stern wave – any deeper and I think the prop would be affected.

“Vital Byte”  gave a very powerful performance and no flames out of the stack this time! Fuel consumption was 4.3g/min – excellent considering the thrashing it was given – this would give about 1 hour 45 minutes per 460g gas cylinder and perhaps 3 hours just on tick-over.

In the mean time the building jig for “Vital Byte II” has been completed with the keel under construction.

Ian
Logged

TurboTyne

  • Full Mayhemer
  • *****
  • Offline Offline
  • Posts: 113
  • Location: Newcastle upon Tyne, England
Re: Flash steam plant control.
« Reply #249 on: September 21, 2012, 01:41:19 PM »

Hi Ian

It's a while since I've looked at this thread. For some reason not known to me, I stopped getting notifications of new postings so I'd assumed there were no new messages.



Now that I've broken the back of the problem, it would nice to know that other enthusiasts will take up the challenge and build a similar (perhaps better) monotube boiler system.



I just wanted to say, in response to your comment quoted above, that there is at least one person who has been inspired by your work, namely yours truly. I am still plugging away at my PIC-controlled boiler. Following your advice, I bought the 100rpm motor and then designed and built a PIC-based speed controller, using the application notes from Microchip. A PIC16F1827 measures the time to complete each revolution of the motor shaft via a Hall-effect sensor as per your VitalByte. It then calculates the speed as rpm and compares to the target rpm. It then uses a PID  feedback control routine to adjust the pulse width modulation output from the PIC.  Actually I had to turn off the D component to avoid speed cycling high/low – again similar to your notes on gas control. This output (at  PWM freq of 10KHz and a 400 step resolution) is sent to the motor via a MOSFET and MOSFET driver.  The target and actual pump speeds are communicated via the I2C bus to a PIC which drives the LCD  display and another which allows input of target speed and values of PID constants etc.   This seems to work O.K as far as I can tell at present. Eventually I plan that the pump control PIC will communicate with the Master PIC that measures temperature and controls the gas valve.  The gas control/temp part is already built and functioning. I'm currently nearing completion of a twin-cylinder feed pump (but only one of them for the present) that is coupled to the motor. So I am looking forward to connecting pump to boiler tube and burner control to see if I have anything approaching a working system.

Regards,  Mike
Logged
Pages: 1 2 3 4 5 6 7 8 9 [10] 11 12   Go Up