Author Topic: Talking Thermodynamics  (Read 68761 times)

Offline MJM460

  • Full Member
  • ****
  • Posts: 528
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #705 on: February 10, 2018, 11:41:52 AM »
Hi Willy, there is more to see in that Thermomotor every time I look at it.  Like many drawings of the time, it does not show the mechanism for the valve actuation.  I guess they were all very secretive, and wanted to hide the critical details from those who would copy their design and sell in competition.  The top cylinder, being a simple air compressor, would work with normal plate valves, with light springs to close them.  But then the power cylinder, the lower one, is an expander and needs valves to be mechanically opened in the same way as a steam engine, though because of the combustion, most of the pressure comes from combustion of fuel, and only enough air is admitted for combustion.  However, it seems to me that still requires active valve opening mechanism.  There seems to be a governor in one of the drawings, but I can't make out how it is linked to the fuel admission valve or air throttle valve.

The heat exchangers come close to true counter current flow, so quite good at heat recovery.  Even the beam is a little unusual to me, in that instead of a solid column, the flimsy looking column seems to rock, and instead of any Watts linkage, the piston end of the beam appears to be guided in a straight line, cross head style.

I don't think our understanding of thermodynamics has changed so much as we have learned lots more stuff.  As I have said before, I am in admiration of these early pioneers who achieved so much with intuition and trial and error.  I suggest lots of error was involved, but matched by their persistence.

I suggested that it was a bit like a gas turbine cycle, though shifting the combustion to inside the expansion cylinder means that most of the pressure comes from combustion, and the engine is not completely reliant on the compressor to provide all the pressure, as it does in the gas turbine.  This means that instead of a large portion of the expander work being used to drive the compressor, leaving only a little, if any, for external work, the high pressure comes from combustion with much less compression work required.  In that respect, it has similarities with the diesel cycle, as you originally suggested, even though it has separate compression and expansion cylinders, and all those heat exchangers to make it a regenerative cycle, in a way I am not sure is possible when compression and expansion occur in the same cylinder.

If we made one today, improvements in materials and pressure vessel technology would almost certainly allow for some more optimisation of operating conditions.  Similarly machining technology and sealing technology allow more accurate construction and less seal friction and leakage.  But I have no idea whether we could make it a huge amount better.  I suggest that the heat exchangers and regenerator would make it heavy and expensive for its power output.  So low fuel costs of our recent past mean there is little economic incentive to go to such lengths in the quest for efficiency.  As fuel costs rise, the economics change and perhaps we will see them reappear.

Well I tried two tests with my modified boiler arrangement.  What a difference!  I don't have any doubt that the water tubes are now working very well.  So much so, that they are producing plenty of steam, well before the water in the main shell reaches 100 degrees.  Clearly things are not in thermal equilibrium.  I tried exchanging the safety valve and filler plug/thermowell locations to put the measurement point over the down comers end, but this made no real difference.  Had trouble with the engine again.  I suspect partly the valve is still a little tight on the nut, and partly because the port face is vertical, Gravity is not helping to seat the valve.  At one point, the valve suddenly seated and the engine took off.  Obviously plenty of steam, and I was only using the small 50 mm burner.  Calculations will come, I hope tomorrow, as will the photographs, which I managed to take, but time ran out on me before I could download and resize them.

I definitely need a regulator valve, so I need to make one of those, and modify the steam pipe to accommodate it.  I think it would be an advantage to be able to retain the pressure high enough to inhibit boiling until the whole boiler capacity is up to temperature.

Similarly, the exhaust separator has not been working too well.  It was gooed up with a very sticky water-oil emulsion.  Clearly I am suffering the penalty for using normal lubricating oil.  I have cleaned it all out, and started using real steam oil, so we will see if that improves things. 

I don't know whether to install a screw in the valve chest cover so I can press the valve onto the seat before I start, or perhaps a light plate spring to hold the valve against the port face.  Any recommendations?

In the spirit of experimentation, not necessarily in gentle equilibrium, though I hope avoiding banging and crashing, I will try one run with the 90 mm burner if time allows, tomorrow or Monday.

Thanks for looking in,

MJM460


The more I learn, the more I find that I still have to learn!

Offline MJM460

  • Full Member
  • ****
  • Posts: 528
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #706 on: February 11, 2018, 10:48:48 AM »
Not the progress I hoped for today, but progress none the less.

Decided to try a run with the larger (90 mm) burner to compare with the 50 mm burner before drawing too many hard conclusions.

It was quite an adventure.  I am always surprised at how much I get out of these tests compared with just watching the engine run, even though that is always the first highlight with a new engine.  But knowing the actual temperatures at each location does say so much more about what is happening.  After the two runs with the 50 mm burner, I was not hopeful about the larger one.  Expected it might be more of the same, with the possibility of being a bit on the wild side.  When the temperature rises too rapidly, it is hard to get all the readings recorded before the next one comes up.  But once again, it was not what I expected.

Before starting, I removed the valve chest cover, now I know why I should have noticed that studs are preferred, (but my next engine was made with studs to hold the valve chest).  Carefully oiled the valve with steam oil, rotated the flywheel a few times to move the valve back and forth, and made sure it was pressed to the valve face, hoping the oil film would stick well enough while I replaced the cover, and tightened the cap screws.  I hope I can eventually solve that problem.  Strange it did not arise in earlier runs.  I am assuming a lot to do with lubrication, and gradual buildup of that sticky goo.

I filled the boiler with the normal quantity of water and the burner with Meths.  (I am starting to doubt the repeatability of the kitchen scale, the first sign that the battery is getting a bit flat.  A frustrating characteristic.  I am thinking of a new scale with 0.1 g resolution, and I hope better repeatability, but not today).

Switched on all the meters, set up the iPad as a timer, lit up the burner, and prepared myself for quick action.  At first nothing seemed to happen.  Then the temperature started to rise, just a little, and I could hear bubbles bursting at the surface.  Nothing dramatic, perhaps 3 or 4 a second and the temperature started to rise steadily in a very predictable manner, so I was able to take some sensible readings.  Nervous as it approached 80 deg, but it continued to near enough to 100 before a some steam leak was apparent.  When it reached 110, I flicked the flywheel and away it went.  The non-contact tachometer gave me readings of 1066 rpm.  I checked some temperatures on the outside of the insulation and the stack temperature.  It rose to over 300 degrees, obviously plenty of scope there for a feed water heater for heat recovery, (if I had a feed pump), or perhaps more area for steam generation.  I took readings of 1160 rpm then 1300 when a sudden noise from the engine let me know things had gone pear shaped.  Unfortunately, the crank pin had unscrewed itself, screwed up the run somewhat.  Nothing went well after that, but I adjusted the valve to let steam blow through and eventually the burner ran out of fuel.

Cool down was quite orderly, and I was able to take readings from 95 all the way down to 35, so more good data to look at on lower temperature differential heat transfer.

Very clearly, with the extra heat available, the water tubes worked as intended, and the whole boiler heated up, more evenly.  An interesting difference from heating with the smaller burner, but very clear on demonstrating the importance of which end the down comers are located.

I took some more temperature readings (with the infra red instrument) for the outside of the insulation.  It was still over 100 degrees over the top of the boiler, though in the 50's low on the sides.  Definitely room for improvement with more insulation.  Another experiment to try in the near future.  I will try and find some rockwool as temporary insulation for a future test.  The future test list is growing, I had better write it down while I remember it.

When I had a closer look at the crank disk, I had marked it out to cut off the usual sections to improve the balance a little, but now I remember that I could not resist running the engine first.  Never got back to do it, or to lock-tight the pin.  Have already hacked at the disk with a hacksaw, ready to mount on the mill for a neater job in the next day or so.  Now I know why it is called a hacksaw!

Then family called, so still have to reduce those photos and input the data for the calculations.  Fortunately the same headings and formula I used for previous runs can be copied down again, so it won't take long when I get a chance to put in the data.  Perhaps after the dentist tomorrow.  Only a check up, I hope.

Thanks for looking in,

MJM460
The more I learn, the more I find that I still have to learn!

Offline steam guy willy

  • Full Member
  • ****
  • Posts: 1701
Re: Talking Thermodynamics
« Reply #707 on: February 11, 2018, 11:51:37 PM »
HI MJM, yes there is always another thing to do with engines !!! Do you take the ambient temperature when firing the engine up and can you take this into account on the spreadsheet ?. As the water is at a higher temp as well as the meths. And does the meths evaporate at a different amount when warmer or cooler, and will it boil at a different rate to H2O ?  I notice you take the highest temp to 110 degrees ...is there something significant with this temp or is this when the safety valve lifts ? Interesting to note that if something is "wrong" then the gauges will let you know. this is why there are so many sensors all brought together on huge ocean going liners....it beats having to go round the engine  with a stick in your ear !!! Also my mum always said don't hack at it "let the tool do the work'...!!

Offline MJM460

  • Full Member
  • ****
  • Posts: 528
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #708 on: February 12, 2018, 11:00:18 AM »
Hi Willy, somebody once said "if you have not measured it, you don't really know it", or something like that.  But even an approximate measurement is better than just a guess.  And changes in measurements are more important than the actual values.  Not to be too hard on guesses though.  A good guess can save a lot of time and effort by helping decide what to measure.  Not really practical to measure everything.

I do measure the ambient temperature, though in what I have done so far, the main influence of ambient temperature is the temperature at which it all starts, so how much heat required to get it to rise through the first temperature interval.  It is always interesting to see that the temperature in the boiler and at the engine inlet differ before any heat is applied.  Only a few degrees and I don't really know what it means.  Certainly, the water is nearly always cooler than air temperature, and at that low temperature difference, it takes a really long tome to equalise.  I usually run out of patience and just wait until the readings seem stable.  However, now I am measuring the stack temperature.  This leads to the interesting possibility of estimating the temperature of the hot gases resulting from from combustion and possibly the air flow rate by a further application of that heat balance.  I will come back to that in a future post after I have tried some calculations.

The other main application of the ambient temperature measurement is in highlighting the differences in experimental conditions.  For example, in your electric boiler, the heat loss to the atmosphere is determined by the difference between the water temperature and the air temperature, along with the insulation thickness and properties.  I suspect it is possible to compensate for the difference in the temperature difference in some of the calculations to make them comparable whether you conduct the experiment in high summer or the depths of winter, though I have not carried that through so far.  Extra insulation reduces the loss to atmosphere,  hence reduces the importance of the atmospheric temperature.

Boiling ethanol (Meths) is similar in principal to boiling of water, though the temperatures and latent heats are different.  So, at atmospheric pressure, ethanol boils at 70.3 deg C, and the latent heat to boil it is 840 kJ/kg, compared with 2257 kJ/kg for water.  Working out the calorific value for Meths as a fuel requires allowance for the sensible heat and latent heat to evaporate it.  It looks like this is allowed for in the published data, so you don't have to do it again.  So when the ambient temperature is higher, it takes less heat to raise the fuel to its evaporation temperature, but in practice, this makes very little difference as the heat release of 26,000 kJ/kg is very large compared with the differences in sensible heat requirement, unless you are conducting a very accurately controlled laboratory experiment.

The 110 degrees is a bit arbitrary.  It is only equivalent to about 6 psig, but is enough to run the unloaded engine at 1000 - 1500 rpm, so it is adequate.  Without a regulator valve I have very little choice, but by this stage, the steam generation is enough to mean that the heat up calculation assumption of no work done is no longer valid.  So to is an appropriate point to stop.  With the engine running, the temperature varies up and down a little but when I look closely at the steam tables, it makes little difference to the heat required for each kg of steam.  It does not move far from that 110.  The difference in latent heat is compensated for by the heat up or cool down heat associated with any temperature change.  So I just assume the steam is generated at the average rate at a steady 110 degrees.  A bit rough, but close enough to help our understanding.  Steaming calculations are based on a steam rate that is approximate, but the realistic order of magnitude.  Obviously I had to ignore most of the result on those two runs with the boiler the wrong way around.  When there is obviously steam produced, while the boiler temperature reading is only 85 - 90, you can't place any reliability on the calculations.  The safety valve is set much higher, in line with the boiler design pressure.   I did test it way back, and I should have a record of it somewhere.  I lift the stem to ensure it is not stuck before each run, but without a regulator, I cannot raise enough pressure to lift it.  Another project to add to the list.

Not only ocean liners have all those instruments but also chemical plants and power stations.  Not to mention jet aircraft.  They used to even put a few key instruments in cars.  However, they assumed we were all to dumb to know what they meant, so proved that no one read them by leaving them off.  Another stupid economy perhaps?

Ok, I asked for the comment about the hacksaw.  Actually, the problem was that while I usually have good success with a timber file guide, the guide slipped on the small component so I finished the cut without it.  I should know not to do that by now.  Even a violin apprenticeship would not give me the eye/hand coordination to do that.

However, today I pressed on, set up the rotary table in the mill and had a go at neatening up the cut.  Actually made some swarf as you can see in the attached photos.  Had to reintroduce myself to the mill as it is the first swarf I have made since I started this thread.  The modified crank "disk" is now fitted with the crank pin, and will be good to go tomorrow.  I carefully cleaned the threads of oil and applied some loctite to lock the threads.  Had to move the whole shaft to ensure I did not get loctite in the bearings, so will have to retime the engine tomorrow.  That motivated me to make those studs at long last.

Last picture shows it all set up, almost ready to go.  You can see the simple bent wire clip I made to hold a sheathed thermocouple in a fixed position in the stack.  The insulation does not look too bad from that angle.

Started putting the data into the calculations, and will continue that tomorrow.

Thanks for looking in,

MJM460

The more I learn, the more I find that I still have to learn!

Offline paul gough

  • Full Member
  • ****
  • Posts: 366
  • Tropical Queensland, Australia
Re: Talking Thermodynamics
« Reply #709 on: February 12, 2018, 11:31:25 AM »
Hi MJM, Your adventure with the little boiler is proving to be very interesting, and demonstrates that even this simple steam generator is not just a pot of water with a flame under it, it is a system and each of the systems components have to work effectively. For this to be achieved sufficient understanding of the behaviour of each component and of the whole system needs to be understood if optimal conditions are to prevail. When a teenager I came across this type of boiler in model form and thought I knew how it worked, but did not fully understand the full implications of tube design/arrangement until a little later in the year when I was 'playing' with a friends 6"x8" single cylinder stationary engine and steaming its companion boiler. It was an ex- army cooks field kitchen boiler of essentially the same design with two under tubes and about 15" dia. barrel. But, the long leg or down comer was reversed to what I understood to be the 'normal' position, i.e. normally  rising away from the front of the boiler. Something my mate nor I grasped immediately.
All was revealed when we decided to wash out the boiler which had been standing for a long time and conveniently had a handhole at the front and allowed a good view of the proximal tube openings in the bottom of the barrel from the short riser. These tubes were still completely full of water as they had no provision for a drain. So we 15 year old first year apprentices decided to dry them out with a small fire. That was something of a revelation, it instantly revealed the flow direction with water oozing and then flowing out of the tubes at the front end. We immediately grasped the reason for the 'reversed' tube slope in our 'light bulb' moment. Most of the heat from the fire was always in the front two thirds of the firebox and the radiant heat in the front half, the rear end being a sort of dead spot and so relatively cold. My first lesson in boiler design! Ever since then I have had as much interest in boiler design as in engines, perhaps more so, as there are so many factors to consider. I look forward to seeing more data and conclusions from the tests. It might be worthwhile recording feed water temperature and when you finally settle on a standard method of relevant parameters to run a repeat test on a real Melbourne Winter morning and see what differences this environmental parameter has on result. This might reveal something for us Gauge 1 people who operate in the full range of outside conditions. Willy's question about temperature effects on alcohol has got me to thinking whether there is any significant difference in vaporisation rates and hence wick outputs in hot or cold conditions. Regards, Paul.

Offline MJM460

  • Full Member
  • ****
  • Posts: 528
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #710 on: February 13, 2018, 11:29:40 AM »
Hi Paul, good to have you back again.  I assume no longer limited to the phone screen.

I do always record the ambient temperature, though I have previously not had much direct use of it in the calculations.  However, now I have some insulation, so I can assume most of the unaccounted for heat goes up the stack, and I am measuring the stack temperature, I can calculate  the air temperature rise, so I am exploring how I can use that.  In the simple closed system with no feed pump, the main effect of atmospheric temperature is in determining the starting temperature.  For the simple furnace casing with no insulation, it also effects the heat loss from the casing.  That in turn cools the flue gas a little more, so must reduce the temperature difference available to transfer heat to the boiler.  Important with a small burner, but less so with a larger burner where the area is not sufficient to transfer all the heat anyway.

Further to what I said yesterday on the effect of ambient temperature on the Meths, I have noticed that if I remove the burner as soon as it goes out, so I can weigh for any remaining Meths, the burner part is very hot, too hot to place on the plastic scale pan without an insulating protective layer.  I generally zero the scale with a piece of 20 mm thick timber on the pan, on which I place the burner.  However, the tank end, which stays outside the furnace enclosure, is barely more than slightly warm to touch.  So I conclude that most of the heat in the burner comes from the conditions in the firebox, which are obviously much warmer than the outside air temperature.  Also, while my data book does not have a specific heat for ethanol, but water is a whisker over 4 kJ/kg.  So assuming ethanol is a similar magnitude, even with a 20 degree ambient temperature range, 80 kJ/kg is not important beside the 26000 kJ/kg heat of combustion.  So I suspect the Melbourne ambient temperature range has little noticeable effect, other than the burner might be a bit slower to start when everything around it is colder.

Had quite a successful day today.  I finished those studs.  Must have been putting off the trip to the bolt shop for too long, I could only find 3 of the necessary 4 M3 nuts.  Couldn't even find anything to steal one from, so three studs and a capscrew until I get to the shop.  Still, three studs do hold the steam chest in place when the cover is removed for valve timing adjustment.  I was able to get the engine reassembled, so it again looks just like in the previous pictures.  And reset the valve timing.

Now I wanted to run it, to ensure everything was ok again, but I had really run out of time.

One question nagging at the back of my mind was Willy's question about the importance of the height of the boiler above the burner.  I was starting to wonder if the little 50 mm burner was just getting lost in the larger firebox, and if perhaps the flame was barely high enough to reach the boiler shell.  Could that explain the unexpected steaming performance?

I decided to do just a quick run, minimal measurements, just boiler temperature and stack temperature. Measured water and fuel by the volume marks in the containers, and only timed roughly with my wrist watch.  And I lifted the burner by about 20 mm (the top of the burner was previously about 50 mm below the boiler shell).    I wanted only to do enough to see if the burner height made a significant difference, and of course, hopefully to demonstrate that the engine was again a runner.  I hoped to be able to complete this much in the limited time available.

A remarkably successful test.  When I first lit the burner, the boiler temperature took quite a while to start responding.  I started to hear those tell tale bubble bursting sounds, and finally the temperature was rising.  What a difference.  It continued all the way to 100 then 105 when there was clearly steaming happening, a couple of flicks of the flywheel and the engine was away.  Slowly at first, then as the boiler temperature (pressure) rose a little more to 110 to 115 deg C the engine revs increased to around 1000 rpm.  Not quite as steady as previously, possibly the engine is a bit stiff after the extensive rebuild.  As the burner died away, the temperature slowly dropped, and the engine continued to tick over beautifully down to around 300 rpm.  Slowed over a longer time period than previously.  Not sure what that means.  I noticed the superheater was not having any measurable effect.  (Yes I know, I said minimal instrumentation, but it was still set up).  I wonder if I lifted the burner too far.  I will try another quick test tomorrow with a lower packer, then I will try a full test with the height that seems best.

Also made a bit more progress on the calculations, which I will continue tomorrow.

Thanks for looking in,

MJM460
The more I learn, the more I find that I still have to learn!

Offline steam guy willy

  • Full Member
  • ****
  • Posts: 1701
Re: Talking Thermodynamics
« Reply #711 on: February 13, 2018, 11:57:32 PM »
Hi MJM , more interesting stuff going on there and i like the pictures !! If you did use the waste heat for the boiler feed would that reduce the efficiency of the boiler somewhat, or would it be negligible ? perhaps this is measurable as well ?? When the meths burner is alight what shape does the radiant heat take coming off, assuming no draughts. could one see this with an infrared camera ? What is the speed of the radiated heat to atmosphere or is that a silly question?? Are there figures from heating types available for different burners from the makers of commercial and model boilers?  more questions i'm afraid...

Offline MJM460

  • Full Member
  • ****
  • Posts: 528
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #712 on: February 14, 2018, 12:33:02 PM »
Hi Willy,

I am glad you liked the photos.  I will try and lift my game and include a few more pictures, sketches, graphs or something. 

You can see that my machining and set up skills are not up to so many on this forum, but I am enjoying trying and learning.  That very nominal effort at balancing the crank disk made a huge difference.  You can't balance the reciprocating mass of the piston and cross head so easily, but the rotating force from the big end and crank pin also introduces a vibration at right angles to the piston travel.  This part of the vibration is almost totally gone, so that the mass of the baseplate and wooden base is now barely moving.  Clearly that normal shaping of the disk comes very close to what is actually required.  The crank pin is almost at one end of the large wooden base, so introduces a rocking movement, which was previously quite obvious.  The moment of inertia around that horizontal axis is relatively small.  The piston in the horizontal plane has to move the whole mass of the baseplate, engine and boiler, so this results in less response to the forced vibration.  F=m x a is easily transposed to a= F/m, so large mass means small acceleration.

Waste heat is almost by definition heat that would otherwise be rejected to atmosphere, so wasted.  So if you put the feed water heater in the smokebox, or at the base of the stack, you are using heat that would otherwise be wasted, so you improve the overall efficiency with no detrimental effect.  Obviously if you put it in the firebox, with the superheater, you reduce the temperature of the combustion gases which would otherwise be available to generate more steam, so any benefit in higher feed water temperature is offset by reduced steam generation.  Nominally a zero sum game, but the second law says you always loose some.  Similarly, if you recover some heat from the engine exhaust, so long as you don't introduce extra back pressure, the heat recovery is an overall benefit.  In fact you may be able to slightly lower the exhaust back pressure with some condensing, which would be a further benefit, even if limited to a very small difference by the need to remove air from the system.

It would be really interesting to take a picture with an infrared camera.  I used to have a film camera with a special focus mark for infrared, but I never purchased the necessary filter.  I can't remember if special film was also required.  I looked through the settings on my digital camera, and could not find anything.  This evening I tried a photo in the dark garage of a candle as a quick test, but not convinced it showed anything special in the way of an infra red image.  Also searched the Ap store, but everything that looked relevant was just an interface to a special camera.  I am tied up for the next few days, but then I could try taking some photos of the burner say every 30 seconds for five minutes or so so that we can see the flame developing.  That will have to be next week.

Radiated heat, not a silly question, it travels at the speed of light and does not require air in between.  It is electromagnetic radiation, just like light but a longer wavelength.  The amount of heat transmitted from a surface is proportional to the absolute temperature raised to the fourth power, but the cold surface, unless at absolute zero, also radiates back, proportional to the fourth power of its absolute temperature, but at a different wavelength.  So the driving force is the difference in the fourth powers of the absolute temperatures.  The next step in calculating the amount of heat transferred involves absorptivity, emissivity, view factors and so on, so it gets very complicated.

I don't know what data is provided with commercial burners, but if they tell you fuel consumption, you can multiply that by the calorific value of the fuel to get the heat rate.  You use the lower calorific value, as you do not use the heat down to the temperature at which water is condensed, so get no benefit from the latent heat of the water in the flue gases.  If you need data for your fuel, I may be able to provide it.

More questions are always welcome.  I hope I am generally providing satisfactory answers, but I am always willing to have another go if I miss the point of your question.  So please continue asking.

I should have said a little more about burning Meths in the last few posts about the effect of temperature.  One of the reasons that Meths is a good safe fuel for our purpose as it is quite difficult to light.  You cannot easily light vapour at the vapour pressure at normal temperatures, nor can you easily set the liquid alight.  You have to put a match, or electric spark right in a critical zone near the surface, where the fuel air mixture is just right.  This happens quite close to the surface, at a wick, or a free surface.  The heat to continue the vapour production comes from the flame.  In a vaporising burner, you have to supply the heat necessary to evaporate the liquid, and apply the lighter quite close to the vapour jet from the burner, again where the air fuel ratio is just right.   This is not to say that the fire won't flash to the liquid surface if you apply enough heat, and I don't advocate pressing the testing of this too far.  But in a suitable safe area, with a small amount of liquid in a tin lid, you can easily see how close you have to put the match to light the Meths.  If you have enough depth to submerge the match, the liquid will extinguish it, though you will probably light the surface in the process.  If you spill much liquid under a lit Meths stove, it will eventually get enough heat to catch a light, in fact, before the kettle boils, don't ask how I know,   Finally, water is all you need to put out the fire.  The liquid absorbs the water, is cooled to the point where the vapour pressure is too low for continued combustion, and the flame is snuffed out.  This is unlike most liquid fuel fires, where water will spread the fire, so not at all suitable.

Remember, if you must try it, a clear area with no flammable materials, a surface that will contain any liquid spills, and a bucket of water as an extinguisher.  And the fuel bottle capped and put away before getting out the matches.  Remember safety first and always.

I tried the burner at an intermediate height today.  Started to produce steam at 85 degrees again, but after a few seconds, the engine slide valve decided to seal, and the temperature then continued to rise to 100, then on to 110 as normal.  The engine is a little tight after the rebuild and required about 115 degrees to start, but as the burner died down, the engine just slowed but continued to run down to around 105, when it was ticking over beautifully at about 300 rpm.  Then the flame went out with a small pop, and the engine stopped. 

Probably now worth doing a full test at each of those two heights, even at another in between, preferably after I make a stop valve.  Any suggestions for a simple design with an good through passage?  I don't need or even want to regulate, just want to be able to isolate the boiler for safety valve tests and then to control when steam production really starts.

Also continued experimenting with the calculations.  This post is getting awfully long and I have no pictures, so I will tell you more about it tomorrow.

Thanks for looking in,

MJM460
« Last Edit: February 14, 2018, 12:41:13 PM by MJM460 »
The more I learn, the more I find that I still have to learn!

Offline MJM460

  • Full Member
  • ****
  • Posts: 528
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #713 on: February 15, 2018, 11:53:21 AM »
I have often thought there must be a way to use a heat balance to calculate the combustion temperature.  The equation is simple enough, heat flow = mass flow times specific heat times temperature difference.  Unfortunately the equation has two unknowns, so is not sufficient to solve the equation.  Similarly, looking at the flue gas side of the boiler, the heat transferred to the boiler = mass flow times specific heat times temperature difference.  Again two variables, in fact, if we assume the flue gas is very like air, it is mostly air after all, it is the same two variables.  Two equations, two unknowns, can be completely solved, providing the equations are really independent.

If we follow the gas path through, air at ambient temperature is heated to a combustion temperature and absorbs the heat released by the combustion of the fuel.  The temperature difference in this case is the difference between atmospheric temperature and the temperature reached as a result of combustion.  Note that this is not the same as the combustion temperature data available from some sources.  That temperature assumes the exact theoretical air quantity.  Excess air will reduce the maximum temperature reached, as part of the heat is absorbed by the extra air.

The combustion gases then pass over the boiler and heat is transferred to the steam and, in the process, the gas is cooled to the stack temperature.  This time the temperature difference is the difference between the temperature reached in combustion and the stack temperature.
 
Thus, the equations use quite different data, so can be considered independent.  I set up the equations in my trusty spreadsheet, and sure enough was able to solve for the mass flow and the combustion temperature.  And as I know the fuel flow rate, I can also calculate the air fuel ratio.

The process is of course rather simplified.  The combustion gases contain a fair amount of air, but also water vapour and carbon dioxide.  And the mass flow is the sum of the air flow and the fuel flow.  As a first approximation, I assumed the mass flows equal.  As the fuel mass flow is only 2-5%  of the calculated air flow, this approximation is reasonable, though it is a source of a small error. 

The formula also includes the specific heat of the gas, and another approximation, I assumed the flue gas has the same specific heat as air.  I need to have a look at the flue gas composition and the specific heat of each of the components, to check the accuracy of this assumption, but I am making the assumption for the moment.  When I have some better data, I will have a closer look at the likely error introduced by these simplifications.

Finally, I assumed that all the heat of combustion was either absorbed in the copper or the water or steam, or lost up the stack.  In other words, I assumed the heat lost by the boiler casing was zero.  My infrared thermometer reads around 100 degrees on the outside of the casing above the burner location and 70 to 80 degrees towards the stack, so the heat loss is not going to be zero, despite there being some insulation.  However, I can find some rockwool and tie it around the casing and reduce the heat loss to near enough to zero.  If I use only the data from during steam production the casing temperature will be essentially constant, so the heat absorbed by the casing can be ignored.

I used representative numbers from the tests so far, though you will remember that the engine lost its crank pin after some minutes of running, so I did not get a complete satisfactory run after turning the boiler around.  I need to get back to the shed, now the engine has been rebuilt, and do a few runs to get some consistent results.  Then I will produce some graphs

In the lack of really good data, I tested the calculations with some representative figures from previous runs.  I expect the numbers are not reliable enough to post, but I was amazed to see that my simple rig actually provided enough information to calculate the air flow and flame temperature.  That will allow me to experiment with blocking up some of the airflow, particularly the part that enters where it probably just dilutes and cools the flue gases, and does not actually help combustion.  The first very rough figures suggest two to three or more times the theoretical air requirement, so some reduction should be possible, and reduced excess air should result in higher gas temperatures and higher heat transfer to the boiler, so more steam.  More exciting work ahead.

I won't be able to achieve a lot until next week, but in the meantime I will still look in every day, so keep up the discussion.

Thanks for looking in,

MJM460

The more I learn, the more I find that I still have to learn!

Offline paul gough

  • Full Member
  • ****
  • Posts: 366
  • Tropical Queensland, Australia
Re: Talking Thermodynamics
« Reply #714 on: February 15, 2018, 12:29:02 PM »
I will be very interested to see some results and conclusions regarding the air requirements. I fiddled about with this aspect by fitting some metal mesh, (flyscreen), to the bottom of the firebox on my loco and found no observable difference to performance. But, this was really only to see if it choked the fire  and in no way could be considered a real test  nor provided any further suggestions. I feel it did appear to lift the flame a little but could not determine if it was due to increased draught, screen induced turbulence, increased air velocity through the screen or maybe some combination. Looking forward to your tests and thoughts. Regards, Paul Gough.

Offline steam guy willy

  • Full Member
  • ****
  • Posts: 1701
Re: Talking Thermodynamics
« Reply #715 on: February 16, 2018, 01:39:09 AM »
Hi MJM , wondering about the ambient temp ... as heat rises is there a bit of a natural upward draft when it is really warm that can affect heat loss from any of the surfaces. ?? Still trying to absorb all this info btw.!!

Offline MJM460

  • Full Member
  • ****
  • Posts: 528
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #716 on: February 16, 2018, 10:22:31 AM »
Hi Paul, like you, I have tried fiddling with the air vents to reduce the air flow, but was unable to see any obvious difference.  However, there is no doubt that excess air reduces the maximum temperature of the flue gas, and reduces the efficiency of the boiler of any size.  I was only doing a simple test,  time to produce steam and total time of the run, based on weighed in water and fuel.  The little boiler did not have a totally enclosed furnace, so no stack temperature measurement.  I am hoping that with the more detailed tests with measurements during heat up, and also measuring stack temperature, I might be able to detect a real difference.

All furnaces require some excess air in order to achieve complete combustion, and the common wisdom is that Meths requires plenty of air, implying more excess than other fuels.  I am not sure what observations this is based on, however we will see.

I am also thinking of a series of tests of the burner, starting with say 10 g of Meths, then increasing the amount each run by 10 g each test.  My thinking is that the burner seems to start a bit slowly, then more vigorously after a few minutes.  Finally it tails off until it extinguishes.  I am thinking that a 10 g run includes all these stages.  If I add a further 10 g, I assume the extra fuel will extend that middle vigorous phase, so the difference in time will give a fuel consumption for that phase. A few more tests might converge to a uniform incremental rate.  Can you think of any variations to this that might give me a better idea of how the heat rate varies through a run? 

I did think of weighing the burner at intervals, or by modifying the base plate, continually, but I don't think my scale has adequate resolution or repeatability.  So that method will have to wait for a new scale.

Hi Willy, I hope that you are finding the challenge interesting, and similarly the many other silent readers.  The aim is that the knowledge should both be developed and shared, so I hope many are learning something.

I know the conventional wisdom is that "heat rises".  However, this is really quite mis-leading.  Heat travels from a hot area to a cooler place under the influence of the temperature difference.  Gravity does not influence the heat flow.  Now that is the total picture for heat flow by conduction.  Radiation is even more interesting.  A hot object radiates heat, proportional to the fourth power of the absolute temperature, whether there is anything to receive it or not.   However if there is something receiving the radiant heat, it's temperature increases and it radiates back but proportional to the fourth power of its temperature.  So the resulting heat transfer is proportional to the difference in those fourth powers.

It is in convection that Gravity has an influence, but not on the heat.  When a fluid is heated, it expands, which means it's density decreases.  And of course lower density stuff tends to float on higher density stuff, so the heated air rises.  It is the rising air that then transports the heat.  A bit pedantic, I know, but every now and then you meet a situation where assuming that heat rises gives you the wrong answer.  Certainly with radiation, it can definitely flow downwards, and also with conduction.  If you put a hot brick on an ice block, the heat definitely travels down to the ice, just as surely as it flows horizontally if you place the brick beside the ice block.  And of course it flows up if you put the ice block on top of the brick.

With the boiler in an enclosed furnace, the inside of the furnace wall receives heat by radiation and convection within the firebox.  Heat travels through the wall material, which preferably includes a layer of insulation, raising the temperature of the outside surface.  This surface is in contact with air, so the air is heated, and consequently expands.  Also the viscosity reduces, so we have the interface between thermodynamics and fluid mechanics.  The hot air rises, and is replaced by more of the cooler surrounding air.  This process is called convection.  The rate at which heat is transferred by convection is proportional to the difference in temperature between the furnace wall and the surrounding air, which is of course the ambient temperature.  And if you hold your hand above the furnace wall, you will feel the warmer air rising.  So the ambient temperature influence is its effect on that temperature difference.

So if your atmospheric temperature is low, say 5 deg, the temperature difference will be higher than if the ambient temperature is say 30 deg.  This larger temperature difference means more heat transfer, the furnace wall is cooled more by that cooler air. 

But then, more heat loss means more temperature difference between the inside and outside walls of the furnace.  As a result the outside wall temperature is lower with a lower outside temperature.  Of course the higher heat loss also affects conditions inside the firebox, but that had better wait for another day.

You can draw a temperature profile between the inside and the outside air to illustrate all this.  Our internet has defaulted to clockwork performance, I will be lucky to get text through tonight, so I will wait until the wind changes, before I try and post a drawing.

Thanks for following along,

MJM460

The more I learn, the more I find that I still have to learn!

Offline steam guy willy

  • Full Member
  • ****
  • Posts: 1701
Re: Talking Thermodynamics
« Reply #717 on: February 16, 2018, 08:35:32 PM »
I MJM two questions to ponder over Dose the height above sea level dictate how much oxygen there is available in the air so the fuel is less fuelish  ?  foolish question? !!!  and as the meths evaporates does the cooling effect ,affect the performance of the fuel?

Offline MJM460

  • Full Member
  • ****
  • Posts: 528
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #718 on: February 17, 2018, 09:28:47 AM »
Hi, Willy, the height above sea level affects the air pressure and hence density as we saw in the earlier discussion on filling the boiler at the top of the mountain.  It does not affect the fuel quality, though it does affect the boiling point, just as it does for water.  At lower atmospheric pressure, the vapour pressure will match the atmospheric pressure, where boiling starts, at a bit lower temperature.  Just as with water, the latent heat is a bit higher at the lower temperature, but it takes about the same amount less heat to reach that lower boiling point, so not much overall effect.  It does not affect the calorific value, so the fuel is just as fuelish as at sea level.

However, at the higher elevation, lower air pressure, means that a given volume of air has less mass, so less nitrogen but more importantly, less oxygen.  It would be more accurate to say the air is less airish.  You need a greater volume of air to get the same amount of oxygen for complete combustion, as you need at least the same amount of excess air as at sea level to ensure complete combustion.  May even need a touch more.

I believe that accurate laboratory standard data for calorific value uses a measured mass of vapour  with a measured mass of air ignited in a closed calorimeter submerged in liquid in an insulated container.  This gives a calorific value for ethanol.  Methylated spirits is a liquid with 5% water, the highest purity obtainable by simple distillation.  So to get a true calorific value for methylated spirits, you have to allow for only 95% ethanol, heating the ethanol to boiling point, evaporation of the ethanol, heating and evaporating the water.  This results in a slightly smaller value, and probably explains the differences in value quoted from different sources.  I don't have access to my data book for the moment, but I can look it up Tuesday.  But as always, heat released by burning the fuel is only available for generating steam after you have allowed for the heat used to evaporate the fuel.  And of course, at the lowest flue gas temperature that is useful for generating steam, the water vapour in the flue gas both from that initial 5% and from combustion of the hydrogen content of the fuel is still vapour, so again you cannot take credit for the latent heat of this water vapour which is only released when the water condenses.  Calorific value is quoted as two values the higher calorific value or hcv, and the lower calorific value or lcv.  You need to use the lower value for steam generating purposes.  I hope that makes it all a bit clearer.

Well I must have held my tongue at just the right position yesterday when I pressed post, and the text went through, phew!  So I have sketched that temperature profile I talked about  yesterday.  I have assumed the same temperature inside the furnace, though in fact, when atmospheric temperature is lower, the heat loss is greater, so the flue gas in the furnace is a little cooler.

The outside of the furnace wall finds its own value between the flue gas temperature and the ambient temperature so that the heat transfer through the insulated wall is the same as the heat carried away by convection.  You can see the linear profile through the insulated wall, due to a constant thermal conductivity of the insulating material.  The parabolic profile on the air side is due to the velocity profile of the rising air at the wall surface.  No velocity at the wall so higher temperature by conduction through the air, further from the wall the air rises at an increasing velocity, thus replacing the warmed air with cooler air from below, and lowering the air temperature at that distance from the wall.  The change of air velocity with distance from the wall results in that curved part of the temperature profile.  I will try attaching the picture, and see if it goes.  If it does not, I will try again later, or tomorrow.

Thanks for following along,

MJM460
The more I learn, the more I find that I still have to learn!

Offline MJM460

  • Full Member
  • ****
  • Posts: 528
  • Melbourne, Australia
Re: Talking Thermodynamics
« Reply #719 on: February 17, 2018, 09:31:45 AM »
Success.  First time, it timed out before completion, second try, the picture included twice!

No understanding the Internet.

MJM460
The more I learn, the more I find that I still have to learn!