Talking Thermodynamics

[paul gough]

Chris, Injectors are for direct feed, pumps offer less quantity and only work when a loco is moving in the case of axle pumps. This is why very old locos before the advent of injectors had to run up and down the track if they were standing for any time. Later types were driven by other means and independent of loco movement, these later types were often used in conjunction with a feed water heater usually with exhaust steam. Very modern engines even had exhaust steam still having some superheat so as to ensure dry steam and sufficient heat for auxiliaries like feedwaterheaters. On very small locos, e.g Gauge 1, injectors would be difficult to configure and probably somewhat cantankerous in operation, though maybe not impossible, so axle feed pumps are pretty much the sole feed mechanism, other than some manual feed arrangement. Injectors have always been regarded as something of a dark art when it comes to making them, but probably really only a thorough understanding of them combined with extreme care and exactitude in machining and assembly. A good clean injector and properly adjusted should not have much water, if any, discharging from the overflow. Regards Paul.

[crueby]

On the full size Shay my model is based on (Kozo New Shay, 1920s) there was a feed pump on the left side, steam driven not axle driven. Same setup on the ones at Cass I rode on. Any idea why they would use a steam pump vs injectors then? Curious as to the tradeoff.

[paul gough]

Hi again Chris, As far as I was aware Shays had lifting injectors and pumps on the side were for 'taking water' from trackside streams or reservoirs to refill the locos tanks/tender. I should think they were probably used for other purposes as well such as fighting fires along the track. I'm not a Shay expert but I bet there would be someone on this forum who could give a detailed answer covering all the different Shays and whether boiler feed pumps were ever used. Regards Paul.

[Dan Rowe]

Chris,
I have looked at a lot of Shay records and I have never seen a steam pump listed as original equipment. They always had one or two injectors. The device Paul is talking about is known as the siphon. It is a low pressure injector to add water to the tender tank from a pond or stream.

I have seen several photos with a duplex steam pump on the left running board and these were added to make sure there was a way to add water to the boiler. Shays mostly operated remotely in the woods and not all the crews trusted injectors. The crews also favored steam brakes for the locomotive over air brakes because once you run out of air you are in real trouble on a grade. There was always steam for the steam brakes.

Dan

[crueby]

Hi again Chris, As far as I was aware Shays had lifting injectors and pumps on the side were for 'taking water' from trackside streams or reservoirs to refill the locos tanks/tender. I should think they were probably used for other purposes as well such as fighting fires along the track. I'm not a Shay expert but I bet there would be someone on this forum who could give a detailed answer covering all the different Shays and whether boiler feed pumps were ever used. Regards Paul.

Interesting how many combinations there were. On the Lombard they had a pair of injectors to feed the boiler, and a third one to pump water into the saddle tank from streams. At Cass, I could hear the side pump going while it was sitting on the siding waiting to go, no water supply there, so it had to be using it to replenish the boiler. Given all the different manufacturers and models, there was probably every combination used at one time or another.

[10KPete]

And sometimes the water quality was such that strainers couldn't handle the volume of detritus that would plug an injector. A pump has less of a problem with junk.

Injectors need clean water.

Pete

[MJM460]

Hi Paul, you are right on target with time not being the issue, research cannot be done to time or budget.  Pity our politicians do not understand that.  I will be very interested to follow your investigation of those wicks, they are very suitable for these tiny boilers and the small engines I tend to prefer.  If they will do one of your locomotives, surely they will do a small boat.

Hi Chris, adjusting those gas nozzles and air openings seems to be quite critical for clean combustion.  My little centre flue boiler from Marine Steam has a whole page instruction on how to adjust the gas orifice position, and that is after they have done all the work to get the sizes right.  Not to mention matching it to the ceramic burner and the boiler flue.  Not sure about the differences (other than appearance) between the ceramic burners and the poker ones you use.  Your description of the time to heat up, and the stack temperature seems to indicate that the boiler heating surface is adequate, just needs more heat from the boiler.  The hotter stack from the G1 suggests they opt for more heat to maximise the steam from a limited boiler size.  On a larger model, with room for more heating surface, a lower stack temperature from the same burner  means you are getting more steam from the same heat input.  But if you need more steam, the principle holds.  If you want to see this carried to the extreme, check out those flash hydros in Benson and Rayman's book, three pressure fed petrol blow lamps firing one relatively short steam coil.  The boiler flue gas outlet would surely singe your eyebrows.  But there are times when efficiency is not the issue.

Hi Willy, those injectors are really fascinating, and a great illustration of the meaning and use of the energy equation.  The energy for the work done in forcing water into a boiler compared with the energy contained in the steam used would give a true measure of the efficiency.   I suspect it is limited by the energy equations and not easy to optimise, just getting a model injector working is enough.  But they seem to move a lot of water quickly, and don't require the engine to be running as would the alternative of a shaft or axle pump.  You can actually calculate the steam flow as accurately as you know the throat diameter in the steam nozzle and the upstream pressure, and most seem to be rated on water rate.  I have not calculated a figure, just laziness I guess.

Hi Paul, I am another not familiar with the Stephenson name in that context, another interesting historical detail coming out.  But every boiler needs a pressure difference to drive the flow through the tubes.  With coal firing, you need enough draft to allow atmospheric pressure to drive the air through the coal bed.  This almost certainly involves more resistance than those wicks, which also need adequate air.  So a coal bed almost certainly requires more active draft production, steam blown, exhaust blown or one of those induction fans to get it all started.

With a gas burner, the gas vapour pressure in the fuel container means the gas is delivered at high velocity through the burner orifice, which gives energy to transport the flue gas, just like those injectors.

Come to think of it that gas pressure is probably enough to cause sonic flow in the orifice, but as the outlet is generally square edged, or just a little chamfer, sonic velocity is the upper limit.  Perhaps we need to try a 13 deg expanding exit to further increase the velocity, though that would require a very tiny tapered reamer!

I hope that adds a little more to what you were saying in slightly different words.  Once you know where the energy is coming from, to force the flue gas flow in a particular boiler setting, it becomes more obvious how to control it. 

It sounds like your colleagues using gas firing use square ceramic burners, similar the the round one in my centre flue boiler.

Hi again Chris, I am glad that Paul was able to explain the smoke door issue.  If leaving it open a little seems to make no difference, it suggests that the energy mainly comes from the gas supply, and the stack is large enough not to provide undue restriction.  So it is back to the size of your boiler.  Presumably either a mixing problem or an air/fuel ratio problem.  Mixing is about the location of the air holes relative to the fuel jet, while the air fuel ratio is about size of the air holes.  Some burners also require separate primary and secondary air flow.  I am really hoping that one of the burner experts on the forum will join in with more information on this one.

Hi Gas Mantle, there have been magazine articles on blower designs, obviously an area worth a little investigation.  Start simple as you have suggested, and experiment with the nozzle diameter, which would be different for a steam blower to an exhaust blower.  Location and distance from the stack seem to be critical parameters to induce maximum flow.  Don't be sad about all that extra coal being burned, that is where the heat comes from to generate your steam.  It is a sign of success.   I can see your next project will be a star wheel gated feed chute, to save on the shovelling.  That video could be interesting as you say.  There is nothing like a video to help understanding of just what it all looks like in operation.

Hi again Chris, that site has some really good information on full size injectors.  I expect the efficiency is actually determined by the energy equation and the fine detail of the configuration.  But the adjustments on those full size injectors might enable a little tweaking, not so practical in model sizes.  You are allowed all the questions you want.  If there are too many in one day, I might have to make a list and answer over a few days, but I can do that.

I seem to remember someone mentioning that regulations require two independent sources of feed water, it is really important not to run out.  I think an axle pump would actually use less energy, so great when the vehicle is running, but no help when stationary, for a red light or perhaps picking up passengers.  If you get low level due to a pump failure, or from a sudden change in operating conditions, you want water NOW!  And don't worry about the efficiency.  Just get shovelling.

That's about enough for another day, thank you everyone for contributing, and also for just dropping in,

MJM460

[MJM460]

Oops!  I even updated the thread immediately before posting but did not get the usual warning.  I wonder if I was just on the wrong page, as they came in over a long enough time period, so I should have seen them.

My apologies to those who I have not acknowledged, I just did not see them, so I will respond tomorrow.  Thanks for all those contributions.

MJM460

[MJM460]

Thank you Paul, Chris, Dan and Pete for that additional information on the sort of feedwater makeup systems provides on various prototypes.  I am sorry that I did not notice your contributions  when I posted yesterday.  I think that I was on the previous page, and while I refreshed it before posting, I did not get the warning, and did not notice an extra page appearing.  I will keep a better look out for that in future.

In summary, different prototypes had different combinations of injectors, axle pumps and steam driven pumps, and I was interested to learn, even an injector like device designed to lift water out of a creek or dam into the tender, and even for fire fighting was mentioned.  Obviously a mixture of devices increases security of feed water supply, by having a suitable device which will work when the vehicle was stationary, as well as axle driven pumps which very efficiently pump water roughly in proportion to consumption when the vehicle is moving.  Interesting that even the crew's degree of trust  and area of operation influenced what equipment was provided.  Great background for us all.

When I was talking about Chris' boiler on the Lombard, I suggested checking the stack temperature as a means of indicating whether the boiler heat transfer area was adequate for the burner.  I mentioned it more in passing as the reason for concentration on the boiler, but possibly should have made the basis for this simple observation more clear.

A boiler with a centre flue in particular is a relatively simple heat exchanger, where the boiling water  in the boiler operates at constant temperature, while the flue gas starts at the firebox temperature, where the heat released by fuel combustion is taken by the incoming air and combustion products, and cools down as it travels through the flue, transferring heat to the water.  At the Firefox end, the  temperature difference between the flue gas and the boiling water is very large, so the heat transfer equation, q = U x A x delta T gives a high transfer rate because the local temperature difference driving heat transfer at that end is very large.  At the smoke box end, the flue gas is cooler due to the heat transferred to the water on the way through, hence the temperature difference at this end is much smaller.  So as the flue gas proceeds through the flue, not only does it get cooler, but the heat transfer rate becomes slower.  It is very like Willy's coffee cooling experiment.  The flue gas never quite gets to the water temperature, no matter how much area we provide.  The overall average heat transfer is that logarithmic mean rather than just a simple average, and as we previously discussed, the answer is biased towards the higher temperature difference end, where a larger portion of the heat transfer happens.  This logarithmic mean temperature difference, or LMTD, is calculated with the formula
  LMTD = (inlet end delta T - outlet end delta T)/ln(inlet end delta T)/(outlet end delta T)
That "ln" term is short for natural logarithm, or logarithm with a base e, and is available on any scientific calculator or spreadsheet function.

However, if you imagine a very long flue, and just assume that the draft is sufficient, then most of the heat is transferred in the first part of the flue, but as the flue gas moves towards the stack end, less and less heat is transferred, and the temperature change is minimal.  This is when the flue gas will feel barely warm.  Remember that it is above steam temperature until it starts mixing with outside air so don't push your luck with putting your finger down the stack.  A stainless steel sheathed thermocouple is a much better idea.

Now if you consider a shorter flue, the flue does not have enough area for the temperature to get quite so low, so is warmer when it exits the flue and the stack.  On those little gauge 1 locomotives, it actually feels quite hot.  Heat is being lost up the stack.  Without burning any more fuel, you can generate more steam if you could have more area.

On the other hand, Chris observed that on the Lombard, the stack gas was quite cool.  So clearly the heat transfer area was adequate to get the flue gas to an area where the diminishing returns make further area ineffective.  And there is scope to generate more steam by using a bigger burner.

If on the other hand, Chris found that the Lombard boiler stack temperature was hotter than any he had experienced on his other models, we could conclude that the burner is already generating about as much steam as the heat transfer area will achieve.  In that case, we need to look back at that engine to see if enlarging the ports gives enough extra shaft power from the available steam, or do we need to somehow get more heat transfer area into that scale outline boiler.  At the very least, there would be an indication of whether the scale Marion boiler might contain enough area for its scale engines.

So a simple observation, waving our hand over the stack tells us whether the boiler and burner are well matched.  If the stack gas is relatively cool, then by enlarging the burner to burn more fuel, we can force the boiler a bit, and get more steam, which means the pressure can be better maintained with the same engine load.

At the end of the day, modifying the boiler is a big job that most of us would be reluctant to start.  So it is quite comforting that the heating surface in the scale outline Lombard boiler looks like it possibly has capacity to generate more steam with just a bigger burner.   However it is worth looking at the details and see of the boiler for the next model could accommodate more heat transfer area.  And "more" is relative to the boiler size, hence the value in a test to derive a figure such as Harris has used for the type of boiler and burner.

I hope that clarified the basis for that stack temperature test, and hence added to our understanding of our boiler performance.

Thanks for looking in,

MJM460

[crueby]

Great clarification - thanks!

[steam guy willy]

Hi MJM,....With a coal fire you need air above the fire to compleat combustion  ..available through the grate and fire door. so do you also need a similar arrangement with meths or gas burners ??
Willy

[crueby]

Hi MJM,....With a coal fire you need air above the fire to compleat combustion  ..available through the grate and fire door. so do you also need a similar arrangement with meths or gas burners ??
Willy

Thats something I was wondering about with the butane jet in the poker burner style - the commercial G1 locos don't have a seperate air intake aside from the one in the side of the burner next to the gas nozzle, but they don't seal around the burner to the flue well either, not sure if they depend on that gap or if it is just cheaper not to seal it. Don't know about the ceramic style ones, though the wick type would definitely need an air supply.

[MJM460]

Hi Willy, I would like someone with more experience to comment on your statement about needing air above the coal bed.  But if you remember your mother's fire lighting trick with the news paper, there are two things she is achieving.  First, by reducing the air flow, the temperature of the flue gas is higher, so less density which means more draft to draw in more air, more velocity, hence the roar.  But also, she has the gap at the bottom of the fireplace, not the top.  So with the newspaper in place the air is directed into the fuel she has set for the fire, just a bit easier and less likely to lead to hyper-ventilation than blowing to get the fire going.  But I have always understood that the stoking door should normally be closed, and you get enough air to cool the fire and destroy the draft drawing the air through the coal bed if you leave it open too long.

All fuels require adequate oxygen for combustion, whether liquid, gas or solid.  And most have a quite narrow range of air fuel ratio which will actually make a flammable mixture.  Liquid fuels generally have to be vapourised as it is only the vapour that will adequately mix with air and burn.  It is certainly true for Meths, and candle wax, and I suspect this is also true of coal, even though the vapour is only very close to the surface, and not obvious.  If the air is restricted too much, there may not be enough to adequately mix and burn completely.  There might be some unburned fuel, or some that is partly burned to carbon monoxide, the source of the danger in confined spaces.  So to get good mixing, the air is often admitted in two stages, primary air which gives good mixing and allows ignition to occur, and secondary air which admits additional air to ensure complete combustion, which actually requires excess air over the exact chemically correct (stoichiometric) amount.

There are also velocity effects, the flame has a characteristic flame velocity.  The flame will sit where, if it moved closer to the fuel source the gas velocity would carry it away, while if it was further away, the flame easily propagates back to the burner.  That does not look like a very good description.  Look at a soldering burner.  The gas velocity in the jet is way above the flame propagation velocity, so the flame cannot move back into the jet or worse.  The jet has enough metal to actually cool the gas below ignition temperature if the flame enters the jet when you turn it very low or off.  In the main burner tube, the primary air plus fuel velocity is enough to carry the flame out to the end of the tube, but the velocity is not enough for the flame to lift off further unless you have way too big a jet.  And more air comes in at that point.  But I don't know enough to suggest how much primary or secondary air is required.

With a Meths burner, using a wick, the wetted surface of the wick encourages vapourization, and it is the vapour that mixes with air and burns.  The flame is generally a bit softer, and the division of primary and secondary air is not so obvious.  But the yellow flame and occasional tiny spark from a wick or a candle indicates incomplete combustion.  If there is space above the flame, there is enough air drawn in to burn the bright yellow glowing carbon particles.  But if the wick is too close to the boiler, or to a metal spoon, the gas is cooled before it burns, and the black unburnt carbon is deposited on the boiler or spoon.

There are also vapourising Meths burners, where the burner, and primary air holes are not unlike a similar size gas burner.  I even have a stove like this for camping as it is a very safe fuel.  And of course there are those petrol and kerosene blow lamps with their vapourising coils and fierce flames.

As an aside, solid rocket fuels contain the oxygen in one of the compounds used to make the fuel.  So once you have enough heat to start vapourising the fuel, the vapour has the oxygen and fuel intimately mixed, and you know the result.  You can burn a lot of fuel in a very short time when the oxygen is that well mixed without the air flow considerations getting in the way.  And minimal nitrogen to dilute and slow the process.

Hi Chris, I am glad the clarification was helpful.  Always difficult to identify the boundary between too many words and an incomplete explanation.  I will keep trying.

I assume your poker style burner is like those ones described on the Southern Steam Trains website.  They get over the problem of low radiant heat  from a clear or faint blue flame by allowing the flame to heat a SS mesh to a red or orange colour, which then is a better radiator than a pale blue or invisible flame.  If I understand his explanation, he uses much less gas with this arrangement than an open soldering type of burner, and still gets enough steam.  When you look closely at the picture, the air holes seem relatively large, compared with the ones on my ceramic burner anyway, so I assume all the air mixes there and the lower velocities allow adequate mixing for complete combustion.  Then the gap around the burner admitting a bit more air is probably acceptable but not necessary. 

When I look at my ceramic burner from Miniature Steam Models, the primary air holes appear  smaller than on those poker burners, and there are some secondary air holes around the outside but behind the ceramic insert.  The shroud then is a close sliding fit over the end of the boiler flue, not air tight, but not fully open.

I would expect the relative sizes of the primary and secondary air holes has been the subject of considerable experimentation.  The position of the gas jet relative to the primary air holes is critical and the burner comes with instructions on how to adjust it for a flame outside the boiler that becomes optimum when the burner is fitted over the boiler flue tube.  I suspect that is about induction of the right air quantity when installed in the boiler, where the draft obviously has an effect on the air flow.

Someone familiar with the design of full size burners might have access to computer programs that gave suitable answers for the starting point for a tiny burner, but I expect a successful design would still take considerable experimentation.  I for one am happy to pay for the results of all that work, so I can use my time on other projects.

If there is excess air, the total flue gas will have a lower temperature, which decreases heat transfer, but if there is enough draft, there will be more velocity, and this tends to increase the heat transfer coefficient.  So there is almost certainly a happy sweet spot between minimum air with maximum temperature, and more air at a lower mixed flue gas temperature.

I hope that adds a little to the understanding of burners, but you can see that my knowledge is more in the area of fluid flows, pressures and velocities, and boiling of fluids than the details of combustion, so there is plenty of room for those with more experience in the area to come in and help out.

Thanks for looking in

MJM460

[Gas_mantle]

This short video may be of interest to those wanting to know about coal firing.

It's about firing steam locos when the UK had a 'proper' railway but it is informative on the subject of coal firing in general.

<a href="https://www.youtube.com/watch?v=NHo860Q66Gw" target="_blank">http://www.youtube.com/watch?v=NHo860Q66Gw</a>

[crueby]

Getting the balance right on the air hole in the poker burner is why some include a sliding outer bit of tube, so the hole can be partially blocked to tune it. The little butane soldering torch I have for small work has that too.