|
Post by Deleted on Apr 30, 2015 22:52:42 GMT
Hi Julian ,
Problem with the Keiller formula and similar is not actually that they don't work but rather that they only give one answer .
All they basically say is that if you make one successful engine and then make another one not much different then the second one will probably be successful as well .
The reallity is that there are many other answers which could also give a successful result .
I'll probably be abandoning this thread before long but what I have been trying to get to bit by bit is a demonstration of how no one thing in an engine can be designed in isolation and how all elements of a complete engine interact .
For instance the tube layout and the blast pipe system are highly interdependent and there are multiple combinations which work and multiple combinations which don't . Choosing which is best does actually come down to getting correct flue gas velocity and turbulent flow .
Not my intention to get down to the complex physics and mathematics involved - just show the general principles .
Cheers ,
Michael .
|
|
|
Post by joanlluch on May 1, 2015 6:41:34 GMT
For instance the tube layout and the blast pipe system are highly interdependent and there are multiple combinations which work and multiple combinations which don't . Choosing which is best does actually come down to getting correct flue gas velocity and turbulent flow . For the boiler of my -gas fired- loco I intent to use relatively big diameter pipes. Then I will attempt to adjust turbulence by inserting turbulators. This is why and what I chose to do so: In a gas fired boiler the function of blast pipe should be merely helping gases to travel through the boiler, as it does not have an effect on actual power extracted from the burners. Also the blasting system should not disturbe the gas burners. This is in opposition to coal fired locos where draught has a direct and desirable effect on coal combustion power. As a starting point, I will fit big pipes in the boiler. Initially, this should require only a small action from the blast pipe, but efficiency will suffer due to little turbulent flow in the pipes. At this step I should be able to fire the burners with no or only very little blast pipe action. From that basis, I will experiment with inserting turbulators -essentially twisted strips of metal sheet- of several lengths into the pipes. It is expected that the longer and the more turbulators, the stronger turbulence and gas flow restriction in the pipes, thus requiring stronger blast at the smoke box. At some time -possibly after long experimentation- I should come with a combination of turbulent flow at the pipes and blast effect that works from the point of view of easy burner operation, enough power and relative efficiency. Or at least I hope so.
|
|
|
Post by Deleted on May 1, 2015 9:31:57 GMT
Hi Joan ,
It may be possible to make the burners generate hot gas flow patterns which are better shaped for use in big tubes than the simple linear patterns produced by standard burners .
For example a developed version of the Cyclone burner idea would give a helical hot gas flow pattern which would contact the tubes much better than one from a straight burner .
Also burners could usefully be designed to work in the ' roaring flame ' condition to generate true turbulent flow in relatively short tubes . 'Quiet burning' flames will be less effective if you really want to transfer a lot of heat .
MichaelW
|
|
steam4ian
Elder Statesman
One good turn deserves another
Posts: 2,069
|
Post by steam4ian on May 2, 2015 8:31:22 GMT
Joan
If you are going to use gas firing you would do well to look into the experience of those in the USA where gas firing has become the norm in many circumstances.
Issues to consider with gas firing is that the radiant heat input to the boiler is significantly less than with coal or oil firing. Note in the garden gauges gas burners often use stainless steel mesh heated by the flame to get the radiant heat transfer and some use ceramic elements for this effect.
One word of warning. You may find that such a gas system used in public has to have certification as an appliance. I would expect European rules are not much different than those in Australia. This will require purging cycles, flame detection, shut down valves etc.
Regards
Ian
|
|
|
Post by joanlluch on May 2, 2015 12:29:49 GMT
Thanks Michael and Ian. I think that one way to improve radiant heat is having the burner flames directly hitting water walls or pipes. This is what basically happens when you heat a kettle. I have tried to heat a soft drink can filled with water with the direct flame of a sievert kind burner. The can is aluminium so without water it will only stand the burner for very few seconds. Once you fill it with water all the heat goes to the water and the water starts boiling in a matter of about a minute. The can remains unaffected. So I think that this tells us that having a provision for the flames to hit the walls of the combustion chamber is the way to do. What is used in the USA differs from what I want to try because in the USA they use standard coal boilers to which they replace the grate with a series of burners. But the walls of the fire box remain relatively distant from the gas frames. I think that this kind of arrangement is not particularly efficient because of the lack of radian effect from the gas flame, as you suggest. However if you instead look at models from Accucraft they use some sort of marine boiler with water pipes in the path of the hot gasses and a low ceiling chamber where flames can hit the roof and walls.
About legislation, trouble is that it may be difficult to know what to apply. For the boiler considered as a pressurised vessel the rules are quite clear. However the rules applying to gas burners are more difficult to apply. They are meant to cover either industrial burners or domestic hot water / heating boilers. In most cases it goes down to fitting approved devices and connected sensors that perform all the required startup steps and safety checks during operation, but that's not enough because manufacturers of gas fired equipment must also get approvals for their designs. The legislation also allows for approvals of "unique" gas fired industrial "apparatus" that are not standard or that are build to serve a rare application. But getting approval for a non-standard device is usually a very long, tedious and expensive process that the industry tends to avoid by all means. Following that procedure for a small locomotive is possibly overkill to say the least, and will possibly result in the need to over-engineer everything just to meet the "unique apparatus" constraints.
But still, gas fired small locomotives do exist in Germany, which I doubt have officially passed any of the above (I may be wrong as that's just an assumption given the difficulties). So I wonder what their owners did, or whether they are never ignited in public. All this subject is a big question so far.
As opposed to the Anglo-Saxon law system, which is based on jurisprudence and relatively small number of mostly generic laws, the European one is a more 'civil' one which requires all the principles to be codified in a referable system. As such, the Anglo-Saxon system may allow an assurance to cover any uncertainties not directly covered by law, and thus make an assurance a requirement, but the European code does not work like that...
|
|
|
Post by joanlluch on May 2, 2015 13:31:52 GMT
From Memory, so I may fail to get everything right, but these are the requirements that an industrial gas burner must follow:
Required sensors: combustion air flow, secondary air flow, gas flow, flame detector, gas valve limit switch, Other requirements: safety gas valve, gas venting valve.
Startup (venting): - Combustion and secondary air sensors must signal off just before starting venting. - Gas flow sensor must signal off at all times during venting, otherwise stop startup procedure and report an error. - Gas flame detector must signal off at all times during venting, otherwise stop startup procedure and report an error - Switch on the combustion air and secondary air blowers for the required length of time to completely remove any unburnt gas in the system. - Both Combustion and secondary air sensors must signal on during actual venting, otherwise stop startup procedure and report an error.
Startup (ignition) - Combustion and secondary air sensors must signal on, otherwise stop startup procedure and report an error. - Open gas valve, close gas venting valve, open safety gas valve. Gas flow sensor must signal on, otherwise shutdown and error - Spark the burner to ignite it. - After a short while the flame detector must signal on, otherwise shutdown and error
Regular operation (ignition detected, regular operation) - Combustion and secondary air sensors must signal on. - Gas flow sensor must signal on. - Gas flame detector must signal on - Apply regulation to the burner gas valve (optionally to air valves). Control is generally based on temperature.
Shut down procedure. - Close gas valve, close safety gas valve, open gas venting valve - Keep combustion and secondary air blowers on, for the required amount of time to remove any combustion gases from the circuit. - Do not take into account the air sensors, continue with the blowing procedure for the stablished time anyway regardless of the sensors signal.
The above must be carried out automatically without direct human intervention or decisions.
Now, this applies to industrial burners but I do not suppose that gas fired model locomotives are required to perform all that. I suppose, at most they have a flame temperature sensor that switches off the gas, and that's all. Just like an old stove running on bottled gas.
If you think on it, it is possibly far more dangerous to leave a steam boiler without water than to spray some small amount of unburnt gas thought the boiler pipes. I mean, the boiler is open enough to the atmosphere -through the chimney- so IMO it is difficult to think on any big danger that may arise from a gas fired small loco.
|
|
|
Post by Deleted on May 2, 2015 14:29:37 GMT
Anyone interested in ways of transfering lots of heat in small spaces should study heat sinks for electronic devices - some of them are brilliantly well designed .
There is a well known danger with gas fired barbeques . Injuries from explosive flash overs are quite common .
|
|
|
Post by Deleted on May 2, 2015 15:15:23 GMT
--- more ---
An important consideration with heat transfer in fireboxes and nests of fire tubes is what is happening on the water side .
Ideally water side of fireboxes and tubes should be totally immersed in moving water all the time . Also ideally water should be kept moving naturally by thermal convection currents .
Many designs of boiler actually have quite poor water circulation . This usually means poor steam raising power and in full size boilers it can cause a host of problems with plate and tube damage .
Getting the water flow right is a major factor in determining :
(a) the fire tube size , number and layout . (b) the size of water spaces around firebox . (c) whether flow enhancers such as thermic syphons and return tubes are needed/beneficial .
Try doing some sketches to help visualise the water flow in a plain pot boiler , in a water tube boiler and in a locomotive type boiler .
--- more ---
|
|
|
Post by Deleted on May 2, 2015 15:30:27 GMT
Just a bit of light entertainment :
? Is it true that a coal fired steam locomotive hauling a train of loaded coal wagons over a long distance is one of the most efficient ways of transferring large amounts of energy ever devised ?
|
|
|
Post by joanlluch on May 2, 2015 15:45:04 GMT
Just a bit of light entertainment : ? Is it true that a coal fired steam locomotive hauling a train of loaded coal wagons over a long distance is one of the most efficient ways of transferring large amounts of energy ever devised ? Yes, This is (or was) used in Argentina. Now go figure a diesel carrying oil tanks, LOL!
|
|
|
Post by Deleted on May 6, 2015 12:45:30 GMT
//
|
|
jma1009
Elder Statesman
Posts: 5,901
|
Post by jma1009 on May 6, 2015 17:33:50 GMT
as a bit of an aside to michael's comments about boiler water circulation,
i agree that many ME loco boiler designs are far from good in this respect.
on many designs of round top fireboxes the barrel material is split to form the firebox outer wrapper, and extension pieces added to the sides or side. often the strap joining the extension piece is shown in the water space on the firebox sides rather than on the outside.
on many designs both round top and belpaire fireboxes an internal 'piston ring' of copper is added to provide a better joint for attaching the throatplate. this inner ring is in the worst place possible for water circulation effectively blocking off what should be a gap underneath the tubes and greatly impeding good water circulation. a proper double flanged throatplate is really no extra work and far superior as a generous flange can be provided for the barrel to throatplate joint on the outside. if the thoatplate flange to bottom half of the barrel is made of generous radius this will assist water circulation, plus makes the flanging easier.
cheers, julian
|
|
|
Post by Deleted on May 6, 2015 18:28:12 GMT
Hi Julian ,
You raise an important point - as well as being generally well proportioned the internal water spaces of a boiler need to be free from localised restrictions and as far as practical smooth .
Internal butt straps and piston rings joints are certainly bad . If they have to be there then at least taper off the edges a bit .
I had a look on Model Engineer forum today for first time in months and saw the thickness of firetubes question . Neil Wyatt's explanation of why firetubes should be very thick wall is a gem !
For anyone interested :
If the thermal conductivity of a component such as a firetube is very high compared with surrounding gas or liquids then you can within reason make it any wall thickness - it doesn't make any difference as far as heat conduction is concerned .
Heat flows in essentially the same way as electricity and it is possible to draw circuit diagrams for heat flow in much the same way as for electric circuits .
At any representative place in a fire tube you can think of the heat as flowing through resistors in series in getting from flue gas to water .
In a basic version of the circuit there would be two resistors for heat flow out of flue gas , one resistor for the flow of heat through fire tube wall and two resistors for the flow of heat into bulk of water .
Resistor for fire tube wall is very low value compared with others so it's actual value doesn't matter .
|
|
|
Post by andyhigham on May 6, 2015 18:45:31 GMT
Could the lack of internal butt straps and other internal restrictions, combined with generous width of water passages which get wider the further up the firebox be the reason that the "sweet pea" boiler is such a free steamer
|
|
|
Post by Deleted on May 6, 2015 18:58:01 GMT
Could the lack of internal butt straps and other internal restrictions, combined with generous width of water passages which get wider the further up the firebox be the reason that the "sweet pea" boiler is such a free steamer Yes - the excellent internal design of the boiler certainly contributes to it's free steaming qualities .
|
|
|
Post by Deleted on May 6, 2015 19:21:25 GMT
Just add :
(1) Girder stays as normally designed aren't good for water circulation . A modified form of girder stay with lots of cut away areas might not be too bad but will never be as good as simple rod stays .
(2) For background interest only - the electric circuit idea can be extended to an entire steam locomotive if anyone wants to .
|
|
|
Post by joanlluch on May 6, 2015 22:26:56 GMT
For anyone interested : If the thermal conductivity of a component such as a firetube is very high compared with surrounding gas or liquids then you can within reason make it any wall thickness - it doesn't make any difference as far as heat conduction is concerned . Heat flows in essentially the same way as electricity and it is possible to draw circuit diagrams for heat flow in much the same way as for electric circuits . At any representative place in a fire tube you can think of the heat as flowing through resistors in series in getting from flue gas to water . In a basic version of the circuit there would be two resistors for heat flow out of flue gas , one resistor for the flow of heat through fire tube wall and two resistors for the flow of heat into bulk of water . Resistor for fire tube wall is very low value compared with others so it's actual value doesn't matter . This is totally consistent with practical design of heat exchangers. The limiting factor is always the so called "overall heat transfer coefficient". See this www.engineeringpage.com/technology/thermal/transfer.html . In fact, the ability of the tubes to move heat between the inner and the outer walls is usually degrees of magnitude higher for any metallic pipe, including stainless steel pipes of reasonable thickness, so this is rarely a limiting factor. Thus, the use of cooper in boilers has no advantage in this respect.
|
|
jma1009
Elder Statesman
Posts: 5,901
|
Post by jma1009 on May 6, 2015 22:38:12 GMT
hi joan,
i dont think you should place too much reliance on stainless in boilers. it wasnt for nothing that most UK boilers in fullsize had copper fireboxes despite the considerable increase in cost.
last year i silver soldered up the copper tails on my stainless radiant superheater to wet header etc for Stepney made by Jim Scott. i could pick the stainless ends up straightaway by hand after the silver soldering at 620 degrees C!
one reason why i always make steam valve spindles out of stainless.
however i am now a confirmed convert of stainless radiant superheater elements having had another loco fitted with them last month and proved in service.
cheers, julian
|
|
|
Post by joanlluch on May 7, 2015 7:38:41 GMT
last year i silver soldered up the copper tails on my stainless radiant superheater to wet header etc for Stepney made by Jim Scott. i could pick the stainless ends up straightaway by hand after the silver soldering at 620 degrees C! one reason why i always make steam valve spindles out of stainless. Hi Julian, thanks for your input. What you describe is of course explained because the Thermal Conductivity of S.S is much lower than Copper. S.S is at 16 W/(m C) while cooper is a about the most conductive material in earth at 400 W/(m C). Data from here: www.engineeringtoolbox.com/thermal-conductivity-d_429.html So it is clear why you can hold one edge of a S.S pipe without your fingers being burnt while you are soldering the other edge. HOWEVER, my point (and the actual fact) is that this is almost irrelevant for a (tubular) heat exchanger. I will try to explain that with a theoretical case: Lets consider a pipe or set of pipes with a total transfer surface of 1 square meter and a wall thickness of 1 millimetre (expressed in metres, the thickness of said pipes is 0.001m). These pipes can be made in S.S or Cooper. Now, look at the figures for typical "overall heat transfer coefficients" of tube exchangers. Data can be found for example here. www.engineeringpage.com/technology/thermal/transfer.html . For Gases-Water exchangers, or gases-steam exchangers the figures range from 30-300 W/(sqm C). Lets assume we have a highly efficient exchanger and take the highest possible value of 300 W/(sqm C). Be aware that calling this "overall" can be misleading but this is actually something that has a reason to exist and is used thoroughly to actually design heat exchangers see this en.wikipedia.org/wiki/Heat_transfer_coefficient#Overall_heat_transfer_coefficientSo we have so far: Heat transfer surface: 1 sqm Pipe thickness : 0.001 m Stainless Steel thermal conductivity : 16 W/(m C) Cooper thermal conductivity : 400 W/(m C) Overall heat transfer coefficient for Gases - Water : 300 W/(sqm C) Let's assume we have an average difference of 100 C between the fluids. Now, the power (heat) that can be transferred in this system based on the fluids being exchanged will be: 300 W/(sqm C) * 1sqm * 100 C = 30,000 WThe heat transfer capability of Copper pipes in this system would be : 400 W/(m C) * 1sqm / 0.001 m * 100 C = 40,000,000 W The heat transfer capability of S.S pipes in this system would be : 16 W/(m C) * 1sqm / 0.001 m * 100 C= 1,600,000 W So we can see that regardless of pipe material, their heat transfer capability is far above what the heat exchanger can do, and both cooper and S.S well satisfy the requirements. You can change any figures you like, but essentially the heat transfer coefficient of the fluids involved would be always the limiting factor, not the pipe material. Unless you would use 50 mm thick S.S. pipes, of course, but that's not even possible. I hope this helps to clear my point about cooper not being a specially advantageous material over other metals for usual designs of heat exchangers or boilers. Joan.
|
|
|
Post by Deleted on May 7, 2015 9:50:22 GMT
Joan & Julian ,
Stainless steel was not used in many coal fired steam locomotive boilers because :
(a) During most of the history of locomotives stainless steel and the means to make tubes and flanged plates from it were just not available . When stainless steels did slowly arrive it was a long time before suitable grades and effective means of using them evolved .
(b) Stainless steels initially available performed very badly with many coals due to corrosion and cracking problems .
(c) There was no obvious advantage in using stainless steel anyway in a conventional design boiler .
Copper (and its alloys) were extensively used in coal fired locomotive boilers because :
(a) Copper was always available and could easily be made into flanged plates and tubes even in the early days .
(b) Copper conducts heat so well (1) . Copper is very good for distributing heat and this is a good thing in boilers . In the firebox for instance the conductive properties of Copper take some of the intense heat of the fire away - by conduction in the Copper itself - to other relatively cooler areas of firebox . This aids overall heat transfer into the water and reduces the problems with burnout associated with firebox plates actually in contact with fire .
(c) Copper conducts heat so well (2) . What was said in earlier postings about conductivity not being that important overall in transfer of heat to water is certainly true but nevertheless copper is better than steel so may as well use it - every little helps .
(d) Copper performs well with many coals and is resistant to corrosion and cracking .
Just note that there are a few coals that damage Copper fireboxes badly - some GWR engines working off region and some preserved engines have had to be taken out of service for this reason .
As always in engineering nothing is quite simple and there is certainly a lot more involved in the mild steel/stainless steel/copper story .
|
|