|
Post by Deleted on Apr 17, 2015 13:47:44 GMT
/Deleted/
|
|
|
Post by Deleted on Apr 21, 2015 13:28:43 GMT
Michael. The reawakening of this topic has prompted me to go back to the beginning. This in turn has led me to ponder steam locomotive development. The question then becomes; has any body/group bit by bit evaluated the form and function of each part of a steam locomotive? By this I mean as an example transmission between driving axles, are side coupling rods the best thing? What is the best way of coupling the engine or engines to the driving axles? If each engine or engine pair drove one axle how is slip managed? My point, there is still room for assessment of the functional efficacy of each part of a steam locomotive even at model level. Graham Driscoll's latest effort is a turbine driven locomotive. Many people have tried to make a better steam locomotive but very few have ever achieved much . Problem is that performance of a steam locomotive is decided almost entirely by the temperature\pressure of steam that can be safely used and by the loading gauge . Pressure and temperature are decided by the metalurgical limit of steel used for boilers and superheaters . Amount of output power of engine depends on size of boiler . Size of boiler is limited by loading gauge . On the mechanical side a simple configuration of wheels , cylinders and valve gears does almost all that is possible with a steam locomotive . Any attempts to move away from the traditional limitations and configuration of a locomotive lead to extra weight and great complexity for very little return . Complex boilers , compounding , condensers , economisers and sophisticated valve gears have all been tried and soon abandoned . The clever CME's of yesterday recognised that the way to go was to optimise all aspects of the standard simple locomotive configuration and leave it at that . Turbines in simple format have no performance advantage over reciprocating engines . They only have advantage when multistage and used with condenser so that steam can be expanded over most of the possible range and maximum energy extracted . Basically the same as several stages of compounding in a reciprocating engine working with condenser . No-one has ever come up with a practical arrangement of condenser on a locomotive . The chassis layout of a locomotive could be arranged as electric motor type power units on bogies . This has been done but most high power engines just have two or more conventional engine units articulated under a common boiler . There are ways forward though : Eliminate the boiler as such and all it's attendant problems . There are possibilities for closed cycle steam engines packaged like diesel engines . Only really practical with oil firing but could be quite efficient . Doubt whether they could actually be better than diesels though . Leaving locomotives aside there is considerable scope remaining for developing land and sea based steam plant . Most coal and nuclear power generating plant is steam based and likely to continue so for a long time . More interestingly perhaps steam technology lends it self very well to use in alternative power generation schemes both on large and small scales .
|
|
|
Post by Deleted on Apr 21, 2015 14:19:50 GMT
thank you michael, to get down to basics, and the nitty gritty, can i please mention superheaters? i am not (as is well known) a fan of coaxial superheaters, and have fitted stainless radiant superheaters to Stepney (untried as yet) and LINDA (tried and tested) last weekend. my general policy is determined by my 5"g GWR 0-6-0 loco which has 3 x 1" OD superheaters with 1/4" dia pairs of elements in each. however i have used the same arrangement and size of elements in 7/8" OD flues and in Stepney's case 3/4" OD. do you have a view please on the subject of superheaters? we discussed coaxial superheaters awhile back, and i know pete (greenglade) will be interested in this subject re his DONCASTER, as i think this is one big error on the DONCASTER design, and many of don young's later designs. hope im not going off topic again! cheers, julian I don't see that co-axial superheaters have any merits at all compared with simple return bend types . Apart from any thermodynamic considerations they just make it near impossible to clean the flues . Probably someone somewhere used them in full size but I have not come across any examples myself . The only things which are superficially similar are steam powered immersion heaters used in chemical and petrochemical processing . I'll get back to the broader subject of superheating later .
|
|
|
Post by joanlluch on Apr 21, 2015 14:34:06 GMT
Julian, about co-axial superheaters my gut feeling for them to be effective is this:
(1) Wet steam must enter the outer tube and hot steam must return through the inner tube. (2) The inner tube must have an insulation sleeve to prevent already heated steam to be cooled down again.
This keeps the return pipe (inner one) out of the equation and the only intervening surface for heat exchange is the outer pipe, which then works exactly as it should.
This follows the principle of heat exchangers, where the fluids being exchanged travel in opposite directions through the exchanger. In this case steam goes from the front to the rear, hot combustion gases go from the rear to the front, and the exchanger surface is the outer tube. The inner tube just conveys steam to the outlet but it does not intervene in the heat exchange because it was insulated.
I do not know of any case where (2) has been attempted. In my opinion this may be crucial for it to work.
|
|
|
Post by Deleted on Apr 21, 2015 16:51:38 GMT
(See full text in earlier thread) I have been thinking for some time on the subject of making a simpler boiler design, and I will make tests at some time based on some ideas I have. Fresh ideas or theoretical background on the subject of simple boiler design is something that interests me. Joan (1) Most heat transfer in boilers occurs in firebox and about first third of tubes length . (2) Heat is transfered from fire and flue gasses into boiler by three processes : (a) Radiant heat transfer - mainly from the fire bed . (b) Conductive heat transfer - again mainly from fire bed . (c) Convective heat transfer from flow of flue gasses over firebox plates and in tubes . All three processes are active in firebox . Convective heat transfer is what mostly occurs in tubes . Heat transfer directly from the firebed is relatively big and other heat transfer is relatively small . Radiant heat transfer depends upon a firebed being above dull red heat - it alway is so no problem . (3) Combustion should ideally be complete within the confines of the firebox and certainly within confines of boiler . This does not occur naturally so brick arch is fitted in firebox to give longer flow path and tubes are sized to give an appropriate gas flow velocity . (nb . Brick arch has other uses such as deflecting flue gasses so as to contact more area of boiler plates and to protect tube ends from extreme heat . Flue gasses tend to take a short cut from firebed into tubes if brick arch is not fitted ) --- more ---
|
|
|
Post by Roger on Apr 21, 2015 18:04:21 GMT
Hi Michael, What's the reason for the bulk of the heat transfer happening in the first part of the flues? Is this because convection is very strong there and water is drawn along from further forward? Maybe it's because there's such a large temperature difference between the flue gasses and the tubes that the flue gasses are cooled very quickly so they can't do much useful work further on? It does beg the question, why not have fewer bigger flues so that heat transfer happens more evenly along the barrel.
|
|
|
Post by Deleted on Apr 21, 2015 19:45:20 GMT
Hi Michael, What's the reason for the bulk of the heat transfer happening in the first part of the flues? Is this because convection is very strong there and water is drawn along from further forward? Maybe it's because there's such a large temperature difference between the flue gasses and the tubes that the flue gasses are cooled very quickly so they can't do much useful work further on? It does beg the question, why not have fewer bigger flues so that heat transfer happens more evenly along the barrel. Hi Roger , Proper answer should drop out of next few postings but certainly some of the principles that you mention are involved . MichaelW
|
|
steam4ian
Elder Statesman
One good turn deserves another
Posts: 2,069
|
Post by steam4ian on Apr 23, 2015 23:53:26 GMT
Roger/Michael
The question of optimum tube diameter is very interesting. More small tubes means more transfer surface but less effective heat transfer at the smoke-box end of the tube. For some reason tubes seem to be about 1.5"to 2" diameter no matter how big the boiler
The decline in heat transfer along a tube is a direct result of the gas temperature being lower by such transfer.
Ideally the gas temperature in the smoke-box should not exceed the water temperature on the boiler.
Regarding coaxial superheaters; there were semi coaxial superheater designs tried. These were much more complex than the one tube inside the other approach used by modellers; their complexity probably countered any other advantages they might have had.
|
|
steam4ian
Elder Statesman
One good turn deserves another
Posts: 2,069
|
Post by steam4ian on Apr 24, 2015 0:09:39 GMT
As previously stated there have been numerous attempts to make quantum leap improvements to steam locomotive performance. Few have been a sustained success.
The downfall of the simple open steam cycle used in a locomotive is the relatively low temperature and enthalpy difference between the steam supplies to and exhausted from the expander. relatively too much energy is used just making steam and too little goes into producing work. Superheating was one almost quantum leap.
Squarer better timed valve events is only tinkering at the edges (sorry Don). Did the Caprotti gear on the Duke really make an improvement over Walshearts or outside Stephensons (for Don)? Unfortunately a second loco with Walshearts gear was never built for comparison. Likewise feed water heating, exhaust steam injectors, Kylchap/Lempore exhausts, gas producer fireboxes and Franco Crosti boilers only made marginal improvements. The success, or otherwise, of these tinkerings was more based on improvements to operability and/or lowered maintenance than quantum leap efficiency improvements. That said, I do recognise the improements made by Porta et al could make what were otherwise slugs into "really useful engines".
|
|
|
Post by Roger on Apr 24, 2015 6:53:14 GMT
Presumably the failure to produce an effective closed system is due to space limitations as much as anything then? I suppose the boiler could be a lot shorter is the heat transfer can be done that quickly and the rest of the space used for a condenser? There's still the problem of getting rid of enough energy to condense the steam, so a pretty large radiator would be required. I imagine that even if all this could have been achieved, the complexity of it all would wipe out any fuel savings in the high initial cost and maintenance.
|
|
|
Post by ejparrott on Apr 24, 2015 8:08:16 GMT
There were various engines built with condensers and cooling equipment, I've seen photo's of some pretty large tenders covered in them, but it seems to not have been a great success.
|
|
|
Post by Deleted on Apr 24, 2015 8:34:57 GMT
I'll continue with the main theme shortly but to pick up on several points raised :
(1)
There is no doubt whatever that just fiddling with minor design details of a simple steam engine produces hardly any improvement in thermal efficiency BUT quite small changes in detail design have improved the operational efficiency of steam engines markedly .
Thermal efficiency is next to nothing anyway and it will stay that way with standard simple configuration engines but the way that steam energy is converted into useful power is another matter all together .
Vast subject but in essence it's all down to getting a favourable shape for the engine torque-speed characteristic at the wheels .
(Conceptually same as for an i/c engine - you can have engines which produce good torque at low speeds , engines which produce useful torque at a range of medium speeds and engines which produce best torque at higher speeds - depends what you want the engine for .)
Basic torque speed characteristic for a steam engine is normally rather flat in the mid range and tends to fall off badly at higher speeds . Some quite simple detail design changes have given some engines a much improved characteristic and there is considerable scope for further work in this direction .
(2)
A close cycle steam engine which is turbine based is quite feasible and some demonstrators have been built . Never applied to a locomotive though as far as I know .
By using gas turbine technology rather than old type steam technology all sorts of fascinating things can in principle be done .
A quite high power closed cycle engine working at very high pressures and temperatures need only have a few cup fulls of water in it and a condenser can be made quite small . Wouldn't look anything like normal condenser though - it would either be a film condenser or a 'bootstrap' type turbo condenser .
(3)
Effectiveness of heat transfer in firetube part of boiler does not depend on number , length or diameter of the tubes in any simple way . Several conflicting requirements for the number and sizes of the tubes have to be optimised at the same time .
|
|
|
Post by Deleted on Apr 24, 2015 9:34:56 GMT
There were various engines built with condensers and cooling equipment, I've seen photo's of some pretty large tenders covered in them, but it seems to not have been a great success. Yes there were a few - totally unsuccesful in most cases . Proper information is a bit scanty but I understand that condensers were fitted to some engines to conserve water rather than increase efficiency . Some railways crossed desserts with limited and unreliable fill up water available so a condenser might make sense . Unfortunately hot dry regions are not the best places to make condensers work .
|
|
|
Post by andyhigham on Apr 24, 2015 11:35:58 GMT
Ideally the engine would have several stages of compounding to extract all the energy from the steam. The exhaust from the lowest pressure stage would be hot water to be fed back into the boiler. The down side would be fans would be needed to provide draughting for the boiler. Every chuff of a steam engine is wasted energy going up the chimney
|
|
|
Post by Deleted on Apr 26, 2015 12:32:24 GMT
Carrying on with heat transfer :
What happens in the firebox is more or less obvious - but what happens in the fire tubes is much more complicated .
The combustion products from fire consist of gasses and quite a lot of solid particles and larger lumps .
This mixture may be just hot products or it may be still burning and continuing to produce heat . Usually some of each at entry to tubes . Ideally any burning is complete or quenched before exit from tubes .
Fire tubes have two main purposes :
(a) to provide a clear exit path for the combustion products . (b) to transfer heat into boiler .
One very important design consideration for tubes is getting the flow velocity of the combustion products through them correct .
There are several reasons for this :
(a) Heat transfer takes a finite time - if flow is too fast then heat just goes clean through tubes and uselessly to waste . (b) Combustion products (ideally) should be in a condition of turbulent flow and this depends in part on velocity .
Any fluid flow can exist in two conditions :
(a) Laminar flow - sometimes called streamline flow - which is tranquil . (b) Turbulent flow which is highly disturbed .
Turbulent flow is much more effective in heat transfer situations .
--- more ---
|
|
|
Post by Deleted on Apr 30, 2015 13:31:29 GMT
--- more ---
With turbulent flow the combustion products as well as flowing from one end of the fire tube to the other are being churned up all the time and this results in three useful effects :
(a) Heat is constantly being moved from core of flow to outer layer in contact with tube .
(b) Very good heat transfer to tubes due to scouring action causing very good contact between flow and tube .
(c) Minimises formation of boundary layer between flow and tube . Boundary layer is a layer of inactive flow which adheres to inside of tube and makes heat transfer less effective . Tends to get thicker as flow progresses along tube .
--- more ---
|
|
|
Post by Roger on Apr 30, 2015 14:43:00 GMT
From an outsider's point of view, it seems that it's very difficult to translate the theory into practice unless there is a FEA program that can be used to model the flow of gasses and boiler water in a variety of scenarios. Are the rules of thumb that have been posted elsewhere are about as good as as we're likely to get, or has anyone modelled this?
|
|
|
Post by Deleted on Apr 30, 2015 15:01:52 GMT
Heat transfer has been studied extensively in full size engineering .
FEA is used in modern times but quite good results can be obtained from simple hand calculations .
Like many things in engineering it is not nescessary to do exact calculations anyway to get good results . Just knowing that there is a difference between laminar and turbulent flow and what design factors are likely to ensure turbulent flow is good background information and can be incorporated into broader design requirements .
Purely aside - the big differences in steaming quality sometimes obtained by quite small adjustments to blast pipe and nozzle arrangements are due to the minor change flipping the gas flow from laminar to turbulent - doesn't take much if you are near top limits of one and bottom limits of the other .
The 'rules' used in model boiler design aren't even empirical - they are just fudge factors .
|
|
|
Post by joanlluch on Apr 30, 2015 15:11:00 GMT
From an outsider's point of view, it seems that it's very difficult to translate the theory into practice unless there is a FEA program that can be used to model the flow of gasses and boiler water in a variety of scenarios. Are the rules of thumb that have been posted elsewhere are about as good as as we're likely to get, or has anyone modelled this? Fluid dynamics is very hard to model. Most mass or heat exchanging systems involving fluids are designed based on practical rules and real world experimentation. Think on F1 cars. If car aerodynamics or hot gases in the cylinders could be reliably simulated through computers, the sport would turn a very different game. .
|
|
jma1009
Elder Statesman
Posts: 5,901
|
Post by jma1009 on Apr 30, 2015 22:14:28 GMT
michael, you are being a bit of a 'tease' re boiler flues! ive used the Keiller formula with the emphasis on generous bore tubes verses length. plus i aim for a generous free gas flow through the tubes as a proportion of grate area. cheers, julian
|
|