Compounds ex time machine thread

Has there been a dedicated NP thread about 'compounds'? I'm intrigued to know more about what was tried, and why it never caught on in the UK. Not that I'm fully across the theory, but it seems like a good idea.

 

Well, there are a number of reasons but one which needs to be taken into consideration is technical education. A quick look at the C.V.'s of such as Du Bousquet, Mallet, Chapelon and de Caso will show this.

PH

 

Wardale also pointed out that the British loading gauge restricts the cylinders a large compound might need.

AFAIK the most powerful compound in the UK was the Midland one & that was only a 4P

 

Deleted - my error

 

Surely the W1 was more powerful?!?!
There was also a Lanky Dreadnaught converted to flour cylinder compound (5P) in preparation for the Fowler compound Pacifics which were never built.

 

Mea Culpa, but the Midland Compounds were certainly the largest 'production' loco

 

To expand (groan) on it a bit, in a multi stage expansion engine the 'output' is effectively the low pressure cylinders working on HP steam if that makes sense

So take the triple expansion engine in the paddle steamer Waverley, 3 cylinders 24, 38 & 62 inches by 66 inch stroke, 2100hp@56rpm 180psi saturated steam

The HP & IP could be removed and you could simply run on the 62 inch LP taking steam at 180psi - the ships motion would be abominable but thats another story.

So the problem you have on a railway locomotive, especially a big one is fitting in all that extra cylinderage within the UK loading gauge

The next issue is over cost, as we have discussed before coal isnt the only cost and you have the cost of installing all that extra machinery, maintaining it and hauling it around. It may well be - well, clearly there is much more that can be done to improve fuel efficiency/reduce running costs before you get to building a compound loco

 

The GNR(I) V class was more powerful (with original boiler pressure) at 23,762 lbf against the 21,840 lbf of the Midland locos.

Keith

 

He didn't need to, they did the job for which they were designed. They were not superlative, or even as good as an engineer as able as Webb should have been able to manage, but they did what was required. Not all the abuse hurled their way is deserved, in the context of the time they were built. Later of course he did switch to a four cylinder layout which was much better.

 

You can't get something for nothing in thermodynamics - there's no reason why you can get more energy out of steam by expanding it in two stages rather than just one. If the valve gear allows full expansion, you can get the maximum energy out without the added complication of compounding (and inherent energy loss of the steam moving between cylinders). So better valve gear design to allow a full range of expansion is a better route on a locomotive to maximising efficiency without introducing significant complication and additional heat losses.

There are two really big energy losses on a conventional steam loco, with neither of which compounding helps. The first is that the exhaust from the firebox is thrown away at several hundred degrees centigrade. That represents a huge mass of air heated form ambient to high temperature and then discarded. Moreover, for every unit mass of oxygen that is drawn through the fire usefully, you drag four times as much nitrogen through, heat it to several hundred degrees, and throw it away (which is why you need to control airflow so that you have enough for combustion, but no more more). (*)

Secondly, the really big energy consumer in producing steam is the latent heat of evaporation of water at its boiling point to steam at the same temperature (**). Without a condenser, you take water as input and discard steam as output, throwing away all that energy (***). Compounding doesn't help there.

The really high efficiency steam engines (for example, plant in power stations or ships equipment) overcomes the second of those by having condensers; and the first by very efficient heat extraction from the combustion gases. Neither of those is possible within the mechanical and physical constraints of a steam locomotive.

(*) At a crude level, burning 12g of carbon requires 32g of oxygen, but that is associated with 112g of nitrogen. When it burns, you form 44g of CO2 and 112g of nitrogen remains, but that whole mass has to be heated up to the exhaust temperature. So to a first approximation, burning a ton of coal means heating up 12 tons of air and throwing it away.

(**) Consider turning on an electric kettle and holding the switch down so it won't turn off automatically. When you turn it on, it heats the water from room temperature to boiling quite quickly, but will run for ages (at the same power) without ever getting close to boiling dry. Which shows that far less energy is needed to heat water to its boiling point than to then evaporate the same amount of water to steam at the same temperature.

(***) You get a bit of that energy back in an exhaust steam injector, which uses that latent heat of evaporation to heat the feed water and push it into the boiler, but certainly nowhere near all the energy in the exhaust steam is used that way.

Tom

 

Possibly done more to conserve water than for overall thermal efficiency but weren't there a few condenser equipped steam locos built in various countries (including the UK)

 

Much good sense from Tom there, as usual, but I would quibble with two points.

To get the same range of expansion in a simple engine you need to work at a much shorter cut-off, and that places more stringent demands not only on the valve gear but on the valves themselves and the associated steam passages; though on the other hand a simple has less steam passages to be fitted in.

There is also the matter of heat flowing from the steam into the cylinder walls and back again, which wastes more energy in a simple with a wide temperature range than in a compound where the range is split between the HP and LP cylinders.

If it were just as easy to get the same efficiency with a simple Chapelon wouldn't have bothered with compounding.

When you condense the steam you get back the latent heat but you then need to get rid of that heat. The huge cooling towers of power stations are there for exactly that purpose. If you want the benefit of it you use it for district heating. The efficiency benefit from condensing is mainly that the steam can expand to a pressure below atmospheric. You also get the benefit of pre-warmed water to send back into the boiler, but there are other ways of pre-heating on steam locos and they mostly turned out to be not worth the bother.

 

Sorry, I beg to disagree! Compounding is about curbing losses and having ports that are wide open is inherently better that a single stage expansion.
Single stage cut-off at 25% is comparable with compound cut-off at 50% or so.
As for compound usage, countries with small coal supplies loved it: Austria, Switserland, France, Bavaria. Not having to drag large additional quantities of coal
to coaling stations is a big issue in that case.
Kind regards
Jos Koopmans

 

Quite often done to reduce steam in tunnels I think - many Victorian locos built for suburban work round London had condensers.

However, you can’t condense all of the exhaust as you still need some exhaust steam to provide the energy to move the mass of air through the fire. (You also run into a practical problem with some kinds of condenser that the feed water gets contaminated with cylinder oil).

Tom

 

This is the real crux of it... There is only so much energy to be had in the steam supplied to the cylinders and it doesn't matter if you convert that energy in one 'bite' or two - there is only so much there.

So what really counts is the efficiency of that conversion process from thermal energy into the mechanical form. If this was 100% efficient there would be no point in compounding - but of course it never is. So compounding is really just an attempt to make up for the inefficiencies of this process. So - when a designer sits down to improve a simple expansion design he has two options - improve the valve gear and steam circuit or add compounding. The latter will always win out in pure efficiency terms - but if the simple valve gear is well designed then the difference is not all that great. Pulling some figures out of the air to illustrate - Lets assume that your conversion process is 50% efficient. A compound can pull 50% of the available energy out of the steam in the HP cylinder then 50% of the residual 50% in the LP side - a total of 75%. However - if you can push the simple engine up to say 70% with a superior steam circuit etc then the benefit of compounding remains but is much smaller.

But compounding does not come for free - there are additional manufacture costs and maintenance costs to deal with as well as additional weight. At what point does that small improvement exceed the extra costs from the compound? This is a subjective decision that designers have struggled with. The first to make a real comparison of the two strategies was Churchward when he compared his new Saint 4-6-0 with the accepted 'best' compound in the world at the time - the DeGlenn atlantic from France. He had no doubts as to the outcome and no-one has really had any data to challenge that since. Over time, as maintenance costs have risen even more in proportion there was just no reason to return to compounding. At least in the UK...

The other factor to consider is that coal in the UK was cheap - much cheaper than say in France. So you didn't save all that much in economic terms by squeezing that last few % of efficiency out of the design and so the downsides of a compound were more significant still. If you look at the US as another example, fuel in general was relativity cheap and although they persisted with compounding for a while, almost all of the late designs were using simple expansion.

So compounding had its place and over UK rail history had a fair trial - but in the specific circumstances of rail locomotives in this country, was not worth the trouble.

Mike

 

A translation of mr Nordmans compound versus simple report can explain a lot.
I have downloaded it as a PDF file and do not know how to tranfer it to one of the AI translators from Google or whereever.
It can be found here ,somewhere,

http://anno.onb.ac.at/cgi-content/anno-plus?aid=lok&datum=1941&size=45&pos=193


and I have it as a 8.41 MB pdf file.

 

As paulhitch points out part of the reason that compound expansion locomotives were not more common, did not catch on, in the U.K. was the the level of education to be found in the locomotive engineers working here.

The development of the compound took many years much like the development of any other similar concept. There were misconceptions and confusion when superheating came onto the scene and it took quite some time before somebody understood how the locomotives should be worked with respect to the use of cut-off in the combination of high and low pressure cylinders.

In Europe the demands made on locomotives reached a far higher level than here. If you look at what was required of the Chapelon Pacifics and 4-8-0s it will give some idea of how underpowered U.K. machines were in comparison. True we thought that we did not need such power and efficiency at the time but as the years went by people wanted to travel faster and to meet this demand you can either run faster or accelerate better. Or use a combination of the two. Compounds accelerate better than simples. Speed is not a problem either.

When the great 4-8-4 experimental engine sent the French electric locomotive designers back to the drawing board, when the 2-12-0 showed that high power outputs at low speed were perfectly feasible and far superior machines were already being designed what was going on here?

I have posted before my view that by and large U.K. designers were technically too limited and not good enough. Though to be fair they were labouring not only in the darkness but also within complicated circumstances. But this does not excuse the poor power to weight ratios of the engines designed under Riddles whose obsession with simplicity means that we need a new build to equal what Chapelon achieved some eighty years ago.

 

There was an article in 'The Railway Magazine' some years ago on the lines of "Why not a British version of Chapelon's 242A1?" which produced a follow-up letter pointing out that to fit the British load gauge the outside LP cylinders would have to be reduced to a maximum of (say) 20" diameter and the inside HP one reduced to keep the proportions the same. From memory this would reduce the power output to not much over 50% of that from the SNCF 242A1 which could be bettered by a Peppercorn A1. The only advantage that a British version of 242A1 would have would be 4 coupled axles rather than 3.

 

Hang on a minute! Are you trying to say that every time the issue of compounding was floated in the UK by any or all of the loco design teams, from any or all of the railway companies, (and no doubt it was floated often because it was being done successfully elsewhere in the world), the answer was invariably, "forget it, we can't possibly work that out because we are not smart enough - in fact we are not even smart enough to send a man over the channel to find out how it is done well." Or not smart enough to buy a French compound to test it....no, wait, someone did do that.

Surely the answer lies elsewhere.

 

It is here that Webb is important.
Use two outside high pressure cylinders and one big inside low pressure.
For a Duke like thing two 18inch outside and one 36 inside low pressure can easily run within UK loading gauge.
An A2/2 version with first axle drive a la Raven 22 inch outside and 44 inside will pass.
If four driving axles is a must:
https://imgur.com/whbIOWW