Driving a compound locomotive

Just because it hasn't done a lick of revenue earning work* for at least the best part of a century, how could I forget about Aerolite? That little loco has to be both the strangest and luckiest of UK 'big railway' survivors, bar none, in every way imaginable.

A thrice rebuilt loco that started out as a clone of itself and survived grouping as the 'personal chariot' of LNER Chief Assistant Mechanical Engineer A.C. Stamer** from 1923 until that gentleman's retirement in in 1933 and even then, somehow contrived to be preserved, escaping the WWII fate of Mr.Drummond's The Bug. Talk about a charmed existence!

* moving the odd wagon out of the way to exit Darlo doesn't really count as 'revenue earning' in my books!
**how does a Chief Assistant Mechanical Engineer come to enjoy such seemingly lavish provision? Evidently, he was a highly valued senior company official. All I could find was this: http://www.thepeerage.com/p58246.htm

 

Yes - I have driven K1 a few times...

It has lever reverse and has been said up-thread you don't notch it up very far. Get the train moving, bring it up one notch and that is about it. You cant notch it up any further and maintain satisfactory progress.

K1 has a simpling valve like that mentioned above that provides (from memory) up to 80 psi into the LP cylinders to overcome train resistance. I only had cause to use it the once - at Hendy crossing in the days that this was stop-and-proceed. We stopped and the loco did not want to proceed. A nudge from the simpling valve however did the trick and away she went.

As effectively a prototype, K1 has a few quirks and things that aren't quite right. For instance, the two ends aren't well balanced in terms of the work that each is doing. The rear HP end is always providing a lot more of the traction that the front LP unit. As a result the rear end is always the first to slip in poor (typical welsh) rail conditions. And the backhead mounted injectors are (or at least were) a real PIA. An interesting machine none the less. At least you can see where you are going an awful lot better than on a 16!

Its a shame it is simply the wrong size to be very useful to the WHR (rated max 6 cars - on a good day). The line never runs trains that short which means you always need to double head with the crewing problems that creates and with what do you pair it?

 

I was in contact with John Startin in 2014 and he was of the opinion that the K1 was a work in progress.

The valve gears are linked to the same cut off in each unit and the h.p. cylinders are inside admission with the low pressure outside which makes sense. The reversing quadrant offers five settings but the three shorter settings do not work well. (That should be a red flag from the off - the original Talyllyn locomotives wouldn't link up either but they wouldn't be on their own)

The high pressure ball joint was leaking very badly and so steam pressure to the h.p. unit had to be restricted 140 psi in spite of the boiler being capable of working at 210 psi the engine was at the time working at 190. The offending joint was replaced but not enough experience had been accumulated at the time to assess the impact of this.

He wrote that there had been hints of what the K1 could do. The group tested the engine in some secrecy on 11 coaches to the summit of the line and it kept time which is quite an achievement for an engine which is happiest at its design speed of 15 mph though it will run at the WHR required 25 mph. The impact of redesigning the valve gear and making the two units capable of independent adjustment would be worth discovering because this would help to equalise the power outputs of the two units. And attention to valve gear would give other benefits including making the locomotive "happier" at higher speeds. Working at say 70% hp cut off and 45% in the l.p. - now there's a thought.

 

If (with the present configuration) the HP power unit tends to deliver more power than the LP, wouldn't it make sense to run at a shorter cutoff in the HP and longer in the LP?

 

Maybe even a South African/Australian Garratt?

 

I'm not sure the PW guys would be too excited about K1 running at 25mph on a regular basis. With an axle load of 8 tons vs 6 tons something for a NGG16 - and no pony truck - K1 had a reputation of being hard on the track at 15-20mph. Slight track imperfection were rapidly hammered into large ones - or so I was told....

 

Many thanks for the contributions so far guys.

Ha, I am 99% this one is a simple loco that masquerades as a compound, but it was enough to spark my interest! Thanks for th book recommendation as well.

 

The cylinder dimensions for K1 are listed as H.P. 11"x 16" and L.P. as 17"x 16".
Ray.

 

I have a copy that I bought second-hand (from the WSR), which has a few damaged pages. If anyone would like it for the cost of the postage PM me.

 

You need to have sufficient steam flow into the receiver for the l.p. unit to develop its share of a contribution, the h.p. side cannot be allowed to exhaust at too low a pressure since the h.p and l.p. cylinders should have similar outputs. We suspect that the valve gear design might need attention with h.p. and l.p. needing independent adjustment though once the exact relationship is worked out you can design a mechanism which removes the need for this independent control. The idea of short h.p. cut offs plagued pre-Chapelon compounds and their use. Four cylinder compounds had volume ratios l.p./h.p. of between 1.94 and 2.69 for French practice, 1.63 and 3.18 for British practice, 2.18 and 2.89 for German. The Chapelon 240P was 2.32 - the K1 has the cylinder sizes that it has and the ratio is 2.39 so that does not present any problems.

Regarding the ride of the engine and the lack of a leading bogie, the Double Fairlies weigh little less than the K1, number 11 weighs 31 tons and the Garratt 33.5 tons and with these two engines in mind I suspect that balancing could be the issue not the lack of a bogie. As has been said before the engine is a work in progress.

 

I spoke to Fred Howes at Port earlier this year and the topic of wear and tear on the track came up. To be fair, he was no fan of fairlies driven at full line speed either....

 

It is worth adding some comments, partly based on Van Riemsdijk's book:
Basically, the amount of steam used in a compound is controlled by the cut off in the high pressure cylinder(s). For a compound with a receiver and independent reversing gears, the low pressure cutoff determines the proportion of power developed in the high pressure and low pressure cylinder(s), but it is the other way from what one might expect in that lengthening the low pressure cutoff reduces the pressure in the receiver and thus increases the power developed in the high pressure cylinder(s) and reduces that in the low pressure. I have not heard of it being done, but it should be an easy experiment in a locomotive with independent gears to adjust the low pressure cutoff for any given value of high pressure cutoff so that the total power developed is a maximum, and then to do this throughout the cutoff range. It is surely better to do something like this rather than suppose that the optimum position for the low pressure is the same as, or a simple ratio based on the high pressure and control both from a single reversing lever. it is of course possible that the optimum position depends on the speed.
For four-cylinder compounds without a receiver is seemed normal practice was to have the same cutoff in both sets of cylinders, though I am not sure if there is any practical basis for this. In fact, without a receiver the low pressure cutoff is a bit of a misnomer as steam supply is also controlled by the point of compression in the high pressure cylinders, i.e. when the exhaust valve closes. Once the high pressure valve has closed there is no more steam for the low pressure to use even if the valve is still open; on the other hand, once the input valve for the low pressure has closed theere is no point in keeping the high pressure exhaust valve open as there is nowhere for the steam to go...Many people did not consider locomotives without receivers as true compounds.

 

You can only expand the steam once, whether you do it in one stage or two. So as I understand it, the primary advantage of a compound, at least in the days of saturated steam, was to minimise the heat drop across one cylinder. By getting the same total heat drop across two thermally independent cylinders, you avoid the issue of the cylinder being at some constant, but intermediate, temperature between the high inlet and low exhaust steam temperatures, which means the steam is cooled at inlet on contact with the cylinder walls, but ends up being heated at the exhaust end of the expansion - with associated entropy losses.

Is that correct? (And if so, is it negated by the heat loss to keep all the intermediate metal hot, i.e. the receiver)? For a saturated loco, does compounding allow greater expansion before condensation occurs, relative to an equivalent "simple" - otherwise it feels like a lot of complication for not much thermodynamic gain? It seems to me that, in the UK at least, compounding largely fell out of favour once superheating was introduced and it was possible to run with a high expansion ratio without associated cylinder condensation.

As an aside, did anyone ever build a compound in which the exhaust steam from the high pressure steam was re-heated by the flue gases before passing to the low pressure cylinder? Or is that just further mad complication?

Tom

 

The LNER Hush-hush had a superheater and a reheater.

 

I could be wrong on this, but i think the receiver on Francis Webb's three cylinder compounds was inside the smokebox, and a desire to gain some heat from there was the reason. Doesn't sound too efficient but the effort was there!

 

Thanks for that. The Douglas Self site says "An intermediate superheater (surely really a reheater) was added to raise the LP steam temperature by 100 degF" but it isn't quite clear from that whether he means "it was added as part of the design" or "it was added following initial trials". Either way it is interesting.

(Also of interest - and moving away from compounding: it looks like it also had a form of pre-heating the air before it was fed under the grate to the fire. Shades of the Hammond air pre-heater fitted to a Stroudley Gladstone for a while).

Tom

 

i stand to be corrected here but ..

the original 5 Smith/Deeley Compounds had separate cut off adjustment for the HP and LP cyls. in that form they did some spectacular work.

when it came to the production engines the cut offs were linked so individual adjustment was not possible . other changes were made , for instance none of the cyls were subjected to the full BP ., and the first 5 were rebuilt to conform.- probably done to make the machine easier for the ordinary driver to manage .

it would be very interesting to have a newbuild compound in the original configuration.

LMS 2968 will no doubt correct my errors

 

au contraire, I think you're right, but the Midland Compounds aren't my specialist subject either. But certainly the production batch were much simplified from the first five, and these were then made to conform. An oddity is the numbering. The production batch were numbered from 1000 onwards, but the Midland renumbering scheme saw the first engine take that number and its sisters became 1001-04, so all the production engines had to have five added to their first numbers.

For contemporary spotters, could you claim 1000 twice, and how would you know you'd seen both?

 

I think these also had some common ground with NER no. 1619, but I don't know whether this had separate cut-off adjustment.