Current and Proposed New-Builds

As I've repeatedly banged on about a new build "K", I hope you'll accept that there was no intention to come across as being harsh on Lawson Billinton, whose robustly built locos seem to have withstood the rigours of operation far better than many others. If my posts come across thus, it's as I've been attempting to understand why little more of LBSC standards, other than the pneumatic pull-push system, were adopted by the Southern. Given their survival, for the most part virtually unmodified and in seemingly sound condition, until en-bloc withdrawl at the close of 1962, the history of the "K" class is of particular interest.

Never having clapped eyes on a GA drawing, I confess I was unaware that the "K" continued the practise of short travel valves, which Churchward's work in the early years of the century clearly demonstrated were inferior in performance to the long travel variety. Billinton was certainly in good company in that department!

Were a new build to become reality, I wonder whether it would be worth adhering to original spec, in the name of authenticity, or adopting something of the A1SLT approach by subtly "breathing" on the front end to devlop the full potential of this demonstrably sound design? There are arguments for following either route, though, even for heritage only operation, I confess a preference for ironing out as many known issues as possible, short of noticeably altering the external appearance.

The matter of redesigned piston valves raises another question, namely that optimum best practice would include straight ported valves, rather than the older Z ported layout and that this would make a considerable difference to the sound of the exhaust. Ahead of the howls of protest, might I point out that my previous new build ramblings have always indicated preference for a "next in sequence" (354*/B354/2354/32354) rather than a straight recreation. Coal isn't going to get any cheaper! Common sense, or tin helmet time?

*OK, probably not 354, coz that would have to be to original LBSCR loading gauge.

 

I'm biased, obviously, but I rather liked Thompson's Q1 and the story behind it is something else!

 

Not forgetting that the S15, even in its pure Urie form, was a better and I think more economical heavy goods machine than either of the Moguls. The K suffered from being bookended by those two excellent classes - the N for general work and the S15 for heavier duties where the track permitted.

Still good engines, by all accounts.

 

Short travel valves are less of an issue at low speed, because the ports remain open for sufficient time anyway. So you could redesign, but I doubt you'd see any noticeable improvement in efficiency or coal consumption at heritage line speeds.

When it comes to new builds, I tend towards the "build a replica" school of thought rather than "improve on the breed", especially where the design is for heritage line running. (I can understand the logic of the improvements in the P2 project for the far more exacting requirement of the mainline). Part of the purpose of a new build is to help experience a long-lost design, and to my mind that ought, as far as practical, include any operating quirks. After all, you could build a replica of "Sans Pareil" and add a multi-tubular boiler, since it would steam better and the advantages are well proven. But it would no longer be a replica of "Sans Pareil", so what would be the point?

Tom

 

When talking about a new build "K", I've made it clear I'm speaking in terms of a loco for heritage railway use, not a museum replica, therefore "Sans Pareil", or anything else from that era, wouldn't even be on my list, so that aspect of Tom's agument for simply reproducing an original isn't vaguely applicable.

The "K" is of reasonably modern concept, certainly in LBSC terms. Had one of the original locos survived, I'd be numbered amongst those arguing for any restorations to be as per original spec, though probably to the SR composite loading gauge. Since none did, my emphasis switches to what is in keeping with a line, yet suitable to take steam operations on to the end of this century.

This approach places emphasis on optimising a design's potential. Under these conditions, I see no virtue in perpetuating identifiable issues in what is needs to be more than a fully operational 12in/ft scale model. Before anyone screams "heresy", this is no more than the approach the L&B have taken with "Lyn", a recreation designed with today's requirements firmly in mind.

 

I'm afraid I struggle with that argument. Steam locomotives are fundamentally out of date. Many of the locomotives in daily service on heritage lines today are over 100 years old, and a few are nudging towards 140 or more. So I am struggling to see what else is needed to make them "suitable to take steam operations to the end of this century", given that whatever you do, they are fundamentally obsolete already.

The business model is predicated on the availability of significant volunteer labour. The other existential threat is probably the supply of suitable lump coal. If either of those disappear, no amount of monkeying around with valve gear design is going to keep heritage lines in operation. If on the other hand, both of those survive, the existing designs (either originals or replicas) are perfectly adequate without needing any redesign.

Tom

 

Mmm, maybe. But I bet you're in favour of using modern lubricants in preference to the considerably inferior ones available 120 years ago. I agree there's probably little point in redesigning valve gear, especially as its probably a considerably more difficult exercise than the average enthusiast would think. But if the original design had marginal bearing designs and lubrication arrangements that could be improved on with what we know now, what what you think about upgrading those? I suppose its all down to the definition of perfectly adequate, but some lines back in the day seemed quite prepared to accept poor mechanical reliability on the grounds that it was how things were, but if we're building new shouldn't we seek to improve that sort of thing? It will be more and more the case, of course, if you go back in time.

 

I might be able to shed a bit of light on the Walschaerts -v- Stephensons gear debate and re the LBSCR J1 (Abergavenny) and J2 (Bessborough) and the LBSCR K class.

Walschaerts valve gear is pretty much indifferent except for outside verses inside admission requiring the combination lever top pins being reversed. It can easily be applied to long travel valves because the lap + lead element of travel is provided by the combination lever and crosshead.

In Stephensons there are 2 types of gear using different types of expansion link; the launch type link, and the loco link. Both have quite different geometrics and idiosynchrasies.

Any discussion of Stephensons valve gear requires an appreciation of whether loco or launch type links were used. There is also a significant effect on suspension offset with each type of link depending on whether outside admission or inside admission and whether rocker arms reversing travel are employed.

So for example a Churchward/Pearce valve gear with launch type links works very well with direct drive and inside admission piston valves. Very little suspension offset required.

Similarly a loco links valve gear with direct valve gear works very well with direct drive and outside admission (usually slide valves). Very little suspension offset.

Lead is simply a product of eccentric throw compared to eccentric rod length and can be mathematically calculated or these days you can use a computer valve gear simulator.

Don Ashton (who is the leading expert on valve gears) considers the GWR Stephensons arrangement to produce near perfect valve gear events. The 'constant' lead in Walschaerts is a hindrance when starting.

A loco link in Stephensons gear for long travel valves becomes problematic unless a rocker arm is introduced increasing the travel (as per the SECR E1s). The eccentric throw otherwise becomes too big and creates far too much friction. Hence why Churchward used the launch type link that requires less eccentric throw, and the design adopted by Churchward (worked out by Willie Pearce) could be easily standardised and suit different locos via rocker arms that did not reverse travel.

I havent seen the J1 'Abergavenny' drawings for some time (I believe the GA is at York NRM) or that matter the LBSCR K class GA, but I reckon they both used loco links. The H2 Atlantics certainly use loco links. The H2 Atlantic valve gear is fundamentally flawed because they use loco links and direct drive and inside admission piston valves requiring a large suspension offset creating large die block slip to get equality of events. The original gear was for the original GNR Atlantics with outside admission slide valves where very little suspension offset was required.

The Js both J1 and J2 were hampered by being constipated due to small piston valves as were the B4X rebuilds, and short travel valves. The effects on the K class would have been less pronounced due to it's more hum drum duties on freight, but with inside admission piston valves and direct drive and loco links it would also have had hampered valve gear.

Suspension offset = the amount by which the bottom of the lifting link suspension point on the expansion link is moved for or aft of the centre line of the gear.

Cheers,
Julian

 

Yes, I take the point, but I was commenting more about the possibility of doing a fundamental redesign on the valve gear.

If I had was pushed into a corner, I'd suggest that invisible improvements that enhance the reliability of a loco are justified (as in your lubricating oil example); improvements to the capability (such as redesigning valve gear) I struggle with, not least because the capability of steam locos on heritage lines (i.e. the ability to shift a viable load at heritage line speeds) is rarely an issue; and those kinds of fundamental deviations from original drawings would not only get away from the spirit of creating a replica, but would also likely lead to additional cost that would be hard to recover by very marginal efficiency savings.

Tom

 

Fundementally obsolete?....... Unarguably so, however I'm standing by the points I've made. By your statement Tom, can I assume you'd have been more in favour of the L&B recreating "Lyn" to original spec rather than the path of improvement adopted?

As originally stated a few posts back, I can see valid and coherent arguments on both sides. My own suspicion is that close examination will bear out the case for the odd tweak here and there. This is no more than is already being done by some projects.

Regarding coal and other fossil fuels, my thoughts on that have been posted elsewhere. Sufficing to say we appear to agree supply issues are likely to circumscribe operation of conventionally fuelled steam operations at some point. Taken to one possible conclusion, your arguments could be taken as a case against any further newbuilds. Again, there would be a logical case to say "That's it...... the working steam loco is dead", but obtaining 100% acceptance of such a sad event seems highly unlikely.

Unless one or other of us scoops a substantial jackpot (Bob willing!), the performance of the Keefe Baldwin will just have to serve to clarify matters either way. Bitter experience, however, suggests we could continue argue our respective jackpotless cases until our keyboards wear out. I think we're just going to have to agree to disagree on this one.

Edit: In light of Julian's compendious post (for which, thanks), I'd best just clarify that my original point questioning the Lawson Billinton's selection of Stephenson's gear was solely with respect to any additional maintenance costs incurred resulting from any inside gear, when J2 No.326 had been turned out with outside Walschaerts and purely an enquiry into the reasons for the SR not developing the Brighton design. In any event, Tom's point about the kits of wartime Maunsell moguls, kicking around to be had for a song, seems to provide the answer to that one.

The issue of any new build is seperate in my mind, and I'm not for a minute suggesting redesigning any newbuild "K" with Walschaerts (or Caprotti for that matter!)... inside or out! That would indeed be a tweak too far.

 

Ah, but it wouldn't be much of a forum were everyone to agree all the time ...

But anyway, we're probably boring everyone else, so yes, agree to disagree.

Tom

 

At risk of stringing out the "is improvement worth it" debate, I think the Ffestiniog double Fairlies make an interesting case study. I'm pretty sure they weren't superheated with piston valves as built, or even as preserved, and yet now they can be found with both features. The difference is, I suppose, that the Ffestiniog can't just bring in something bigger - the double Fairlies are it - and they have to run reliably for most of the season.

On a more frivolous note, the Unlimited class racers at Reno are certainly obsolete technology, yet that doesn't stop the owners honing their mutant Mustangs and Sea Furies. If it's somebody's own money and they wish to put improved piston valves on their West Country, or tweak its inside valve gear geometry a bit, since there are a few of them around I wouldn't consider it objectionable and I wouldn't have a say anyway. Sometimes improving something without violating the spirit of its obsolescence can be an interesting engineering exercise.

 

I'm amused by "fundamentally flawed"... Like many locos with sub-optimal valve gear, they performed a highly satisfactory job of work for many years...
I'm an engineer myself, and am always dissatisfied by sub-optimal design. But the harsh reality is, often it doesn't matter!

 

Interesting that Blodge have opted for introducing monobloc engine units for Fairlie bogies when experience with Gresley's V2 saw the same technology removed in favour of individual castings.

Why does Brunel's comment on brakes suddenly leap to mind?

 

Talyllyn on the TR had its loco links changed for launch links with the same direct drive and slide valves without appreciating that this change was far from optimal ie 'fundamentally flawed'. The same could be said of the Fairlees converted to inside admission piston valves on the FR retaining loco links compared to outside admission slide valves.

The fundamentals are really quite simple but rarely understood!

Direct drive loco links outside admission

Direct drive launch links inside admission (piston valves)

Indirect drive loco links inside admission

Indirect drive launch links outside admission

The Maunsell and Bulleid Q and Q1s used loco links with indirect drive and outside admission which was 'fundamentally flawed'. Showing Maunsell and Bulleid had no grasp of Stepehensons gear fundamentals... had Holcroft been around in the drawing office at the time this error would not have occurred!

Cheers,
Julian

 

Different operating requirements and, like all engineering solutions, a compromise one way or another.

Thompson reintroduced the three separate cylinders idea initially on the P2 rebuilds - Gresley had opted for monobloc on the P2s and V2s as this introduced an extra level of strength across the frames.

However this added considerable time to overhauls as you'd effectively have to lift the boiler to do the whole cylinder block.

Thompson in WW2 wanted to reduce overhaul time and intensity - so separate cylinders on the rebuilt and new Pacifics and V2s became the norm. The big drawback here was keeping the cylinders tight to the frames.

On the A2/2s the three separate cylinders were also over the frame partition. So in theory a good idea to have separate cylinders but the working loose of the separate cylinders over time also affected the frames, leading to cracking and possibly warping.

The arguments for and against leave me to conclude that both Gresley and Thompson were right in theory for monoblocs and separate cylinders, but operating conditions and the difference in circumstances made both solutions a compromise ultimately.

V2s with monoblocs and separate cylinders both made it to the end of steam on BR. That to me says both ways were good solutions ultimately.

One wonders why there was no wholescale standardisation though - Edward Thompson's words "I have much to do and little time to do it in" ring true.

 

The suggestion seems to be that there were few engineers who 'understood' valve gear resulting in many designs (Q and Q1) being fundamentally flawed

 

The Q1 ran 75 mph backwards with mr Bulleid on board.
Flawed valve gear?

 

Mr Riddles dreamed as old man of having made the Clan class as Caprotti 4-8-0 and thus having beaten mr Chapelon.
LNER could have done before war and would have looked much better

https://imgur.com/mv7Jp6V