P2 discussion (split from train length reduction thread)

I think the hamstrung A1-style front-end of the P1s may have held them back from passenger work; however, there is nothing inherently wrong with a 5'2" wheel loco being used on such. As an example, the NZR JA class 4-8-2s were built for a very similar duty to the putative use on passenger trains to Aberdeen: hauling heavy passenger trains over twisty and heavily graded lines. Yet when they were given their head on the flat, straight Canterbury plains, they were often recorded at over 70mph and anecdotally at up to about 80mph - from a loco with 4'6" driving wheels and 26" stroke pistons!

I suspect the big problem against even trying such a use was more to do with the small numbers. In order to make a transformative improvement on a route, you need sufficient locos to run all the diagrams - if you only have one or two, loads and schedules still have to be kept at levels that the original locos can maintain, which then means any putatively improved locos are under-utilised for their capabilities. I suspect with such a small class; the mixed traffic V2s proving to be a big success; new-build A4s taking the cream of the long-distance passenger work and a major programme to rebuild A1s into A3s, it simply wasn't worth the development effort that would have been needed to improve the P1s and then build sufficient more to have been used on the route - given that workshop capacity and capital would have been scarce even if a suitable scheme could have been developed. So a "near miss" becomes a "what might have been".

Tom

 

S.A.C. Martin has already alluded (in post 16) to Gresley's fondness for specific designs for particular duties, sometimes in small numbers. Given the wish for something a bit better than A1s/A3s for the Aberdeen line, and the existence already of the P1s, it is hardly surprising that he came out with the P2s. The only thing that he might have done but apparently didn't would have been a few experimental runs to and from Aberdeen with a P1, analogous to the trial high speed runs with an A3 (?) before introducing the A4s. Even if he had done that, the next step would have been a new design for that duty, not additional P1s.

 

I'm sure a P1 designed and built a generation or two later may well have benefitted from improved valve events and roller bearings as did the NZGR Ja class but it was a product of the 20s and it can only be judged as it was built. It's test on passenger seems have been enough to convince those who mattered that it was not a suitable passenger design. Maybe that in itself was proof enough that something a bit different was needed. For laymen - and I include myself - to post here that this should have been done, that should have been altered and so and so should have tried this is somewhat pointless IMO.

 

It is striking, to an amateur enthusiast like me, how often known problems just weren't fixed... Seems amazing that the P1s were left with known inadequate valve gear or so long or, to balance lines, the GWR 47s never had any work done to resolve their "nosing about" at speeds above 60mph, and Holcroft's book has some similar examples from the Southern, and there are examples from the LMS too... But my long ago industrial experience was in the light side of the chemical industry, proper heavy engineering and the finance/production/maintenance side of it is pretty much a closed book to me.

 

One reason for leaving less-than-ideal features unchanged may have been simply that there were other things to do that were, or were perceived to be, more important: the "up to your arse in alligators" situation. Another could have been the scale of the organisations: someone at the sharp end being well aware of a problem and a possible solution but unable to get the message through the hierarchy. I hasten to add that these are speculations on general principles, not based on any actual knowledge of the goings on in the railway workshops and drawing offices.

 

One reason why the P1s did not receive the long lap/long travel gear was the speeds of the trains they were working. While it's true that there were gains to be made even here, these really come into their own at higher speeds since they increase the port opening to admission and exhaust, allowing more steam to enter the cylinder in the time available. At low speeds, this isn't an issue, but becomes one as speed rises and the ports are open for a shorter period. Allowing a greater flow of steam through the ports as the available time reduces paid bigger dividends with the A1/A3s than it would with a goods engine.

So we had the classic Catch 22 situation: they didn't get better valve evnts because they weren't needed at the speeds being run; and they weren't given faster work because the valve events wouldn't allow it.

 

I think that various posts on here have given good reason why a P1 would be no use as a passenger loco but you just don't seem to accept this fact. Your thoughts about small wheels being good for acceleration are ill founded. Acceleration is simply a function of tractive effort, not wheel size. These locos had a good starting T.E. but this, in itself is a function of m.e.p., cylinder size and wheel diameter. Note the use of m.e.p. (mean effective pressure) and not boiler pressure. It is a fundamental that m.e.p. reduces with rotational speed and, with small wheels, the rotational speed increases quite rapidly so the fall off in T.E. is going to be much faster compared with a large wheeled loco. The fall off in m.e.p. can be reduced (but not eliminated) with long lap valves allied to large and carefully designed steam passages but long lap valves on there own aren't the solution. The P1's had none of these necessary design features and would simply choke themselves at any reasonable rotational speed. At 50 mph, a short lap loco with 5'-0" wheels would be lucky to produce an m.e.p. of 20% of boiler pressure, with a T.E. proportionate to that. Not much good for hill climbing or acceleration.
With any steam engine, maximum torque occurs at low revs and that fall off in torque is less with big wheels, all else being equal. Steam loco design is always a compromise. If you want high T.E. at low speed, small wheels help you achieve this. If you want high T.E. at high speed, big wheels are a better bet but the most important thing to consider is steam circuit design.

 

Consider the "Decapod" to understand this blind alley of development.

 

No I have asked for comparisons and some information. Fully prepared to change my view accordingly and I've been very happy to see your further thoughts which are informative.

Thank you for this, I will consider it further and probably ask more questions if I may?

One has occurred though - why do we accept the 9F was fine and yet the P1 is not - specifically? Surely the 9F is not entirely dissimilar in its set up as a goods engine?

 

The 9F had long travel, long lap valves for one thing so that helps. I'm sure others will be able to point out other advantages the design had.

 

As Spamcan says, much better valve events and probably better steam passages. Certainly the 9Fs were more than good enough at 60 mph and possibly 70, but a problem was their excellent ride qualities, which would have masked a lot of issues going on below footplate level. The recipricating masses are only partly managed by the balance weights, and the forces generated rise with the SQUARE of velocity, so these forces would be much, much greater at 90 mph than they were at 70, which is why the upper management clamped down on their use on passenger workings, even though crews would not have felt any unease, and in contrast to the enthusiats' views. A similar situation would become manifest on the P1s as speeds rose.

 

Sombeody worked out that at 90, a 9F's wheels would be doing in the region of 1800 rpm and that was mighty close to 4468 at 126 mph, i.e. 1880 rpm. IIRC another factor in the decision to limit their speed.

 

It would be in proportion to the ratio of the wheel diameters, i.e. the "equivalent" speed of a 9F to an A4 doing 126mph would be 126 * 60/80 = 94.5mph (give or take any tolerance in wheels being under their nominal diameter). Peak piston speed on the 9F would actually be marginally higher on account of the slightly longer stroke. Effectively, a 9F at 90mph is equivalent wheel rpm to an A4 at 120mph and equivalent peak piston speed to an A4 at 129mph. (90 * 80/60 * 28/26)

Lubrication also becomes a factor: one reason that many Victorian passenger engines had very large wheels was to minimise piston speeds as lubrication was a bit suspect, especially cylinder lubrication. Twentieth century advances in lubricating oil technology allowed a general reduction in driving wheel diameter while maintaining (or even increasing) maximum speeds. (Of course, that is only one aspect and enhancements in steam distribution were also very important as others have noted).

Tom

 

Mr Steve..then it is I believe broadly accepted that the smaller wheeled locomotive will reach its peak output at a lower speed, as you have I believe eluded to above.
Therefore ;
if two locomotives are identical apart from wheel diameter; the smaller wheeled loco by virtue of its greater tractive effort will have greater initial acceleration but will ultimately not be capable of as great a speed...
But Also If two locomotives have identical tractive effort but different wheel diameters (with size /no of pistons being different) then as long as they are driven/ wound back proportionately to the wheel speeds in rpm to give roughly the most efficient use of steam / expansive working then the smaller wheeled locomotive will reach its optimum power output speed first (though initial acceleration would be identical).... makes sense ?

 

Sounds about right, yes!

 

The P1 design owed something to the original O2, the coupled wheels wheelbase was the same and the valve gear 5 1/2" travel with 1 1/4" lap. The Group Standard O2 had a 6 5/8" valve travel with 1 9/16" lap but the P1 remained unaltered. The LNE could have improved these engines but shorter, faster freight trains were becoming the order of the day. Track capacity, integration of freight with passenger traffic, the usual suspects.

The LNER was never a wealthy company and the lack of resources was responsible for some design decisions. Looking at the P2 Spencer wanted to fit a Bissell truck arrangement, he had formed a high opinion of it whilst the River class was under testing in 1927. The swing link was cheaper to construct and appeared to work well enough on the GN mainline, so swing link it was. Windle, the Doncaster Chief Draughtsman realised that changing the leading truck to a design fitted with sprung side control would solve the bulk of the problems facing the design. This would have been far cheaper than rebuilding as Pacifics and would not have impacted on what was so valuable in the class, tractive effort and adhesion. The frame spacing would have offered the chance for controlled lateral movement of a coupled axle or two.

The 9F benefited from a long wheel base as well as the cylinder and valve design. It rode well and could maintain a high rotational speed. The days of the 8' driving wheel for express passenger work are gone. 6'2" will normally be seen as sufficient today. The slow loose coupled freight train has gone, faster, fully fitted is the order of the day. And 5' is closer to 5'2" than it is to 4'8". It was realised long ago that faster was the way forward. Hence bigger boilers, the occasional improved exhaust, tweaks to valves and cylinders to obtain higher power output and better speeds. And adhesion, if you have to deliver the power there is no easy escape from weight on the driving wheels.

Thinking about the 9F, it should be possible to modify the design to fully flanged if fitted with Krauss-Helmholtz, Zara or similar.

 

Certainly more room to work withbetween the cylinders than any other large locomotive in the uk...