GCR compound Atlantics

[John of Paris]

I have always been interested in the Smith 3-cylinder compounds and more recently in the little-known Robinson Atlantic versions (GC class 8D, then 8E; LNER C5).

In the table at the bottom of the Encyclopedia page dealing with these the valves are described as "slide for the inside and 10 inch piston for the outside, whereas in line with all Smith 3-cylinder compounds the opposite was the case as it says in the main body of the text.

That small niggle aside, the history of these engines does seem to be something of a mystery. Cecil Allen in his "British Atlantic Locomotives" book complains that, "Very little was ever made public about the workings of the Robinson compounds" adding that: "(the valve gear) presumably allowed independent adjustment of cut-off in the h.p. and l.p. cylinders respectively."

I find that last statement questionable for the following reason: Smith's own rebuild of NER 1619 and the very first two Johnson Midland compounds did have independent cut-off with two parallel reversing rods worked by a single handle that could be so locked as to give either simultaneous or independent adjustment. However it was found in testing the Johnson Compounds that the driver only tended to make simultaneous adjustments so the side-by-side rods were replaced by a single one when the next four Johnson compounds came out. This was perpetuated by Deeley, which seems a pity as these engines were hamstrung at speed, exactly the problem the GCR Atlantics were said to have, which suggests to me they too only had simultaneous adjustment. One from the first batch of Johnson compounds (no. 2632) with independent cut-off reached 92 mph on test in 1902 and sustained it for two miles, a performance no other British compound ever approached as far as I know.

Anyway I would be very grateful if someone could tell me where I could access a general arrangement drawing of a Robinson compound to clear up this question of reversing rods.

[Atlantic 3279]

Whilst I haven't looked up the place of construction of these locos (Gorton or Beyer-Peacock?), as nobody else has offered a reply I wondered if you had enquired of the Museum of Science and Industry in Manchester, or checked their website. I believe they have many drawings from B-P.

[John of Paris]

Thanks. Following up on your post, I visited the Manchester museum web site but found nothing of help. Do you have a direct contact?

As far as I know the four compounds were built at Gorton, but still Robinson may have initially had the bulk of the design work done at BP. Certainly his predecessors on the MS&L, Parker and Politt worked very closely with the big locomotive manufacturers. Close examination of a drawing of the standard simple expansion version of the Atlantic on page 85 of O.S. Nock's "British Locomotives in the 20th Century", vol 1 suggests that the design was made right from the start with the intention of building a compound version. The only way to easily identify the compounds in a photo (apart from the nameplates) is the Smith regulating valve on the RH side of the smokebox, otherwise they look pretty well identical.

[Atlantic 3279]

Sorry, I don't have a direct contact. I subsequently checked in the relevant RCTS book and it too states that the locos were built at Gorton, not BP, and mentions that a couple of them were built on frames that were initially ordered for simple atlantics and then altered to suit compounds. Logically (?) enough too, the RCTS book states piston valve HP, slide valves LP. Do you have E. M. Johnson's books of the locos of the GC. There are various complete and part GA drawings in the back of one or other of the two volumes. I suspect one is an atlantic, though I cannot remember if it is a compound or a simple.