RETURN TO INDEX

DONCASTER WORKS

LOCOMOTIVE DRAWING OFFICE

IVATT DESIGNS

Inheritance

When Ivatt arrived at Doncaster in March 1896 he inherited just over a thousand engines.  The exact number cannot be determined as there were quite a few on the Duplicate List.  Stirling’s own engines were still intact, if one excludes the four puny 0-4-2WTs thoroughly rebuilt in 1873-74 and given Doncaster Works numbers.  However, there were still large numbers of Sturrock, and indeed pre-Sturrock, engines running around which Stirling had continually refurbished, so most had his standard 4ft.-2½in. diam. boilers.  Stirling had been about to introduce 4ft.-5in. as the new standard diameter for his boilers, despite having first applied it back in 1871 for his 174 Class 0-6-0 mineral engines. Ivatt straightaway adopted 4ft.-5in. for future use, and the boiler with this diameter, 10ft.-1in. long barrel and 5ft.-6in. long firebox became known as his “standard boiler”.  This was applied to numerous Stirling and Ivatt classes, and indeed was used by Gresley for his Class J23 (LNER J50) 0-6-0Ts.

Ivatt soon started withdrawing older engines, and by the end of 1899 he had scrapped one vintage 2-4-0 dating back to 1849 (No. 100), two vintage 0-6-0STs from the same era, five Sturrock 2-4-0s from 1865 (Nos. 250/4/5/7/60), two Sturrock 2-2-2s (Nos. 266A/69), one Sturrock 0-4-2WT and thirty Sturrock 0-6-0s; and if that was not enough to send out clear signals of his intentions, one Stirling 4-2-2 (No. 60).  There were still some eighty-seven Sturrock 0-6-0s around, which took until 1910 to get rid of.  With this in mind Ivatt could be excused for ordering eighty-three 0-6-0s from contractors as well as building further similar engines at Doncaster.  But if only he had waited a while, for in 1898 he introduced 4ft.-8in. diam. boilers, firstly on his first 4-4-2 6ft.-8in. passenger engine (later designated Class C1) No. 990, then on some 4-4-0 6ft.-8in. passenger engines (later designated Class D1).  It was left to Gresley to introduce the 4ft.-8in. diam. equivalent of Ivatt’s “standard boiler” which resulted in a lengthy reboilering process in subsequent years with his Diagram 8 boiler.  On the bonus side, it will be noted that Ivatt seemed to favour 6ft.-8in. driving wheels, and this became the Doncaster “standard” for express passenger engines to the end of steam.

Ordering system

For accountancy purposes, a new system of recording Engine Orders (EO), Stores Orders (which were mainly for boilers) (SO), Tender Orders (TO) and Engine Repair Orders (ERO) was introduced about January 1897.  Whilst there is no specific evidence, it is seems likely the system was devised by Matthewman, who had been Stirling’s Accountant.  Current engine orders on the books were then numbered EO201 to EO204, boiler orders SO201 to SO206, and tender orders TO1 to TO4.  Earlier engine and boiler orders whose construction had been completed needed no recognition, though there were some new boilers in stock that had not yet been allocated.  In particular there were four domed boilers for Stirling Single-wheelers (fitted between August 1897 and June 1899), and nine Ivatt domeless "standard" boilers for Stirling 0-6-0s (fitted between June 1898 and December 1899).  New boilers were afterwards either marked off to a specific EO, or (if built as replacements) to a specific ERO.  Later, Ivatt Engine Orders ran from EO205 to EO268, Boiler Orders from BO500 to BO607 in a new series, and Tender Orders from TO5 to TO28.

Early orders

After placing orders for both his first Klondyke 4-4-2 and 1321 Class 4-4-0s, his next order was for a 7ft.-6in. 4-2-2 Single-wheeler, based largely on the final Stirling 2-2-2s, but incorporating “improvements”.  Significantly it had a 4ft.-5in. diam. boiler and it seems strange that Ivatt should have put so much effort into persevering with this design, until it is realised a 4ft.-8in. diam. boiler was unsuitable for an engine with such large driving wheels.  It was also to be his last new design with a 4ft.-5in. diam. boiler.  The engines only lasted until 1917.

After placing orders for both his first Klondyke 4-4-2 and 1321 Class 4-4-0s, his next order was for a 7ft.-6in 4-2-2 Single-wheeler, based largely on the final Stirling 2-2-2s, but incorporating “improvements”.  Significantly it had a 4ft.-5in. diam. boiler and it seems strange that Ivatt should have put so much effort into persevering with this design, until it is realised a 4ft.-8in. diam. boiler was unsuitable for an engine with such large driving wheels.  It was also to be his last new design with a 4ft.-5in diam. boiler.  The engines only lasted until 1917.

 

Drawing Office

 

In August 1901 Engine Repairs Registers were started, with one page per engine.  Details of existing engines were entered in the books by a young clerk, Walter William Lake (1875-1958), with repair details copied from the Works Stock Books, so the entire process shouldn’t have been too difficult.  The current stock as at August 1901 was entered into 3 new registers.  Book 1 covered Nos. 1 to 499 (with several blank pages), Book 2 covered 500 to 1000 and Book 3 covered 1001 upwards.  Afterwards new engines, for example those numbered below 500, were entered in Volume 1 if the page was blank, otherwise in Book 3, onwards.  In addition the Works maintained Works Repair Books with entries made in date order. Around 1957 Edward Windle (1892-1960), who was by then Chief Draughtsman, told Kenneth Leech (1892-1995) that around 1942 Edward Thompson ordered the destruction of all the old material. He managed to secrete the Engine Repairs Registers, thereby running a serious personal risk had Thompson found out. 

 

These Books also included information concerning repairs carried out at the major sheds.  For example Book 1 recorded that Nos. 16, 193 and 389 (at least) were reboilered at a general Repair at Peterborough (July 1889).  In the same month No. 63 received new cylinders at Peterborough.  Apparently the resources at Doncaster Works were being stretched around this time. 

 

One difficulty for historians is that the clerk only recorded an engine’s history from its official renewal date, where applicable.  This mainly (but not exclusively) affected the Stirling 4-2-2s (e.g. No. 1), and enabled Stirling to inform his Directors that the average age of their top line engines was only such and such, whereas in reality it was greater.  The Engine Repairs Registers were deposited at the National Rail Museum around 1975, after being in my possession for safety for a numbers of years, fearful they would disappear after the Locomotive Drawing Office (LDO) at Doncaster closed.  They are now available for inspection on request in Search Engine.

 

In addition to the Engine Repairs Registers, in August 1901 the LDO commenced keeping details of the Works construction orders for engines, boilers and tenders, mentioned earlier.  At that date surviving details of 1896-1900 orders were somewhat sketchy and the clerk had difficulty deciphering them, evident by the frequent crossings out in his books, with the result that my interpretation is at best still an interpretation.  The 1896 orders in fact pre-dated the 1897 recording system, so were even harder to track down.  As far as I can make out, these tattered copies of the 1896-1900 orders have not survived.

 

Engine Classifications

 

A meaningful engine classification was introduced in June 1900.  A letter was allocated to each wheel arrangement (tender and tank versions taken together), in a somewhat arbitrary fashion as one would have expected A to have been allocated to the Atlantics.  This was followed by a number for each variation.  New wheel arrangements introduced subsequently by Ivatt were 0-8-0, 0-8-2 and 0-6-2, whilst Gresley introduced 2-8-0 and 4-6-2.  The latter fortuitously could be allocated the letter A, the 4-2-2s having by then become extinct, in running stock at least.  Several wheel arrangements were not perpetuated by Ivatt, 2-2-2, 0-4-2, 0-4-4 and 2-6-0, though the latter was taken up by Gresley.

A 4-2-2: A1, Stirling 1003 Class [1894]; A2, No. 1 Class [1870]; A3, A2 with Ivatt boiler [1896]; A4, Ivatt No. 266 [1898]; A5, Ivatt 267 Class [1900]

B 2-2-2: B1, Stirling No. 92 [1870]; B2, Nos. 232/8 [1885]; B3, 234 Class [1888]; B4, B3 with Ivatt boiler [1898]; B5, No. 4 Class [1868]; B6, Sturrock double-frames; B7, B5 with Ivatt boiler [1897]

C 4-4-2: C1, Ivatt 990 Class small Atlantic [1898] and No. 251 Class large Atlantic [1902]; C2, 1009 Class side tank [1898]

D 4-4-0: D1, Ivatt No. 1321 Class [1898] and 51 Class [1911]; D2, 400 Class [1896]. 

E 2-4-0; E1, Ivatt 1061 Class [1897]; E2, 86 Class [1874]; E3, 261 Class [1871]; E4, E3 with Ivatt boiler [1899]; E5, Sturrock double-frames

F 0-4-2: F1, Stirling 218 Class [1868]; F2, 18 Class, larger version of F1 [1868]; F3, F2 with Ivatt boiler [1900]; F4, 126 Class Metro-tank [1868]; F5, Sturrock double-frames; F6, F4 with Ivatt boiler [1900]; F7, 501 Class saddle tank [1874]

G 0-4-4: G1, Stirling 766 Class bogie side tank [1889]; G2, 120 Class Metro tank [1872]; G3, 658 Class bogie side tank [1881]; G4, G3 with Ivatt boiler [1898]; G5, Nos. 629/30 Metro tank [1880]; G6, G1 with Ivatt boiler [1902], reverted to G1 [1905]

H 2-6-0: H1, Baldwin 1181 Class [1899]

J 0-6-0: J1, Sturrock 5ft.-3in. double frames; J2, 5ft.-0in. double frames, 26in. stroke cylinders; J3, 5ft.-0in. double frames, 24in. stroke cylinders; J4, Stirling 174 Class, 28in stroke cylinders [1872]; J5, Ivatt 1081 Class [1896]; J6, Stirling 171 Class, 26in. stroke cylinders [1872]; J7, 369 Class, 24in. stroke cylinders [1869]; J8, J7 with Ivatt boiler [1899]; J9, Stirling/Ivatt 1021 Class [1896]; J10, 374 Class, 4ft.-0in. wheels [1883]; J11, 392 Class saddle tank [1868]; J12, J11 saddle tank with smaller firebox [1899]; J13, Ivatt 1201 Class saddle tank [1897]; J14, Stirling 921 Class saddle tank [1892]; J15, 494 Class saddle tank [1874]; J16, J17 with Ivatt boiler [1897] and J15 with Ivatt boiler [1906]; J17, 606 Class short wheelbase saddle tank [1876]; J18, 684 Class saddle tank [1882]; J19, 471 Class saddle tank [1872]; J20, Nos. 604/05 enlarged version of J19 [1875]; J21, Ivatt No. 1 Class 5ft.-8in goods [1908] and 71 Class goods [1912]; J22, Ivatt 31 Class 5ft.-2in. goods [1909] and 521 Class goods [1911]

K 0-8-0: K1, Ivatt 401 Class [1901]

L 0-8-2: L1, Ivatt 116 Class [1903]

M 0-4-0: M, Ivatt Nos. 2, 1 rail motors [1905]

N 0-6-2: N1, Ivatt 190 Class [1907]         

Tender Classifications

At the same time a simplified classification for introduced for tenders, as follows:

Class A Ivatt flat-bottom tenders
Class B Ivatt well tenders
Class C Stirling flat-bottom tenders, iron frames
Class D Stirling well tenders, iron frames
Class E Stirling flat-bottom tenders, wooden frames
Class F Stirling well tenders, wooden frames
Class G Sturrock tenders
, extinct February 1922 (and Gresley re-used this classification for his 8-wheel tenders).

Doncaster Works References

These were mentioned in the Stirling chapter.  The Works References for Ivatt engines, as recorded by Geo. Fredk. Bird and only complete as far as his final (1910) volume, were as follows:

Table 8 – Ivatt Doncaster Works References

 

Works
Reference


Introduced

Wheel
Arrangement

Wheel
diam

First
engine

Number
Built

E4

1896

0-6-0

5ft.-2in.

1081

10

E5

1898

ditto

ditto

315

10

E6

1899

ditto

ditto

343

40

EE

1908

ditto

5ft.-8in.

1

15

EE2

1909

ditto

5ft.-2in.

31

20

M6

1897

0-6-0ST

4ft.-8in.

111

12

M7

1901

ditto

ditto

1251

20

M8

1905

ditto

ditto

1271

10

M9

1908

ditto

ditto

1281

10

MM

1906

0-6-2T

5ft.-8in.

190

21

S

1896

4-4-0

6ft.-8in.

400

11

S2

1897

ditto

ditto

1301

20

S3

1898

ditto

ditto

1341

10

S4

1899

ditto

ditto

1351

10

T

1898

4-4-2T

5ft.-8in.

1009

10

U

1898

4-4-2

6ft.-8in.

990

1

U2

1900

ditto

ditto

949

10

U3

1903

ditto

ditto

252

10

LU

1902

ditto

ditto

251

81

LUS

1910

ditto

ditto

1452

10

V

1898

4-4-0

6ft.-8in.

1321

5

V2

1898

ditto

ditto

1326

20

V3

1900

ditto

ditto

1366

20

V4

1903

ditto

ditto

1386

10

V5

1907

ditto

ditto

1396

5

V6

1909

ditto

ditto

41

10

W

1898`

4-2-2

7ft.-8in.

266

1

W2

1900

ditto

ditto

267

11

X

1899

4-4-2T

5ft.-8in.

1501

20

X2

1901

ditto

ditto

1521

30

Y

1901

0-8-0

4ft.-8in.

401

55

YT

1903

0-8-2T

4ft.-8in.

116

41

Z

1902

4-4-2

6ft.-8in.

271

1

ZZ

1905

ditto

ditto

292

1

ZZ2

1907

ditto

ditto

1421

1

MC

1905

RMC

3ft.-8in.

2

2

 

General Arrangement drawings

Table 9 - Engine General Arrangement Drawings

P38 (9/1896) 0-6-0 Nos. 1081-90
P39 (3/1897) 0-6-0ST Nos. 606/10 rebuilt
P40 (8/1897) 0-6-0ST Nos. 1201-10/51-70
P41 (10/1897) 0-6-0 alteration at trailing end, 5'-0" Standard Goods
P42 (6/1898) 0-6-0ST Nos. 1201-10
P43 (8/1898) 0-6-0 1101-73, 313 etc.
P44 (9/1899) 0-6-0 ditto, and 3170 gallon tender, painting details
P45 (5/1900) 0-8-0 Nos. 401-55
P46 (2/1896!) 0-6-0ST Nos. 1046-60 outside elevation
P47 (3/1891!) 0-6-0ST ditto               painting details
P48 (-)             0-6-0 Baldwin Nos. 1181-1200 elevation
P49 (11/1907) ditto                                           plan view
P50 (26/9/1910) 0-8-0 No. 420 superheated
P51 (26/9/1910) ditto                end view
P52 (18/5/1911) 0-6-0 Nos. 21-40
Q53 (1/1897) 4-4-0 Nos. 400 etc
Q54 (7/1897) 4-4-40 No. 1321 etc
Q55 (8/2/1898) 0-4-4T Stirling Nos. 763 etc domed
Q56 (-)              4-4-2T Nos. 1009 etc
Q57 (8/1897) 4-2-2 Nos. 92 etc
Q58 (8/1897) 5-5-2 Nos. 990 etc
Q59 (1/1899) 4-2-2 Stirling Nos. 22 etc, domed
Q60 (10/1897) 2-2-2 Stirling Nos. 21 etc, domed
Q61 (3/1903) 4-4-2 No. 271
Q62 (21/5/1903) 4-4-2 No. 251
Q63 (27/10/1903) 0-8-2T No. 116, large boiler
Q64 (13/1/1905) 0-8-2T Nos. 116 etc, small boilers
Q65 (24/1/1906) 0-4-4T Stirling No. 766, domed
Q66 (19/4/1906) rail motors 1, 2
Q67 (23/7/1906) 0-4-4T Nos. 682, etc, domed
Q68 (2/1907) 4-4-2 No. 292
Q69 (2/1898) 2-4-0 Stirling Nos. 702, etc, domed boilers
Q70 (2/1898) 2-4-0 Stirling Nos. 89, etc, domed boilers
Q71 ex P24 (5/1875) 0-4-2 Stirling Nos. 36, etc
Q72 ex P31 (9/1882) 0-4-2 Stirling Nos. 103, etc
Q73 (11/1907) rail motors Nos. 1, 2
Q74 (3/1908) 0-6-2T No. 190
Q75 (30/4/1908) 2-4-0 Stirling No. 702, spring gear
Q76 (6/5/1908) 0-6-2T No. 190
Q77  (11/6/1908) 4-4-2 No. 1421
Q78 (23/11/1908) 4-4-2 Vulcan No. 1300
Q79 ex P31 (9/1882) 0-4-2 Stirling
Q80 (5/1910) 4-4-2T
Q81 (7/1910) 4-4-2 Large Atlantic superheated
Q82 (8/1910) 4-4-2 Klondyke superheated
Q83 (24/5/1911) 0-6-0 Nos. 1-15
Q84 (19/7/1911) 4-4-2 No. 271 superheated


Running Numbers

Ivatt attempted to number his engines in a meaningful way, and not just fill empty blanks.  But to achieve this quite a number of Sturrock and Stirling engines were transferred to the Duplicate List.  This did not work quite right for his J5 0-6-0s, which filled blanks between 315 and 328, then between 165 and 192, then finally between 302 and 389 in three batches.  His own No. 1 was an 0-6-0 goods engine certainly without the charisma of Stirling’s No. 1.  By the time of his retirement in 1911, the highest number reached was 1580.

Frames

Ivatt frames were 4ft.-1½in. apart at the firebox end, as on all the Doncaster classes with narrow fireboxes.  (The Baldwin 2-6-0s are another story.)  They were usually 11/8in. thickness, but not necessarily parallel throughout their length.  On his 4-4-0s, there was a 1in. “set” (two adjacent reverse bends) ahead of the leading coupled wheels, making them 3ft.-11½in. apart alongside the cylinders, to provide extra clearance for the bogie wheels.  This made it awkward as the distance across the cylinders was greater than in for example the Stirling 2-4-0s.  Slots had to be cut in the 4-4-0 frames to clear the cylinders, with the holes hidden by thin cover plates.  Ahead of the cylinders the frames tapered inwards slightly to become 3ft.-7½in. apart at the front buffer plate, so that they would not be caught by the leading bogie wheels rounding curves.  The frames for his first 4-4-2, No. 990, were formed of two sections, so that instead of the “set”, the front section was bolted to the inner faces of the main frames ahead of the leading coupled wheels to provide the necessary clearance for the bogie wheels.  They did not taper inwards at the front, presumably anticipating that the steadying effect of having trailing wheels would make this unnecessary.  The frames on the remaining standard Klondykes C1 (LNER C2) and Large Atlantics C1 (LNER C1) had the 1in. “set” instead of a separate front end.

The frames of 4-cylinder Klondyke No. 271 had a separate front end section, 1¼in. thick instead of 11/8in. to take the weight of the heavier cylinders, at which point the frames were 3ft.-11in. apart.  They then tapered inwards towards the front to clear the front bogie wheels, where the distance apart was 3ft.-9in.  The engine was rebuilt as a 2-cylinder “simple expansion” engine (July 1911) and appears to have had its original front end frames altered rather than replaced.  The frames of 4-cylinder Nos. 292 and 1421 had the standard arrangement of 1in set just behind the cylinders, but in addition they tapered inwards ahead of the cylinders, presumably having concerns over the front end weight.  No. 1300 built at Vulcan Foundry had its frames parallel throughout their length, the usual 4ft.-1½in. apart, with no provision for bogie side play, considered unnecessary with their smaller diameter wheels.  In later years the front ends of Atlantic frames were occasionally replaced, by whichever method was most appropriate at the time.  One method was to bolt a new front end to the sound rear end, but if the rear section had a “set” this had to be first removed under the press in the boiler shop.  From 1923 another method was to butt weld a new section to the rear section, provided the rear section still had its “set”.

The Atlantics also had an additional outside frame at the rear end, to support the trailing carrying wheels.  These were 5ft.-6½in. apart on the early Klondykes, then 5ft.-7½in. on both the Large Atlantics (which needed the extra clearance for the wide fireboxes) and the last batch of Klondykes (probably simply because they had originally been ordered with large boilers).  The last batch of Klondykes could therefore have been given wide firebox boilers at a later date, but this fact seems to have been overlooked.  The outside frames on the Doncaster engines were 8ft.-5¾in. long, parallel alongside the wheels, then tapering inwards at the front.  From 1902 the outside frames were 8ft.-91/8in. long, with a vertical front edge, to make room for the longer springs.  On No. 1300 the outside frames were only 6ft.-0in. long.

Trailing Carrying Wheels

These were 3ft.-8in. diam. in all the Ivatt classes, i.e. 4-4-2, 4-4-2T, 0-8-2T and 0-6-2T (and also applied to the Gresley LNER 2-6-2T).  On the tender engines adequate side-play was provided by allowing a small amount of lateral movement in the axle boxes, this being sufficient with the steadying effect of a tender to the rear.  The bearing springs were 3ft.-6in. across their hangers, then 4ft.-6in. from 1902.  The position with the tank engines was complicated because there was no tender to steady the back of the engine. This batch was ordered 18/1/1897, suggesting preparatory work was done in 1896. The trailing carrying wheels were 3ft.-8in. diameter, and the the axle had limited side play, with outside laminated leaf springs.  The first engine (1009) appeared in February 1898 and was quickly found to be unsatisfactory.  Nine set of Bottomley's radial axle boxes were ordered for the later engines. So the answer appeared to be a radial “truck”, which followed the lateral movement of the axle boxes, but remained stable in the horizontal position.  Truck was a bit of misnomer, but the expression has stuck.  Almost immediately an amendment was made to the frame drawing O53 (15/2/1898) entitled “Alteration to trailing end of frames 5’-6” ten wheeled side tank engines to suit radial axle box arrangement”.  There is little detail on the drawing, which was presumably intended to be read in conjunction with the full O53 drawing.  The second engine (No. 1010) had already left works, so the the 3rd engine (No. 1013) became the "guinea pig", with the result that the 4th engine (1014) left works first unaltered (3/1898), and the modified 3rd engine (1013) followed (5/1898).  The remaining engines of the batch (1015-20), duly modified, started to appear five months later.  It is not known when the earlier engines (presumably 1009/10/13) were brought into line.  In the new arrangement, inside helical bearing springs were employed, though for the last ten engines, the inside springs were the laminated leaf type.

In this arrangement the main frames were parallel throughout, spaced 4ft.-1½in. apart as usual.  Alongside the trailing wheels there was a 1ft.-11in. radius cut-out, adequate to clear these wheels.  Extra spacing plates 7/8i.n thick, 6ft.-8in. long and cut to follow the contour of the main frames, were bolted to the inner faces of these frames from the drag box forward.  Finally, bolted to the inner faces of the spacing plates were two 11/8in. thick inside frames, with much smaller cut-outs, just sufficient to surround the axle boxes, total length again 6ft.-8in. to match the spacing plates.  The distance between the inner frames was 3ft.-9½in, now adequate to clear the trailing wheels.  The arrangement seemed rather untidy, but adequate to resolve a problem discovered after the initial frames had been cut.  Later engines had their frames “set” 2in on both sides of wheels, to give the same result, i.e. same distance between frames at this point, but standard width in front of and behind the wheels.

In the following description, the components are described using the official jargon on the drawings.  The axle boxes were connected to a 3ft.-3in. long Controlling Spring Spindle, over which were placed the two side control helical springs separated by two Controlling Spring Spindle Washers.  Above the short gap between the two Washers was suspended the Bracket for Controlling Spring.  So any sideways movement of the spring spindle resulted in compression of one or other spring, tending to restore the spindle to normal and hence axle-boxes to their central position.  This worked regardless of whether the spindle was high up, low down or skewed to one side, as there was complete freedom within the Bracket.  The surfaces of the axle boxes had rubbing plates such that, regardless of their position, they were in contact with the laminated bearing springs above them.  A further feature was that the axle boxes were also connected to two Radius Bars (also known as radial arms) which had their pivotal point 4ft.-3in. forward of the centre line through the Radial Axle.  The slack connection between Radius Bar and axle box (which were on different, constantly changing, planes) permitted a small amount of side-play.  The entire radial arrangement was arranged horizontally, on a slightly higher level than that of the side spring control to allow for the axle boxes bouncing up and down.  The pivotal radius of 4ft.-3in. was not arrived at scientifically.  The pivot had to be placed on a Radius Bar Stay running between the frames, and the best site was below the twin vacuum brake cylinders, with holes provided for the brake cylinder pistons to pass though.  Placing the radial pivot half way between these holes meant they could arrange the narrowest Stay possible (1ft.-3in.), saving on weight and materials.  Ivatt used a similar arrangement for his L1 (LNER R1) 0-8-2T and N1 (LNER N1) 0-6-2T, and then Gresley for his N2 (LNER N2) 0-6-2T and LNER V1/V3 2-6-2T, though in all these cases the radius swing was around 5ft.-9in.  Again the exact figure depended on where it was most convenient to fit the pivot.

Boilers

Under Stirling, boilers remained with the engine for life, so effectively when an engine was scrapped so was its boiler.  Steel boilers had been introduced in 1885, opening up the prospect of boilers outliving the engine.  The first example of re-use was in October 1896 when 2-2-2 No. 239 received the 4-year old boiler taken from No. 874 which in turn had just received a new boiler.  There were two further cases during December 1897, when 0-6-0 No. 723 received No. 743’s 15-year old boiler, and 2-4-0 No. 864 received No. 206’s 13-year old boiler.  During 1898 there were ten more examples among a variety of classes, after which it became common-place.  This begged the question of how to keep track of boilers, which up until then bore no identification numbers.  The first sign of a possible problem was in February 1901 when 2-2-2 No. 239 received its second consecutive second-hand boiler.  The boiler taken out was presumably checked for possible re-use (and in fact was used again, though a year later) and it was perhaps realised that checking the history of a removed boiler was no longer straightforward, and could only get worse.  It was likely that this resulted in the decision to introduce boiler numbers, which were allocated around July/August 1901.

Ivatt made certain changes to the boiler design.  For a start, the 1081 Class 0-6-0s appeared virtually as intended by Stirling, though the boilers conformed to N71 (18 February 1896) which had been prepared during the interregnum, with raised working pressure, greater overall heating surface and larger firebox grate area.  There were a further ten spare domeless boilers on order and for the time being their completion was delayed while preference was given to construction of domed boilers to N77 (July 1896).  The ten spares were eventually completed (rather slowly) in 1897-99 and the last one was fitted to a vintage Fowler 0-6-0 of 1867, No. 476 (December 1899).

Ivatt replaced Stirling-style three-ring barrels by two rings, with the rear one overlapping the front one.  Steam was collected inside a dome which also contained the regulator valve, with the dome located on the rear ring.  The width of the foundation ring, at the bottom of the firebox, between the outer and inner casings, was reduced in width from 3in. to 1¾in. to increase the grate area.  Not all the changes took place straightaway, but by the end of 1897 these were the norm for the time being.  The narrower foundation ring was not a success, as it quickly accumulated sludge.  The width reverted to 3in from 1902 on all classes.  This only slightly reduced the firebox grate area.

Perhaps Ivatt’s chief concern was the small grate area of the Stirling boilers.  The width was limited by the distance between frames, 4ft.-0½in. in the case of the Stirling 4-2-2s, but the length could be improved upon.  Ivatt’s changes were immediate.  The firebox was redesigned for the 8-footers (N72, completed February 1896, before he had taken over at Doncaster) with an inner firebox 6ft.-8¼in. long by 3ft.-5¾in. wide. This provided 23.26 sq. ft. of grate area, which was an improvement over Class 1003 (20 sq. ft.).  Ivatt had in mind an even longer firebox, and the boiler drawing for his first Atlantic (N80, 16th December 1896) showed the firebox casing 8ft.-0in. long, and the grate area 26.786 sq. ft.  To go any further needed a more radical approach, and his BG (broad grate) firebox had 31 sq. ft. of grate area.  The BG firebox required carrying wheels under the cab to support the weight, so that for engines without them, e.g. 0-8-0s, an 8ft. firebox was as far as Ivatt wished to go, without making the job of firing more difficult.  As will be seen later, Gresley had no such inhibitions.

For some time Doncaster continued to construct boilers for the Single-wheelers. . It would seem likely that ten boilers had been ordered in 1895 by Stirling, of which three were completed in 1896 to the original drawing, but the Works had managed to delay construction of the remaining seven until they had reordered different size boiler plates.  As a result, the first three boilers were domeless, but the last seven had two-ring barrels, with domes.  Details of these boiler orders (BO) are as follows, set out in order of boiler construction.  Fitting dates are noticeably out of order with construction order, inferring that the boilers were completed promptly and dumped in the boiler park, then afterwards randomly selected for fitment.

BO - (1895) (3-ring barrels, domeless).  Nos. 234 (January 1896), 773 (March 1896), 874 (September 1896).

BO - (1895, altered) (2-ring barrels, dome on rear ring.  Last four had plain casings around the safety valves, and Ivatt-style cabs).  Nos. 776 (July 1896), 544 (May 1897), 93 (June 1896), 95 (August 1897), 34 (December 1897), 22 (January 1898), 221 (June 1899).

BO204 (1897) (3-ring barrels, domeless).  Nos. 876 (March 1899), 240 (July 1898), 873 (April 1898), 871(August 1899), 233 (September 1899).

BO 507 (1897) (3-ring barrels, domeless).   Nos. 771 (July 1898), 69 (June 1898), 94 (May 1898).

BO 514 (12th July 1898) (3-ring barrels, domeless).  Nos. 548 (March 1899), 7 (June 1899), 670 (July 1899), 778 (August 1900), 664 (September 1900 at Ardsley shed), 2 (December 1900).

BO 522 (27th April 1899) (2-ring barrels, dome on rear ring).  Nos. 877 (November 1901), 872 (May 1901 at King’s Cross shed), 236 (March 1903), 880 (May 1903), 235 (August 1902).

BO 537 (14th March 1901, amended 20th October 1902) (3-ring barrel, with dome on centre ring).  Nos. 872 (May 1910), 1007 (May 1907), 116 (August 1903), 871 (June 1906), 877 (July 1906), 128 (December 1910).

Notes on above allocations:

The boiler fitted to No. 664 (September 1900) was probably intended for No. 549, which instead received a second-hand boiler.  The boiler fitted to No. 771 (July 1898) was afterwards fitted to 2-2-2 No. 234 (August 1904), thus requiring wider expansion brackets.  The boiler fitted to No. 778 (August 1900) was probably intended for No. 98, which was instead withdrawn; and this is the boiler presently fitted to No. 1.  The boiler fitted to No. 872 (May 1910) was later fitted to Ivatt 0-8-2T No. 117 (November 1911), then later (after shortening) to Gresley 0-6-0T No. 3175 (December 1924).

The final batch of domed boilers (BO 537) was ordered “for 8ft-0in passenger engines” then amended as “Made for 7ft-6in engines 119 N drawing”.  Drawing N119 was a new drawing for a three-ring barrel arrangement, with the dome on the centre ring, which was a surprising backward step, the reason for which is unexplained.  Furthermore it was indexed as being for the 4-2-2s, in other words with the narrower firebox, but at least it gave greater versatility.  Surprisingly, the first use for one of them was in 1903, when a replacement boiler was needed in a hurry for 0-8-2T No. 116.  The firebox needed a wider expansion bracket, and the Chief Draughtsman specifically noted that the firebox was (only) 3ft.-11½in. wide.  The boilers fitted to Nos. 116 (August 1903), 128 (December 1910) were afterwards shortened and fitted to Gresley 0-6-0Ts Nos. 174 (February 1915), 3178 (January 1924).

Stirling-type domeless boilers having small diameter barrels continued to be made, but with two rings, for the 0-4-2ST, 0-4-4BT and 0-6-0ST classes.  The last domeless boilers were ordered on 3rd August 1911 just before Ivatt retired.  The first significant exceptions to the norm for general use were the 4ft.-8in. diam. boilers for the 1321 Class 4-4-0s.  They had their largest ring at the front, 4ft.-91/8in. diam.  Ivatt’s larger boilers were all different.  The barrel of C1 (LNER C2) No. 990’s boiler (May 1898) was 14ft.-85/8in. long, with three rings.  This length was necessary on account of the long wheelbase, and the smokebox tubeplate was in fact recessed into the front ring, so that the distance between the tubeplates was 13ft.-0in.  Similar length boilers were used for the 0-8-0s and 0-8-2Ts, though the firebox proportions differed, see later.  C1 No. 271 (new July 1902) had a different boiler, 15ft.-4¼in. long overall with two rings, with its dome on the front ring.  The smokebox tubeplate was recessed into the front ring so that the distance between tubeplates was 14ft.-0in.  Its replacement boiler (February 1909) was a standard Klondyke boiler with a dummy extension piece at front end.  After serving 3 years on No. 254, followed by a further 4 years laid up, No. 271’s original boiler returned to this engine in December 1918.

The Long Tom K1 (LNER Q1) 0-8-0s had similar Klondyke boilers with an important difference.  The firebox needed to be 4in shallower at the back so that the ashpan could be inclined steeply to clear the rear coupled axle.  Long Tom boilers could be fitted to the Klondykes, though not vice versa, so that after the last batch of Klondykes appeared, only Long Tom boilers were constructed.  The boiler was also fitted to the first L1 (LNER R1) 0-8-2T No. 116 (June 1903).  After being declared unsuitable for work on the Metropolitan Widened Lines, it was fitted with a small diameter boiler (August 1903), one which had been ordered two years before for the Stirling 4-2-2s.  This solved the problem and a further forty 0-8-2T engines were built with similar small boilers.  Later, No. 133 received a Klondyke boiler (September 1909), and the remainder were eventually brought into line under Gresley.

Ivatt broke new grounds with his large Atlantic No. 251 (C1).  The boiler barrel was 5ft.-6in. diam. and its length was just over 15ft.-6in.  The firebox had a broad grate, and the grate area was 31 sq. ft.  Commencing with No. 1421 (August 1907) the depth of the firebox at the front end was increased by 3in so that the grate sloped down more than before, tending to make the firing easier by shaking the coal forward to the front.  From then on there were two types of wide firebox, shallow and deep.  Deep firebox boilers could be fitted to Nos. 1421 onwards only, though later several earlier Atlantics had their frames cut to suit.  Shallow firebox boilers from the early Atlantics could be fitted to later engines by fitting a special packing bar to bridge the gap.  Both shallow and deep firebox Atlantics continued to be constructed, even within the same boiler order.  The width over the firebox casing was 6ft.-9in. so that elaborate arrangements were necessary to support it on the main frames whilst allowing for heat expansion.  Traditional expansion brackets on the firebox sides were not an option.  As the frame arrangement was the same as on the Klondykes (apart from the shorter outside frames on the first eleven engines), these frames could take Klondyke boilers with expansion brackets, and likewise the early Klondykes could have been given the broad grate boilers.  It was intended to fit Klondyke boilers to new 0-8-2Ts, but only No. 116 was fitted (July 1903) before the proposal was banned on grounds of excessive weight for operating “underground”.  The resulting snap decision to fit this engine (July 1909) with a spare boiler originally intended for a Stirling Single-wheeler had surprising consequences.  A further forty of these boilers were built for later engines.  No spares were built, and any needed were taken from those intended for Single-wheelers.  This left a large quantity of surplus Klondyke boilers, of which there were still seven lying around in the Works in mid-1909, when a surprising turn-around resulted in one of them being fitted to 0-8-2T No. 133 (September 1909), following the departure of the class from the London scene.  It was left to Gresley to capitalise on this event.

Foundation Rings

The foundation ring separated the inner copper firebox from the outer iron (later steel) firebox.  Ring is a misnomer as the shape was approximately oblong with four straight sides with curved corners.  It was traditionally 3in wide, because if made any narrower there was the chance that solid matter in the water would collect on its surface and not be swirled away.  However from 1896 the width was made 1¾in. for new boilers and it can only be assumed this was a directive from Ivatt before he arrived at Doncaster.  This slightly increased the firebox grate area and heating surface.  There appears to have been a problem with the first Klondyke No. 990 (May 1898), for there was a return to 3in. wide foundation rings in all later boilers for Klondykes and Long Tom 0-8-0s.  Meanwhile fireboxes in other boiler types continued to have narrow foundation rings down to BO 534 (August 1900) for 0-6-0ST Nos. 1251-70, whilst BO 535 (February 1901) and onwards had 3in. wide foundation rings.  The general instruction seems to have been unclear and the question appears to have been raised when the boilers for 4-4-0s 1386-95 were ordered (October 1901). The LDO was instructed (21 December 1901) that they had to have 3in. foundation rings.  Klondyke No. 990 was eventually brought into line (November 1903) with its first boiler change.

Safety Valves

Ivatt continued to fit Ramsbottom “duplex” safety valves and for a short time his earlier boilers also had brass trumpets.  This applied for example to his 2-4-0s Nos. 1061-70, 4-4-0s Nos. 400, 1071-79 and 0-6-0s Nos. 1081-1100.  Commencing with 4-4-0 No. 1080 he substituted cast-iron casings.  However, the numerous domeless boilers he continued to provide for the Stirling 0-6-0 saddle tanks continued to have brass trumpets.  No doubt these were all recycled trumpets.

No. 271 to check

The small Atlantics Nos. 250/52-60 (1902) had a pair of Ramsbottom “duplex” with two cow’s tails.  This applied to the similar boilers fitted to new 0-8-0s and 0-8-2Ts from 1903, but it has not been established if the earlier boilers were brought into line.  Similar safety valves were fitted to the large Atlantic boilers [No. 251 when new to confirm].

Ivatt’s last design was his Class D1 (51 Series) 4-4-0 and one of these, No. 61 (new 5/1911), was fitted with Ross “pop” safety valves.

Boiler Classifications

The LDO attempted to classify the boilers in 1902.  The original list was complex, then rearranged, but still remained a lengthy list with about the same number of boiler classes as engines classes.  The difference between any two classes could have been height of firebox, slope of firebox grate or height of expansion bracket on the firebox side.  It was just as easy classifying boilers the same as the engines they were used in.  No evidence has been found of this list being used in practice.  As an example of its complexity, the following boiler classes covered what was basically Stirling’s standard boiler, with 4ft.-2½in. diam. barrel, 5ft.-6in. firebox casing and differing slightly in barrel length.

Table 10 - Specimen boiler classes, Diagram D

  

Original
Boiler
Class

Later
Boiler
Class


Engine
Class


Barrel
length

C6

D3?

B2, E2, E3

10ft.-2in.

D2

D3

B5

10ft.-2in.

D4

D6

J6

10ft.-1in.

D5

D7

J7

10ft.-0in.

D6

D8

G3

10ft.-0in.

D7

D5

G1

10ft.-1in.

D10

D11

J10

10ft.-1in.

P1

K2

J1, J2, J3

10ft.-0in.

Notes:
Class D1 (not changed) was allocated to engine Class B7 No. 21 etc with Ivatt “
standard” 4ft.-5in. diam. boiler with dome.  Class E1 No. 89 etc added.  Other classes with these boilers presumably added later
Class D9 (later D10) was allocated to engine Class J9, Nos. 1021-30, which had the Ivatt “standard” 4ft.-5in. diam. boiler though without domes.

Superheating

The water in the boiler is heated by hot gases passing from firebox to smokebox, through an arrangement of smoke-tubes, the diameter of which varied as a result of experimenting, and there was no firm ruling on the subject.  The tubes lower down heated the water whilst those higher up heated the steam.  Based on this reasoning the next step was to introduce superheating, whereby steam from the regulator could be routed through the upper tubes on its way to the cylinders, and thereby heated further.  Ivatt introduced this method in April 1909 on Klondyke No. 988, using the second-hand boiler from No. 253.

Instead of the 191 x 2in. diam. smoke-tubes there were now only 97 such tubes.  Higher up were 18 x 5¼in. diam. flue tubes arranged in two rows of nine each above the water level.  The steam went from the regulator valve in the dome to the “saturated header” in the smokebox.  Here it branched out into 18 x 1¼in. diam. small tubes (called “elements”).  Steam in the elements passed through the flue tubes, where it was further heated.  Two complete passes were made so there were in effect four small tubes within each flue tube.  The steam was collected in the “superheater header” and sent to the steam chest.  There were variations in element length, e.g. long loop or short loop, worked out by trial and error. The main concern with this system was the method of securing the elements in the headers which were close together and subject to wide temperature variations. Firstly between the adjacent headers, and secondly as a result of running with regulator valve open or closed.  Ivatt favoured the Schmidt system of expanding the element ends into flange blocks bolted to the respective headers, i.e. one in the saturated header and the other in the superheater header.  Initially, heat damage to the elements was avoided by fitting smokebox dampers which shut automatically when the regulator valve was closed, sealing off the flue tubes from the hot gases.  (The dampers could also be operated manually from the footplate.)  Anti-vacuum valves allowed air to be drawn into the cylinders when the regulator was closed, and later it was realised these obviated the need for dampers.  The total evaporative heating surface of small tubes and flue tubes, plus superheating surface of the elements was less than the original evaporative surface of small tubes in the saturated steam boilers, suggesting heat waste, but the benefit was higher temperature steam in the steam chest.  Comparative heating surface figures below are taken from the LNER engine diagram for Class K1 (LNER Q1).

  Table 11 - Class K1 boiler heating surface comparison

Saturated boiler
(sq. ft.)

Superheated boiler
(sq. ft.)

Firebox

137.00

135.50

Small tubes

1302.00

660.25

Flue tubes

-

321.00

Superheater
elements

-

254.00

Total

1439.00

1370.75


 
Ivatt fitted the Schmidt superheater to 50 engines as follows:
Existing Class C1 (LNER C2)  No. 988 (1909)
Class K1 (LNER Q1)  Nos. 451-55 (new 1909)
Class C1 (LNER C1)  Nos. 1452-61 (new 1910)
Class D1(LNER D1)  Nos. 51-65 (new 1911)
Class J22 (LNER J6)  Nos. 521-35 (new 1911)
Existing Class K1 (LNER Q1)  Nos. 405/16/20 (1911)
Existing Class C1 (LNER C2)  No. 271 (1911)

Ivatt retired shortly afterwards and subsequent development was left to Gresley.

Cylinders and Valves

Under Ivatt there was a gradual, though not universal, application of balanced slide-valves.  The exhaust steam passed directly through the backs of the valves to the blast pipe with minimum resistance.  To assist the blast in drawing hot gases through the smoke-tubes, a cowl was then fitted inside the chimney.  Then for his 4-cylinder Klondyke, No. 271, he used piston valves in a novel arrangement.  On one side of the engine, the two adjacent cylinders (straddling the frame) worked at 180 degrees to each other.  They shared a common piston valve, with four heads.  One pair controlled (by inside admission) the port openings for the outside cylinder, whilst the other pair controlled (by outside admission) the port openings for the inside cylinder.  The same arrangement applied on the other side of the engine except that their cylinders were out of phase by 90 degrees.  Very much later piston valves gradually replaced balanced slide valves following the widespread use of superheating, though this was mainly under Gresley.  On inside cylinder engines there was no room between the cylinders for two steam chests with balanced slide (or piston) valves and they were then located above their cylinders.  But on outside cylinder engines, steam chests continued to be provided between the frames, with the inevitable lengthy live and exhaust steam passages connecting with the cylinders.  But under Gresley, who favoured outside cylinders and Walschaerts valve gear, steam chests with piston valves could be located above the cylinders, in view.  Another point to note about engines with piston valves was that the steam chests had to be longer, and noticeably so on outside cylinder engines. 

Four-cylinder engines under Ivatt presented problems.  Steam chests having balanced slide (or piston) valves could not be located between the frames, and in all such cases these had to be located above the cylinders.  Unusually four-cylinder Atlantic No. 292 had ordinary slide-valves for the inside cylinders, and these could be arranged vertically as in the traditional Stirling manner.  Clearly fitting balanced slide instead of ordinary slide-valves was not the universal solution.  For example, it is recorded that J50 No. 3167, the only superheated member of its class, lost its balanced slide-valves in 1935 after complaints of sluggish running downhill (see “Locomotives of the LNER”, Part 8A, p. 10).  The 4-cylinder Atlantics were arranged as follows:

Table 12 – Ivatt and Gresley 4-cylinder Atlantics


No.


Date

Outside
cyls  (in.)


Valves
(c)

Steam
chest
position


Motion
(b)

Inside
cyls
(in.)


Valves
(c)

Steam
chest
position


Motion
(b)

271

7/1902

15 x 20

PV

Above

(a)

15 x 20

PV

above

(a)

271

9/1904

15 x 20

BSV

Above

WVG

15 x 20

BSV

above

SLM

292

3/1905

13 x 20

BSV

Above

WVG

16 x 26

SV

alongside

SLM

1300

7/1905

14 x 26

PV

Above

WVG

23 x 26

BSV

above

WVG

1421

8/1907

13 x 20

BSV

Above

WVG

18 x 26

BSV

above (e)

WVG

279

5/1915

15 x 26

PV

Above

WVG

15 x 26

PV

below

(d)

(a)     Combined valves operated off inside valve motion

(b)    SLM = Stephenson Link Motion;
       WVG = Walschaerts
Valve Gear

(c)     SV = slide valves arranged vertically between cylinders;
       BSV = balanced slide valves above the cylinders;
       PV = piston valves above or below the cylinders

(d)    Inside PV operated off outside valve motion

(e)     BSV for inside cylinders sloped at 5 degrees from the horizontal when viewed head-on.

Cabs

Commencing with Class D2 (LNER D4) No. 400 (December 1896) cab roofs were extended back, further still than on the Stirling 1003 Class 4-2-2s, and the roof had a flatter appearance.  Similar cabs were generally fitted to those Stirling engines that were rebuilt with larger boilers, but not in all cases.  Sometimes the original cabs were simply altered to fit, as a cheaper alternative to fitting a new Ivatt-style cab with extended-back roof.  A short horizontal handrail was usually fitted if not already provided, but there was no consistency on this point.  Single-wheeler No. 1007’s cab retained its original short roof extension.

Bogies

Ivatt’s bogie had 1in thick frames, but otherwise differed from that of Stirling in a few significant details.  The wheelbase was 3in. shorter at 6ft.-3in., pivoted at 1½in. to the rear of the bogie centre line, and the wheels were smaller, 3ft.-8in. diam.  Instead of having laminated springs for suspension they had Timmis helical springs, one located at each side of an axle box, thus there were eight springs per bogie.  Ivatt also introduced bogie side control by means of swing links.  On his 4-4-0s the engine main frames narrowed inwards alongside the leading bogie wheels to prevent them coming into contact on curves, a feature introduced by Stirling on his 4-2-2 No. 1 (rebuilt 12/1880).  C1 No. 983 had a special bogie with outside frames (1903-1909) afterwards carried by No. 983 (3983) (1909-1936).  C1 No. 271 (3271) (rebuilt July 1911) and also D1 4-4-0s (new 1911) had longer bogies, with 6ft.-5½in. wheelbase, to clear the larger cylinders.  The Atlantics, both Klondykes (LNER C2) and large Atlantics (LNER C1), had their bogies positioned ¼in. further forward for clearance purposes when piston valve cylinders were fitted.  Vulcan 4-4-2 No. 1300’s bogie was non-standard in having 3ft.-2in. diam. wheels.  The Atlantic bogies gave trouble, and strengthening patches had to be fitted.

Reversing Gear

The general rule was lever reverse fitted to engines with Stephenson valve gear, practically universal with Ivatt, hence for example the large Atlantics always had this arrangement, though strangely the Klondykes (apart from No. 990 at first) had screw-reverse.  Screw reverse though was essential with the heavier Walschaerts valve gear.  It is possible that over the years some engines were altered from lever to screw and vice versa.  Ivatt also fitted screw-reverse to Stirling 4-2-2s Nos. 544, 95, 34, 22 and 221 (1897-99).  Unlike Nos. 1003-8 on which the driver turned a wheel, these five engines had turn-handles (requiring a new drawing, dated May 1897).  Thereafter turn-handle reverse became standard.

Tender Types

Ivatt introduced a tender classification:

Table 13 - Ivatt Tender Classification

Class

Wheel
 diam

Flat Bottom
 or Well

Frames

Constructed

A

4ft.-0in.

FB

iron

1896-

B

4ft.-0in.

W

iron

1896-

C

3ft.-6in.

FB

iron

1883-96

D

4ft.-0in.

W

iron

1883-96

E

3ft.-6in.

FB

wood

1866-83

F

4ft.-0in.

W

wood

1866-83

G

various

both

sandwich

pre-1866

Tender Drawings

In appears likely that in the early part of Ivatt’s reign detail drawings for tender drawings continued to be prepared in the Works Manager’s Drawing Office (W.M.D.O.).  As before, the LDO dealt with details that overlapped with engine design, e.g. wheels, brakes, couplings, and (later) water pick up gear.  The first two Ivatt drawings are of particular interest, as follows:

R22 (April 1896) titled “Water Scoop & Details Tender (Flat bottom) 3280 3170 Galls.” in the drawing register.  This was possibly a general Arrangement drawing prepared by the W.M.D.O. on which the LDO had superimposed details of a proposed water pick-up gear, though at that time the GNR had no water troughs.  It had probably been requested by Ivatt (see later under “Water Pick-up Gear”).  The words “Water Scoop & Details” were later struck out in the drawing register, and the drawing as issued showed no water pick-up gear.  It did however show the capacity as 3140 gallons which was correct only for the tenders which had this gear. 

R23 (July 1896) was another general Arrangement drawing, initially for “Tender to carry 3400 Gallons capacity”, but later amended to “Tender to carry 3900 Gallons capacity (Well Tender)”, in other words an updated version of the Stirling well tenders attached to his 1003 Class 4-2-2s.  This drawing no longer exists and it was reissued (R23, June 1898) as “Tender 3670 gallons”.  It now incorporated the new central coupling, see below under “Engine-Tender Intermediate Coupling”, though the first tender to this drawing, had already appeared a month earlier with Klondyke 4-4-2 No. 990.  Another interesting feature, peculiar to this tender type, was its vacuum brake. There was now a deep 21in. diam. brake cylinder located above the rear axle, actuating the brakes blocks in the front of the wheels instead of the rear.  The next development was the relocation of the 21in. diam. cylinder above the front axle, placed at a slant so that the piston would be well clear of the axle.

From about 1898 it appears that the LDO usually took responsibility for producing the more important detail drawings, such as frames and tanks, though details of minor importance, e.g. handrails, could still have been dealt with by the W.M.D.O.

Engine – Tender Intermediate Coupling

The traditional Stirling coupling between engine and tender comprised a screw coupling which passed through a 10in. wide slot (to permit a little side-way movement) in the tender buffer-beam, connected to a 3ft.-long horizontal laminated spring which controlled fore and aft movement.  The screw coupling was held in place at the engine end under the “fall-plate” by a vertical bolt.  Ivatt’s first tender general Arrangement drawing (R22, April 1896) showed the usual screw coupling.  The next arrangement was a solid drawbar (officially known as a central coupling) having larger lateral movement and this was depicted on the next general Arrangement drawing (R23, redrawn June 1898).  In both arrangements the coupling was initially backed up by a pair of safety chains, and there were also two small buffers.  The LDO recorded their early usage on Ivatt engines as follows.  Screw couplings for the small-boilered 4-4-0s Nos. 400, 1071-80, 1311-20/41-50 (and possibly though not recorded for the 315 Series 0-6-0s).  Central couplings for 4-4-2 No. 990, large-boilered 4-4-0s Nos. 1321-36-40/61-65 and 4-2-2 No. 266.  The LDO also noted that the fitting of the central coupling reduced the overall length of engine and tender by 1½in.  It is possible that the central coupling became standard thereafter.  Photographic evidence does suggest that Ivatt 4-2-2 No. 270 could latterly have had a screw coupling.

Tender Frames

Ivatt eliminated the use of wood in tender construction, which were latterly under Stirling just the buffer beams.  In the basic tender (R22, April 1896) the frame plates were 20ft.-8½in. long with steel bufferplates and a distance of 22ft.-115/8in. between buffers.  The wheels were 4ft.-2in. diam. with 3in. tyres and the wheelbase 6ft. + 6ft.  The bearing springs above the axle boxes were the laminated type, 3ft.-6in. between hangers. The top leaf was 5/8in. thick and a further 10 leaves were ½in. thick. The vacuum brake cylinder was in front of the leading axle and tilted at a slight angle for clearance purposes.  Drawing R23 (July 1896) was the well tank version.  The R23 tender first appeared with Class C1 (LNER C2) No. 990 (new May 1898).  The drawing was specially revised after the event (June 1898) to incorporate design changes that had been made during construction (the Works having exercised their initiative), in particular the vacuum brake cylinder had been located above the rear axle as the well tank was in the way at the front.  In later designs the LDO took care to ensure the vacuum brake cylinder was in its rightful place, at the front.

The subsequent introduction of water scoop gear needed an important design change, which was the wheel-spacing. The general arrangement drawing (December 1902) was probably drawn out in the Works, and a copy handed over to the LDO in 1904 where it was given the next available number R43 (but its date of issue was left blank in the register).  R41 was for the frame plate (October 1904), but it is not clear if this had been drawn out in the Works.  The wheel-spacing was 6ft.-10½in. + 6ft.-1½in. and the lightening holes were smaller.  The tenders were ordered 11th October 1904 as the main detail drawings (frames and tank) were now complete.  Despite the odd wheel-spacing further order were placed in January and March 1906.  These were the last flat-bottom tenders built at Doncaster.

The next design was the 3500 gallons tender with a well and water scoop between the frames.  From hereon all drawings appear to have been made in the LDO  The first tender order (TO21, 6th September 1906) was closely followed by the first drawings e.g. frames (R50, October 1906) and general arrangement (R54, November 1906).  The introduction of the well resulted in the wheel-spacing being 7ft. + 6ft.  The vacuum brake cylinder was now back in its old position above the front wheels and tilted to one side to give adequate clearance away from the axle.

Tender Tanks

Ivatt’s first tender design (R22, April 1896) was the so called horse-shoe tank, with no well tank.   This was 4ft.-8in. deep at the rear, the “horizontal” portion gently sloping down to 4ft.-7in. at the front.  The sides extended forward right to the front of the tender (apart from a portion 2ft. long and 11in. deep at the front, providing space for a tool box).  The overall length of the tank was 18ft.-7¼in. The side extensions were 2ft.-1in. wide and the depth at the front was 3ft.-8in.  Between the side extensions the tank sloped down shallowly at about 18 deg. from the horizontal towards the front over a distance of 6ft., then dropped vertically.  This provided a coal space 2ft.-10½in. between the tank side extensions and 4ft.long on the level to the coal gate.   The shovelling plate was back again at its low level but set at angle as on the final Stirling design.  The tender design was revised (R23, June 1898) to incorporate a well tank located forward of the vacuum brake cylinder which was at the rear. The well was 4ft. wide and 1ft.-7¼in. deep, but the length measurement is unreadable on the drawing.

The next major design change was the self-trimming tender, introduced with tenders T.601-20 (TO17, 12 December 1901).  There was delay in their construction and Malcolm Crawley shows them completed March to August 1903 (his source is unknown).   As far as the drawings were concerned, there are big problems.  The first drawing was general arrangement R37 (October, 1902) probably prepared by the LDO and endorsed in the register at a later date as “T.O. 17 only”.  Work was then handed over to the Works for them to prepare the necessary detail drawings.   Three Works drawings were handed over to the LDO probably towards the end of 1904 where they were then given R numbers.  The clerk recorded them as 1902 stressing the year.  These were R38 scoop gear for the Ivatt R37 3000 gallons tenders, R39 scoop gear for old Stirling flat-bottom tenders (signed by F. Wintour, the Works Manager) and R40 scoop gear for old Stirling well tenders (specifically the R20 3850 gallon tenders).  The original date of R38 is not known because it was replaced in 1923 by a revised version, whereas R39 and R40 were both dated October 1902.  Whether or not any Stirling tenders were actually fitted with scoop gear is open to conjecture, but as these two latter drawings were never revised (unlike R38) this suggests they may never have been used. 

The LDO then prepared two drawings for the next 3000 gallon tender order TO18 (issued 19 October 1904). These were R41 (frame plate) and R42 (tank) both dated October 1904.   At this point another Works drawing turned up and was given the number R43.  This was the general arrangement drawing used for TO18 onwards, but had been drawn out in December 1902.  The drawing date was left blank in the LDO register.   R44 (November 1904) was a LDO drawing for older Stirling tenders. Then another Works drawing turned up (dated December 1902) for the scoop gear for 3000 gallon tenders, and this was registered as R45 and recorded in the register as being a 1902 drawing.  Two further drawings turned up and were registered by the LDO in 1904, R46 (water gauges “see Z-31”) and R47 (water gauges for tenders for “17 & 18 T.O.”).  The latter drawing was dated March 1904.  No R drawings were registered in 1905.  Another drawing turned up in November 1906, and was allocated R55.  This was for the brake gear on the Baldwin tenders (dated February 1900).   No drawings were registered from November 1906 until after Ivatt’s retirement in 1911.  All this suggests that there is insufficient conclusive evidence regarding the 3000 gallons tenders, but the following seems probable.

The R22 tender design was altered (R37, October 1902) and therefore applicable to TO17 (issued 12 December 1901). This was revised by (R43, December 1904) which altered the wheel spacing and was therefore applicable to TO18 (issued 19 October 1904) and the two 1906 orders (TO 19 and TO20).  It is assumed that the basic tank details were the same for these four orders.  This was 4ft.-8in. deep at the rear. The sides extended forward right to the front (apart from a portion on the left side only, about 11in. long by 11½in. deep, providing space for an off-centre toolbox).  The overall length inside the tank was 18ft.-8in.  The side extensions were about 2ft.-1in. wide.  Between the side extensions the tank sloped down shallowly towards the front where there was a horizontal section about 1ft.-6in. deep and the full width of the tank.  This provided space for a further 10 cwt coal, at the expense of around 150 gallons of water, which was less critical with the advent of water scoops.  Between the side extensions a 3ft.-7in. long shovelling plate was fitted, on a slight rise, so this was once again on a level with the fire-hole door.  Several dimension were absent from the drawing (those marked “about” in this paragraph) presumably because they were unaltered from R22.  This was followed by the well version (R57), which was virtually identical apart from the addition of the well tank which was located behind the vacuum brake cylinder. The well tank measured 13ft.-1½in. long, 4ft. wide and 1ft.-7¼in. deep, giving a calculated cubic capacity of 85 cu. ft., representing 529 gallons of water. 

Tender Orders

The Tender Order (TO) system was introduced about 1897.  The LDO commenced keeping tender order records around 1900 and recorded earlier orders that had been completed after Stirling’s death.  The LDO order book disappeared around the 1970s, but two rough copies (which do not exactly agree with each other) had been made for various purposes and these have survived.  For example, the copies noted the current allocations of the tenders, ignoring the various crossings out caused by tender exchanges.  In the early days tenders carried the same number as the engine to which currently attached, so on occasions they were renumbered. The following is believed to be an accurate interpretation of the events.  In the version I have used below, Completion Dates were recorded down to TO4 and Ordering Dates were recorded from TO5.

TO1.  Reconditioned Class A, 3ft.-6in. wheels, 2654 gallons.  Classification corrected to Class C in March 1922.  These tenders were probably taken from: Nos. 540 (6/1894), 435 (12/1894), 329 (2/1895), 290 (7/1895), 341A (c.7/1895), 468 (8/1895), and after reconditioning they were then probably attached to: Nos. 540 (9/1894), 749 (8/1895), 290 (11/1895), 299 (2/1896), 21 (10/1897), 89 (3/1898).  Order completed 3/1898.

Part TO-.  Reconditioned Class D, 4ft wheels, 3450 gallons, intended for No. 400 (see later), but attached instead to No. 1061.  Engine completed 2/1897.

TO-. Class A, 3ft.-6in. wheels, 2690 gallons, attached to Nos. 1081-90.   Classification corrected to Class C in March 1922.  Order completed 12/1896.

TO-. Class A, 4ft. wheels, 3280 gallons (first nine) and 3170 gallons (last one).  Intended for Nos. 1061-70, but actually fitted to Nos. 400, 1062-70.  Order completed 5/1897.

TO2.  Class A, 3ft.-6in. wheels, 2864 gallons.  Order most likely for ten tenders using up surplus material in stock for Engine Order (EO) 208, but later reduced to six when it was realised there was insufficient material. TO7 (modern tenders) substituted for the last four engines of this EO.  Attached to Nos. 315/16/18/29/31/32.  Classification corrected to Class C in March 1922.  Order completed 12/1897.

From here onwards all tenders had 4ft. diam. wheels.

TO3.  Class A, 3170 gallons, for Nos. 1071-80.  Order completed 6/1897.

TO4.  Class A, 3170 gallons, for Nos. 1301-20.  Order completed 6/1898.

TO5 (1/7/1897).  Class B, 3675 gallons, intended for Nos. 1321-30.  Attached instead to Nos. 990, 1321-27, 266, 1328.  While No. 1323 was in Works for a general repair (17/3/1900-25/5/1900) its tender was taken for new 4-2-2 No. 267 (ex-works June 1900) and accordingly renumbered 267.  No. 1323 emerged from Works with a second-hand tender, details not known.  Either the replacement tender was never renumbered 1323 (as it should have been), or if it was then it was scrapped early, because “1323” has never subsequently shown up in the records.

From hereon, the relevant Tender Order book has survived at the NRM.  The earlier records were entered retrospectively in another book which has not survived.

TO6 (6/1/1898).  Class B, 3675 gallons, intended for Nos. 1331-35. Order then increased to seven tenders for Nos. 1329-35 to balance the tenders taken from the previous order for No. 990 and 266.

TO7 (6/1/1898).  Class A, 3170 gallons, for Nos. 334/36-38, which were originally to have had hybrid tenders from TO2, until that order was reduced.

TO8 (21/2/1898).  Class B, 3670 gallons, for Nos. 1341-50.

TO9 (8/11/1898). Class A, 3170 gallons, for Nos. 343-45/48-53/59-64/67/68/71/75-81.

TO10 (21/3/1898). Class B, 3720 gallons, for Nos. 1336-40/61-65.

TO11 (6/1/1899).  Class A, 3170 gallons, for Nos. 1351-60.  No. 1356 shortly afterwards received water scoop gear (February 1900), reducing the water capacity to 3140 gallons.

TO12 (14/4/1899).  Class B, 3670 gallons, for Nos. 949/50/82-89.

TO13 (26/10/1899).  Class B, 3670 gallons, for Nos. 1366-85.  Due to a shortage of material caused by the ongoing industrial dispute in the steel industry, several engines acquired second-hand tenders instead.  Nos. 1377/83/85 still had Class A tenders at the 1901 tender census.  There could have been others which had also received second-hand tenders when new but which had been rectified before the census was taken.  It is even possible that this TO was finally cut to seventeen tenders, as there seems to be no trace of three Class B tenders bearing other numbers that were never allotted to such tenders.

TO14 (26/10/1899). Class B, 3670 gallons, for Nos. 1386-95.  The order was cancelled and the engines instead acquired second-hand tenders, due to a shortage of material.  Ten Stirling Class A2 4-2-2s in sound condition were withdrawn late-1902 to mid-1903 to balance the ten “renewal” Class D1 4-4-0s which had been under (slow) construction throughout 1902.  The displaced tenders would have been in good condition and suitable for transfer to the 4-4-0s.  It is also believed these reconditioned tenders were then numbered T.1386-95.  TO14 was later reissued.

TO15 (26/10/1899).  Class A, 3140 gallons, for Nos. 165/77/79/80/92, 302-04/06/08, 384/86-88/90/92/94/96/98/99.

Allocating numbers to tenders that agreed with the engines to which they were attached caused problems when they were switched around.  It was decided to seek out blocks of engine numbers that were currently allocated to tank engines and use these numbers in future for tenders.  Suitable blocks at that time were 241-50, 270-79, 494-503, 601-10, 611-20, 621-30, and these were allocated to the next four tender orders, though oddly 270-9 came before 241-50.  The system fell down in 1902 when the number 271 was given to a new engine just over a year after T.271 was put into traffic, and a year later a new engine was given the number 250.  The problem for the historian is that no permanent records were kept of which engines received these tenders, and the tender record cards covering this era have long since gone.  The next three orders were the last of tenders with evenly spaced wheels, 6ft.-6in. apart, as follows.

TO14 (reissued 6/4/1900).  Class B, 3670 gallons, T.270-79.

TO16 (28/4/1900).  Class A, 3140 gallons, T.241-50, 494-503.

TO17 (12/12/1901).  Class A, 3000 gallons, first self-trimming tenders, coal capacity increased from 5 to 6½ tons, T.601-20.

There was then a gap of almost three years before the next tender order was placed in 1904.  Tender construction lagged well behind Order dates and in the intervening years TO16 was executed in 1902 and TO17 in 1903.  However no new tenders appeared between September 1903 and July 1905, during which time approximately the same number of tender engines were withdrawn as new tender engines were built, which must have kept the Tender Shop fully occupied refurbishing displaced tenders.

TO18 (19/10/1904), TO19 (23/1/1906) and TO20 (27/3/1906) Class A, 3000 gallon, similar to TO17 above the frames, but middle wheels moved forward 4½in., because of the fitting of the water scoop gear, so that the wheel spacing became 6ft.-10½in. + 6ft.-1½in. The first ten tenders were numbered T.621-30 using numbers allocated to tank engines.  By then it was realised a more sensible idea was to number new tenders well away from engines numbers, so the next two batches became 5001-10, 5011-20.  These were the last flat-bottom tenders built at Doncaster.

The next development came in October 1906 when drawings appeared for a 3500 gallons tender: R49 (tank), R50 (frame) and R51 (scoop gear).  The general arrangement drawing (R54) followed a month later.  R54 was later revised (R57, 24 November 1911).  The design was based on the 3000 gallons R43 tender, but with a well fitted between the frames and altered wheel spacing 7ft. + 6ft.  This was a most successful design and altogether 220 of these tenders were built between 1907 and 1925, T.5021-5210, 5213-22, 5233-52.

TO21-28 Class B, 3500 gallons water, 6½ tons coal. There were certain detail differences within these orders, as described in Malcolm Crawley’s booklet.  Their construction is summarised as follows:

Table 14- Ivatt Class B Tenders


TO

Order
date

Tender
Nos.

 


TO

Order
date

Tender
Nos.

21

6/9/1906

5021-30

 

25

25/1/1910

5061-70

22

14/5/1907

5031-40

 

26

13/9/1910

5071-80

23

20/1/1908

5041-50

 

27

31/1/1911

5081-95

24

27/1/1909

5051-60

 

28

2/8/1911

5096-5110

                                                                                   Further Class B tenders were constructed under Gresley, before and after Grouping.

Tender Brakes

There were two alternative methods for fitting full length 21in brake cylinders. It is possible some of the Stirling tenders were similarly fitted during refurbishments though such information is no longer available.  They were either fitted towards the back of the tender (the position of the brake blocks in front of the wheels is a “give away” on photographs) or placed at an angle towards the front thus both clearing the axle and increasing the distance between the sloping piston and axle.

Water Pick-up Gear

When Ivatt took over at Doncaster (March 1896) he was aware that the GNR had no water-troughs, which limited the range of their express passenger engines, and he prepared a tender drawing (R22, April 1896) for “Water Scoop & Details Tender (Flat bottom) 3170 Galls.” (Also see earlier under “Tender Drawings”).  At that time the GNR had no suitable engines for the long distance run London to Leeds, too much for the Stirling Single-wheelers.  Ivatt had a 4-4-0 in the pipeline as a first step and very likely a 4-4-2 more suitable for the task he envisaged.  The 4-4-0 No. 400 appeared later that year (December 1896) and his prototype 4-4-2 No. 990 (authorised February 1897) appeared in May 1898.  Another tender drawing was prepared (R25, December 1898) for “Arrangement of Water Scoop for 3170 gallons Tender”, which meant the time was now ripe to discuss the matter with the Locomotive Committee, as the provision of water troughs needed Board approval; and the scheme was agreed. 

The tender (from TO11) attached to 4-4-0 No. 1356 (built December 1899) was shortly afterwards with water scoop gear (February 1900).  Twenty 3140-gallon tenders (with water scoop gear) were ordered (TO15, 26 October 1899) and fitted to new 0-6-0s (April to August 1900).  Meanwhile the first batch of the production Klondykes (949 Series) appeared in the first half of 1900.

These events were followed by the opening of the three troughs essential for long distance working at Werrington (November 1900), Muskham (1901) and Scrooby (1902).  (The fourth water troughs, at Langley, did not open until October 1918.)  A certain amount of tender switching may have taken place before through working became possible and then the final batch of Klondykes Nos. 250/52-60 (EO230, issued 17 December 1901) appeared in May-June 1903 with tenders having water pick up gear (TO16, issued 8 April 1900).  No new tenders were ordered in 1902-03.  From 1904 (TO18 onwards) new tenders had an unevenly spaced wheelbase to provide extra clearance for the scoop gear, when fitted.  Also from hereon tender orders did not show whether or not scoop gear was to be fitted, and this decision may have been taken shortly before each tender was completed.  It is doubtful if the pick up gear was fitted to tenders which were unlikely to require it, as it was known that if the gear was not used frequently it would seize up.  Sometimes air vent pipes were provided towards the front of the tank though not in all cases and the reason for this distinction is not known.  In a few cases baffle plates were fitted in front of those vent pipes that were well forward, i.e. on the “self-trimming” tenders, to avoid firemen getting drenched when picking up water.

Postscript

Ivatt’s period at Doncaster is best remembered for his Large Atlantics which quickly became the mainstay of top-link working on the Great Northern main line, even after Grouping. This was particularly so on the Leeds route with its prohibition on Pacifics on the approach to Wakefield until the bridge was strengthened over the River Calder.  His experiments with compound steam working were not a success, but his introduction of superheating was capitalised upon by his successor.  He is also remembered for his 4-2-2s which he optimistically hoped would eclipse the Stirling Single-wheelers.  With a life span of under ten years they became casualties of war, when they were put into storage in December 1917 so that their tenders could be utilised for new engines.  He was succeeded by Herbert Nigel Gresley (1st October 1911).

RETURN TO INDEX