Flying Scotsman

Please note that I have condensed the following pages so please use the refresh button on your browser to make sure you are seeing the new pages!

This loco was started by my father some years ago and passed to me on his death. The loco was a two cylinder version of the design by Paul Wiese and Don Collin that appeared in the Model Engineer volumes 147 and 148. The build quality was not brilliant (sorry Dad!) and the boiler was a bit suspect so I decided to virtually start again from scratch salvaging what bits I could and adding the middle cylinder. I also decided to build the original A1 version with the higher cab as I thought this would give a bit more access to the controls.

Original chassis


New castings from N21/2GA

Chassis stripped down

New Boiler Materials

Work started by cutting out a new set of frames from 3/32" mild steel by the usual method of marking out one side and then rivetting both plates together using a few copper rivets in suitable holes. This went all to plan until I realised that I had cut the horn slots to the depth for cast horns when I was in fact using the built up type made from 3/8" square steel bar. This meant they were 1/8" too deep. Fortunately I had made the frames deeper by adding a curved top over the horn slots, as suggested in the ME articles, so this did not matter too much. I just made the steel horns a bit longer! The straight edges were finished by milling in the lathe and the curved edges by milling in the vertical mill using a rotary table. All a bit fiddly actually and probably not worth the extra trouble involved in setting up. It would have been much easier to finish by hand with a file! The horn slots were milled to finished size, again in the lathe using the vertical slide. The front of the frames were reduced to 1/16" thickness where they protrude through the running boards to give a better scale appearance.

Beforehand, I had drawn the frames full size with all the hole positions using Cad. I printed out this onto thick paper and used this as a template to centre pop the hole positions onto the frames. At the time I only had an A4 printer so had to print the template in 3 pieces and carefully join them together to give the full size template. This worked out fine and the template was exactly the right size, which says a lot for the accuracy of modern printers. Using this method saved a lot of time and fiddly marking out.

New frames and full size drawing

New frame stretchers were machined from 1/4" by 2" mild steel flat, the drag box from 3/4" square and the front buffer beam mounting stretcher from 3/4" x 1/2". I much prefer this method of fastening buffer beams to the frames. The 'traditional' method using bits of angle seems a bit fiddly to me and quite often the angle is not perfectly square which throws out the frame alignment. Fixing holes in the stretchers were carefully spotted through from the holes in the frames and drilled and tapped for the bolts. The frames were temporarily assembled and found to be square and true.

Trial assembly of frame and stretchers

Horns were machined from 3/8" square mild steel. A length sufficient to make one pair was milled in the vertical slide and then cut into two. After drilling holes for the mounting bolts the working surfaces were polished and case hardened to resist wear. They were then bolted in position using hex head bolts tapped into the frames. The advantage of using bolts rather than rivets is that the positions can be adjusted slightly to compensate for any inaccuracies in the frame slots. The original axleboxes were OK so were used again.

New axles were turned from 1/2" ground mild steel as the originals had been made from ordinary bar and were slightly undersize. I then had to make a crank axle for the new centre cylinder and I was a bit aprehensive about this. However the job went quite easily. The two blanks for the crank webs were soft soldered together and bored and finished as a pair so that they were identical. The axle was left in one piece for the time being, the centre piece would be removed when the webs and crankpin were assembled thus ensuring accurate alignment. To help positioning of the webs in the centre of the axle and to make it easier to saw out the middle piece, two grooves were machined halfway through the axle with a parting tool. The outside edges of the grooves corresponding to the inside edges of the webs. I decided to use Loctite for assembly and used type 603. Most people however seem to use 638 which I think takes longer to go off and gives a bit more time for assembly. The axle was assembled in stages rather than all in one go. Firstly one end of the crankpin was glued into one crankweb. This was nearly a disaster as the Loctite went off much quicker than I had expected and I only just got it in in time! The pin was then glued into the second web using the axle to align the two webs and a piece of steel between the webs to get them the right distance apart. The webs were then glued onto the axle using the two grooves to get them in the right place. The whole was left overnight to set properly and then two 1/16" steel taper pins fitted through the webs into the axle to be on the safe side. The centre part of the axle was then sawn out using the two grooves to guide the saw blade. Incidently the crankpin had been case hardened before assembly. Success, one crank axle!

On checking the driving and coupled wheels I found that some did not run truly and they were slightly different diameters as well. I then spent some time remachining the treads to correct this by bolting them to a small faceplate with the axle bore running truly. I also decided to cone the treads by 2 degrees as they had been turned parallel before. I did a trial assembly of the crank axle and axle driven pump in the frames and found that the heads of the hex bolts used to fix the horns to the frames fouled the pump eccentric and the water pump casting so the bolts in the relevant horns were replaced with recessed cheesehead screws. Also I had to mill a bit off the side of the pump casting. This would not be a problem with the two cylinder version as the pump is central instead of to one side.

Clearance problems with the pump assembly

The next problem was fixing the wheels to the axles and accurately quartering them. In the ME articles Paul Wiese/Don Collin describe a quartering jig for accurately setting up the crank axle. On studying this, however, I realised that it was impossible to use it as it relied on the axle being clamped in two bearings which are part of the jig. By the time the pump eccentric and axleboxes are fitted on the axle there is no axle left to clamp to! After a bit of thought I decided to adopt the method of mounting the axle between centres in the lathe and using accurate pillars to set the height of the crankpins.

For a three cylinder loco with the cylinder centres all in line, the angle between the crankpins is 120 degrees but on the Gresleys, the centre cylinder is at an angle to the outside cylinders so its crankpin must be rotated by the same angle to give the correct valve timing from the conjugated gear. The angle between the outer crankpins remains at 120 degrees but the angle of the centre crankpin in relation to the right hand outer and left hand outer becomes 112.9 and 127.1 degrees. I drew out a full size diagram in CAD showing the positions of all three crankpins and drew in a baseline which corresponded to the top of the lathe crosslide. It was then a simple matter to measure accurately the height of the bottom of each crankpin above the topslide and turn some pillars from steel bar to set the crankpin positions.

CAD drawing for crankpin setting jig

As a final check the CAD drawing was printed out onto thick paper and used as a template to ensure that the pillars did indeed set the crankpins in the correct positions.

Using the CAD template to check crankpin positions


When I was happy that all was ok and I had the crankpins in the right order (i.e. RH leading followed by the LH and then the centre) the pump eccentric and axleboxes were fitted and the RH wheel Loctited on followed by the LH one. The pillar for the LH crankpin was made from the one used for the centre crankpin by gluing a piece of steel flat to the bottom to make it the right length. The lathe crosslide was only just wide enough to take the two pillars for the outside crankpins!

Loctiting the RH wheel


Loctiting the LH wheel using rubber bands to hold the crankins against the pillars


The coupled wheels were then loctited to their axles in similar fashion.


The completed axle assemblies were fitted to the frames and I had a rolling chassis at last!


< Return to 2-1/2" Locos.......Next Page >

Page 2 3 4