'Canadian Pacific' in 7.25" gauge,
a new build Merchant Navy locomotive - Part 2.
by Andrew Giffen and Martyn Dix.
In the first instalment, we got as far as the running chassis on the digital drawing board, having tweaked the rod centres and return crank angles to get the timing as good as possible.
But what about those Bullied-Firth-Brown wheels, since they are not available from any of the commercial suppliers?
Yes, that was high on our list of things to tackle, and as mentioned, with my lathe unable to manage the almost 9.5" diameter, it was going to have to be outsourced.
Our success so far has been more about organisational skills than machining skills, and Martyn was soon to prove that he was also an expert in spotting sequential dependencies - a fancy term from project management -
so things we knew had to be outsourced had to be given attention as soon as possible, so that they could move forward in parallel.
Another very transferable skill that you would not necessarily associate with model engineering, but that we all have, is quantity surveying; fundamentally a juggling act of comparing cost and outcomes for all the different sources and methods that can be used, against what compromises are acceptable.
Martyn was a huge help again with this time-consuming task for the driving wheels.
Sometimes you can end up with a spreadsheet of all the available options, pros and cons, and navigating a cost effective path optimises the chances of success.
There are various sources for casting in iron and steel both within the hobby industry and out, leading to other options of whether lost wax casting gives cost effective advantages compared to sand casting, what CNC options would be out there, and how they all compare in approximate lead times.
A few versions of 3D files would be needed to investigate and get quotes for the different options.
I started by copying the BFB wheels from the works drawings as accurately as possible in CAD, with all the undercuts and hollows of the prototype, with a little bit extra on the wall thicknesses for lost wax casting: as much to the model - amongst others improved access to all the valve gear, especially the outside obviously, and using conventional valve setups all helps to keep things a bit more predictable.
In fact, the drawing would have needed another version, with extra thickness to allow for machining back to the final tread profile, if they were to be cast in steel.
For CNC quotes, we removed the undercuts.
Some specialists can CNC cut some of the undercuts, but again, we didn't feel the significant extra cost was justified.
And this 2nd version still has plenty of verticals which would need to be modified to make another version for sand casting, with the minimum 6 degree draft angles to be able to remove the patterns from the sand.
And once the quotes came in, we did in fact go for the CNC option, using a specialist just down the road from me: Lincoln Jigs.
They were the most expensive, to be machined from solid EN8, but they would also be finish machined to 7.25" gauge tread and flange profiles, ready for paint, axles and pins.
If using the other (casting) options, they would still need outsourced machining, and the addition of steel tyres, if cast in iron.
I could estimate these additional costs and steps needed, and the CNC option seemed best.
Upon receiving the machined wheels, the CNC tool marks were minimal, but I set to with various sanding tools to get to the stage shown:
And after a coat of primer, they were ready for the balance weights:
The same company also machined the 8 off trailing truck/ tender wheels, but their quote for the bogie wheels seemed excessive to us, no doubt there being only 4 of them.
However, the laser cutting company that took care of the main frames in EN8 were also able to cut blanks for the bogie wheels.
This was far easier than trying to part off discs from a bar, and was fairly straightforward, taking care to lock the lathe cross slide for each cut for all 4 wheels.
So, I kept that job in-house, it also being a comfortable fit on the Harrison lathe.
After turning from solid, my new dividing head made it easy to drill the holes
To the main frames:
Meanwhile, my usual laser cutting supplier didn't think they could cut the 2.6mm holes in the 5mm steel main frames I needed to tap M3 for the motion brackets, so we looked around and found www.lasercutting.co.uk who were offering to cut 5mm thick EN8 sheet big enough for the main frames, and they could also handle the 2.6mm holes, which would save me a huge amount of drilling.
The wheels were set aside at this point, with some nightmares lingering about crank axles, and I was hoping for inspiration.
Bulleid was keen on welding, and I was only too happy to follow this philosophy.
All the stretchers, motion brackets, running boards supports, drag boxes etc are TIG welded fabrications, using the slot and tab method to ensure alignment.
TIG is a very controllable and precise tool, and tacking up slot and tab assemblies of 1mm steel is very doable.
With all the MN fabrications being 2mm or thicker, this was straightforward, and finally behold the very first parts in metal, the loco spring hanger brackets:
The main frames bolt through to the stretchers, all the holes are pre-cut by the laser and tapped accordingly.
Some people have claimed issues with hardening of the laser cut edges and holes and go to the expense of using water-jet cutting, at least double the cost, but I have yet to find any problems with that.
In fact, I used a hand drill with the torque setting at 2, the tap dipped in cutting oil for each hole, and through it goes in seconds.
The frames were also free of heat distortion, but in anticipation of that possibility, they were cut as left and right-handed, so any distortion would cancel out on assembly.
And the Axle box keeps are also welded assemblies, as on the prototype, but a little simplified again, being invisible behind the wheel hubs.
On the full size, the inside expansion link bracket is integrally welded with the inside motion bracket, and the expansion link is assembled in situ between the frames.
For the model, that requires the expansion link to be assembled and installed with the inside girder bracket removed.
Any subsequent attention needed, such as replacing the die block, would not just be a boiler off job, (which it would be anyway), but also need the removal of the whole inside girder - quite an undertaking to replace something as small as a die block!
So, on my version, the inside expansion link trunnions are modified to be a separate bolt-on piece, facilitating that whole operation.
The same applies to the outside trunnions, the outer ones bolt on, so the expansion links can be complete assembled including the radius rod first.
That removable outer expansion link trunnion can be seen here, bolted to the motion bracket with 5 M3 hex bolts, and located very precisely by slots and tabs:
The end result of the main frames in primer, another big milestone, although I notice that in these photos, the foundation ring support brackets are missing, as well as the cab underframe, so they really belong with the complete main frame static assembly unit:
Introducing the boiler:
I've made very many silver soldered copper boilers for my 16mm and G scale locos, but this one is comfortably in a different order of magnitude, and was always going to be top of the list for outsourcing jobs.
So a Bulleid pacific boiler, and what jumps to mind?
Nicholson thermic siphons of course.
Here I have to hold my hands up and say I suggested we not have them.
What was my reasoning?
For a grate area of this size, they were only ever going to be a "nice to have", and it would be essential to at least every few years inspect them, so requiring the same inspection plugs as the full size, only this time access to them is not possible by climbing between the frames.
For them to really be needed, the boiler would have to be worked to its limit, and that is not likely the case in our scales, where we are lucky to have 3 passenger carriages on the drawbar.
Looking around at copper boiler suppliers in this size, and having excluded steel, the very first decision is between welded copper of the kind that Steam Tech in Kent have been doing for many years now, and the traditional silver soldered type, and in that category, a good handful of options.
I confess straight off that I had already used Steam Tech and each time was very impressed with not only the quality of their work, but their lead times and price were both considerably less than any other source.
I suspect a few years ago there were still some traditionalists who looked at TIG welded copper boilers with some scepticism, in spite of the inherent logic in the strength of fused copper joints, but now that they have been in regular heavy use for many years, it is just another of the copper boiler options.
They are able to work from custom drawings, which was also very handy in that I would be able to draw the boiler in position on my design in CAD, using the works drawings again, albeit with some practical modifications, and they would use my drawings and digital templates to copy it. That would logically mean it would fit in the frames with a minimum of fuss.
It would also mean that while the boiler had gone into their order book, I could continue with the cladding arrangement on my CAD template, instead of having to wait for the complete boiler to start that process.
Here's a look at the 3D design on the frames, with the cutaway giving a good idea of the grate area and heating surface area:
Steam Tech does have one limitation: they don't generally do the extensive forming that you would usually see for the throatplate on these, which is a considerable curved shape blending the barrel radius to the firebox.
I decided on a compromise there, with a flat oblique throatplate which would be disguised by correctly formed cladding.
This does shorten the fire grate, but I don't believe there will be any lack of steam, and the 5 radiant superheaters will also help mitigate that.
The model follows the full size faithfully in most other respects, like the steeply tapered barrel, combustion chamber and the Belpaire firebox, more things which could in theory have been omitted and disguised under cladding.
In the next instalment: more on the boiler, the important crank axle, and quartering (or is that thirding?) the wheels on to their axles.