Genesis of a model locomotive design

[John Bateson]

Genesis - the coming into being of something

Previous designs have been documented on paper - its easy to do, cross things out - make sketches etc but more recently I have begun to use Windows OneNote which is part of MS Office 365 (I use the cloud version). But, in a way, these experiences, both good and bad are private and given the usefulness of the society forum and the way experiences and benefits (plus a few things to avoid) are being shared I have come round to thinking that I should document the next project on the forum.

picture001.jpg

picture002.jpg

First - define the project
This is one I started late last year but events has caused a pause in its progress. It was easier to restart from square one rather than try to remember what on earth I was doing last November.
The project will be the GCR Class 11B, the first Robinson design - although really a continuation of his predecessor's work. This will be completed in 4mm P4 scale only and will be a one-off solely for my own collection.
The notes I will add to this blog(?) will cover

getting the information needed
assessing feasibility
acquiring the necessary tools - discussion on what is needed and what I did in the end
3D design - compromises for the modelling scale world
conversion of 3D design to 2D plates for the etchers
design of additional parts using 3D printing or other tools
prototype build, instructions and error correction

106.jpg

and having started all this I will have to acquire the discipline to keep it going - good thing - discipline!

John Bateson

[Andy W]

Nice, John. This will be interesting to follow.

[andrewnummelin]

Indeed an excellent approach - looking at your notebook I don't think you are short of discipline! My equivalent are odd scraps of paper....
I was planning to do something similar after Scaleforum but it won't be the genesis, it will be more a list of problems and mistakes with an appeal for help in making a better job second time round. I should try your more methodical approach for my next project.......

[jayell]

Indeed an excellent approach - looking at your notebook I don't think you are short of discipline! My equivalent are odd scraps of paper....

In my case it is mostly in my head.

I just don't have the mental discipline to be able to write things down in an orderly fashion so end up looking for the same info over and over again instead of knowing exactly where to find it from 'notes'

John

PS I do scan article from books and print them out, plus printing all sorts of stuff from this Forum and elsewhere but at the moment it is all in one big heap in a cupboard under my work bench. I could try getting my secretary (aka wife) to put it into binders and index it but I'm not sure she'd take too kindly to my suggesting that idea

[John Bateson]

Getting the infomation needed
The first place to search is the archives at the NRM.
From Oxford Publishing Company microfilm collection at http://www.nrm.org.uk/~/media/Files/NRM/PDF/archiveslists2012/railwaycompanyworks/Oxford%20Publishing%20Company%20drawing%20list.pdf a rather tedious search for “11B” and the “D9” reveals the following

4190 Weight Diagram 11B dated 1901
3982 Weight Diagram D9 dated 1901
6103 Outline Sketch 11D
6104 Outline Sketch 11C
6105 Outline Sketch 11B
3982 Weight Diagram D9 dated 1901
10346 Full pipe & rod arrangement of 4-4-0 engine Class D9
10387 G.A. of 4-4-0 passenger engine D9 Class GC/LNE
12439 G.A. of 4-4-0 passenger engine Class 11B (Vulcan) GCR
12440 Full pipe & rod arrangement of above engine

This list is rather longer than I thought it would be but from the descriptions I think that many were re-drawings of the originals viz #3982 referenced as D9, which it did not become until after grouping. It would have been nice to see one marked from Sharp Stewart who made the first 30, rather than Vulcan, who made the last 10, so the first compromise must be to assume that the two manufacturers used the same basic drawings.

Second place to look is E. M. Johnsons books “Locomotives of the Great Central Railway Vols 1 and 2”. It is here we begin to discover that the LNER D9 is not just a single design, the original GCR Class 11B being redesigned twice if fairly short order but with a complicated history of timing including some back-tracking on boiler fitting and some differences in cab sizes.
Third place is the bible, the RCTS Green Book “Locomotives of the LNER Vol 3B”, and this is where I really start to get confused. Ignoring all the difficulties with different types of valves, inside or outside admission, there are

2 boiler sizes – not necessarily linked with the upgrade to superheating
Different cab fronts and spectacles – linked to superheating
Different smoke boxes – linked to superheating and with frame extensions which may have been purely cosmetic to hide the gap a longer smoke-box would cause
Safety valve differences – not always linked to superheating

The nice thing to find out is that these engines could be found in later days from Birkenhead to Great Yarmouth – plenty of scope there then. They also ran into Chester Northgate, but that is a deduction only.

Fourth place

is the newly re-printed (2003) “Yeadon’s Register of LNER locomotives Volume 29”. Lots of pictures and the usual full works history.

From the above sources it is easy to note that photographs of the original Class 11B, before Robinson started to play around with boiler sizes and valves, are relatively few and far between and are of poor quality. The most useful ones all seem to be from the later LNER era. This will be a problem if I decide to build the originals.

Fifth place is the new book from John Quick, “Robinson’s Locomotive Liveries on the Great Central railway”. This has a full chapter devoted to the Class 11B/C/D during GCR days. This will become extremely valuable in the next few months.

We mustn’t forget the shows, such as Scaleforum, where the Stores will exchange a wodge of photos for a very reasonable wodge of cash.

All of the above as a very boring summary of the reading I have done even before putting drawing pen to drawing board.

[Will L]

Getting the infomation needed...

...a very boring summary of the reading I have done even before putting drawing pen to drawing board.

I can imaging getting thus far may have had its moments, but boring.. no.

Will

[John Bateson]

Assessing feasibility
This must cover several areas, date of model wanted, which locomotive, ownership at the time and lining and in particular the gauge.

First – can it be built – I guess the etched parts are fairly routine and a couple of manufacturers have made 7mm models which are still on the market. I don’t know of any 4mm scale models available and if there were any, a design for ‘00’ may not be acceptable for P4 but it would have been considered for purchase.

Are all the parts available?

Wheels – should be 6’ 9” but the only available P4 ones are 6’ 8” from 3 suppliers at a wide ranging cost. I shall fudge over the PIB/PIL options (pin in line and pin in between) since this is a subject of dispute elsewhere but I believe they are PIB. The tyres were 3” so a couple of years wear should have sorted out the size difference
Fitting the motor – should not be a problem for the usual High Level combination but the base of the gear box frame will show through the wheel spokes. The motor will have to drive the leading axle with the motor set rearwards and a 1220 should fit but it may be a challenge to keep the motor out of the cab area ('house' in GC terms.
Safety valves – the first 30 were built with 2-column Ramsbottoms which are not available, will have to scratch build or use 3D printing versions unless I assume that replacements which were started in about 1905 for the more usual 4-column version are valid.
Chimney and dome – since this is a one off it will have to be a hand-crafted job on the lathe since the pitch of the boiler doesn’t match any of my templates. Or I could add a new template to the etch.
Tender – originally built with the 3250 gallon tender – this appears to be the same as the 4000 gallon tender except that it is 4” lower at the top. These were replaced by the 4000 gallon tenders from 1905 onwards and do I really want to create a design for yet another tender?
Budget – must be completed within the cost of the average loco plus tender kit – say £160 – this is to include etches for the loco chassis in NiSi, the loco body in brass and an existing tender. Given the set up costs for etching these days this is very tight.
Inside motion – I want to produce this but it will be a challenging thing with my limited lathe skills to produce the eccentric axle. The valve rods and their eccentric attachment to the front axle I can crib from other models. The motion is quite visible so should be included.
Other small parts – springs, knobs, vacuum cylinders all seem to be straightforward.

Model choice – the options
The D9 includes the GCR Classes 11B, 11C and 11D. When we include a timeline with these classes plus the odd saturated engine with the larger boiler and/or the longer boiler with shorter cab, the smoke-box variations this comes out to 8 (?) possible models even excluding EM options. There are the usual variations in chimney and dome size which were in a few cases mis-matched and outside loading gauge.
I am ignoring for the moment the differences in the ‘green’ livery.

The KISS principle
Since I only want one model and that one in P4 the number of possible variations means that some rationalisation is needed. Accordingly I will build a single model #106 from the Vulcan Works at a time of about 1905 with original chimney and dome, 4-column safety valve, 4000 gallon tender. It will be DCC driven. The livery will be standard GC (the Sharp Stewart engines were initially quite different on the frames and the official (?) livery was not yet defined.)

[John Bateson]

And finally, the last bit of scene setting ...

Acquiring the necessary tools

It should be obvious that a drawing is necessary and that the purpose of the drawing is to provide a set of instructions to the etchers or to the 3D printers. The discussion of what software package is needed or even what sort of pen and ink methods are still valid causes a lot of heated debates. I will simply present my choice and the reasons for that choice.
I do my drawings in full scale in full 3D mode – that is my starting point for any design. I don’t have the spacial skills that would allow me to view a flat design in a folded up presentation. While producing the drawing I introduce the compromises in materials that are necessary to translate an original GA into a modellers GA – such as material thicknesses. For example, a footplate may be 3/8” on the original but to model that exactly means using 5 thou metal and that is basically infeasible for something like a footplate.
Since I end up with a full locomotive drawing in multiple colour and the ability to render it for pictures (and thus for assembly instructions) I need a reasonable drawing package. This is where the concept of deferred design comes into consideration and a definition is useful.

“Deferred design is design within a formal design process to cope with unknowable emergence of things we can’t plan.”

A computer is the epitome of a deferred design process. In other words, we never really know what we buy a computer for, only that we may find it useful in the future for something we are not doing right now. Translated, this means buy something that has spare ‘oomph’ if at all possible.

Hardware - to drive this software and produce complex 3D drawings a reasonable specification computer is needed. I have a quad-processor i7 64bit computer which is quite sufficient for the next few years. A lower specification computer will work just fine, it just allows a bit more thinking time between operations such as picture rendering and that is not always a bad thing in my experience.
Since the chimney and dome is non-standard as far as a purchase commercially is concerned, then a small lathe is ideal. At a pinch an electric drill held securely in a vice can do the job and I have seen successful turnings from such.

The software I use is TurboCAD v20 DL. At one extreme AutoCAD might be an ideal, if rather expensive solution, at the other extreme then something akin to the basic Google Sketchup, which is free, is certainly capable of much, but not a full locomotive especially when needing to export data or do solid modelling. Google Sketchup is the taster for the much more expensive PRO. Where packages such as Google Sketchup score heavily is the modern interface which seems much easier to use than more traditional packages– at least initially!

And finally a couple of soldering irons, one of which should be at least 40w and made in the traditional way with a big copper tip capable of holding the heat. Plus the usual plethora of small tools, which of course are in good condition, well-oiled and sharp where needed.

In a couple of days I will start to cover the setting up of the drawing and its parameters and then load the GA into the drawing.

[John Bateson]

Creating the drawing is quite straightforward. I am going to call it “Robinson 11B V0.2” to reflect that is it is a restart of an earlier drawing.
Set up some base line parameters :-
Set space units to ‘English’, type ‘absolute’, units ‘inches’, precision ‘3’ and format fractional. This ensures that the drawing will look like a GA when finished, measurements will look as presented on the original GA and the precision of ‘3’ will allow measurements down to 1/16” where it is needed.
Set grid to 1/8”, orthogonal style and lines – with major grid axis at ½” and minor grid axis at 1/8”.
The very small grid is a function of the materials we need to use, for example frames on the locomotives are normally 1 ¼” and the nearest NiSi equivalent is 1 1/8” or 15’. This allows the drawing to model the materials used accurately.
Angle style to ‘degrees’.
Layers – create as many layers as needed right at the start – I actually create 25 layers starting with
0 – base layer – default where everything goes unless it is put elsewhere such as datum line and datum measurements
GA – pictures of the NRM drawings
Frames
Frame Dimensions
Etc etc etc
Ending with
Spare1, spare 2 and spare 3
The rest of the options can be left until much later – if needed at all

Add some datum lines to layer 0 – red dashed
These make a ‘frame’ for the drawing
0,0,0 horizontally for about 500”
0,0,0 vertically for about 250”
And from the GA, we can also add lines to show boiler pitch, buffer height, platform height, frame length,
Wheel centre heights (remembering that we will be using 6’ 8” wheels).
All these are usually common to several layers and the best place is usually the default ‘0’ layer.

Import the drawing of the GA
Drawings from the NRM can be purchased on disc in which case they will come as TIFF files. Turbocad V20 can import these directly. I usually split the NRM file into two parts, top view and side view.
Set the drawing to side view, the import the side view TIFF file by selecting a spot and dragging the far corner to a large size of about 40’. Obviously this is not the correct size and it will not be aligned horizontally to any degree of accuracy.
Select the imported drawing, reset the reference point for it at exactly the point of the rail height and the rear of the buffer. Then reposition this to 0,0,0. At this point it should be seen to line up with one of the datum lines drawn earlier.
Now draw a line from 0,0,0 along the top of the rail the full length of the drawing, but do not release the mouse button. The guide numbers will show the angle of the line. Select the drawing, and rotate it by the shown amount.
Now draw a line from 0,0,0 along the measurement on the GA showing the frame length, in this case 30’ 9 ¾” and note what the length is. This will give you a scaling factor and entering this into the appropriate boxes will set the drawing to exact scale.

Now set the drawing to ‘plane’ view, i.e. from the top. Import the chassis part of the GA and repeat the alignment process but with the centre of the frame behind the buffer at 0,0,0..

If you now inspect the location of the wheel centres and other measurements you will begin to see that the original GA has stretched and not in a linear fashion. This will confuse things later!
However, it should be possible to draw circles representing the wheels at the ‘true’ positions, again on layer 0, and then add key dimensions so that they match those on the original GA.

006 Robinson 11B V0.2.jpg

Next – drawing the frames.

[jayell]

Assessing feasibility

Inside motion – I want to produce this but it will be a challenging thing with my limited lathe skills to produce the eccentric axle. The valve rods and their eccentric attachment to the front axle I can crib from other models. The motion is quite visible so should be included.
[

Whilst it would be possible to turn the crank axle on a lathe between centres, I have done it for a 2 inside cylinder 3.5" gauge steam loco, I think that for a 4mm scale loco you are going to need something like watchmaking lathe skills so you might find it easier to fabricate the crank axle.

It would in fact simplify making the connecting rod big ends as they would needn't to be in two parts as in real life, but could be assembled to the crank pins whilst pressing the parts together, it would also be easier fitting the eccentrics plus valve rods/eccentric straps if they happen to be in the middle of the axle on your prototype loco. It is still going to involve making and handling tiny components.

I don't think you would need to solder the axle parts after assembly as it won't be stressed too much so a suitable grade of loctite should suffice to fix it all together.

John

[David B]

Crank axles for the inside motion of Martin Finney kits are made by assembling the cranks, eccentrics and eccentric sheaths then force fitting them on the axle. They are silver soldered on to the axle and the parts between the crank webs removed. A carborundum disc rather than a hacksaw is recommended for this!

I have not actually done this yet - I have the pleasure to come. I have taken this from a MF instruction sheet.

[John Bateson]

Drawing the frames.
It’s now time to do some real work!
While usually I prefer to draw items in 3D first, the frames are one of the three items that are best done from a 2D perspective – at least in the initial stages.
The trick here is to draw lots of arcs (start, end, middle) and straight lines using the GA as a guide. There are very few key measurements on the GA and we should try to get as close to them as possible, after we have allowed for the 6’ 8” wheels. Some common sense is needed to do what is practicable rather than to achieve absolute accuracy.
Another compromise – the underside of the platform is 50” and the platform is 3/8”. We can either mount the 7/8” material to be used for the body on top of the 50”, or set the top of the footplate accurately to 50 3/8”. Each will require some changes to key measurements. I find it easier (especially for measurements when built and testing) to set the top of the footplate accurately and offset parts below this rather than parts above it. So, do we make the buffer beams ½” less in height at front and back, or do we just offset the holes to get the correct height for the buffers? Again, although it means that buffer beams if of the correct overall height will need the holes off-set there will be a defined ‘right way up’ when fitting them. Guide marks will be needed.
And another compromise – in the event that the finished engine will not go round my 4’ curves, the bogie will hit the frames and need to pass through them in extremis. It gives a horrible view through the frames but sometimes it is necessary. A guide circle for the P4 flange has been added to the bogie wheels and a couple of arcs marked on the side of the frame to show the positioning of a half etch that can be cut out if needed.
And another – in the majority of cases, there is no need to make allowances for undercuts when etched, but each case is checked anyway. For the top of the frame and the curve down towards the front footplate the frame size has been increased. A few pictures (where they show anything useful at all) have the frames very slightly proud of the upwardly curved part of the front footplate. This is something to key in and remember for when the etch layers are created.
Horn blocks – I prefer to use the Markits horn blocks, with the front small circle filed off. A couple of these have been positioned over the centre of the wheels for 6’ 8”. The frames can then be marked out for 1½” (0.5 mm equivalent) upwards movement, which I think is the recommended maximum. Previously I have used 2½” to give a little extra movement and simply soldered an ‘L’ in the hole where some restriction is needed.
Springing – small holes above and to each side of the horn blocks are marked and these will become half etches on the inside – to be drilled out if used. I’m still not happy with the CSB idea for a 4-4-0 where the bogie has no roll, pitch or yaw. No doubt it will be ‘explained’ to me!

006 Robinson 11B V0.2 2.jpg

And the last thing to do today is to select all the arcs and lines that outline the 2D frame, make a copy of it, moving the copy to somewhere outside the main drawing area and use the ‘chain polyline function’ on the copy to create a ‘closed’ polyline.

Next - but the frames aren't straight!

[Will L]

...I’m still not happy with the CSB idea for a 4-4-0 where the bogie has no roll, pitch or yaw...

Not sure I understand, with no roll pitch or yaw on the bogie I'd worry about any form of suspension, or lack of it.

Will

[John Bateson]

32. Bogie Springs And Hangers.PNG

There is a leaf spring between front and rear axle with equalising bars. All the bogie can do is turn on a vertical axis or slide left and right about 2.5".
On the model the bogie wheels are essentially individually sprung although on the full size bogie the equalising bar - equalises around a point at the bottom of the bogie frame which rotates independently for each side.

34. Bogie Completed Underneath.PNG

[Will L]

I assume that the equalising beam is only attached to the frame vis the leaf spring (exactly how isn't shown in your drawing but I assume the shackle holding the centre of the spring extends down to and pivots in that hole in the frame near the bottom edge). So both wheels on one side can be pushed up above, or drop down below the mean point at the same time by virtue of flexing of the spring, and as such it is logically no different to a CSB fitted bogie.

Will

[John Bateson]

What the plural of hiatus? Several of these have caused an unscheduled interval in progress.

Some more compromises.
The GA shows that the frames have several bends, just forwards of the leading driver axle box, straightening just in front of the sand boxes and then bending finally inwards again at 4’ from the rear of the buffer beam. This would be tricky to pull off in 4 mm scale. The only really noticeable bend when viewed at 4 mm scale is the front one. Its purpose on the prototypes was to allow the bogie wheels to move further about their centre point, although we decided earlier that an option was to cut the frames to allow the wheel to pass through the frames should smaller than standard P4 radius curves be used.
The second factor complicating this is the frame spacing. This has been much discussed over the past years and can be contentious. It has been explained to me quite forcefully that to produce a good model, the designer should start at the ‘00’ end and add modifications for EM and P4. I choose to start at the P4 end of the design spectrum (more later).
At the rear of the prototype, the frame spacing inside measurement is 4’ 2” equivalent to 16.67 mm. The wheels of the prototype are only 0.56” from the frames. The frames themselves are 1.25”, making a distance over the frames of 4’ 4.5” or 17.5 mm. As the P4 wheel back-to-back is only 17.67 – 17.75 this makes everything impossibly tight. Also to be remembered is that we have chosen to use Markits horn blocks which sit very slightly proud of the frames and this is 0.9” or 0.3 mm in 4 mm scale.
It’s time to draw in the frames by a small amount!
The standard/generic (?) frame spacer for P4 is allegedly 15 mm although many kit spacers are wider than this. Frame measurements are not defined anywhere as far as I can determine, but some kits have thicker frames which must impinge on side play more than others. I have selected a frame spacer width of 15.4 mm which gives a theoretical measurement over the outside of the frames of 16.83. (The etch cusp forms part of this and can be lightly filed off to give an additional, small easement of the frame width.) The outside of the frames including the horn blocks will be 16.83 mm – which should allow more than sufficient side play when added to the one inbuilt with the rail dimensions.
And of course any boss on the rear of the wheel must be removed.

Creating the 3D frames
From the previously generated polyline it is easy to create the 3D version – just look at the properties and define the width as 1.125” which is slightly narrower than the prototype but is the standard 15’ NiSi sheet equivalent. Also define it as ‘steel’, it makes the pictures look good!
And since we have decided to have only the front bend, we should decide on its angle. Simply drawing a line over the GA shows this should be about 1.5 degrees.
Now select the frame, choose ‘slice’ from the menu and choose two points 48” from the front of the frame. Select the front part of the frame which is now split into two parts, specify a reference point at 48” from the front and on the inside edge of the frame, choose the rotate tool and type in 1.5 degrees.
There is a lot more work to be done on the frames and it may even be necessary from time to time to put the bent part of the frame back to the original straight line temporarily.

006 11B Left Frame.JPG

[John Bateson]

Spacers – getting started
Next stage is to work out what spacers are needed and I prefer to put spacers in the same position as they were on the prototype since they are quite visible (especially when the 11Bs were superheated and the boilers raised. In this category I include the buffer beams, since they lock the front and rear of the pair of frames.
The buffer beams are 8’2” wide, the front is 16” high and the rear is 15” high. These are created as half full width boxes into which various holes and slots can be cut later for buffers etc. using the 3D addition and subtraction tools. Centre of buffer is 3’ 6” and centre to centre spacing is 5’ 8”.
It is convenient to reset the front section of the frame to the straight and add a TAB to it and one at the rear of the main frame. These are almost always 6” (2 mm) in size and usually 2” – which is greater than the frame width into which it will slot but it can be filed down flat after soldering.
Adding a TAB is again straightforward, create a box 6” x 1/125” and 2” length, place it against the front of the frame, I usually set this 6” from the top edge of the frame. Now make some copies using the ‘rubber stamp’ facility, one on top of the first TAB and the next two at the rear of the frame. These will be added to the frames and subtracted from the buffers to make a dimensionally accurate TAB and SLOT fitting.
The only thing we have to remember for the front buffer beam is to rest the bend for the front of the frame to 1.5 degrees, then move the box over the intersection of the front frame and buffer beam, and then subtract it from the buffer beam.
Note that the buffer beams are now keyed since the slot is offset vertically, which means that the buffer holes can be done next. The buffers have a square base and this is usually offset by a couple of ‘pads’. I use the Gibson buffers and by fixing one of the pads to the base of the buffer can drill through the holes in the pads to insert some 0.45 mm wire to act as bolt heads – the same wire will locate the buffer head accurately on the buffer beam into which 4 corresponding holes have been made. The hole in the buffer beam is set to affix the base of the AG buffer head and in this format is separated from the buffer body – which is not the usual way of fitting these.
I have also made a hole for the draw hook centred at 3’ 6” and allowing for a double layered draw hook which is 4” wide where it goes through the buffer beam. Simply draw a box and subtract if from the buffer beam.
And finally, the top layer for the buffer beam is simply a copy of the original but with the hold for the TAB of the frame removed. Simply draw a box on the top layer, make sure it is larger than the hole but the same width as the part and then use the 3D addition to remove it.

006 11B Front Buffer Beam.JPG

[Steve Taylor]

John, speaking as a complete ignoramus when it comes to the proper use of cad and its functions, and having only a passing grasp of the design process, I'm finding this a fascinating and enlightening read. Please post more.

Cheers
Steve

[John Bateson]

I thought I had done this part last week but must have misplaced it, probably in a similar place that the “thumbs-up” smiley so recently disappeared to…

The next spacer to be tackled is the one that supports the frame just in front of the fire box. This has a stiffening plate attached to the frames and a complex cross piece. There are three parts,
1. the frame stiffener, which can be located accurately to the frame by a couple of 0.3 mm holes and wires which means a couple of holes in the frames as well. The reason for the peculiar shape is that it must fit around the horn block mount.
2. Vertical cross piece – which will include a TAB and therefore this TAB requires a SLOT in the frame
3. Horizontal cross piece, just another piece of stabilization for the frame.
These are all made up of boxes and cylinders, added together or subtracted where needed. This is different from the complexity of the frames and the resultant solid has slightly different characteristics – about which we will not worry too much.
Having done this on several models to date, I noticed that the vertical cross piece was deeper than the earlier designs and will impinge on the spring wires. Note to self – deal with it #1
One benefit of full documentation is that it becomes harder to forget things.

006 11B Rear Spacer.JPG

Spacers – Part 2

And for this post, there is a frame stiffener just forrard of the front drivers. This is positioned again using a pair of 0.33 mm holes. The cross piece will be dealt with later since it will have a large hole for the connecting rods and two mounted blocks for the valve rods mechanism – angles yet to be worked out. This is why the frame needs stiffening here!

006 11B Front Spacer 1.JPG

Hair shirts – in another place ‘hair shirts’ is a pseudonym for P4 modellers who get carried away with detail, so if we are not too hairy shirted already, we are about to get totally qualified to wear the T-shirt.
Rather than provide simple spacers under the smoke box, it ought to be possible to produce something that looks a lot more like a cylinder block. So I am going to have a go, either as etched fold-ups or a 3D printed part.
And secondly, rather than provide a simple extended ‘U’ under the cab, which locks the frames dimensionally, since the rear driving wheels occupy some of this space, this block is going to be interesting. If I am going to go to the trouble of making the rear frame support anything like the original I should also make the buffers between the cab and tender spring loaded. It should be feasible!

Errata – earlier, under feasibility, I had noted that the final gear for the motor drive would be on the front driving axle. If I want to do the front driving axle complete with eccentrics then this cannot be. But putting the final gear on the rear axle is also problematic.
Solution – first get the High Level diagram showing all the combinations possible, cut out those which offer a possible solution, import them into the drawing and play around with positioning. It turns out that the Slimliner+ which I have used elsewhere will fit as shown in the picture below providing the motor is no longer that 20 mm and is mounted vertically.

006 11B Possible Drive Solution.JPG

[Horsetan]

Crank axles for the inside motion of Martin Finney kits are made by assembling the cranks, eccentrics and eccentric sheaths then force fitting them on the axle. They are silver soldered on to the axle and the parts between the crank webs removed. A carborundum disc rather than a hacksaw is recommended for this!

I have not actually done this yet - I have the pleasure to come. I have taken this from a MF instruction sheet.

I have. It's not the easiest thing in the world.

Here's how I built a crank axle for a Collett Goods

I don't have proper silver soldering equipment, so I used a portable gas torch. I'm not sure it worked all that well, so I ended up having to do a bit of surreptitious pinning. One thing to remember is to ensure that the crank webs should be an interference fit on the axle, so you want that axle hole to be 3.16/3.17mm and not 3.2mm dia.

I don't think a slitting disc is necessary to cut the axle; I found that careful use of a piercing saw blade was just as good and actually more controllable.

I do like inside motion, though, and have a crank axle to build for a D11 "Director". I'm having trouble sourcing appropriate cranks for the LNE J27, and Martin Finney absolutely will not supply the cranks separately - certainly not to the likes of me.

[John Bateson]

I have. It's not the easiest thing in the world.

But isn't that what makes it all lots of fun?

[John Bateson]

Rear platform
The rear platform of the chassis is the first part that will need some bending from the flat etch. Some thought must be given to the width of the half etch lines and their ‘foldability’. A few years ago there were a number of competing recommendations, some of which were, I felt, a little too wide. In the end I chose to go with those of the “Hollywood Foundry” which suggests that the ideal width is the same as the thickness of the sheet to be bent. In the case of 15’ NiSi this would then be – 15’. A slightly sharper bend may be achieved if the half etch is scored before folding.
Hollywood Foundry (http://www.hollywoodfoundry.com/) also have a very useful Scale Calculator which sits on top of CAD and other programs.

This part is a peculiar shape since the rear drivers are set under the cab and the rear leaf springs and axle take up some of the space needed for a really solid box.
It was simple to trace the general outline from the GA and create both the top and bottom layers, TABS were added to these and SLOTS created in the frames to accommodate them. Note that the bottom layer is quite a bit shorter than the top since otherwise it would impact on the rear axle. It was therefore not possible to make this rear box as a single item to be folded up, so two parts have had to be made with a single TAB and SLOT to lock them at the front. This was a little bit worrying since on other designs I have relied on this part to provide a solid, longitudinal lock on the frames.

006 11B Rear Frame Block 1.JPG

006 11B Rear Frame Block 2.JPG

The large hole towards the front end is for the coupling pin, which was simply a large lump of steel dropped through the cab floor and the two layers of this rear frame part, for the coupling rod to the tender. While most modellers may to prefer simple hooks on the rear buffer, I do like the idea of a long rod which, with a 1.6 mm thread on the end can enter the front of the tender and give a true ‘spring’ to starting.
There are also two smaller holes which were on the original and these were simple access holes. These will be used (and filled) with 12 BA screws to fit the upper body.

Access holes for the buffers are needed but only from the underside. These are rectangular slots which would allow the spring to be visible. A small plate at the front will provide the base against which the buffer will sit. The buffer itself will be a simple turning, fitted from the inside, with a small spring. There is a slight fudge here, since the frames are a little closer together than true scale, but it is just possible to have the buffers in the correct position.
The original part was a large, heavy casting something like the attached sketch. The model form is as close as we can reasonably get without also resorting to casting!

006 11B Very rough sketch of rear frame.JPG

And finally, just to be sure about clearances, I imported the leaf spring and fastenings from another drawing only to find there was an unfortunate interference! A few minutes of sotto voce remarks led to a more careful period of quiet contemplation. The leaf springs on this engine are rather smaller than the later Robinson engines, which with hindsight is probably quite reasonable since they are much lighter engines and the largest loads they pulled were five carriages on the London Extension. (The HS2 pre-quel to some of us).
Which means I am going to have to design a new set of springs.

[Horsetan]

I have. It's not the easiest thing in the world.

But isn't that what makes it all lots of fun?

Absolutely. Particularly if you're not using the "right" tools.

[John Bateson]

Cylinder block/spacer
Given earlier reservations about the solidity of the rear platform spacer for setting the frames accuracy, the front cylinder block arrangement must be a good solid fit.
It seems from the GA that if the smoke box is made of the usual concentric rings with the final wrapper being the only part bent to shape around a former then the smoke box cylinder is going to be below the top of the frames by a small amount, so it is important to check this out first.
First, make the ‘Boiler’ layer visible, change the drawing perspective to ‘front’ and create pairs of cylinders. For the first, the outer diameter is 5’ 3” (it is really 5’ 3.25” but I also use 21 mm brass cylinders available from Eileens etc as a measurement guide).
Then the outer diameter for the inner cylinder since we are using 12’ brass which is equivalent to 7/8” will be 5’ 1.25”. The cylinders are then subtracted, the smaller from the larger, and we have the boiler tube. We can adjust the length later.
The smoke box cylinder is tackled in exactly the same way, but with outer diameter 5’ 4.75” and inner diameter 5’ 3”.
And finally, the wrapper, which will encircle only the top 75% of the smoke box and then curved round to slot between the frames. This will be a full half etch equivalent to 7/16”. It is to be supported at the bottom end by a frame, part of which fits inside the smoke box cylinder to provide a flat front into which the smoke box door can be fitted. This will be dealt with later.

006 11B Cylinder Block and Boiler.JPG

Now we can determine whether the actual top of the cylinder block is feasible as a support for the model smoke box. So we must check fixing capabilities for chassis and body and there are two main possibilities.
1. Long screw through the chimney into a nut on the underside of the top plate of the cylinder block, quite easy to do and it works well.
2. Long screw through the base of the cylinder block into a fixed nut at the bottom of the smoke box assembly. This predicates a detachable bogie and an access hole through the bottom of the cylinder block.
Either of these could use a retaining screw to good effect.
For this exercise, the second option seems easier to do since I have a small lathe. The fixing supports for the Adams bogie under the chassis block is basically a large circular lump of metal around which the large circular lump of metal on the bogie can revolve. If the part on the cylinder block has a thread it can be fitted and removed easily and adjustments can be made for ride height using it.

The cylinder block can now be created, and it is simply a set of boxes, arranged in a square, with holes top and bottom for fixing screws/blocks. Add TABS as before, remembering to create a pair at each point, one for adding to the cylinder block and one for subtracting from the frame for the SLOT. It is probably a good idea to make these offset so that front and back of the cylinder block can be easily identified. For the purpose of this drawing I have also added the ¼ circle parts which on the final etch will be the half etch bend lines. These are shown as brass items rather than steel to high light them.

And to add a little more detail at the front, the front buffer support and the guard irons are created as polylines by drawing over the GA and then changing these to steel items 1.125” thick and set in place.
The front buffer support must be fitted using TAB and SLOT into the front buffer beam under layer and at the side into the front part of the frame. The TAB in this second case should be a little larger since it will be used to create a flat plate into which a couple of holes are created for fixing holes to the underneath of the front of the footplate. This is a rather fragile piece – so its belt and braces time. Into it is a small opening just in case we want a sprung draw hook.
The buffer beam support also has a small cut-out so that the usual Gibson buffer fitting can fit tidily. Note the height of this part must clear the central hole so that the buffer spigot can be bent in place to hold the buffer!

006 11B Cylinder Block Inside.JPG

006 11B Cylinder Block Top.JPG

006 11B Front Buffer Beam and Guard Iron.JPG

006 11B Front Buffer Beam and Footplate Fixing.JPG

[John Bateson]

And now for the "******" interlude
To complete the cylinder block I added tubes for the piston rod and the valve rod. The piston rod is is about half way towards the outside of the cylinder block and this can be accurately determined from the GA. The piston rod is somewhat thinner at 3.25” (3.75 on later engines) so fitting a tube through the cylinder block (1.5 mm outside and 1.0 mm inside) is quite feasible and is added to the drawing.
Previously I had indicated that fitting the chassis to the body through the chimney would be good, but once I had done the same action for the valve rods, except that they are inclined at 5.80 degrees, I realised that the fitting nut and these cylinders were trying to share the same space. Panic ensued – for a while and options like forgetting the whole thing to do with operational inside valve gear came to mind.
I remembered finally that the valve gear throw is just under a quarter that of the main pistons, at 6.75” and with the cover over the entry to the cylinder there would be plenty of room for a shortened rod to locate and not fall out when running. This is shown in the drawing along with a modified cylinder block which is now somewhat higher than the first version. The valve rod cover is a simple tube (two cylinders, the smaller subtracted from the larger) and turned through 5.80 degrees.
Raising the height for the top of the cylinder block will have repercussions later for the smoke box design.

006 11B Cylinder Block Modified.JPG

The valve rod has a variable geometry, at the maximum it is 3.75" where is passes through the frame spacer just behind the cylinder block. Thinning it down on the one side as per prototype may just be a cosmetic exercise after assembly of the rest of the motion using a small file very carefully!