Building a Caledonian Pug in S Scale

[JimG]

When I got my CNC milling machine, my original, primary intention was to make driving wheels for locomotives in S scale, until I got waylaid into building coaches. However, I want to start making some locomotives, and one of the ones I hope to start is a Caledonian 498 class - the dock tank designed by MacIntosh. Dave Lockrie posted an excellent picture of one in the Caledonian Railway Association forum

http://www.crassoc.org.uk/forum/viewtopic.php?f=8&t=165#p975

In that same thread, Mike Williams also supplied an extract from a Caledonian plate showing the detail of the wheels with their large balance weights.



I used this picture as a guide to milling out a centre for the wheel and the results of the first attempt have been promising.

First, I dug an ancient bit of half hard brass strip out of the scrap box to use for a blank, set it up in the mill and faced off an area to 3mm thick.



Next came the first shaping of the boss and the balance weight using a 2mm diameter cutter...



Next came the final shaping of the boss, and the flaring of the spoke ends next to the boss. For this I used a 2mm diameter ball end mill. This cutter finished the boss to size, and produced a curved fillet around the boss to reproduce the flare on the end of the spokes.



Next came the main cutting operation, with a 1mm diameter slot drill cutting out all the apertures between the spokes, then making the main cutouts round the wheel to release it from the main strip. Four small tabs were left to retain the wheel in the strip to avoid any possible problems with the wheel moving and causing damage if it was cut right out.



You can see the waste bits in the apertures still in place since the cutter had not quite cut right through, but had left a very thin layer which retained the waste.

I then used a piercing saw to cut through the tabs round the circumference, and I pushed out the thin metal holding the waste bits, and I was left with the wheel centre.



You can see the remains of one of the tabs on the left hand side of the wheel centre. Eagle eyed readers will also note the centres drilled for the axle and the crankpin - a step that I omitted to take a picture of.

I had machined the boss to be 1mm thick to be turned down in the lathe to be 0.8mm thick - primarily to get a turned finish on the boss - although the milled finish came out better than I had expected. So the centre was drilled out and push fitted onto a mandrel on the lathe and the b oss turned down to 0.8mm thick, and the outer diameter was tidied up to get rid of what remained of the tabs.

I took the edge off the spokes with a scraper to get closer to the shape of the prototype's spokes and made a comparison with the prototype.




I've got close but I'm not quite there yet. I am going reduce the thickness of the boss - possible down to 0.5 - 0.6mm, and use a larger radius ball end mill to get a larger radius on the spoke ends. I'm also going to thin the spokes slightly.

I'm also trying to work out the best metal to use. I looked at the 3mm and 1/8" brass strip I had to hand and it was definitely not flat - the bit if old scrap was the flattest bit I had. I might have to source some clockmakers' plate which I believe to be flat.

Jim.

 

[Scale7JB]

Magnificent!

Which program are you using, as a couple of us on here have access to a CNC miller..

JB.

 

[Jordan]

I think we need some new Thread Tags, marked as "Front" and "Back".

You can guess at which end of The Class this Thread is being done already, though.....

 

[JimG]

Magnificent!

Which program are you using, as a couple of us on here have access to a CNC miller..

I start off in 2D in AutoCAD LT then use Vectric Cut2D to generate the CAM code for the mill, which uses Mach3 for CNC control. The CAM generation element is fairly basic so other CAM programs will probably do the same job and might even be less expensive (or even free). Cut2D came bundled with the mill when I got it and I find it very good for this kind of work (and the coach work I've done).

And Mach3 is excellent since it allows me to do offline runs of the CNC code to check that I've done things correctly before letting a cutter get anywhere near the material. That still doesn't stop me punching a cutter through the work support when I've forgotten to set the tool height.

I've also got Cut3D and an old version of Rhino for full 3D work but in that case Cut3D generates raster cutting code which doesn't really suit my machine. The KX1 mill has a slight Z axis issue with the head nodding slightly on reversal of direction giving a similar effect to backlash. It's only a thou or two but that can flatten out small detail. So I'm looking for a way of generating waterline CNC code which gets round this problem. I hand wrote the waterline code for my coach bolections, but that was a much easier task than writing the code for a driving wheel spoke with its tapers and flares. I'm still working on it. I could buy software that would do the job but you start to get into the "four figure sum" category which is a bit of a show stopper.

Jim.

 

[JimG]

Had another try yesterday afternoon with some alterations and it's sort of getting there.



Apologies for slightly rough finish although a ~8x picture isn't going to help matters. This was cut using a larger radius ball end cutter - 3mm - for the radius at the end of the spoke, and the depth of the boss was reduced to 0.5mm. I've just slightly over-cooked things when taking the boss thickness down and the apertures between the spokes are starting to cut into the edge of the boss.

I think the flare on the spokes now has to be increased. The CAD drawing of the spoke flares was done by using a simple circle and two straight tangents as shown in the first drawing and using the radius of the fillet between spoke and boss to provide the flare on the spoke. I don't think this is enough to match the prototype flare. Here's the first CAD drawing



and the second with the more elliptical end to the aperture between the spokes which I will use for the next cut.



This second drawing is still unfinished - I've still got to tidy up the apertures on each side of the crankpin.

This alteration might get the flares to match the prototype - or it might be a bit too much, but there's a bit of wiggle room to decrease the flare effect. I can now sympathise with people agonising over the finer details of complex curves on the front of a diesel cab.

I will now have a bit of a respite from milling until I can source a bit more flat brass strip or sheet, or I might have to mill the slightly distorted 1/8" strip I already have to a flat finish. It looks as though the wheel centres will be 2.7mm thick, so I've got a bit of depth to play about with when "flattening" it.

Jim.

 

[JimG]

I managed to find a bit of 3/4" x 1/8" brass strip in the workshop - just wide enough to make these wheel centres. In fact it is a very nice piece of hard brass which machines veryh well - I think it might have come from Whiston's many years ago.

I machined another centre this afternoon with the new CAD version with the flared spokes.





I haven't done any rounding off on the spokes since I wanted to turn the 2.7mm boss down to 2.5mm thick to see the effect. In the past I have always left my bosses a bit thick until all the work is done on the spokes so that any file/tool marks that get on the edge of the boss will disappear when the boss is machined down to its intended thickness.

This cut looks like it's getting there and with rounding off on the spoke edges, should be getting close to the prototype. I think that the prototype spokes had a slight taper along their length which I might possibly confirm when I get the drawing from the NRM. But to machine that I'll have to hand write some code and I will need a bit of time to get my brain round it.

I'll round the spokes off in this last example to get a feel for how that would look and turn up a tyre to get a feel for the complete wheel. I won't be using this one on the loco since I have not been taking care to mount the wheel accurately in the lathe so it is more than likely not concentric, but it will be good enough to see how it looks.

Jim.

 

[adrian]

That's a big difference and the flare on the spokes looks a lot better, very impressive.

 

[Jordan]

It's all just awesome.

 

[JimG]

That's a big difference and the flare on the spokes looks a lot better, very impressive.

I've got another task on hand at the moment - I started on the wheels while waiting for some items from C&L. But I'll see if I can get a bit of time to get the spokes rounded off to see what that looks like. From previous experience with wheel making, I know that the rounding off can make quite a difference to the look. I've also worked out another way of getting a taper along the spoke which will take a bit of hand coding so that will have to fit in with other things at the moment.

This is all just a way of trying to get a result using the most inexpensive way. Now, if I had one of these very expensive five axis machines with the matching expensive CAD and CAM software, it would be no problem. I might win the Euro Lottery this weekend and that might be a possibility.

Jim.

 

[ScottW]

Very impressive Jim. Are you planning on producing a master for casting or are you going to make all six wheels on the milling machine?

Scott.

 

[JimG]

Very impressive Jim. Are you planning on producing a master for casting or are you going to make all six wheels on the milling machine?.

Scott,

At the moment it takes just over half an hour to mill a centre so I will just make all six required on the machine. If I was going to make several locomotives ( or, more realistically, if the wheel would suit several types of locomotives) then I would make a master and get them lost wax cast. I've actually got some blue wax here which is used for milling masters directly on to wax.

The next wheel I will be looking at will be the five foot diameter one for the 1 class 4-4-0T - which also suits a fair number of other Caledonian classes. I've got a set of the SSMRS cast centres of the correct size but I have to put bigger bosses on them - as you've done on your wheels - and I'm not sure how much work that will be. If I have to fill in the apertures at the top end of the spokes to match the increased boss diameter, then that might be a problem, and uit might be easier to make the centres from scratch. I'm looking at doing some coding to machine a dished wheel to give a taper to the spokes to try out on the dock tank wheels and if that works, I can modify it to do other sizes of wheels.

Here's a cross section of the 1 Class wheel from the GA. There's a definite longitudinal taper on the spokes.



Note also how deep the boss is - almost half the complete thickness of the wheel - and I've noticed this on other Caledonian wheels. So it might be easier to start from scratch.

Jim.

 

[ScottW]

Note also how deep the boss is - almost half the complete thickness of the wheel - and I've noticed this on other Caledonian wheels. So it might be easier to start from scratch.

That is a deep boss. I would imagine that was quite a feature on the real thing. If I remember rightly I think I made the boss on my wheels about 15thou deeper than the thickness of the tyre. Out of interest I'll have to have a look at the GA drawing and see how thick the boss actually is.

Scott

 

[JimG]

That is a deep boss. I would imagine that was quite a feature on the real thing. If I remember rightly I think I made the boss on my wheels about 15thou deeper than the thickness of the tyre. Out of interest I'll have to have a look at the GA drawing and see how thick the boss actually is.

Scott,

I first noticed deep bosses on Caledonian locos when I was looking at a 104 Class tank - the "Balerno Pug". here's a clip from the GA of tyhat class - very similar to the Class 1 wheels



I'm also thinking of getting a drawing of the 782 Class tank which also has 4' 6" drivers - like the 104 Class - and it will be interesting to see if the bosses are similar in depth.

With the 498 Class tank wheels I don't think MacIntosh could have such deep bosses and still keep the cylinders within the loading gauge.

Jim.

 

[David Taylor]

Wow, that's very nice. I can't see why you need to improve on your 3rd attempt, but what do I know?

Can't wait to see them with tyres and on an axle.

 

[JimG]

Wow, that's very nice. I can't see why you need to improve on your 3rd attempt, but what do I know?

Can't wait to see them with tyres and on an axle.

David,

I'm up to my eyes in other things at the moment, so it might be a few days until I get round to the wheels again - I only started work on them while waiting on materials for another job coming in the post. I will round off the spokes in the last example and I've been having thoughts about writing the code to put a dished taper on the spokes, so I might make some progress on that as well.

Jim.

 

[iploffy]

you know that is really something when you can produce a wheel centre from a piece of flat brass, I take my hat off to you Sir

 

[David Taylor]

Well, I'm looking forward to seeing a thread on the loco, this makes me think it will be a very interesting build.

 

[JimG]

Well, I'm looking forward to seeing a thread on the loco, this makes me think it will be a very interesting build.

It will be a while yet - I'm just over two weeks into the NRM's six week delivery schedule for the GA drawing so it could be a month before I can start doing any drawings. But I should have the wheels ready by then.

Jim.

 

[JimG]

Well it's taken a bit longer than I thought, but I had a couple of projects to finish off and clear the workbench.

I did manage to take the guesswork out of drawing the wheels by finding teh official drawing off the wheels in the archive of the Caledonian Railway Association. So I have been able to re-draw the wheel to the prototype dimensions and I also was able to get the accurate dimensions of the spoke flares.

So, as for the first attempt I used my CAD file of the wheel to generate the basic code for the wheel using the Cut2D program - basically the wheel circumference, the balance weight and boss outline and the spoke apertures. The flaring would need full 3D code and I actually spent most of last weekend starting to look at writing a program to generate the code. But I think life is too short. So I opted to write the 3d code by hand, generating the point data in CAD, copying this into a spreadsheet to calculate offsets where necessary, then copying and pasting the values into my hand code. It took a couple of days to do this and it worked out very well. Here's a sample of the GCode doing waterline cutting of the smaller end of the boss



It's a very old programming language and is quite basic compared to the much more sophisticated languages of today but its simplicity means that you can learn it quite quickly.

I did the first cut in Plastikard to save on cutters if my code wasn't quite correct. Here's the wheel after cutting and still in the sheet.



And then being taken out of the sheet and cleaned up a bit.



The boss is 0.2mm too high at the moment so that if its edges are marked while I'm working on the spoke shaping, that marking will disappear when the 0.2mm is machined off. There are one or two small points where the code is not quite right and I'll do some small adjustments, but overall the wheel will look OK when I've rounded off the spokes - that by hand. I haven't worked out the 3D code for that yet.

I'll now have to send off for some nice flat brass plate.

Jim.

 

[JimG]

Got the brass plate a day or two ago and started to cut brass to see how things turned out. the plate is clock plate which should be flat. It should be cut from a large, flat sheet with something like a bandsaw and not a shear or a guillotine, and this bit of plate looked nice and flat when I put a rule along it. It also cost a pretty penny and I almost looked for the hallmarks on it. First I had to modify my hand written 3D code to get rid of the small "dent" next to the boss. You can see it in the lowest picture in the message above - in the flare of the lowest spoke to the left of the boss - it was also on the right-hand side as well. This meant re-routing several of the passes of the cutter to avoid this.

I did try using double sided tape to hold the plate down, as I had done with the Plastikard sheet, but the first attempt didn't work well and the plate started moving, so I went back to clamping the plate down onto some 12mm MDF, and started cutting.



The boss shows up in the middle, and the balance weight and the material for the forming of the flares around the boss is the next layer down. Although the picture doesn't show it clearly, the deeper level round the rim has also been cut. The circular depression to the left was my first start - I had mistakenly used the tool offset method for working with my double-sided tape setup and started cutting 0.8mm too deep.



The next action was to cut the flaring and taper on the spoke area of the wheel using small ball end cutters and my hand written code. The grooving apparent is the result of cutting with small diameter ball end cutters and actually looks worse than it actually is. You can see that there is a bit of waste remaining to the left and right of the crank pin end of the boss but that lies on an aperture between spokes and will disappear when the spokes are cut.



The next operation is cutting out the spokes and here's the 1mm cutter starting on the spokes to the right of the boss.

After all the spoke cutting.......



and then a quick suck with the workshop vacuum......



The cutout round the rim was also completed after the spokes were cut leaving four small tabs to hold the wheel in place.

One thing I didn't take pictures of was the drilling of the centre holes for the axle and crank pin.



The axle hole was drilled undersize at 3mm and a rod turned to to be a push fit to give me a good way of holding it for the spoke shaping exercise. The tools for the shaping are the triangular shaper and thin strips of wet and dry sandpaper. The scraper takes the edge off the spokes and the sandpaper smooths them off. The sandpaper doesn't last long in this job and the worktop finished up littered with umpteen, shredded bits.





The right hand spokes have been got at..........



..and then the left hand ones, and the whole wheel face has been given a rub with a glass fibre brush to get rid of the milling machining marks. The marks at the crank pin hole will disappear when the excess 0.2mm is machined off the face of the boss. I can also see some parts which I might re-visit now that I'm looking at pictures about 6x size.

The next part of the process will be to fit and machine the steel tyre, but I need to make some bits to do that and the first will be a split chuck to hold the steel blanks. I started on that this afternoon



This has to be done out in the garage workshop where the ML10 can handle 1 1/4" steel bar and a 1/2" drill. I doubt if the Cowells would do this - I think the 1/2" Jacobs chuck with the 1/2" drill in it is probably longer than the Cowells.

Jim.