7mm:1ft - 0 Gauge Railway Modelling.

Castings Manufactured by: AGH Miniature Engineering
Sadly no longer trading.

From AGH Investment Castings.            PDF version

(See also David Littlewood's additional notes at the end.)

Turning up cast iron wheels is seen by some people as a black art, indeed, I used to see it that way too however, it is not, but one does however get very black doing it!


Here is how I go about it.  I am an outright amateur so, 'proper' engineers and tool makers are advised now to move onto something else to prevent apoplexy.


Simply put, my method is based upon that advocated by W Castellan, which was published in MRJ No: 44 in 1991, much modified using Alan Harris's excellent advice to avoid overloading the spokes when turning up the flange and tread and avoiding the need for mandrels too.


Casting ready for truing upThe first thing to do is set the casting up in a set of soft jaws by its tread, having first filed off any casting pimples, and true up the spigot and the back of the wheel.


The soft jaws are turned down to be a slightly larger diameter than the wheel itself.  This provides more grip and a very positive location.  One can mark each wheel to indicate jaw 3 (or whichever is convenient for you) but I find it seems now to be unnecessary.


wheel back now trued up concentric with the treadThe next picture shews a wheel with the back and spigot trued up.  I get all the wheels to this stage before proceeding further so as not to keep on changing chucks.         Back to Top

cleaning up the face of the wheelNow change the jaws for a conventional set and mount the wheel in the chuck by the spigot.  The wheel then turning down the treadhas all operations completed to the face, tread and flange before being removed from the chuck.


Where the boss stands proud of the face of the wheel, skim off just enough of the rim to get a smooth surface all over.  In this case the boss stands proud by 1mm so I back off the cutter exactly that and then skim across the boss.


Turn down the tread to within about 0.3mm of finished diameter.  We need to leave enough metal in the corner so that the form tool has enough material to work on when forming the coning.

turning down the flangeThe next operation is to turn down the flange to within 0.02mm of its finished size.

Back to Top

We need now to set the width of the wheel so that the form tool can be set up correctly.  The wheel thickness is 5mm, NOT including the boss.


I set up a new cutter and measure its thickness. Then offer measure the cutterit up to the face of the wheel, note the reading on the dial, offer up cutter to wheelback off the cutter and traverse it along exactly 5mm cut a channel behind the flangeplus the thickness of the cutter.

The next operation is to cut a channel about 0.1mm deeper than the tread was cut.


This is to allow the form tool to shape the flange and tread without having to work hard at removing excess metal.  I then take off the remaining metal from the back too to ensure that it cannot foul the form tool.


measure the form toolNow, similarly to that which we did with the cutter, measure the form tool as shewn to see how thick the portion behind the flange is.


Offer up the form tool to the form too to the face of the wheelface of the wheel, back it off and traverse the form tool exactly 5mm plus the thickness we have measured.         Back to Top

frm the tread and flangeIt is now a simple matter to turn up the flange, coning and tread in one operation.  Measure the wheel diameter on the tread against the base of the flange, not the edge of the tread.

centre drill the wheelBefore we can remove the casting from the chuck, we need to drill out the axle hole.  For this I use a centre drill followed by 2, 3,  and 4.6mm stub drills.  Each is held in a separate chuck so changeover is quick and simple.  (There no room for a quick change tailstock tool holder on my machine unfortunately.)

Back to Top


the 2mm drillthe 3mm drillthe 4mm drillthe 4.5mm drillreaming the holeDrilling no more than 6-7mm deep, this is followed by a 4.74mm reamer held in the chuck and slid in by hand into the work while it is spinning relatively slowly. Once this stage has been reached, polish up the face and boss with some fine wet and dry wrapped around a file.


Do use a handle and DO NOT use wire wool.


Now, remove the wheel from the lathe.  One can now bring all the remaining wheels to this stage. Once all are completed, saw off the spigot for each casting with a hack saw in the bench vice.

turn doen the back until the spoke shew throughChange the chuck for the soft jaws, remount each wheel and turn off the back of the casting so that the spokes shew through.  All that remains now is to use a fine file to clear any residue not needed between the spokes.


There are two operations remaining, to drill out the crankpin hole and insulate two of the wheels in the set of 4. We will begin with fitting insulation.

Back to Top


turning the tufnol rodStart by turning some tufnol rod to 4.76mm, I turn a length of about 7-8mm and then put a slight chamfer on the edge with a file.

force the wheel on to the tufnolOnce this is done, using a flat faced chuck, the wheel is forced onto the tufnol centre.

face of the tufnel and drill out to 4mmThe tufnol is then faced off level with the boss. Using the centre drill, a 3 and 4mm drill, the axle hole is re-drilled through the insulation.  Now the wheel can be removed. face off the tufnal at the back to exactly 1mm deepWhen they are both done, it is back to the soft jaws so that the insulation can be faced of to exactly 1mm deep on the back of the wheel.


Crank-pin hole

a jig 'axle' to hold wheel.Now for the crankpin hole. For this operation an accurate vertical drill is required.  I am using my milling machine for the job.  A length of brass rod was turned up so that it made an easy, but not loose, fit for the size of the axle holes in the wheels and set up in a machine vice.         Back to Top


using the jig to centre drill the wheelThis was securely clamped to bed of the milling machine.  It is now crank pin hole in the completed wheela simple matter of fitting a wheel, lining up the boss for drilling and keeping the setting for all remaining wheels.


I found it easy enough to hold the wheel in place with my fingers but a more elegant solution would be to make a jig with provision for a stop too.


The crank pin hole is started with a centre drill and then drilled out 2.5mm.  The crank pin is made from 2.5mm silver steel rod.

threading the crank-pinA length about 8mm is fitted in the lathe chuck, faced off, centre drilled and then drilled out with 1.4mm stub drill.  Then the hole is tapped 10BA using a taper tap followed by a second tap.


It is now ready to take CPL crankpin nuts.  This will be cleaned with isopropyl before being fitted in the wheel with Loctite 408, but not until after the axle at the insulated end is turned down, fitted and faced off.

wheels with crank-pins fittedSo here then is the finished set of wheels awaiting axles. This set of wheels are to be fitted to an engine by someone else but this, broadly, is the method I am currently using for axles.         Back to Top

turn down axle for insulated holeFirst thing to do is turn down one end of some 4.76mm steel rod to fit the insulated wheel, in this case, 4mm. Turn a chamfer on the end with a file.


This section should be overlength for the thickness of the wheel, which in this instance is 5.56mm thick over the insulation. The next thing to do is cut the axle to length and face off the other end.

press wheel lonto axleI measure from the shoulder.  The length equals the back-to-back plus the thickness of the other wheel at the boss.

Then press the insulated wheel on the axle with a flat faced chuck.


I should have removed the crankpin first of course.

face off the axle in the insulated wheelFace off the end level with the boss and the wheelset is ready to be fitted to its frames.     Back to Top

completed setHere is the completed set of drivers, with crankpins now properly fitted.


One last job of course is to put a meter across the axle end and the wheel's boss to check that the insulation is OK. I would be very happy to hear from any 'proper' engineers out there who have got this far and can offer ideas on better methods.



David Littlewood's notes:


Dear Raymond,

I have read - two or three times - the excellent piece on your website about turning AGH wheels. I have never done this precise job before, though I have profiled Slater's wheels, and I do have a set of AGH wheels to turn.   I do have some thoughts on this process, which I hope you might find of interest.

 (1) When doing any kind of precision work on a lathe or milling machine, one needs to take great care to avoid backlash errors (unless you have a proper* DRO fitted). These can easily be 0.25 or 0.5 mm - sometimes substantially more, depending on how worn the leadscrew and nut are - and at this level of work that is just not sufficiently precise. I adopt the habit of always approaching the work, having first backed off the tool sufficiently to ensure forward movement is required. Doing it in this direction means the slide is always under forward pressure from the screw, and will not set back under pressure. I suspect you already know this, but it may be worth making it clear to any less experienced readers. It is also sound practice to lock any slides which are not in use for the current cut, if you have the facility to do so (and adding that facility if you don't is a very good improvement to the machine).

(2) I absolutely agree with your use of soft jaws for holding the wheels by the rim. What I recommend is that when holding them by the spigot, use a collet (if you have them). This has two advantages: firstly, although not perfect, you are likely to retain better concentricity if you need to remove and replace, than if you use a 3-jaw chuck. Secondly, collets and collet chuck (usually) have no nasty sticky-out bits to gouge your fingers, especially if you need to use a file on the work.

(3) If you want to ream a blind hole of 6-7 mm depth, you really need to use a machine reamer; these cut to exact diameter to the end (apart from a tiny 45 degree bevel) whereas hand reamers are tapered for a substantial part of their length. The taper is very slight, but it is enough to change a running fit into a seized one. Actually, I think reaming the hole is probably overkill for this application; it is really only required for a bearing fit, and where you are going to stick a shaft in with retainer, or a force-fit, you don't need it, and the rougher finish might even help. I know, I often ream holes myself when I don't need to.

(4) On the matter of reamers, you say you use a 4.74 mm reamer and then turn the tufnol to 4.76 mm. I assume this is to obtain a force fit - but where did you find a 4.74 mm reamer? (N.B. I bought it from Allan Harris, RW.)

(5) For drilling the crankpin holes, it is (as you suggest) more elegant - and better - to make a simple jig. Your method does not automatically align all the wheels at the same angle, though I agree that my jig relies on the Slater's square ends to do this, and I need to devise a modification or a new jig for AGH wheels.

(6) For axles, I recommend precision ground mild steel (PGMS). It is made to the same dimensional standards as silver steel, but is noticeably cheaper and much nicer to machine. Ordinary bright drawn mild steel (BDMS) is usually a thou or two undersize, and is best avoided.

(7) for bevelling the ends of the axles, there is some merit in using a 45 degree tool rather than a file, as the finish is smoother. Or maybe I'm just over fussy....   I think you are to be congratulated for making such progress with your lathe - do I recall you've only had it for a few years? - and for making the effort to pass on your experiences. Best wishes,   David.


 * By "proper" DRO I mean one which actually measures the position of the carriage, and not one of the faux varieties which just measure the rotation of the handle. These utterly pointless devices are often sold for use on milling machines, and IMO are an infringement of the Trades Descriptions Act, as they do nothing to avoid backlash errors. -- David Littlewood

Back to Top