Wheel Shop

Discussion in 'Workbenches, including workshop techniques.' started by JimG, 26 July 2019.

  1. JimG

    JimG Western Thunderer

    That might be the way to go in S scale since we prefer all metal driving wheels to use with split axle insulation/pickup. I can obtain wax or castable resins so there's no problem in producing masters. The one problem we have found in the past is getting a suitable brass. We carry a small range of generic cast driving wheel centres and the member who set this up some years ago had to try several casters until he found a brass which would machine well. However there is a trade off since the finish quality is not quite as good as can be achieved with other brasses so a wee bit of finishing work is required.

  2. JimG

    JimG Western Thunderer

    I post cure in a bucket with an UV LED strip stuck round the inside and it seems to do the job. I have also purchased a lamp which is a guaranteed 405nm wavelength and I might set that up to see if it is any better.

    I was cleaning post printing with an airbrush spraying IPA and that seemed to be doing a good cleaning job in quite a quick time and I was able to blow dry the parts to get rid of the IPA - dual action airbrush. :) However I was getting a bit worried with the amount of IPA vapour floating around so I have invested in a small ultrasonic tank and that works very well with IPA in the tank and three minutes of immersion.

    Thanks for the pointer to the Halfords etching primer. I normally use their standard grey and red oxide primers but I'll get a can of the etching primer to use.

  3. Bill Bedford

    Bill Bedford Western Thunderer

    These resins are cured with UV light, so simply painting them will stop them from continuing to cure.

    Castable wax resin is problematical if a commercial foundry is used. The burnout cycle is usually longer than for standard waxes and not all foundries are willing to upset their normal workflow to accommodate the odd pot of brass castings.
  4. JimG

    JimG Western Thunderer

    I might finish up just providing a single master to the casters for them to make their own wax masters.

  5. Brian McKenzie

    Brian McKenzie Western Thunderer

    I'm watching Jim's work with much interest :thumbs:

    As Overseer has suggested, casting driving wheels from printed wax patterns (not wax polymers or resins) has proved very successful for me with many hundreds made. Some recent waxes:

    Hunslet____wax diameter 1.136"

    Sharp Stewart - with bolted on addition to counterweight____wax diameter 1.462"

    Bogie wheel____wax diameter 1.030"

    These have been printed commercially at 16 micron layer height using a '3D Systems' machine. Stainless steel tyres are fitted after turning the brass castings, for which a machining allowance is also added to the side faces.

    -Brian McK.
  6. Overseer

    Overseer Western Thunderer

    It will come down to balancing accuracy, repeatability and cost. 3d printed waxes will be more accurate and more consistent than waxes from a silicone mould. I expect 3d prints will also be cheaper for small numbers as they avoid the cost of making a mould, squirting hot wax into it multiple times, and waiting for them to harden enough before taking them out without distortion (a big problem for wheels but not for rings and other jewellery).

    I was thinking in terms of jewellery casters for production, finding one who is interested at the right price may take time. 3d printed waxes are now very common in low volume jewellery production so printed waxes shouldn't be a problem. Brass can be a problem, someone told the one I used to use that brass contains lead so they stopped casting it but they will do bronze, which can have machining issues. Talk to the potential casters about their investment pot size to work out the most cost effective batch sizes. You could also supply the metal for casting to make sure you get the properties you want.
  7. JimG

    JimG Western Thunderer

    We wouldn't be too worried about high accuracy in castings. Our range of cast brass driving wheel centres, which I mentioned earlier, have to be turned to size and bored so all we are probably be looking for is a good representation of the spokes and boss. Our caster of the range works from one master so he will be using a silicon mould and wax. But I will look at printing centres to use directly in turned steel tyres for those members who can put up with tyre scraper pickups. :).

  8. Rob R

    Rob R Western Thunderer

    Thanks Jim.

    Not all of us have access to a lathe these days

  9. Overseer

    Overseer Western Thunderer

    Jim, I realised I have a sample of a 3d resin print of a S scale wheel made by Lawrence Boul at NZ Finescale (NEW ZEALAND FINESCALE – Doing amazing things in 1:64 scale). I have had it for some years and it seems not to have deteriorated. It is for a New Zealand 3'6" gauge wheel in S scale, so fits a tyre about 12mm diameter. Lawrence has progressed dramatically with 3d printing since this sample. It could be useful to discuss your wheel plans with him.

    NZF wheel1.jpg

    And with one of my 1:48 injection moulded wheel centres.
    NZF wheel2.jpg
    Dog Star likes this.
  10. JimG

    JimG Western Thunderer

    Many thanks for that pointer - certainly worth following up. It will also be worth letting our SSMRS membership know of his range of parts since a good few could be useful for UK modelling - i.e. since the prototypes were built by British producers; like his Mackenzie Holland signal parts.

  11. Tim Watson

    Tim Watson Western Thunderer

    The castings produced by NZ Finescale are exquisite.

  12. JimG

    JimG Western Thunderer

    It's back to the wheels again. :) It took a while for the parts to appear but they did arrive at the end of September -


    ...1000 tyres and 500 axles. :)

    I was held up a bit by a bout of sciatica - a knock-on from carrying a heavy portable sound recorder around on my shoulder for a good few years. But I managed to get started on trying to set up a method for batch production. My previous methods were very much on a one-off basis per axle, taking several minutes to complete the axle with a lot of changing around on the Cowells during the job. With the prospect of making hundreds of axles I started investigating ways of doing things in batches and also on an easier setup than using a lathe,


    The tools developed so far are, from left to right, a holder for pressing axles into wheel centres. next right is the small brass insert to use in the axle press and above it is a piece to set up the press accurately. The next part along is a holder for the wheel when the centre is being reamed for the axle. Next right is a holder for ten wheels to assist putting the resin centres in the tyres and applying the cyanoacrylate. The three holders were CNC machined. Next along is the 2mm hand reamer with an adjustable stop to ream the axle holes in the wheels. The short brass rods are used to locate and press the axles into the centres. Above them is a split chuck to hold a tyre in the lathe, or to hold a wheel by hand. The brass bar beside it is a pusher to eject wheels from the split chuck.



    Just a couple of closer shots of the axle press. In the upper picture, a steel 8BA screw pokes through into the bottom of the recess and this is adjusted to stop the axle at the correct depth to give the correct back-to-back with the wheels equi-distant from the axle centre. In the lower picture, the small brass insert is placed in the recess and this provides a stop for the centre boss of the wheel so that it doesn't move during the pressing action. I found that the resin was distorting longitudinally when pressing the axle home and that was often giving inaccurate results when it settled back at the end of the action.


    Gluing the centres into the tyres using the cyanoacrylate, using a very fine nozzle. This particular glue takes about thirty seconds to set with the steel tyre and resin, so there's plenty of time to adjust things if need be. I started by applying the glue to the centre then pushing it into the tyre but that could get a bit messy at times, so I opted to change to inserting the centres in the tyres and applying the glue to the joint - much cleaner and no stuck fingers. :)


    Lining up the plate used to hold the wheel for reaming. The tool in the drill chuck is a tightish fit in the recess and the plate bolts are tightened while being held in alighnment using the tool.


    Reaming the axle hole.


    Pressing the axle into the first wheel using the press plate.


    ...and the second wheel going on to complete the assembly.

    The press tool being used is a light bench drill that I got many years ago in the 1970s from a model engineering supplier in South Queensferry under the approaches to the Forth bridge. I had gone over for a set of castings for a light bench drill and he offered me this machine ready built. I think I remember that he asked £10 for it. I got a good Jacobs chuck to go with it and the motor driving it is an ancient BTH 1/4 horse which powered a washing machine owned by the parents of a good friend. They got the washing machine when they were married in the 1930s so it's over eighty years old. :) The drill is square and there's no play on the quill so it makes quite a good light bench press.


    The results of the first trial production run using the tools. They were not quite so good as my earlier tests, with the TIRs spanning from 0.003" to 0.008" with the majority being around 0.006" TIR. This is still better than the TIR of wheels supplied by one of our commercial suppliers, but I would like to get down to the 0.002" - 0.003" range, to match our other supplier. Today I ran the process as a production run - i.e. fit and glue all the centres in the tyres, then ream all the axle holes, then press all the wheels on. I'll do a stagger through of the process and check at all stages with the DTI in the Cowells and see where the problems are happening. I have been relying on the resin printed centre hole for accuracy and I may have messed something up in all the other adjustments to outer diameters when adjusting for the new tyres.

    Last edited: 19 October 2019
  13. Rob R

    Rob R Western Thunderer

    Superb Jim.
    Thank you very much.

  14. JimG

    JimG Western Thunderer

    I had a production run with 96 wheel centres off the 3D printer, using the methods shown above. Some of the centres got damaged in the process but I finished up with 84 wheels on 42 axles, I decded to test them all to see how concentric the wheels were with my limit being a maximum 0.003" TIR to match the better commercial wheels we have in stock.

    The first job was to build a better test holder for the axles. With pin-point axles I could hold the axles fairly well in a "W" iron etch with a toolmaker's clamp applying slight pressure to make sure the pinpoints were properly seated. With the parallel journal axles, I couldn't get good holding with an etch, so I built up something a bit more workable.


    The holders are two brass trunnions (there's a good old Meccano word :) ) bolted to an 8mm alloy base. The bearing holders are two 1/8" steel rods with brass bushes pressed in the ends. The farther holder is locked in position with the screw, and the screw locking the closer one has a butterfly wing to allow quick release and locking when inserting and removing axles.


    ...and checking out the wheels with the gauge.

    The results were not quite as good as I had hoped.


    The 22 axles with wheels in spec are in the right hand container. The 16 axles with one good wheel and the 4 axles with no good wheels are in the central container awaiting fitting with more wheels. The out of spec wheels are in the left hand container, along with others from my earlier tests, for reclaiming the tyres. For reclaiming, I use the split chuck in the lathe to hold the tyre and use the tailstock to push out the centres. It takes a bit of effort so proving that wheels should stand up to a fair amount of handling. I then use a boring tool to clean out any resin and cyanoacrylate residue.

    So 60 wheels were in spec and 24 were out of spec. The out of spec wheels were not too far out, most of them being 0.004" - 0.005"TIR with the occasional one being in the 0.007"TIR region. These wheels are still better than the wheels from another of our suppliers which are around 0.010"TIR

    I'm going to have a think about this. It could be that I am fighting the system. The minimum resolution of the LCD screeen in the printer is 47 microns or about 0.002". It could depend on where the wheel centre is placed relative to the screen as to which pixels it lines up to and whether the central holes jumps the same way as the rim. I have applied anti-aliasing which is supposed to fudge the gap between pixels but I am unsure if it is having any beneficial effect. I have done another run of 32 centres to make up some more wheels and I'll hope to get enough good ones to finish off the existing out-of-spec axles.

    My next experiment will be to print centres with no axle holes and to drill the axle holes in the drill press before the reaming process. This extends the process of making up the axles but if it guarantees me a higher level of accuracy, it will cut down the processing time after assembly. It could mean that I have a level of reliability which requires no checking and we can have an exchange policy for any duds which slip through.

    Last edited: 1 November 2019
  15. Overseer

    Overseer Western Thunderer

    Hi Jim
    A couple of questions. Have you checked the wobble as well as concentricity? Or does the press set up ensure they are consistent? Will the resin move over time, could a wheelset take a set if stored with some pressure on it for example? Also, what is the reason for using superglue instead of Loctite 601 or similar to assemble to tyres to the centres?

    Sorry, that is more than a couple of questions.
  16. JimG

    JimG Western Thunderer

    The setup to press the axles into the wheels is set up to push them in square. I haven't actually tested for this so far but they look OK to the eye. I could turn the gauge 90 degrees and have a check.

    I don't know if the resin will move. I did leave a set of centres on the printer's plate for a day or two, uncured, and I found that they were oversize when I remembered about them and got them off and cured them. So I'm aware of the possibility of things changing over time and I'm keeping an eye on it. I've done more tests and found the diameter of the 3D print that just fits in the tyres with no excessive force so that there is room for a bit of expansion. That's the reason for the adhesive.

    I think the resin could take a set if subjected to pressure, and I think that could be said about moulded plastic centred wheels as well. Wheels are stored on the axles so that there can be no undue pressure put on them and I try and pack them carefully when being posted to members.

    On the glue I use, I got the Deluxe Roket glue since the supplier also sells the long thin nozzles for application. The glue has a slower setting time with the resin/steel bond and this helped with my initial methods of assembly when applying the glue to the centre before insertion. I've now changed to using Vitalbond Super Thin CA which is applied after the centres are inserted and it wicks round the whole joint. I'm getting the knack of just applying the correct amount. It is very quick setting and the bottle of acetone is kept handy to unstick fingers. :) I'll get some Loctitie 601 to see how I get on with that. I've been thinking about getting some more glue since the bottle of Vitalbond is getting difficult to use with build up round the nozzle.

    At the moment I have hit a problem with the reaming of the axle hole. I got consistently bad results on the next batch of wheels I did and I've narrowed the problem down to the reaming. I suspect that the setup can be off centre by a thou or two and the reamer can take the hole off centre rather than follow it.

    Rob Pulham likes this.
  17. simond

    simond Western Thunderer


    I believe 601 is one of the products that lock metal bits together by expanding to (over-)fill the gap into which it is introduced. If I’m right, it is not recommended for plastics, as it will simply move them out of the way as it expands. I’m not in the office right now so can’t check, but we had some issues with this, I think it was 601.

    Dog Star and Ian@StEnochs like this.
  18. Lyndhurstman

    Lyndhurstman Western Thunderer

    I believe (memory serves, but generally heads for the bunker when trouble appears..) that Loctite 601 was recommended for 4mm wheel fitting for the Sharman/Gibson black plastic injection-moulded 'push fit' wheelsets.


  19. JimG

    JimG Western Thunderer

    I've ordered some 601 to check it out and I'll see what happens. It looks as though I should leave completed wheels for a month or two to assess if any further "development" happens. I do leave a fair bit of space for the glue to wick into - here's a picture from early on in the thread showing the construction of the centres :-

    With the present dimensions there is a 0.125mm/0.005" gap round most of the rim to accommodate the glue. I'm actually thinking of reducing this gap to 0.075mm/0.003" since the lesser diameter is getting very close to the lip diameter of the tyres and I occasionally get the inserts getting past the lip when inserting them. I'm modifying my way of inserting them to avoid this, but a bit more on the lesser diameter would help. Thinking ahead, if I get another batch of tyres produces, I will deepen the recess for the inserts to avoid this possibility.

  20. Overseer

    Overseer Western Thunderer

    Thanks Jim.
    I will probably pinch some of your techniques so just wanted to check your findings. I have used Loctite 601 for years for the same purpose without any problem, with glass filled nylon and acetal (Delrin) centres. There may be a more suitable product but I just use what is to hand. I wouldn't worry about shelf life as one of my bottles is at least 30 years old and still seems to do exactly what it did when new. I queried the use of superglue because I find stuck fingers annoying and the retaining compound avoids this. It might not suit your assembly method though.

    Plastics and resins can move with heat and moisture changes, as do the metal tyres with temperature changes. I had some acetal bogie frames stored in non ideal conditions that ended up twisted but managed to salvage them with hot water and tweaking. Being able to predict how the material will behave is the important thing for our uses.
    Dog Star likes this.