Wheel Shop

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

  1. JimG

    JimG Western Thunderer

    As Parts Officer of the S Scale Model Railway Society, I've run into a problem with obtaining reliable supplies of wheels from our suppliers. The problem will most likely be exacerbated fairly soon since we have commissioned masters for several wagon bodies for resin casting. I've got plenty of split spoke wheels but the wagon prototypes will be Groupling era, so solid spoke wheels and three hole disk wheels will be required, and I am pretty well out of them. So I could have the embarrassment of having new wagon bodies available with all the bits to go with them - except for the wheels. :(

    With my recent messing around with the 3D resin printer and printing parts for the 1:32 scale loco directly on the base plate, I found that parts were coming out quite accurate and square. I had also seen other people using resin printer for wheels centres. So I decided to have a go at some wheel centres to see how I got on. I had a small stock of turned tyres which had been produced some years ago when the Parts Officer of the day also started investigating DIY methods. I also found out that the 2mm Society made up their wheel sets and I had an excellent exchange of information with their Parts Officer about how they did it. But their turnover in parts is way more than the SSMRS and their use of moulded plastic centres ordered in thousands could not be justified for us who think in tens and occasionally hundreds. :)

    Another test run of the three hole disc centres just after the printing has finished and they are still on the base plate. This is version "L" to give some idea of the fiddling about that has been going on. :) Printing straight onto the plate avoids using the normal thick base raft and supports for parts so a lot of resin is saved and the time to print is much lesss than if rafts and supports are used. The Stanley blade is the best tool for getting them off the plate.

    This is the image from Fusion 360 for the disc wheel. I wanted to glue the centres into the tyres with cyanoacrylate glue so there has been a fair bit of experimentation with allowing clearance for the glue yet still allow accurate location. I found that if the centre fitted too well in the tyre, there was insufficient gap for the glue. If I reduced the diameter to provide sufficient clearance around the rim then there would have been play which could have meant loss of concentricity. So the above design is used where there are four spacers around the centre to fit well in the tyre and the reduced diameter of the main body provides clearance for the glue. There has been a fair amount of experimentation until I got the diameters which seemed to work.

    Here are the centres with a couple of the steel tyres. The cocktail stick is the best tool to hold them. :)


    The small blobs of glue have been placed on the edge of the centre. I'm still trying to acquire the light touch to get small blobs - the one on the top is a bit large. :)


    Six wheels with the centres inserted. The cyanoacrylate takes about fifteen seconds to set so there is a bit of time to make sure that the centre is seated properly.


    The wheel is then held in the lathe and the axle hole is opened out with the taper end of a 2mm diameter hand reamer. The brass collar sets the depth so that the taper hole provide a good press fit.

    The next process is to fit the axles and they are pressed into the wheel using the lathe at the moment to make sure that they are fitted square.

    The second wheel is fitted to the axle and the axle is pressed in until the back-to-back fits a gauge.


    Two axles finished on the back-to-back gauge.

    Finished disk and solid spoke wheel sets with more trial wheels.

    The next test I have to do is set up a small rig to check the concentricity of the wheels. I'm hoping that the wheels can be put together using a bench drill for reaming the axle holes and press fitting the axles. I've use my lathe for this purpose primarily to get the axles pressed in square. But the 2mm reamer can not guarantee that the axle hole will be concentric since it is flexible enough to follow an off centre hole. So I'm hoping that the 3D printed centres will be accurate enough to be assembled concentric in a drill press with no need for special attention to the axle hole. If this is not possible then I will print the centres with no axle hole and use a setup in the lathe to drill and ream the axle holes.

  2. Brian McKenzie

    Brian McKenzie Western Thunderer

    Two thoughts: Do you have a stop to prevent the axle being pushed too far through the first wheel - but primarily to stop the axle being pushed further as the second wheel is being added ?
    To firm up the reamer to get its business end more rigid, perhaps it could be mounted with only a short length protruding from a block held in the toolpost ?

    -Brian McK.
  3. JimG

    JimG Western Thunderer


    I intend to put in a backstop so that the axle is pushed through the correct distance. However I'm using pinpoint axles at teh moment and it could be that we change to parallel journal axles in the future, so I'm holding up on the jig until that change comes about. I'm thinking about milling a holder to use on a drill press and maybe have different backstop inserts to use with different axle lengths.

    I'm hoping that I might be able to use the hole as printed. At the moment I put a 1.7mm drill through the holes when the prints have been cured since holes can get partially blocked with excess resin. The reamer then uses this 1.7mm hole as a guide for forming the taper. If the 1.7mm hole is out of true then the reamer will follow the eccentricity. I will check out the concentricity of the wheels I've done so far and if they are good to +/- 0.001" then I'll stick with the printed holes and ream them in a drill press. If they are more eccentric, then I'll have to print the centres with no hole and drill and ream them in an accurate setup - preferable not in the lathe. All good fun. :):)

  4. JimG

    JimG Western Thunderer

    I set up to test the runout on my wheel samples.



    I used one of my "W" iron etches with pin-point bearings to hold the axles and the toolmaker's clamp was used to ensure that there was no play in the bearings which might affect results. My Verdict gauge did the measuring.

    On my four axles, one of the wheels had 0.001" TIR, which was probably a miracle. :) Six of the wheels had TIRs between 0.002" and 0.003". The eighth wheel had a TIR of 0.006". I remember noting the reamer deflecting slightly when doing one wheel and that was probably this last one.
    But for a process which depended on the resin 3D print being accurate in diameter and which used the printed axle hole as a pilot for drilling and reaming, I was pretty happy.

    I then had a look at wheel samples from our two suppliers. I chose two axles at random of each supplier from our stores. For one manufacturer, all wheels showed a TIR of between 0.002" and 0.004". For the other supplier all four wheels had TIRs of between 0.007" and 0.010".!!!! I had never thought to check TIRs on suppliers' wheels before.:)

    My next action will be to change to a stronger resin. The grey resin I have used so far is the "cooking" resin supplied by Phrozen which has worked well for me so far. But they also supply a stronger resin which is black and I am thinking on using that since we prefer black spokes or discs. The increased strength could be an advantage, although the grey resin seems to be quite strong and stands up to the push fit axles very well.

    But I'm quite satisfied so far, achieving results which are comparable with our better quality supplier.

  5. Tim Watson

    Tim Watson Western Thunderer

    Nice to know that the SGMRS is still talking about wheels, just as it did 40+ years ago...

    PS. Looks good BTW.
  6. JimG

    JimG Western Thunderer

    As ever. :) I remember AGMs at Keen House in the 1980s when the one and only topic for discussion seemed to be wheels. :) We have found suppliers since those times but they all seem to become short term options. In fact the steel tyres I am using for my tests were a batch made for Trevor Nunn ten or more years ago when there was another supply problem and he had started to investigate DIY. About three years ago I found another supplier when we again had supply problems but that source now seems to have dried up. So I am going back to DIY to try and ensure a secure supply. 3D printing seems to be the answer the production of the wheels' centres which has been the main consideration up until now. I know that the 2mm Association use injection moulded plastic inserts for their DIY operation but we could never justify the minimum order quantities which are part and parcel of the injection moulding process, since our membership is probably about a tenth of the 2mm Association.

    I hand over to a new Parts Officer next spring so I'm trying to avoid dropping a mess onto his lap. :):)

    Rob Pulham, Rob R and Tim Watson like this.
  7. GrahamMc

    GrahamMc Active Member

    Couldn't resist pointing out that you've done a fantastic job for the society. Thank you, it's much appreciated.
    D G Williams and Rob R like this.
  8. Rob R

    Rob R Western Thunderer

    Well said sir!

  9. victorianman

    victorianman Western Thunderer

    I second that.
  10. jonte

    jonte Western Thunderer


    Thank you, Jim.

  11. JimG

    JimG Western Thunderer

    Just a quick update. I got a supply of the Phrozen Rock Black resin at lunctime today (ordered yesterday) so I did a run of spoked wheels using it.


    ...and here are the wheels - two axles fitted and four centres in front. As you can see, the black is translucent which is maybe a given since the UV light has to penetrate the resin to cure it. :) Users may be happy with this since the spokes are mostly hidden by "W" irons, but a quick spray with matt black or rust brown could improve matters.

    I also tried throwing them together to see how they came out. The pilot hole in the print was opened out by hand with a 1.7mm drill in a pin vice . However I did the reaming and axle fitting in the lathe since I haven't as yet made a fitting to hold the wheels square on the table of the drill press. The wheels were then put under the gauge and three came out at 0.002" TIR and the fourth one at 0.003" TIR so I was still getting the same readings as in the earlier wheels.

    The next job now is to order up the 1000 steel tyres and make holders for them when I get them. The tyres I am using at the moment are slightly undersize at a scale 3' 0" diameter and the ordered ones will be a scale 3' 1" diameter.

    ScottW, Rob R, Rob Pulham and 4 others like this.
  12. mickoo

    mickoo Western Thunderer

    Where do you order your tyres from, I'm thinking of doing something similar for some 1:45 euro engines but it is the tyres that are tricky to source.

    Cheers Mick D
  13. JimG

    JimG Western Thunderer

    From Steel River Engineering


    I think you might have had some parts from them for Finney7.

    I am providing a form tool to provide our SSMRS profile and you might have to do something similar for your tyre standard. But they have CNC equipment so they might be able to produce tyres on that and machine the profile.

    richard carr and mickoo like this.
  14. JimG

    JimG Western Thunderer

    And to finish things off, and keep the Chairman of SSMRS happy :-


    ...split spoke wheels using the Drawing 6 on this web site as a reference :-

    Model Railways: March 1972

    I might have to beef up the spokes very slightly since they were starting to flex when I was press fitting the axles. At the moment they are 0.3mm thick and I might thicken then to 0.35mm.

    I don't think I will have to produce the split spoke wheels for a while since we have several hundred axles in stock - probably enough for several years' supply. :) But I had better make some just in case there is a run on supplies - oh yeah. I mean this is S scale we are talking about. :):)

  15. Rob R

    Rob R Western Thunderer

    Very nice Jim.

    Do you think that the tecnique would work with loco wheels - in the range say of 4ft to 5ft 6 inch?

    On a slightly different note - do us plebs care if the chairman is happy? Nah!:p

  16. JimG

    JimG Western Thunderer

    Someone else asked me that and I think that it would work. I will have to brush up my CAD skills to cope with all the flare work involved with locomotive wheels then try one. I might have a go at a Midland 5' 2 1/2" driver since I have some suitable tyres available and it's a popular size with a few people.

    [Later] Just had a thought that I might have a go at some Britannia wheels as well. I have a superstructure made in 1951 at the time the prototypes arrived on the scene. I got it from the late Bernard Wright's estate many years ago and it was missing a working chassis. The thing that has stopped me doing work on it was the lack of suitable driving and carrying wheels to go with it. The loco did have a chassis in 1951 since there was a picture of it in the Constructor at the time and I think Bernard might have been rebuilding it to two rail since stud contact or outside third was the preferred methods of the time.

    Makes for an easier AGM. :):)

    Last edited: 1 August 2019
    Rob Pulham, Rob R and AJC like this.
  17. SimonT

    SimonT Western Thunderer

    Does your resin continue to cure from day light? The original Formlabs resin kept on curing and eventualy went yellow, brittle and occasionally warped. Their current resins don't do this, but it would be worth a check with your suppliers or a trial on a sunny window sill. Thin sections were most prone to this behaviour. There are acrylic primers that will stop the UV light getting through to the resin.

  18. JimG

    JimG Western Thunderer


    I'm waitting to see if anything does happen longer term. :) Parts do get a good cure time in my UV setup - typically an hour or more - and I hope this completes the curing. At the moment the wheels are left in the open on my workshop desktop near a south facing window so they are getting a fair bit of natural UV. I'll maybe accelerate any possible further curing by leaving them on the window ledge to get a much UV as possible to see if anything happens.

    I haven't seen any reports from user groups about longer term warping or shrinking of resins - any such problems seem to occur during the actual printing process itself, usually, it seems, from over-long submersion in IPA.

    Since the black resin is translucent, I might look at spraying the inserts with matt black primer which would improve their looks and provide UV protection.

  19. SimonT

    SimonT Western Thunderer

    I left my test pieces in a south facing window for a few months, this was with clear and white resin. My post cure took place in a UV Nail thing that my daughter had stopped using. As it was cheap I presumed that it would be a broad spectrum of the UV bandwidth rather than the wavelength specified for the resin; it worked.
    I discovered the IPA long imersion problem too. I got the best results by buying smaller clip lid containers (0.5L ish) rather than use the 2L job supplied by Formlabs. A vigourous shake in the 'dirty' one, and then a slightly more liesurely soak and shake in the 'clean' IPA container. I also used a oil painting brush to clean the surface of unwanted resin. With use the dirty resin ladden IPA was discarded and the 'clean'became the 'dirty', etc, etc. The Halfords plastic etching primer was recommended as a suitable primer on the Formlabs user group, cue frantic searching by US users!

  20. Overseer

    Overseer Western Thunderer

    For the loco wheels at least, wouldn't it be sensible to print with a castable wax resin and have them lost wax cast in brass? It would cost more but no risk of flexing or the resin breaking down over time. Any lost wax caster should be able to cast them and because the wax is printed instead of cast in a rubber mould there is much less chance of distortion.
    Richard Spoors and simond like this.