1/32 A Loco for the Garden

Discussion in 'G1/32' started by JimG, 28 August 2015.

  1. richard carr

    richard carr Western Thunderer

    Thanks Jim

    I like Mach3 software I haven't tried cut2d before. I have a sherline based system but the Z axis isn't accurate it can drop suddenly due to the weight of the motor being attached to it.

  2. JimG

    JimG Western Thunderer

    I find that the Mach3 works well. I haven't delved into all the operations you can do with it but I find that it always seems to do what you expect it to. I got the Cut2D bundled with the KX1 and I find it to be an excellent 2.5D program. I've only just upgraded to the Cut2D desktop version with a load of additional bits and pieces like more CAD facilities so I'm still apraising these at the moment.

    I have to confess that the Z axis on the KX1 is the one area where I'm not too happy. The KX1 suffers from a problem on all Seig mills where the Z leadscrew is behind the column. When you reverse direction of the leadscrew, the head "nods" and the head assembly pivots on the guides on the column. So if your head is descending and you reverse direction, the head continues down for a small distance then starts to rise, and the reverse happens when changing from rising to descending. So you get an effect which looks like backlash at first, but isn't. You can minimise it by tightening up the taper gib on the head guides, but then you run into the problem of the head sticking if the gib is too tight. I can get round it to a certain extent by using the backlash facility in Mach3 but that only partially gets round the nod excursion and I still get problems like witness marks when cutting pockets when the cutter/head dips slightly when the head is lifted out of the pocket. But the advantages of using the machine far outweigh this drawback and I am finding ways to minimise the problem.

    Rob Pulham, Steve Cook and Dog Star like this.
  3. john lewsey

    john lewsey Western Thunderer

    Some people are just so clever great thread
  4. JimG

    JimG Western Thunderer

    After attending to other things, the wheels came back out on the bench over the weekend.

    At the moment I'm still scraping the spokes to get rid of the square edges. The two on the left were done last year and I've been working on the four on the left. The wheel centre at lower centre has been scraped and the one on the lower right has half of its spokes treated. The centres on the upper right have still to be worked on. After scraping to get the main amount of metal off, I use thin strips of wet and dry sandpaper to get a smooth finish on the spokes.

    All this work was being done while the milling machine was churning out the S scale sleepers mentioned in another thread - one of the advantages of a CNC mill - doing productive work with no human intervention to adversely affect it. :)

  5. JimG

    JimG Western Thunderer

    I can't believe it's been a year and a half since I last contributed to this thread. :) But I was thinking that that there was a danger of a bit of track getting laid in the garden some time soon and it might help it I had something to run on it. :) The spoke centres and the steel tyres had been maturing so it was maybe time to put them together.


    The first job was to open out the steel tyres to the required inside diameter - a scale 3' 10". I opted to try doing this work on the Cowells rather than the Myford. It would take a bit longer but it meant that I could do the work indoors rather than in the garage at the bottom of the garden. The Cowells worked quite well in the fastest backgear speed with up to 15 thou cuts. I also had to remember how to set the reversible jaws on the Cowells three jaw - I'm used to the Burnerd three jaw on the Myford with separate sets of jaws. :) The tyres were spaced out from the chuck jaws with a styrene washer so that i could cut right through without damaging the jaws.


    The next job was to cut the seat for the spoke centre. First I took a light across the faces of the tyre to true it up. There was now no need for the styrene washer to hold the tyre off the jaws so the tyre could be located against the face of the chuck jaws and a light cut across the face would true up any discrepancies caused by the styrene washer. After that I cut a recess to take the spoke centre leaving a lip at the front of the tyre which the spoke centre would butt against. I wanted the spoke centre to be a press fit in the tyre so it was a case of opening out the recess until close to the finished size, then opening out just the first millimetre of the recess until the spoke centre was an easy push fit, set back about one thou/0.25mm and finish the recess off to this reduced size to give a press fit.


    The spoke centre was then located in the first 1mm which had been opened out to be an easy push fit and the pad in the tailstock was used to push the spoke centre into the press fit.

    The spoke centre almost fully pressed home. It is handy that the Cowells tailstock pad is a nice size for 1:32 scale driving wheels. :)

    The next job is to open out the 6mm hole in the spoke centre to 1/4" and provide a bit of taper to allow a press fit. The reamer is a hand reamer with an end taper so it is possible to enter the reamer such that there is a short section of 1/4" parallel bore with the remainder of the bore having a reducing taper to give a press fit. The small brass collar sets the depth of cut to achieve the amount of parallel and taper bores. This position is found by advancing the reamer by about a millimetre at a time and testing with a piece of the axle material then the collar is locked with the 8BA screw.


    The tapered hole has been reamed and a test piece has been entered to check it. The nice thing about this method is that the wheels can locate on the axle ends and be held reasonably square. For adjusting quartering, the wheel can be pressed lightly until the taper starts to bite. This allows quartering to be checked and adjusted by twisting against the taper until all is well when the wheel can then be pressed fully on to the taper.

    The six wheels so far. The next job is to make up a holder to hold the wheel face out on a true mandrel and finish off the face of the tyre to width, then turn the tyre down to diameter. Th e slightly different tyre top right was made from a left over from the slicing operation. I messed up on the internal diameter on one of the tyres and the leftover was pressed into service. I think there is just enough meat to provide a full diameter tyre but if not, then the six of them will be representing a reduced diameter set. :)

  6. JimG

    JimG Western Thunderer

    I had a week or two off doing any 32nd scale stuff to re-jig all the parts in the SSMRS Parts Department before our Autumn meeting last weekend, but I managed to get back to the wheels yesterday.

    The next job I had done a couple of weeks ago was to make a mandrel to hold the wheels for final tyre turning. I got a plate offcut of 10mm cast aluminium some years ago to make an extension table for the mill, but changed my operating practices to work with smaller parts, so this bit of plate has been hanging around in the workshop for a few years now. So I cut a disc out of it on the mill and turned it up in the Cowells to take the wheels. I made the stub in the centre in the same way as I treated the wheel centres, with the 1/4" hand reamer partially put through a 6mm hole then a stub of the axle material push fitted in the hole. There were also a couple of holes drilled and tapped 6BA to us as clamping screws.


    The next job was to reduce the tyres to 4.5mm width - a straightforward job with the knife tool. I had thought of using my bigger ML10 to do this work but I thought I might try the Cowells to see how it handled the work, and it did quite well. I set it up in the third lowest backgear for this work and it handled the job with up to 0.2mm cuts.


    I then set the topslide over by three degrees, stoned a radius on the knife tool and cut all the tyres to radius, leaving the flanges slightly over in height and width - i.e. 1mm instead of 0.9mm in both cases.


    The next job, I thought, would be the tricky one in the Cowells - i.e. using the Mark Woods ScaleOne32 form tool. It's a pretty chunky 10mm square which would normally be way oversize for a small lathe. But as luck would have it, the tool sat pretty well spot on for height when placed on the topslide base, so I went ahead and finished off the tyres to profile. All the form tool is doing is shaping the flange so the load it comparatively light and only starts getting a bit heavy when the whole flange profile is being cut. I had to drop down onto the second lowest backgear (110rpm) to do the job without stalling.


    Here's the setup holding the form tool, with a clamp borrowed from the mill. I only had to do a small amount of filing to open the slot out in the clamp to get clearance on the stud on the topslide.


    The last job was to cut the small chamfer on the front of the tyres and that was done with the back of the knife tool set round to 45 degrees.


    The final products waiting for a chassis. :):):)

    The main reason for making my own was to get the quite deep centre bosses which were a feature of Caledonian inside cylinder locomotives. I think most available products tend to cater for outside motion locomotives where the depth of the crank bosses is usually a lot less.

  7. David Taylor

    David Taylor Western Thunderer

  8. David Halfpenny

    David Halfpenny Western Thunderer

    I like the use of a small Engine Lathe to turn ScaleOne32 wheels.
    While the back-geared Cowells 90ME is scarcely an 'economy' option (£3 000 including 3-jaw, 4-jaw and drill chucks) it's a very practical one indeed.

    [At the other price extreme, and using different machining methods, I've converted commercial wheels to ScaleOne32 using a Proxxon woodturning lathe costing £300, including those three chucks and a ScaleOne32 form tool.
    Foolishly, I shall attempt to demonstrate this at the G1MRA AGM - no pressure!
    I'm hoping to stimulate competition for the lowest-cost Finescale wheel option.]

  9. JimG

    JimG Western Thunderer


    It was really an experiment on my part to see if I could do the work on the Cowells and avoid having to work on the ML10 in my outside workshop. However it was quite a slow operation, being limited to up to 10thou off on each cut. I think I would revert to the ML10 for future work. It probably took a total of two days to do the work whereas I think it would easily have been about one day's work on the larger lathe.

    The use of backgear on the Cowells was necessary to get the torque and I do wonder if it would be possible to do the work on one of these small Seig C0 lathes which use electronic speed control. I don't know what kind of torque is available at low speeds on these small lathes. We have one at our clubroom and I might take a bit of mild steel up to see if it is possible.

    I have to confess that I was using a needle file to clean off rags on the edges of the flanges before using the form tool and I was quite tempted to keep going and shape the flange with the file - a process I've used in the past with quite good results. :)

    There might be a use for a larger version of the Fonly lathe

    Fonly Lathe (Part 1)

  10. Martin Field

    Martin Field Western Thunderer

    With patience, I've turned mild steel and silver steel on my Peatol (Taig) lathe which was the cheapest of the lot. 30 years later, I still have it and have had to replace just one drive belt! I bought an Indian 3 jaw which has so little run out you can barely measure it and a 4 Jaw for those awkward bits. It still doesn't total what even that Proxxon cost. It's made in Arizona, so you know it has good quality workmanship and materials.
  11. David Halfpenny

    David Halfpenny Western Thunderer

    Well since I run online support groups for the C0 and have one sitting in the middle of my kitchen, I'll have a go too, but not necessarily before the AGM. I'm reluctant to promote the C0 these days - things have got so bad that its main champion, Arc Euro Trade, have been obliged to discontinue it, (still do the accessories though) and Axminster have ramped the price up dramatically to pay for the heavy-duty support that purchasers need.
    If anyone cares, a C0 outfit with all three chucks mentioned is now £600.
    Having said that, if somebody already has a C0, or is prepared to treat it as a 'lathe kit' needing serious reworking to do what it says on the tin, or (like me) is prepared to forget slide-rest turning and use it for hand-rest turning, then I reckon it's up to the job, given ingenuity.
    To answer your question on torque, at the lowest speeds (control knob pointing to 9 o'clock), roughly 100rpm, I can stall the spindle with my bare fingers. It will still peel off a shaving, but not a very big one, and the current-trip often pulls out. Meanwhile, the DC motor is very stressed because, with the motor virtually stationary, the Back EMF is zero so the brief current pulses are huge, and the cooling fan is useless.
    The forté of the C0 is its high speed, but anyone with a C0 they don't mind cutting up could re-motor it for high torque / low speed.
    In terms of Economy Lathe Tactics:

    The first step is to ditch those handy offcuts of BDMS (Bright Drawn Mild Steel) bar.
    It's not only seriously work-hardened, but in bar form it has higher carbon content than BMS (Black Mild Steel), both factors making it harder to machine. (Opposite might apply if you want to hack washers out of plate though.)
    Better still is a free-machining steel like EN1a or Leaded EN1a.​

    The next step is to optimise tool efficiency.
    A sharp HSS tool with the right rake for the job is a joy. (The best tool grinding teacher is a treadle lathe - your leg tells you instantly whether you've got it right.) My own preference is for hand gravers, which respond instantly to the slightest tweak of rake.
    What gravers can't do, of course, is get a series of tyres identical first time round. So that needs dealing with at some stage.
    But don't write gravers off as mere 'ticklers'. To start with, a tickle is all we need when modifying wheels, but also a decent graver in a long handle on a big lathe can peel off angry smoking blue swarf just as effectively as a slide-rest tool (leather gloves and a face mask help!).
    Then there's chucking, which I'll pontificate on when I've done more wheels.

    Finally Feeds and Speeds.
    - Your back-geared lathe is having slow success with high-torque / low speed.
    - My Pulse Width speed controlled lathe is having slow success with high-speed / low torque.
    I'm probably making even finer swarf than you, but it's coming off in quantity!
  12. David Halfpenny

    David Halfpenny Western Thunderer

    Brilliant choice for this job, Martin!
    The like-for-like price on Peatol's website today (lathe + tailstock and three chucks) is £297 PLUS the cost of motorising (which many people will be able to lash up from stock). Vastly more solid and better value than the aluminium and plastic Proxxon (which is going to Godmanchester primarily because I'm an old git and can carry it with one finger.)​

    This entry is going to be tough to beat, but keep them coming, please :)

  13. Martin Field

    Martin Field Western Thunderer

    I can see that they've gone up a bit, but it is at least 30 years. It was priced at £160 when I got it, but I'd swapped it for the biggest piece of junk Emco ever foist upon an unsuspecting public, the Unimat PC, Ye Gods what a pile of rubbish. The motors burned out free running, the speed control failed, twice the three step plastic toothed pulleys fell off the too short shaft, und, und, und....
    The Peatol is the best machine tool I have ever bought. Pity their miniMill is so expensive by comparison, but I have a vertical slide for the Peatol and I made a milling attachment for t from an old minidrill and a dead fax machine decades ago with which I've made a LOT of wheels for model cars.

    BTW, if you want one, they come up on Ebay. My son bought one with every little bit of kit available, even the book on a nice stand with a British Thompson Houston motor for 200 quid! Only last year. It had hardly any work and was super clean.
  14. David Halfpenny

    David Halfpenny Western Thunderer

    It's a shame, because it so easily have been their best mini-lathe ever if those crucial points had been given attention. The bed was far stiffer than the original Unimat.
    I passed it over in favour of the Compact 5 which has a huge bombproof motor.

    The Taig lathe is easy to use because (unlike almost every other small lathe) it has rack saddle feed, not screw feed. OK it has problems too, but there's a community of fixers out there.

    The Taig mill is indeed more expensive, because it's a different scale of machine from the lathe it appears to resemble. And very suitable for CNC if one is inclined.

  15. Martin Field

    Martin Field Western Thunderer

    I'm no turner, David. I describe my work on the Taig as "lathing", or "lavin'". I get where I need to be with it, with no notion of whether I'm doing it right or not. I've worked that way since metalwork lessons at school, when old Doss Shepherd left me alone to make boiler fittings for model locos on the Boxford and Myford lathes that no other pupils knew how to use! I used to spend every PE lesson in there too. If the PE master wanted professional quality PA systems for sports day for nothing, he kept Mum about my hatred of his stupid subject!

    I've had Unimat SL90 (borrowed), Unimat 3 (special offer on the last day at the ME exhibition), Myford ML7 (too big and rocking with its linked leather belt, even on 2 layers of 3/4" ply floor), Simat (Flexispeed copy from Brian Wexler in Nth. Walsham), Unimat PC and finally the lovely little Taig, which has never given a moment's trouble. I bought a spare belt and an indexing attachment from Carter Tools, Stateside and have been completely happy with it. When I mentioned that the guy delivered it in his black MG Metro he told me the actual date of the delivery and it was indeed 30 years ago!

    I shall never build that 1/4 scale working petrol engine now so I'll stay with the little Taig. CNC doesn't interest me one jot!

    If I could do CAD, I'd be bloody dangerous!
  16. JimG

    JimG Western Thunderer

    A year and a half later, and I'm back to the Caley pug. Last year was "lost" to all the work on the US switching layout and the start of this year was "lost" to messing around with the 3D printer and S scale parts.

    But the more I worked with the 3D printer, the more my thought processes turned to using it for structural parts. It's easy to consider these wee printers only for things like loco chimneys, wagon axleguards, etc., - i.e. the cosmetic side of modelling. One of the methods I want to use in the locomotive is split axle insulation and with the loco having inside motion, I thought about using all metal hornguides and axleboxes and insulating the frame sides. I've done this before with an S scale locomotive where the frame spacers were made from machined Perspex and I was thinking about doing something similar in 1:32 scale. I then started to think about using resin from the 3D printer.

    So the last two or three weeks has been taken up by me messing around with Fusion 360 and starting a 3D model build. I've now got the first basic bits done and I aim to get the printer started work on frame parts tomorrow to see how they turn out. T he main consideration will be to see how accurate they turn out. I see a lot of comments about resin shrinking after printing so I might have to assess the results and apply correction if that problem emerges.

    Here's the view of the work so far...

    ...and an exploded view to show a bit more detail. The frames will use 0.9mm nickel silver which is slightly over scale. 0.8mm would have been spot on, but that gauge of nickel silver doesn't seem to be too common. The frame extensions to the rear are purely down to length restrictions on my CNC mill - I'm beaten by about 5mm to cut the complete side on my table, so I've cut the frames immediately behind the rear driving wheel and the joint will be covered by the wheel. The rear spacer is pretty well what the prototype locomotive had although I don't know if it had diagonal bracing. The GA drawing I have is annoyingly bereft of details in that area.

    With the cylinder and motion plate, I've succumbed to adding a few details. I was going to make bare bones structures to hold the frames apart, but decided to add some slide bars and cylinder detail to see how they look. I'm actually not sure how the underside of a Caledonian inside cylinder setup looked. One drawing I have of a Drummond Jumbo cylinder block shows what I think is a sheet covering all the warts on the underside. Drmmond, and his successors on the Caledonian, used two castings with a hefty joint down the middle with big flanges and lots of nuts and bolts. I've chickened out and depicted the sheet covering. :)

    I could put more detail around the slide bars but this area will be pretty well hidden by splashers and side tanks and I'll wait and see how it looks. If more detail would be worth it, I could modify the drawing and print another part.

    Fusion 360 also has facilities to produce CNC code so I'm going to investigate producing the Gcode to mill the frames from it as well. That might take another two weeks until I beat the program into submission and produce something workable. :) It really has been one step forwards and two steps back for a lot of the time and sometimes deleting all and starting all over again has probably been the best action.

    P A D, Dikitriki, Rob Pulham and 10 others like this.
  17. JimG

    JimG Western Thunderer

    And here's the result of a few day's work.


    ...three attempts on the cylinder block and a couple on the rear member. The main problem with the 3D printing was finding the best orientation of the part. The successul attempts are the closest to camera. The first attempt was done almost horizontal with the underside supported from the build plate. The result was this...


    The part was only orientated by five degrees from horizontal and what seems to happen with surfaces close to horizontal is that excess resin gathers around the support points (the small blips) and gets partially exposed, sets and covers the intended surface with a pretty lumpy and uneven surface. I had thought that my five degree orientation would have been enough angle to let the resin roll off but apparently not. You might think that being on the underside might not matter but that's the only side of this part which will be seen by anyone who wants to turn the loco over and have a look. :)

    So I flipped the part over in the slicer and printed again...


    ...but this time the build up was on the top surface and all the slidebars when I thought that the resin might have run off these relatively narrow surfaces.

    So I bit the bullet and tilted the part to fifty degrees which got rid of most of these problems. Why the reluctance? Well the five degree orientation prints were taking about two and a half hours to print; the fifty degree orientation was sticking up a lot more and took almost six hours to print. :(

    While the last print was churning out I started work on the side frames and got the CNC machine fired up to start cutting the 0.9mm nickel silver sheet.


    ... the second side in the process of being cut...


    ... then the end pieces


    ...and the final result before breaking the parts out of the sheet. All the drilled holes are 1mm which will be opened out to 1.4mm as tapping size for 10BA to use as pilots to drill the resin parts. Not the most efficient placement of parts on the sheet, but I had spent a fair bit of time setting up the spoilboard for the sheet and I decided to leave everything as was and live with the inefficiency. :)


    All the parts after cleaning up with the two jointing plates cut from 1/16" brass. All the pilot holes have since been opened out to 1.4mm (10BA tapping) and the next job will be to join the side parts using the jointing plates. I'm going to be interested to see how accurately they go together since I'm relying purely on the accuracy of the CNC machine for the hole positions, then I'll get the frame assembled.

  18. JimG

    JimG Western Thunderer

    A bit more progress.


    I started setting the frames up on the mill's table with a temporary cheesehead screw on the nearer frame member and the toolmaker's clamp on the farthest.


    When both sides were temporarily screwed together, I checked for squareness on two silver steel rods on the table and all was well.


    ...and now the frames right way up. The bottoms of the hornguide slots were all machined level so there's a quick check of squareness when sitting right way up.


    The rear spacer has now been fully fitted to the frames using 10BA countersunk screws, but the front cylinder block is still temporarily fixed since I expect to have to re-design this part for various reasons. I'm also considering putting another spacer between the middle and rear axle. There wouldn't have been one on the real thing since there was a firebox in that gap, but the frames are a bit flexible between the rear axles and another spacer would help matters. I'll have to tie that in with whatever motor and gearbox setup I use, so no work on it at the moment.

    The next development is to 3D print hornguides. I've been wondering if 3D printing resin would be suitable for this purpose and I decided that the only way I would find out was to try it out. The centre guides have been temporarily fitted in place but the outer hornguides will have to be adjusted to clear the joint plates at the rear and the cylinder block at the front. In fact, at the front, one of the adjustments to the cylinder block will be to accommodate the hornguides. I will machine the axleboxes from hard brass and I'll see how they feel when put in the guides. I'm not sure if they will need some form of lubrication or will work well "dry".

    All wheels will be sprung and I'm lookng at using individual springs for each wheel and not a CSB setup. I'm also thinking about underslinging the springs and hiding them behind cosmetic 3D prints of the prototype's springs. I'm aining to incorporate the spring holders in the hornguide prints. If this works out, it means that I can adjust the springing relatively easily by swapping the spring wire gauge. Russ Elliot's formula will get a bit of use again if I can second guess what the weight of the loco might turn out to be. :)

    In all my messing around with getting experience at 3D printing, I've printed the rear spacer three times and the cylinder block six times. :) The last print of the rear spacer was done directly on the bed of the printer with no raft and supports to get a really flat surface, and it worked. Printing on supports does not necessarily produce flat and square surfaces as I've found out. :)

  19. JimG

    JimG Western Thunderer

    Some more progress although there has been a fair but of head banging - principally with Fusion 360 as I find out all its small surprises. I got "circular references" the other day - I haven't seen that since I did programming. :) Rather than try and debug the timeline to find out what was causing it, I started over again. :) That was the fourth time of re-starting the cylinder block. :)


    The frames with a new cylinder block print. It was getting a bit too complex fitting the front hornguides with the slidebars in place so the new cylinder block has lost them. All the hornguides have been re-done to fit in the frames for the outer ones, and to have spring holders included.


    ...and all the parts in place on the frames. I have depleted my stock of 10BA countersunk screws - there are forty-two of them in the frame. :)


    Part of the time spent on Fusion 360 has been to use it to design the parts for placement in an assembly. It was used for the front hornguides and here's the real thing and the same assembly in Fusion 360. I left 0.5mm clearance around parts and that has worked well.


    There will be individual springs on each axlebox using guitar strings. This is a bit of an Ernie Ball 0.018" string and that could be around the size I will use after dabbling around with Russ Elliot's formula and second guessing the possible weight of the locomotive.


    The spring now fitted in the spring holder. The spring is retained in the holder by the "U" bends on the wire at each end.


    The wire is located by the two lugs and the upper end of the wire sits in a slot. So the movement of the spring is not constrained by the holding method. You can just discern a "V" groove this side of the lugs above the wire. A piece of wire will be glued into the groove to provide a harder fulcrum point than the resin. I suspect that the wire would start to work its own slot in the resin after a while. :)


    The rubbish bin so far. :) And there's a chimney and dome in there as well. :) It's probably half and half designing problems and 3D printing problems although there are another one or two bits which found their way directly into the rubbish bin beside the 3D printer.

    The next job will be to machine up some axleboxes and see how the chassis looks with some axles in place. All parts are straight off the CNC mill and the printer, and are tight fits. I haven't done any fitting of parts so I will be intrigued to see how accurate the setup turns out to be.


    Attached Files:

    Last edited: 30 April 2019
    ScottW, JohnFla, TimC and 13 others like this.
  20. adrian

    adrian Flying Squad

    Thanks for posting - it's interesting to see what is possible with these new techniques

    That is one drawback - if it was white metal casting then at least you could remelt and recycle the scrap. :D