brianCAD

Will they fit decent looking wheels to the Hunslet? This horrid thin tyred rendition would be condemned in real life. Some spoke fillets (especially where they meet the counterweight) need to be split radially in the design software, and a smaller radius used on the lesser angled side to get the fillets evened up.

If only it was that simple. The Chinese are very able and can turn out a first rate product (e.g. Brawa) when all goes well, but the operation of their business model is something else and isn't as strait-laced as in the UK. Speak with anyone who manufactures in China. After experiencing varying success with model manufacture in two western countries outside of UK (operator ran a machine too fast causing issues, elsewhere another employee fitted an incorrectly radiused carbide tool tip when turning thousands of wheels), my hand is poised nervously over the 'go' button for a first venture in China. I do hope that Keith Revell is able to maintain some oversight as the manufacturing progresses, and is not relying merely on placing an order, and waiting for the models to turn up. I'm mindful that Canadian, Christopher Howard operating as 'RailFlyer', lost his shirt by not employing the right attitude, unlike his successful countryman, Jason Schron. Sorry, it certainly wasn't intended that way - and your subsequent retort was uncalled for. I am genuinely concerned with some of the views expressed on this forum regarding mechanisms, being partly my stock in trade.

Chrisp p bacon, sorry old chap but you seem a bit confused about the characteristics, merits or otherwise (mostly otherwise as it happens) of the mechanisms that were/are being considered for the GT3. We can agree on a preference for worm and worm wheel drive direct to one axle, which arrangement served us pretty well for decades. Meanwhile, there is a clear difference between the practicalities of the gear arrangement in Rapido's 'Royal Hudson' and that originally proposed for the GT3. There is no need (and undesirable) for the Hudson's side rods to be loose fitting (assuming the gear teeth are properly aligned to the wheel quartering), whereas it is more problematical for the GT3 with gear backlash having different amounts at 1st and 3rd axles, thus making it harder for the side rods to function reliably. Loose rods would not improve this scenario. With regard to the simplified GT3 proposal, having only one geared axle, it doesn't matter a jot how many gears, or backlash are present - as it doesn't impact operational performance. Whether or not any further inefficiency might creep in through the use of extra gears, would need to be judged against the inefficiency of any alternative worm gear drive. Case in point: Bachmann's sometimes use of a large diameter worm leads to greater transmission losses than for a small diameter worm possessing a faster pitch. This is all fairly basic stuff to an engineer, and it's a bit disappointing that a Chinese design office has wandered off down this garden path, while not treading very carefully.

Agree. But isn't this where KR Models are now going? I haven't looked at the Facebook page, but an image of a revised drive has appeared on this thread.

On spur gear mechanisms, ideally there should be the same number of gears in the gear train leading to each axle, so that backlash is similar at each axle. This doesn't matter for a Co-Co diesel where the wheels are not connected with side rods, but for a steam loco with all axles geared, the addition of side rods presents a double whammy. Not only is accurate quartering required, but this also needs to be synchronised with the alignment of teeth in the gear wheels. The European manufacturers construct very elaborate jigs and fixtures for the assembly of their models, but the Chinese seem to rely on nimble fingers from what I have seen. The drive arrangement for the Rapido 'Royal Hudson' is better than that originally proposed for the GT3, as the Hudson's outer driving wheel axles are powered similarly, enough to keep rigid side rods on the level. Whereas, for the GT3, the leading axle was freer to do its own thing, suffering excessive backlash compared to the second and third axles, and the side rods would detract from a smooth performance. The revised GT3 drive with gears powering only one axle is a much better scheme, and there shouldn't be any loss of traction as the side rods will function in the same manner as the real thing.

The mechanism appears to be the same design as used by for Heljan (O scale A3 and assorted diesels), so experience there, good or bad (split gears), should provide some pointers. I would prefer to see an old fashioned X04 or can motor powering one axle directly.

And that may happen. But DJM's custom is quite insignificant when viewed against the volume of model railway equipment the Chinese are catering for globally. Eventually, failing a good offer, they may consider past production has sated the market - and be pleased to move on. There is still good value in the many parts associated with a mould set, even after the expensive cavity work is removed, and it would ease the need for storage, especially where there are multiple tools for a single model. The sudden closure of Affa Technology last year caused quite a re-jigging in the supply chain, with plenty of opportunity for manufacturers to replace an errant customer (until such time as higher US tariffs kick in perhaps? ).

Yes, I'd fully expect that the Chinese businesses involved will show that they are also creditors, or at least successfully construe to do so. I don't foresee any return from there to the people who funded Dave's lifestyle. Furthermore, unless the Chinese can make a quick sale of existing tooling, the cavity plates within the mould sets will be stripped out, and the sets modified to accept new cavity plates for perhaps a European, American or Australian prototype. Pandering to the whims of a meddling enthusiast has exacted a toll on all parties - as it also did for Dave's former employer.

Mike (Stationmaster), I'm puzzled by your constant references to .STL files - and that they have value (very limited in my view). STL files are a byproduct format ex CAD, used primarily for 3D printing processes. The files are no more than a mesh of triangles, which if not saved at a high resolution can significantly degrade the information contained in the vastly more valuable originating CAD file, comprised of vector information. Perhaps they could be likened to a pixelated reproduction of a line drawing (which they are when the triangles are shaded) ? While an STL file can transfer data as a "dumb solid" (and might be used where a designer was required to conceal how a design was arrived at), the vector data of CAD files is the real McCoy for the generation of G code CAM data for any CNC work, and that is where the value lies.

Have tried some of the foregoing methods, but now will not do anything other than turn and fit separate tyres. Tyre insulation has mostly been thin rings turned from PVC (more easily turned to exacting dimensions than Delrin), and secured with super glue. Two locos have ordinary notepaper for rim insulation (0.004" or 0.1mm thickness), soaked in Araldite , with the tyres wrung on. Another set used a variation of Michael's method above, but by cutting around the thickness of the rim instead of the spokes, filling one half length at a time with warmed up Araldite - to make it run and fill the groove fully. These wheels were from my own nickle-silver lost wax castings, which had a small slot cast in the rim for starting the piercing saw blade. (Slaters wheels have the plastic centres moulded into the rims, which are grooved internally - so don't see how those tyres could come loose.)

And in an earlier life, Dave devised a Manor to be much much bigger. . . so its 'power' was fit for a King. Then people revolted and it cost a King's ransom to get the Manor back to its rightful space before any beheading. https://www.modelrailforum.com/forums/index.php?showtopic=6723

With the mess as it is on UK shores, how could we imagine it might be any better in China? Surely, no manufacturer is going to be interested in producing anything now for DJM unless CAD design can be approved in good time (if based correctly on a comprehensively prepared dossier of information). That Dave, apparently is reliant on others for CAD skills makes me think he is ill-equipped for managing these projects. Only after the completion of the CAD stage can the manufacturer get started generating full income from the efforts of his factory workers - assuming the requisite hefty step up in payments is made. Somewhere in this area is where Dave seems to have difficulty in progressing - and I would be asking why if a crowdfunder. In a practical sense (given the geographical areas involved) the registration of designs is nonsensical and not worth discussing. OK, so there is argy bargy going on with some existing tooling. For a reason I keep envisaging a pot calling the kettle black. -Brian M.

What this visiting colonial has done previously is; take the 233 bus from Sidcup railway station directly into Hilda May Ave (for the Leisure Centre). This allows for visiting Invicta Model Rail to the north side of Sidcup station and for buying that new lathe (yes) at Home and Workshop Machinery in Maidstone Rd near the the Ruxley roundabout. -Brian

Thanks Dave. Your parts are indeed small at sub fingernail size. I sometimes find myself defending Shapeways when parts have been produced from elementary CAD or minimally sized STL files. No chance of that from you, using CATIA. Can't be many modellers so skilled or equipped with that package. The blends in your file of the three-legged component are very nicely executed. A model engineer local to me, uses the same machine as Shapeways for waxes. Most are made at the 25 micron setting (for 5inch gauge) and the castings are sold globally. Being a fussy bu99er, I prefer the finer layer thickness setting of 16 micron, albeit it being more expensive. I'm not sure Shapeways can always be relied upon to use the finer setting. For a time (when it was available), I'd just buy the waxes and get another party to do the casting - for a better result. Output from in-expensive DLP type printers is now proving superior to the processes described above - and is worth checking out for that 'special little bit'. ATB - Brian

Dave, What's the physical size of your components ex Shapeways? Unless the parts are very small, I'd venture your STL files were a bit light on number of polygons. I'm a bit disturbed by the very coarse finish evident. Shapeways services my part of the world from New York, so possibly results from there are different to European production? I crank up the polygon count so that STL file size approaches Shapeways maximum of 64mb - and leave it to their software to crunch them down to whatever is needed for printing waxes at 16 micron layer thickness. Sometimes I wonder if they have used a coarser machine setting to speed production. . . . Whoever converts the waxes into brass castings (contracted out apparently), insists on rumbling the castings - so any desired sharp edges can be lost - and certainly so if the option of 'Polished Brass' is chosen. Adding a small machining or filing allowance to any flat face that can be accessed readily later, is one way of preserving crispness. -Brian

I'm not familiar with the prototype, but a few small details that the talented toolmakers might be able to tinker with: Remove rounded top edge to driving wheel splashers? Expansion Link looks a bit 2-dimensional. Heavy rims to moulded portion of front bogie wheels and perfunctory appearance of spokes within. Has there been any change to the drive mechanism since the initial CAD presentation? -Brian

Håvard, In a nutshell, use HSS for brass and for any fine finishing work in steel. > I am not entirely happy with the finish of the inside bore. > I used a boring tool with carbide inserts ( 8 dollars on eBay...) Reserve your (positive rake) carbide insert tooling for heavier cuts in steel. Ignore negative rake tools which require more power and rigidity than hobby sized machines possess. HSS takes on a sharper edge than most carbide inserts provide - although a carbide tip can be reground to resemble the geometry of a HSS tool, its edge might soon crumble microscopically. The pointed tip of a HSS tool should not be left sharp. The vertical edge, where the angled side and end faces meet, will benefit by having the tiniest radius applied. A few wipes of an oil stone suffices - but take great care to maintain a positive rake cutting edge, and not round off any of the tool's horizontal edges. Until about two years ago, I was using Shapeways to provide large numbers of assorted driving wheel centres as brass castings. These came from their New York base and my experience from there differs to yours. The castings were always a very close match to the CAD file size, varying only by two or three thou in diameter on wheels typically of about 1inch in diameter. On one occasion the castings were one thou larger than the CAD, so I assumed any scaling necessary for the casting process was accommodated by one of their automated processes. The waxes were made with the wheels standing upright (presumably to maximise part count on the machine platen). This meant that the spokes close to vertical alignment showed no 'stepping', but those spokes extending out close to horizontal alignment had ever increasing evidence of 'fishtailed' stepping as they became nearer to horizontal. On counting up these visible steps, the resolution sometimes seemed coarser than their claimed 16 micron layer thickness. To avoid any cheating (if indeed it was occurring), I resorted to purchasing only the waxes (which I could inspect first) and then had them cast by other parties for a better result. Shapeways (NY) always insisted on tumbling their brass castings which rounded off any desirable sharp edges (but not actually a problem for wheel castings with machining allowances built in). One batch arrived 'polished' which damaged counterweights, but was quickly replaced by Shapeways. Cheers, brianCAD

I'd be very wary of using that chuck. Check to see if it runs true and is not out-of-balance at full motor revs. Check to see if cutters run true when gripped in the chuck. By now the business end of the cutter is a long way below the support of the motor/spindle bearings and the whole unit might vibrate horribly when attempting to cut anything no matter how soft. Suggest seeking out a skilled lathe operator who can make short 'adaptors' to mount cutters of assorted diameters directly to the motor shaft - similar to that supplied for 1/8" dia tools. I've only just come across this thread, and would have preferred at the outset that you made some use of your TH pantograph machine first, to gain machining experience helpful for evaluating CNC possibilities. A lingering concern is that while 'CNC' control systems are well sorted and robust today, the same can't be said for some inexpensive hobby machines leveraging off the technology. -Brian M