Guy Rixon

The early SECR was fond of "Hill's patent Invicta brake" for wagons, but used it in a weird, variant form. Mr. Hill's original patent was specifically for a method whereby the brake could be taken off from either side, and adds a lot of gubbins to achieve this. The SECR version (at least the one for which I've seen the details; there were a few revisions) removes the ability to release from either side. It reduces to just a way to apply one set of blocks on one side from levers on either side. It's logically equivalent to DC or Morton arrangements, but much more complex than either. This makes sense if the SECR had already paid the patent fee for the Invicta brake but wanted to remove the emergent safety-risk.

Pull rods are most common. However, I've recently come across some SER and SECR stock in which the vacuum brakes have a pull-rod to one axle and a push-rod to the other. The brake shaft is offset toward one axle and that one has the push-rod; the latter is quite short.

Hollowing out the axleboxes to allow the bearings to move inside: I do it with a 1.9 mm router bit in a mini-drill ("electric adze") , and the drill spins fairly slowly to reduce the risk of melting the plastic. You need room for about ±0.5 mm movement about the actual position at which the suspension settles when you make the wagon up the desired weight and pack out the axleguards to get the buffer height correct. Since it's tricky to estimate this position to tenths of a millimetre, it's best to hollow out for as much movement as possible. This is easy enough in plastic and no fun at all in whitemetal. Therefore, I avoid replacing the moulded axleboxes with castings. Usually, the axleguards need about 0.5 mm packing to make the buffers right with a weight around 45g. It only works well with waisted bearings, as noted above, and you need to file down the pip on the end of the bearing as far as you dare. It's probably worth making one into a wasted bearing to find out the depth at which the conical cavity breaks through. The waisted bearings reduce the required width of the cavity at its bottom, of course. But you need the rim of the cavity to be opened out to clear the 2.0 mm width of the bearing where it runs in the axleguard. You may need to get bearings from a few different suppliers to find ones in which the cone depth is right for the axleguards. Or get relatively deep ones (e.g. Markits) and shim them to the right depth. You'll need a way to align the axleguards in the chassis. The jig in the Bedford range does the job well; buy some 2.0mm drill bits to use as jig axles if you get this jig. There are also baseplates that help with the alignment. Palatine do etched ones. Or you can find printed ones in my Shapeways shop. But you absolutely don't need the baseplates if you have the jig.

OpenSCAD.

It should be possible to couple the driving axles with a belt. I would guess that is feasible in 4mm scale and very easy in the larger scales. If the belt drive really is easy to set up, it might even help with models of rod-coupled locos. The rods could be made a much looser fit, as they would carry no load, and would not then be likely to cause binding.

Try Worsley Works. They do most of the types for the Trio sets.

Some new products are now in the shop. Brake assemblies for four-wheeled coaches and vans of the SER and SECR. Nice if you want a little more of the rigging than a typical model, or if you simply can't face the soldering of clasp brakes. Might do for other railways if you're flexible about the fine details. One print does two vehicles. Boiler fittings for the "small England" engines of the Ffestiniog Railway. Designed to suit David Eveleigh's kits for these engines, the print equips one engine with either early-pattern or late-pattern fittings. Various products, notably LNWR Emmett buffers and LNWR wagon axleboxes, are now available in S scale. Almost any of the 4mm-scale products can be done in S if there is interest. Things in prospect: LNWR no. 4 axleboxes, as used on most of the fitted wagons in the first 10 years of the 20th century. This will become available as a sprue of 20 boxes with springs and spring shoes when I have checked that I have the axlebox in the correct vertical position to allow springing of the model. Brake assemblies for LNWR vacuum-fitted wagons, similar to those noted above for the SER coaches.

The train of SECR 27' stock moves glacially towards completion. I've been working on the other 3rd-brake. This is newly built from an unstarted, fossil kit and involves significant upgrades to the kit from home printing: buffers, brake rigging, spring hangers, and droplights. The detail is over at the S4 Society forum. If the lining and lettering works out, then this one should be finished some time in the next month or so and I can then move on to refurbish the other, built coaches shown earlier in this thread using what I've learnt from the new build. Having stalled for 25+ years, the glacier is now moving relatively fast. Must be global warming.

Slightly earlier on LNWR fitted wagons. E.g. D95 vans seen with reversible blocks c1906.

If I had that space, I would concentrate on two opportunities: long trains and models of structures that can't readily be split or compressed. To combine the two, in the scope of a one-man project, I'd pick a serious bridge and model it to near-scale, with not too much scenery in front of the bridge, so that viewers could stand close enough to see the trains. It could be a viaduct somewhere out in the bleak moorland, but an urban setting could also be attractive, with the buildings at each end hiding the sharp curves to the return loops. I'd also look into a automated control-system that runs sequences of trains over the bridge, respecting sections. (What I'd really want to do is a proper model of a vast, urban station that would take 40 years to build and stock. But I don't have 40 years.)

E.g. Windrush. A government, with democratic oversight and through several parliaments, deciding that they dislike immigrants and non-citizens so much that they start deporting persons --- including those who have right of residence --- because those persons can't produce the right papers. You talk as though oppression implies an absence of democracy.

The problems with the RFID element are, AFAIK, unique to the British scheme of a few years back. Take off the information storage, the remote readability and the easy trackability, and the cards become more attractive, at least to me. The issue with id-carrying turning mandatory has a specifically British aspect. We don't have a legal arrangement in parliament whereby a right of citizens can be made constitutional and incapable of being revoked by a subsequent, normal vote. C.f. the constitution of the USA in which such rights have special protection from rogue governments. IIUC, some rights "anciently" have acquired this status, by way of expensive test-cases in higher courts, but it happens rarely and it's really hard to add a new right that way. Effectively, if governments decide to use the ID cards for oppression we're stuck with it.

I know four practical objections to British ID cards. The first is that it might become mandatory to produce them when challenged, giving police and other authorities a whole new way of hassling people they don't like. The second is that, per the last pilot scheme, the cards will encode information about the carrier that should remain private but will almost certainly be insecure and will be trivially stolen via the RFID mechanism. The third is that the RFID mechanism will be used to track movements of citizens, which the government will record and then leak through bad security. This gives problems of targeted crime. The fourth is that cards will be stolen and used for identity theft; and that too many will accept the stolen cards at face value, not even checking the photo. Objection 1 needs guarantees that our constitution can't provide. Objection 4 needs properly thought-out laws applied with common sense (not holding my breath). The others need the scheme details to be designed for the benefit of the people, not the government (oink-flap).

G W Models do "nut spinners", which are a posh kind of box spanner, in small BA sizes. One can drop the nut into the spinner and use the tool to carry it onto the stud. For holding machine screws, jewellers' three-pronged stone-tweezers can be useful. They don't work very well for holding nuts.

Polystyrene is a good insulator for low voltages (it insulates high voltages too, but is too mechanically fragile to be reliable). Whether it's safe thermally around a motor is unclear. It actually melts at ~ 240 Celsius (according to Wikipedia), but as we modellers know, it softens below 100 Celsius. I would worry about the motor-cladding pieces warping and flowing out of shape over months and years. If it were easy to replace damaged boiler-pieces, as it seems to be here, then I think i would go ahead and fix them as and when they did distort. However, if they were hard to replace then I would change them to metal from the start.

York Modelmaking do a range of valancing and canopy brackets made by laser cutting. Scale Link fretcetera sell some etches of valances (search their site on "valance"; the relevant products are hard to find by browsing). To make valances at home, a diecutting machine seems suitable.

Joking aside, they are of course the regulators to go with roof ventilators. One can see the operating handles for the hit-and-miss slides, to limit the flow of air. These things were the subject of a patent at one point (Laycock's?) and it would be of passing interest to know if the GWR are manufacturing under license, or if they're evading the patent, or if had lapsed at the time of this photo. I would guess that the photo was taken because the regulators were a new design.

They look to me like spun metal, enamelled. Bakerlite is a resin, so probably clear or milky coloured in its natural state. The black colour in bakelite 'phones and such would be a dye.

Chitubox is very useful software for operating home printers, and the free edition is very capable, but you get the UI/UX refinement you pay for. Some necessary operations are hidden, not documented and can't easily be discovered because they use atypical UI. Here's some notes to ease a beginner's frustrations. The Mac port of Chitubox seems to work, but the UI hasn't been converted properly to Mac conventions. The Chitubox menu in the menu bar doesn't have the usual new/save/save as... commands. These appear instead on a menu linked to a button top-left in the Chitubox window: the one marked with three horizontal lines. To prepare a new model for printing, start a new project with the menu command: this gets you an empty build-space. Then use the Open... menu-command (not Open project..., which is different) to import the .stl file for the model into the build space. Then you can position it and add the supports. When slicing for a Mars printer, there are two kinds output files: .chitubox and .ctb. The .ctb file is the sliced model that you transfer to the printer. You get this file via the Save button that appears at the end of a slicing operation. The .chitubox file records the whole project, including the model, its orientation and the support pattern. You get this one using the Save project... command from the main menu. The manual I was given says to transfer the .chitubox file to the printer, which is wrong for a Mars printer. Other brands of printer may consume files of different kinds. To refine the printing of an old, saved model, use the Open project... command from the menu and load the .chitubox file for the project. However, if you just want to print it again with the same settings and supports, just keep and re-use the .ctb file. You don't need to re-run Chitubox for a verbatim reprint. To place supports manually, you need to zoom, rotate and pan the view so that you see where to put the support tip. These view-change movements are all mouse/trackpad actions: right click and drag to rotate; left click and drag to pan; zoom with the moose scroll-wheel, if you have one (I think; I have a trackpad, so can't test the mouse thing); zoom with two fingers on a trackpad; move up to zoom out and move down to zoom in. Changing the view is a separate matter from moving/tilting the model in the build space. The latter process uses the buttons on the main screen. You'll often need to remove supports manually. The UI looks like supports should be removed by selecting the "remove" mode in the GUI and then clicking on each doomed support. Not so! In this mode, clicking on a support only selects it (and turns it red in the rendering). To actually remove it, select it then press the DEL key. Only one support at a time can be selected in this mode. Where's the DEL key on a Mac keyboard? There isn't one. Use Fn+Backspace to get the DEL operation. (I.e. hold down the fn key and then press the backspace key, the one with the backward-facing arrow.) Yes, DEL and Backspace are different; thank Microsoft and IBM for that. Often, one wants to print multiples of the same model in one print-run. This can be done by opening the .stl file of the model multiple times, but that's slow and seems to confuse Chitubox. There's a "clone current model" button in the main screen that does this better. Rotating the model in the build space may be buggy. There's a button in the main screen to turn on rotation mode, and you can either drag the model around each axis with the mouse or type in the angle of rotation. If you type in the angle, Chitubox sometime applies the angle relative to the bottom of the build space and sometimes relative to whatever angle the model is already at. This seems to be some subtle bug. However, it always seems to print to the angle at which it's drawn, so you can tell what's going to come out. Opening project (.chitubox) files also seems to be buggy. If a project is already loaded when another is opened, Chitubox sometimes tries to add the model of the newer project to the older one. Starting a new project first, to clear the print space, seems to fix this.

Inches are fine as a tool for modelling railway vehicles specified in imperial measure, but feet less so; there are vastly more dimensions given in inches than in feet on a wagon GA. Mixed feet and inches is unhelpful when computers are involved. I'm currently finishing CAD for a wagon chassis and my OpenSCAD code is full of macros like middle_bearer_thickness=inches(4.5) and solebar_height=inches(10.67). When it comes to the longer lengths I put wheelbase=inches((9*12) + 9) and similar.

Possibly, but then the buffers will just break again. Metal is needed for the thinnest bits. This is why I don't normally print buffer heads and rams: they would be too fragile.

OK, those buffers; one loses track. Yes, Miss Prism's drawing should do nicely.

If you have decent drawings (or dimensioned sketches) and some photos, I can arrange the prints. They would be the buffer guides and mounts, to take turned-metal heads and rams.

There's also the issue of returning baskets that originally arrived by road, but these would have come from nearer London and perhaps not from cattle country.

Fruit baskets returning in cattle wagons is known from a period photo reproduced in Southern Wagons vol. 3. Doing it this way makes sense when originating stations need to get the baskets back to the consignors when the fruit vans are needed somewhere else. Bear in mind that "fruit vans" on the SECR means PLVs or PBVs or Grande Vitesse vans, all of which are in high demand.