Downton Train Crash of 1884 in EM Gauge

[Stevers]

A cheap and cheerful smartphone photo of the Lion Class chassis etch. The mistake that I immediately noticed is that at 1.2mm the holes for the knuckles to articulate the coupling rods are larger than any commercially available 4mm rivet - what was the idiot thinking who drew that? In addition since the knuckle ends were half etched, slightly more material had been removed than was allowed for, leaving perilously little material remaining around the hole to work with.


A shot of progress to date. The large knuckle rivets (there was no going back) were turned up in my cordless drill using files, from a random suitably sized bit of non-magnetic silvery rod that was harder than nickel silver. It could well have been stainless steel and happily it riveted OK. The leaf spring layers were lined up using a couple of lengths of 0.35mm phosphor bronze wire. The leafs of the springs are finely engraved, but still overscale. There are three layers for the leaf springs sandwiching a reinforcing wire for the rods and an extra layer for what I'll call the 'buckle' - similar in appearance to the springs I scratchbuilt for my Wills/Persy M7 thirty years ago, but much easier! It's a shame they'll also be out of sight behind wheels. The springing arrangements for the rear axles on this and the Vesuvius Class are tucked up out of sight on the prototypes, with some drawings hinting at a form of coil springing.

The firebox grate was aligned using the hole for the brake rigging and then by eye, but it might be better to provide more positive alignment for that and the hopper as per the leaf springs. All completed with lead multicore solder and phosphoric flux. Here's hoping that when I fit the frame spacers it won't all fall apart. With that hurdle safely over I'll consider switching to 145C solder to fit the hornblocks, the rear ones of which I'll need to trim to clear the grate. The HLK axleboxes are all made up ready for the next stage. I was short of one 1/8" axlebox, but was able to open out a 2mm axlebox to the right size with tapered broaches, finishing with a reamer. On reflection this is how I should have opened out the 2mm MiniBlox to 3/32" for the tender.

 

[Stevers]

The frame spacers are in with nothing falling off, and the coupling rods have been opened up to fit my GW Models axle alignment jigs. Given that there's not a lot of meat on those connecting rods, I standardised the inner steps on the jig at 1.46mm which appears to be about the typical size for the AGW crankpin bushes - previously they were a little larger than that on one side. The idiot who designed the etch put the wrong spacers in the etch for the body mounting at each end. He thought he'd specified standard Persy spacers throughout when the ones at each end were definitely not standard and the poor old builder had to improvise with parts of some spare Persy brass ones. I'll pad the axlebox at mid-point in the guides before soldering the guides in. Another mistake spotted in the etch that won't prevent a running chassis, is that when I changed to having twin brake rods both sides in the 3D render, I didn't double the number of rods on the etch! One of my GWM springs pinged off into Narnia, so until its safe return, I had to wind up another from 0.4mm piano wire on a 4mm(?) drill bit in my old Footprint hand drill.

 

[simond]

“Footprint”, there’s a name from the past. I have a pipe wrench with that trademark.

 

[Stevers]

It was one of the first tools that my late father bought me, possibly nicer than his own 'Stanley' hand drill. Still working faultlessly, but now relegated to making springs and drilling holes tight up against things. I went to Uni in Sheffield and have lots of other Sheffield made tools including a pair of 'Elliot Lucas' needle pliers and chromed 'Footprint' gas pliers (always referred to as the 'Footprints') that I bought as a student that both still get a lot of use. I have a 'Paramo' vice that is rather good, as well as three 'Record' vices - my first one being a poorly machined Record 'Imp' - another unusual childhood purchase.
[Edit] I used the 'Footprint' gas pliers on some plumbing today only to be reminded that they were made by Elliot Lucas of Cannock!

 

[JimG]

“Footprint”, there’s a name from the past. I have a pipe wrench with that trademark.

Like this?


Came from my late father's toolbox and I suspect they came from his father's since he was a plumber. Still in occasional use by me and very effective as a pipe wrench.

Jim.

 

[Herb Garden]

Like this?

View attachment 241260
Came from my late father's toolbox and I suspect they came from his father's since he was a plumber. Still in occasional use by me and very effective as a pipe wrench.

Jim.

That's almost identical to the one I made during my secondment to technical college.... It's a design that teaches you all the nesscecary workshop skills

 

[simond]

Like this?

View attachment 241260
Came from my late father's toolbox and I suspect they came from his father's since he was a plumber. Still in occasional use by me and very effective as a pipe wrench.

Jim.

exactly like that, but without the blue paint

 

[Stevers]

It's been a while with gardening and DIY getting in the way of more important things, but these have just turned up and are showing a lot of promise:

The first couple of tasks will be to make up one of the etched smoke box fronts to see if the push through rivets work and if it all fits, and then to see if the MayGib style sprung buffers fit the 3D printed housings. If they do, the tender will get them too.

 

[AJC]

Did the box have 'Airfix' or similar written on it? That all looks rather good. I especially like the non conductive brakeshoes and the double decker firebox doors (not seen those before).

Adam

 

[Stevers]

Mr MM does indeed make exceedingly nice prints - the box was a brown Airfecks one - that might turn blue!

I now have my own Anycubic Photon Mono 4k (little used yet excellent value), and will be able to really appreciate just how good Richard is at this stuff. I've settled on Anycubic's Ultra Tough resin for my first attempts with the Stevens signal levers (and brakeshoes) in mind and will wash in BioEthanol rather than IPA. I'm feeling the need to build a glazed cabinet/fume cupboard vented to the outside (on the other side of the room). As with everything I tackle, if it isn't mind bogglingly difficult - I'm not trying hard enough.

 

[AdeMoore]

Looking good Steve, very interested in how this one goes together.
The print looks very crisp.

 

[Stevers]

This is my 'spare' smokebox front as there is also a nickel silver one. I did have to shave a teeny bit from the etch round the outside of the smokebox, but on the plus side the smoke box door was an interference fit on a rebate invisible to my naked eye. Even with the benefit of my GW Models rivetter it was impossible to get the push through rivets located correctly on such small landing areas. On the plus side handrail brackets and lamp irons are on securely with appropriate rivet detail! Version 2.0 of the boiler will have a little more room in front of the rivets and a hole to help locate whichever chimney is chosen. Can't help feeling there ought to be a second small round cover between the cylinders, though that would be out of sight behind the buffer beam.

As for the functional buffers in the 3D printed buffer shanks. I borrowed the KM buffers that were intended for my Q, but these are the bent wire sort rather than the Slaters threaded bar and nut type that I had in mind. The cylinders at 3/32" internal were too tight in the 3D printed shanks, and the resin too soft to easily drill out for a sliding fit with a 3/32" drill, 3D printed resin also lacks lubricity so there were multiple problems in adapting the KM buffers so I'll source a set of the AG ones with the brass collar and redesign the buffer housing to exactly fit those.

 

[Stevers]

A change of tack (currently reading Hornblower) to look at a solution for springing those troublesome trucks. The Masokits units should have fitted, but it is hard to determine which type(s) would be appropriate and any 'W Iron' outline would be a compromise perhaps needing bits cut off. The Bill Bedford sprung units still don't seem to be available (the S4 Society having the license to produce them), but I was concerned (quite rightly as it turned out) that they wouldn't fit anyway and of course the outline would also be wrong.

Having had the experience of designing the etches for my Lion Class, it seemed logical that I should try my hand at designing an etched suspension unit that would provide some much needed comfort to my resin and whitemetal passengers on their fateful journey. I quickly determined that a 26mm axle would mean that there would not be room for cosmetic Irons - everything being more or less in the right place for a narrow prototype. However a unit taking a 25mm axle would fit between thinned cosmetic Irons and give me something to attach the stretcher bars to. A waisted bearing was measured and it was determined that the bottom of the V was 1mm below the face - I soldered a wire into some scrap etch to measure this, only to discover that this was probably a standard that might have come from the EMGS. The flange was 0.2mm and the waist 1.86mm. I'm happy with the 0.4mm sheet material that I've used so far, so started to draw up the most basic of suspension units that would do the job and fit my coaches.


Wheel clearance is provided for 00, EM and P4 wheelsets and 0.5mm travel is provided in both directions. The low stiffening folds just clear the Newell brake rack and cylinder unit that optionally goes in over the sprung units. I'm reluctant to go down to the readily available Lima axle length of 24.5mm although I have my EM Lima HST running on Lima length axles without issue - I do wonder how well that would have gone for P4. Dapol use a 25mm length axle so I'm currently looking into sourcing some of those, or commissioning a batch from Alan Gibson.


The units placed in the 1859 Brake Van - a type of van that shouldn't have been in our fateful train, thus making my life more complicated. The Newell brake bars (where fitted) will need to be attached to the cosmetic axleguard irons, and I still need to put a cut out in the springing unit to cater for that together with a hole to mount the stretcher bars.

 

[Stevers]

Just for background information, this is a low resolution copy of the detailed plan associated with the BoT Accident report held in the NRM as photographed by a helpful friend and stitched together (with some difficulty) by yours truly. It shows the waterlogged ditch that featured heavily in the report and the first class coach that came off its underframe. One of the coaches was left 'hanging' in the tree that is visible. The initial derailment was probably between the two first class coaches. Most of the deaths and severe injuries seem to have been in the third class coach in the water and the severely damaged first class coach next to it. It is interesting to speculate how accurately the hand tinting reflects the livery of the train. The artist has shown the loco domes as being polished brass. But, if the locos were green (presumably 'Goods Green') as appears to be indicated, then I would have expected them to be painted. The coaches appear to be in a brown and cream, and from reference to the HMRS LSWR/SR Livery Register, this is pretty much what I would have expected with the actual colours being darker than shown giving latitude for the cream to be 'redder' rather than 'yellower' and so perhaps on the way to being the salmon colour that the LSWR became famous for later.

 

[Stevers]

Courtesy of a very helpful friend, one of our block sets at Salisbury said to be in 1882. Very similar to how our Weymouth train should have been formed just two years later. Looks like the 1864 style Pass. Brake Vans at each end, and two 1862 style First Class coaches in the middle. Going by the number of oil lamps we have two 1865 style Third Class coaches this end. Down the far end there should be an 1862 style Second Class and by rights an 1865 Third, but it's hard to reconcile the width of coaches or number of oil lamps visible to there being anything but third class down there. What is clear is that it's close coupled in two sets of three coaches with long couplings between the first class coaches. The passenger coaches look surprisingly uniform considering the two very distinct styles employed. The Newell brake control rod is just visible on the back of the van, but I can't see any evidence of handrails on roofs and therefore end steps being fitted. I have added the lamp brackets to my brake vans, but am reminded that I need to make holes for some safety chain eyes that I'll include in the suspension etch, and think about how these coaches can be close coupled yet still sprung buffered. Too close and there'll be derailments, and that would never do.



Have been going through the coaches getting them ready for 3D printing. Still hoping for a better class of render, but in the meantime.I've added a top lip to the sides aligned with the roof intended to hold the sides straight and slotted out for the slighty shortened roof ribs Comfy seats in first, but no intermediate armrests are shown in drawings. Also shown are the glazing pockets intended for glass but plastiglaze or similar will fit.

A side view showing the missing safety chain eyes and that previously missing stage coach like bottom curve on the tumblehome. Having contemplated the sort of trig. employed on the domes, that is just a flat curve with a slight lean to align both ends! To print everything straight and true I shall continuously support those footsteps (and W Irons) on knife edge supports that will only be removed once everything is fully cured.

 

[AJC]

Interesting vehicles - and impressive renders - a couple of thoughts:

1. I'll be interested to see how the footboards hold up to handling - I think I'd etch them, given the choice (I hate having to mend things once they're finished), but that's what test prints are for. That said, these are tiny, you might be ok.
2. Close coupling - what did the prototype do? What the NLR did (and what London Road Models replicate in their kits for those vehicles, and yes, I've built one), was to have concave cups at one end of a coach and very short buffers at the other. This ought to work, but I guess depends on the amount of play in the coupling.

Adam

 

[Stevers]

1. I'll be interested to see how the footboards hold up to handling - I think I'd etch them, given the choice (I hate having to mend things once they're finished), but that's what test prints are for. That said, these are tiny, you might be ok.
2. Close coupling - what did the prototype do? What the NLR did (and what London Road Models replicate in their kits for those vehicles, and yes, I've built one), was to have concave cups at one end of a coach and very short buffers at the other. This ought to work, but I guess depends on the amount of play in the coupling.

Good points Adam!

The footboards make me a little nervous, and the oversized yet still delicate brackets bother me too. Even if they don't ping off, will they stay straight? However that's what test prints are for - 3D printing them is the easiest option if we can get away with it. There are tougher (as in bendier) resins than the ABS style ones that Richard has been using so far. I'm going to experiment with the AnyCubic Tough Ultra resin in my own 3D printer (mainly with my Stevens lever frames in mind), but as this is real life there must be a downside to those, and it could be dimensional stabilty or the ability to carry detail that suffers. Our preference for metal to be added to 3D prints wherever possible, isn't shared by many modellers who seem to want everything 3D printed right down to the solid buffers - a conversation I've had with Richard in relation to doing a wholly 3D printed Lion Class body from my etch based renders - presumably to stick it on a wagon as a curiosity, rather than on a ready-to-run chassis as a working loco. In contrast, one prominent modeller I know isn't keen on any part of a loco being 3D printed - you can't please all of the people all of the time!

If etched, the brackets could all fold up from a part of the etch that held them at the right distance and straight until the delicately notched separate foot boards could be soldered on - that would be a way forward. They would look lovely in nickel silver, and the soldered assembly would just need to be separated and plugged in. If the bracket spacings vary between coaches and I'm sure they do, that's probably a slightly different etch for each coach, but as we already have two wheelbases and at least three lengths it doesn't add much complexity - you can tell I'm tempted...

Gordon Weddell's drawings show two styles of close coupling, in each case I'm assuming side buffers and central coupling. The first being very short buffers both ends, and the second a longer buffer on one coach and a pad on the other. To me the latter seems much easier to implement when adapting the prototype and on a model. With close coupling, those Victorian engineers must have been torn between trying to add stability by restraining the coaches at the buffers, whilst retaining sufficient flexibility to safely navigate humps, hollows, and curves. This was a compromise that they might well have got wrong at Downton. For my coaches that could be a 3D printed pad that plugs into the hole intended for the buffers when close coupling with probably a single link semi-permanent coupling - experiments required...

 

[Stevers]

The AG LSWR Drummond pattern buffers have turned up and the heads look about right, although there's no centre hole in them yet. For the loco buffer beam I drilled through the housing right to the back of the beam, in the process removing just enough material for the buffers to move freely. I then countersunk the back to take the cuff. Due to how far the cuff is from the head, I had to double spring them from my stock of even tinier MayGib style buffer springs. The wires are only partially bent so that I can recover them to fit to the final version of the buffer beam. If you look closely, in trying to straighten it, I exceeded the limited amount of flex in the area of the coupling hole, and had to roughly (in close up) epoxy it back together. On the back of the buffer beam can be seen the poorly formed lamp irons, and angles that on the real thing join the buffer to the frames. In this case it's the slots that take the etched cosmetic front frames that are soldered to the running plate. The beam will be better printed in two halves with both detailed sides facing away from the plate. This gives an opportuny to increase the compression on the buffer springs via the brass cuff being moved in, or via a fully 3D printed enclosure as per the tender.

A problem with outside frame tenders seems to be that the buffers are almost in line with the frames. My solution was to disgard the brass cuff and hollow out the inside of the frame. I don't fancy trying to bend the wire from in there, so I'll use a short length of brass microtube to retain the buffers and set the travel. Having now thought of that, I'll do the same on the loco for the final fit. Also visible are the formers for the safety chain holes, and the external steam brake cylinder fitted to some of the these tenders.


Given that metal sprung buffers have been shown to work in 3D printed housings, I added the buffer housings to my Beattie tender together with holes for the safety chain eyes. I did open out the inside of the housing - although in theory it was already too large, and slightly reduced the ring round the shank. This tightness of 3D printed holes (and oversizing of cylinders) is due to 'additive kerf', and it seems another reason for this is over exposure during printing. Due to light bleed, it's not really individual pixels (or layers) that are being set, and that's presumably why the layer height can a bit more than the pixel size without losing detail. Some of the 3D print exposure test models come with parts that you break off that should fit inside other parts of the print if correctly exposed. If your exposure test piece has a Dungeon and Dragons origin (as mine does) it will involve the manipulation of resin swords, rocks and foaming jugs of ale, and they probably think we're weird...

Although the tender print was excellent overall, I wished to address the wonky W Iron straps and missing step. I knew that better support for the model was required during printing, but could see that there is a limit to what point supports could do. It was convenient that I came across a thread in that other place where 'edge' supports had been used to address this problem. Sadly Google won't take me back there, so that I can credit the author* here, and it seems such supports are often employed on 3D printed functional engineered parts. 0.2mm seems to the optimum thickness for the tip, and the rest I made up. For this application I have created a generic capability to support any model by creating a configurable grid of horizontal and vertical members on which I can plant my edges in appropriate places. In this case many of them are offset to support printing the tender brake cylinder side tilted up, whereas before the optional cylinder was added, it was clearly tilted the other way. The model (and supports) will need to be further supported, but the edge supports should keep everything straight through printing and curing. The plan would be to fret saw them off from the front until the support tucked up under the buffer beam can be gently flexed off.

[Edit] *It was Paul Sterling and his project 65033 (which I'll 'borrow' another idea from):

https://www.rmweb.co.uk/forums/topic/164432-project-65033-a-3d-printed-lner-j21/?do=findComment&comment=4438911

 

[Stevers]

At the start of February I finally fired up my secondhand Anycubic Photon Mono 4K. After practising on my Class 156 interior and lever and ground frames I thought I'd have a go at something more challenging. With the Beattie 1950g Tender reoriented at 20 degrees rear end up with the knife edge supports changed to suit, my tiny little printer whirred into action and quietly produced this:
Anycubic's Water Washable ABS-Like V3.0 resin was my poison of choice and I see no reason to try anything else. I had been using Lychee Slicer for supports, but that seemed quite random in how supports were added. in a thread in that 'other place' I came across a picture of a coach with some very organised looking supports added via 'Blueprint'.


My 1859 brake was therefore tipped up at 20degrees and supported in Blueprint, but sliced in Anycubic Photon Workshop since Lychee Slicer reckoned there were over 17,000 'holes' in the Blueprint supports and couldn't repair them.


Well, not the disaster it could have been (and was half expecting), but a number of lessons learned and revisions in hand.
Pros:
- It's undoubtedly an 1859 style LSWR Passenger Brake van with body and underframe printed in one piece.
- On top the brake wheel, platforms, steps, desk, sorting rack, glazing pockets (now all glazed) came out perfectly.
- Underneath the Newall brake shaft could be installed and all the bits of the complicated brake mechanism were created correctly.
Cons:
- The panels between the frames (and sides overall) are clearly too thin, so 0.2mm has been added internally all round.
- Water was trapped in the sorting rack causing it to distort when cured despite hours of drying time on a radiator.
- The footboards left much to be desired, these have been amended to be wedge shaped (thickening towards the rear) and have an extra couple of supports not on the real thing as this is preferable to wavy footsteps. The knife edge supports for the footboards were too light and themselves distorted so they have also been beefed up.
- The 3D printed W irons are too delicate. and two have broken off. To thicken them I'll need to move guides, axleboxes, springs and possibly other items out to maintain a credible look. They cannot be thickened inwards as there's already only room for 25mm axles in the etched suspension units.
- It was not possible to drill out the holes for the horizontal handrails along the side - so the holes will need to be enlarged and their mounts simplified.


Sadly, this was a perfectly formed roof when left to dry overnight. The roof with caboose was always going to be a challenge, so I've run some beefy knife edge supports down each side that should keep it aligned for curing and give it the strength needed for the glazing to be fitted. The roof sections and ribs have also been beefed up. Despite this the knife edge supports will almost certainly be needed until the glazed roof is ready to be glued in place.


What a difference a day makes! This was left on the built plate overnight in the printer (thus soaked in resin) so that I could keep an eye on it as it dried after washing. After washing almost all of the Blueprint supports were removed, and it was left to dry on my knife edge supports and grid. Once thoroughly dry there were no problems with curing it. In front of it are the N/S templates for the glazing that I made using paper templates printed out of QCAD. These were adjusted for final fit on the 3D printed model.


Some supports were left in as they weren't in the way of painting or glazing and might actually help retain the shape. The glazing templates are in position paper side in. Double sided tape will be used to secure the templates to the glazing material so that I can cut round them accurately. I'll paint the roof a nondescript grey as even if white lead paint had been used, it would have got grubby pretty quickly, and the roofs certainly don't look white on a photo of one of these sets taken at Salisbury. This roof won't fit the the original body print due to the thicker roof, and extra ribs. It seems likely that the caboose glazing bars and framing were brown though, so that needs to be looked into before I paint and glaze the thing.

 

[Stevers]

The 1859 Passenger Brake Van on eight spoke wheels and 0.3mm wire handles fitted in the opened out holes. The buffers are Kean Maygib and intended for PO/MR Wagons (from a pack of 20) - the last pack that Branchlines had in stock. Despite the two 3-sets being close coupled, twenty will still not be enough for the rake of eight coaches. I could switch to AGW 'Early Coach' buffers, but really need to try 3D printing my own housings to match these to take the AGW buffer heads and springs (available in a pack of 20). The caboose windows have been painted Precision Paints LSWR Coach Brown. I'm itching to glaze the caboose so that I can test fit the roof, but sadly no-one in my little Mendip town is currently able to sell me any suitable double sided tape to hold the templates in place so that I can accurately cut round them.


A big change is that wheels have been fitted with no sign of brass W irons, sprung or otherwise. To enable this, starting with this 1859 Brake, all of the 3D models have been modified to have the option of thickened W irons having pockets for pinpoint bearings. For EM and P4 models I can then print the brake operating mech in place as shown above. There is enough flex in the ABS-Like resin to permit wheels to be fitted. For a static model pinpoint bearings will be ideal, also suitable for a friend's 00 Gauge models and anything that just trundles in and out of a branch terminus, but springing will still be useful for the lumpy South Junction baseboard joints. I hadn't noticed that the wheels I'd bought had split spokes, but I think I think this was a happy accident for this model. The footboards with five brackets and the knife edge supports are just about OK and should look better when painted.


On the tail end of the train was an '1864' passenger brake van, the ends of my 3D model now being heavily modified to match a photo of such a van in one of these sets at Salisbury (below). From the mouldings I suspect that this is a teak bodied coach. This is a version for pinpoint bearings and the Newall brake operating mechanism can be seen.


And the real thing at Salisbury in 1882. The caboose is taller than in the Weddell drawing so my model has been adjusted to match. The sides are obviously two tone, so presumably brown and a pinkish cream being some precursor to the famous salmon. The end appears to be a single colour allegedly vermillion for brake vehicles, if so, not showing black as with some photographic emulsions. The grubby underframe was no doubt originally some form of 'chassis black'. Given the primitive braking arrangements it's possible that brake dust was not quite as prevalent on the underframes as in later years, but that they'd still be a muddy sort of colour in traffic. What I find most interesting is that the windows in the doors appear to slope in as if they were a 'hopper' window rather than a droplight. If so I'd really like to model them open! In any case my model still doesn't capture the look of the windows seen here. What I haven't seen anywhere is a note of the internal finishes for such a brake van.