tynewydd

In the film “The Long Drag” - 1963 - the Appleby Dairy cheese drying shed and its chimney are shown along with the attachment of ?six? 6-wheelers full of milk to the rear of a stopping train of about 4 coaches bound for Carlisle. The commentary says that they would there join the train to London. The Express Dairies sign says “Milk for London”.

Its also a question of the size of the non-platform areas being related to the choice. Those long trains will probably need staging/fiddle yard unless you are roundy-roundy, and its also less than convincing to have your 10 coach train with only 10 coaches space before arriving at the next station unless each one has a proper scenic break. I personally like the look, as others have said, of 5 or 6 coaches plus tender engine in OO. DMU 6-car Pullman is also fine. They substitute in my minds eye for the 11 or 12 coaches I would count out on the Irish Mail as it went about 1/4 mile away through my summer-cottage window past the Valley down distant signal to meet the boat in the 60s. I think the factor has got something to do with the length relative to the size of the human viewing it. That may explain why 3 or 4 looks fine in O to me, but 8 or 10 coaches is better in N. On an OO branch line? Platform for 3+, actual train only 1 or 2, please. Of course if we go back in time to the Edwardian era the prestige coaches were about two-thirds to half our modern length - a reason Buckingham was set then as the consideration there was the Denny rotating fiddleyard length needing to have a clear rotation path and not need too much engineering of the pivot while still allowing a realistic semi-fast train. Even with long platforms in a nice space, I agree with the late Rev., I think it important to have some visual break in the un-interrupted platform if possible with things like station buildings, roofs, overbridges, footbridges, gantries, signalboxes, etc to prevent the viewer easily seeing everything at once. A gently curved platform is also very helpful, this allows the mind to do some work and it will increase both the actual "coaches" length in the same space and even more increase the perceived length. This is especially true when standing nearer one end than at the middle, so having one or more of those visual "breaks" split the platform up into at least two if not three mini "scenes" will help a lot.

Apparently for steep enough hills, an extra lever could be added to prevent roller-coaster runaway type incidents - https://www.old-bus-photos.co.uk/?cat=212 I don't know if it was a Foden, but my dad remembered the issues when the single decker bus in Denbigh was converted to coal gas during the war, with a coke furnace towed at the back and a giant bag of gas on the roof. At the foot of the steep hill in Denbigh, first all the passengers had to disembark at the bottom, the fire was stoked up in the generator - which made it belch copious amounts of smoke - and then the bus would inch up the hill finally meeting up with the passengers who had all reached the top well ahead of it and were sitting on their parcels.

As I recall, the original impetus for the South Hams railway was the LSWR wanting a competitive way to get from Plymouth to Newton Abbot/Exeter in the 1860s - that scheme eventually got as far as Yealmpton only in, but even that was built as a through station as the initial terminus on the modified scheme was supposed to be Modbury. Given that the ambition was a link across the Dart to the Kingswear side - and with the Admiralty usually being sticklers for unimpeded navigation to 100ft (e.g., Menai and Royal Albert bridges), it seems to me that a crossing near Ditisham at a high level was a likely outcome using the local topography of very steep sides to the estuary to help level the approaches to the bridge. Perhaps it could have added a spur to a terminus/docks nearer Dartmouth, although Kingswear already supplied that to a great extent. This would also have avoided the need to breech the town itself. Many a station was sited several miles away from its nominal place after all. But in any case, the idea of a line from Totnes by the GWR, although easier from an engineering standpoint, would not have met the competitive case except as a pre-emptive strike - potentially leading to a dueling Dartmouth stations. From a modeling POV, the vista of through trains passing by a terminus but at a higher level is available. Of course, just as at Yealmpton, the scheme could simply have run out of money at any stage desired.

I have 122 points/turnouts all Seep motor controlled and ElectroFrog wired into the (in my case DCC Concepts Digital) point motors for switching. There are two powering wires from the accessory bus to each motor and one short wire from the frog to the included switch on the motor and two wires to from that switch to the secondary bus. In some cases the switch is the other end motor (double slip) or is an auto-electronic switch (diamond). It is a process, but by having the switch with onboard DCC, it is manageable where for me, wiring everything back to a central point is not. Chaçun à son goût. The point is you need to pick a systematic approach and follow through. In my case two DCC buses to every board, and then dropper wires to each point motor and the electrofrog special dropper to the point motor. It took about an evening a baseboard to do on average - but some of them had 20 motors. I guess my total count was about 600 dropper wires. Looking back now, I would have used a different joint system to make the bus to dropper wire connections rather than soldering each one, but at least I always worked with the board inverted or on its side. In full disclosure, then there are the detection sections wired via windings on DCC detectors to do as well if you intend to be able to automate or semi-automate sections. And you will need more for signals.... Still the aim should be, IMO, short wires to the common DCC accessory bus. Adam

In V5 flextrack is automatically transition curves and if you set Easement to "Cornu" then Joins are as well. If you set that and then take out the last curve section on each curve and replace with a Join, you'll get a result you can adjust. By adjusting the end points along the fixed and curved ends, the program(me) will fill in the gap with a dynamic transition curve rather than fixed radius or fixed transition. Hopefully the improvements in UI highlighting in V5 have made the learning curve just a little easier.

On the original topic, I was told by my father that the road from Valley to Wylfa was improved a lot for construction traffic - especially as some especially heavy components such as the transformers were sent by ship to HolyHead and transported to the site along it. There were one or two very large cranes at Holyhead opposite the station near the dry dock and I recall they were also used to put locos stranded by the Menai Bridge fire onto ships to get back to the "the mainland". So given that road was being improved anyway - adding to the yard at Valley was probably a natural for a road/rail transfer when compared to reinforcing and in some cases widening the Amlwch branch. My Dad has an interest in both areas as his father had been a curate at Llanerchymedd and had bought a cottage near Valley (before the days of Y Fali or Y Dyffryn) in the 1920s.

There is a Group dedicated to talking about BR carriage workings (including catering) which has posted many useful files to Google Drive - You can start by subscribing at https://brcoachingstock.groups.io/g/main - the links to the carriage working timetables is at the top of the messages page. Prepare to spend a lot of time reading!!

I have usually used the soldering iron bit on the side to melt the insulation for a section. Any hot item will do. Adam

https://www.atlasobscura.com/places/starcross-engine-house-and-brunels-atmospheric-railway States that some of the exhibits may have moved to the Newton Abbot Town and GWR museum

And XTrackCAD does work on a Mac without needing an emulator or a VMware/Parallels setup. Completely free to use (open source). Adam

For folks who are not building their own track, XTrackCAD (free) has a flextrack tool that produces eased transition curves that minimize the rate of change of curvature along the entire path between the ends using some heavy duty maths. The ends of the eased section can be straight or curved (fixed radius) as desired. Further tricks allow additional constraint points (equivalent of adding track pins in real-world flextrack) such as requiring the track path to "slew" to give clearance past an fixed obstacle like a signal, bridge abutment, etc. The eased path changes dynamically as the ends and constraints are adjusted so you can play with it until you have the result you like. Full disclosure - I wrote the code but borrowed the maths.

Wow! -> Limerick Junction is like old Chester General with the central platform crossover welded to Templecombe Junction for the "all passenger trains reverse" part on steroids - but the crossing is on the level to boot rather than with a bridge over the S&D. A wonderfully Irish layout!

The Summer 2021 SRS Signalling Record had a detailed article on the Bridge of Dun station. The station had an island platform on the Down side, and in the article there was mention of an intriguing working practice. Several times a day, three trains would converge on the island until at least the 1950s - a Down train to Montrose and an Up to Brechin would meet a Down Aberdeen express with the Brechin train being shunted backwards from the Up main over to the outside of the island the other end of which was already occupied by the Down Montrose train facing the other way. The Down express to Aberdeen would then arrive on the other side of the island and all the passengers and luggage could quickly transfer between trains without using the bridge or a level porter's crossing. The express would leave, and then the Brechin train, and finally the Montrose train following the express as far as Dubton Junction. This seems like an interesting "move" operationally and I have such an island with the right sort of multiple lines joining and leaving to mean that Up and Down trains might benefit from a transfer - in my case I could actually see 4 trains meet as the local branch has a dock in the island as well. The question is - was this a one-off for BoD or were there other examples of creativity in use of an island platform to minimize passenger transfer time so that this would look interesting rather than out of place? Adam

This is what has happened to the US Post Office as well. They have to deliver to everywhere and have a social contract that the cost only depends on distance door-to-door, not on degree of difficulty. Companies like Amazon can and do pick or choose and just use them for the most uneconomic deliveries of the "last mile(s)". One can observe, however, that when the boot was on the other foot and private companies on rail were the only way to go in the US, they operated as you would expect giant monopolies to do - they price gouged the little guy when there was no competition and acted to stop any new competition. There were similar complaints about the LNWR and then the LMS (and probably all the big four) in the UK in areas where they were effective monopolies like North Wales. When my father travelled to school in Leatherhead from St Asaph in 1940 he went via Wrexham and observed that the LMS and GWR trains were perfectly timed to miss each other to reduce the convenience of the route to Paddington compared to Euston. I think two major problems with Beeching were the lack of understanding of the value of the network effect to traffic, which leads to devaluing the "twigs", and having a planning horizon that didn't properly value returns from investments in efficiency to reduce the overall cost. The same standards were not applied to the road network, of course, either in the ROI for all the rural roads per mile or in setting up a true comparison of different investment strategies for how freight and passengers would most efficiently travel the longer routes once the novelty wore off. But the overall sentiment is correct - Beeching just reaped the whirlwind as a totem of the policies implemented both before and after him. He was asked to do a job - he did it and he was not about to produce something that was not PC.

On our way to Devon to stay with Gran, we used to take out a Cornish Pasty in a flat tin (same mixture as a usual but flat) that my mother had made cut into squares, spread out a tartan rug on the verge or central reservation depending on which lane we were stopped in and have a relaxing lunch! There was always a bit of a warning of movement starting again and so time to pack up. This pattern persisted on the new M5 at whatever point it had ended as it was built south, and then the A38 as it was "dual-ed" as well for a bit. I preferred the train journey from Derby to Plymouth as a youngster (travelling alone at 7 yo) but that meant relying on Granny for transport in her Morris Minor with the wood back - and she drove a car like a tractor - needing an average of three clutch rebuilds a year.

If we compare to the 1963 picture from roughly the same position on wikipedia, https://en.wikipedia.org/wiki/Bromyard_railway_station#/media/File:Bromyard_railway_station_1921287_0d209dac.jpg. We can see that this was not a temporary repair - but rather a decision to remove the edge that persisted. I also notice in passing that the wartime white painting of the platform edge is retained on the entire remaining up platform in 1959 but only seems to be present on a very small portion of the down, just near the platform building shelter even then - maybe where the awning protected it.

Some LNWR kit survived a lot longer than 30 years. There was a working ex-LNWR LQ starting signal beyond the end of a platform at Holyhead well into the 1980s. It was specially sized arm (short), which gave enough clearance for the train on the adjacent line to go past it when it was on. Probably closer to 100 years service in that case. Holyhead had LQ, UQ and Color Light all at the same time.

Similarly, XTrackCAD (always free) uses a Cornu curve algorithm for its flextrack tool that does real-time curve fitting to produce the smoothest curve in the space available. You can drag the ends around or nudge connected tracks about and see the results in real-time. Technically it is ensuring the radius varies linearly along the length of the arc which is the underlying characteristic of all smooth transitions. In the latest version, you can add fixed track pins at key points to allow "slewing" around objects as part of the overall smoothed result. It is true that transitions can be approximated by a series of fixed curves as well. For simple transitions that is easy. Things can be more complex to calculate as the number of constraints (angles and radii of connected tracks, desirable offsets) increases - so letting software take the strain may be easier. Adam

Be aware that the radii (I can say for certain at least for Peco) are "nominal". This means that the angle and radius of the track ends especially for for wyes and curved points will often not be accurate for real life, but that the end position will be roughly correct using the radius listed. The points are compounds made up of straights and curves of different radii so a single radius for each path will not work. These discrepancies are usually not important until complex sets of points are used close together in tight formation. At least in the case of the latest XTrackCAD, in HO/OO and N (because I did the work with kind help from Peco) the included definitions are accurate for position, angle and radius at the ends of the points. Adam

From a reliability POV, energized is the way to go - but you'll have to weigh the hassle factor against that. If you are going to use point motors such as the Cobalt, the additional cost for normal turnouts is some extra wire only (rather than a new component such as a juicer) as the switching is onboard the motor. It is a really only another dropper wire to include and the wire from the bus to the other parts of the switch. You will need to prepare the points for energized frogs, but once you are underway it is simply repetition. For diamonds, which have no motors, juicers will needed for live frogs. If those loops reverse, you'll need auto-reversers (like a double juicer) to ensure the sections polarity is correct. When you do that, make sure the isolated loop is long enough to accomodate all the powered parts of the trains so that both ends are not bridging the isolated section to the rest of the layout and so wanting to change polarity at the same time. If not you will have equipment that you can't use on some of your layout. (I'm looking at you, Blue Pullman!!). Adam (256 count)

The book "An Historical Survey of Chester to Holyhead Railway - Track Layouts and Illustrations" by Anderson and Fox published by the OPC contains both a signalling diagram and a trackplan for Menai Bridge along with nearly all the other stations on the line. For more color, and information on additional lines including Bangor to Caernarfon, the late Bill Rears book "From Chester to Holyhead - The Branch Lines" is highly recommended. Adam

It is certainly true that the Peco track shapes are often both compound curves and incorporate transitions to "fit it all together". Peco explained as much when we were discussing the dimensions with them. The XTrackCAD V5.2 upgraded "turnout builder" tool that we use to create the track templates incorporates a Cornu curve fitting engine -> which does the heavy lifting to use the mathematical family of curves from which railway easement/transitions come. We also use the engine to solve for these conditions in real time while users dynamically join their tracks together with transitions/easements as the user drag the ends or move additional "track pin" constraints along the desired track path. The externals (positions, angles, and centers and radii) of the points are fed in and the resulting track curves are smooth transitions between them. This is not to say that our curve solution for a turnout is the one Peco used of course, which is why we don't claim that the internals are identical, but it is the smoothest curve achievable with a linearly varying curvature given the end constraints. I'm sure that the Rev PD method of laying out his station throats and so on allowed him to achieve superior results because his track lines would all be smooth curves throughout the frogs and switch blades. The scratch track builder has the advantage of a fully customizable solution. This degree of design freedom, as I understand it from the excellent help material at Templot, would have been uncommon in much of the prototype because the trackwork was usually assembled in parts from sets of stock items that each had fixed geometry rather than being fully "bespoke". But the real thing was also not anywhere as compressed as the typical layout, allowing for much more languid curves - even accounting for scale - in which the inserted fixed elements would prove much less of a "jolt" anyway.

Just a note to add to this discussion that the latest version of XTrackCAD V5,2 (multi-platform and free as always) has adjusted several Peco OO and N turnouts/points based on the good graces of the technical bureau at Peco. The turnout dimensions were corrected for curved, three way and Y points. XTrackCAD does not aim for precision within the point but does for the external offsets, angles and radii so that the overall design fits together as it will when laying the real track. Our files will not help attempts to copy internals by others as we did not obtain or model those details. I would say that Peco has been much more open to providing this level of information than other manufacturers we asked. We previously found that trying to accurately measure exact dimensions from the PDF templates available from Peco website was very hard especially for end radii and angles on these more complex types. We also had some suspicion that the PDFs and the products might have diverged very slightly over time as the molds were repeatedly retooled and replaced and improvements made to enhance running. Adam

I presume that "Heaven" is something like a port or a much bigger commuter city, because having that number of platform faces in use simultaneously with equipment that can simply reverse ends and leave again, means a lot of arrivals in a row (or departures one after another) otherwise they will nearly always be empty. Such a pattern does suit a port aka H(e)aven - which would have a tidal pattern coinciding with boats arriving and departing. It could be that the lower one or two platform faces are about parcels traffic, perhaps, while only the upper 3 or 4 are passenger from the time when the BR Parcels service commanded its own sections of stations. Those parcels might even still be loco hauled or even a DPU - which could haul other parcel stock and need to run-around. Southampton Terminus would have been such a station with a parcels/mail side, but it was omitted from electrification and closed instead. Agree about the reverse curves being nice to remove, but that's not as important if there is not going to be shunting or propelling over them going on. 10 cars will be a stretch for the longest two (7' platforms) in OO even if EMUs.