Warning: "continue" targeting switch is equivalent to "break". Did you mean to use "continue 2"? in /homepages/45/d549400858/htdocs/nicklebyhousewebsite/templates/sdr_template_nov17/functions.php on line 199
Planning the South Downs Railway
This section the design of the railway, and how it came to be.
The railway is designed to fit into a part of my 15' x 15' shed, allowing other uses for the shed. It needs to be compatible with the existing workbench; it needs to permit the childrens' railway to be stored and operated. because of that, a 15x10 feet space (approx 5m x 3m) was set aside for the railway, but the "hollow" in the middle would have to be multipurpose.
It is not intended to be representative of any real railway, either visually or in terms of train operation. Instead it is designed to allow an interesting range of train movements. To fit in a reasonable amount of track, "N" gauge is to be used. It will represent modern image UK railways, but with no particular regard for types of trains expected to be in any one area.
The design was sketched out so that a range of train movements could be planned. The track design is shown below.
A continuous run double track main line runs around the baseboard at varying altitude. Much of the run around the middle of the layout is hidden at low level (deliberately so at the front of the layout so as to make it accessible to humans in case of problems). A fiddle yard is a part of this track loop; the aim is for around 10 lines in total.
two reverse loops are available, they allow trains to change direction from one line to the other and vice versa. They also have a few storage tracks to allow trains to be delayed.
A piece of timber marked with the lengths of various trains has been very useful for final checking of the design. It is planned to run trains with a locomotive and 8 carriages. Some typical train lengths are:
- Class 47 loco + 8 65' carriages: 1275mm
- Class 47 loco + 8 bogie oil tankers: 1133mm
- 8 carriage Eurostar: 1060mm
The design needs to take account of train size for automatic control: the sections between signals need to be slightly larger than the maximum train length and "blocks" for automatic operation will probably follow the signal positions. Each block needs two train detectors: one to sense occupancy, and another to sense the train being at the end of the block so that it can be told to stop.
I planned to use Peco code 55 track from the outset. This looks nice, has electrofrog points and has the "scissors crossover" available.
I wanted to avoid curves that looked daft, while space constraints did not permit prototypical radii. Most rolling stock is specified to operate down to 9" radius; I tested several trains and they negotiated it with ease. (I do know that some co-co type diesel locos will NOT run reliably around 8.5" radius bends). My plan was to keep radii to 12" minimum if possible, and use larger radii on visible sections.
The Peco 26.5mm track spacing leads to collisions on tight corners, and that spacing can only be used on 15" and above radius. Realistically, to allow margin for error, the minimum radius used for that spacing should probably be 18". At smaller radii, the gap needs to be opened out by 3-4mm.
In practice I have tried to avoid 12" radii except in hidden areas, and most curves are nearer 18" radius. All visible pointwork will be medium radius (18").
The layout has multiple levels. I wanted reasonable sized trains to be able to run without difficulty, so I needed to know that the slopes would be manageable.
To find out what slopes could be used, I conducted some tests using three locos with realistic loads:
- Kato "Eurostar", pulling its 8 carriage rake;
- Farish class 47, pulling 4 carriages + the weight of 4 more;
- Farish 0-6-0 PT, pulling 4 carriages + the weight of 4 more
The conclusions were that in a straight line, these trains would start from standstill on a 1:35 slope and 1:40 was comfortable. I wasn't able to try the effects of a full axle count, or try pulling trains around corners. Consequently I've concluded that I still need to allow some margin.
In practice, my worst case slopes are around 1:60. They were planned to be better than that, but I've kept changes in gradient away from points etc and that resulted in shortening of some of the slopes themselves.