British Railway Modelling (BRM)

Roof Constructi­on

-

Smart Models, is actually a representa­tion of the west wall. In reality, the east side of the station was the main access and the wall should be peppered with Gothicstyl­e windows and doors. This was simply done to save a lot of work and, instead, I decided to construct an island building to cover passenger facilities. The near side support wall is a simple strip of 6mm ply covered in the plain Carlisle wall texture. This includes a 6in-wide viewing aperture; big enough for a camera to poke through. I agonised over the correct height of the walls initially, but after having inspected prototype pictures, I figured that the roof beam to platform measuremen­t was somewhere between 28 to 32ft. I eventually settled on the lower dimension and also set the screen accordingl­y.

The platforms were next and were made from 12mm MDF that was covered in a combinatio­n of Scalescene­s and Smart Models textures pasted onto card. Again, this is where the model differs from the real location as there should be an additional bay platform for local services on the far side; instead, I plumbed for a single 4in-wide platform that butts directly up to the East wall and an 8in-wide island platform that splits the layout down the middle. The tracks were then laid in quick order using a straight edge to ensure accuracy.

The rear four tracks are Code 100 and are the link running lines; the cruder style track simply chosen for reliabilit­y and easy maintenanc­e. The six tracks to the front are all Code 75 Bullhead and lend the layout a little finescale credibilit­y; I particular­ly like the fishplate connectors as they give a good account of those used on the prototype.

The track was then painted black and given washes of paint. Ballasting costs could have been exorbitant given that there is nearly

70ft of line on the cameo. However, I am lucky to live near some ancient volcanic beaches here in the South West of Ireland, and once washed, dried and sieved, the sand collected provides a good impression of scale ballast. This was carefully laid and glued down with a watery mix of matt varnish.

The last of the main scenic components to be constructe­d was the Victoria Viaduct and access ramp. These were made from a combinatio­n of card, Plastruct, Peco Girder Plates and wooden dowels for the ramp supports. The decks are set at 65mm above the track bed and in order to ensure that other locomotive­s running through the link aren't impeded, especially those with

pantograph­s, the rear bridge is removable. Getting the right look for the ramp was difficult and a first attempt at making a panelled version was a disaster. However, after reviewing prototype photograph­s of the 1950s, I noticed that the Victorian panelling had been replaced with a speartoppe­d fence that had been painted white. A strip of 1mm thick white plastic sheet was cut and carefully scored to represent this. It was painted matt black, then scraped back with the tip of a scalpel blade once dry. This emulates a reasonable chipped paint effect. Weathering was achieved with powders.

The remaining details, apart from the roof, were the platform building and footbridge. The island passenger building is 8in long and has been covered in the Carlisle texture. The laser-cut windows and doors have been ‘borrowed’ from a Metcalfe

Castle Keep kit and closely match Carlisle’s neo-Gothic style. The LNER type lattice footbridge, placed over the far tracks, was from LCUT Creative and is a truly beautiful model, albeit a little tricky to construct.

Finally, it was on to the most difficult aspect of the build - the roof. This was very involving and it required a lot of planning beforehand. I wanted to do justice to the cavernous expanse of Carlisle’s ridge and valley structure and was determined to achieve a good match from the onset. While the structure is a mere 3ft 4in long, it is butted up against a mirror in order to double its scenic length; this can be easily slid out the way during running sessions. The roof structure is also removable to allow for track maintenanc­e.

The layout has presented an interestin­g challenge, albeit a little frustratin­g at times.

Building it has given me much food for thought about grouping and pre-grouping locomotion. With a credible backdrop, it would seem a bit remiss not to explore this further, so I’ve already got my eye on a Bachmann LNWR Precedent 2-4-0 and some six-wheeled coaches. There are also a plethora of pre-grouping kits from London Road Models to consider. During Carlisle’s heyday, seven railway companies used the exchange lines. The general arrangemen­t being that all follow on passenger and freight traffic had to be transferre­d to company locomotive­s – plenty of scope for a possible shunting project. Lastly, while the layout represents only a small section, I hope it goes some way in conveying the sprawling nature of the prototype – a station that, in my opinion, surely has to take top billing as a Cathedral of Steam.

Carlisle Citadel Station’s roof is unique in that it is a cantilever ridge and valley structure. Most roof trusses are supported by their heel points; however, at Carlisle these ends appear to float in mid-air. This is because the main support beams run through centre of the trusses instead. The three circle webbing is also unique and was patented by William Edgecumbe Rendle who incorporat­ed it into the overall design. It’s also worthwhile noting that the roofing contractor, Messrs Arrol of Glasgow, went on to construct the Forth Rail Bridge; another cantilever structure whose main supports bear an uncanny resemblanc­e to that of the Carlisle truss profile.

This roof design presented me with a difficult challenge; do you create both truss and beam together or retrofit them as separate components? I have no idea how the builder erected the original, but it took me many hours of speculatio­n to figure out a suitable method for the model.

 ??  ??
 ??  ?? This scene conveys the cavernous nature of Carlisle Citadel Station. By the late 1950s, the dilapidate­d Victorian end screens were gone, along with nearly two thirds of the roof – another sad
casualty of a network badly worn down by two World Wars.
This scene conveys the cavernous nature of Carlisle Citadel Station. By the late 1950s, the dilapidate­d Victorian end screens were gone, along with nearly two thirds of the roof – another sad casualty of a network badly worn down by two World Wars.
 ??  ??
 ??  ?? Two of the most famous Express Trains associated with Carlisle – The Royal Scot and Thames Clyde Express. A Hornby Coronation and Bachmann 7P 'Royal Scot' simmer under the screen awaiting departure. The beautifull­yrendered 4mm:1ft scale nameplates are from Fox Transfers.
Two of the most famous Express Trains associated with Carlisle – The Royal Scot and Thames Clyde Express. A Hornby Coronation and Bachmann 7P 'Royal Scot' simmer under the screen awaiting departure. The beautifull­yrendered 4mm:1ft scale nameplates are from Fox Transfers.
 ??  ?? A convincing factory-weathered Hornby 2P Fowler. The streaming sunlight effect was
achieved using an old spotlight picked up from a retired projection­ist.
A convincing factory-weathered Hornby 2P Fowler. The streaming sunlight effect was achieved using an old spotlight picked up from a retired projection­ist.
 ??  ?? Strips of cereal box card were cut and glued down to emulate the ridge caps and valley gutters. The metal truss joiners were also hidden under card plates that had been carefully duplicated by using a template; these mimic the actual joining plates as seen on the prototype. The structure was then sprayed matt black and given several soot and rust colour washes. The next operation was to install the roof light panels. The window texture was downloaded from Textures.com and printed out onto acetate. Several windows were randomly cut out to represent broken and missing panes of glass; a problem that plagued the prototype. These panels were then glued into place with superglue and given a weathering wash. The final operation was to place support columns on the platform. These are made from 4mm metal rod and are merely cosmetic additions because the roof structure is quite robust and self-supporting.
Strips of cereal box card were cut and glued down to emulate the ridge caps and valley gutters. The metal truss joiners were also hidden under card plates that had been carefully duplicated by using a template; these mimic the actual joining plates as seen on the prototype. The structure was then sprayed matt black and given several soot and rust colour washes. The next operation was to install the roof light panels. The window texture was downloaded from Textures.com and printed out onto acetate. Several windows were randomly cut out to represent broken and missing panes of glass; a problem that plagued the prototype. These panels were then glued into place with superglue and given a weathering wash. The final operation was to place support columns on the platform. These are made from 4mm metal rod and are merely cosmetic additions because the roof structure is quite robust and self-supporting.
 ??  ?? To make the process easier, the trusses were soldered together in a jig of their own because trying to join them up in situ would have been awkward and time consuming. They were laid in line with the beam positions to ensure accuracy. Each run was held in place by a flat piece of card and pins. The crude joint plates are slithers of wire; these were eventually hidden by carefully duplicated card plates in the final model. It was a messy business but the process made for an extremely stable arrangemen­t.
To make the process easier, the trusses were soldered together in a jig of their own because trying to join them up in situ would have been awkward and time consuming. They were laid in line with the beam positions to ensure accuracy. Each run was held in place by a flat piece of card and pins. The crude joint plates are slithers of wire; these were eventually hidden by carefully duplicated card plates in the final model. It was a messy business but the process made for an extremely stable arrangemen­t.
 ??  ?? Once complete, individual rows of trusses were brought across and offered up to their predetermi­ned positions to check for alignment. Then, using a Dremel, a small slit was created in top of each beam that matched the line on the jig. The trusses were then simply slid down into the slot and the opening soldered close. Note how the bottom of the large circular web sits down upon the girder bottom plate and how the ends of the curved chords butt up against the outside of the plate; just as on the prototype. These were also soldered to create a solid and rigid arrangemen­t. As on the real structure, the ridges, bottom plates and valleys had to all line up perfectly so there was a fair bit of MacGyverin­g during the process. However, and no matter how accurate I tried to be, some of the truss joints were out of level by up to 3mm in places. Luckily, these errors were evenly distribute­d and were therefore cancelled out. Once all of the trusses were in position, lengths of wire were soldered in place to represent the valley gutter positions and window bars. Once happy with the structure it was washed thoroughly to remove any flux residue.
Once complete, individual rows of trusses were brought across and offered up to their predetermi­ned positions to check for alignment. Then, using a Dremel, a small slit was created in top of each beam that matched the line on the jig. The trusses were then simply slid down into the slot and the opening soldered close. Note how the bottom of the large circular web sits down upon the girder bottom plate and how the ends of the curved chords butt up against the outside of the plate; just as on the prototype. These were also soldered to create a solid and rigid arrangemen­t. As on the real structure, the ridges, bottom plates and valleys had to all line up perfectly so there was a fair bit of MacGyverin­g during the process. However, and no matter how accurate I tried to be, some of the truss joints were out of level by up to 3mm in places. Luckily, these errors were evenly distribute­d and were therefore cancelled out. Once all of the trusses were in position, lengths of wire were soldered in place to represent the valley gutter positions and window bars. Once happy with the structure it was washed thoroughly to remove any flux residue.
 ??  ?? The roof required 78 identical trusses and each one needed approximat­ely 2 foot of wire to complete. The truss is not an exact copy; with no architectu­ral drawing I had to design a workable ‘look alike’ arrangemen­t between the chords and circular webs. Therefore, a template was first drawn out and cut from card before being transferre­d to a piece of plasterboa­rd. Lines were then drawn around the template and scored out. These channels held the wires firmly in place during the soldering process. This method ensured near accurate duplicatio­n despite the jig getting very messy with repeated use.
The smaller circles were made by wrapping wire around a broom handle. For the larger centre circle, an old postal tube was employed.
The roof required 78 identical trusses and each one needed approximat­ely 2 foot of wire to complete. The truss is not an exact copy; with no architectu­ral drawing I had to design a workable ‘look alike’ arrangemen­t between the chords and circular webs. Therefore, a template was first drawn out and cut from card before being transferre­d to a piece of plasterboa­rd. Lines were then drawn around the template and scored out. These channels held the wires firmly in place during the soldering process. This method ensured near accurate duplicatio­n despite the jig getting very messy with repeated use. The smaller circles were made by wrapping wire around a broom handle. For the larger centre circle, an old postal tube was employed.
 ??  ?? Six identical spans were required with an additional girder that matched the depth of the front screen. Each girder was made in two halves that were joined together. This was done because the strength of the whole Roof, as at the real location, is in the girder’s ability to resist sag across such a large span. The depth was dictated by the diameter of the centre circle web and its length by the span between each wall. They were assembled in their own jig and it took over 14 pieces of wire to complete each one. The bottom of the lattice type span is comprised of three wires soldered together to form a flat plate as seen on the prototype.
Six identical spans were required with an additional girder that matched the depth of the front screen. Each girder was made in two halves that were joined together. This was done because the strength of the whole Roof, as at the real location, is in the girder’s ability to resist sag across such a large span. The depth was dictated by the diameter of the centre circle web and its length by the span between each wall. They were assembled in their own jig and it took over 14 pieces of wire to complete each one. The bottom of the lattice type span is comprised of three wires soldered together to form a flat plate as seen on the prototype.
 ??  ?? The easiest way to construct the roof, I found, was to retrofit truss to girder. Firstly, a large piece of ply was laid and levelled atop an old wooden dining table to act as a master jig. A plan of the roof was then carefully drawn out on the surface; the girder beam positions were determined by the wall columns and the truss positions by the top windows joints in the screen. The girder beams were held in place by pieces of MDF and thumb tacks to ensure they remained upright during the joining process.
The easiest way to construct the roof, I found, was to retrofit truss to girder. Firstly, a large piece of ply was laid and levelled atop an old wooden dining table to act as a master jig. A plan of the roof was then carefully drawn out on the surface; the girder beam positions were determined by the wall columns and the truss positions by the top windows joints in the screen. The girder beams were held in place by pieces of MDF and thumb tacks to ensure they remained upright during the joining process.
 ??  ?? I used 1mm dia galvanised wire for the roof. Using material like this is not everyone’s cup of tea, but I simply cannot afford brass profiles or the like; especially given that the structure needed over 150 lengths to complete! The wire was stretched straight by using a large furniture clamp in reverse; this method introduces tension and lends the malleable wire a bit of strength.
I used 1mm dia galvanised wire for the roof. Using material like this is not everyone’s cup of tea, but I simply cannot afford brass profiles or the like; especially given that the structure needed over 150 lengths to complete! The wire was stretched straight by using a large furniture clamp in reverse; this method introduces tension and lends the malleable wire a bit of strength.

Newspapers in English

Newspapers from United Kingdom