Model Railways On-Line - Ashprington Road - Part 2
ASHPRINGTON ROAD - Part 2 A 00 layout based on the Western Region Main Line between Exeter and Plymouth By Graham Plowman Time has been passing and it is now 1984 on the South Devon Railway Above: An HST bound for Penzance passes Class 50 No.50043 ‘Eagle’ with a train for Bristol. The Exeter Area Resignalling Scheme is under way and the semaphore signals at Ashprington Road will soon be replaced by colour lights. A paper template of the layout The next step was to build a paper ‘model’ of the layout. For this I used sheets of newspaper held together with sticky tape and cut them to the shapes of the boards. These were laid out on the floor of the railway room to give a full size representation of the layout. This is also a good technique for refining the board shapes. I had worked out the board shapes mathematically, but it wasn't until I created the newspaper templates that I found that one of my calculations was incorrect and needed revising. This could have been a costly error if I had been constructing with timber! With the newspaper templates on the floor, I laid track loosely on them to get an idea of how the layout would appear. Fortunately, the track fitted on the ‘boards’ without any problems. The paper templates were also used to determine the position of board bracing to avoid conflict with point motors. Building the baseboards The baseboards were built using 12mm plywood because this material stands up to changes in temperature and humidity much better than the traditional chipboard and 2-by-1 method. I had also obtained a large amount of Dexion (the giant Meccano used for shelving racks) to form the bracing beneath the boards. With 12 boards to build, construction took a couple of months. The boards are held together with bolts and each joint is supported by a pair of legs. When all of the boards were completed and erected, they needed to be levelled. I used a builder’s spirit level and found that I needed to make a number of adjustments to ensure that all of the boards and particularly the joints, were level. After levelling, the track layout was drawn out on the surface of the boards. Large curves were marked out using a measuring tape. I positioned my folding workbench in the middle of the layout and held a piece of timber in a vice with a nail in the top. A tape measure was hooked on to the nail and the arc marked out on the boards. This was done for curves at both ends of the layout but proved more difficult on the ‘London’ end because the board level is 20icm lower for the viaduct. Years ago, my father was lucky enough to obtain some railway drawing curves, which were being disposed of by a drawing office. These have been invaluable. Viaduct Construction I have always felt that having model trains going over a bridge or viaduct adds an extra dimension to the appearance of a layout. Before commencing construction, I undertook some research examining pictures of viaducts in Devon and Cornwall. The purpose of this was to identify the little details and techniques used in building the prototype that are so often missed in models: There is rarely much distance between the top of the arches and the underside of the track. From the underside of the sleepers to the top of the brickwork can be as little as six inches of ballast! The locating holes used to position the timber formwork used during construction are often missed. The stonework ‘style’ often changes between the pillars and the viaduct arches. Many viaducts have refuges for track staff to stand in safety when a train passes. Some viaducts have fencing instead of parapet walls. Those with fencing have low parapet walls usually only just high enough to retain the track ballast from falling off. Most viaducts (in the UK) have some form of guardrail device to restrain a derailed train. I chose to base my model on Clinnick Viaduct, which is a six-arch structure on the 6imile bank between Doublebois and Bodmin Road in Cornwall. The structure is straight, so some modeller’s license has been applied by building the model to suit a curve on my layout. Clinnick Viaduct is a very good example of the standard design of viaduct used in the Devon and Cornwall area, representing a tidy, yet interesting structure. Building the model turned out to be a mammoth task. The basic structure was constructed using plywood, nailed and glued together. It is a complete self-contained unit, which can be removed from the layout. It has a ‘novelty’ feature in that because there would be a point motor located between two of the arches, there is a cable ‘pull string’ that can be used to pull cables up and down the inside of the pillar. The structure was covered with Polyfiller. This was left for two days to dry before starting the carving of the stonework. I used a modelling drill with a mini ‘countersink’ and this worked very well. However, the carving process took an inordinate amount of time to complete to the point where I almost lost interest! The moral here is to build a smaller viaduct! Capping stones were made of balsa wood. These were attached first and then rubbed down to shape. Side fencing, made by Faller has been purchased but not yet fitted. A London bound HST emerges from the tunnel to the west of Ashprington Road. Preparing the trackbed I decided from the outset that all of the main running tracks would be canted (tilted) on curves. To achieve this requires a different approach to construction from the conventional method of laying track directly onto the board surface. I used a layer of 6imm plywood to form a sub-base for the trackbed. At the outer edge of the sub-base, on curves, a 1icm wide strip of 3imm thick hardboard was tucked under the edge to provide a tilt. The assembly was held in place by screwing to the main baseboard. Where curves pass through a transition into a straight, the construction becomes more complicated requiring the hardboard to be progressively increased, or reduced in thickness. Suffice to say that prototype geometry has been adopted. This was achieved by using professional railway alignment design computer systems and plotting the plans to 1:76.2 scale for laying directly onto my baseboards. Because we have to make significant compromises for the radius of curves on our model layouts, all of the transitions on my layout correspond to the minimum length permissible on the prototype. This scales to about 26cm long, or about the length of a coach. The 3mm packing under the trackbed described above, creates an angle of super elevation, which is equivalent to 2 inches of cant on the prototype. I have chosen to have both tracks canted in the same plane as this significantly simplifies construction, and enables crossovers to be located on curves. There are several other possible configurations for canting pairs of tracks. The emerging standard appears to be for the two inside (sixfoot rails) to be at the same level and for the outer (cess) rail of the outer track to be raised through transitions while the inside (cess) rail of the inner track is lowered through transitions. A set of points and a pair of working ground signals are positioned on the viaduct. The cables to the point motor and the pull rods for the signals pass down through the pier of the viaduct. The electrical system Once a trackbed has been prepared, there is a great temptation to start laying track. However, this can lead to all sorts of problems such as the track having to be pulled up to fit insulating joints or the use of rotary saws to cut rails to create isolating gaps. It becomes necessary for wires to be soldered to the sides of rails, resulting in unsightly solder globules and melted plastic sleepers, which destroy any possibility of realism. Planning beforehand prevents all of these problems. I designed the electrics in the form of a feed diagram. This included all isolating sections, overrun dead sections, signal run past sections and switching between controllers. Using this diagram, I would cut the rails appropriately and fit isolating fishplates where required. Although supplied as ‘live frog’ turnouts which can be connected up and used ‘as is’, the Peco code 75 turnouts are designed so that they can be re-wired to meet the modellers’ requirements. Essentially, the switch rails can be connected electrically to the stock rails and isolated from the crossing vee. This enables the vee to be fed separately and not to rely on the contact of the switch rails. To prevent any possibility of shorting on the backs of the switches caused by out of gauge wheels, I have rewired all of the turnouts to provide separately switched feeds to the crossing vees. Point motors are mostly Peco and use the Peco switches. In certain situations, Seep motors were later used as these proved to have much more reliable switches. Signals are GW lower quadrant semaphore, using mechanical relays. Since the layout is also required to represent the BR blue period, the use of semaphore signals dates it to 1984 at the latest. After this date Exeter Control Centre came into use and all signalling in the area was replaced with colour lights. Track laying Track laying commenced once the electrical diagram had been completed. Track was glued down using either Evo-stik or PVA after the appropriate wires had been soldered to the undersides of the rails. Computer control system As previously mentioned, the layout is computer controlled. The system controls all movements of points and signals but it does not drive the trains. Trains are driven using conventional Gaugemaster WS handheld controllers. I have used the ‘Remote Panel Control’ hardware system available from the Model Electronics Railway Group (MERG). The core of this system is a control board, which connects via a cable to the RS232 port of a PC computer. A number of other boards connect to the control board to form a stack. There are boards containing 8 relays, 32 logic level outputs, 32 logic level inputs and 8 track circuits. The relays are used to control track section switching between controllers and switching of isolating sections. The logic level outputs are used for controlling the point motor modules and signals. The signals are mechanically powered from relays connected to the logic level outputs. Using this system, there is no need for a control panel – on this layout it is a computer screen. Computer software The software used is the ‘Solid State Interlocker’ (SSI) software from GPP Software. The website address is www.gppsoftware.com/ssi/ssi.asp. This is the only software available, which represents British practice. SSI simulates modern IECC computer control systems used in UK signalling centres and looks and operates like the real thing, including graphics and the mouse operation of Entry/Exit route setting. The SSI software supports a fully signalled and interlocked layout and these capabilities have been fully utilised. HST's pass at Ashprington Road. Signalling Signals are constructed from Ratio kits suitably modified. They are all fully working including ground signals and are operated from the mechanical action of relays controlled by the computer system. Scenery Once the signalling was completed, a start was made on the construction of the scenery. Wood-fibre insulation board was used for the formers, which was cut to suit the profiles of embankments and cuttings. The gradient of cutting and embankment slopes is 1 in 1½. This is about the maximum value for normal soils. Rock can be steeper and clay is shallower. I made up a 1 in 1½ triangle from card as a template for slopes. The scenery formers were covered with chicken wire. Layers of newspaper soaked in plaster have been laid over this to create a reasonably thick surface with some strength. When dry, this was all painted in a light green, which can be seen in the photographs. This is a base colour for the scenery and will later be covered with scenic materials. Buildings ‘Ashprington Road East Signal Box’ is a Ratio ‘Highley’ kit and the west signal box is from Hornby. The Hornby model is of Hagley Signal Box and their GWR footbridge kit is from the same location. I will be installing the footbridge at a later date but temporarily I am using an Airfix/Dapol kit footbridge. Other buildings to be constructed will include a stationmaster’s house, which will be made from Linka. A goods shed and the abutments for the bridge over the railway at the western end of the station will be constructed using the same techniques as I have described for the viaduct. The steel plate girders of the bridge are MDF and will be detailed with plastic and metal strips. The tunnel mouths are from Merit. Rolling Stock My collection of rolling stock falls into two distinct periods, late BR steam with early diesels and the blue period of 1984. Steam locos are a mixture of Bachmann, Hornby and Mainline with early diesels from Bachmann, Mainline, Lima and Heljan. All have been weathered. The 9F and the Britannia are hand built. The construction of the Britannia is featured on page 16 of this edition of the magazine. Diesel locos in the blue period are from Mainline and Lima. Coaching stock for the steam era is predominantly Bachmann MK1’s with a few Replica and Hornby coaches, while for the later period they are mostly Hornby and Lima. Wagons are a mixture of Bachmann, Replica, Hornby and Dapol. Class 50 No.50043 ‘Eagle’ coasts over the viaduct as it approaches the station with a train for Plymouth. A Class 31 A1A-A1A emerges from the West Tunnel and enters the Up Yard with a freight composed mostly of lwb open wagons. Operation The computer control software provides for timetables and schedules. I am currently in the process of testing a schedule prior to entering it into the software. Summary I would like to thank my wife for all of her support and for her tolerance of my hobby. Thanks also to my father for his professional expertise and assistance in designing the track layout.
The next step was to build a paper ‘model’ of the layout. For this I used sheets of newspaper held together with sticky tape and cut them to the shapes of the boards. These were laid out on the floor of the railway room to give a full size representation of the layout. This is also a good technique for refining the board shapes. I had worked out the board shapes mathematically, but it wasn't until I created the newspaper templates that I found that one of my calculations was incorrect and needed revising. This could have been a costly error if I had been constructing with timber!
With the newspaper templates on the floor, I laid track loosely on them to get an idea of how the layout would appear. Fortunately, the track fitted on the ‘boards’ without any problems. The paper templates were also used to determine the position of board bracing to avoid conflict with point motors.
The baseboards were built using 12mm plywood because this material stands up to changes in temperature and humidity much better than the traditional chipboard and 2-by-1 method. I had also obtained a large amount of Dexion (the giant Meccano used for shelving racks) to form the bracing beneath the boards.
With 12 boards to build, construction took a couple of months. The boards are held together with bolts and each joint is supported by a pair of legs. When all of the boards were completed and erected, they needed to be levelled. I used a builder’s spirit level and found that I needed to make a number of adjustments to ensure that all of the boards and particularly the joints, were level.
After levelling, the track layout was drawn out on the surface of the boards. Large curves were marked out using a measuring tape. I positioned my folding workbench in the middle of the layout and held a piece of timber in a vice with a nail in the top. A tape measure was hooked on to the nail and the arc marked out on the boards. This was done for curves at both ends of the layout but proved more difficult on the ‘London’ end because the board level is 20icm lower for the viaduct. Years ago, my father was lucky enough to obtain some railway drawing curves, which were being disposed of by a drawing office. These have been invaluable.
I have always felt that having model trains going over a bridge or viaduct adds an extra dimension to the appearance of a layout. Before commencing construction, I undertook some research examining pictures of viaducts in Devon and Cornwall. The purpose of this was to identify the little details and techniques used in building the prototype that are so often missed in models:
I chose to base my model on Clinnick Viaduct, which is a six-arch structure on the 6imile bank between Doublebois and Bodmin Road in Cornwall. The structure is straight, so some modeller’s license has been applied by building the model to suit a curve on my layout.
Clinnick Viaduct is a very good example of the standard design of viaduct used in the Devon and Cornwall area, representing a tidy, yet interesting structure. Building the model turned out to be a mammoth task. The basic structure was constructed using plywood, nailed and glued together. It is a complete self-contained unit, which can be removed from the layout. It has a ‘novelty’ feature in that because there would be a point motor located between two of the arches, there is a cable ‘pull string’ that can be used to pull cables up and down the inside of the pillar. The structure was covered with Polyfiller. This was left for two days to dry before starting the carving of the stonework. I used a modelling drill with a mini ‘countersink’ and this worked very well. However, the carving process took an inordinate amount of time to complete to the point where I almost lost interest! The moral here is to build a smaller viaduct! Capping stones were made of balsa wood. These were attached first and then rubbed down to shape. Side fencing, made by Faller has been purchased but not yet fitted.
I decided from the outset that all of the main running tracks would be canted (tilted) on curves. To achieve this requires a different approach to construction from the conventional method of laying track directly onto the board surface. I used a layer of 6imm plywood to form a sub-base for the trackbed. At the outer edge of the sub-base, on curves, a 1icm wide strip of 3imm thick hardboard was tucked under the edge to provide a tilt. The assembly was held in place by screwing to the main baseboard.
Where curves pass through a transition into a straight, the construction becomes more complicated requiring the hardboard to be progressively increased, or reduced in thickness. Suffice to say that prototype geometry has been adopted. This was achieved by using professional railway alignment design computer systems and plotting the plans to 1:76.2 scale for laying directly onto my baseboards. Because we have to make significant compromises for the radius of curves on our model layouts, all of the transitions on my layout correspond to the minimum length permissible on the prototype. This scales to about 26cm long, or about the length of a coach. The 3mm packing under the trackbed described above, creates an angle of super elevation, which is equivalent to 2 inches of cant on the prototype.
I have chosen to have both tracks canted in the same plane as this significantly simplifies construction, and enables crossovers to be located on curves. There are several other possible configurations for canting pairs of tracks. The emerging standard appears to be for the two inside (sixfoot rails) to be at the same level and for the outer (cess) rail of the outer track to be raised through transitions while the inside (cess) rail of the inner track is lowered through transitions.
Once a trackbed has been prepared, there is a great temptation to start laying track. However, this can lead to all sorts of problems such as the track having to be pulled up to fit insulating joints or the use of rotary saws to cut rails to create isolating gaps. It becomes necessary for wires to be soldered to the sides of rails, resulting in unsightly solder globules and melted plastic sleepers, which destroy any possibility of realism. Planning beforehand prevents all of these problems.
I designed the electrics in the form of a feed diagram. This included all isolating sections, overrun dead sections, signal run past sections and switching between controllers. Using this diagram, I would cut the rails appropriately and fit isolating fishplates where required.
Although supplied as ‘live frog’ turnouts which can be connected up and used ‘as is’, the Peco code 75 turnouts are designed so that they can be re-wired to meet the modellers’ requirements. Essentially, the switch rails can be connected electrically to the stock rails and isolated from the crossing vee. This enables the vee to be fed separately and not to rely on the contact of the switch rails. To prevent any possibility of shorting on the backs of the switches caused by out of gauge wheels, I have rewired all of the turnouts to provide separately switched feeds to the crossing vees. Point motors are mostly Peco and use the Peco switches. In certain situations, Seep motors were later used as these proved to have much more reliable switches.
Signals are GW lower quadrant semaphore, using mechanical relays. Since the layout is also required to represent the BR blue period, the use of semaphore signals dates it to 1984 at the latest. After this date Exeter Control Centre came into use and all signalling in the area was replaced with colour lights.
Track laying commenced once the electrical diagram had been completed. Track was glued down using either Evo-stik or PVA after the appropriate wires had been soldered to the undersides of the rails.
As previously mentioned, the layout is computer controlled. The system controls all movements of points and signals but it does not drive the trains. Trains are driven using conventional Gaugemaster WS handheld controllers.
I have used the ‘Remote Panel Control’ hardware system available from the Model Electronics Railway Group (MERG). The core of this system is a control board, which connects via a cable to the RS232 port of a PC computer. A number of other boards connect to the control board to form a stack. There are boards containing 8 relays, 32 logic level outputs, 32 logic level inputs and 8 track circuits.
The relays are used to control track section switching between controllers and switching of isolating sections. The logic level outputs are used for controlling the point motor modules and signals. The signals are mechanically powered from relays connected to the logic level outputs.
Using this system, there is no need for a control panel – on this layout it is a computer screen.
The software used is the ‘Solid State Interlocker’ (SSI) software from GPP Software. The website address is www.gppsoftware.com/ssi/ssi.asp.
This is the only software available, which represents British practice. SSI simulates modern IECC computer control systems used in UK signalling centres and looks and operates like the real thing, including graphics and the mouse operation of Entry/Exit route setting.
The SSI software supports a fully signalled and interlocked layout and these capabilities have been fully utilised.
Signals are constructed from Ratio kits suitably modified. They are all fully working including ground signals and are operated from the mechanical action of relays controlled by the computer system.
Once the signalling was completed, a start was made on the construction of the scenery. Wood-fibre insulation board was used for the formers, which was cut to suit the profiles of embankments and cuttings. The gradient of cutting and embankment slopes is 1 in 1½. This is about the maximum value for normal soils. Rock can be steeper and clay is shallower. I made up a 1 in 1½ triangle from card as a template for slopes.
The scenery formers were covered with chicken wire. Layers of newspaper soaked in plaster have been laid over this to create a reasonably thick surface with some strength. When dry, this was all painted in a light green, which can be seen in the photographs. This is a base colour for the scenery and will later be covered with scenic materials.
‘Ashprington Road East Signal Box’ is a Ratio ‘Highley’ kit and the west signal box is from Hornby. The Hornby model is of Hagley Signal Box and their GWR footbridge kit is from the same location. I will be installing the footbridge at a later date but temporarily I am using an Airfix/Dapol kit footbridge.
Other buildings to be constructed will include a stationmaster’s house, which will be made from Linka. A goods shed and the abutments for the bridge over the railway at the western end of the station will be constructed using the same techniques as I have described for the viaduct. The steel plate girders of the bridge are MDF and will be detailed with plastic and metal strips. The tunnel mouths are from Merit.
My collection of rolling stock falls into two distinct periods, late BR steam with early diesels and the blue period of 1984.
Steam locos are a mixture of Bachmann, Hornby and Mainline with early diesels from Bachmann, Mainline, Lima and Heljan. All have been weathered. The 9F and the Britannia are hand built. The construction of the Britannia is featured on page 16 of this edition of the magazine.
Diesel locos in the blue period are from Mainline and Lima.
Coaching stock for the steam era is predominantly Bachmann MK1’s with a few Replica and Hornby coaches, while for the later period they are mostly Hornby and Lima.
Wagons are a mixture of Bachmann, Replica, Hornby and Dapol.
The computer control software provides for timetables and schedules. I am currently in the process of testing a schedule prior to entering it into the software.
I would like to thank my wife for all of her support and for her tolerance of my hobby. Thanks also to my father for his professional expertise and assistance in designing the track layout.