Ashburton and Totnes

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 23

 

No 4706 2-8-0 c.1930

The big one – designed for use on fast fitted freight turns (but often filled-in on weekend excursions).

Built 1923 Withdrawn 1964

 

The model was my first effort in completely scratch-building a loco. The chassis and mechanics are of sheet brass and nickel-silver but the superstructure is of plastic card.

Power is provided by a tender-mounted Portescap 1219 motor, with a flywheel and a simple bent wire cardan shaft to a worm gear in the engine firebox. This feeds a double-reduction gear drive, giving a reduction of 53:1. Both engine and tender have split frames with wheels from the late Mike Bryant range, all wheels collecting current. The drive is taken to both the third and fourth axles, reducing the stress on coupling rods. As a conceit, there is operating outside and inside valve gear, the latter almost totally invisible under such a massive boiler! See link below:

https://youtu.be/257UZ3nXnck

The superstructure is of plastic card; adhesion weight is hardly a problem with such a volume to fill up with lead. Most details are of metal.

Current collection from the tender chassis is transferred to the engine by phosphor-bronze wire springs, soldered to the tender and bearing on the insides of the engine frames. Of course, this only works with split-frame construction.

The combined weight of the engine and tender is a substantial 118g.

 

The model won the 2mm Scale Association Groves Trophy in 1983 and a Bronze Medal at the Model Engineering Exhibition in 1984.

 

More details of this locomotive to follow.

 

John

Edited November 21, 2018 by JohnBS

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 24

 

No. 4706 A Great Western 2-8-0

An essay in scratch-building to 2mm Scale, Part I

Based on two articles from "The 2mm Magazine" October & December 1985.

 

Perhaps like many others, I became attracted to N-gauge modelling by the scenic breadth that the scale offered - the opportunity to put a railway in its context. I immediately rushed in where angels fear to tread and became heavily committed to a large home layout, based loosely on the Totnes area of South Devon. Although I became a member of the 2mm Scale Association and I am most grateful for all the assistance that I have received, the prospect of scrapping all that proprietary track and rolling stock and beginning again in fine-scale was too daunting to contemplate. I therefore continue to model in N-gauge, albeit to 2mm scale. However, the techniques that I describe below would be equally applicable to 2mm fine-scale.

 

Over the years, a collection of suitable locomotives had been built up by the well-tried method of modifying commercial models. However, I was running out proprietary mechanisms with suitable dimensions and adequately reliable performance; in addition, the necessary modifications were becoming more extensive and time-consuming. The 2mm Scale Association's London Group locomotive building sessions provided the necessary impetus and guidance for me to take the plunge into scratch-building.

 

Prototype selection was relatively straightforward; I needed a large goods engine with considerable pulling power. The massive proportions of the 47xx Class 2-8-0s had long been attractive and no suitable commercial chassis was available - or likely. Therefore, despite the problems of beginning with an eight-coupled tender engine, the choice was made.

 

The Mechanism

Russell's well-known work "A Pictorial Record of Great Western Engines" provided the necessary prototype information for the preparation of a general arrangement drawing to 4mm scale (Figure 1). There was adequate space for the excellent Portescap 1219 coreless motor to be mounted in the tender, driving through a steel wire cardan shaft to a worm gear mounted in the firebox. Initial sketches showed that a 32:1 worm reduction and a 22:14 spur gear could be accommodated, giving an overall reduction of about 50:1. Also it became clear that the drive could equally be taken to the third or to the fourth axle, so I compromised and did literally that, driving both! The rationale for such a double drive was that it spreads the torque over two axles, reducing the stress on the gear-to-muff and muff-to-axle joints. Also, as the one-piece coupling rods would be held horizontal by the pairs of crankpins, each rod would impart some turning movement to the coupled wheels throughout each revolution. To my pleasure, this turned out to be correct, all the driven axles could be rotated quite sweetly with only one rod in place.

 

Chassis construction followed well-tried techniques perfected by the Association; metal wheels with integral half axles are fitted in insulated muffs and supported in frames separated from one another with insulated spacers. Clearly, this approach neatly avoids all the decisions of which form of wheel scraper pick-ups to adopt and allows the use of all wheels for electrical contact with no friction penalty.

 

The frames were made of 0.025" hard brass with 0.020" nickel-silver coupling rods, all sweated together into a four-layer sandwich with the rod blanks on top of the frames. Wheel centres were marked out on the coupling rod centrelines and drilled through all the layers to crankpin diameter - painstakingly in my case with a pin chuck. Then the pair of rod blanks were separated from the frames by a sharp end-on blow with a screwdriver while slightly softening the solder with a warm iron; somewhat nail-biting but it worked.

 

The rod and frame blanks were then marked out to profile; a black spirit-based marker pen rubbed over the metal prior to scribing greatly increased the visibility of the lines. The tender's inside frames were treated similarly and both set of frames were drilled out to 1.5mm diameter for the axles, fretted and filed to shape, leaving about 10mm extra material at each end to take 10 BA bolts through tubes as temporary spacers to aid assembly. Double-sided copper-clad fibreglass, cut to 5mm overall width and with the copper scribed through longitudinally to provide insulation, was used as frame spacers (note the restricted width between frames in N gauge; 2mm fine scale would give an additional 1.5mm). Initial assembly was by the bolts and tube spacers; when the axle alignment had been checked with 1.5mm rods through the holes, the copper clad fibreglass spacers were tack soldered in place, one by one, rechecking the alignment at each stage. On completion of the soldering, the frame extensions were cut of and the ends filed to shape.

 

The worm pinion and intermediate gear, both 0.125" bore to fit on an insulated muff (all obtained from the 2mm Scale Association) were checked for concentricity and then sweated together so that the torque was transmitted directly from gear to gear without relying on the bond to the muff. The latter was trimmed to 4.5mm length to allow some side-play and, with the gears superglued in place, two short lengths of 1.5mm dia steel were inserted through the frames and bonded into the muff to form an insulated idler axle.

 

I have always found quartering driving wheels to be a particularly awkward and tedious task and was not looking forward to the prospect of tackling four axles, two of which were, in addition, linked by gears.

 

Firstly four muffs were cut to 4.5m length and drilled mid-way at right angles to allow superglue to be injected. Next a gear was mounted at one end of each of the two muffs for the driven axles. The wheels, obtained from Mike Bryant Models, were prepared by removing any burrs or chuck marks from the axles - otherwise they act as very effective reamers - and by turning down the flanges slightly. Then two pairs were inserted into the third and fourth axle muffs (the ones fitted with the gears) and were quartered by eye, using slotted shim spacers marked with radial lines at right angles to aid sighting. The quartering was then "tweaked" first with one rod in place and then with both until the wheels ran smoothly and the rods did not work against the gears. Superglue was then coaxed into the muff holes with the aid of a pin to lock the quartering and the process was repeated for the second and first axles in turn. In all, it was a lot longer and more frustrating procedure to carry out than to write about.

 

Part II to follow.

 

John

Edited November 23, 2018 by JohnBS

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 25

 

 

No. 4706 A Great Western 2-8-0

An essay in scratch-building to 2mm Scale, Part II

Based on two articles from "The 2mm Magazine" October & December 1985.

 

The Mechanism (continued)

Having at last got to the position of a freewheeling chassis, attention moved to the completion of the drive transmission. Firstly the worm was reamed out to be an interference fit on to a length of 1/16" diameter shaft of pivot steel. A bearing of 3/32" outside diameter brass tube, readily obtainable from model shops and a good fit for the pivot steel shaft, was fabricated by soldering a single length of tube through two bent brass angle plates; These in turn were soldered to a stout baseplate of 1/16" nickel-silver cut to 6.0mm width and slotted to clear the worm pinion and idler gear (Figure 2). The baseplate was drilled and tapped 12BA at each end to take mounting screws through frame spacers from underneath. The centre length of the bearing tube was then out to receive the worm, thus ensuring a true alignment of the halves of the bearings.

 

The worm was located in place and its front face was supported on the vice jaws to enable the shaft to be driven in from the rear without distorting the assembly. Mesh with the worm pinion was adjusted by filing the bottom of the baseplate and the plate was insulated from the frames by a thin layer of self-adhesive tape. A coupling for the wire drive shaft was made by cutting a slot across the diameter of a short length of 3/32" tube and soldering it to the end of the worm shaft to form a keyed socket and then using a push-on sleeve of plastic tube to prevent the end of the wire shaft from slipping out sideways.

 

The motor was mounted on the tender chassis solely by using stiff wire leads from the split frames to the motor tags at the rear. The motor casing required a layer of self-adhesive tape underneath to prevent shorting across the tender frames or wheel flanges. The turned brass flywheel had an integral coupling socket for the wire drive shaft and, after checking for balance, it was superglued to the motor shaft; inject glue from the front only! The dumbbell-shape steel wire drive shaft completed the drive system. Bending the wire to the correct overall length (about 1mm shorter than the distance between the socket inside faces) was very much a matter of many trials and many errors, however the resulting rejects may come in useful one day for future models!

 

The physical connection between the locomotive and tender was by a drawbar soldered to the tender and attached to the locomotive chassis by a screw into a tapped plate on the rear spacer. The small amount of side-play necessary was obtained by filing off the threads near to the screw head, thereby reducing its diameter. Electrical connection was by a pair of phosphor-bronze wire springs soldered to the insides of the tender frames and shaped to rub against the insides of the locomotive frames opposite the drawbar screw. This method permits the easy separation of the locomotive and tender and avoids the use of fragile or stiff jump leads, without restricting rotational and pitching movements.

 

The pony truck was of brass sheet frames and brass milled tee-section bars on a copper-clad spacer. The pivot point was set out by using “Baldry's Rule" as in Roche's book “Building Model Locomotives". Surprise, surprise, Churchward got it right on the prototype!

 

When it came to the cylinders, I cheated and used a Graham Farish "Hall" set, modified by cutting out a section in the centre to reduce the overall width to scale and reinforcing the join with a plate of brass epoxied underneath. The slide bars were filed to the correct profile and holes were drilled in the block for the piston valves. Slide bar support frames were simply of brass tee-section, shaped and fixed to the footplate. Other details - drain cocks and snifting valves were then added.

 

The coupling rods were generally made of nickel-silver, detailed by sweating on pieces of oversize shim to form the bosses, drilling through from the back for the crankpins and then filing to shape. The fluted connecting rods were also fabricated from 0.020” nickel-silver and filed to the profile of the web in a pair. After separation, bosses were added as above and the flanges were formed by sweating on strips of nickel-silver shim, again oversize, and filing down to size in place. The crossheads were made from sheet material, again drilled and filed to shape and then soldered to piston rods of appropriate size steel pins. They were then fixed to pivot on the connecting rods on filed-down track pins soldered to a rear retaining plate of shim. Tissue paper between the crosshead and the connecting rod prevented the whole assembly becoming rigid when soldering the pins. The coupling rods were retained temporarily on the crankpins with collars of plastic insulation from electrical wiring until chassis fabrication, testing and painting was complete. They were finally fixed with soldered rings of 5amp fuse wire, again using tissue paper to prevent accidents! The bosses on the rods for the front crankpins were countersunk slightly and the pins filed back to improve crosshead clearances.

 

The model sports simulated working inside and outside valve gear, perhaps something of a pointless exercise on a locomotive with so little daylight under its massive boiler but a worthwhile experiment and not too difficult. Figure 2 shows the main details.

 

The inside fore and back rods were of nickel-silver wire; lengths were bent right round a 2mm diameter former until the shanks crossed over at about 30deg. The crossing point was reinforced by binding with 5amp fuse wire before soldering. Expansion links of nickel-silver sheet were drilled, slotted and cut to the radius shape and then soldered to the gear rod ends, one either side. The extension rods were shaped to profile but were made over-length so that their forward projections could slide backwards and forwards in grooves formed in a plate below the smokebox saddle.

 

They were then detailed and pivoted to the expansion links. To generate the required movement, a pair of eccentric grooves were simply filed into the second axle muff, set to the correct phasing. A section of the rear part of each gear rod loop was then snipped out so that the rods could be sprung into the grooves. A little tweaking and filing until a good running fit was achieved resulted in a simplified inside motion.

 

The outside gear operates independently; nickel-silver wire valve spindles were each flattened at one end and bent and soldered to form long closed loops. Fine phosphor-bronze wires were soldered to the backs of the crosshead retaining plates, projecting upwards to engage in these loops and sliding the valve spindles backwards and forwards at the ends of each piston stroke. The wires and rear sections of the loops were chemically blackened to reduce their visibility. The chassis was completed with the addition of brake gear, balance weights of black plastic card, tender water scoop, couplings and other details.

 

Part III to follow,

 

John

Edited November 23, 2018 by JohnBS

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 26

 

No 4706 A Great Western 2-8-0

An essay in scratch-building to 2mm scale. Part III

Based on two articles in "The 2mm Magazine" 1986

 

The Superstructure

The superstructure is where the story starts to diverge from normal scratch building practice. Having progressed over the years from fairly minor butchery of commercial models to wholesale rebuilds, I had acquired a degree of experience in modifying and building locomotive superstructures from plastic card.

 

Compared with sheet metal, the usual material, plastic card offers some considerable advantages - principally the ease and speed of cutting, shaping and joining. It is a material with which most modellers are familiar and can use with confidence; in all, it is "user friendly". Of course, the disadvantages are also fairly clear - limited strength and, particularly important to modellers in 2mm scale, low density. However it is capable of being used in thinner sections and with finer details than the average white metal kit and it takes oil-based paints better than any metal. Of course, with a prototype like the 47xx, we are not short of volume to fill with lead so adhesion weight should not be a problem.

 

Clearly, then, it is important to make the most of the material's attributes and to minimise its shortcomings; a literal translation of sheet metal techniques is likely to be only partially successful. It is vital to begin with basic structures of fairly stout material, typically 0.030" card, with adequate cross-bracing. On to these structural frames, the outer sheets and wrappers, usually 0.010", are fixed. The basic structures can be kept short of the exposed edges of assemblies such as cabs, tender sheets and bunker tops to avoid showing excessively thick edges. Wherever it is necessary to bond large areas of thin sheet on to these structural panels, small vent holes in the underlying material will allow the solvent adhesive to evaporate fully from the centre of the area without softening the thinner sheet. Otherwise irreparable distortions or blisters can develop, sometimes weeks later, as I have found out to my cost! For the same reason it is important that no part of the structure forms an unventilated closed cell.

 

Figure 3 shows the superstructure components. Black plastic card was used for parts liable to abrasion during handling to camouflage any wearing-away of the dark paint finishes; otherwise white card was used for ease of marking-out. Generally all the main structural components were of 0.030" card, the wrappers and plates of 0.010" and the beading of 0.010" rodding.

 

Part IV to follow

 

John

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 27

 

No 4706 A Great Western 2-8-0

An essay in scratch-building to 2mm scale. Part IV

Based on two articles in "The 2mm Magazine" 1986

 

The superstructure (continued)

The main structural element of any small scale model locomotive superstructure, except perhaps tank engines, is the smokebox/boiler/firebox assembly. This forms a rigid cylinder of considerable strength, a backbone which supports footplate, cab and ancillary components. Figure 4 shows the construction of this element and I hope it is largely self-explanatory.

 

To retain sufficient flexibility for rolling and to give adequate strength, a double lamination of 0.010" plastic card was used for the smokebox and boiler, formed "swiss-roll" fashion. The developed shapes of the wrappers were derived from the general arrangement drawing using schoolboy mathematics - and a pocket calculator. An allowance for the material thickness was made to the inner layer circumference, effectively a reduction of about 1.5mm from that of the outer layer for 0.010" material, irrespective of the diameter.

 

The inner layer of the smokebox wrapper projected some 3mm into the front of the coned boiler section to act as a locating plug, the inner layer of the boiler wrapper being cut back by a similar amount. I used old felt pen barrels of suitable diameter as rollers to coax the wrappers into the right curvature and to act as temporary internal formers to hold the sections circular while the adhesive was evaporating and the plastic was hardening - allow at least 24 hours!

 

The smokebox and boiler sections were filled with rolled lead sheet ballast, carefully shaped so as neither to rattle nor to distort the plastic. I did not want to use any adhesive to avoid the risk of it softening the wrappers. The circular permanent formers, all the internal ones drilled with vent holes, could then be trimmed and fitted. Finally the front and rear planes of the coned boiler were trued up to be perpendicular to the boiler base line and the boiler and smokebox were united.

 

If a coned boiler can be a bit tricky to get right, then a Belpaire firebox can be the very devil. The geometry is very complex with subtle rearward taper both in plan and in elevation, springing from a parallel-sided base, intersecting curves on the top and front edges and varying curvature to the waist. A careful study of detailed drawings and photographs repays the effort and makes one wonder how they managed to build the prototype ones - and make them work as well!

 

Figure 4 also shows the main components of the model firebox. The top edges of the structural core were rounded before applying the wrapper which was left over-length for trimming to the footplate, in place. After joining to the boiler's rear former and checking for square and level, final shaping was left for a couple of days to allow the adhesive to evaporate and the plastic to harden.

 

The footplate was relatively straightforward; the main and dropped front end plates were formed from plastic card the thickness of the valance, linked with curved sections and topped with 0.010" black card as the footplate decking. The buffer and drawbar beams were added and the smokebox saddle was made up from two laminations of 0.060" plastic card, filed to shape. The boiler assembly was then mounted, checking carefully for fit, level and parallel. Finally, the boiler support brackets were inserted and the splashers formed from 0.020" discs with black card tops.

 

A 0.020" transparent plastic card spectacle plate and 0.030" plastic card sides and floor, scribed to represent planking, formed the main cab structure. The black card roof was supported on an inner former, set in from the rear edge. The lower parts of the inner, or structural, side-sheets were projected below the footplate level to form the rear dropped valances; the visible front and side sheets were of 0.010". Finally the steps, of black card reinforced with brackets, were added below the cab footplate.

 

The Tender

The tender structure was simple enough in concept, a box with open bottom and with a recess between the tanks for the coal space – or perhaps more precisely for the motor space! However, appearances can be deceptive; the difficult part was to ensure that the planes of the box were set up square and parallel and stayed that way when the adhesive had dried. Small inaccuracies are very noticeable in such a simple rectangular shape.

 

The plates in the coal space and all the components below footplate level were in black card. The tool boxes, dome and tank filler were laminated and filed to shape before fixing while the main tender wrapper was cut slightly oversize and trimmed down after fixing. The upper coal sheets and the rear flare were spaced on 0.015" sheet and were also filed to finished shape in place.

 

The Details

Generally the details and minor assemblies were of plastic card - principally "lumps" were of thick card, laminated where necessary, plates were of thin card and attached pipes and beading were of appropriate diameter plastic rodding. The backplate, gear lever housing and other cab details were made separately and painted before fixing. Boiler bands and the reinforcing angle around the tender tank base were of self-adhesive tape - the brown sort used for packing, not sellotape which has a tendency to fall off after a few months.

 

The smaller details were generally of metal - phosphor-bronze wire for handrails with 5amp fuse wire stanchions fixed with superglue, filed down dressmaking pins for tank vents, bent wire and solder for brake and water scoop standards, soldered wire for the smokebox dart and cylinder drain-cocks and nickel-silver shim for the gear rod (operating) and the sanding gear rod (dummy !). The chimney, safety valve casing and whistles were of brass, turned in a mini-drill; no need to use phosphor-bronze for the former as No 4706 did not boast of a copper capuchon! 2mm Scale Association tender axle-box castings and proprietary turned buffers and wound wire vacuum hoses completed the details.

 

Painting and Weathering

The paint scheme for No 4706 was suitably austere - luckily; no lining and very little bright metal to worry about. The chassis was painted before the final fixing of the side rods. I used scouring powder and an old toothbrush for cleaning the metal, then, after washing down thoroughly, the wheel treads and flanges were masked with two coats of dilute PVA glue and the gears masked with tape. The chassis was then primed grey and finished matt black using aerosol car paints. On the superstructure the only areas which required masking were the spectacle plate windows (inside and outside) before spraying the whole with Precision Paint's pre-1928 green (I like the colour better than the later green!) and brush-painting the black areas, the red buffer beams and not forgetting the red of the inside frames and motion.

 

Before their final fixing, the coupling and connecting rods were chemically blackened, lightly burnished with a fibre-glass brush and then smeared with a film of Teflon grease; this process gives a realistic and permanent impression of slightly rusted oily metal.

 

Weathering was concentrated on the wheels, chassis, other areas below footplate level and on the tender interior, using dilute black and rust oil paints and finished off by dry brushing with lighter shades. After applying Methfix transfers and etched number plates, the whole superstructure was sprayed with semi-matt varnish as a final finish. The cab fittings, crew and tender coal load completed the model.

 

The concluding part to come.

 

John

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 28

 

No 4706 A Great Western 2-8-0

An essay in scratch-building to 2mm scale

Based on two articles in "The 2mm Magazine" 1986

 

Conclusion

Now came the proof of the pudding. On the tracks No 4706 pulled as well as I had hoped - forty plus wagons up the 1 in 25 gradients on the 450mm radius hidden curves of my old "Totnes" layout were no problem (the current "Totnes" layout has no gradients). Low speed running was equally satisfying, a scale 5 mph could be sustained through hell or high water. Still, with 14 wheels to choose from, current collection should not have been a major problem and the superb performance of the Portescap motor at low voltages is well known. At a scale 50 mph, the model's maximum speed was perhaps a little on the slow side for use on an excursion passenger link but it was fine for the more common fitted goods traffic. The rubbing electrical contact between locomotive and tender has showed no sign of causing current connection problems but I wouldn’t like to try it without tender pick-ups. Inevitably, the long coupled wheelbase, without compensation, (food for thought in 2mm scale) gave rise to some slipping at any sharp changes of vertical alignment, but, as noted above, this is no longer a problem.

 

Well, with the benefit of hindsight, what would I have done differently? As it turns out, relatively little except for a few significant points.

Firstly, a keeper plate to allow the removal of the driving axles would have been invaluable for wheel quartering, worm meshing and painting and would have reduced the risk of disturbing the quartering during later stages of the chassis construction. Even the relatively low temperature reached by an axle when soldering nearby areas of the frames can be sufficient to break the bond of the superglue in the muff.

 

Secondly a technique used by the late Denys Brownlee would have simplified the installation of the idler gear axle. He used insulated bushes of mica-filled PTFE, drilled and turned in the lathe and mounted in the frames to support simple steel idler shafts. This avoids the need for, either a split axle and muff-mounted gear, both increasing the risk of eccentricity, or the construction of a complex subsidiary frame, insulated from the main frames. Perhaps I only got away with it because 64DP gears are reasonably tolerant of poor meshing and a degree of eccentricity.

 

Finally, when the locomotive was in for a recent major service and clean, I followed a method used by John Greenwood which avoids the ever-present risk of losing the cardan shaft when lifting the loco. I anchored the front end of the shaft by drilling a transverse hole through the worm shaft sleeve and formed the front loop of the shaft to fit through this, thus holding the shaft captive at one end, while still allowing the necessary movement.

 

Well, that's the end of this story; on to another scratch-built tender locomotive in the next posting.

 

John

 

© John Birkett-Smith

March 1985, November 2018

Edited November 29, 2018 by JohnBS

[Donw]

A veryuseful set of posts John

 

 

Don

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 29

 

No 3824 “County of Cornwall” c.1922

A pretty engine, though its short coupled wheelbase and outside cylinders made for a rough ride.

Built 1911 Withdrawn 1931

 

 

The model is entirely scratch-built, using methods similar to the previously described No 4706 - i.e. with a split-framed metal chassis and a plastic card superstructure.

 

Power is again provided by a tender-mounted Portescap 1219 motor with a small flywheel and a simple bent wire cardan shaft to a worm gear in the engine firebox. This feeds a double-reduction gear drive, giving an overall reduction of 37:1. Both engine and tender have split frames with wheels from the late Mike Bryant range, all wheels collecting current. Transfer between the engine and tender is by phosphor-bronze springs, soldered to the tender frames and bearing on the insides of the engine frames. This permits the easy separation of the loco and tender.

 

The engine and tender superstructures are of plastic card with metal details and fitting and rolled lead sheet ballast in the firebox and rear of the boiler, taking care to ensure that the centre of gravity falls within the coupled wheelbase.

 

The combined weight of the engine and tender is 84g.

 

This model won the 2mm Scale Association Groves Trophy in 1984.

 

More scratch-built tender locomotives to come.

 

John

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 30

 

 

No 3373 “River Plym” Bulldog class 4-4-0 c.1930

A vintage engine with beautiful proportions.

Built 1903 Withdrawn 1948

 

Again, the model is entirely scratch-built, but this time, all in metal.

 

Similarly, power is supplied by a tender-mounted Portescap 1219 motor, with a small flywheel and a simple bent wire cardan shaft to a worm gear in the engine firebox. This feeds a double-reduction gear drive, giving an overall reduction of 37:1. The geared drive was taken to both the coupled axles as I didn’t want to risk over stressing the cranks on the extended axles. The crank pins were silver soldered to the webs by Tim W so that wasn’t a problem but the webs were only soft soldered to the axles. (The 2mm Scale Association depthing tool was invaluable in setting the gear meshing distances for this.)

 

Both engine and tender have split frames with wheels from the late Mike Bryant range, including extended-axle drivers, all wheels collecting current. Transfer between the engine and tender is by phosphor-bronze springs, soldered to the tender frames and bearing on the insides of the engine frames. To avoid excessive sideways movement on curved track, the rear axle of the bogie is in the main frames, with some additional side-play, and the front axle is in a separate pony truck. In essence it is built as a 2-2-4-0. The bogie outside frame is supported at the rear by a pivot on the main chassis and, at the front, a pin-in-slot on the pony truck.

Above. A general arrangement drawing with dimensioned larger scaled details of boiler fittings - invaluable when turning and finishing.

 

The engine and tender superstructures, details and fittings are of sheet and turned metal, with rolled lead sheet ballast.

 

The combined weight of the engine and tender is 95g.

 

This model won the 2mm Scale Association Members' Choice Cup.

 

A couple more (largely) scratch-built tender locos still to come.

 

John

Edited December 5, 2018 by JohnBS

[queensquare]

ASHBURTON AND TOTNES

 

Locomotives 30

 

8759D92A-669D-4B68-ACCD-E1E60033A90C.jpeg

 

No 3373 “River Plym” Bulldog class 4-4-0 c.1930

A vintage engine with beautiful proportions.

Built 1903 Withdrawn 1948

 

Again, the model is entirely scratch-built, but this time, all in metal.

 

John

 

The Dean goods and 517 run it close but this is my favourite in the stud - as John says, a vintage engine with beautiful proportions which are captured to perfection.

 

Jerry

[Donw]

ASHBURTON AND TOTNES

 

Locomotives 30

 

8759D92A-669D-4B68-ACCD-E1E60033A90C.jpeg

 

No 3373 “River Plym” Bulldog class 4-4-0 c.1930

A vintage engine with beautiful proportions.

Built 1903 Withdrawn 1948

 

Again, the model is entirely scratch-built, but this time, all in metal.

 

Similarly, power is supplied by a tender-mounted Portescap 1219 motor, with a small flywheel and a simple bent wire cardan shaft to a worm gear in the engine firebox. This feeds a double-reduction gear drive, giving an overall reduction of 37:1. Both engine and tender have split frames with wheels from the late Mike Bryant range, including extended-axle drivers, all wheels collecting current. Transfer between the engine and tender is by phosphor-bronze springs, soldered to the tender frames and bearing on the insides of the engine frames. To avoid excessive sideways movement on curved track, the rear axle of the bogie is in the main frames, with some additional side-play, and the front axle is in a separate pony truck. In essence it is built as a 2-2-4-0. The bogie outside frame is supported at the rear by a pivot on the main chassis and, at the front, a pin-in-slot on the pony truck.

1565DB26-92F3-4F4B-9C17-0BBC498BAA74.jpeg

Above. A general arrangement drawing with dimensioned larger scaled details of boiler fittings - invaluable when turning and finishing.

 

The engine and tender superstructures, details and fittings are of sheet and turned metal, with rolled lead sheet ballast.

 

The combined weight of the engine and tender is 95g.

 

This model won the 2mm Scale Association Members' Choice Cup.

 

A couple more (largely) scratch-built tender locos still to come.

 

John

 

 

I saw this at the 2mm AGM suberb piece of modelling.

 

Don

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 31

 

No 111 “The Great Bear” c.1922

The Great Western’s only 4-6-2 ‘Pacific’, for its time a huge locomotive.

Built 1908 Withdrawn 1924 though the frames were re-built as a 'Castle'

 

As far as I can determine, the prototype never made it to South Devon, though it is reported that it travelled as far as Newton Abbot. However, of course, Rule 1 applies.

 

The model is based on a Fleischmann powered tender drive, selected for the approximately correct wheelbase and wheel diameter. Originally, the locomotive used the complete (butchered) Fleischmann chassis but the wiper pick-ups proved somewhat delicate and my phosphor-bronze alternatives tended to generate too much friction. From time to time, "The Great Bear" became "The Great Sledge".

 

So an alternative approach was necessary. Now, the unpowered engine chassis is of scratch-built split-frame construction with wheels from the late Mike Bryant range, on half-axles with insulating muffs – the typical 2mm Scale Association method - so no wiper pick-ups. Current collection was transferred to the tender. Connecting and coupling rods, slide bars and crossheads were scratch-built in nickel-silver (scrap Peco track).

 

The engine and tender superstructures are entirely scratch-built in plastic card, as described previously for No 4706, with metal details and fittings and a little rolled lead ballast in the engine.

 

The combined weight of the engine and tender is a fairly modest 82g but, with traction tyres on the tender, it can cope quite well with seven or eight coaches.

 

When Churchward was told that Gresley was building an Atlantic, he reportedly said that, "If he had asked, he could have had ours."

 

One more locomotive still to come.

 

John

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 32

 

No 2937 “Clevedon Court” c.1930

A curved frame "Saint"

Built 1911 Withdrawn 1953

 

The model is a first essay in 3-D printing. I drew out the basic design in 2-D CAD and my son then converted the superstructure and tender in Solidworks for printing by Shapeways.

 

The chassis is a fairly conventional scratch-built split-frame type with a Nigel Lawton motor mounted in the tender, with a copper tungsten flywheel, connected by a wire cardan shaft to the two-stage reduction gear (47:1) driving the centre axle. Driving wheels were from the late Mike Bryant range, similar to the 2mm Scale Association type but with N-gauge tyres. The bogie was from the N-Brass range and the rods and cross-heads were scratch-built.

 

New boiler fittings, metal steps, handrails and other details were added. Current transfer between the engine and tender was by my usual method - phosphor-bronze springs soldered to the tender frames and bearing on the insides of the engine frames. The tender chassis again has MB wheels and metal steps and other details and a load of real coal bonded to a sheet lead bed as ballast.

 

The combined weight of thr engine and tender is a respectable 88g. Thanks to the coreless motor and the flywheel, the loco runs on for about ten centimetres from powering-off at a normal running speed - lovely !

 

More details of the loco construction to follow.

 

John

[JohnBS]

ASHBURTON AND TOTNES

 

Locomotives 33

 

No 2937 “Clevedon Court” c.1930

A curved frame "Saint"

Built 1911 Withdrawn 1953

 

Four more photos of the Saint at different stages.

 

Firstly, the 2D CAD drawing which was the basis of everything. The top elevation is a simplified scale drawing of the prototype. In the centre, the drawing is a 2mm scale version of the model and at the bottom, and at the sides are the tender and various details.

 

Then, a photo of the scratch-built chassis. Typical 2mm Scale Association methods but with N-gauge wheels. The split frames (20 thou brass) were spaced with double-sided copper-clad fibreglass, gapped for insulation, and the wheels were on half axles and fit into insulated muffs. I like to be able to remove the driving wheels so I made a keeper plate, complete with cosmetic springs, using the same technique and this was bolted to the main frames. The Mashima motor shown here was later swapped with a Nigel Lawton 1012 one, with flywheel and with excellent results.

 

The third photo shows the loco in grey primer and almost complete except for connecting rods, cross-heads and sundry details. The 3D printing includes the loco superstructure, with loose cylinders and firebox back-head. The separate printing of the 3500 gallon tender comes complete with frames and axle boxes. Loco and tender steps were scratch-built in nickel-silver as, in my view, they would be too vulnerable in printed material.

 

The final photo is of the right hand side of the finished loco The loco and tender are available from Shapeways in 2mm FS (1:152) and N-gauge (1:148), under designer SteamPrint nbs3000. A "straight frame" Saint is also available.

 

Best wishes,

 

John

[richbrummitt]

John,

 

Would you be kind enough to show elaborate on the gear assembly and show some more angles on the worm shaft mounting please.

 

Thank you,

 

Richard.

[JohnBS]

John,Would you be kind enough to show elaborate on the gear assembly and show some more angles on the worm shaft mounting please.Thank you,Richard.

Richard,

I’ll take the Saint apart and photograph the worm mounting in the next day or two. At the moment I am without my phone - waiting for a new SIM card.

Basically, the mounting is the same as for No 4706, described earlier, but mounted on to one of the side frames, not on to the frame spacer. This will only work if you have fairly hefty side frames i.e. 20 thou.

John

[JohnBS]

ASHBURTON AND TOTNES

 

Richard,

Surprising how the memory plays tricks ! The photo in the earlier post shows how I originally made the chassis but I totally forgot that it was subsequently changed. It is now almost identical to the arrangement for No 4706, but with angles rather than tubes.

Above. Side view showing the angles forming a "shoe" to support the worm and shaft.

Two small angles were bent out of 20 thou brass and sweated to a brass plate. The angles were set apart by the distance of the length of the worm, then were drilled 1.5mm dia to take the the worm shaft. A pair of frame spacer of double-sided copper-clad fibreglass were added between the frames, double gapped to provide an electrically isolated centre section and spaced to clear the worm and gears.

Above. Front view, showing the double-gapped frame spacer and the worm "shoe".

The spacers were then drilled centrally, the angles and "shoe" assembly offered up and drilled through to take a 14BA tap and the whole test fitted. When all was OK, the centre part of the plate was cut away and the worm meshing checked and adjusted by filing the plate or adding shims.

 

I hope that this makes sense.

 

John

Edited December 11, 2018 by JohnBS

[JohnBS]

ASHBURTON AND TOTNES

 

Sorry folks, I can’t seem to add photos to this post - I will try again later.

 

John

Edited December 13, 2018 by JohnBS

[JohnBS]

ASHBURTON AND TOTNES

 

Buildings 1

 

Illustration 1 – General view of the Totnes model - “The Plains”, (quayside), town and castle

 

"Horses for courses"

 

This section of the thread is based on an article in issue 256 of the Model Railway Journal and is on how I made the buildings for the Ashburton and Totnes layouts. As such, I describe some of the methods that I have used but it is not intended to be prescriptive. Of course, much of the following will be the statement of the obvious but some aspects may be of use in the smaller scales – 2mm, 3mm and 4mm.

 

I begin from the viewpoint (literal and conceptual) of where the buildings are in relation to the viewers of the layout. I can divide locations into three categories :

Foreground or "Specimen" buildings. This is fairly self-explanatory - buildings that are in front of the tracks or that you want to draw particular attention to. These are modelled in relative detail.

The middle ground - beyond the tracks. Here, I start to reduce scale - subtly - or, if appropriate, omit a bay from a multi-bay structure. I take a few more liberties with details and simplify where possible, using paint rather than adding components to suggest details.

Background buildings and beyond. For structures adjoining the backscene, this approach is taken further; they also may benefit from scale reduction and selective compression.

 

Modelling materials

Throughout, I use the following materials.

Mount Board - available in a very wide range of colours and differing thicknesses, usually measured in microns (μ), where 1000 microns = 1 mm. Typical thicknesses are 1200 μ, 1350 μ, 1500 μ, 2000 μ and 3500 μ. Sheet sizes are from A4 to A0. Framing shops often have loads of small and medium-sizes off-cuts which they are happy to dispose of for a nominal payment or for a donation to their favourite charity!

Cereal packets – free. Typical thickness 500 μ.

Card - again available in a wide range of colours; weight usually measured in grams per sq. metre (gsm) 180 gsm is approximately 210 μ and 240 gsm is approximately 280 μ. 180 gsm stock is about the maximum thickness that can be persuaded to go through most printers.

Plain and coloured paper - copier paper 80 gsm is approximately 90 μ and 100 gsm is 110 μ.

Printed paper (brick, stone, etc) - A very wide range is available from on-line sites such as Scalescenes. It is worth having these printed at a print shop with a colour laser printer to avoid smudging ink with water-based adhesives and paints.

Acetate sheet - various thicknesses

Transparent plastic card

Plastic card - plain and textured, especially useful for timber boarding and for some roofing materials

Adhesives - PVA, paper adhesive (Pritt-Stick), impact adhesive (Evostik), (Uhu), plastic solvent adhesive (Mekpak)

Modelling clay/epoxy – DAS, Milliput or similar

Paints - originally, I used enamels but now I often use emulsion (test pots) for undercoats and artists' acrylics for finishing

Drawing pens - Rotring Rapidograph or similar - expensive but good. Sizes 0.13mm and 0.30mm will probably suffice.

Detailing components - plastic and metal sections, proprietary etches, white metal castings as required.

I am not going to say much about tools - they are pretty much what you would expect to see on a modeller's workbench. However, a small drawing board with a parallel motion and an adjustable set square are very useful.

 

References

Before starting, I gather reference material. If the building is still in existence, your own photographs or on-line sources are all that is needed, otherwise a bit of historical research is necessary. Again, on-line sources are a good starting point, also specialist associations and your local library.

 

Illustration 2 - The Royal Seven Stars, Totnes – hand tinted postcard c 1890.

 

From the data, I prepare a drawing. If the building is simple and straight-forward, this can be done directly on to the structural card but for more complex constructions, it is worth preparing a separate drawing. If you have access to a CAD program, then use it - there are many fairly simple ones available as free downloads.

 

It helps sometimes to be aware that older buildings change during their lifespan; they can be a bit like Grandfather's Axe - two new heads and three new handles. Locally, (North Somerset) I know of buildings which started life in the 1690s as modest terraced cottages; a ground floor and a low first floor with gables facing the street and a door and three stone-mullion casement windows. Then the facade was raised to full two stories, reusing the original stone roof slates as walling and obliterating the gables. A new clay-tiled pitched roof parallel to the facade and two dormers to light the new attic rooms were added. Then, in the late 18th century, along came the fashionable sash windows (the Georgian equivalent to uPVC glazing) so out went the casements. At the same time, the facade was raised with a parapet and cornice, half hiding the dormers. Finally, in the Victorian era, the front living room was converted into a shop, with a new door, display window and lift-out shutters. Hardly a stone was left unturned.

 

Building structure

In all buildings, I start with a core structure of Mount Board (usually 1200-1500 μ, depending on the depth required for door and window reveals). I found that board was a very familiar and un-daunting material, easy to mark-out and cut, cheap and readily available. It will take a wide range of finishes and adhesives and is sufficiently durable and robust in 2mm scale. I add dividing walls as required for structural or detailing purposes and allow about 5mm below outside finished ground level to plug the building in to the layout. Cut out all window, door and other openings and make at least one floor, preferably at roof level, to allow for easier access to insert windows and doors. Make up a roof structure of gables, intermediate rafter and hip formers and roof slopes, again of card with overhangs at eaves and verges, to suit the prototype. Assemble it all with PVA glue and see if it looks right on the layout.

 

When all is in order, paint everything inside and out with a dilute coat of PVA (one part glue, two parts water) to reinforce and stabilise and to provide a base for finishes.

Then it's time to apply finishes. For brick or stone paper, cut roughly to size then paint the back of the paper with dilute PVA glue and apply. When wet, the paper will stretch slightly and then shrink back when dry to give a nice flat finish. With brick paper, be careful to get the courses level and wrap around corners and make joins were least conspicuous. Cut down the centre of each of the door and window openings and at head and sill and fold back the sides to form the reveals and paste to the inside of the walls. If the building has a projecting plinth, make this up out of appropriate thickness card and cover with paper before fixing. If you are using inkjet-printed papers, use Pritt-Stick.

 

If the final finish is to be painted (e.g. stucco, render) I usually add a layer of plain paper in the same way as the printed paper to ensure that any open joints are covered.

 

More to come,

 

John

Edited December 14, 2018 by JohnBS

[JohnBS]

ASHBURTON AND TOTNES

 

Buildings 2

 

Illustration 3 - A group of buildings with various roof finishes. A lot of the work is done by simple painting, such as quickly masking the ends of corrugated sheets, then over-painting the next layer.

 

Foreground buildings - windows and doors

The next stage is to prepare the windows and doors. For windows (and glazed doors), I draw them out on paper with at least 2 mm margin all round the opening size, then temporarily stick a piece of acetate sheet over and scribe the glazing bars and perimeter with a suitable point (compass point, old gramophone needle in pin vice or similar) and then apply a layer of paint. When the paint is almost dry, wipe off with a tissue and then paint the window frame margins. For sash windows, I don't make separate panes for the upper and lower parts (unnecessary in 2mm scale) but merely scribe a broader line for the bottom of the upper sash.

 

For timber doors, again I generally just use paint on card to indicate any relief but sometimes it is worth ruling basic detail on the card with a fine ball-point pen before painting. This is a good way of simply providing a subtle texture of panels or boarding. Then it's time to cut out the windows, complete with margins for sticking. I use Uhu or a similar colourless adhesive to stick them to the inside of the walls.

 

Foreground buildings - slate roofs

The next stage is to work on the roof. Because models are often seen from above, the appearance of the roof is of real significance.

 

As my prototypes are set in Devon, most roofs are slate. In Totnes, the main local source of roofing and wall-hanging slate was Harbertonford Quarry. These slates, a very attractive silvery-blue colour, are usually slightly thicker than Welsh slates but are no longer available and are often replaced with Welsh (or Spanish!) slate.

 

Typically, slates are about 1/4" (6mm) thick, which equates to 0.04mm (40 μ) in 2mm scale, about half the thickness of normal copier paper, so be careful! First, I add paper strips for valley and abutment gutters. Then I use overlapping paper strips for the slates. It makes life simpler to use 1mmx1mm squared printed graph paper, available in A4 pads - it saves a lot of measuring. I rule lines with a fine ball-point pen across the sheet at 2mm centres - this gives a subtle texture to the side joints between slates. Then I cut the sheet into strips at right angles at 3mm centres. Using dilute PVA and starting at the eaves with a 1.5mm strip to tilt the first course to the correct angle, I continue with 3mm strips overlapping by 1.5mm, each offset horizontally from the row below. I work-up opposing roof slopes in step and check progress towards the ridge.

 

When a pair of roof slopes are finished and dry, I carefully cut the excess strip lengths off at the verges or hips. On completion of all the roof slopes, I add paper strips as ridge tiles or narrower ones as cement hip cappings.

 

Illustration 4 - two buildings with Welsh slate roofs and one with Double Roman clay tiles

 

Foreground buildings - other roofs

I have little cause to model tiled roofs, however, on Totnes, I have one building with a Double Roman tile roof. For this, I used 4mm scale Corrugated Iron plastic card, cut into strips across the corrugations and overlapping very slightly. These were stuck down with Evostik and then carefully washed with a light coat of solvent adhesive to bond everything together. Valley, hip and ridge details are as for slate roofs.

 

For "flat" roofs in lead, I form a base out of board with card steps, add ridges parallel to the fall (lead "rolls") with paper strips (or sometimes sewing cotton) then cover with tissue paper.

For felted roofs, do away with the ridges and steps but lay the tissue paper in horizontal courses about 6mm wide, starting at the eaves and with a very narrow overlap.

 

For corrugated iron or asbestos roofs, I normally use card, ruled out with a fine ball-point pen to indicate the corrugations grooves. The grooves in most bought plastic materials are too coarse for 2mm scale.

 

For glazed roofs (station canopies, greenhouses) I use transparent plastic card, typically 500mu (20 thou) thick. Scribe and paint as for widows and overlap at the purlin centres. Plastic card is normally tougher than acetate and the slight lack of transparency is acceptable as the prototype examples rapidly got pretty grubby.

 

Finally, for thatch I make the base of board, carefully shaped to conform to the curves of the prototype and with an adequate overhang (gutters are not normally used on thatched roofs). This is then covered with Surgical Lint BPC (more difficult to get nowadays but available on line), stuck with Evostik and brushed downwards with an old toothbrush. Trim eaves and dormers carefully to give a crisp drip to the edges and paint varying shades of grey/green/brown. I find that diluted matt enamels are the best for this as they soak into the fibres - acrylics tend to sit on top.

 

Foreground buildings - details

Finally details are added. I add thin card strips for wall copings and stone or timber window sills but I usually just paint lintels (stone, timber, brick or concrete) and any dressings or quoins. Also, I fix card strips for eaves and verge boards, as the prototype, black card strips for gutters, stuck edge on, and any other relief detail. My criterion is "Add it if it casts a shadow".

 

Some details, like railings and balcony balustrades are best made from etches, of which a wide range is available. For rainwater, drain and soil pipes, I use lengths of copper wire from old mains electric cable, stretched to straighten, then bent into swan-necks and shoes as required and stuck with Evostik. Chimney stacks are made from board, as the main building, with cast white metal or shaped wood chimney pots. Many details (plaques, brackets, hoods and finials) are from scribed, carved or filed plastic card.

 

More next time.

 

John

Edited December 15, 2018 by JohnBS

[JohnBS]

ASHBURTON AND TOTNES

 

Buildings 3

 

Illustration 5 – The Royal Seven Stars – CAD drawn elevations

 

Background buildings - walls and windows

I cheat where I can. Again I start with the drawing, but this time producing a paper copy, or preferably a CAD print, in ink. (See illustration 5 above). The drawing is of outlines of all the surfaces of the building (floors, wall elevations, internal structure and roof surfaces) with doors and windows. I fill all the glazing in solid black or dark grey, leaving the frames and glazing bars white. I then paste the paper drawing on to the Mount Board to be used for the structure and cut out the panels. I do not cut out the windows. If the frames and glazing bars are other than white, I give them a wash of dilute acrylic colour or print on coloured paper. Then I cover each window with a patch of shiny transparent film - this gives a gleam to the glass. Next comes a set of overlays - card or board, depending on the window reveal depth required - and finally any finishing layers of printed or plain paper, grooved where required with the ubiquitous fine ball-point pen.

 

Roofs and details

For slated roofs, I merely rule plain or graph paper with a fine ball-point pen at 1.5mm horizontal centres, then paste the paper to the roof slopes and trim when dry. After adding ridge and hip cappings, l rely on the painting to express the detail.

 

Detailing is as minimal as can get away with - verge boards and gutters, chimney stacks, etc. I make chimney pots out of shaped sections of cocktail sticks. I don't bother with rainwater pipes and other plumbing unless they are particularly prominent.

 

Illustration 6 – The Royal Seven Stars - model. The first floor of the entrance porch was constructed around a cut-down Peco wagon box, which had the correct curved corners.

 

One of the virtues of CAD is the ability to copy and paste repeats of windows and other details, even borrow from earlier drawings, and that details can simply drawn which will show through translucent paint washes. By using Layer commands, I can draw each of the card/paper details on top of each other, then selectively print them. Also, it's simple to add lettering and sign details direct to the wall surfaces.

 

Painting

For any surface not covered with pre-printed paper, I usually paint an undercoat of dilute emulsion paint (test pots). This gives a degree of texture. Then a coat of body colour, often emulsion tinted with artists' acrylics. Beware of over-assertive colours, particularly in the background - tints of the earth colours (ochre, sienna and umber) seem to work well and white is best let down with a tint of grey. Greys are best made with a combination of Ultramarine and Light Red, perhaps with a little Hookers Green - black and white gives a very dead, artificial result. Lighten greys with white as required.

 

For slate roofs, after an overall coat of grey body colour, I mix small amounts of slightly different versions and pick-out random individual slates - say 10% of the total - "variations on a theme". I don't strive to be super precise on this as much will be softened by the subsequent weathering.

 

Weathering

Finally, this is where the story really starts (see Illustration 6). It is amazing how much detail can be implied with judicious use of the paintbrush. First, wherever possible, carefully look at the original - building materials weather in different ways. Walls beneath projections are sometimes "as black as your hat", especially in Millstone Grit and Portland Stone before about 1960. However almost all weathering is generated by rainfall and, in general, striations are vertical. Horizontal surfaces (parapet copings, tops of window sills) are usually well washed and are therefore bright. Areas below lead or copper dressings can have particular colours because of chemical reactions. The bases of walls are the exception to the vertical nature of staining. Here the spattering of rain on the ground produces a parallel band of thrown-up dust or of lichen growth, depending on the nature of the surface.

 

Roofs are a bit more problematic because they are relatively difficult to observe, though high points such as church towers or tall buildings provide improved viewpoints. Binoculars or cameras on "selfie" sticks can sometimes help.

 

Slate colours, even from the same source, can vary subtly. In rural areas with clean air, lichen slowly grows but metal flashings around chimneys and other abutments tend to inhibit this. Bird droppings concentrate at favoured landing points, on chimney stacks and at the ends of ridges, and promote the growth of lichen.

 

So how do we paint all this. For walls, I mix a colour to more or less match the body colour, then tint it by adding white and shade it by adding (usually) Hookers Green and Burnt Siena. The darker shade is then applied in very diluted streaks below projections, etc and the process repeated until satisfied. Then the lighter tint is applied, usually using the dry-brush technique, downwards on to projecting ares.

For roofs, the same basic process with dilute but unmixed Cadmium Orange and Sap Green applied to a pre-wetted roof slope to represent lichen and a pale grey-green dotted on to indicated bird droppings. All is then finished off with rust and other staining from railings, chimneys and flashings.

 

Illustration 7 – Drone’s eye view of buildings and back-scene; compare with illustration 1. Note the compression of the buildings and streets at the back (top) of the photo and the flat cut-out of the buildings beyond.

 

Buildings and back-scenes

For buildings right at the back of the layout, I take this approach a step further (see Illustration 7 above). Scale is reduced to say 1:200 and the buildings are compressed in plan in the direction at right angles to the back-scene. For example, a building with a 2mm scale footprint of 80x60mm would be "squeezed" to something like 60x30mm. Of course, the "squeezed" elevations of the buildings must not be easily visible - roofs at 70° pitch do not look convincing - so the selection of prototypes and their grouping is of importance. Therefore obscure any end-on views parallel to the back-scene with view blockers (in my case, Totnes Castle keep and motte). I then finish this off with a layer of thin MDF directly in front of the back-scene, cut out to the profile of more distant buildings and painted.

 

A bit more to come.

 

John

Edited December 18, 2018 by JohnBS

[JohnBS]

ASHBURTON AND TOTNES

 

Buildings 4

 

Illustration 8. Model of a Regency Terrace, Totnes.

 

USEFUL DATA

Note: where relevant, dimensions are given as follows -

Actual metric sizes in millimetres - no brackets

2mm scale sizes in millimetres - square brackets [ ]

Imperial sizes - curly brackets/braces { }

 

BRICKS:

Modern - post 1960 h=65, w=103, l=215 with 10 mortar joints [0.43 x 0.68 x 1.41]

(Old - Victorian and up to 1960) Up to h=80, w=120, l=230, often with thinner mortar joints, usually of lime mortar, with significant regional variations. [0.53 x 0.79 x 1.51]

Bonds are designed to increase the strength and stability of brickwork. Almost all brick bonds rely on alternative courses being offset by 1/4 or 1/2 brick from the course below.

Common bonds include - English (alternating courses of stretchers and headers), Flemish (alternating bricks as stretchers and headers), Stretcher (mainly seen in cavity walls), Header (mainly seen in industrial chimneys and curved works), Garden Wall (three courses of stretchers and one course of headers), and many others.

 

WALL THICKNESSES

Brick : 110 [0.8] {4½ "} (outhouses and sheds only), 225 [1.5] {9"}, 290 [1.9] {11"} (cavity walls), 335 [2.3] {13½"}, 440 [3.0] {18"}

Ashlar stone : 100+ [0.7+] {4"+} (facings, e.g. Bath stone)

Rubble stone : 300-600 [2-4] {12-24"}

 

SASH WINDOWS

Sizes (domestic) Normally two equal-sized sashes. Georgian sashes usually glazed with six to ten panes per sash, with no horns on the frames.

Victorian sashes often glazed with one or two panes per sash, with horns on the frames.

Typical sash windows: w=1200, h=1500-1800 [8 x 10-12] {4'0" x 5'0"-6'0"}

w=1500, h=1800-2100 [10 x 12-14] {5'0" x 6'0"-7'0"}

 

LINTELS

Stone : height 1/6 opening clear width plus 150 [1.0] {6"} minimum bearing at ends

Brick : height 225 [1.5] {9"} or 330 [2.25] {13½"} with shaped bricks (flat or arched soffit)

Timber (usually oak) : height 1/10 opening clear width plus 150 [1.0] {6"} minimum bearing at ends

Concrete : height 1/10 opening clear width plus 150 [1.0] {6"} minimum bearing at ends.

 

SLATES

Roof slates: lots of different sizes, traditionally named after female nobility. Common ones include - Wide Countess 508 or 457 x 305 [3.3 or 3.0 x 2.0] {20" x 12" or 18" x 12"}, Countess 508 x 254 [3.3 x 1.7] {20" x 10"}, Wide Viscountess 457 x 254 [3.0 x 1.7] {18" x 10"} and Wide Lady 406 x 254 [2.7 x 1.7] {16" x 10"} Thicknesses vary from about 6 to 8, [40 μ to 50 μ] {1/4" to 3/8"}

 

TILES

Three main types -

Plain tiles : double lapped and set staggered, various sizes, depending on locality, Typical size 265 x 165 x 13 thick [1.7 x 1.1 x 0.1] {10½" x 6½"}, gauge 85-100 [0.6-0.7] {3½"-4"} (the shallower the pitch, the less is the gauge), cover width 165 [1.7] {6½"}. Half tiles under bottom course and on top of upper course, tile-and-a-half at ends.

Pantiles : interlocking size 343 x 242 x 13 thick [2.3 x 1.6 x 0.1] {13½" x 9½"}, gauge 295-305 [1.9-2.0] {11½"-12"} cover width 200mm [1.4] {8"}.

Double Roman tiles : interlocking size 418 x 330 x 12 thick [2.7 x 2.2 x 0.1] {16½" x 13"}, gauge 318-343 [2.1-2.3] {12½"-13½"}, cover width 300 [2.0] {12”}

Glazed roofs : glazing bars at a normal maximum of 610 centres [4] {2’0”}, purlins normally at 1800-2400 centres [12-16] {6’0”-8’0”} with a step over each purlin.

 

ROOF PITCHES

"Flat" (lead) : up to 5° with 50 [0.3] {2"} steps across fall at 1800-2400 [12-16] {6'0"-8'0"} centres and ridges parallel to fall at 600mm [4.0] {2'0"} max centres

Corrugated iron or Asbestos : 5°+

Slate : 22.5° - 35°

Plain tile : 35° - 40°

Pantile, Double Roman : 30° - 35°

Thatch : 40° - 50° (note - often replaced with Plain tiles at the same pitch

Stone slate : 50° (note - often replaced with Pantiles, Double Roman or Plain tiles at the same pitch

Note that very few roofs were constructed at 45°

 

CHIMNEY STACKS

Brick-built - minimum sizes

One flue: 450 x 450 [3 x 3] {18" x 18"}

Two flues: 450 x 800 [3 x 5.3] {18" x 2'7½"}

Three flues: 450 x 1150 [3 x 7.5] {18" x 3'9"}

Four flues: 450 x 1485 [3 x9.8] {18" x 4'10½"}

or 800 x 800 [5.3 x 5.3] {2'7½" x 2'7½"}

Stone-built chimney stacks are usually significantly bigger

______________________________________________________________________________

 

REFERENCES

Most titles are available from on-line outlets, some are expensive and best sought via your local library. Where there are several editions of the same book, I have given details of the latest one that I can find, as that is the most likely to be available.

 

There are many books published on regional vernacular building; again your local library can help. Note that, with a few exception (Brighton, Bath, Bristol, Harrogate, York, Edinburgh, etc), provincial fashions tended to lag a generation behind those of London.

 

Vernacular Architecture – an Illustrated Handbook. Dr R W Brunskill. Faber & Faber 2000.

Excellent explanatory text and lots of lovely line drawings of plans, diagrams and details

 

Houses of Britain – the Outside View. John Prizeman, Mark Prizeman Hutchinson & Co 2003

Many colour photos and superb coloured elevations which just happen to be to 1:150 scale

 

The pattern of English buildings. Alec Clifton-Taylor. Faber & Faber 1987 Difficult to get.

An authoritative work with sections on all the main materials used in building.

 

Six English Towns. Alec Clifton-Taylor. BBC Books 1986 Now out of print.

Chichester, Richmond, Tewkesbury, Stamford, Totnes and Ludlow

 

Six more English Towns. Alec Clifton-Taylor. BBC Books 1985 Available

Warwick, Berwick, Saffron Walden, Lewes, Bradford-on-Avon and Beverley

 

Another Six English Towns. Alec Clifton-Taylor. BBC Books 1984 Available

Cirencester, Whitby, Bury St Edmunds, Devizes, Sandwich and Durham

 

Well folks, that’s all from me about buildings.

 

John

Edited December 20, 2018 by JohnBS

[bécasse]

I particularly recommend Brunskill's Vernacular Architecture. It was first published in 1971, my 1978 paperback edition cost me £ 2,95 (and is much thumbed) and it is still in print - Amazon have it in paperback at £ 18,95. 

It isn't just worth it for the drawings and diagrams, Brunskill explains clearly how materials and construction styles vary over the UK, the two are related, of course, and will stop you making many a howler in model building construction.

 

One thing that I would add that I don't think Brunskill mentions is that, while bricks are typically 3" by 4½" in plan (with mortar included) and even modern metric bricks aren't far off that size, the height varies (or rather varied until about 50 years ago) with 3" (so four courses are 1' high) being typical in south east England, but as one moves north and west the brick height tends to increase (four courses equalling, say, 1'-1½" or even more). This can make quite a difference if one uses brick courses to estimate the height of walls (and ultimately buildings).

[JohnBS]

ASHBURTON AND TOTNES

 

A couple of ships

 

Clyde Puffer "Starlight" c.1935

Registered in Glasgow

Built 1929, scrapped 1953

 

"Starlight" at Totnes quayside is at the southern extremity of her range.

 

In my interpretation, times were hard in the 1930s so she was loaded with a cargo of timber baulks from Forestry Commission plantations in Scotland for J & R Reeves & Co Ltd, timber importers. When unloading is complete, she is due to call at Par to pick up a return cargo of china clay, then off to South Wales for bunkering, before offloading the clay at Connah's Quay for onward transfer to the Potteries at Stoke.

 

The model

 

The model was from a Langley kit; the major components in resin with white metal details. The hull was cut down to waterline and the hold was drilled and ground out, with a floor of 60 thou plastic card added. Various details were also added - etched handrails, plastic card hatch coaming, boards and canvas (tissue paper), ladder and companionway, dinghy with thwarts and oars, stays (wire) and rigging (cotton), flag (painted tissue paper), navigation lights and crew. The cargo of timber was made from large matchsticks.

 

The inspiration

 

The Clyde Puffer in its protective box-file. "Starlight" is currently moored at Padstow on John Greenwood's excellent North Cornwall layout.

 

Another one to come.

 

John

[JohnBS]

ASHBURTON AND TOTNES

 

Topsail schooner "Freya V" c. 1925

Registered in Lübeck

Built 1905, scrapped 1935

 

An iron-built three-masted topsail schooner with a cargo of Baltic pine planks for Reeves of Totnes in 1925. According to my story, the ship was built in 1905 at Neptun Werft AG, Rostock, Germany and was registered in Lübeck. She had a displacement of 560 tons and is shown waiting to be warped into a berth at St Peter's Pool Quay.

 

After the Great War, she was seized by France as reparations but was bought back by her German owner. Her sail plan was revised to reduce crew requirements and she was fitted with an enclosed wheelhouse, an auxiliary diesel engine (ex U-boat) and a donkey engine on the foredeck, driving a hydraulic winch and windlass on the forecastle.

 

The model won the Chairman's Trophy at the 2017 2mm Scale Association Annual General Meeting.

 

The model started life as a fairly ancient and rudimentary Czech Republic plastic kit of the Cutty Sark, to 1:180 scale. The hull was cut down to the waterline and a couple of transverse slices were taken out to reduce the slender "clipper" shape. Masts were slightly reduced in height and the whole was totally re-rigged, with just two square sails on the foremast and fore-and-aft gaff-rigged sails on the main and mizzen.

 

Sails were made of thin paper, shrouds and rigging were of button thread and sewing cotton. The windlass, donkey engine, wheelhouse and other details were scratch built in plastic card and the hatch covers and deck housings were modified.

 

The inspiration. A Russian (?) ship at Totnes with a cargo of Baltic timber, c. 1910.

 

Merry Christmas to you all,

 

John