Or... Scratchbuilding a Garratt in Sn3.5 from styrene
INTRODUCTION
When my parents returned from a holiday to Zimbabwe some 10 years ago, they asked me whether I had ever heard of "Ingulungundu locomotives". Of course I hadn’t. How could I have known that Ingulungundu was the word for "bush pig" in Ndebele language, and that this was also the name of a certain 15th class Garratt locomotive that they had seen derail (a switch had been thrown when the loco was running over it) at Victoria Falls station. My parents had voluntarily stayed over eight hours at the station, fascinated by the huge 4-6-4+4-6-4 steam loco and by a troop of African workers finally pushing the giant back on the rails only with their own hands and some levers. Since then, my parents ask me regularly if there are any models available of "their Ingulungundu", no matter which scale, and I was sorry each time to say that there weren’t. The only commercial Garratt models I was aware of where the DJH white metal kits, but even their GMAM (although this type was used in Zimbabwe for some years as well) looked too different from typical Rhodesian types to justify the expense and effort. And I didn't like their gauge compromise to build the model of a cape gauged loco to H0 standard gauge. A few months ago, when I went to South Africa to a rail conference myself, the unavailability of Zimbabwean railway models came to my mind again, and returning to Germany I decided to change this and to present my parents their model.
DECISIONS
So I planned to scratchbuild a NRZ Garratt on the base of two commercial steam loco’s chassis. As railways in the south of the African continent are built to cape gauge (3' 6" or 1067 mm), this could be either done in H0 scale (1:87) is using the chassis of 12 mm TT locos, or in S scale (1:64) using 16.5 mm H0 components. I went for S scale on 16.5 mm rails (this combination being called Sn3.5 or Sm), as H0 locomotives should be easily available, and as I had 16.5 mm rails for testing already at home. I didn't have any S-scale equipment before, but neither had I had anything suitable for Africa. Thus there was no need to build this singular model compatible to anything already existing. Finally, during construction, I found out that the slightly bigger scale made building and detailing considerably easier. When I was still searching for cheap big-wheeled (due to the scale conversion) steamers with the motor in the right place to fit under the bunkers, I accidentally discovered an old East-German PIKO model of a Saxon XIV HT (later Br 75.5) from the early 80's in my spare parts box. It had a very simple but robust chassis, with worm drive and all axles being coupled by gears. Not to spend too much effort in searching for suitable donor locomotives, I took it as a start, and a second one could be bought soon for little money at a swap market.
But considering scale conversion from H0 to S, the PIKO locos had too small wheels. So I decided not to model a 15th class Garratt, but its little brother, the slightly smaller 2-6-2+2-6-2 14A class.
RESEARCH
Finding plans or drawings was the next problem. Despite some correspondence to Africa I didn’t manage to get any scale drawings at all. But I had some photographs and drawings of other Garratts built by Beyer-Peacock for several other countries from some books. Comparing them, I found out that many of them were built in kit-style from standard or similar components. So I was able to somehow reconstruct the dimensions of my 14A. I didn’t want to reproduce the loco in absolute fidelity anyway, but just to build some kind of impressionistic model superficially similar to the prototype and giving the impression of a "typical Zimbabwean Garratt".
But just after I has finished the model, I got A. E. "Dusty" Durrant’s wonderful book "The Smoke That Thunders", and to my shame, there I found a scale drawing of the 14A. But comparing it to my drawings, I realised that my reconstruction had been almost correct, as in most dimensions it had a discrepancy of less than one (model) millimeter.
CHASSIS
Each PIKO original chassis consisted of two solid steel plates that I sawed to shape to reduce their size. As the original motor was situated at the wrong end of the loco in the former cab, I mounted the cylinders and the frame extensions carrying the buffer beam the other way round to get the motors to the Garratt’s outer ends. The European style buffer beam was replaced by a pilot plate made from brass, including a bracket to fit a Kadee #30 series knuckle coupler, a cow catcher, steps and a lip to snap in the bunker’s outer ends. The articulation joint was made from a small brass block acting as additional frame spacer with a vertical 6 mm hole for the pivot pin. This pin has a ball-shaped end to allow all angular movements and to constrain only the longitudinal and lateral displacement. A small brass plate was soldered on top of the block as support for the boiler frames to float on.
MOTORS
The original motors were extremely bulky, rough and noisy old-fashioned monsters. No way in keeping them. I replaced them by coreless Faulhaber can motors with flywheels (from sb-Modellbau) which finally allowed building the front water tank to scale.
AXLES AND VALVE GEAR
I kept the wheelsets and the valve gear almost unmodified. The only greater discrepancy between the PIKO model and the 14A Garratt is the second driving axle being directly driven by the cylinders (like the 15th class, and unlike the 14A class), but this can be still changed later. The driving wheels were ok, but the pony wheels had too big flanges to run on my code 83 rails. So I reduced the size of the flanges by chucking the wheelsets in a micro drill and shaping the flanges by means of a file.
BOILER FRAMES
The entire middle unit was scratch built from Evergreen styrene sheets (2 mm for the frames, 1 mm for the cabin). In the big ash pan was enough space for some additional 150g of lead ballast to keep the center of gravity low, get good adhesion and to avoid pitching of the head-heavy engine units to the outer ends. A pair of wires beneath the frames connects both engine units electrically. The boiler was made from a piece of wooden broomstick rolled in paper to bring it to the correct diameter and covered with a final layer of 0.2 mm styrene for easier cementing of additional parts and painting. Rotational parts like chimney dome and headlights were turned freehand from styrene: I cemented several layers of styrene to the approximate shape of the parts, chucked them in a drill, and machined them with a sharp knife. The smoke box door was the only part of the superstructure that I recycled from the PIKO loco.
TANK AND BUNKER
The water tank and coal bunker were built from styrene as well. The tank with it's characteristic rounded edges was made quite solid from 5 mm sheets and filed and sanded to its shape. Tank and bunker were then covered with a layer of 0.2 mm styrene in which I had embossed the rivets from the backside with a scriber. The screw for fixing the tank to the chassis is hidden under the water filler lid; the one of the bunker is hidden between the coal lumps.
DETAILING AND FINISH
Boiler piping was made from styrene rods for the bigger and brass wire for the slimmer pipes. Handrails were made from 0.6 mm steel wire fixed in the head parts of railing posts as used for naval models. Crooked styrene parts, as cabin roof or tank front were formed over a suitable master (e.g. an empty wine bottle) in boiling water. I etched the cowcatcher, ladders, number- and nameplates from brass sheet, a very basic job in a plastic bowl after exposing the photo resist in the sun, but sufficient for those few simple parts. Although the real 14A Garratts didn’t carry nameplates, I gave my model the name "Ingulungundu" as tribute to the original story. Finally, after sanding, filling and cleaning the superstructure was airbrushed in matte colors.
PERFORMANCE
I designed the angular clearance of the articulation such that the loco can run through curves with 400 mm radius. It has a good rail contact of all wheels even on slightly misaligned rail due to the angular flexibility of the articulation in all directions. The running behavior and noise is excellent, which truly can be expected when using Faulhaber motors: It runs very steady and smooth even when creeping at the slowest speed. Due to the power pickup from 12 wheels and the extremely long overall wheelbase the electrical contact is excellent. And finally, as the assembled loco weighs some 600 g, the traction force fulfills my personal expectations despite the lack of adhesion types. Even if my model is not built to the finest detail, I think it has the recognizable appearance of a fascinating type of locomotive which has been characteristic for Africa. But most important, building it was fun, even though it was the first locomotive I have built from scratch and it took most evenings of almost two months.
|