building railcars
a preliminary glimpse on how you can produce a railcar
railcar 4403 in blue livery
Method
In the early 90's I did an attempt to built a railcar of the 604 series but this was never fully finished. This was done in brass, hand drawn and etched. Therefor the windows lacked accuracy and being all slightly different I never got around to glazing them. Thus having a rainy holiday day and looking for something to do I decided to produce something better with the current tools. Normally the start is finding correct information about the item to built, in this case a railcar of the 603 series in its 1955 livery. I could only locate a small dimensional sketch with the main dimensions of length and bogie, the originally used information about window positions and sizes is hidden until today. But a trip on a dry day to the preservation railway in Maldegem with ruler, sketch book and camera was very helpful in producing correct dimensions and it made a nice ride. Backdating this to the 1955 layout was done with help of photos from books and lots of information provided by very helpful Fremo members.
When you have the dimensions you can produce a drawing and from this you can produce the necessary files for production. The drawing was produced in Corel Draw and the producting method is cnc milling.
The cnc milling requires a dedicated drawing that takes account of the radius of the cutter to be used. This requires that a contour is made at exactly the right distance of the original line. This can be easily done in Corel because it has an effect called contour function (F9) which does exactly that. Either outside, inside or the center with a given contour step. Exactly for this function I still prefer Corel above other packages. With this function you can create milling patterns on the fly. For plain jobs like this it is ideal. The above drawing shows the resulting pattern for the two sides and fronts including doors and some partitions and a false roof. You can feed this to the milling software as HPGL plotter file and this will produce the wanted output. In this case an 0.8 mm cutter was used for the major production with some additional cutters of 0.1 and 0.3 for surface details. The depth of cut and/or cutter can be stored as colour information and you may use up to 10 different colours. These can all be individually cut, thus I normally use that also to indicate the cutting order from inside to outside. You should generally first cut the windows and than the contour for the sidewall. The basic cutting order is normally top down following the list of objects, thus sorting them in the right order is another option. Further I find it useful to document the list of the cutters and depths to use as you generally tend to forget this sort of thing very fast.
drawing with correct dimensions on top, milling pattern of all parts below.
The blue lines give surface detail like door contour lines and those on the front as help for painting its 'moustache' of yellow warning colours.
This drawing also shows the alternative fronts for the various rebuilt and updates the railcars got. The version with 2 headlights in single dark green, double headlights with 3 original windows and the later version with 2 windows in the cab.
Put a sheet of material on the machine bed, mount the particular cutter and let it do its work. This item was cut from .25 mm sheet, maybe thin but strong enough. Additional drawings were produced for the frame to be cut from 1 mm double sided pcb and the windows from 0.4 mm clear Evergreen sheet.
More effort takes the curved roof. The small contour in the right hand lower corner was a construct for the double curved profile to be cut. This profile was imported into a 3D drawing program and extruded into a 3D object for the roof.
roof drawn in a 3D software with added bevels.
The same holds for the interior although this was produced with 3D printing in mind instead of milling.
A sketch with simplified interior but with integrated passengers.
milling pattern of the short bogie
Also one of the bogies was drawn up. Essentially this was started out as a plain 2D sketch of which the parts were extruded into this 3D pattern.
The resulting parts glued together together with part of interior at the front and just at the edge the old metal part to be replaced.
Pilot version with colours
This is the pilot painted and with the windows in place. The light green should be a bit subdued. The yellow lines meant lots of trouble, painting didn't look good, airbrushing doesn't give sufficient cover. This here shows painted sellotape over the earlier attempts, also the green on the front was redone by brush with a bit darker green. Cutting thin strips is no problem using an angular board. For better cover it is probably better to first airbrush a white base layer before the yellow top layer.
Three of a kind, essentially copy the process 3 times and stick the lot together. The pilot version on front, 3 copies and a version of the newer 605 railcar which differs in details around doors and roof.
Frame construction
The initial attempt shown above was a pcb frame plate using an Arnold Köf with upgraded motor (6mm Maxon) as bogie. However this turned out not to be a great success.
initial frame with Arnold Köf as bogie and original Arnold motor
Lack of weight and not too well running properties were reason to try to built something better. A new homebuilt bogie with an 8 mm Chinese coreless motor mounted direct on top with 30:1 drive was tested. But this was not to satisfaction as well, lack of weight and vibrations with the thing dancing along the track. Thus clearly a much heavier base was necessary. Re-using some leftover scraps of brass and a bogie from an old surplus model from a starters kit made for a heavier attempt.
2 frames each using a bogie from Roco starters model
Having 2 bogies this allowed to build 2 frames along the same lines. Necessary is integrating a step down gear box. The bogies have only 15:1 gearing and this is too low for slow running. Having 6-6.5 mm wheels requires about 30:1 for decent slow running. The new frames are built from a strip of 15 x 2 mm brass which is too narrow of course. Therefore these are reinforced with 2 strips along the entire length that also allows to integrate the bogie. There are a series of 3 holes in the original Mazac frame which allows to put in some pins for positively fixing its location and then glue it permanently with 2 component epoxy. One further thing that needs to be done is to replace the original plastic bearings for the worm. Originally this axle has a large angle but in this case horizontal is better. The bearings are just 3 mm bushes with a 0.9 mm thick disc from Delrin made on the lathe. The disc is cut square and sits in the original slits, with a bit of fiddling and filing you can get the axle horizontal.
Apart from the length the main difference between these frames is in the gearbox. Because I still had a recovered 10 mm FH with a 12t M0.2 on the axle I decided to try that using a single 28 t M 0.2 gear. This gearbox also functions as motormount which sits in a slot in the base frame. The coupling is by rubber hose from a bycicle valve.
With respect to speed and smoothness this runs a lot better than the earlier attempts, but unfortunately the step down gear generates some noise when running at high speeds. Maybe a plastic gear would be better, other alternative is an inline gearbox with 4 gears, than the relative speeds are very much reduced and will lie below noise generating levels. That is what I used in my first series 59 diesel, but this is not documented on my site.
side view of motorisation, the M 0.2 28t gear can be seen in the box before the motor.
View of the drive with rubber band. A class 70 is a very short loco with bogies with a long hood but ridiculous short cab on one end.
copyright: Henk Oversloot
date: 28 aug 2014
updated: 3 Dec 2020
