3D printing the bogies
My Lightwave 3D model of the S-303 was not originally intended to become an actual 3D model, but due to Covid Lockdown dragging on, I decided to 3D print its ‘Flexicoil, 2nd Pattern’ bogies… maybe I’ll print the rest of the loco or do another balsa/3D print combo (as with my X-Class model).
Flexicoil bogies were delivered to Victorian Railways in two forms: 1st Pattern and 2nd Pattern. 2nd pattern (the type I have modelled) is easy to identify as it has a series of holes on the frame sides. For more info about which VR locos had which Flexicoil bogies, see this link.
Modelling objects in virtual 3D for rendering in 2D (i.e. as a graphic image to be viewed on a screen or printed on paper) does not necessarily mean that the virtual 3D object will print correctly on a 3D printer in the real 3D world.
Intro: Virtual 3D vs. actual, real world 3D
Virtual 3D objects are usually constructed as a series of flat faces known as polygons. Polygons have a minimum of three edges and vertices (corners), and unlike in the real 3D world, virtual 3D polygons can exist with only having one side. In the real world, everything has to have depth – viz. cutting out a triangle polygon from a thin piece of paper; no matter how thin the paper, the triangle will have some depth to it.
If a virtual single-sided 3D polygon is flipped around away from the virtual camera (i.e. the point of view), the polygon will seem to disappear, unless its surface properties have been set to make it a double-sided polygon. However, the surface properties of a polygon only affect how it is seen by the virtual camera; making the polygon double-sided doesn’t give it any depth so it still won’t print in 3D.
Virtual 3D objects
A simple virtual 3D box will appear to have three dimensions when viewed on a 2D screen or printed as a 2D image on paper; all it takes to sell the 3D illusion is to have at least two, but preferably three of its six faces visible.
Below: 2D drawings pretending to look like 3D. The drawing on the left might or might not look like a three dimensional box, but with three faces visible, the one on the right looks more convincing.
Virtual and physical 3D worlds
A virtual 3D object can exist just fine in virtual 3D space without having any real depth to it, and more than one object can occupy any given space at the same time. But in the real world all objects have height, width and depth (even the thinnest piece of paper has some depth or thickness to it), and separate objects can’t occupy the same space at the same time. These physical constraints add extra consideration when modelling for 3D printing.
Virtual and actual 3D compatibility issues
Consider a virtual 3D box that is made from only six faces without anything inside – it would 3D print as a solid piece of plastic; but if you want to print the box as a hollow shape, the virtual 3D box will need to have walls with thickness; i.e. the six wall faces will need to have depth modelled as part of their geometry; the box will need an inside and an outside separated by depth (or thickness, if you prefer that term.)
Compatibility issues will also arise where two or more virtual 3D objects intersect or overlap – just fine in virtual 3D space, but in the real world only one object can occupy any given space at once. Therefore, virtual 3D objects would need to be reworked, cut and trimmed so that they intersect properly in the actual 3D world.
Below: 3D virtual box, with its top face removed for clarity. Single-sided polygons. Notice the rear polygons seem to be missing. Being single-sided, they are facing away from the camera and are not visible. As is with the top removed, this object would not print to 3D because its polygons do not have any thickness (or depth) to them. If the top was added, the box would print as a solid piece of plastic. Click image for larger view.
Below: The same box with its surface attributes set as double-sided polygons. The rear faces are now visible to the camera, but as with the box above, it would not print in 3D because the polygons do not have any thickness; but putting a top face on to the box would still result in a solid 3D plastic cube. Click image for larger view.
Below: This version of the box has side walls and base with thickness (depth), and would print just fine in 3D as an open-top box. To create walls with depth, extra polygons have been added to make the inside of the box. If a top with depth was added, the box would print as a hollow cube. Click image for larger view.
Reworking the loco geometry for 3D printing
As with the example of the simple 3D box and overlapping or intersecting objects, the same principles apply for the model loco and bogies, which are made from hundreds of separate 3D objects. All the 3D geometry must have ‘wall thickness’ or depth in order to print in the real 3D world, and multiple objects must intersect correctly in order to be assembled into the final real 3D model.
This meant that all the virtual 3D model components had to be checked for ‘real world 3D compatibility’ and were necessary, altered or modified to make sure that they would print properly on the 3D printer.
Below: The Lightwave 3D bogie; hopefully the 3D print will look something like this when its parts are assembled. Click image for larger view.
3D printing the bogie components
Below: The series of holes in the frame sides identify these as Pattern 2 Flexicoil bogies (the frame side at the rear has its printing supports still attached); the H-Shaped bolster and frame joining rails. The hole in the bolster is for a large pin to join the bogie to the loco sub-frame and act as a centre pivot point in order to let the bogies turn in relation to the loco body so it can negotiate curved track. Click image for larger view.
One of the three traction motors set up ready to print; the Snapmaker 3D printer gives a preview based on the Gcode created from the Lightwave 3D model, and reports an expected print time of 4hrs 18 mins. It’s a slow process!
Another of the traction motors just about finished – 5hrs 7min!
Finally done. Parts of the traction motor component had to have removable support material printed.
Traction motor components were designed with tongue and slot connectors to make location and gluing easier.
Traction motors joined. Not having access to detailed resource material, I’m not sure how the real traction motors are mounted, but for the model, the pair of connecting drawbars do the job. The transverse beams will support the H-shaped bolster and help to tie the frame sides together. The traction motor axles don’t turn, and will have their ends glued into recesses on the frame sides. The wheels will rotate freely around the axles, but the fake bearing ends on the outside of the frame sides remain static.
Assembled bogie. Although designed as a non-functioning static model, the wheels can turn freely allowing the bogie to be pushed along by hand.
The brake pistons need some connecting rods. Otherwise, the bogie just needs a paint job. The assembled bogie is 265mm L x 135mm W x 60mm H – approximately 1:20 scale (same as my X-Class model)