In my previous post, I showed the brass components that will form the ‘hard’ skeleton of my planned model of an Armstrong Standard Goods engine. The fireman’s side of No.661 is shown below to complement my previous illustrations of the driver’s side of No.31:
Armstrong Standard Goods No.661
Now, I shall describe my procedures for producing 3D-printed parts to form the ‘flesh’ of the model, which will carry the details of the outside frames, footplate, boiler cladding, smoke-box, and firebox. The methods I used are similar to those that I have previously described for my model of a Gooch Standard Goods.
The frames should have been easy but I hit a glitch for which I have a work-around. Since I had a drawing of a frame, including all the rivet details (shown in my previous post), I used ‘Silhouette Studio’ software to trace this drawing. I used this software because I am familiar with it but I also know it has limitations when saving files.
Fortunately, there is a website that can convert ‘studio3’ files into SVG format on-line. Unfortunately, however, for reasons that I haven’t yet investigated, the resulting SVG files do not import correctly into my ‘Fusion 360’ 3D-modelling software. My ‘work-around’ was to open my SVG file in ‘Inkscape’ and re-save it as a DXF file. The resulting DXF file opened correctly in ‘Fusion 360’ and allowed me to continue with creating the 3D model.
My first step was to select all the raised rivet heads and raise them by 1.33 mm above the drawing surface. I then selected the rest of the frame outline and raised this by 1 mm. This leaves a solid frame with the rivet-heads protruding from the surface. My reasons for doing things in this order is that the original drawing remains visible whereas, if I raised the whole frame first, I would not be able to access the rivet details. Once drawn, I exported the file into my ‘Cura’ slicing software, where I could preview the appearance of the frame before printing:
Preview of 3D-model of an Outside Frame
For the other components, I used ‘Autosketch’ as an intermediate step towards creating my own 3D drawings. In principle, the footplate was the simplest of these parts but needed care in ensuring clearance was provided for the 00-gauge wheels that intend to use. It is basically a rectangular plate with cut-outs for the wheel apertures. I decided to add the outer faces of the wheel arches and the sand boxes to the plate since it was convenient to add these at this stage.
Autosketch Drawing of Footplate
My 2D drawing is shown above and, after saving it as a DXF file, I opened it in ‘Fusion 360’ to raise the 3D elements above the footplate. This is the first time that I’ve added details from a drawing in a different plane from the original, then using the Push/pull tool from the new plane. So, another step along my personal learning curve.
Adding wheel arches from a drawing in the vertical plane
A big attraction for me, as an impatient modeller, is that the print times for these individual components are quite short. This makes it easy to review my designs and to make corrections and re-print as necessary. I found the printing process was extremely tedious when I was printing entire vehicles at one go, which usually took several hours to complete. The print time for one of my outside frames is only about 9 minutes, while the footplate, complete with sand boxes and wheel arches, takes less than 45 minutes.
My design for the smokebox followed the same process that I have used previously for the Gooch Goods, in that I simply extruded the length of the smokebox from a 2D drawing of the front face. The shell includes openings for the boiler barrel and the chimney.
Similarly, the boiler cladding was designed as for the Gooch Goods, with the addition of the sides of the firebox in the relevant area. Both these component proved straight-forward to print.
Before actually physically printing anything, I realised that it was possible to open several components at once in the ‘Cura’ software and arrange them on the build plate, in their correct relative positions, in order to check that they all fitted together neatly. I made a couple of screen-shots of these ‘virtual test builds’:
Print previews of 3D Component Assemblies
Because it’s rather too warm in my work-room at present, I’m going to leave things on a ‘cliff hanger’ for the moment but I feel reasonably confident that I have a good way ahead, towards a physical model. I also need to wait for my wheel-sets to arrive from ‘Alan Gibson’, so that I can check that all the clearances I have allowed will be sufficient but not excessive.
Whilst in the 3D-modelling mood, I decided to investigate the possibility of adding boiler fitments such as domes and safety-valve covers to my boiler designs. I recall some correspondence, several years ago, when several fellow modellers were experimenting with ‘Blender’ for 3D modelling. I must admit that my impression that it was all rather difficult has remained with me ever since.
Fortunately, times have moved on. I see that, back in 2014, I wrote a comment “It will be nice when there's a 3D equivalent to the Silhouette cutter!” to which @JCL replied “Yep, there's no way I'll be able to afford a printer“. How much the world has changed in six years!!! Fortunately, 3D software has also advanced so that, when I set out to try creating a dome, it turned out to be much easier than I expected.
In ‘Fusion 360’, it is possible to make a 2D cross-section drawing and use the ‘Revolve’ tool to rotate it around a selected axis, to produce a solid component. This works in a similar way to the ‘Push/pull’ tool that I now use routinely for linear extrusions. There are one or two caveats, in that the axis has to be outside the area enclosed by the drawing but that is easily overcome by splitting the profile down the centre-line. I created a ‘profile’ by combining drawings of a rectangle and a circle and then ‘revolved’ the profile through 360° to create a dome-shaped object.
Extruding a Dome in ‘Fusion 360’
The next move along my ‘learning curve’ was to add this newly created object to my existing boiler model. This was not an intuitive process but, fortunately, the solution was to be found on the web.
The first requirement was to select and copy the dome ‘body’ and then to open the existing model of my boiler. The tricky step is to select ‘Create Base Feature’ from the ‘Create’ menu. After that key step, you can ‘Paste’ the body of the dome into a ‘base feature’. Finally, click on ‘Finish Base Feature’ and the dome becomes a new ‘body’ alongside the ‘body’ of the boiler. I have no idea what all that means but it works.
Once the two ‘bodies’ are in the same drawing window, it is only necessary to use the ‘Move’ tool to slide the dome into its desired location on the boiler. The lower part of the dome can be slid ‘through’ the shell of the boiler until the height above the boiler matches the required specification. Any excess length inside the boiler can easily be removed by selected the circular end face of the boiler then extruding this face through the length of the boiler, where it acts as a ‘reamer’, to remove any intrusions!
The next step is to use the ‘Combine’ tool, to join together the dome and the boiler as a single ‘body’. As the finishing touch, the ring where the dome meets the curved surface of the boiler can be selected and then the ‘Fillet’ tool applied, to achieve the desired radius around the join. Once you know how, it is all very straight-forward and, flushed with success with the dome, I quickly added a safety valve cover as well:
Adding Boiler Fittings in ‘Fusion 360’
Well, I may not not have produced any new hardware yet but I have learned quite a lot about 3D modelling, which will hopefully be useful in all sorts of future projects.
In the meantime, we must all wait to see how this will turn out in practice, once I start printing
P.S. Some time ago, I expressed concern about non-availability of spares for my Geeetech E180 printer. I am now pleased to report that I have received spare print heads directly from the manufacturer in China. It took a while but I’m now set up to continue for a reasonable time ahead.
Edited by MikeOxon