I hadn’t expected to make quite such rapid progress with this model – I must be getting slightly more proficient at using the ‘Fusion 360’ CAD software!
In part 1, I showed how I created the rather complex profile of the sides of the Posting Carriage by extruding a drawing of one end. I drew the end profile as a series of straight lines and arcs, traced over a drawing contained in one of the very useful Broad Gauge Society data sheets.
While this process gave me the basic outline of my planned carriage, there are many aspects that needed further thought, especially since I have to create a model that can be printed with my Geeetech E180 FDM printer. In particular, this means avoiding overhangs wherever possible and adding support structures where there’s no alternative.
With the above considerations in mind, my first step was to split the body horizontally into two. I created an offset plane parallel to the base of the carriage but raised to the top level of the window openings. I then split the body around this plane and worked on each of these two new ‘bodies’ separately, as described below:
Lower Carriage Body
The first step was to open out the interior of the carriage. To do this, I selected a drawing plane at the top of this body and used the ‘offset’ tool to draw a rectangle set back from the edges of the body by 1 mm all round. I then used the ‘push pull’ tool to push the rectangular opening down through the body to the lower level of the windows.
Below this point, the carriage body becomes narrower, so I offset another smaller rectangle and pushed this downwards to ensure sufficient material was left around the edges to provide strong sides. I repeated the process again for the lowest and narrowest part of the carriage. This gave me an open body shell, which could be printed upwards from the floor, as shown below. There may be more elegant ways of doing this but this simple procedure gave me an adequate result.
There are some small overhangs along the sides, where the upper part of the body is ‘jettied’ out from the lower part (rather like a mediaeval house!). I decided to do a trial print, to check whether my printer could cope with this amount of overhang, as support structures on the outside of the body are difficult to remove cleanly. I am pleased to report that my printer produced the curved overhangs with no apparent problems.
To create the window openings, I drew a series of rectangles on one side of the carriage and used the ‘push-pull’ tool to extrude these through to the opposite side. By choosing the option to extend the depth of the extrusion to the opposite outer face of the carriage, the window opening will remain in their correct places if I make subsequent alterations to the width of the carriage. This ‘parametric’ approach simplifies editing of the model, if I decide to make changes later.
The Whishaw drawing shows that the ‘jettied’ appearance of the upper sides was continued around the ends of the carriage. To reproduce this feature, I used the ‘push-pull’ tool to inset the lower panels of the ends and then used the ‘chamfer’ tool to provide a rounded transition between the upper and lower parts of each end.
Although it is not clear from the Wishaw drawing, I also used the ‘chamfer’ tool to round the corners between the upper body sides and the ends. This was a common feature in many early carriages, so I felt it was appropriate in this case.
Adding the Doors
The next unusual feature to tackle was the doors. These appear to be flat and flush with the upper side panels. Thus, the lower door-frames protrude outwards from the in-curving lower body.
I tried a few experiments to work out how to add this feature. At first, when I tried using the ‘push-pull’ tool, I simply succeeded in making openings in the lower side panels! There may be settings to alter this behaviour but my own solution was to extrude the door and frame outwards from the carriage side, designating them as ‘new bodies’. I could then use the ‘move’ tool to slide these new bodies back into registration with the upper body.
Finally, I used the ‘combine’ tool to ‘join’ these new bodies to the original body of the carriage, so that the whole structure became a single entity again. By now, the body was beginning to look like a model railway carriage rather than a solid block!
These steps have completed my basic design for the carriage body. I could, of course, add many additional features, including interior fittings, such as seats and table, but I’ve decide to pause here, to see whether the design works in practice.
Returning to the second of the two ‘bodies’ that I created by splitting the original model, I used similar procedures to open out the interior of the clerestory and to create the rows of 6 windows on each side. I also had to add the upper part of the door frame, which protrudes from the lower sections of the roof and has its own ‘roof’ extending from the sides of the clerestory.
There are significant differences between the Whishaw drawing and the Bourne lithograph (both shown in my previous post) regarding the shape of this part of the ‘porch’ but I feel that the curved top shown in the lithograph looks to be a credible representation of the prototype. In other matters, such as the number of windows in the clerestory, I have preferred to follow the Whishaw drawing and description, so I must point out that several parts of my model are conjectural.
Just as I did for the lower parts of the door, I created the porch tops as separate ‘bodies’, which I then moved into their correct positions by means of the ‘move/copy’ tool. This part of the model is shown below:
It now looks a bit like a cathedral roof with transepts!
Preparation for Printing
I exported the upper and lower bodies from ‘Fusion 360’ onto my computer as two separate STL files, which I could open with my ‘Cura’ slicing program.
At this point, it is important to check the design carefully on the computer screen, to ensure that all the details will print as expected. Because the interior of the clerestory is hollow, it is necessary to include some supporting structures, to ensure that the uppermost roof will print correctly.
This support is generated automatically by the ‘Cura’ software but there are several parameters that can be adjusted. Experience has shown me that this support structure needs to be as ‘skeletal’ as possible since, otherwise, it can prove surprisingly difficult to remove.
Because I have created some poor prints recently, I gave careful attention to the setting-up of the printer. I replaced the blue masking tape on the bed and then carried out a careful levelling process.
I was now ready to create some test prints. The clerestory roof took about 1 hour to print, while the lower body took almost 2 hours. On examining the test prints, I found that my printer had performed very well, with good surface finish and clean openings for all the windows. I believe that it showed the importance of checking all the printing parameters before starting.
It was quite surprising to see how small this carriage appeared, when seen alongside some of my other Broad Gauge carriage designs from the 1850s. The lower part of the body of the prototype was only 6’ 6” wide, so there was no advantage in dimensions over contemporary standard gauge carriages. It is clear that Brunel’s ideas for his Broad Gauge did not include any thoughts of more commodious carriages!
In spite of my comments about keeping the support structures to a minimum, I still found it difficult to clean out the inside of the clerestory and, in doing so, I damaged one of the frames between the clerestory windows. I have described my removal method in an earlier post; my main tool for the job is a set of wax-carving chisels. For the next print, I shall try reducing the amount of support even further.
Now, I’ve started the process of checking the test prints and refining the model as necessary. This involves checking dimensions against the available data (not much of that!), and cleaning up lots of details, such as rounding the corners to all the windows and adding bolection mouldings.
The great thing about computer modelling, as opposed to ‘real’ modelling, is that all this ‘fettling’ can be done from the comfort of my armchair, with a laptop on my knee! I find it quite a therapeutic activity, slowly nibbling away at small details, checking and re-checking, with a restorative glass close by my side By the time I had finished, the model on the computer screen looked like this:
Computer Model after ‘Fettling’ (rounded corners, etc.)
Next stage will be to design a chassis, with it’s characteristic ‘outside’ wheels.