A couple of months ago I blogged about the possibility of scaling some of my 3D printed models down so that they were N scale in size (1:148 instead of the 1:76 scale I've been working in). The problem, as I mentioned at the time, relates to the difference between scale and gauge, which would require moving some parts around to fit the different gauge to scale ratio; put simply, making sure the wheels would fit! It turns out that it isn't that hard to tweak the 3D model and so here I present both an OO gauge and an N scale print of the same wagon.
Rather than scale down one of my existing models I thought I'd start work on another model; LNWR Diagram 103, 1 plank open goods wagon. As it only has one side plank it's fairly cheap to print so even if the scaling down went drastically wrong it wouldn't be the end of the world. This was a good plan as unfortunately neither the OO or N scale model is usable, although this isn't entirely due to the scaling.
Lets start with the OO gauge model. Firstly I managed to accidentally fill in one of the axle holes which meant I've had to try and drill it out, which hasn't worked very well and certainly won't lead to a smooth running model. Secondly I tried to simplify the building of the model by using the bevel modifier in Blender to model the gap between the floor planks. Unfortunately this hasn't resulted in a deep enough mark and so has disappeared from the print due to the resolution of the method used to print the model. On the plus side the more detailed suspension has come out much better than on the wagons I've printed in the past.
Having built the OO guage 3D model I simply scaled it by 76/148 to get the base N scale model. I then moved the axle boxes outwards to accommodate the different gauge ratio. While the resulting model is printable, using the more expensive frosted ultra detail material, it is very fragile. In fact it's so fragile that I broke it trying to fit the wheels. Essentially I'd made the wheel supports so thin that they snapped rather than flexed. I glued them back on for the photo and that seems to have worked reasonably well as an interim solution. The other problem is that I seem to have fitted the N scale NEM pockets for the couplings in the wrong place -- they are two far back and so can't pivot properly. I thought I'd followed the specification for their placement but I guess not (the spec is unfortunately only available in French and German so I could have missed something). Also the fitting is very loose so the couplings don't really stay attached very well. Fortunately all these problems are fairly easy to fix by slightly tweaking the model, but as a first attempt at modelling in N scale I'm more than happy with the result.