Even if I had chosen a different printer, range of filaments or design software, there is one thing that I am sure would be true: it is almost impossible to get a design perfect on the first go. That is doubly true for items that have to interface with real world things, like the lawnmower pinion gear I posted about last month. Something are hard to measure, you have to work around how the printer extrudes the filament and often there are design aspects you forget until you start trying to use the ‘finished’ item.

It would be a huge waste of time and material to try and run off completed items too soon. What I will often do is spend a while sketching things out on paper and trying different ideas out. Once I settle on something, the next questions are whether it can be simplified and whether it will be strong / flexible enough. Once the design idea is settled, I will then think about whether there are any critical parts that need to fit and look at producing a smaller version so I can try them.

For example, I was recently working on a cap for the end of a plastic pipe. The measurement should be easy enough but I still ended up printing three short test rings. The first, at the measurement, was too small as was the one 1mm wider. 2mm wider was just a little lose and so, when it came to printing the cap, I ended up going for 1.5mm wider. The draft pieces took about 3-4g of filament and about 15-20 minutes of printer time but that meant the final piece (40g and just over an hour) landed where it needed to be.

Design, physically test critical parts of the design and then commit. Even then, there may well be later versions once it has been tried in practise but that is always the benefit of hindsight.