following on from here.
no matter how nicely my rhino models rendered, i couldn’t get them into a printable state. boolean operations were temperamental at best and often failed, apparently at random – working one time and failing the next with identical objects. my models were teeming with horrible jagged edges.
moment of inspiration came to my rescue. it was written by the same person who wrote rhino and all the functions are familiar, but this is a pared down version aimed at the beginning modeller which excels at producing “hard surface” models – ideal for 3D printing.
here you can see my finished model along with some of the construction elements. the tube on the right is for cutting central holes for the magnets (larger 6mm diameter ones to increase the “grab” between them and hold the heavier final materials). much as i would have liked to use a heightfield image for the top i couldn’t get it to work so i resorted to extruding my line drawings and punching details through cookie-cutter style. this process still required a fair amount of trial and error – i was unable to include some detail i’d have liked on the moth’s body.
i flowed the moth shape across the surface at top left, constructed from the same curve initially used to create the medal profile, after which i just slid it across onto the top. having tried this process in rhino i was left with the eternal boolean union problem – looks okay in render, looks awful when taken into makerware to print – but moi joined them perfectly without complaint.
my plan had always been to powder print the finished model, but i discovered the printer at college had died a while back. the pressure of sending the job out to a commercial printer really focussed my mind – no opportunity for experimental test prints. 3dprint-uk looked like my best bet price-wise, with the added benefit of having an automated quote system that previews the validity of your model and offers some basic fixes. when your two-piece model shows up as having several more you know it has a chance of failure. unfortunately having a test print come out okay on the makerbot – which my glitchy files did – is no guarantee it would also work on a different printer.
i used a piece of software recommended on their site – netfabb basic – to check and repair my stl file. initially it came up with several holes – easily fixed – and thousands of inverted triangles – not so easily fixed in the free version. it turns out this was a result of exporting the mesh with the maximum number of polygons, once i reduced that number down i managed to make an honest-to-goodness valid file \o/
while this process has often been a frustrating one i feel i’ve come a long way in a relatively short period of time. i’ve identified the modelling software best suited to my needs and abilities, with the bonus that what i learn on there should be transferrable to rhino if i find i need something with more features in the future. i did try a number of other pieces of free or trial software as well as taking advantage of autodesk’s student downloads to trial maya and mudbox, but each had its own steep learning curve. in terms of producing printed finished pieces moi works best for me. the tutorials on 3dprint-uk have been very helpful in narrowing down what makes a valid model that has the best chance of printing as expected. i feel well enough equipped now that i would happily use 3d modelling and printing in future work.
can’t wait to get my print back and see how i did.