If you were to believe the mainstream media, you obtain a 3D printed part by loading up the part, sending it to the printer, and a few minutes later, it pops out. Presumably with a little “ding” noise.
The reality is that there is a lot of preparatory work needed to ensure that you not only get a part, but that you get the part you want, out of the machine. There are, of course, all of the setup and preparatory steps for each machine. These are as arcane, and often undocumented, as the very best of the black arts and we don’t have time for those here.
There is also the question of setup in the software. Every 3D printing machine has setup software. It’s here that you need to ensure that your part is scaled correctly for the build process at hand. Right at this first stage, you need to balance the fact that your prints will, in most cases, shrink after they’ve finished. Smart setup software takes this into account. That at the lower end won’t, so you need to account for it manually.
Then you need to ensure that your part is oriented or laid out for the build process. This is best illustrated with examples. Resin-based (those built from a vat of resin ala Formlabs or many Stereolithography machines) systems often require that you have your part angled at 45 degrees to the build platform (so that you don’t get localised build up on heat from the exothermic reaction of the resin curing). Sintering machines also have sweet spots in their build chambers where you get the best results. You need to make sure your critical parts are in that zone.
Then in many cases, you need to add in support structures (sintered parts don’t need these as the parts are supported in the powdered material). Again, some software does a good job of this, some doesn’t.
3D model with support material displayed.
At the same time, you need to make sure that your part is oriented for the function you want it to perform. 3D printing is a layer-based process. As a result, the parts are, by their nature, anisotropic, and you have weaknesses introduced into the parts in line with the layers.
Critical features must be balanced out with part orientation. For example, if you have snap fits, you typically want the layers running vertically through them, otherwise they’ll snap off. But what if you’ve got multiple snap fits that don’t all conform to the same direction?
You’re starting to get the idea.
And this is just with basic machines and pretty standard materials. We’re heading into a period where it’s possible to build parts on these machines with segmented material properties. Stratasys can already do it with its Connex machines. See the work of Neri Oxman as the most extreme (and extremely beautiful) examples. This will become more commonplace in the coming years.
Successful 3D printing is a balancing act—one that
Successful 3D printing is a balancing act—one that extends way beyond the remit of the 3D design system. Changes are afoot in this space and several vendors are (PTC is one of them) introducing tools to make the process easier, more reliable, and repeatable. Will it ever be a case of pushing a single button and getting your print out of the machine a few hours later?
[Ed. Learn more about PTC Creo and 3D printing on our Design for Additive Manufacturing page.]
The only answer has to be “Probably not.”
Then of course, once you’ve got your print, it’s finished, right? Crack on with the next job? Not so fast, cowboy. Post processing of 3D printed parts is a similarly, if not more, complex business rarely understood outside the small proportion of folks that do this daily.
If you want further reading, I recommend a look at Spencer Wright’s work. If you’ve not come across Spencer before, he sends out an email digest of design-, engineering-, and materials-related goodness called The Prepared every Monday and it’s since become something of a regular read for me. Alongside his current Kickstarter project (the Public Rad.io), he’s also documenting his experiences with additive manufacturing with metals – I’d recommend starting here.