A Warning: Proprietary Textures and Geometry for 3D Printing

3D printing is a number of technologies that collectively build objects layer by layer. Famously this gives us more design freedom and lets us make more varied and unique shapes than other technologies. We’re also faster at iterating and going from an idea to a part. Less lauded but no less important is the ability to be able to apply textures to a part to improve its performance, as well as apply textures to internal surfaces. We can change infill patterns as well, and the geometries of infill. We can also dynamically change these things, or change them at one particular point, creating gradient parts. All of these abilities are criminally underutilized except for lattices, which are like tattoos on mumble rappers, everywhere and almost universal.

Launcher combustion chamber.

We find ourselves in a sandpit of endless hope, variation, and possibilities. We can, with the physics of the sandpit, quickly and cost effectively build many shapes. In my Flow article, I talk of optimised internal structures that change how we design and make things. Something is not a collection of assemblies, subassemblies and parts, but rather one holistic thing through which energy flows. In my Functional Textile Libraries article I discuss how optimised textures can be an opportunity to improve many parts through a separation of concern and specialisation. Each of these articles is trying to optimistically reassess what we will start to do when we fully optimise products with 3D printing in mind.

Unicore of a 20-kilowatt microturbine engine, by Sierra Turbines, made with VELO3D.

Now we are working in prototyping often and series production sometimes. We do indeed have a lot of design freedom, and unbound, we make the best shape for the one application. And partially this is where a lot of excitement in 3D printing originates. We can make one single valve and we can even optimise that valve to perform best for that one individual use case. This way of thinking is fun and very “problem solvey” which is nice. But, this approach will soon lead to series of tens of thousands to millions of parts being printed. With the industrialised applications we have now—hearing aids, bridges and crowns, Invisalign, polymer dental parts, eyeliner—we are already making tens of millions of parts. But in each case the one defining characteristic that makes 3D printing special is that the shape is optimised to the task. So a unique hearing aid fits you better, the shop just needs to scan an impression, the manufacturer saves money when compared to milling or making it by hand, and overall the process is efficient and benefits everyone in the value chain while making the customer happier. Pure joy. But, the part itself is useful because it can be made specifically for your right ear. If we look towards orthopaedics, we’re seeing components have unique porous textures that promote osseointegration, or unique structures that in effect change the modulus so that an implant works better. Here, it is not just our ability to print unique items, but the geometry on and of those items which is important.

A ColorFABB LW PLA wing.

In speaking with friends, I’ve noticed that a tacit warning was not interpreted as such. So to be a bit clearer: if I can through ANSYS, nTopology and Mary, my genius in house designer, develop an optimal geometry for osseointegration of a Ti64 bone implant, then this one texture could be the highest performing texture for bone ingrowth. This texture would then benefit any company making implants where this is an important goal. Likewise, if I can with Sergey the polymath and Maria the surgeon develop a unique lattice design that gives us good wicking properties to get an optimal blood flow to the implant, then this will also be a huge advantage. If you’re a hobbyist and are working with Colorfabb LW PLA, which expands to make a good lightweight material in RC plane wings, you may just share your settings and the nifty infill structures that you made. But, if at one point you happen upon the most optimal infill for the lightest best performing wing, amateur hour may just be over. If you are developing a filtration mesh for taking impurities out of water using powder bed fusion printed filters, then you may have just found the optimal structure for RO water filtration.

Additive Drives motor winding.

The geometries that we are finding can create huge commercial advantages for the owners of their IP. And yes, this can be a big opportunity for us. At the same time however, this is a threat. In the future I would expect companies to start to patent, obtain design rights for, and try to somehow protect specific textures and geometries. I would expect patent trolls and firms set up to do this to claim large swathes of the geometric landscape we all have access to now. A lot of it may not ultimately be legally defensible, but who will have the money to try to combat this? So we can expect that a lot of the best patterns, geometries, structures, and internal structures will be unavailable to the broader 3D printing community.

(a) Different separator infill patterns that can be obtained by using classic 3D printing slicer software (40% infill density); (b) Various infill densities of the same infill pattern (Hilbert curves); Capacity retention plots at 4.25 mA.g−1 (C/40) for the complete assembled battery after 1 h impregnation: (c) using 100% infill density separator and (d) using a 70%infill density Archimedean chords pattern. Here, note that each layer is 200 µm thick.

What we have is a technology that can search for and build the best individual shape for almost every use case. Obviously, we would assume that people will want to then claim that one single geometry as their own. The infinite meshes will therefore be reduced to the available ones.