There is some very interesting research at the Hasso Plattner Institute 3D-printing metamaterials. Normally, complex mechanical properties arise from assembling components made of various materials; in contrast, a metamaterial object obtains its properties within a single part using a single material without assembly. Their work centers on using a deformable cell pattern, which they said would work with ABS and PLA, but all their demos are made with NinjaFlex... which deforms easily, but lacks the rigidity needed for many kinds of devices.
To test the feasibility of combining rigidity and deformability, I created a simple design for metamaterial longnose pliers. The black pliers printed in ABS copy the metamaterial cell design they used, but the ABS proved too rigid; the nose barely moved before the ABS began to crack. For my second attempt, I reinforced the stress points, thinned the walls of the deformable cell structure, and printed using blue T-Glase. This more flexible material works better, but still the nose only closes about half way before the pivot points start to crush and tear. As one last test, I tried printing one out of NinjaFlex... and the thin nose simply flops around; the material is never truly rigid (here's a video of PLA HingePliers picking-up the NinjaFlex cell-based metamaterial version). Update: As of February 7, 2017, HassoPlattnerInstitute_HCI has posted their design on Thingiverse.
One potential fix is to use different materials for the deformable and rigid parts; that approach was used in Thing 1388783. However, that requires assembly or dual extrusion... and the resulting tool clearly isn't a metamaterial.
Our trick is to use a single, fairly stiff, material and to design the deformable and rigid regions so that operation does not develop damaging stress between them. The result is my HingePliers design, published here. This mechanism is not merely "printed assembled" without supports -- it is literally one piece of material and doesn't even have any structural spans! (Technically, the University of Kentucky logo embossed in the bottom side has tiny spans in it, but that shouldn't really count.) The central hinge uses 45-degree angles to avoid spans and the spring, which deforms harmlessly into free space (rather than against a rigid part), connects the two handles so that the entire device is literally one object made of a homogeneous material: blue PLA. On that basis, I would argue that my HingePliers thing is actually a metamaterial object, extracting very different behaviors from the same material by controlling structure. The pliers operate very smoothly, with good tension to restore them to their open position, yet can exert a reasonable amount of force through the rigid nose.
A very simple print with no supports taking less than an hour to print and using under 20g of PLA. Just about any material can be used, but PLA is one of the better choices in balancing stiffness of the nose with flex in the spring. I used blue PLA as usual, on a heated glass bed. I wouldn't recommend printing with less than 25% infill for good stiffness, and the part is oriented so that a regular 45-degree grid internal fill pattern (Cura's default) will result in good strength for the thin nose. The print-assembled hinge contains gapped interlocking surfaces, but the design incorporates a fairly generous 0.75mm gap, so most printers should have no difficulty printing a working structure. Incidentally, the gap doesn't cause significant play in operation because the spring keeps the interface area in the hinge aligned and loaded.
This part shouldn't need a brim, but the nose section of the HingePliers gets very thin, so I have incorporated a rectangular pad as a partial brim to keep that portion from lifting during printing. That pad must be removed after printing; simply clip it off and perhaps clean the edge with a few passes of a file.
It should go without saying that you can easily break these pliers by applying too much force; these are only useful for relatively gentle work. However, normal operation does not seem to cause measurable wear and they are effective for picking and placing small parts, etc. Printing with higher fill in other materials, such as high-temp PLA or stainless steel PLA, might produce a more robust tool.
This is just a first experiment in metamaterial design for Professor Dietz's computer engineering research lab at the University of Kentucky (better known as Aggregate.Org). Our goal here really is not just to create interesting metamaterial structures, but to make the underlying principles semi-automatically usable in creating new designs... which is something the folks at the Hasso Plattner Institute seem to have done very nicely with their cell library and a design tool that can simulate operational deformations. Although we are currently only publishing an STL file, the designs shown here were created as parametric OpenSCAD code which we hope to evolve into more generic metamaterial design tool support; for us, this really marks the first step toward combining of metamaterial concepts with design-for-manufacturability insights like the print-assembled hinges in our HingeBox (although those hinges pivot vertically, whereas the hinge here operates in the horizontal).
The principle that our HingePliers demonstrates is that one can build more effective metamaterials by using larger, more complex, specialized structures instead of repeating structures from a small library of fixed-shape cell designs. The various manufacturability constraints on 3D-printable metamaterial structures are what make this sort of design really difficult. For example, extrusion-based printing has the constraint that extruded material must be self-supporting. There are also less obvious constraints that are just as important; for example, a printed spring cannot be under tension as it is printing, so it is much more obvious how a metamaterial spring can hold the pliers open (as here) than how it could hold them closed.