A new 3D printer uses light to shape objects

Several schemes now exist to assemble, or “print,” three-dimensional objects through repeated 1D or 2D operations. But such point-by-point or layer-by-layer approaches often create ragged-edged figures, constrain their shapes, and can take hours. A group led by Brett Kelly and his thesis adviser, Hayden Taylor , both at the University of California, Berkeley, and collaborators at Lawrence Livermore National Laboratory have developed a 3D printing technique that circumvents those limitations. The method, called computed axial lithography, selectively solidifies parts of a photosensitive liquid polymer by exposing it to a dynamically evolving light field. In practice, the researchers project a series of computed 2D images from multiple angles onto a rotating cylinder of the liquid. The superposition of exposures produces an energy dose sufficient to solidify the material into the desired shape—the 40-mm-tall version of Auguste Rodin’s The Thinker pictured above, in this case—in a matter of minutes.

The process, which prints an object all at once instead of serially, was inspired by the image reconstruction procedures of computed tomography (CT). Those procedures, widely used in medical imaging, ensure that the light projections contain the necessary information to produce the object. Kelly and his colleagues used a video projector to output the computed intensity-modulated projections, which they then synchronized with the rotation rate of the uncured photopolymer material. The technique is scalable to large volumes of material—beyond the 50 mm to 100 mm sizes they have printed to date—requires no support structures, and can print soft biopolymers and hydrogels more quickly than conventional 3D printing techniques. It can also “overprint"—that is, construct components that encase other, preexisting objects. The latter capability might come in handy for mass customization of such objects as dental crowns, prosthetic devices, and grips on sporting equipment. (B. E. Kelly et al., Science, 2019, doi:10.1126/science.aau7114 .)