Digital Manufacturing of Lightweight and Efficient Structures via Reconfigurable Lattice Printing

The general approach of this project is based on a concept in which a continuous pre-fabricated chain of links is deposited and connected to itself at joint nodes, producing structural lattices in nearly arbitrary configurations that can be disassembled and reconfigured when desired. We will implement the concept as a sort of 3D printer for sparse lattices, in which a robotic manipulator arm lays down the links of a passive (non-robotic) continuous chain to create programmed truss lattices using a hierarchy of sub-modules all formed from a single chain of links and joints. The concept will be similar to modern 3D printers that lay down a heated and extruded thread of ABS thermoplastic (i.e. Fused Deposition Modeling) onto a substrate to create arbitrary 3D structures. Instead, we will lay down a chain of rigid links and joints (with appropriately designed connectors) to create lattice structures. In order to ensure structures are rigid while being built, sub-units consisting of links folded to planar triangles will be built upon and used to three-dimensional rigid substructures, which will be expanded to produce “large” (i.e. many-unit) target three-dimensional structures, all using a single compact general-purpose platform. In terms of space travel and exploration, this work will allow more efficient construction of structures that must be built on demand in space, by providing a means of custom fabrication of structural trusses and lattices based on a simple and generic base material. By starting with a densely-packed spool of linkage chain, structures can be efficiently “deployed” in customizable configurations and geometries to meet immediate fabrication needs, and then reused/reconfigured for successive applications. The approach can be implemented in any number of size scales and materials, thereby cutting across many application domains, from millimeter-scale segment lengths for small part construction, to meter-scale segments for civil structures. This project investigates the implementation of that concept, including the development of the chain and “printer” subsystems, and will include a number of stages of proof-of-concept prototypes. We are beginning by examining the design of the reconfigurable chain, with a primary goal to come up with a design that will allow a continuous, unbroken long series of links and joint nodes to be “extruded” and connected onto itself at joint nodes in order to produce much larger spatial target structures of the desired final geometry. Primary challenges related to that goal include the design of the connector nodes and rotational mobility of the nodes/links in order to allow rigid connections at the desired link orientations.
We have discovered a promising node design that will facilitate robotic assembly (e.g. through being self-aligning and not requiring complex movements or high forces), and have prototyped a large number of modular links (24.5m in length containing 700 nodes) that can serve as a fully functional testbed for investigating the following two important aspects of our project. First, the physical model consisting of sufficient links provides a tangible platform in order to efficiently explore different folding strategies. As shown in Fig. 2(a), different types of connection joints can be easily assembled by using our node design. Thanks to the prototype of the long chain, we have found several basic construction units and variations of their combinations that can be used to tessellate a variety of 3D shapes. Secondly, in contrast to the manual folding process demonstrated in Fig. 2(b), we have investigated the possibility of using a stable, open-source software package to automatically generate the folding path of different target structures. Eventually, the combination of the software can help us to convert STL file directly to 3D lattice structures composed of our proposed construction unit(s).