SCF '18: Proceedings of the 2nd Annual ACM Symposium on Computational Fabrication

Industrial knitting machines form fabric by manipulating loops of yarn held on hundreds of hook-shaped needles. Transfer planning algorithms generate a sequence of machine instructions that move loops between their current needles and given target needles. In this paper we describe how to compute the run-time cost of a transfer plan in terms of machine passes, and compare the plans generated by several existing and new transfer planning algorithms under this metric over a a large benchmarking set of transfer operations taken from example flat lace patterns, along with synthetically generated patterns.

A predominant issue in the design and fabrication of highly non-convex polyhedral structures through self-folding, has been the collision of surfaces due to inadequate controls and the computational complexity of folding-path planning. We propose a method that creates linearly foldable polyhedral nets, a kind of unfoldings with linear collision-free folding paths. We combine the topological and geometric features of polyhedral nets into a hypothesis fitness function for a genetic-based unfolder and use it to map the polyhedral nets into a low dimensional space. An efficient learning strategy is used to optimize the fitness function to produce the optimal nets. We experimentally demonstrate that the proposed method can find linearly foldable nets for highly non-convex polyhedra with substantial complexity. The technique presented in the paper will provide a powerful tool to enable designers, materials engineers, roboticists, to name just a few, to make physically conceivable structures through self-assembly by eliminating the common self-collision issue. It also simplifies the design of the control mechanisms when making deployable shape morphing devices. Additionally, our approach makes foldable papercraft more accessible to younger children and provides chances to enrich their education experiences.

This paper presents an exploration of computer-controlled embroidery design and stitching with the goal of making purposeful, precision changes to material properties of a base elastic textile. Two techniques are proposed for adding stitches through designed microstructural cells and stitch-level planning. For the latter, a novel path planning algorithm is proposed to serialize stitches that is similar to a greedy solution for the travelling salesman problem with a set of domain-specific constraints that dictate edge cost. We show the efficacy of the concept through a set of simple design examples that undergo mechanical load testing and discuss the value of the technique in future applications using computational design.

Highly specified digital control tools for the customization of knitted fabrics is an existing technology, widely used industrially by trained knitters. Some fabrication challenges inhibit the extent of use of digital customization for knitted products on a mass scale, specifically when complex multiple structured knits are involved. This is due, in part, to physical changes that occur in the overall fabric dimensions, when stitch combinations with different physical attributes are combined within the same fabric. Existing computational tools fail to assist in the simulation and prediction of these overall deformations and the industry as a whole relies on the expertise of highly trained individuals. The outline shape of the fabric is of specific importance as it is commonly preconfigured to a specific shape and dimension that must be reproducible. In this work, we propose a computational parametric tool for digitally designing and industrially producing knitted fabrics, creating a direct link between design and manufacturability. The tool includes an integrated physical simulation component for estimating the fabric deformation. It allows users to visualize and control compensations in the fabric structure, to better match between the initial graphic intent of the design and the actual physical knitted fabric outcome. Our approach aims to reduce the number of iteration cycles for knitting material samples, especially when knitting highly varied designs.

We present Compositional 3D Printing, recasting the 3D printer as a tool for expression that responds to real-time design decisions, analogous to composing a piece of music using a mixer. Our paradigm supports a wide range of inputs and interactions for designers, to be used in the moment; not only before printing, but anytime during production. We propose the design space of this digital fabrication paradigm, and outline methods and technical details with which researchers and practitioners can expand this space.

The recently revived interest in thin shell architecture demands an efficient method to quickly prototype non-developable surface models for design evaluation. This paper focus on improving the traditional papercraft technique of approximating smooth surfaces with discretised planar meshes. The necessary step to unfold strips of planar faces often creates many separate pieces. These pieces require tedious manual work in gluing and assembling them together. The addition of gluing tabs also results in an aesthetically unpleasing underside of the surface.

This paper proposes a double-shell construction to overcome the challenges of edge jointing and eliminate the need for gluing tabs. A double-shell is created using two sets of unfolded planar strips that were split at different seams. When the two sides are folded and glued back together, the folded connections bridge the seams on another side and vice versa. Spray-on contact adhesive can be globally applied to the underside of each set, thereby speeding up the assembly process.

Models of trimmed and untrimmed NURBS surfaces and triangular mesh surfaces are constructed using this method as demonstration. Quantitative and qualitative effect of model size and paper thickness are explored to provide baseline application parameters. As a result, this method allows anyone with a laser or drag knife cutter to rapidly create mesh models that are easy to assemble and aesthetically clean on both sides.