SCF '19: Proceedings of the 3rd Annual ACM Symposium on Computational Fabrication

Full Citation in the ACM Digital Library

Human-in-the-loop fabrication of 3D surfaces with natural tree branches

We propose a design and fabrication system leveraging the diverse shapes of natural wood. With our system, precise geometries are fabricated from non-standardized, naturally curved branches. In this way, low-valued tree branches are up-cycled to a construction material. The process is implemented as follows. Taking a set of branches and a user-defined parametric target surface, the system makes a reciprocal pattern of curves that lie on the surface. Then, it automatically matches the shapes of the scanned branches with these curves. There was a similar attempt to build a large beam structure from tree trunks using industrial equipment such as cranes and robot arms. In contrast, we target smaller surface structures from tree branches using more accessible tools (2.5D CNC) with human-in-the-loop operation. We ask the user to manually place a branch to a defined orientation with the help of audio-visual guidance. Then a CNC-machine can mill out the joint. Finally, the structure is assembled by hand without screws or adhesives.

Exploring mechanical meta-material structures through personalised shoe sole design

Mechanical meta-material structures (MMS) are designed structures with mechanical properties not found in ordinary materials. MMS can now be created far more easily using digital manufacturing. We explore how different MMS can be combined, through the design of a shoe sole. Thereby showing the potential of using MMS to create personalized and sustainable footwear. We analysed the phenomenon of foot deformation and mapped different structures with different behaviours to meet the needs of different feet. Consequently, a shoe sole was generated by an algorithm and 3D printed in one single material with multiple properties (e.g. stiff and soft) and responsive behaviour, making it easy to recycle. We report the design phases which required using six types of software. Our findings reflect the complexity of this process given the limited availability of software tools that support it. We conclude with a list of requirements regarding tools to further explore MMS.

Automated reconstruction of smoothly joining 3D printed restorations to fix broken objects

In this work, we provide an approach to automatically reconstruct a 3D printed restoration piece for a broken object from 3D scanned meshes of the broken object and an original counterpart. Our approach provides two contributions to reconstruct a restoration with a smooth join to the broken object, necessary for object functionality such as liquid containment, injury prevention, and visual aesthetic. As our first contribution, we leverage the original counterpart mesh to grow an exterior surface for the restoration piece that approaches the broken object within a small tolerance. As our second contribution, we project the exterior surface boundary onto the broken object to create a fracture surface boundary whose vertices satisfy the constraints of proximity, normal alignment, and tangency to vertices on the exterior surface boundary. Our approach prevents artifacts of volumetric Boolean subtraction, such as floating components and thin long slivers at the join, and avoids ruts at the join region introduced by Euclidean distance thresholding. We show 3D printed restoration results for 14 objects and 3D printable results for 8 objects.

Rough carving of 3D models with spatial augmented reality

Carving is a subtractive process where we get the shape by removing materials. While most people can get roughly the right intended shape, it is usually challenging not to over-cut the model. We propose a method that helps an unskilled user to carve a rough physical replica of a 3D model using the minimum number of cuts while only using manual cutting tools. The method starts by analyzing the input 3D model and generates the minimum set of cutting steps that remove most of the material. Then using a projector, we project the instructions sequentially onto a block of material to guide the user in performing them. We use the projector-camera setup to 3D scan the object after cutting and automatically detect the changes to reflect them on the digital model. We demonstrate a complete system to support this operation and show several examples of manually carved 3D models while using the system.

Echidna: mixed-domain computational implementation via decision trees

Custom mechatronic devices offer personalized functionality, but also come with many non-functional requirements that are unfamiliar to those inexperienced with electronics such as current draw and servo power. The Echidna prototype system enables non-electrical engineers to move from conception to implementation with their mechatronic ideas by generating and searching through a design space that automatically fills in supporting requirements, such as PCB placement and wiring, around their functional specification. The space is modeled as a decision tree whose root is the user's list of lights, motors, sensors, and other functional components that need to be connected, powered, and controlled. Once found, a complete and valid design can be used to synthesize geometry for 3D printing, circuits, and firmware resulting in a set of "plug and fabricate" files for creating their device. We demonstrate how Echidna realizes several designs and discuss how it can be further customized to task-specific applications.

Volumetric Michell trusses for parametric design & fabrication

We present the first algorithm for designing volumetric Michell Trusses. Our method uses a parametrization-based approach to generate trusses made of structural elements aligned with the primary direction of an object's stress field. Such trusses exhibit high strength-to-weight ratio while also being parametrically editable which can be easily integrated with parametric editing tools such as Autodesk Fusion. We show a number of examples that demonstrate that the output of our algorithm produces truss structures that are aligned with an object's underlying stress tensor field, are structurally sound and that their global parametrization facilitates the creation of unique structures in a number of domains.

A method to evaluate the formability and fluidity of concrete based materials for 3D printing

Use of 3D printers for construction brings a new possibility of creating walls with shapes that were not possible with conventional construction methods. However, existing construction materials can not be used for 3D printing as they are, and the exploration for new materials is the key to the realization of this concept. Formability and fluidity are the key factors in evaluating whether or not materials are suitable for 3D printing, including use within the field of architecture. A standardized quantitative method to evaluate construction materials is required for the comparison between materials for various projects. For example, the concrete slump test is one of the universal standards that is used to evaluate the quality and suitability for a specific construction method. In the emerging field of construction with 3D printing, however, a standardized method is yet to be established. This paper proposes an efficient method to evaluate the formability and fluidity of concrete-based materials, while only using simple instruments. The proposed method and instruments were tested by evaluating geo-polymer-based concretes, which are relatively difficult to evaluate due to its high intrinsic viscosity. By testing with these materials, we aim to show the methods ability to be applied to a wide range of materials. Notably, materials that were evaluated to be favorable under this method were confirmed to be suitable for use in 3D printing robots, being able to be used for building large structures.

3D-printed formwork for bespoke concrete stairs: from computational design to digital fabrication

Non-standard stairs have an important role in architecture, but their complex details and three-dimensionality pose significant fabrication challenges. One of the preferred materials for custom stairs is concrete, which can be cast in complex shapes, the only limitation in this case being the fabrication of the necessary formworks.

This paper reviews the opportunities and challenges of using 3D-printed formwork for fabricating custom concrete stairs with complex geometries (Fig. 1). 3D printing can unlock an entirely new vocabulary of shapes for concrete structures, previously unavailable with traditional formwork systems. Only a minimal amount of 3D-printed plastic is required to deliver a very thin, stable shell. With these formworks, complex topologies can be achieved in concrete elements. Such elements can optimize the structural performance or improve functional aspects, as well as introduce a radically different aesthetic.

Computational laser forming origami of convex surfaces

Laser forming is a fabrication method that uses laser to fold sheets into 3D structures. To overcome the limitations in the traditional practice that relies on tedious manual design, this paper advances laser forming by developing computational methods that procedurally convert a polyhedron P into laser formable 2D patterns and folding instructions or report that P is not laser formable. Due to the limitation of the low-cost laser cutter considered in this paper, we will focus our discussion on laser forming convex surfaces. A 3D surface is called convex if the entire surface lies on the boundary of its convex hull. Our theoretical analysis shows that, even for convex surfaces, the laser formability can be expensive to determine. We then present a framework that efficiently computes patterns and motion instructions for laser forming convex surfaces. An end-to-end laser forming pipeline is presented with several fabrication results to demonstrate the capability and current limitations of the software and hardware framework.