Strategies for Encoding Multi-dimensional Grading of Architectural Knitted Membranes

This paper introduces the knit as a special case of functionally graded materials. The ability to specify knitted material both structurally and materially allows for multi-dimensional grading of the knit. This ability to functionally grade knit is defining for knitted material’s relevance to architecture and construction, as it allows for combining performance multiple criteria in one material by smoothly transitioning between the properties in a single manufacturing act. However, the scaling of knit to architectural scale brings computational challenges for automating material specifications. This paper investigates the methods for solving these computational challenges when operating with multi-dimensional material gradings that increase the scale and resolution of knitted textiles and therefore increase the complexity of their encoding.
The method for multi-dimensional material grading of CNC-knitted membranes is developed through the project of Zoirotia, a large-scale textile installation (Fig. 1). Zoirotia is made of 88 unique membranes, where each is uniquely graded at the levels of both material structure and yarn composition. This results in surfaces of varied expansion properties and color transitions across the entire structure, achieved by changing yarn morphology and material. The project is understood as a testing ground for solving the challenge of large volume material graded specification through simplified numerical data and the process of binarisation. A dithering technique is used to translate rich design-driven gradients for material specification into binary gradients for CNC fabrication. This paper presents the gradient numerical map as a solution for handling the material specification spanning across the multi-element structures.