Computational design of cold bent glass façades

Abstract

Cold bent glass is a promising and cost-efficient method for realizing doubly curved glass façades. They are produced by attaching planar glass sheets to curved frames and must keep the occurring stress within safe limits. However, it is very challenging to navigate the design space of cold bent glass panels because of the fragility of the material, which impedes the form finding for practically feasible and aesthetically pleasing cold bent glass façades. We propose an interactive, data-driven approach for designing cold bent glass façades that can be seamlessly integrated into a typical architectural design pipeline. Our method allows non-expert users to interactively edit a parametric surface while providing real-time feedback on the deformed shape and maximum stress of cold bent glass panels. The designs are automatically refined to minimize several fairness criteria, while maximal stresses are kept within glass limits. We achieve interactive frame rates by using a differentiable Mixture Density Network trained from more than a million simulations. Given a curved boundary, our regression model is capable of handling multistable configurations and accurately predicting the equilibrium shape of the panel and its corresponding maximal stress. We show that the predictions are highly accurate and validate our results with a physical realization of a cold bent glass surface.

Using isometries for computational design and fabrication

“…Examples include flat pre-stretched tilings assuming a desired shape upon release [Guseinov et al 2017], auxetic micro-mechanisms [Konaković-Luković et al 2018], micro-mechanisms based on spiral springs [Malomo et al 2018] and deployable structures of a gridshell nature [du Peloux et al 2013;Panetta et al 2019]. Most recently, and relevant to our work, Gavriil et al [2020] used learning methods to panel architectural freeform skins by cold-bent glass, taking the restrictions of the material into account.…”

Computational design of weingarten surfaces

“…Very recently, [Gavriil et al 2020] developed a design system for glass facades where cold bent glass panels are used wherever possible. The judgement whether a panel can be produced via cold bending comes from a neural network that has been trained on a large number of simulation results.…”

Mixed Integer Neural Inverse Design

“…Perhaps photonic design is the front-runner field in using neural surrogate models for computational design [Jiang et al 2020]. Also, in computational fabrication we are seeing a surge in the use of NSMs, such as in computational design of cold-bent glass façades [Gavriil et al 2020], appearance-preserving tactile design [Tymms et al 2020], or fine art reproduction [Shi et al 2018].…”