Computational Design of Cold Bent Glass Fa\c{c}ades

TL;DR

This paper introduces a real-time, data-driven design tool for cold bent glass facades that enables non-experts to create feasible, aesthetically pleasing curved glass surfaces by predicting stress and shape using a neural network trained on extensive simulations.

Contribution

It presents a novel interactive, parametric design approach utilizing a differentiable Mixture Density Network for accurate, real-time stress and shape prediction in cold bent glass facade design.

Findings

Model achieves high prediction accuracy for equilibrium shapes and stresses.

Designs can be optimized to meet aesthetic and safety criteria interactively.

Physical prototype validates the effectiveness of the proposed method.

Abstract

Cold bent glass is a promising and cost-efficient method for realizing doubly curved glass fa\c{c}ades. They are produced by attaching planar glass sheets to curved frames and require keeping the occurring stress within safe limits. However, it is very challenging to navigate the design space of cold bent glass panels due to the fragility of the material, which impedes the form-finding for practically feasible and aesthetically pleasing cold bent glass fa\c{c}ades. We propose an interactive, data-driven approach for designing cold bent glass fa\c{c}ades 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. 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 predictions are highly accurate and validate our results with a physical realization of a cold bent glass surface.