Towards Voxel-Based Algorithms for Building Performance Simulation

This paper explores the design, coupling, and application of algorithms relying on grid-based or voxel-based representations of buildings. Such algorithms may encourage the adoption of building performance simulation in design practice by avoiding the timeconsuming and somewhat arbitrary process of manually partitioning a building’s interior into polyhedral spaces. They also support detailed visualizations that may open up new possibilities for design tools. This paper presents two grid-based methods: a hierarchical pathfinding algorithm to support detailed simulations of individual occupants, and a very simple heat transfer algorithm to motivate future alternatives to zone models. The coupling of the pathfinding and heat transfer algorithms demonstrates the need for a simulation framework supporting different time advancement patterns. The coupled model is applied to simulate a building with individual occupants that move between rooms and open windows in response to temperature changes.