FORMAL VARIATION EXPLORATION FOR PERFORMANCE-BASED BUILDING MASSING DESIGN OPTIMIZATION

Due to the requirement of design abstraction and simplification for early-stage building design optimization tasks, most applications of performance-based building design optimization are based on the designs generated using orthogonal and cubical geometries, which allows for simpler geometrical operations and sufficient design variability and differentiation in terms of geometrical configuration. However, these applications are only able to produce coarse solutions with room for improvement. In order to address this issue, this study proposes a method focusing on the formal variation in performancebased building massing design optimization to produce more detailed and precise solutions. In this study, two formal variation algorithms are developed using a volume-based and a boundary-based approach, which can modify the input orthogonal geometries produced by EvoMass, a design tool for building massing optimization and exploration. To demonstrate the efficacy of the proposed approach, a case study is presented, which shows that the use of the two formal variation algorithms can further improve the design performance and also allow designers to extract more accurate information related to the design strategies and performative design implications.