Performance Based Generative Facade Workflow for Large Scale Projects

Over the last two decades, in what has been coined “The Digital Turn”, the introduction of parametric design software has afforded increased accessibility to practitioners within the profession for creation and manipulation of complex 3D geometries (Carpo, 2013). These tools have been adopted with great enthusiasm, but their use has generally trended towards the generation of fluid and organic forms with mostly aesthetic drivers. As environmental regulations and sustainability standards become more stringent, design teams can benefit from leveraging a workflow that engages simulation and analysis tools from within a parametric modeling environment to affect the trajectory of the design from the earliest stages.

The exterior envelope of Kuwait Children’s Hospital (KCH) is presented as a case study where performance guided the development of an aesthetic concept and contributed to the generation of a facade system. The project adopted an experimental methodology intended to simultaneously maximize occupant comfort while increasing the effectiveness of the project design team. By streamlining the movement of data from various design, simulation, and documentation platforms, the team reconciled a conceptual aesthetic consideration leveraging crystalline growth patterns with daylight performance. This process allowed the project team to generate solutions responsive to visual comfort originating from contextual restrictions relative to climate and orientation. A parametric shading scheme was developed on an artificial fractal with quasi semi-similar structures, and with the specific sizing of facade units controlled by their impact on daylight performance in the space.

Design and analysis was facilitated by an interconnected workflow consisting of custom-scripted geometry creation (written in Python and C#) which was then fed into an analysis feedback loop accessing Radiance and Daysim via the DIVA and Honeybee plugins available for Grasshopper. The resulting design geometries were made available to the rest of the project team via the web-based data exchange platform Flux.io and translated into the Revit production environment through a series of custom scripted Dynamo components accessing the Revit API. This method of data exchange allowed for smooth collaboration from remote locations across different time zones and enabled precise data translation between different computational applications and among different disciplines. This workflow was proven to save time and increase the accuracy of simulation and modeling translation, while enabling cross-referencing of data between otherwise non-compatible software, resulting in the deployment of a performative facade proposal.