A PIXELS-BASED DESIGN APPROACH FOR PARAMETRIC THINKING IN PATTERNING DYNAMIC FAÇADES

Abstract

In today’s Architectural design process, there has been considerable advancements in design computation tools that empowers designer to explore and configure the building façades schemes. However, one could formally argue that some processes are prescribed, lacks automation and are only for the purpose of visualizing the aesthetic design concepts. As a result, these design concept explorations are driven manually to exhibit variations between schemes. To overcome such limitations, the development presented here describes a proactive approach to incorporate parametric design thinking process and Building Information Modeling (BIM). This paper reports on an ongoing development in computational design and its potential application in exploring an interactive façade pattern. The objective is to present the developed approach for exploring façade patterns that responds parametrically to design-performance attractors. Examples of these attractors are solar exposure, interior privacy importance, and aesthetics. It introduces a paradigm-shift in the development of design tools and theory of parameterization in architecture. This work utilizes programming script to manipulate the logic behind placement of faced panels. The placement and sizes for the building facade 3D parametric panels react to variety of Analytical Image Data (AID) as a source for the design-performance data (e.g.: solar exposure, interior privacy importance, and aesthetics). Accordingly, this research developed the PatternGen(c) add-on in Autodesk ® Revit that utilizes a merge (or an overlay) of AID images as a source to dynamically pattern the building façade and generate the façade panels arrangement rules panels on the building exterior. This work concludes by a project case study assessment, that the methodology of applying AID would be an effective dynamic approach to patterning façades. A case-study design project is presented to show the use of the AID pixel-gradient range from Red, Green and Blue as information source value. In light of the general objectives in this study, this work highlights how future designers may shift to a hybrid design process.