January 12, 2016
By Daniel Inocente (DC) & Kevin Vandeman (DC)

The Idea Fellowship continues as the Holistic Design Workflow team makes advances in the research and application of computational strategies. As buildings and the industry become increasingly technological, architects are challenged to continuously develop new ways of mastering the intricacies across ever-changing elements and systems. The architecture of today has not only been infused with technological growth, but it has also become dependent on it. Technology has enabled numerous design processes and yet not one single process can fully solve the thousands of contingencies that present themselves as a project evolves. Today, the design of a building is a challenge of bringing together various technological systems, outside of our domain, that rely on computational processes.

Through computational thinking, we can influence and lead the necessary cross-disciplinary collaborations, approaching architectural problems with analysis and automation strategies . Establishing critical partnerships with industry leaders in other disciplines (fabrication, computation, engineering, science, etc.) is only the beginning toward achieving a holistic vision.

 To highlight the focus of the research and the trajectory, we want to investigate the role of computation and simulation in design today. Computation allows us to describe geometries, analyze performance, specify materials and simulate physical behaviors. In architecture and other disciplines, computation and simulation have been used extensively, in many cases enabling new discoveries and shaping the way we give form to matter. By understanding the foundations of these technologies, mathematically and abstractly, the work we do moves toward increasing efficiencies and forms of expression. We have many questions of computation’s influence on our work, but for now, we are using the tools and knowledge that we have to continue answering these questions.


Through iterative design, our pilot project takes advantage of dynamic physics to study the behavior of surface geometry under tension, gravity and topological deformation. These studies resulted in finding an optimized distribution of vertices for mesh generation and topological optimization. There is craft in guiding and automating how geometrical performance is achieved, this is another topic we should look at critically. Computational craft is becoming more prevalent than ever before and this project aims to utilize some of these shifts. The tools used allowed us to both digitally craft and analyze geometrical decisions in the process. After finding a fit geometry, the model then goes through parametrization, where the mesh is converted into a NURBS surface. This NURBS (Non-Uniform Rational B-Spline) surface is calculated through equations of multiple curves with small degrees of precision. The resulting geometry allows us to perform further analysis of our design, toward additional structural and material investigation.


Curvature analysis in this scheme was used to analyze areas where curvature might affect structural performance. The degree of curvature along the surface is constantly changing, which makes identifying high-stress areas critical in developing a structural logic. Through these studies, we identified the guiding splines which would help achieve both a structural logic and an aesthetic quality to best describe the geometrical complexity.  Iterative thinking in the generation of the guiding splines produced several options, from which we selected one solution. By using Grasshopper, we can make changes to the geometry and continuously change how we approach the problem. The end result of this investigation yielded optimized spacing and flow of structural guides for translation into Inventor, where structural modeling begins to take place.


A major strategy in our design workflow is data management. The structuring of disciplines with their corresponding systems and sub-systems becomes a key component in how we communicate our design. This structure will enable us to implement a master- model approach, through which we can maintain all assembly and part associations (Figure 2) from the early stages of a design engagement. The master model is a holistic representation of any design, where each discipline and its corresponding elements can be stored and managed. The active contribution of each collaborator’s efforts will be logistically managed by giving the design team a hierarchy of access to this master model. Each author/collaborator will also have the ability to isolate the individual element, or groups of elements, they are working with. With the ability to isolate the driving design elements and their detailed components, all users can navigate and manage the master model independently, without the need to load the entire model.

The increased complexity of architectural and engineered elements makes the management of all the interrelations critical in the process. Using a systems engineering approach will make the management of heavy sets of data increasingly more manageable. Inventor allows for multiple modeling techniques, including the ones illustrated (Figure 1). Through these modeling techniques, the use of design drivers – such as design surfaces, bodies, wires, vertices and working planes – allows for high-level control of all model elements. Once we have established the design drivers, we can then publish and distribute these drivers across the design team, to continue developing and refining, while maintaining a high level of control over the entire project.


Through the use of derived geometries, which can maintain associative relationships with structural and mechanical elements, all elements that are generated can be continuously updated. In this study, a design surface and its curves are derived in order to generate the structural members which are all linked to parameters for sizing and position. These members are then sent to the engineer for analysis before continuing to detail additional assemblies.


The process for this structural system begins with instantiating structural members along the diagonal curves. This model gives valuable information to the structural engineer to begin the process of optimization, detailing and sizing each member. Through our research fellowship, we have established a relationship with engineers in order to develop a workflow that bridges together design and engineering processes. By analyzing the relationship between form and structure, challenges and opportunities can be identified during our collaboration.


The use of ETFE (Ethylene tetrafluoroethylene)in the design for this pavilion allows for a great amount of flexibility and a reduction in structure, due to the lightness of the system and its ability to span large distances. We have established a relationship with Vector Foiltec, an industry leader and pioneer in engineering, fabricating and installation of ETFE systems, to continue our research.  Vector Foiltec has the ability to work with a large range of model types, such as Revit, Rhino, Inventor and Catia in their workflow which allows us to be flexible with how we issue models. In addition to this, their use of Tekla, a structural modeling platform, will allow us to streamline the structural engineering and calculation process. The design model provided includes the overall geometry and the surface area for each one of the ETFE pillows, including additional design drivers. Proposals from each of our collaborators will be applied to the design for further coordination between the structure and ETFE cladding. As we narrow down the solutions for achieving structural performance, we will continue to investigate how ETFE can be treated to generate added environmental performance in the overall design.


Vector Foiltec will be providing custom detailed aluminum extrusions, which include additional mechanical air inflation systems for maintaining the rigidity of the ETFE pillows. Due to the design’s complexity, the ETFE system will be made up of unique members at every intersection and optimization will have to take place to minimize energy usage. All of the detailed components generated by our collaborations will be incorporated into our design model and as the project evolves we aim to achieve a holistic approach to this design. (mgx)

Posted in Research, Research for Research Page

Add your comment