Full text
| |
| Author(s): |
Gabriele do Rosario Landim
Total Authors: 1
|
| Document type: | Master's Dissertation |
| Press: | São Carlos. |
| Institution: | Universidade de São Paulo (USP). Escola de Engenharia de São Carlos (EESC/SBD) |
| Defense date: | 2019-05-13 |
| Examining board members: |
David Moreno Sperling;
André Luís de Araujo;
Maria Gabriela Caffarena Celani;
Márcio Minto Fabricio
|
| Advisor: | David Moreno Sperling; Cláudio Fabiano Motta Toledo |
| Abstract | |
This research is contextualized in the Computational Design field, an area that explores computation as an architectural design tool. To declare and develop projects, it is common to use visual programming languages. However, limitations in the use of visual languages are well known, such as the constraints found in the flexibility and scalability of codes. To extend access to problem-solving methods in these systems, users often resort to the use of textual programming languages, which allow the use of more control structures. However, it is common for architects to be induced to make changes to processes to accommodate the lack of control and unexpected design results. This happens partly due to the lack of clarity in how to use programming and the full potential of computational methods. In addition, there is a lack of empirical evidence using real architectural designs that support discussions about the future of programming tools for architects. This research investigates the comparison between the use of visual and textual programming languages and aims to suggest guidelines to improve the programming interfaces currently used in the area. To achieve this, we perform analyzes based on evaluation metrics of programming languages and Cognitive Dimensions (GREEN; PETRE, 1996). In addition, we explored a Performance-driven Architectural Design as a study case, the Vancouver Academic Building, from Perkins+Will. The office initially implemented the design using Grasshopper visual language to optimize the design of a façade by exploring the balance between aspects of illuminance and energy efficiency. We reimplement the case study using a textual language and the Rosetta programming environment (LOPES; LEITÃO, 2011). As we attempt to solve the existing gap in the comparison of workflows between visual and textual languages using a real project, we select this case study because it covers a performance-oriented process rather than a parametric design only. There are fundamental differences found between these two workflows in relation to scalability and data flow strategies since the performance-driven one requires more control structures such as iteration and recursion. We identified relevant differences between both languages and the potential in the use of textual ones in the significant increase of code and model scalability; increased reliability of project performance results and better control and clarity of the process. We also point to the inherent qualities of interaction and expressiveness of visual languages. From the theoretical-practical cross, we indicate the guidelines that make up the extent to which the preferential use of textual languages and hybrid languages (which integrate both) can improve the reliability of the Performance-driven Design, as well as the possibility of applying clearer and more effective methods in the development of the Algorithmic Design. (AU) | |
| FAPESP's process: | 16/05427-0 - Data use in the design process: visual programming and information modeling |
| Grantee: | Gabriele Do Rosário Landim |
| Support Opportunities: | Scholarships in Brazil - Master |