This thesis presents a new framework for Web-based parametric modeling for design collaboration, towards allowing multiple users to work on the shared Web-based model in the process of drafting, design, simulation, and optimization. The framework consists of a WebGL-based model Viewer (Autodesk Forge Viewer), two visual programming tools, Grasshopper and Dynamo, and two software prototypes developed for this thesis. The prototypes, one for Grasshopper, the other for Dynamo, provide the communication between the Viewer and the visual programming tools. The Web-based 3D model and design data can be viewed in the Viewer. The Web-based model is controlled and modified through the visual programming tools using the prototypes. The embodiment of the Web-based information technology, WebGL and networking, makes it possible for users to view the Web-based model, collaborate, and participate in modeling through Web browsers. The full-fledged visual programming environments, Grasshopper and Dynamo, enable users to interact with the parametric Web-based model; the plugins of the visual programming tools allow users to implement building energy performance (BEP) simulation and optimization while the Web-based model enables collaborative design exploration. Two case studies with three tests each were conducted on a simplified residential building model. In Case Study 1, two simulated users (actions done by the author) tested the parametric capabilities of the shared Web-based model using Grasshopper and Dynamo. The basic geometric transformations including scale, translate, and rotate were tested in Tests 1.1, 1.2, and 1.3 respectively. In the Case Study 2, the collaboration of two Grasshopper users on the shared Web-based model in the process of optimization for different building performance objectives -in terms of daylight, energy use, and roof coverage- was tested. Case Study 2 was conducted through three tests of optimization. (i) In Test 2.1, the objectives were maximum preferred daylight and minimum roof shape with shading. (ii) In Test 2.2, the objectives were minimum energy and minimum roof shape with shading. (iii) In Test 2.3, minimum energy use was the first objective, maximum preferred daylight and minimum roof shape with shading was the second objective. The findings demonstrate that the framework successfully provides interoperable Web-based architectural models which could be controlled parametrically through different visual programming tools. The framework also constitutes an environment of collaboration between Grasshopper users during daylight, energy, and integrated daylight and energy optimization along with the optimization of the roof shape. In both collaboration cases, the users retrieve data from the Web-based model to update their visual programming files. Major technical limitations of the framework are also found and discussed. Future work includes automation of the data flow through Web to the local visual programming files and contributing to the interoperability solutions by incorporating other tools.
