Representation and reasoning for integrated conceptual design of building structures

During conceptual design the most salient characteristics of a design artefact are defined. A team of specialists is usually required to carry out this process. For buildings, the architect is the lead designer and his/her work is supported by engineers and other specialists. The distribution of work among specialists with different backgrounds and priorities and the lack of a shared understanding of each participant's view usually result in inefficiencies, redundancies, omissions and errors in the design. This translates in cost overruns, poor quality and lack of competitiveness of the building industry. Computer programs for performing structural analysis and detail design calculations have been in the market and used by practitioners for many years. However, there is still a lack of computer support to assist properly engineers in the conceptual design of building structures. One explanation for this lack of support lies in the fact that during conceptual design the engineer's work flow is highly dependent on the amount, quality and type of information that is exchanged with the architect. To address this problem, the interactions between architects and structural engineers during early design are studied and formalized in a model of conceptual design of building structures. In this model, two stages of structural support are identified which are differentiated by the type and refinement of the architectural design representations. The first stage corresponds to the earliest phases of the architectural design process in which the architectural representations are tentative, ambiguous and imprecise, whereas in the second stage an unambiguous but still imprecise three-dimensional computer model of the building architecture is made available to the engineer. This research project concentrates on the second stage of computer support. A methodology for architecture/structure integration is devised for this stage. For the implementation of this methodology a computer representation integrating the building architecture and the structural system is developed. Synthesis algorithms are also proposed that rely mostly on geometry and topology of the architecture and the structure to help identify structural opportunities, detect potential structural problems and respect constraints coming from the building architecture. A proof-of-concept software prototype is developed and implemented, and two test cases demonstrate the validity of the proposed approach for supporting conceptual structural design. It is expected that the results from this research project will lead to more integrated and efficient conceptual structural design practice for improved building performance.