Dr. Shuai Lu

School of Architecture and Urban Planning, Shenzhen University



Performance-driven design methods of building shape: investigations based on acoustics and daylight





Shape design has significant influence on building performance, and architects that are unfamiliar with technical issues need efficient aiding methods so that they can improve building performance through appropriate shape design. However, existing simulation software and aiding tools are not accordant with the features of shape design including incompletion of design information, frequency of design modifications and multiplicity of design goals, and thus are not effective in helping architects with shape design. This research integrates digital methods including parametric model, performance simulation and machine learning, in order to propose a new aiding method of shape design, which contains functions including evaluation of the performance potential of shapes, tendency analysis of building performance against shape parameters, recommendation of similar shapes with better performance, shape optimization within a predefined region, and analysis of feasible ranges of shape parameters. Then the proposed aiding method is implemented by developing a tool taking 1) concert hall design and its acoustic performance, 2) office building and its daylighting performance, as examples, and the feasibility and effectiveness of the aiding method is tested. The results show that the aiding method is able to help architects with their shape design by providing multiple forms of suggestions from the viewpoint of performance, and thus architects can improve building performance through appropriate shape design. Meanwhile, under most circumstances the time delay of the suggestions is within 1 minute, so architects can get real-time feedback.