Improving interoperability between architectural and structural design models: An industry foundation classes-based approach with web-based tools

摘要

Medium and large construction projects typically involve multiple structural consultants who use a wide range of structural analysis applications. These applications and technologies have inadequate interoperability and there is still a dearth of investigations addressing interoperability issues in the structural engineering domain. This paper proposes a novel approach which combines an industry foundation classes (IFC)-based Unified Information Model with a number of algorithms to enhance the interoperability: (a) between architectural and structural models, and (b) among multiple structural analysis models (bidirectional conversion or round tripping). The proposed approach aims to achieve the conversion by overcoming the inconsistencies in data structures, representation logics and syntax used in different software applications.\ud\udThe approach was implemented in both Client Server (C/S) and Browser Server (B/S) environments to enable central and remote collaboration among geographically dispersed users. The platforms were tested in four large real-life projects. The testing involved four key scenarios: (a) the bidirectional conversion among four structural analysis tools; (b) the comparison of the conversion via the proposed approach with the conversion via direct links among the involved tools; (c) the direct export from an IFC-based architectural tool through the Application Program Interface (API), and (d) the conversion and visualization of structural analysis results. All these scenarios were successfully performed and tested in four significant case studies. In particular, the conversion among the four structural analysis applications (ETABS, SAP2000, ANSYS and MIDAS) was successfully tested for all possible conversion routes among the four applications in two of the case studies (i.e., Project A and Project B). The first four steps of natural mode shapes and their natural vibration periods were calculated and compared with the converted models. They were all achieved within a standard deviation of 0.1 s and 0.2 s in Project A and Project B, respectively, indicating an accurate conversion.