Automatisierte Erzeugung geometrischer Modelle für die BIM‐basierte Gebäudesimulation

  • Automated generation of geometric models for BIM-based building simulation

Fichter, Eric; van Treeck, Christoph Alban (Thesis advisor); Nytsch-Geusen, Christoph (Thesis advisor)

Aachen : RWTH Aachen University (2022)
Dissertation / PhD Thesis

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2022


Driven by the digitalization of the construction industry as well as the expanding integration of regenerative energy systems in building technology, and the increasing use as a verification method, thermal building simulation is gaining in importance. In combination with simulations of thermo-fluid dynamics, buildings can be designed more economically and ecologically while ensuring thermal comfort. To an increasing extent, simulations are integrated into Building Information Modeling. In the context of openBIM, this includes the exchange of information using IFC files, which ideally contain a comprehensive data set for the creation of the simulation models. However, in practice, the data is often erroneous and incomplete. Consequently, the advantage of semi-automated model creation leading to a reduction in effort and increased use of simulations is lost. This also affects Space Boundary models, which are relevant for creating simulation geometries. Therefore, various methods for their downstream generation have been proposed, most notably in thermal building simulations. Disadvantages of these methods are, to varying degrees, the high demands on data quality, limitations in the accepted scenarios, and the insufficient attribution of the Space Boundaries. In addition, the requirements of Computational Fluid Dynamics are mostly not considered, although the coupled simulation can significantly increase the quality of the results. Concerning fluid dynamics simulation, only a few approaches for automated transfer of IFC data have been presented already. Existing methods often represent the building geometry in a simplified way or ignore the handling of faulty models. This thesis deals with a new method for the automated generation of Space Boundaries (IFC2SB). It is characterized by high robustness concerning model quality and comprehensive consideration of simulation- and scheme-specific requirements. In two introductory chapters, the thesis presents the state of the art in building simulation, Building Information Modeling, and Space Boundaries. Subsequently, the challenges and opportunities of existing approaches are discussed. Based on this, the new methodology IFC2SB is designed. It provides a variety of geometric and topological algorithms. Using example models and a prototypical implementation in the \CC programming language, the validity, robustness, and performance of the method are proven. Finally, the findings are discussed critically. Also, corresponding limitations, as well as potential research topics, are pointed out.


  • Institute of Energy Efficiency and Sustainable Building [312410]