Project Details
Projekt
Representation and transfer of detailing operations on the basis of formal graph transformation systems
Applicant
Professor Dr.-Ing. André Borrmann
Subject Area
Structural Engineering, Building Informatics and Construction Operation
Term
from 2016 to 2024
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 271444440
The process of detailing plays an important role in the design process, as design decisions are made (often implicitly) that have a significant impact on the resulting building in terms of energy efficiency, usability, realisation costs, etc. The chosen level of detail correlates with the progress of the planning process; the fundamental decisions necessary at the beginning of the design process are represented by a correspondingly abstract design, which is successively enriched with details resulting in a more concrete model. The development of the meta-model for the formal specification of levels of development (LOD) conducted in the first phase of the subproject, provides an important basis for supporting model-based design processes in early phases. However, the question of how detailing decisions can be transferred between different design variants remains unsolved. At the same time, this aspect is of decisive practical importance, since often one design variant is worked out in detail, but later, due to changed design boundary conditions, another variant is followed up and has to be detailed again. In the course of the second phase of the subproject, it will be investigated how detailing processes can be formally recorded and flexibly transferred to alternative variants. Methods based on graph transformation systems will be developed and their coupling with parametric design systems will be investigated. The concept pursued is that a graph represents a section of the building model in a certain level of detail and the application of graph transformation rules manipulates this graph in such a way that it subsequently describes this section in a finer level of detail. In this way, it is possible to represent detailing processes and the associated knowledge in a computer-interpretable way and to make it reusable. A complete automation is explicitly not aimed at. Instead, the user has a number of detailing rules at her disposal from which she selects the most suitable one on the basis of experience and contextual knowledge. The methodology is elaborated in a generic manner and implemented exemplarily for a limited number of design entities and use cases. The proposed method is prototypically implemented by combining the graph transformation methodology with a parametric BIM system. For the realization of this subproject, an intensive cooperation with all other subprojects is planned, e.g. for developing procedures of decision support and documentation, for exploring suitable graph representations, for integrating simulation-relevant aspects related to energy and structure as well as for validating the overall methodology by means of the reference project.
DFG Programme
Research Units
