DIVING – Digital coupling of multi-scale analyses in modelling and monitoring
Safe operation of built infrastructure requires not only regular or continuous monitoring of structures, but also robust prognosis models for predicting the system or structure properties. By constantly adapting the prognosis models to the actual situation of the structures, it is possible to take deviations between modelling assumptions and reality as well as uncertain or strongly varying parameters into account within a numerical model. In addition to a corresponding monitoring of the structure by measurements, a suitable numerical model is required in this context. For this purpose, a multi-scale approach is used in the project. Within the framework of a hybrid modelling, critical areas that are identified on the basis of the monitoring data on the macro scale are mapped with a significantly higher resolution on the meso scale. The necessary information on the current condition of the structure is obtained by means of non-destructive testing methods.
For the information flow between the monitoring system and the numerical model, a digital linkage of both components has to be established, for which the approach of theory-inspired machine learning is used. In this way, a digital link is created between experimental investigations on the structure and the numerical structural model. At the same time, the basis for the control of mobile non-destructive measurement technology based on the prediction results of the model is laid.
The methodology to be developed includes the creation of numerical multi-scale models, tools for digital coupling and the experimental investigation of a structure on different scales. The individual components are first examined on structural elements that can be investigated in the laboratory under controlled conditions. In a second step, the methods will be validated on the reference structure of the priority programme SPP 2388.