Large scale hybrid models for damage detection in arbitrarily shaped concrete structures

TUM2在第一个资助期的工作产生了基于fe的损伤定位方法。对于我们的基准测试的给定维度和使用频率，它被证明工作得非常好。具体而言，实现了以下目标：-利用尾波开发基于fe的损伤定位方法。-开发了一种允许任意几何形状描述的非结构化网格的使用公式-通过嵌入两种先进的解决算法，改进了已建立的尾轴损伤定位问题的解决算法-将新的基于fe的方法成功应用于实际实验。使用60kHz的频率代表了基于尾波的灵敏度损伤定位的新方法。它将可用频率范围扩展到更低的频率范围，更适合在大型结构中应用。在实际实验中的应用取得了很好的效果，可以很好地检测和区分出多个紧密间隔的裂缝。沿着这条道路，目前建议的目标是巩固和具体发展我们的方法。㈠捕捉复杂几何形状对尾波测量的影响；㈡根据测量到的尾波数据确定结构模型的参数。这将通过进一步调整该项目第一阶段开发的技术来应对现实世界的挑战来实现。（iii）找出现场损伤检测的阈值。（iv）开发数字孪生，使用来自结构的所有测量数据来识别和量化特征模型参数。数字孪生还将用于研究结构力学对尾波信号的影响。

The work of TUM2 in the first funding period results in a FE-based damage localization methodology. It was shown to work very well for the given dimensions and used frequencies of our benchmarks. Specifically, the following was achieved:- Development of an FE-based damage localization methodology using coda waves.- Development of a formulation for the use of unstructured meshes that allows description of arbitrary geometries- Improvement of the established solution algorithm for the coda damage localization problem by embedding two advanced solution algorithms- Successful application of the novel FE-based methodology to a real experiment. The used frequency of 60kHz represents a novelty in sensitivity-based damage localization with coda waves. It extends the range of usable frequencies into a lower frequency range that is more suitable for application in large structures. The application in real experiments showed very good results, as several closely spaced cracks could be successfully detected and distinguished.Following this path, the goal of the current proposal is to consolidate and specifically develop our methods to.(i) capture the influences of complicated geometries on the coda measurements(ii) identify the parameters of the structural model based on the measured coda wave data. This will be accomplished by further adapting the techniques developed in Phase 1 of this project to meet real-world challenges. (iii) find thresholds for damage detection in the field.(iv) Develop a digital twin that uses all measured data from a structure to identify and quantify characteristic model parameters. The digital twin will also be used to investigate structural mechanical effects on the coda signal.