Entwicklung von isogeometrischen und Deep Learning-Verfahren für das Design akustischer Metamaterialien für praktische Anwendungen

• Antragsteller: Dr.-Ing. Ahmed Mostafa Shaaban Mohamed
• Fachliche Zuordnung: Akustik ; Angewandte Mechanik, Statik und Dynamik ; Mathematik
• Förderung: Förderung seit 2023
• Projektkennung: Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 524106968

In most of the previous research works, acoustic metamaterials were treated as infinite sound-hard structures ignoring their complex 3D shapes and their structural properties as well, where the results cannot enrich the real applications. From that point, this research project is aimed to overcome these deficiencies by investigating acoustic metamaterials as a coupled acoustic-structural system with fully 3D geometry considering the acoustic-fluid interaction with total losses using two approaches separately. The first analysis will be performed with DNNs to solve PDE strong form of the coupled system. Then, the same coupled system will be solved with the discussed IGA approaches. After that, a comparison will be implemented between the efficiency of the two suggested approaches. It is also planned to implement a dynamic analysis on the metamaterials for high frequency problems. As an additional task, DNNs will be utilized as a surrogate model to identify multivariate uncertainties in acoustic-vibration interaction of the metamaterials system where DNNs can improve the sampling efficiency of uncertainty methods.

• DFG-Verfahren: WBP Stelle