Experimental-numerical approach to detecting manufacturing- and load-induced damage in fiber-metal laminates based on changes in vibration and damping behavior

Fiber-metal laminates (FML) combine fiber-reinforced plastics (FRP) with metals to specifically improve properties such as specific strength and fatigue strength. In this project, glass fiber-reinforced plastic (FRP) is processed with aluminum to form a so-called GLARE composite. Due to its properties, this hybrid material is particularly popular in the aviation industry.

Reliable production of complex geometries is necessary to enable its use in other fields of application such as automotive engineering. As this increases the risk of defects, a methodology for defect detection is being developed and investigated in this project.

The non-destructive detection of defects is necessary to ensure the reliability of components in further operation after production or after impact events. Certain defects, such as so-called "kissing bonds" (unwanted adhesion-free interfaces), pose a particular challenge for the current state of the art. These are particularly difficult or even impossible to detect using ultrasound and computer tomography. Other defects such as pores and intralaminar damage are also part of this work.

The defect-related mechanical property changes are to be detected, localized and characterized by deviations in the vibration and damping behaviour of the component. Experimental investigations of the dynamic and modal behavior of the FML are carried out as part of the project; this part of the project takes place at the IAM-WK of the KIT. In parallel, a computer-aided simulation of the system is being developed here at FAST. Numerical simulations are used to develop methods for mapping the defects, which are then used to generate a training data set.

Finally, a model is created that is trained on the basis of the simulation data. This model is used to evaluate the vibration behavior of real components. The prediction variables should initially include the type of defect and then its location and size. The method is then validated and verified on real components in tests to assess its accuracy and sensitivity.