In a new paper published in the Journal of Composite Materials, SUSTAINair partner JKU has identified a new, efficient approach for strain-based damage detection in aerospace composite sandwich structures, by introducing a data-smart framework for physics-driven generation of strain data.
Recent developments in damage detection in aerospace
In recent years, Structural Health Monitoring (SHM) has been increasingly applied to composite sandwich structures, as typically used in aerospace applications. In addition, machine learning approaches are increasingly popular for damage detection, localization and size estimation, due to their advantages in pattern recognition and anomaly detection.
However, a major disadvantage of machine learning techniques is that these algorithms generally require large amounts of realistic data. In general, these data are expensive or even impossible to obtain within a feasible time.
Efficient framework for physics-driven strain data generation
In order to overcome this obstacle of machine learning approaches, the JKU researchers, including SUSTAINair scientists Christoph Kralovec and Martin Schagerl, introduced a computationally inexpensive framework for physics-driven feature generation of strain data for the training of ML-based SHM methods, by using sub-structuring and reanalysis.
First, they subdivided the global FE model into a monitored part, i.e., a smaller submodel, and a global model. Second, the JKU researchers extracted the stiffness matrix of the submodel from the finite element software. Then, they performed static condensation to further reduce the computational effort. Afterwards, selected eigenvectors were derived by the JKU team in terms of displacements of master nodes, and the corresponding strains were calculated. Finally, the Austrian researchers performed a statistically varied linear combination between the different characteristic eigenvector load cases, based on the superposition principle.
This procedure enabled the efficient generation of a large number of different physics-driven determined strain solutions for a subsequent training of a ML algorithms. The proposed framework was evaluated by means of a damage detection approach, based on an artificial neuronal network classifier algorithm. The applied approach utilised strain measurements from selected positions as physical quantity, and was demonstrated by using a composite sandwich structure imitating an aircraft spoiler.
Change in relationship between sensors indicates damage
The key principle of the JKU damage detection algorithm is based on the fact that a change in the relationship between sensors indicates the presence of damage. Additionally, to the numerical healthy strains resulting from the framework, synthetically generated damage data was used for training the neuronal network classifier. The synthetic data was obtained by statistical modifications of the healthy strains, to avoid time-consuming and expensive damage simulations.
The JKU researchers validated the feature generation framework and health monitoring approach by using experiments and numerical simulations of a glass fiber reinforced polymer sandwich structure with a hole considered as damage. Their presented numerical and the experimental results clearly show the high potential for the efficient approach for damage detection in a aerospace sandwich structure.
Find out more about JKU's research on damage detection
The full version of JKU’s paper “A framework for physics-driven generation of feature data for strain-based damage detection in aerospace sandwich structures” is available online here since 17 September 2022. Find out more about SUSTAINair partner JKU’s research on Structural Health Monitoring (SHM) and damage detection also in the Scientific Publications section on our website and in our Zenodo circular aviation community.