Structural Integrity Assessment of Army Ground Vehicle Structures for Predictive Maintenance

Project start date Aug. 15, 2024.

The Army requires a robust nondestructive testing (NDT) tool to inspect Ground Vehicles (GVs) upon their return to bases, ensuring structural integrity and combat readiness. This tool must effectively detect embedded damages, such as delamination in composite armor or fatigue cracks in critical components, which may compromise protective capability or mechanical functionality. Ultrasonic NDT offers promise for imaging damages in GV components and structures. Effective imaging could streamline failure analysis and decision-making, preventing mechanical failures, reducing risks to human life, and minimizing maintenance downtime and operational costs.

Current methods for ultrasonic NDT face challenges. Optimization-based approaches (Waisman et al. 2010; Jung et al. 2013; and Zhang et al. 2019) require significant processing times, limiting swift imaging. While artificial neural network (ANN)-based methods (Jiang et al. 2021 and Latête et al. 2021) offer faster processing, they rely on prior knowledge about the quantities and shapes of damages, restricting their applicability to damages of arbitrary geometries and quantities in 3D structures such as GVs.

Our project seeks to address these limitations and develop a novel ultrasonic NDT method capable of imaging cracks and voids of arbitrary shapes and quantities in GVs. It is crucial that the ideal ANN-powered ultrasonic NDT can image structural damages of any shape and quantity in a GV structure of various geometries and material properties, such as conventional or composite materials.

The project aims to develop a novel computational and experimental method for imaging damages in GV components using ultrasonic responses measured by a laser Doppler vibrometer (LDV). We will investigate an innovative grid-wise classification algorithm, which can facilitate accurate and rapid identification of structural damages hidden within GV structures, and validate the algorithm using LDV-experimental data from real sample GV structures (e.g., composite layered-panel structures, mock armors, cylindrical-shaped structures, and enclosures of critical components like the battery of an electrically-powered GV).

Prior publications:

• Kim et al. (2024) and Pranto et al. (2023) introduced AI-based algorithms for Ultrasonic NDT to image structural damages, such as delamination cracks, of arbitrary shapes and quantities.
• Lloyd and Jeong (2024) presented a wave-source characterization method utilizing elastodynamic response measurement.
• Guidio and Jeong (2021), Guidio et al. (2023), and Kim and Kang (2024) investigated wave-based material characterization algorithms.
• Lloyd et al. (2023) conducted experimental validation of the wave-source characterization algorithm using ultrasonic measurement provided from a lab experiment.

3.A118