Font Size: 

Structural health monitoring (SHM) is a growing field of tracking and evaluating the health condition of structures. 3D frame structures are one of the most common types of structures these days, and over their life spans, they could experience damage such as corrosion and cracks due to material degradation, or overloading. This study aims to develop a reliable and real-time damage identification method for 3D frame structures. The aim of this study can be achieved by using deep learning techniques and vibration parameters. A new damage identification method is proposed by utilizing convolutional neural networks (CNNs) to detect damage from modal strain energy. Modal strain energy is well-known for its high sensitivity to damage, and therefore, it can be utilized as a damage index in SHM. CNNs have the ability to self-learn and make predictions in an instant manner, making them suitable for real-time monitoring of the structural condition. To validate the effectiveness of the proposed method, a space frame is used as a case study. The results demonstrate that the proposed method accurately detects, localizes, and quantifies damage in the frame. This study also paves the way for further research on the application of deep learning techniques for SHM and other areas of structural engineering.

Convolutional neural network, Damage identification, Modal strain energy, Space frame, Vibration