Loading...

Experimental and Numerical Investigations on Damage Location Detection of Multi-Damaged Beam with Different Intensities using Vibration Responses

Khodabakhshi, Nahid | 2025

131 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 58411 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Khaloo, Alireza; Khajehdezfuly, Amin
  7. Abstract:
  8. In this study, a structural health monitoring method called the Sensor Network Frequency-Based (SNFB) method is presented, which uses the structure’s frequency response function to identify multiple damage locations. This method is not only accurate but also simple and cost-effective, making it suitable for preliminary structural health monitoring. Due to limitations in the existing technical literature, comprehensive experimental and numerical studies were conducted on steel beams with various damage scenarios. In the experimental study, beam specimens were excited using a modal hammer, and the structural response was recorded by a sensor network. To determine the damage locations, the frequency response function of the undamaged structure was calculated by solving a numerical model of the undamaged beam in MATLAB software. Then, by comparing the frequency response functions of the undamaged and damaged structures, the SNFB method identified the damage locations. Data fusion was used to combine the sensor network data through hypothesis testing theory, and a threshold was introduced to allow more accurate interpretation of the results and precise damage localization. The results showed that, compared to common methods such as the damage location vector and transmissibility index, the SNFB method can accurately identify multiple damage locations with an intensity above 5% without loss of accuracy due to damage severity, quantity, or proximity. To evaluate the method’s performance on more complex structures, SNFB was also tested on a numerical model of a two-span box-shaped bridge. The results demonstrated that, unlike the transmissibility method—which loses accuracy despite a higher number of sensors and is affected by damage severity and location—the SNFB method maintains satisfactory accuracy. Only in the presence of noise and damage near supports its accuracy slightly decrease. Overall, the SNFB method can accurately identify multiple damage locations with intensities above 5%
  9. Keywords:
  10. Structural Health Monitoring ; Frequency Response ; Data Fusion ; Hypothesis Test ; Multi-Damage Structures ; Probabilistic Damage Localization

 Digital Object List

 Bookmark

No TOC