Loading...

Damage Detection and Structural Response Estimation of Reinforced Concrete Columns Using Data-Driven Methods

Sheikhi Yousefabad, Majid | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57215 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Mohtasham Dolatshahi, Kiarash
  7. Abstract:
  8. The main objective of this study is the damage detection and structural response estimation of reinforced concrete columns using data-driven methods. After a severe earthquake, the first signs of structural damage appear as surface cracks on the structural elements. The visual features of the damage encompass a wealth of information, which can be utilized for structural damage assessment. Recently, the use of computer vision tools for automated assessment of structural damage based on visual damage features has been increasing. This method offers higher speed and accuracy compared to traditional methods such as visual inspection while being immune to human errors and judgments. In this research, computer vision algorithms are employed to extract visual damage features of reinforced concrete columns, and genetic algorithms are used to find predictive equations for predicting the strength degradation and stiffness deterioration of reinforced concrete columns. Additionally, finite element modeling is another method employed to predict the response of structures. However, in practical applications, significant discrepancies between finite element predictions and experimental results are often observed, especially in the range of nonlinear behavior. This study proposes a hybrid finite element-machine learning approach using the Extended Kalman Filter to update finite element model response. Ultimately, the parameters obtained from this study can be used as engineering demand parameters for assessing the condition and losses of concrete structures. Furthermore, the proposed hybrid framework can serve as a reliable method for accurately estimating structural response with lower computational costs compared to conventional methods
  9. Keywords:
  10. Machine Vision ; Reinforced Concrete Columns ; Machine Learning ; Kalman Filters ; Finite Element Model Updating ; Visual Damage Features

 Digital Object List

 Bookmark

No TOC