Loading...

Structural health monitoring of nonlinear structures using real-time hysteresis loop analysis

Hosseini Tehrani, Amin | 2021

395 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 53954 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Rahimzadeh Rofooei, Fayaz; Mahdavi, Hossein
  7. Abstract:
  8. Under conventional seismic design strategies, civil engineering structures are designed to undergo inelastic deformation to dissipate earthquake energy, leading to residual displacements and nonlinear deformations in the structure. As a result, there is an essential need for structural health monitoring methods to detect and process these nonlinear behaviors. In this dissertation, a model-free Hysteresis Loop Analysis (HLA) method is developed to identify damage and monitor the structure's performance online, under the earthquake excitations in both severe linear and nonlinear scenarios. The structural models used to validate this method includes three types of one-story, five-story, and ten-story moment resisting frames in prepared in OpenSees software. Different noise levels of 5%, 10%, and 15% were added to the earthquake acceleration amplitudes to evaluate the robustness of the implemented method against the intensity of the ground motion. The purpose of this study is to design and implement an efficient real-time structural health monitoring system using a new method of hysteresis loop analysis for structures with severe nonlinear behavior. In most cases, the structural failure directly affects the amount of structural stiffness, so one of the appropriate methods to explore the structure's extent of nonlinear behavior is to use the structural hysteresis curve which requires predicting the behavior and structural properties using existing hysteresis models. In the method developed in this research, in order to analyze the hysteresis behaviour of the structure, the reconstructed hysteresis cycles are obtained using the acceleration and drift record of each floor, then the hysteresis loops are divided into smaller half cycles and each of the half cycles are divided into one-part, two-part, three-part, and four-part categories using a method based on optimization and probability testing. Finally, the failure indices (which in this study are changes in elastic stiffness) are identified for each of these half-cycles and its changes are displayed over time. The results obtained in this study indicate the ability of the proposed method to detect the structural failure online in structures with nonlinear behavior with noise. Also, this method can be considered for pinching behavior in the structural models and has a low computational cost
  9. Keywords:
  10. Structural Health Monitoring ; Hysteresis Ring ; Damage Identification ; Real-Time Damage Detection ; Real-Time System Identification ; Inelastic Deformation ; Elastic Nonlinear Behavior

 Digital Object List

 Bookmark