The Reliability Analysis of 3D Steel Structures by Constructing Artificial Records Compatible with the Determined Intensity

نوع: Type: thesis

مقطع: Segment: PHD

عنوان: Title: The Reliability Analysis of 3D Steel Structures by Constructing Artificial Records Compatible with the Determined Intensity

ارائه دهنده: Provider: Masoud Ghamari

اساتید راهنما: Supervisors: Dr. Mohammad Shooshtari

اساتید مشاور: Advisory Professors:

اساتید ممتحن یا داور: Examining professors or referees: Dr. Azad Yazdani, Dr. Amin Mohebkhah, Dr. Mostafa Moghdasi

زمان و تاریخ ارائه: Time and date of presentation: 2024

مکان ارائه: Place of presentation: آمفی‌تئاتر دانشکده مهندسی

چکیده: Abstract: Structural reliability assessment is a tool for assessing the effect of uncertainties on the structure. Seismic excitation is one of the uncertainties that influence structural reliability. Many random intensities are generated, and earthquake records are scaled to those intensities. Time history analyses are performed using the scaled records, and the structural failure probability is calculated. An intensity measure is needed to perform this process. The available intensity measures have been developed for 2D frames. In addition, scaling causes response bias and weakens the record-to-record variability. So, the process of structural reliability assessment under the seismic excitation has double uncertainties. The main goal of this research is to remove those double uncertainties. To this aim, artificial records whose structures (nature) are compatible with the generated random intensities are constructed. The abstract term "earthquake nature" is made tangible using the geometrical and temporal (topology) structures. For research, four 3D steel structures were designed with two structural systems (two 8-story and two 12-story structures). In the first stage, the suitability of 22 intensity measures was investigated for 3D steel structures. After selecting the suitable intensity measure for 3D steel structures, the structures of the seismic records were studied. Structural similarities between the two horizontal components of an earthquake, the acceleration response spectra of the two horizontal components of an earthquake, two two-component earthquakes, and the acceleration response spectra of two two-component earthquakes were computed. The effects of intrinsic earthquake characteristics and intensity on the record structure and the acceleration response spectra were studied using the resulting similarities. The temporal structure of earthquakes was extracted recurrently, considering an earthquake as a chaotic phenomenon and a memory signal. The selected random intensities were induced in each record by setting the two initial values to allow the intensity-compatible artificial records to be built. In the last stage, the structural reliability of an 8-story structure was assessed using the Conventional (CSRA) and intensity measure-compatible artificial records (ICSRA) (the proposed method) methods. The results of the first stage of the study showed that there is no intensity measure which can satisfy all criteria for all engineering demand parameters and structures. The suitable intensity measure should be selected based on the circumstances and priorities of each project. The combination method of the intensity measures of the two horizontal components does not influence the suitability of an intensity measure. Soil-structure interaction is crucial in selecting the suitable intensity measure; ignoring it falsifies results. I_(∆-PGV), as a function of the strong shaking duration and the peak ground velocity, is the only IM that is not among the last three IMs in any criterion and structure. Also, I_(∆-PGV) is among the best three IMs in some cases. So, I_(∆-PGV) is the most suitable IM for 3D steel structures. Morphological studies showed significant structural similarities between the two horizontal components of an earthquake, between the acceleration response spectra of the two horizontal components of an earthquake, between two two-component earthquakes, and between the acceleration response spectra of two two-component earthquakes. Also, earthquake intrinsic characteristics and intensity influence the earthquake's geometrical and temporal structure by more than 38 percent. According to the results, intensity equality between two earthquakes is not a sufficient condition for the same structural responses. Both the same temporal structure and the same intensity should be fulfilled to have two earthquakes cause the same structural responses. As a sample, the observed difference in the stored plastic energy in the buildings between earthquakes with the same intensity and with different structures is up to 7 times. In addition to this, scaling the earthquakes changed the temporal structures up to 17 percent. Change in the temporal structure means creating uncertainty in the record-to-record variability and imposing bias on the failure probability of the building. So, the earthquake's destructive power depends on both the temporal structure and the intensity. Reliability assessment of the 8-story structure showed that the ICSRA method converges sooner than the CSRA method. The faster convergence is due to the removal of double uncertainties. The failure probability and its standard deviation of the 8-story structure using the ICSRA method are 0.0243 and 6.07×〖10〗^(-5), respectively. Therefore, removing the double uncertainties in record-to-record variability and scaling the earthquakes causes a faster convergence in the reliability assessment. By referencing the results obtained in the morphological studies, the accuracy and validity of the ICSRA method are far higher than those of the CSRA.