سيد ميثم علوي

The progressive use of geosynthetic reinforced soil (GRS) walls, shortcomings of the accepted methods of designing these walls in the guidelines, and limited numbers of wall model configurations and seismic loading characteristics in previous studies indicate the necessity for more detailed studies in this field. As a result, in this study, seismic performance of short and tall GRS walls has been eva‎luated using 2D and 3D finite element analyses and fragility curve approach. Furthermore, the distribution of maximum reinforcement loads in the wall height has been eva‎luated based on numerous machine learning simulations. Next, the eva‎luation of seismic deformations of single- and multi-tiered GRS walls has been conducted using the fragility curve method, founded on the results of finite element analyses and machine learning algorithms. In this regard, a new seismic intensity measure (dimensionless parameter) has been developed to be used in generating such fragility curves. The results of this study reveal a significant difference in the performance of short and tall GRS walls, considering he deformations and reinforcement loads (by comparing normalized indices). Considering the perpendicular to the face deformations, the results of modeling the walls in 2D and 3D models at the end of construction or under seismic records are nearly the same (the maximum difference in displacements was 0.4% and 0.07% of wall height in the case of dynamic and static loading, respectively). Besides, considering two horizontal directions of seismic loads in the 3D assessments of the walls resulted in the observation of bulging deformation mode; although, seismic loadings in only one direction (perpendicular to the facing) is more conservative. Increasing the reinforcement stiffness resulted in better results in controlling the wall deformations (and decreasing the damage possibility), considering increasing the reinforcement length. In this study, a new dimentionless intesntity measure has been defined based on the PGA, PGV, PGD, SED, and Ia of seismic records to use in the fragility curves of the GRS walls (with maximum facing deformations as the damage index). The accuracy of thi parameter has been eva‎luated using the previous damages of real GRS walls due to various earthquakes. The proposed multiple analysis method in this study for investigating the seismic performance of reinforced soil walls based on seismic finite element analyses, machine learning algorithms, and performance-based fragility curves, can serve as a template analysis method for assessing other geotechnical structures.

ديوارهاي خاك مسلح كوتاه و بلند , تحليل المان محدود لرزه‌اي , يادگيري ماشين , ديوارهاي پلكاني , مدل‌سازي سه بعدي , منحني‌هاي شكنندگي , مطالعات پارامتريك , شدت زلزله

Short and tall reinforced soil walls , Seismic finite element analyses , Machine learning , Tiered walls , 3D modeling , Fragility curves , Probabilistic parametric studies , Seismic intensity measure

Seyed Meisam Alavi

Dr. Hamid Reza Razeghi - Dr. Ali Akbar Heshmati Rafsanjani