MS7.9 - Dynamic Soil-Structure Interaction and Wave Propagation

Dynamic soil-structure interaction studies are widely used in geotechnical earthquake engineering to evaluate structural capacity and component’ safety requirements under seismic loading. Given the uncertainties and physical complexity of the seismic excitation and soil domain, as well as computational constrains, very often soil-structure interaction studies are bounded to vertically incident plane-waves and horizontal soil stratification. However, these simplifications may not be always adapted to all site conditions and seismic scenarios. In order to tackle this issue, the domain reduction method (DRM) proposed by Bielak et al. (2003) is considered in this work. It consists of a two-steps weak coupling approach where the complete 3D wave field obtained from an auxiliary domain is replaced by equivalent nodal forces to be exerted on the boundary surface of a reduced domain, providing a more realistic definition of seismic excitation. In this framework, the complete problem can be decoupled and solved in two separate models with adapted numerical approaches. Therefore, an auxiliary domain defined in a regional scale integrating the seismic source and wave propagation from the source to the site of interest is proposed. This first step can be solved in a spectral element framework, which is adapted for large-scale wave propagation studies. By obtaining the equivalent nodal forces on the boundary surface of a reduced finite element model, soil-structure interaction analysis studies can be conducted integrating all the relevant aspects from source and path characteristics of a given seismic scenario. Consequently, the “incompatibility” of the different scales of the problem (regional, local) can be efficiently solved, by maintaining a sufficiently sophisticated numerical model for the local scale, where the hypothesis of a nonlinear soil behavior can be examined. The validation of the coupling is discussed in Korres et al. (2021), while this work focuses on the performance of the second step of the coupling approach and the FEM resolution. The main objective is to present the feasibility of the FEM resolution with the DRM excitation in a high-performance computing (HPC) framework. A simple case study is chosen so as to demonstrate the gains in terms of computational performance. Comparisons are also made with traditional approaches utilizing direct parallel solvers.

Sedimentary materials enclosed in basins complexify the ground motion by energy trapping, resonance and surface wave generation at the basin edge. In general, effects of basin presence are estimated by aggravation or adjustment factors (AGF). The present study evaluates the influence of basin effects on the seismic damage of a nonlinear structure, representing a bridge column, with a complete analysis from the earthquake source to the structure. The numerical simulation of wave propagation is performed with a coupled 3D SEM-FEM approach using the Domain Reduction Method, including non-linearities in the structure with a multifiber beam. To evaluate different seismic scenarios affecting the structure, a parametric study is performed with simplified basin geometries and homogeneous material properties. Two types of seismic sources are simulated: plane waves with vertical incidence and deep and shallow double-couple point sources. Besides, the spatial variability inside the basin is analyzed by inserting the bridge column in different positions, altering the ground motion wavefield arriving at the base of the structure consequently. The structural damage is estimated by the plastic energy dissipation index. The results show that basins' effect on structural damage can be estimated in a simplified form using a combination of a structural behaviour predictor and AGFs computed from the free-field ground motions. However, these factors should be correctly predicted, including both basin and source variability.

The fast and practical computation of dynamic stiffness of foundations in layered media is a very important task to evaluate the dynamic response of critical structures. In particular, the design of wind turbines in offshore regions are very sensitive to the stiffness and damping characteristics of the foundation system. It is critical in fatigue life prediction, and ultimate limit states. In this communication, a fast assessment method is proposed via a meta-model based on Deep Learning techniques. The model is designed to detect enriched physical domain, based on Transfer Learning from low-resolution physical domain to most refined physical schemes (complex layered systems and poroelasticity). To generate the required database of dynamic stifnesses, a three-dimensional Green function for multilayered elastic and poroelastic half space in the frequency domain has been developed and adapted for source and receiver locations at the top free surface. The fundamental solution is built by potential displacements, angular Fourier transform, and radial Hankel transform. The Julia programming language has been explored to implement and optimice the computation of the Inverse Hankel Transform. A Boundary Element Method code has been developed with traction singular shape fuction for rectangular foundations. In order to consider the applicability of the Deep Learning metamodel to complex physical domains (many layers, poroelasticity), increasing the number of features, a first deep neural network has been built and trained, suitable for two-layers elastic half-space configurations. Network hyperparameters have been optimized based on error analysis. A Transfer Learing approach is explored to build deep neural networks for improved physics, including layers and poroelastic properties. Numerical tests confirm that the proposed methodology is suitable for fast computation of the dynamic stiffness of foundations, with low computing times compatible with industry requirements.