Research Projects

India has set 500 GW of renewable energy by 2030, including 37 GW of offshore wind energy by leveraging its extensive 7500 km coastline. Build plans for offshore wind projects off the Gujarat and Tamil Nadu coasts will be paramount for this purpose. Current geotechnical investigations in these seismically active regions are insufficient. In such a context, granular materials at the soil-structure interface experience severe deformation, particle fragmentation, localised strain accumulation during operational phase as well as installation stage of the foundations of these critical infrastructure units subjected to repeated wave loading. Crushable granular media, predominantly calcareous in nature occur along the western coast and southern part of the Indian subcontinent. Unlike siliceous and quartz sands, crushable sands are light and have inherent voids that facilitate easy particle breakage. Subsequently this introduces more finer particles that transform the soil response from a dilative to a relatively contractive one. Monitoring the evolution of grain size distribution is indispensable in strain localisation studies where shear strain accumulation might significantly influence the overall system response. Particles within the deformation bands tend to be finer, strongly oriented in a preferred direction while undergoing grain crushing under sustained shearing. In such settings, the wave forces, and seismic loads expose crushable sands to cyclic shearing. Cyclic Simple Shear (CSS) loading accurately mimics the dynamic conditions of the soil during such a wave action. It ensures precise shear strain control during testing while considering several repeated cycles of continuous shearing of relatively low frequencies (0.05 to 1 Hz). This range simulates the wave loading scenario encountered by the soil surrounding the offshore structures and helps in assessing the liquefaction resistance, shear strain accumulation, etc. The present study examines the material behaviour of calcareous sands under various CSS loading conditions with varying frequency, density of the material, shearing rate, stress states etc. Experimental insights are restricted to only bulk response from measurements at one point on the specimen boundary. However, with onset of localised zones of shear strain accumulation on sustained loading, specimen deformation does not remain uniform. Conventional Discrete Element Method (DEM) offers valuable grain-scale insights into the micromechanical behaviour of the granular ensemble. However, a complete discrete setting to facilitate this is computationally expensive. Further, they do not capture grain breakage accurately and is often idealised as spherical clumps that introduce a biased fracture path. Although, continuum frameworks XFEM etc. provide means to model fracture-related problems, they suffer from pathological mesh dependency issues. A hierarchical multiscale discrete-continuum model complimented by CSS experiments is proposed that links the micro scale particulate features to the macro scale material response. It enforces DEM to model for inter-particle interaction, while Peridynamics for intra-granular interaction once breakage initiates. Most DEM studies idealise particle geometries as regular shapes, spherical, ellipsoids or clumps at most. Circumventing this, we investigate realistic particle shape effects on multiscale behaviour using avatar paradigms derived from 3D X-ray CT images as well as 3D reconstructed particles from 2D images of soil grains using Fourier descriptors. Machine learning aided algorithms are also leveraged to reconstruct these 3D avatars from 2D images using controlled morphologies. The results obtained will enrich our understanding in improving the existing damage based constitutive relationships that are used in traditional FEM numerical packages for foundation designs and geotechnical engineering recommendations for resilient offshore energy units and associated installations.