Geothermal systems and geostructures, as sustainable energy sources, undergo daily and seasonal temperature fluctuations that significantly influence their mechanical response. Reliable prediction of thermally induced deformations therefore requires advanced thermo-mechanical constitutive models. Existing approaches often address constant elevated temperatures but fail to capture multiple thermal cycles or the coupled effect of mechanical cycling under heating. This study presents a hypoplastic thermo-mechanical model enhanced with the extended intergranular strain anisotropy concept to reproduce small-strain behaviour. Experimental evidence shows that normally consolidated fine-grained soils, when subjected to repeated thermal cycles, exhibit a transition to an overconsolidated state after the first heating–cooling cycle. To capture this, the model introduces a temperature-dependent preloading surface, enabling the evolution of the three-dimensional overconsolidation ratio under thermal cyclic loading. In addition, the original viscous strain-rate mechanism at ambient conditions is preserved, ensuring a consistent representation of rate effects under coupled thermal and mechanical actions. The proposed model is validated against diverse thermo-mechanical loading paths, including monotonic and cyclic scenarios, across different soil types. The results demonstrate its capability to capture key aspects of the complex response of fine-grained soils under combined thermal and mechanical loading, indicating its potential applicability to energy geotechnical problems.
M. Ashrafi, M. Tafili, T. Wichtmann· Geotechnique· 0 citations
Estimating long-term soil deformation is crucial for the safe design of offshore foundations, railway substructures, pavement layers, and other cyclically loaded geotechnical systems. Traditional constitutive formulations, such as the High-Cycle Accumulation (HCA) model, provide reliable simulations but rely on empirically calibrated parameters that require extensive experimental campaigns for their determination. To overcome this shortcoming, it is proposed to design a Geomechanics-based Artificial Neural Network (GANN) that bypasses the need for calibration parameters and instead uses common soil descriptors. Two steps are required for this purpose: (i) develop a foundational GANN capable of accurately predicting accumulated strain evolution for a given soil; and (ii) generalize this GANN so that it can operate for a wide variety of soils using simple input variables such as the ones related to grain size distribution and state boundary surface. Accordingly, this manuscript aims to address this first step by considering two different soils as examples, namely Karlsruhe fine sand and Australian superfine silica sand. For each soil, a synthetic database is generated from the HCA model to train a GANN architecture, enabling it to learn the fundamental relationships governing strain accumulation. The methodology integrates data-driven learning with a physics-informed loss function, ensuring that the predicted strain evolution remains consistent with soil mechanics principles. This GANN comprises multiple fully connected layers using several activation functions, with outputs structured to capture strain accumulation over a high number of cycles. Validation against synthetic and experimental data confirms excellent performance, supporting this first step toward a generalized GANN framework.
R. Polo-Mendoza, M. Tafili, Jose Duque et al.· E3S Web of Conferences· 0 citations