Borehole Heat Exchanger (BHE) can be installed in different infrastructures to store or extract energy, such as pavement cooling in summer and de-icing in winter. Although there are approaches to estimating the performance of a BHE based on analytical, semi-analytical or numerical simulation methods, predicting the performance of a BHE in the perspective of Machine Learning (ML) is still necessary. In this investigation, four ML algorithms including Linear Regression (LR), Polynomial Regression (PR), Artificial Neural Network (ANN) and Random Forest (RF) were used to explore the underlying relationship between the Heat Extraction Rate (HER) and the influencing factors. In total, the annual HERs of 400 Thermal Performance Tests (TPTs) covering 12 major factors were computed in a validated numerical simulation framework. After providing necessary database, the 4 ML approaches were trained and evaluated. The results showed that the trained PR approach had the best performance. Specifically, the trained PR approach had a Root-Mean-Square Error (RMSE) lower than 1.74 W · m⁻¹, with the Coefficient of Determination (R²) higher than 0.99. Additionally, the trained approach was compared to an in-situ TPT in Shijiazhuang (China) and two numerical simulated scenarios in Alsace (France). The results showed that the trained PR approach predicted well the in-situ TPT by having 3 W · m⁻¹ of RMSE, and it predicted correctly the numerical simulated HER of a BHE with 0.09 and 0.72 W · m⁻¹ of RMSEs for two scenarios with 5 and 1 °C of inlet fluid temperatures. The proposed prediction model can be applied to estimate the performance of a BHE, providing another option to fast evaluate the BHE performance.

Granular materials are often used in pavement structures. The influence of fine content and water content on the mechanical behaviour is very important. In this work, first, we present the resilient behaviour of compacted clayey natural sands with two different fine contents using the repeated load triaxial tests. The samples were compacted at different water contents with the same dry density. It can be stated that an increase in the fine content (from 4% to 7.5%) causes an increase in the volumetric strains and shear strains for all the water contents. The influence of water content can be considered negligible comparing to the influence of fine content. Then, the experimental results were simulated using the nonlinear elastic model (modified Boyce model) generally used for the resilient behaviour of the unsaturated granular soils. The results show that the model can correctly predict the variation of the volumetric strains and the shear strains with the mean stress for all the stress paths. Finally, finite element calculations carried out with CAST3M were performed to compare the deflection of a low traffic pavement with sub-base layers containing two different fine contents. The surface asphalt concrete layer and the base layer were taken the same for both structures. The modelling results show that the sub-base soil with higher fine content produces 37% higher deflection in the whole structure and that the effect of the water content is lower than that of the fine content.

This paper presents the influence of the unsaturated state and especially the different hydraulic behaviours during the wetting and drying cycles (hysteresis phenomenon) on the resilient behaviour of unbound granular materials for roads. In this context, the soil water retention curve (SWRC) of a compacted clayey sand was initially obtained for both wetting and drying paths. These experimental results were then complemented by a series of direct shear tests. Finally, repeated load triaxial tests (RLTTs) were also performed at the same water contents on both wetting and drying paths. It can be stated that the samples prepared on the wetting path showed higher resilient volumetric strains compared to those prepared on the drying path, especially at low water content. However, the influence of the hysteresis phenomenon is less evident for the resilient shear strains. Different effective stress concepts based on the SWRC and shear test results were also added to the classical Boyce model generally used to estimate the resilient behaviour based on the total stress concept. The estimated effective stresses of the RLTT give the best correlation coefficient for all water contents of the wetting and drying paths. Finally, finite element calculations, carried out with CAST3M, were performed to compare the deflection of a low traffic pavement at saturated/dry conditions for the subgrade soil. The asphalt concrete layer and the base layer were taken the same for both conditions. The influence of the hysteresis phenomenon on the deflections is more pronounced at the dry condition.

Granular materials are often used in pavement structures. The influence of anisotropy on the mechanical behaviour of granular materials is very important. The coupled effects of water content and fine content usually lead to more complex anisotropic behaviour. With a repeated load triaxial test (RLTT), it is possible to measure the anisotropic deformation behaviour of granular materials. This article initially presents an experimental study of the resilient repeated load response of a compacted clayey natural sand with three fine contents and different water contents. Based on anisotropic behaviour, the non-linear resilient model (Boyce model) is improved by the radial anisotropy coefficient γ₃ instead of the axial anisotropy coefficient γ₁. The results from both approaches (γ₁ and γ₃) are compared with the measured volumetric and deviatoric responses. These results confirm the capacity of the improved model to capture the general trend of the experiments. Finally, finite element calculations are performed with CAST3M in order to validate the improvement of the modified Boyce model (from γ₁ to γ₃). The modelling results indicate that the modified Boyce model with γ₃ is more widely available in different water contents and different fine contents for this granular material. Besides, based on the results, the coupled effects of water content and fine content on the deflection of the structures can also be observed.