This thesis presents a numerical formulation grounded on continuum physics to study different coupled phenomena, such as thermomechanics and high-frequency thermoelectricity, in nanoparticles (NPs) for their applications to nanofluids (NFs) for both thermal energy storage and direct solar absorption applications.This work falls within the current scope of energy transition from conventional fossil fuel sources towards alternative energy ones to fight against global warming. The use of renewable energies appears to be the solution for reducing polluting emissions and thus, constitutes a research field which concentrates great efforts these days. Among all the existing renewable energies, the focus is put on solar thermal energy, in which solar radiation is harvested by Concentrated Solar Power (CSP) plants to convert it into electricity. Since a major drawback of solar energy is its intermittence due to weather conditions, Thermal Energy Storage (TES) systems are commonly incorporated to mitigate this inconvenience.One of the technologies under research for TES systems is the use of nanofluids. Commonly, base fluids with poor thermal properties are combined with metallic oxide NPs in order to enhance the thermal properties of the nanofluid. Recently, metallic nanoencapsulated Phase Change Materials (nePCMs), exhibiting a core@shell structure, were proposed as the solid medium to take advantage of the latent heat storage in addition to the sensible heat storage. It was observed that one of the issues arising in nePCMs was the eventual failure of the shell when subjected to thermal processes and then, the leakage of the molten core, which is not confined anymore.Therefore, the need for a rigorous analysis of the failure of the shell of nePCMs has motivated the formulation of a thermomechanical model with phase change in this thesis. This thermodynamically consistent model, discretised within the Finite Element (FE) method and implemented in a research code, allows to predict the thermal stresses arising during thermal processes. Then, the numerical tool is used to predict the failure of spherical and cylindrical nePCMs for different pairs of core@shell materials (Sn@SnO 2 and Al@Al 2 O 3 ) and to assess the influence of the shell thickness on the energy density capability and mechanical strength of nePCMs.As experiments and real applications are not exempt from uncertainties, the formulation of a probabilistic numerical tool appears to be of capital importance to incorporate measurement dispersion into the numerical analysis. For this purpose, a tool combining the FE thermomechanical model with Monte Carlo techniques is developed: i) to identify the physical parameters exerting a major influence on the mechanical failure of the shell and on the energy density capability of the nePCM, ii) to predict its probability of failure and iii) to select the optimal materials for industrial applications.The developed numerical probabilistic tool is used to analyse the probabil-XIII