Ces travaux ont pour but d’étudier le comportement au mouillage spontané (statique et dynamique) ainsi que le mouillage forcé, sous l’effet de la compression, de différentes structures textiles hydrophobes. Le mouillage forcé a permis d’évaluer la robustesse de l’hydrophobicité des structures textiles. En parallèle, un nouveau dispositif automatisé et plus approprié à l’étude du mouillage forcé a été mis au point. Les principaux résultats ont montré qu’une bonne hydrophobicité ne conduit pas forcément à une bonne robustesse : spontanément, la structure floquée est la seule à favoriser une configuration de Cassie-Baxter, cependant, sa robustesse est plus faible que celle des tissus. Les différents essais ont mis en évidence l’influence importante et majeure de certains paramètres, appropriés à chaque structure textile, sur son hydrophobicité et sa robustesse comme la densité et la finesse des poils pour les tissus floqués. Ils ont également montré que certains facteurs pouvaient améliorer l’hydrophobicité mais pas sa robustesse ou inversement. Ainsi, la robustesse de l’hydrophobicité n’est pas prévisible à partir des mesures du mouillage spontané. / This work aims to study the spontaneous (static and dynamic) and the forced (under the effect of compression) wetting behaviour of different water-repellent textile structures. Forced wetting allowed to evaluate the robustness of the hydrophobicity of textile structures. In parallel, a new automated and more suitable device was developed for the study of forced wetting. The main results showed that a good hydrophobicity does not necessarily lead to a good robustness: spontaneously, the flocked structure is the only one to foster the Cassie-Baxter state, however, its hydrophobicity’s robustness is lower than that of the tissue. The different tests have highlighted the important and major influence of some parameters, adapted to each textile structure, on its hydrophobicity and its robustness such as the density and fineness of bristles for flocked fabrics. They also showed that some factors can improve the hydrophobicity but not its robustness or vice versa. Thus, the robustness of the hydrophobicity is not predictable from the measures of spontaneous wetting.

The dynamics wetting of fiber is of crucial importance in the fields, such as composites, optical fiber or textile industries. It is therefore valuable to acquire a clear understanding of the fundamental physical mechanisms which govern this phenomenon. In the case of partial wetting, it is assumed that the loss of energy due to the change in shape of the liquid-fluid interface (surface tension) is balanced by two channels of dissipation. One is associated with the viscosity of the liquid (hydrodynamics), whereas the other is due to the friction between the liquid and the solid (molecular-kinetic theory). Translated into equations, this original approach leads to the conclusion that the friction regime should precede the hydrodynamic one for a low viscosity liquid. The crossover time between the two regimes is calculated and shown to be material dependent. To validate these theoretical predictions, both experiments and large scale molecular dynamics simulations of the spontaneous spreading of a liquid along a fiber are run. The experiments consist in capturing images of meniscus formation around the fiber via a high speed camera. For each image, the liquid-air profile is extracted and fitted to a solution of the Laplace equation yielding the contact angle and the height of the meniscus as a function of the time. For low viscosity liquid it is found that the measured dynamic contact angle follows the friction regime, whereas for higher viscosity liquid the viscous regime is recovered as presented theoretically. The same kind of procedure is followed to study the wetting of a nanofiber by molecular dynamics. The properties of the liquid are first assessed (viscosity, shape of the molecule, molecular volume). Because of the very low viscosity of the model liquid, it is expected that the friction between the liquid and the solid is the dominant channel. Indeed, the data from the simulation validates this assumption. Moreover, according to the results of the simulation, it is also confirmed that for a given equilibrium contact angle, a maximum of speed wetting occurs. Actually, a low (or high) equilibrium contact angle involves both a strong (or weak) driving force and adhesion of the liquid molecules to the solid atoms. These opposite effects do not simply cancel out and therefore lead to the existence of a maximum rate at which a liquid can wet a solid. To examine in detail this last statement, the forced wetting of fiber is studied by molecular dynamics. The fiber, at a constant velocity, goes through the meniscus of a liquid which is consequently elongated. Stationary receding and advancing contact angles are then measured as a function of the fiber velocity. It is found that the contact angle dependence on the fiber velocity follows the molecular-kinetic theory, thereby confirming the existence of a maximum. Moreover, a comparison between the values of the microscopic parameters obtained via the adjustement of the theory and a direct measurement of these parameters permits us to check the validity of the chosen theory as well as the reliability of the simulation tool. Finally in order to study the wetting of fibrous materials like fabrics, an effective system is studied via molecular dynamics. It is shown that the measurements of capillary imbibition and droplet spreading are well modelled by a set of equations taking into account the conservation of the volume of the liquid, the influence of a dynamic contact angle inside the pore and the spreading on top of the surface. This single pore geometry is extended theoretically to the case of multiple non-interconnected pores. The time required to absorb the droplet completely is then calculated.