Abradable materials are used to improve the efficiency of turbomachines by reducing the gap between the tip of the blade and the casing. The abradability property is the ability of a material to be worn out easily, in a controlled way, during a contact with the blade. In addition, abradable materials must resist erosion caused by particles swallowed in the engine, including dust, sand, water droplets, ice and volcanic fly ashes. Thermally sprayed materials made of aluminum and polyester seem to comply with these apparently conflicting properties of combining good abradability and erosion resistance. Yet, the origin of these good performances is not well understood. The purpose of the thesis is to identify the physical parameters that influence the abradability and the erosion resistance in order to, either guide the design of more reliable systems based on these materials and to offer more innovative and more sustainable solutions. To achieve this goal, hybrid model abradable materials are manufactured, studied and characterized using microstructure analysis, mechanical and fracture testing as well as tribology measurements. Four families of model architectured materials were processed and tested, including (i) metallic foams infiltrated with polyester, (ii) powder aggregates, (ii) polyester matrices with aluminum reinforcements and (iv) the classical thermally sprayed material. Experimental results on these materials confirm that, in terms of combining erosion and abradability, some model materials exhibit much better performances than others. In parallel with this experimental approach, physics-based models of abradability and erosion are proposed, including elasticity, plasticity and fracture mechanics and adapted to three different wear mechanisms. These models are then refined and validated on the basis of experimental results and used to guide the selection and optimization of abradable materials. Modelling results show that the key parameters are the quality of the interfaces between the aluminum and the polyester as well as the maximum elastic strain attained before entering plasticity. Our results also show that the dominant wear mechanism is affected by the penetration depth and by the geometry of the blade. Finally, the calculation of the optimal set of properties required to improve both abradability and erosion resistance points towards materials with low Young’s modulus and large ability to deform elastically before entering plasticity, including high temperature elastomers, polymers and porous materials.