The hypersonic entry of a body into the upper layers of a planetary atmosphere leads to the formation of a plasma resulting from the intense compression of the incident gas. This compression takes place within a shock layer in non-equilibrium, the knowledge of which is partial. This prevents a precise assessment of the convective, radiative and catalytic parts of the parietal heat flux required for the sizing of the thermal protection system of the entering body. The latter contributions strongly depend on the excited states population density, the distribution of which departs from Boltzmann equilibrium. In these circumstances, the purpose of this work was to study, for entry situations into Mars and Earth atmosphere, the excited states behaviour of complex mixtures (CO2 -N2 -Ar and N2 -O2 -Ar, respectively) based on the development of electro-vibrational specific collisional-radiative (CR) models. Two CR models have been developed : CoRaM-MARS for the Martian atmosphere (22 species, 10^6 elementary processes) and CoRaM-AIR for the Earth's atmosphere (13 species, 500 000 elementary processes). These models are implemented in a Lagrangian approach at constant pressure and temperature in representative thermodynamic conditions of entry situations (especially the FIRE II test case for Earth entries). We have shown that radiation plays a minor role in the kinetic mechanism of the studied mixtures, and that departures from the Boltzmann equilibrium are systematically observed. The very large number of states to be considered prevents a direct implementation of these CR models in aerodynamic codes. However, their reduction dedicated to nitrogen has been performed. The CR model thus elaborated (CoRaM-N2 , 5 species, 150 states, 40000 elementary processes) has been integrated in an 1D Euler code dealing with post- shock or divergent nozzle flows. The agreement with high Mach number experimental results obtained in shock tube is very satisfactory. For 2D or 3D applications, global rate coefficients have been determined theoretically for (1) the ionization / recombination by electron impact of nitrogen, oxygen, carbon and argon, (2) the dissociation / recombination of N2 , O2 under N, N2 , O, O2 impact and (3) the dissociation / recombination of CO2 under heavy impact. Comparisons with experimental results globally show a very satisfactory agreement. The forward and backward global rates have been calculated independently. We show that their ratio departs from the corresponding global equilibrium constant as the temperature increases.
