Experimental and computational analysis of condensation phenomena for the thermal-hydraulic analysis of LWRs containments

This PhD research aims at improving the understanding of condensation phenomena of interest for the safety analysis of nuclear reactor containments. Particular attention if focused on the effect related to the possible presence of hydrogen, which could be released in the containment atmosphere during a loss of coolant accident with loss of core cooling and fuel pin clad oxidation. The way steam condensation phenomena have been investigated in this work is multiple: theoretical, experimental and numerical analyses have been carried out. A theoretical analysis of steam condensation is firstly proposed to clarify fundamental issues related to the modelling of diffusion phenomena in multicomponent mixtures. Experimental data available by the CONAN and COPAIN separate effect test facilities were analyzed. Further experimental activities have been performed, aimed at collecting specific data concerning the effect induced by the presence of a noncondensable gas lighter than steam . Three different condensation models are proposed, basing on the two main strategies adopted for wall condensation modelling in CFD codes: a fine approach based on the detailed resolution of the concentration, temperature and velocity gradients near the condensing wall and a less expensive approach adopting coarser discretization in the proximity of the condensing wall, based on the heat and mass transfer analogy. The HMTDMMSD and the HMTDMEBD approaches are proposed based on the first strategy, but different diffusion models. The HMTAM model is a fast running model based on the heat and mass transfer analogy. The capabilities of several turbulence models in reproducing turbulent transpiration phenomena have been formerly evaluated. A first stage consisted in analyzing the capabilities of turbulence models to reproduce nondimensional velocity profiles in the presence of pure turbulent mass and momentum transfer. A second stage consisted in a numerical analysis of suction effects induced by condensation. The condensation models have been finally applied to the modelling of the CONAN and COPAIN steam-air and steam-air-helium tests. An extensive comparison with local and integral experimental data is proposed. A theoretical and numerical analysis is finally proposed, aimed at assessing to what extent helium is an appropriate substitute for hydrogen. Concluding remarks are drawn and future activities are suggested, either under the experimental than the numerical point of view.