[en] During the course of a hypothetical severe accident in a light water reactor, molten liquid may be introduced into a volatile coolant, which, under certain conditions, results in explosive interactions. Such fuel-coolant interactions (FCI) are characterised by an initial pre-mixing phase during which the molten liquid, metallic or oxidic in nature, undergoes a breakup (fragmentation) process which significantly increase the area available for melt-coolant contact, and thus energy transfer. Although substantial progress in the understanding of phenomenology of the FCI events has been achieved in recent years, there remain uncertainties in describing the primary and secondary breakup processes. The focus of this work is on the melt jet and drop breakup during the premixing phase of FCI. The objectives are to gain insight into the premixing phase of the FCI phenomena, to determine what fraction of the melt fragments and determine the size distribution. The approach is to perform experiments with various simulant materials, at different scales, different conditions and with variation of controlling parameters affecting jet and drop breakup processes. The analysis approach is to investigate processes at different level of detail and complexity to understand the physics, to rationalise experimental results and to develop and validate models. In the first chapter a brief introduction and review of the status of the FCI phenomena is performed. A review of previous and current experimental projects is performed. The status of the experimental projects and major findings are outlined. The first part of the second chapter deals with experimental investigation of jet breakup. Two series of experiments were performed with low and high temperature jets. The low temperature experiments employed cerrobend-70 as jet liquid. A systematic investigation of thermal hydraulic conditions and melt physical properties on the jet fragmentation and particle debris characteristics was performed. The coolant temperature was found to significantly affect the shape and size of the debris. The maximum fragment size was found to increase with reduction in coolant temperature. No effect of coolant voiding on the fragment size distribution was observed. A series of high temperature melt jet experiments were performed, in the MIRA-20L experimental facility. Three types of oxidic melts, namely; CaO-B₂O₃, MnO-TiO₂ and WO₃-CaO were employed as melt jet liquid. The melt jet fragmentation was classified into two regimes, the hydrodynamic-controlled regime and the solidification-controlled regime. The delineation between those regimes was observed from the size characteristic and morphology of the solidified debris which was formed. The temperature of the coolant was the primary parameter in determining which regime the jet fragmentation would fall into. It was found, at low subcooling, the fragments are relatively large and irregular compared to smaller particles produced at higher subcooling. The melt density was found to have significant effect on the particle size. The mass mean size of the debris changes proportional to the square root of the coolant to melt density ratio. A systematic investigation of the performance of statistical distributions which may be used to describe the size distributions of fragments obtained from molten fuel coolant interaction (MFCI) experiments was performed. The statistical analysis of the debris produced in both experiments showed that the sequential fragmentation theory fits best the particle distribution produced during the jet fragmentation process. In the second part of the second chapter, analysis of the jet breakup experiments are performed. The low temperature jet fragmentation experiments are simulated with a recently developed Multiphase Eulerian Lagrangian Method. The effect of particle diameter and particle drag on the jet dynamics and penetration behavior is investigated. The third part of the second chapter deals with simulation of Kelvin-Helmholtz instabilities. A high order Navier-Stokes solver is employed along with the front tracking Level-Set algorithm, to eliminate numerical diffusion. The effect of surface tension and viscosity on the development of instabilities is investigated. Three regimes are identified, and delineated, based on Weber and Ohnesorge numbers. The third chapter is devoted to breakup of liquid drops in water. The emphasis is directed towards delineating the roles which melt to coolant heat transfer, melt solidification, melt fusion heat and melt mushy zone play in the fragmentation process. Coolant temperature is found to have a significant impact on the droplet fragmentation behaviour for subcooled conditions. The melt superheat greatly affects the characteristic time for solidification, and thus strongly affects the deepness of the fragmentation process. The fusion heat of the eutectic melt contributes significantly to the solidification time scale, and thereby enables the eutectic melt drop to feature deeper fragmentation. The presence of the mushy zone during the phase change of the non-eutectic melts significantly prevents these melt drops from completing the deformation and fragmentation process, especially when the melt superheat is small. An instability analysis on the crust breakup was performed. A modified dimensionless Aeroelastic number Ae_* is obtained as a criteria for breakup of the plain crust. It is found that the modified Aeroelastic number can be employed to evaluate the breakup behaviour of a droplet with a thin solidified layer on its surface

[en] The objective of this work is to address the modeling of the thermal hydrodynamic phenomena and interactions occurring during the progression of reactor severe accidents. Integrated phenomenological models are developed to describe the accident scenarios, which consist of many processes, while mechanistic modeling, including direct numerical simulation, is carried out to describe separate effects and selected physical phenomena of particular importance

[en] This thesis presents the work involving two broad aspects within the field of nuclear reactor analysis and safety. These are: - development of a fully independent reactor dynamics and safety analysis methodology of the RBMK-1500 core transient accidents and - experiments on the enhancement of coolability of a particulate bed or a melt pool due to heat removal through the control rod guide tubes. The first part of the thesis focuses on the development of the RBMK-1500 analysis methodology based on the CORETRAN code package. The second part investigates the issue of coolability during severe accidents in LWR type reactors: the coolability of debris bed and melt pool for in-vessel and ex-vessel conditions. The first chapter briefly presents the status of developments in both the RBMK-1500 core analysis and the corium coolability areas. The second chapter describes the generation of the RBMK-1500 neutron cross section data library with the HELIOS code. The cross section library was developed for the whole range of the reactor conditions. The results of the benchmarking with the WIMS-D4 code and validation against the RBMK Critical Facility experiments is also presented here. The HELIOS generated neutron cross section data library provides a close agreement with the WIMS-D4 code results. The validation against the data from the Critical Experiments shows that the HELIOS generated neutron cross section library provides excellent predictions for the criticality, axial and radial power distribution, control rod reactivity worths and coolant reactivity effects, etc. The reactivity effects of voiding for the system, fuel assembly and additional absorber channel are underpredicted in the calculations using the HELIOS code generated neutron cross sections. The underprediction, however, is much less than that obtained when the WIMS-D4 code generated cross sections are employed. The third chapter describes the work, performed towards the accurate prediction, assessment and validation of the CHF and post-CHF heat transfer for the RBMK-1500 reactor fuel assemblies employing the VIPRE-02 code. This chapter describes the experiments, which were used for validating the CHF correlations, appropriate for the RBMK-1500 type reactors. These correlations after validation were added to the standard version of the VIPRE-02 code. The VIPRE-02 calculations were benchmarked against the RELAP5/MOD3.3 code. It was found that these user-coded additional CHF correlations developed for the RBMK type reactors (Osmachkin, RRC KI and Khabenski correlations) and implemented into the code by the author, provide a good prediction of the CHF occurrence at the RBMK reactor nominal pressure range (at about 7 MPa). Transition and film boiling are also predicted well with the VIPRE-02 code for this pressure range. It was found, that for the RBMK-1500 reactor applications, EPRI CHF correlation should be used for the CHF predictions for the lower fuel assemblies of the reactor in the subchannel model of the RBMK-1500 fuel assembly. RRC KI and Bowring CHF correlations may be used for the upper fuel assemblies. For a single-channel model of the RBMK-1500 fuel channel, Osmachkin, RRC KI and Bowring correlations provide the closest predictions and may be used for the CHF estimation. For the low coolant mass fluxes in the fuel channel, Khabenski correlation can be applied. The fourth chapter presents the verification of the CORETRAN code for the RBMK-1500 core analysis. The model was verified against a number of RBMK-1500 plant data and transient calculations. The new RBMK-1500 core model was successfully applied in several safety assessment applications. A series of transient calculations, considered within the scope of the RBMK-type reactor Safety Analysis Report (SAR), were performed. Several cases of the transient calculations are presented in this chapter. The HELIOS/CORETRAN/VIPRE-02 core model for the RBMK-1500 is fully functional. The RBMK-1500 CPS logic, added into the CORETRAN provides an adequate response to the changes in the reactor parameters. Chapters 5 and 6 describe the experiments and the analysis performed on the coolability of particulate debris bed and melt pool during a postulated severe accident in the LWR. In the Chapter 5, the coolability potential, offered by the presence of a large number of the Control Rod Guide Tubes (CRGTs) in the BWR lower head is presented. The experimental investigations for the enhancement of coolability possible with CRGTs were performed on two experimental facilities: POMECO (POrous MEdium COolability) and COMECO (COrium MElt COolability). It was found that the presence of the CRGTs in the lower head of a BWR offers a substantial potential for heat removal during a postulated severe accident. Additional 10-20 kW of heat were removed from the POMECO and COMECO test sections through the CRGT. This corresponds to the average heat flux on the CRGT wall equal to 100-300 kW/m². In the Chapter 6 the ex-vessel particulate debris bed coolability is investigated, considering the non-condensable gases released from the concrete ablation process. The influence of the flow of the non-condensable gases on the process of quenching a hot porous debris bed was considered. The POMECO test facility was modified, adding the air supply at the bottom of the test section, to simulate the noncondensable gas release. The process was investigated for both high and low porosity debris beds. It was found that for the low porosity bed composition the countercurrent flooding limit could be exceeded, which would degrade the quenching process for such bed compositions