Hiroaki Ohira

Prediction of Release Rate of Burnt Sodium as Aerosol

Evaluation of the consequences of sodium pool fire requires a prediction model that can calculate the fractions of the aerosols of Na2O and Na2O2 being transported into the atmosphere and onto the sodium pool surface from the reaction zone. In the present model, we applied the laminar diffusion flame theory in a thin layer in the neighborhood of the pool surface. The combustion was modeled assuming the instantaneous chemical equilibrium because the chemistry is fast compared with the mixing. The aerosol transport was assumed to be due to a sum of the drag force, the thermal force and the gravity. The calculated results are in good agreement with experimental data obtained by other investigators for the burning rates and the aerosol release fractions of sodium pool fire. This analysis forms the basis of prediction of aerosol release fraction in the sodium pool fire.

Numerical Simulation of Aerosol Behavior in Turbulent Natural Convection

This paper describes the detail numerical methods for the aerosol behavior in the cover gas regions of LMFRs by considering the aerosol growth as one of the space coordinates. The governing equations for the compressible gas were approximated by removing sound waves for the efficient calculations of the large temperature different fields. The convection terms of these equations were discretized by the 3rd order upwind scheme which was derived for irregular staggered mesh. In order to verify this method, a natural convection in the vertically heated enclosure was calculated and the results were compared with the benchmark results by de Vahl Davis. In this calculation, temperature distribution in the enclosure agreed well with each other. The basic heat transfer tests for the cover gas region of LMFRs were also calculated by this method and the calculated vertical average temperature profile of the turbulent cover gas natural convection and the aerosol average mass concentration were compared to the experimental ones. The calculated temperature agreed well with the experimental ones. It was also estimated that the aerosol were well mixed in the natural convection field and the number densities distributed the similar profiles in almost all region except for the region near the top surface.

Thermal Fluid-Structure Interaction Analysis for the Upper Structure of LMFRs

This paper describes a thermal fluid-structure interaction analysis code FLUSH that calculates both thermal-hydraulics and thermal structure response at the same time. This code has been developed to evaluate the thermal responses of the upper structures of LMFRs, using two different analysis codes of α-FLOW and FINAS. The heat flux on the boundary surface of the fluid region and the temperature on the boundary of the structure region are exchanged in every iterative cycle as the new boundary conditions, and finally the unified thermal fields are calculated. The different mesh method and the detail thermal radiation model were also developed to apply for the large scale models. The 2-D model of the basic experiment for the cover gas thermal-hydraulics was calculated to verify this iterative method. The calculated average temperature on the boundary agreed well with the experimental results. The 3-D large scale models of the out-of-pile experiment for MONJU shield plug were also calculated to verify this method. The calculated temperature both in the annulus and the shield plug agreed well with the experiments. These studies showed that this iterative method of FLUSH was very effective for the predictions in the strong coupled thermal fields.

Magnetohydrodynamic Turbulent Model for LMFRs

In order to study magnetohydrodynamic behavior in electromagnetic pumps, electromagnetic flow meters, etc. for Liquid Metal Fast Reactors (LMFRs), a large eddy simulation method using an artificial wall boundary condition was developed. In this study, Spalding’s law of the wall and the eddy viscosity for uniform magnetic fields, which was proposed by Shimomura, was applied to Finite Element Method of Generalized Simplified Marker and Cell (GSMAC-FEM). We calculated MHD channel flow in various element sizes on the conditions of Hartmann numbers of 0, 52.5 and 125, whose Reynolds numbers based on the average velocity were all about 29,000. These results showed the average velocity profiles were in good agreement with both the experimental results by Brouillette-Lykoudis and the detail calculation results by Shimomura, although farther calculations were needed to verify the turbulence intensities.Copyright