Validation of CFD models for jets and plums in fission reactors

Validation of CFD models for jets and plums in fission reactors. The cooling system of nuclear reactor is very important, as the safety of the whole system significantly depends on it. Nowadays, it’s design and the design of the fission and fusion power plants depends on the models that can simulate a fluid flow in a given geometry taking into account all the environmental and other conditions. There are variety of models available that are results of different approaches, thus the computational time will also vary depending on the approach. Nevertheless, most of these approaches are based on the Navier-Stokes equations. So having the availability of too many models, only some of them will be implemented to simulate a fluid flow and then verification and validation process will be performed too. The analysis concerns mostly the fluid flow in an upper plenum of the cooling system. Verification and validation process (V&V) will be implemented on a scaled-down (1/16) version of a High-Temperature reactor. During this thesis work different models and approaches were tried and implemented which consequently led to diverse and not so much results. In the beginning, the several RANS models were used in the simulation and results demonstrated that they can be pretty precise and sufficiently reliable if the initial conditions, boundary conditions, mesh size and right distribution were set correctly. Although the RANS models are simple and do not require high computational rate and large computational time, the initial conditions can have dramatic effect on the final results, for example: even a 1 degree of temperature difference will change the viscosity of the fluid and also some other less important parameters, leading to the results which differ more from the experimental data. so that the initial condition is not adequate anymore. Situation is much more complicated with the boundary conditions, because there is no data which can be considered as ultimately correct, that is why here several approaches can be considered. With the trial and verification procedure the more suitable boundary conditions can be chosen, although having a good knowledge of fluid dynamics, physics of turbulent flow at different Reynold’s number or just a considerable experience, can save much time to choose the correct boundary conditions. The LES model is much more complicated not only in the terms of the numerical solution and consequently computational time, but also in terms of preparation of feasible parameters and settings for the model. During the work it was observed and discovered that meshing parameters and delta time can be considered as key factors of the results. Being a transient process, it involves more field functions and some of them can help to understand the validity of the model, including mesh sizing and determination of the time step. Moreover, it would be much helpful to set some appropriate monitors which can also help to estimate the validity of the transient model and in case of necessity to motivate to make some changes in order to obtain better results. Through multiple trials and errors, finally, it has come up to the point that in the case that all the initial settings are set in a correct way, by Wale Subgrid Scale (LES) model it is possible to obtain very precise results which can be very useful in the design of a cooling system of the 4th generation fission reactors.