Thermal fatigue is a relevant problem in the context of life-time extension of nuclear power plants (NPP). In many piping systems in NPPs hot and cold water is mixed, which leads to high temperature fluctuations in the region close to the solid wall and resulting thermal loads on the pipe walls that can cause fatigue. One of the relevant geometric test cases for thermal fatigue is the mixing in T-junctions. In this study we apply large–eddy simulations (LES) to the mixing of hot and cold water in a T-junction. We perform a set of simulations by using different formulations of the LES subgrid scale model, i.e. standard Smagorinsky and dynamic procedure, to identify the influence of the modelled subgrid scales on the simulation results. The results exhibit a large difference between the models, which is caused by the use of turbulent viscosity wall–damping functions when applying the standard model.

This content is only available via PDF.
You do not currently have access to this content.