Quantification of Turbulence Statistics for the Near-Wall Region in Unstructured Pebble Beds Using Direct Numerical Simulation
MetadataShow full item record
The aim of this study was to employing high-fidelity computational fluid dynamics to investi-gate and quantify the turbulent flow effects for incompressible, isothermal fluid flows, in the near wall region of unstructured, randomly packed spheres. The flow domain treated in this study is a replication of an experimental setup and analogous to those encountered in pebble bed based high-temperature reactors. A new meshing strategy and meshing assumptions have been employed to decrease the mesh size and these were validated with the experimental results. Quantifying turbulent flow effects sever a dual purpose: One, it assists lower-fidelity engineering tool development, such as those incorporating Reynolds averaged Navier-Stokes based methodologies, and two deepens our fundamental understanding of the physics involved for in-compressible flows over complex geometries. Nek5000, an open source spectral element computational fluid dynamics code, developed by Argonne National Lab was used to conduct this study. The code was used to perform a series of direct numerical simulations on the experimental geometry at low to moderate Reynolds numbers, matching the experimental flow parameters, to validate the model. This was done for a the full and section of the geometry to investigate the cross flow dependence in the problem. Presented results include the comparison between experimental and numerical findings, the development of a high-fidelity database of the experimental geometry at low and moderate Reynolds number, identification of possible flow phenomena present in random packed spheres as well as the calculation of the first and second order statistics in the near wall domain of random packed spheres.
Botha, Gerrit (2019). Quantification of Turbulence Statistics for the Near-Wall Region in Unstructured Pebble Beds Using Direct Numerical Simulation. Doctoral dissertation, Texas A & M University. Available electronically from