James Slaughter

PhD Candidate

Thesis: Towards High-Reynolds Number Large-Eddy Simulation on Complex Geometries within a Spectral/hp Element Framework

Abstract:

The project aims to extend the current capabilities of Nektar++ by including wall-modelling capabilities to reduce the near wall cell requirements for LES. Currently, fully resolved WRLES grid-requirements scale as a funciton of Reynolds Number squared making most industrial simulations, especially for external aerodynamics impossible. For the forseable future, until computational hardware allows otherwise, near wall-modelling - conducted by either imposing wall-shear stress through boundary layer approximations or more complicated boundary conditions that account for the sub-grid scales in both the buffer and viscous sub- layers must be utilised.

The project aims to assess within the Nektar++ framework, current wall-modelling techniques - assessing their applicability, accuracy and efficacy within the Spectral/hp Element construct - and implementing and applying these models to complex geometries and industrial scenarios. The hope is that the project both extends the available research space that Nektar++ can serve, whilst also improving current pipelines within existing research: reducing the computational requirements by offering another initialisation method that could reduce the required interrogation period for fully resolved resolutions.