CCES Unicamp

Analysis of turbulent flows over airfoils using large-eddy simulations

Date: Set 03, 2021

Candidate: Gabriel de Barros Nogueira

 

Advisor:

Prof. Dr. William Roberto Wolf, Instituto de Computação
 
Abstract:
Large eddy simulations with wall resolution are performed with the objective of investigating turbulent flows and coherent structures using statistical tools. The studies are conducted using the NACA0012 airfoil with Reynolds numbers Rec = 200,000 and 800,000 for an angle of attack 5º and Mach number M = 0.115. The simulations are solved considering two overlapping meshes that allow an adequate resolution of both the boundary layer and aerodynamic wake. High-order compact schemes for differentiation, interpolation and filtering are employed in the solution of the flows and a random tripping methodology is proposed to transition the boundary layer. Flow properties such as the Rotta-Clauser parameter, boundary layer thickness, shape factor, mean velocity profiles and Reynolds stresses, besides friction and pressure coefficients are presented. The proper orthogonal decomposition (POD) technique is also used with different norms for calculating the covariance matrix in two flow regions: the wake and the suction side boundary layer. Along the wake, two-dimensional structures highly correlated with the v-velocity norm are observed and these structures are responsible for tonal noise generation. In the boundary layer, streaks are identified by the norm of u-velocity fluctuations. The POD technique was used to reconstruct the flow fluctuation field with different percentages of energy in order to verify the role of coherent structures in the statistics of the fluctuation field. In this case, Reynolds stresses, flow anisotropy, the trajectories along the Lumley triangle, turbulent kinetic energy balances, scatter plots and probability density functions were analyzed to understand the role of streaks in the present turbulent flows.
 

Related posts

Alternative to the solution of the Schrödinger equation for problems of atomic electronic structure

cces cces

Automatic Offloading to FPGA Accelerators

cces cces

Improving Hardware/Software Transactional Memory Codesign: A Phased-based and Over-Instrumentation Elimination Approach

cces cces
WP Twitter Auto Publish Powered By : XYZScripts.com