Caracterização do escoamento do túnel de vento de circuito fechado do laboratório de aerodinâmica experimental da UFMG via fluidodinâmica computacional e experimentação utilizando um modelo em escala real

Abstract

The airflow in a wind tunnel section does not behave in the same way as it does in an undisturbed environment. Therefore, characterizing this flow, as well as the wall interferences and supports in aerodynamic measurements, is of great importance to ensure the reliability of the design of an aircraft, for example, from the early stages of testing to the most executive phase, regarding the performance, stability, and safety of this aircraft. This research work presents a comprehensive study of the turbulence and characteristics of the closed-circuit, atmospheric, low-speed (V less than 135 m/s) Göttingen-type wind tunnel. To characterize the flow, a three-dimensional model of the air flow field inside the entire wind tunnel was developed at full scale using the finite volume method, employing the Reynolds-Averaged Navier-Stokes (RANS) approach with turbulence models to solve the Navier-Stokes equations in a steady state, namely: i) k-omega; ii) k-εpsilon; iii) Spalart-Allmaras. Subsequently, with a temporal variation applied, the resolution methods by scale were studied, namely: iv) Detached Eddy Simulation (DES); v) Large Eddy Simulation (LES). The main objective of this work is to assess the accuracy of virtual models, as computational processing capacity was provided by Siemens, allowing for the use of unstructured discrete grids containing up to approximately 200 million elements. To calibrate the theoretical CFD models and later correlate them, an experimental testing campaign was conducted in the mentioned wind tunnel. The aim of this master's research is to characterize the flow in the test section, especially with respect to axial turbulence and boundary layer thickness. The steady-state model, using k-ω SST, presented an accuracy of 88.92% for the pressure drop of the flow that crosses the honeycomb. The transient analyzes showed even better results, with the LES method achieving a correspondence of 94.84% of the values.