Listing 1 - 2 of 2 |
Sort by
|
Choose an application
Transportation represents a 20% of green-house effect gases emissions worldwide nowadays. If none solutions are proposed, with air traffic increasing dramatically fast, none of the global politics to reduce the carbon footprint generated by human activities would be fulfilled. Aeronautic can get involved into this new global renovation via improving the aircraft's efficiency thanks to a reduction in the structural weight, an improvement on the aerodynamics efficiency or an expansion of free-fossil fuels powered systems. The combination of the first two lead the design of light and low stiff, highly loaded wings, which are subjected to significant deformations. Therefore, the aeroelastic deformations of these light wings is of paramount importance as it affects both the structural design and the aerodynamic performance. The present Master's Thesis is aimed at assessing the transonic aerodynamic and aeroelastic performance of a full aircraft configuration with full potential aerodynamics low-fidelity modeling techniques that are designed to suit the low computational cost of the preliminary stage of an aircraft design process. The benchmark full aircraft configuration of the present project is the Common Research Model with its wing-body-tail arrangement developed by the National Aeronautics and Space Administration. First, the model is adapted and validated to fit the requirements of a full potential aerodynamic solver. The later includes the generation of sharp trailing edges of the lifting surfaces and the inclusion of wake boundaries to enforce Kutta condition. The full potential Common Research Model is afterwards validated via three-dimensional aerodynamic simulations that compare results of three different fidelity levels: Reynolds-Averaged Navier-Stokes, Euler's aerodynamics and full potential aerodynamics. The results prove a validation of the full potential model and evince that at the transonic flight design point, the capturing of the position of the shock is moved downstream when decreasing the level of fidelity. Finally, fluid-structure interactions are evaluated in the context of static aeroelastic computations. The results illustrate sufficiently reliable static deformations of the wing at the design flight condition with a low-fidelity fluid solver.
Aerodynamics --- Aeroelastics --- DART --- Full potential aerodynamics --- Common Research Model --- CRM --- Full aircraft configuration --- Ingénierie, informatique & technologie > Ingénierie aérospatiale
Choose an application
The preliminary aircraft design is often performed based on low-fidelity aerodynamic models facilitating the evaluation of best-suited aircraft configurations thanks to low computational costs and reasonable accuracy at this early design stage. The Full Potential equation, based on the inviscid and isentropic assumptions, has demonstrated its ability to meet those requirements. However, the mathematical nature of this partial differential equation highlights that when the flow switches from subsonic to supersonic, it converts from elliptic to hyperbolic. This flow physics change needs to be reflected in the numerical implementation. DARTFlo, a full-potential solver, is implemented based on a physicsdependent solution experiencing mesh-dependency. Thenceforward, the present thesis aims at characterising the mesh-dependency of this physics-dependent solution and to propose alternatives to withdraw it. The current physics-dependent implementation is studied through a mesh convergence analysis in three different test cases to characterise the mesh-dependency. The analysis relies on two comparison axes, the first is a study of global flow parameters and the second treats the problem from a local point of view. The three test cases are constructed to study the behaviour of each solution in different situations. The original DARTFlo implementation illustrates its mesh-dependency by local flow parameters which do not converge with respect to the mesh refinement as well as by instabilities appearing in the supersonic zones when the mesh is highly refined. In parallel, three alternatives are derived and compared with the original implementation to assess their improvements in removing the mesh-dependency problem. The first alternative demonstrates improved mesh convergence and enables to partially remove the results mesh-dependency according to the case studied. However, the two others do not reveal to act on the mesh-dependency of the physics-dependent solutions.
Listing 1 - 2 of 2 |
Sort by
|