|
Executive summaries of Action Groups of Group of Responsables in Aerodynamics
As part of the GARTEUR AD AG24 project of “Navier Stokes Calculations of the Supersonic Flow about Slender Configurations”, 3 ogive cylinder configurations in supersonic flow have been calculated d using various Euler/Navier-Stokes methods of the GARTEUR AD AG24 participants. ONERA has provided two test cases, a cylinder body without fins with respectively a cylinder and an ogive nose section. In the first case the flow on the body is laminar whereas for the second case the flow is turbulent. For this configuration ONERA has also delivered basic and fine grids to be used/adapted by the participants and experimental data for comparison/verification. The third test case has been provided by DERA (now QinetiQ) and is a configuration with fins present. The following organisations participated in the activities: ONERA, QinetiQ, EADS-Launch Vehicles, Aerospatiale-Matra-Missiles, NLR, CIRA, FOI and SAAB-Dynamics. Various participants used their own methods to calculate these test cases like EULER codes and Navier-Stokes methods (full Navier-Stokes, Parabolised Navier-Stokes, Thin-Layer Navier-Stokes). For the first test case (laminar flow without fins) the results are generally in a good agreement. The computed surface pressures, the overall forces, the position of the primary flow separation and the boundary layer profiles agree satisfactorily. However, the computed skin friction coefficients and the friction component of the computed axial forces are only in a fair agreement. Comparison of the computed results obtained on the basic grid and those obtained on the fine grid shows only minor differences for each method. Comparison between results obtained using the Euler equations and those obtained using the Navier-Stokes equations show deficiency of the Euler equations in modelling the flow separations, which occur on the smooth surface of the body. Results of a pressure integration up to a certain axial position start to show differences between the Euler and the Navier-Stokes results at x/D »4.5 where the flow begins to separate from the body. For the second test case, a smooth body with turbulent flow, the results show much larger differences notably in the prediction of the separation that initiates the formation of a vortex. It also appears that the results are very sensitive to the turbulence model used. These differences are clearly reflected in a large variation in the predicted drag and the centre of pressure location. Like for the turbulent case, Navier-Stokes codes are essential to calculate this flow, For the third case, a body with fins present, the agreement is much more satisfactorily although there are still significant differences in the axial force. The in general better agreement can be understood from the presence of the fins in this case that fix the formation of the vortices geometrically. Details of the turbulence models are then less important. For the same reason, the Euler calculations show much better agreement for this case.
|
