Holger Frahnert

Application of Stereo PIV to the VFE-2 65° Delta Wing Configuration at Sub- and Transonic Speeds

The present paper describes the application of Particle Image Velocimetry (PIV) to the flow around a delta wing in a pressurized transonic wind tunnel. These investigations are the second part of International Vortex Flow Experiment 2 (VFE-2) measurements carried out at DLR Gottingen on a NASA wind tunnel model. The first part comprised surface pressure measurements by means of Pressure Sensitive Paint. These results were used to select the test cases for PIV. Flow fields for the Mach numbers of 0.4 and 0.8 and for the Reynolds numbers of 2 and 3 million at four incidences of 11, 13, 20 and 25 degree were captured within different planes perpendicular to the model axis. Delta wings with sharp as well as with rounded leading edges were investigated. Details of the stereoscopic PIV system such as the overall arrangement, the image evaluation and the flow seeding technique will be described. Techniques to overcome problems caused by density changes within the wind tunnel are described too, such as deflections of the laser light beam and camera viewing. Laser light flare on the model surface is reduced by a specially developed fluorescent paint which allows for the detection of small flow structures very close to the model surface.

Examination of the Eddy-Viscosity Concept Regarding its Physical Justification

In industrial application, the eddy-viscosity concept is the root of nearly every turbulence model in use since the whole class of zero-equation up to two-equation models relies on this hypothesis. Even in large eddy simulation, the eddy-viscosity hypothesis is the base of many subgrid-scale models. In view of that a clear understanding of the theoretical basis and practical perspective of the concept is desirable. Valuable information on this issue is accessible by a comparative study of the principal axes and the eigenvalues of both the real and the modeled Reynolds-stress tensor. Conclusions drawn from this analysis are independent of any particular turbulence model but result from the structure of the underlying eddy-viscosity ansatz. The procedure is exemplified with plane channel flow. Important in practice is the modeling of complex flow problems with separation. Despite of the apparent simplicity of the case regarded, the insights gained from the present analysis are influential to problems of practical interest. In summary, the eddy-viscosity ansatz is judged unsustainable from both a physical and an engineering predictive point of view.