Clinometric measurements by means of high-accuracy PIV in the ONERA F1 Low Speed Pressurized Wind Tunnel
Résumé
Recent needs in the understanding and the quantification of wind tunnel flow characteristics led to the development on an innovative use of an existing PIV system for clinometric measurements at the ONERA F1 low speed pressurized wind tunnel in le Fauga-Mauzac center. The large industrial F1 wind tunnel is part of a full set of wind tunnels monitored by ONERA and ranging from low speed to high speed and high enthalpy flows. The motivation of this development originated from the need to better quantifiy flow up and side wash in the empty test section of the wind tunnel. In usual use of 3C-PIV systems, the 3 components of the velocity vector are evaluated from particle displacements of the order of 10 pixels in the image plane. Considering a typical uncertainty of 0.1 pixel, which is often a realistic estimate, the relative uncertainty on the velocity components cannot be better than about +/- 0.1 pixel/10 pixels = +/- 1 % of the flow velocity.This translates into uncertainty of the order of +/- 0.6° on flow angles, which is insufficient to qualify upwash and sidewash in an empty test section. To overcome this issue, benefit was taken from the fact that the flow in the empty wind tunnel section is nearly aligned with the wind tunnel axis and has a very low level of turbulence. As a consequence, iti is possible to keep patterns of particles within the laser sheet for very long times, and therefore to lengthen their displacement in the image plane. It was found that displacements of 100 to 200 pixels could be obtained, while keeping a high correlation factor between images, of the order of 50 % using the ONERA FOLKI-SPIV image processing software(1). PIV displacement estimation now limits the relative accuracy only to +/- 0.1 pixel/200 pixels = +/- 0/05 %, which translates into flow angles of +/- 0.03°. This high accuracy was also used to assess empty tunnel velocity, therefore complementing and even replacing tunnel calibration using static pressure measurements. Since 200 pixels corresponds to a displacement of a few centimetres, this comes at the cost that the laser sheet cannot be placed transversal to the flow, as done usually to investigate transversal distribution of flow angles. Results obtained with this technique were compared to measurements of the average flow upwash carried out at the test section centre thanks to the standard aircraft model inversion approach. Also, cross-comparison of PIV measurements from different test campaigns and different optical setups (both 2C- and 3C-PIV) were carried out. These consistent comparisons, together with uncertainty assessments, allow estimating that uncertainty on upwash and sidewash measurements is better than +/- 0.05°, strongly linked to image processing, occurrence of peaklocking and positioning of the calibration body in test section. This PIV technique allows investigations of flow angularity in areas where model inversion is not possible. For example, measurements near the test section floor have been achieved in order to determine upwash corrections to be applied on a model tested in ground effect. Compared with traditional 5-holes probing, this PIV-based technique has two important advantages: non-intrusiveness and high spatial density of measurements, enabling a better understanding of upwash and sidewash distribution in the test section and the capability to act towards its improvement. These measurements follow continuous progress in the capability to perform productive PIV measurements in the stringent environment of the F1 wind tunnel (optical access, pressurized conditions, remote-controlled optical system). Measurements have been performed fo Mach numbers between 0.15 ad 0.23, in majority at the maximum tunnel stagnation pressure of 3.85 bars.
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|---|
| Licence |