Low Speed Tests using PSP at ONERA

Low speed tests using the PSP technique were performed by Onera on a delta wing and on a model of a 206 PSA Peugeot Citroen car. A careful choice of the paint as well as a specialized procedure and a devoted software have provided accurate results, with a maximal error of 1mb using the "a priori" calibration method. The main sources of errors have been identified and their relative importance has been assessed. As a result, a very good agreement with conventional pressure taps has been obtained. Fine details can be seen on the final pressure maps, showing the benefit of PSP for low speed tests. The presented results show that PSP can be used as an usual tool for low speed tests. The used paint degrades during the test. The degradation is taken into account in the data reduction procedure. The degradation rate correlates well with the qualitative shear stress distribution obtained with the conventional sublimation technique. Introduction the flow behavior has to be considered at high Reynolds numbers '. Hypersonic applications are more difficult because of the temperature sensitivity of the PSP ' and of the .short flow duration . A large effort has been paid for low speed applications '? but with a limited success because of the sources of errors (see below) which bound the accuracy of the technique. Nevertheless, low speed applications are very important for the aircraft industry as well as for the car industry. This is the reason why a strong effort is underway at Onera in order to provide a reliable PSP technique for low speed tests. Limitations at low speed conditions The involved phenomenon in PSP is a photochemical process: some luminescent coatings absorb energy of the incident light (UV or visible light) and release part of it at longer wavelength . When this luminescence is quenched by oxygen, the increasing air pressure decreases the intensity of the coating luminescence. The relationship between this intensity (noted I in eq. 1) and the pressure P is known as the Stern-Volmer law: Pressure Sensitive Paint (PSP) is an efficient technique for measuring surface pressure in wind tunnel testing. The time saving in model manufacturing and the resolution of this method has originated a strong interest in the United-States \ Russia 2 and Europe. Since the 1980's, the technique has been mainly applied for transonic and supersonic flow studies '' and it is now commonly used in industrial wind tunnels . The obtained accuracy can be as good as 0.01 in Cp value. . However the effect of the paint on Copyright© by Y. Le Sant. Published by the American Institute of Aeronautics and Astronautics, Inc. with permission Iref(Pref,Tref)/I(P,T)=A(T)+B(T)*P/Pre, (1) where the subscript ref represents the reference conditions (usually at ambient pressure Pref and ambient temperature Tref). The B coefficient provides the sensitivity of the PSP and varies from 0.3 to nearly 1 for the most sensitive PSP formulations. The basic data reduction comes directly from this equation: take the reference and run images, subtract the darkness, divide the two images and convert the ratio into pressure using equation (1). PSP is an absolute measurement technique as shown by the equation (1). This is the reason of limitations at low speed conditions where the required accuracy 1 American Institute of Aeronautics and Astronautics c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization. is typically 1mb. This represents an absolute pressure accuracy of 0.1% (assuming Pref=lbar). This high accuracy is difficult to reach, because of sources of errors related with the optical PSP technique. The main sources of errors are the temperature and the variation of the excitation intensity during the test. In order to make the following discussion clear, let us consider a paint very sensitive to pressure, i.e. the slope of eq. (1) (the coefficient B), is equal to 1. In this case, a pressure change of 1mb leads to an intensity variation of 0.1%. All the PSP formulations are temperature sensitive. For a typical sensitivity of 1%/K, an error of 1mb is induced by a temperature change of O.IK, if no temperature compensation is applied. A common way to correct for the temperature effect is to apply the so-called "in-situ" calibration, when the model is equipped with conventional pressure taps. The discrete pressure measurements are used to calibrate the paint during the test assuming an uniform temperature on the model. However this requirement is nearly impossible to fulfill within an accuracy of O.IK. Therefore better results will be obtained using low sensitive temperature paints. The second important source of errors is the model motion between the reference image and the test image. There are two induced effects. The first one is slight changes in the model position between the reference and run images. The run image must be aligned on the reference image before dividing them. This is done usually by dedicated software ' known as "resection" software. The second effect induced by the model motion is the changes in the illumination pattern. They are not corrected by the resection software and may introduce significant errors. Therefore, the illumination pattern changes have to be modeled or to be measured by another way. Temporal (or global) intensity variations of the excitation light source (lamp or laser) do not change the illumination pattern on the model. Therefore, the "in-situ" calibration method compensates for this global intensity change between reference and run measurements. There are other sources of errors which can be usually neglected with transonic flows. They are related with the excitation lamp stability and the accuracy of the camera. The required accuracy of 0.1% is typically of the same magnitude of the noise due to the camera (readout noise, shutter repeatability). Averaging a large number of images enables noise reduction (ex: 10 images decreases the noise level by a factor 3). However, the final pressure image remains noisy. This is due to the readout noise of the camera and also to noises introduced by the data reduction software. A spatial filtering is always required and its parameters have to be tuned to keep the final accuracy better than 0.1%. From all of these remarks, it comes that low speed tests are usually performed using the "in-situ" calibration method and several images (more than 10) are acquired and averaged to reduce the level of noise. The drawback of the "in-situ" calibration is the need for pressure taps on the model. Then the following question arises: how many of them are necessary and where should they be located? There is no satisfactory answer to this question when the pressure distribution is not known (before any PSP testing). So, the research groups developing and using the PSP technique at Onera have decided to systematically apply the "a priori" calibration method: the PSP is calibrated in a devoted calibration chamber before the test and the resulting calibration law is applied during the test. Only a few pressure taps are implemented in the model to evaluate the PSP accuracy. This approach requires a PSP with very low temperature sensitivity and a way to measure the local excitation light (or to correct for both spatial and temporal excitation variations). These two features are described below. PSP hardware and software