FIG 7 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Attenuation spectra, (a) ZPG, U 1 ¼ 30 m/s; (b) ZPG, U 1 ¼ 60 m/s; (c) FPG -12 ; U 1 ¼ 30 m/s; (d) APG 12 ; U 1 ¼ 30 m/s.
Source publication
A correction model is presented for sensor-size-related high-frequency attenuation when measuring the wall pressure fluctuations beneath turbulent boundary layers. The model is developed based on the wall pressure spectra measured on a flat plate model, using sensors of different sizes and types. The measurement covers the range of Reynolds numbers...
Context in source publication
Context 1
... that the flush-mounted Kulite sensor without screen is not considered any further due to the strong spectral disturbances present for all measured flow conditions. Figure 7 shows the obtained spectral attenuation for different sensors. Apart from the low-frequency disturbances, the attenuation is negligible at low frequencies (below 1 kHz) except for the 1/4-in. ...
Citations
... However, in practice, the sensitivity of a sensing element is larger in the middle area and becomes smaller approaching the edge [4,5]. The non-uniform sensitivity over the sensor surface can significantly reduce the effective sensor radius which decreases the attenuation level compared to that calculated with a uniform sensitivity [6,7]. Furthermore, is frequency and convection-distance dependent, and it also depends on the flow condition, i.e. ...
... Furthermore, is frequency and convection-distance dependent, and it also depends on the flow condition, i.e. Reynolds number and pressure gradients [7][8][9]. In the Corcos model [3], the value of was not explicitly defined. ...
... In a recent study, Hu [7] calculated the effect of non-uniform sensitivity over a sensing surface on the resulted attenuation, using a sensitivity distribution of a condenser microphone [4,11]. The results showed that the effective radius of a sensor regarding the attenuation spectra is significantly reduced with a factor of 0.73, due to the non-uniform sensitivity. ...
The paper presents the correction of sensor-size-related attenuation for wall pressure fluctuations measurements beneath turbulent boundary layer flows. The wall pressure spectra were measured with three commonly used flush-mounted sensors with protection screens, one flush-mounted surface sensor, and one pinhole-mounted sensor that measured the unattenuated spectra as reference. Zero, favorable and adverse pressure gradient flows were realized on a flat plate model, where the pressure gradients were produced by a NACA 0012 airfoil installed above the plate. The flows cover the Reynolds number in the range of 2200 < < 9100 and the boundary layer shape factor of 1.3 < < 1.85. The Hu model is applied to correct the attenuation measured with the flush-mounted sensors. Convincing correction results are obtained for all the sensors and flow conditions.
... In [17], Hu extends and generalises the idea to correct the measured test spectra of surface microphones based on assumptions of a microphone response kernel and correlation functions of an attached turbulent boundary layer flow above the sensing surface. In this paper, we vary this approach and outline how a finite sensing surface can be included in simulations and improves the match between simulation and test results, independent of the flow situation. ...
... A first study of the resulting attenuation based on the assumptions of a uniform membrane sensitivity and a zero-pressure gradient turbulent boundary layer was already given in the 1960s by Corcos [25]. Very recently Hu discussed [17] several more general approaches to correct the attenuation of the measured spectra which incorporate spatially varying microphone sensitivities as well as more advanced semi-empirical models to reconstruct . ...
... We first focus on the near field validation and assess the results of the LES runs on at the Knowles microphone and the selected 1/4" UTP sensor on the airfoil. The Knowles sensor was mounted behind a pinhole with diameter 0.5 mm, which, according to Hu [17], leads to an almost non-attenuated power spectrum and hence an evaluation with a point probe should show a good agreement. Since the mesh's surface size on the airfoil is about 0.375 mm, there would be also no additional advantage of looking at other response kernels. ...
Whenever numerical simulation approaches must be validated, it is required to correlate them either with test or with analytical solutions. Due to the turbulent noise generation mechanisms, such a quantitative verification in aeroacoustics is usually limited to comparisons with test data. In this paper we attempt to demonstrate how aeroacoustics simulation can be compared with a test of a rod-airfoil experiment. Therefore, the 2004 experiment from Jacob et al. was revisited and equipped with additional microphones to allow a validation of the numerical method in the near and far field. We demonstrate that near field measurements obtained with surface microphones can only deliver insights into the validity of the simulated turbulent pressure fluctuations and that it is required to incorporate an approximation of the microphones’ response function in the simulation. In the far-field, we demonstrate that modelling installation effects is important in order to capture all details of a spectrum. In simulations of real-world scenarios, it is often required to compromise accuracy and computational costs. In order to explore these aspects we show how different turbulence models and grid resolutions impact the quality of the results and outline how to setup efficient simulations that are still accurate.
... The Kulite sensor can measure a wide frequency range, covering the high-frequency spectra which cannot be properly measured by the MEMS microphones. Due to the small size of the pinhole, the signal-average effect can be effectively reduced, which would cause a dramatical spectral attenuation at high frequencies [14,15]. However, the Helmholtz resonance will be excited due to the pinhole configuration, which affects the high-frequency spectra and the effect is also flow condition dependent [16,17]. ...
The paper describes the experimental setup of a wind tunnel test in the DNW's low speed wind tunnel in Braunschweig (DNW-NWB) to investigate the boundary layer in the cockpit and fuselage area for the DLR project INTONATE. In order to gain knowledge about this accelerated flow during the test, the sensors and the sensor installation play a major role. The goal of the experiment is the measurement of the turbulent boundary layer's unsteady pressure autospectra and respective cross-correlation [1] with high spatial resolution. Therefor the sensor itself has to be small to avoid spatial averaging on the sensor surface and to allow for minimum distances between sensors in order to measure small correlation lengths. These challenges are met with a microphone MEMS array which will be fully integrated into the cockpit surface. The paper gives an overview of the model construction, the test setup and the onboard instrumentation consisting of static pressure ports, the MEMS microphone arrays and pinhole-mounted Kulites, the latter serving as reference sensors to measure the boundary layer induced pressure fluctuations.
... Pinhole-mounted sensors are often used to measure wall pressure fluctuations beneath turbulent boundary layers (TBL). The advantage of this mounting type is that the small pinhole can significantly reduce the sensing area compared to the flush-mounted sensors, which would measure considerable attenuation for the wall pressure spectra at high frequencies, depending on the sensor size [1,2]. However, the advantage of the pinhole-mounted sensors is limited by the Helmholtz resonance, which is generated due to the pinhole and the volume behind it. ...
... The velocity for a spacing of the pinhole size and at frequencies around the resonance frequency is assumed to be more meaningful for the scaling. For estimation of , the Hu model [2] is applied, expressed as the acoustic excitation, depending on the sensors. At frequencies / > 5, the amplitude ofˆ, /ˆ, remains constant, but at different levels for different sensors. ...
... At high frequencies, the value remains constant which can be determined solely with Eq. (10). A scaling factor was proposed in [2] for presenting the spacing effect on the value of . Thus, the convection velocity used for the scaling is calculated with ...
... The cost is the high-frequency attenuation for the one-point spectra measurement caused by the signal averaging over the 'large' sensor surface. This effect has been reported by several studies [1,2], which showed that the attenuation increases with decreasing flow velocity and adverse pressure gradients. Besides the one-point spectra, the wall pressure coherence is an important feature that can significantly affect the noise generation. ...
The wall pressure fluctuations beneath turbulent boundary layer flows can be measured by flush- and pinhole-mounted sensors. The pinhole-mounted sensors have a small sensing area that can resolve the high-frequency pressure fluctuations induced by the small-scale structures of turbulence. However, the resonance of the pinhole structure and the difficulty of mounting are the disadvantages for practical use. In contrast, the flush-mounted sensors are mostly more accessible for practical application and do not face the resonance problem. The cost is the high-frequency attenuation for the one-point spectra measurement caused by the signal averaging over the 'large' sensor surface. This effect has been reported by several studies, which showed that the attenuation increases with decreasing flow velocity and adverse
pressure gradients. Besides the one-point spectra, the wall pressure coherence is an important feature that can significantly affect the noise generation. To the best of
the author’s knowledge, the sensor-size effect on the wall pressure coherence has not been systematically reported. In this work, the coherence was measured by pinhole and flush-mounted sensors on a flat plate with and without pressure gradients. The sensor-size effect is presented using different pairs of sensors with equal sensor distances in the spanwise direction.
... Second, the sensing area of the probe results in spatial averaging of high-frequent pressure fluctuations. Models such as that of Corcos [5] can be used to partly correct this effect but rely on several assumptions [6]. Reducing the sensing area mitigates the impact of potential errors introduced by such models. ...
The empirical calibration of remote microphone probes (RMPs), used to acquire unsteady pressure fluctuations in a wide range of measurement campaigns, often introduces spurious resonance into the estimated frequency response, i.e. the transfer function (TF), of the probe. To prevent this spurious resonance from affecting the unsteady pressure measurements, the TFs tend to be manually post-processed. Yet, such a procedure can constitute an additional source of uncertainty that hampers the accuracy of the results. A semi-empirical calibration method was developed in a previous study to tackle this problem: ASSIST (BAyesian proceSsing of SpurIous reSonance in calibraTion data). Through this technique, spurious resonance is removed and replaced with a physically correct alternative in a much less operator-reliant manner. Recently, a repository containing the source Python code of ASSIST has been made publicly available. However, its application requires a thorough understanding of the model and related assumptions , as well as Markov chain Monte Carlo. This paper serves as a practical guide that provides the reader with all the knowledge required to apply ASSIST to the calibration of RMPs. It covers the acquisition of the calibration data, the setup of the method, the analysis of the results, and the iterative tuning of the input parameters to achieve the optimal fit. These steps are demonstrated on an acoustic finite-element method simulation dataset, allowing the analytic line-cavity model to be compared to a 3-D model. Several examples are provided throughout the manuscript to show the impact of the relevant parameters constituting the method.
... Typically, the Corcos's and somewhat improved correction factors are applied to account for this effect and provide useful spectral measurements at up to frequencies several times the normal 3 dB damping point. 33,34 Figure 11(a) shows the typical unprocessed pressure sensor signal, and Fig. 11(b) shows the Fast Fourier Transform (FFT) amplitude of this signal calculated over a t ¼ 0.4-1.2 s time domain. ...
This study focuses on the details of the geometry and dynamics of sidewall vortices observed in supersonic wind tunnels with a rectangular cross section of the nozzle and the test section. The formation of sidewall vortices limits the accuracy of the data measured during wind tunnels' testing due to a reduced area of uniform core flow results. Most of the test data presented in this work are generated using Mie scattering visualization for M = 4 flow, with CO2 seeded up to 7% mole fraction. The Mie scattering results are complemented by data from fast pressure sensor and schlieren visualization. It is shown that the formation of vortices is caused by a transverse pressure gradient realized in the supersonic nozzle due to the gas under-expansion. The vortex external mixing layer is strongly perturbed in time but remains globally geometrically similar with streamwise distance. The vortex-generated dominant flow disturbances are in the frequency range of f = 10–50 kHz, doubling the magnitude of baseline power spectral density. The authors' viewpoint is that sidewall vortex generation is a more generic phenomenon than was thought previously.
