JP2008082735A - Image measuring method of unsteady pressure field by pressure-sensitive paint - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a procedure for performing image measurement of a distribution of a fluctuating pressure on a field wherein its fluctuating component has a component spread over a wide region, which is a periodical unsteady pressure fluctuating field, and a procedure of measuring a distribution of a fluctuating pressure component even in the case of a minute pressure field having little pressure fluctuation. <P>SOLUTION: In this image measuring method of an unsteady pressure field by a pressure-sensitive paint, which is used for image measurement using the pressure-sensitive paint or a pressure-sensitive coating, the distribution of the pressure fluctuating component in a frequency dimension in light emission data is made into a picture and displayed based on the acquired light emission data from the pressure-sensitive paint or the pressure-sensitive coating. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

The present invention is an image measurement of pressure distribution using pressure-sensitive paint or pressure-sensitive coating (these are collectively referred to as PSP), and the phenomenon can be regarded as a periodic unsteady state. This is a technique for measuring and imaging up to a minute pressure level how and at what frequency component the pressure fluctuation component is distributed on the object surface. The distribution on the object surface such as the amplitude, power, and phase of the unsteady component is measured for each frequency band.
Applicable to a wide range of applications, such as transport machinery such as aerospace machines, railways and automobiles, and fluid machinery such as air conditioners and fans. This is a technique that can be used for the purpose of grasping the behavior of the unsteady field existing on the surface of the object, and designing the machine and improving the efficiency.
Although this method measures periodic unsteady fields, the information of unsteady pressure fields measured by this method is a state quantity closely correlated with the dipolar sound of aerodynamic noise. It can be effectively applied to the aerodynamic noise field.

High-speed response feeling such as pressure-sensitive coating with anodized film that can achieve response in the order of kHz, polymer / ceramic PSP with improved responsiveness by mixing particles and polymer, and TLC PSP with silica gel particles fixed to adhesive tape Up to now, unsteady pressure-sensitive paint measurement using pressure paint has been carried out. In Non-Patent Document 1 and Non-Patent Document 2 using a high-speed response type pressure-sensitive coating, a high-speed response type pressure-sensitive coating is coated on a model, and light emission from this is measured as a plurality of images by a high-speed camera. Stern-Volmer formula, which is a general data processing method for pressure sensitive paints;
Alternatively, based on this higher-order empirical expression, the pressure is measured using the measurement image I (wind-on image) which is an unknown pressure field and the measurement image I _ref (reference image or wind-off image) of the known pressure field. The distribution is calculated, this is repeated for the number of images in the time direction, and the obtained pressure distribution image group is arranged with respect to the time axis, and the result is displayed as a video image or an image group corresponding to each time. However, in these measurements, the camera measurement images are individually converted to pressure, and each image has an acquisition rate of 1 kHz and the exposure time of each frame is very small, about 1 ms, and the light emitted from the pressure-sensitive paint is fluorescent or phosphorescent. Coupled with the weak light, the S / N ratio of noise components such as shot noise to the measurement signal is large. As a result, the significant pressure resolution on the obtained individual images is on the order of kPa or higher, and pressure fluctuation components of the order of 100 Pa or smaller are buried in noise and cannot be detected. In addition, the processing result image group is displayed as a change in pressure distribution with respect to the time axis. What frequency component the pressure fluctuation has in each part is not shown in the frequency domain, and the part on the object where the fluctuation component is big or small is visually qualitative from time axis data It depends on judgment. In Non-Patent Document 1 and Non-Patent Document 2, analysis of pressure-sensitive paint data in the frequency domain at four points around the pressure sensor for comparison is performed. This evaluates the time response of the pressure-sensitive paint. This is done for the purpose, not for the purpose of grasping the distribution of the unsteady phenomenon in the frequency dimension on the object surface.

Non-Patent Document 3 and Non-Patent Document 4 also measure unsteady fields using a high-speed response pressure-sensitive paint, but this method does not use a temporally continuous image acquisition device. In this method, the phase of the phenomenon is phase-locked to specific information such as a signal from a reference sensor, and an image is acquired using integration with a sufficiently small exposure phase using a pulse LED. Non-patent document 3 uses a field with very good frequency separation, and the main purpose is to evaluate the response of a high-speed response pressure-sensitive paint. Non-patent document 4 is used for a smaller pressure fluctuation field. It is what. This method is an unsteady pressure fluctuation measurement method that has a single frequency component and can be applied to phase lock. However, it is a periodic flow characteristic that is a general characteristic of flow fields, but many frequency components are mixed. Since it is very difficult to perform phase locking when the target field is targeted, it is very difficult to measure a general flow field. In addition, since the target field is a single frequency field, only the fluctuating pressure distribution phase-locked to one frequency is presented, and what the field has in the frequency direction is known. The point is not considered. Non-Patent Document 3 and Non-Patent Document 4 also show examples in which pressure fluctuations in the frequency domain are measured by irradiating only one point of the pressure-sensitive paint application site with laser and measuring this using PMT. However, as with Non-Patent Document 1 and Non-Patent Document 2, this was done for the purpose of evaluating the time response of pressure-sensitive paints, and grasping the distribution of unsteady phenomena on the object surface It is not intended for.
In addition, if the fluctuation component has a single frequency, a method for measuring the fluctuation component image using a method such as phase lock has already been proposed. There is no technique for image measurement of the distribution of fluctuating pressure components in a pressure field having frequency components.
Japanese Patent Laid-Open No. 2006-10517 “In-situ Measurement Method and Apparatus for Pressure-Sensitive Paint with Temperature Dependence Correction” Published on January 12, 2006 Japanese Unexamined Patent Publication No. 2006-64600 “A-priori / In-situ Hybrid Pressure Sensitive Paint Data Processing Technique” published on March 9, 2006 Shoji Kameda, Takahiro Tabe, Tomohiro Hanya, Takaho Kawakami, Kazuyuki Nakakita, Hirotaka Sakagami, Yusuke Asai "Image Measurement of Surface Pressure Field in Unsteady Flow Using Anodized Aluminum Pressure Sensitive Coating" Proceedings of the Japan Society of Mechanical Engineers B Volume, Vo1.71, No.710 (2005) pp.2486-2493 M. Kameda, T. Tabei, K. Nakakita, H. Sakaue and K. Asai, "Image measurements of unsteady pressure fluctuation by a pressure sensitive coating on porous anodized aluminum," Meas. Sci, Technol. 16 (2005) pp. 2517-2524 James W. Gregory, Hirotaka Sakaue and John P. Sullivan, "Fluidic Oscillator as a Dynamic Calibration Too1," AIAA2002-2701, 2002. James W. Gregory, John P. Sullivan, Sameh S. Wanis and Narayanan M. Komerath, "Pressure-sensitive paintas a distributed optical microphone array," J. Acoustical Society of America, Vol.119, No.1, January 2006, pp.251-261.

  The present invention provides a method for image measurement of a distribution of fluctuating pressures in a field that is a periodic unsteady pressure fluctuation field, the fluctuation component of which has a component over a wide band. In a steady fluctuation field, it is an object to display an image as mapping data for each frequency with respect to amplitude, power, energy, phase, and phase velocity obtained by performing statistical processing while paying attention to only the fluctuation component. Furthermore, by statistically processing a large number of images, it is possible to perform measurement with higher pressure resolution than when handling as individual images, that is, even in a minute pressure field with small pressure fluctuations. It is also an issue to present a method that can measure the distribution of the fluctuating pressure component.

The method for measuring an image of an unsteady pressure field using a pressure-sensitive paint according to the present invention is an image measurement using a pressure-sensitive paint or a pressure-sensitive coating, and is based on light emission data obtained from the pressure-sensitive paint or pressure-sensitive coating. Thus, the distribution of the pressure fluctuation component in the frequency dimension in the emission data is imaged and displayed.
Specifically, a plurality of time-series images, which are light emission data from a pressure-sensitive paint or pressure-sensitive coating obtained by an image acquisition device, are converted into a series of three-dimensional data consisting of a two-dimensional axis and a time axis on the image. The state quantity related to the unsteady fluctuating pressure is obtained by frequency analysis or time frequency analysis.
In addition, as the frequency analysis or time frequency analysis applied to the time series image group, either FFT, DFT, wavelet, or short-time Fourier transform is adopted, and the amplitude distribution and power corresponding to each frequency component are determined from the result. Any one of distribution, amplitude spectral density distribution, power spectral density, phase difference distribution from the reference point, and correlation region distribution of the same component is obtained.
In addition, as frequency analysis or time frequency analysis to be applied to time series images, either FFT, DFT, wavelet, or short-time Fourier transform is used, and averaging is performed by integrating the data processing results over multiple times. Reduced noise by improving the minimum fluctuating pressure level that can be measured.

In addition, the method for measuring an image of an unsteady pressure field using the pressure-sensitive paint according to the present invention includes an amplitude distribution corresponding to a frequency component, a power distribution, an amplitude spectral density distribution, a power spectral density, a phase difference distribution from a reference point, Correlation region distribution measurement results are displayed for each frequency, for each frequency band with an arbitrary frequency width, or for various frequency band display methods such as 1 octave, 1/3 octave, 1/8 octave, 1/12 octave, etc. The intensity distribution such as the power spectrum, the amplitude spectrum, and the energy spectrum, and the state quantities such as the phase distribution and the phase velocity distribution are displayed as an image in one frame of each image or video.
The pressure-sensitive paint or pressure-sensitive coating used in the method for measuring an unsteady pressure field image with the pressure-sensitive paint of the present invention is any of a porphyrin, transition metal complex, polycyclic aromatic compound, rare earth complex, or phthalocyanine. These pressure-sensitive dyes were included.

The method of measuring an image of an unsteady pressure field using a pressure-sensitive paint according to the present invention applies frequency analysis or time-frequency analysis to time-series pressure-sensitive paint measurement data obtained by measuring a periodic unsteady pressure fluctuation field. Therefore, it is possible to obtain the distribution of the amplitude, power, phase, etc. of the fluctuating pressure component for each frequency band and to measure the image thereof. This also makes it possible to easily grasp the spatial distribution on the object in the frequency band corresponding to the frequency distribution of the unsteady pressure fluctuation component on the measurement target object as an image. The unsteady pressure fluctuation field is not limited to a field having a single frequency component, but can be used for any field such as a broadband component or a field having a plurality of frequency components.
In addition, the image measurement method of the unsteady pressure field using the pressure-sensitive paint of the present invention cannot apply the noise reduction method based on the data integration to the measurement result of the unsteady field if it is time-series data. Since it is possible to perform data integration on the spectrum of multiple images, a large number of measurement images are used, and these are divided into a plurality of image number units that can achieve the required frequency resolution. By performing integration, it is possible to improve the S / N ratio between the noise component and the signal component, and to improve the measurable level up to a small pressure fluctuation. Taking advantage of this point, the distribution of the fluctuating pressure component can be measured even in a minute pressure fluctuation field in a low-speed flow.

The method for measuring an image of an unsteady pressure field using a pressure-sensitive paint of the present invention is to apply or coat a pressure-sensitive paint on the surface of an object. The pressure-sensitive paint used is a transition such as a porphyrin or ruthenium complex as a pressure-sensitive dye. Metal complexes, or polycyclic aromatic compounds such as pyrene-based molecules, rare earth complexes, or phthalocyanine-based molecules with oxygen sensitivity are used, which are based on anodized films based on aluminum or silica gel TLC (silica gel fixed on adhesive tape) plate, porous material such as silica gel, aluminum oxide particles and titanium dioxide particles, Poly (TMSP) (polytrimethylsilylpropyne), Poly (IBM) and poly (IBM -co-TFEM) and various other organic polymers, or polymers and the above A combination of porous particles, and to constitute the form of a mixed or adsorbed or attached to such.
In addition, this pressure-sensitive paint is a 2-color or 3-color containing TSP for measuring temperature, or a component equivalent thereto, a reference dye that is insensitive to or small in both temperature and pressure, or both. It is also possible to take the form of

  Excitation light sources for obtaining light from pressure-sensitive paints include xenon light sources, discharge lamps such as halogen lamps and mercury lamps, various lasers such as LEDs, Ar lasers, YAG lasers and semiconductor lasers, and organic or inorganic light sources. EL (Electro-Luminescence) or the like can be used. All or a part of these light sources are used in combination with an optical filter for transmitting a specific wavelength and blocking other wavelengths.

  In order to measure light emission data from the pressure-sensitive paint, an image acquisition device such as a CCD camera or a CMOS camera, or an image acquisition device attached with an image amplification device such as an image intensifier is used. As an image acquisition device, it is necessary to continuously acquire a plurality of images. As the number of images increases, the accuracy of pressure measurement improves. All or some of these image acquisition devices are used in combination with an optical filter that transmits a specific wavelength and blocks other wavelengths.

If the pressure-sensitive paint is 2-color or 3-color that includes not only pressure-sensitive components but also other temperature-sensitive components and reference components, use multiple image measuring devices or use 3-CCD A multi-chip device such as the above, a color image measuring device, or a filter wheel that switches a filter to be applied can be used.
Further, in the present invention, instead of the image acquisition device, a sufficiently dense spatial resolution is generated by irradiating the pressure-sensitive paint application portion on the object surface with a point light source such as a laser, and light emitted from the pressure-sensitive paint is emitted from a photodiode or The same measurement can be performed by a technique of obtaining a continuous pressure distribution by performing time-resolved measurement using a photomultiplier tube (PMT) or the like and scanning the object surface on the point light source side.

One of the data processing methods is a step of constructing a reference image I _ref by using a plurality of images acquired by an image acquisition device, averaging the whole or a part of the images and performing noise reduction. A step of converting the pressure into a pressure using the expression (1) or an extended expression which is a general form thereof together with a wind-on image I which is a set of averages or moving averages of individual or partial images in the image is performed.
Different data processing methods include using individual images corresponding to wind-on images as reference images, 2-color (PSP component + TSP component or PSP component + reference component) and 3-color (PSP component + TSP component + reference component) In the case of using a multi-component pressure-sensitive paint such as the above, the reference image is an average value of individual, part or whole of a plurality of images in which the reference component is measured instead of the average image of the PSP component. Take the form. The reference image may be constructed using a pressure-sensitive paint image or a group of images under a known pressure that is obtained separately, as in a general pressure-sensitive paint data processing method.

In some forms, alignment to correct the movement of the object during these processes, reflection correction to correct the reflection of the luminescent component on the object surface (Self-Illumination correction), and pasting of the data to a three-dimensional grid In addition, various processes such as smoothing by a spatial filter, pixel combination, temperature correction, and excitation light correction are included.
As a method for calculating the pressure, the formula (1) or its extended form to a higher order formula is used. In these equations, the coefficients A and B in equation (1) or the various coefficients in the expanded expression in equation (1) can be calculated using the a-priori method, in-situ method, or processing with temperature correction applied to them. The method, the a-priori / in-situ hybrid method disclosed in Patent Document 2, or the temperature correction in-situ method disclosed in Patent Document 1 are used.

  Similar to the result shown in Non-Patent Document 1, the result obtained by the processing so far is merely a result of converting a plurality of measurement images into pressure and arranging them in time series. In the next stage of processing, individual pixels, or some pixel groups, or the whole of multiple time-series images are captured as a series of three-dimensional data spanning the time axis and two axes on the image, and these are used as frequencies. Adding analysis and processing time-series data into frequency-axis data (this frequency analysis includes Fourier analysis techniques such as FFT and DFT. This frequency analysis includes wavelet and short-time Fourier transform) Similar effects can be obtained with time-frequency analysis.) This frequency analysis or time-frequency analysis is repeated over the entire image or part of the image, and the amplitude, power, energy, or phase with respect to the frequency axis or time-frequency axis is repeated. To obtain data in the form of (such as these frequency or time frequency analysis techniques Not only applied to one-dimensional data obtained by taking out averaged data over one or more pixels and arranging them in time series, but also two-dimensional or three-dimensional data over the whole or a part of the image. 2D or 3D frequency analysis or time frequency analysis may be applied.

  During this process, data such as pressure sensors, microphones, hot wires, hot films, etc. other than pressure-sensitive paint data are introduced, and these are used as reference pressure or phase references for reference phenomena, or are mutually evaluated with pressure-sensitive paint data. There are also forms in which correlation processing such as cross-correlation is performed. In statistical processing of pressure-sensitive paint data, a window function is required, or the entire time-series image group is divided into small parts, and statistical processing is applied to each small part and then integration is performed on the spectral axis. Processing may be performed. In order to unify the phase information for a plurality of small portion data, it is necessary to unify the phase information by using the phase of the representative point and the data of other sensors for reference.

  Information such as amplitude, power, energy, phase, and phase velocity obtained by these processes is organized in the frequency domain, and is organized and used as a measurement state quantity on an image or a three-dimensional grid corresponding to a measurement target object. As a basic expression method of these frequency regions, an array having a frequency value, a spatial coordinate value, and a value such as amplitude, power, or phase information is obtained. As a display method of processing result data, it is possible to display values such as amplitude, power, and phase for each pixel for each frequency value, but as a display method suitable for visualization, a frequency with a certain width A method for integrating and displaying data of frequency values included in each band is effective. As this frequency band, a constant value such as every 10 Hz or every 20 Hz, or various frequency band display methods such as 1 octave, 1/3 octave, 1/8 octave, 1/12 octave, or the like can be used. These data are stored, displayed, and used as digital data, images, image groups, or various video data.

  As a result of the frequency analysis or time frequency analysis so far, there is a distribution in the excitation light intensity, so the shot noise component is relatively small in the area where the excitation light intensity is high, and conversely, the shot noise is large in the area where it is small. If the data processing result is displayed as an image for each frequency band as it is, the shot noise component increases in inverse proportion to the 1/2 power of the excitation effect intensity, and it appears as if there is a distribution in the magnitude of the fluctuating pressure. In order to solve this problem, not only a series of multiple images in the wind-on state, but also a series of measurement images of pressure-sensitive paint obtained by stopping the wind tunnel immediately after the wind-on image acquisition, The frequency analysis or the time frequency analysis result in which only noise exists is also created. It is also possible to subtract the result of no wind from the result of wind-on for each data processing unit such as a pixel to correct the variation in offset component due to shot noise for each part. By using this processing, offset components due to shot noise resulting from the distribution of excitation light can be removed at a considerable rate, and improvement in quantitativeness of processing result data and measurable fluctuating pressure can be expected.

  Also, in frequency analysis or time-frequency analysis using this method, noise reduction processing such as averaging of acquired data cannot be applied to non-stationary data containing general broadband frequency components, but frequency analysis, or By applying the time-frequency analysis, the amplitude, power, etc. can be integrated on the spectrum of the frequency axis, so that there is a merit that noise reduction processing by averaging can be performed. Divide a large number of acquired images into multiple image number units that can achieve the frequency resolution necessary to understand the phenomenon, and integrate the frequency analysis or time-frequency analysis results for each unit on the frequency axis spectrum. In principle, it is possible to reduce noise components at a rate of 1 / n ^ 0.5 (n is the number of units to be integrated). The noise component referred to here is not a noise component such as white noise on the time axis, but a fluctuation of data such as amplitude and power on the frequency spectrum. When integrating the processing results of multiple image count units, the level of shot noise remains constant and cannot be reduced by integration, but variations in the amplitude and power of the shot noise component on the frequency axis can be reduced. As a result, the separability from the component due to the phenomenon is improved, and it is possible to obtain the merit that the measurable variable pressure limit is improved. By using this, the same unsteady field is measured over many images, and integration on the frequency axis is performed, so that the pressure is much smaller than the method of converting individual images in time series into pressure. Measurement up to level fluctuation components is possible.

  Next, the preferable form of this invention is shown. A high-speed response type pressure-sensitive paint that can generate a sufficient amount of light and a response that is sufficiently faster than the frequency of the unsteady phenomenon that is to be measured, and a sufficient number of thousands in a continuous time series with sufficient spatial resolution. A high-speed camera capable of acquiring tens to tens of thousands of images and an excitation light source having a sufficiently large excitation light amount including the absorption line wavelength of the pressure-sensitive dye contained in the high-speed response type pressure-sensitive paint to be used. At this time, if the total number of images of the high-speed camera is 2 ^ n, since the FFT method can be used instead of the DFT method, high-speed processing can be performed in the subsequent data processing portion. The more the number of measurement images, the more the excitation light source, in the case of a laser, the high coherent light source causes speckles on the surface of the object to be measured, causing noise and the like. It is effective to use a light source that does not cause light emission.

  A reference image is obtained by averaging the entire image group measured using these devices. The reference image may be an image of a static and known pressure field acquired separately. Using this reference image and individual images (wind-on images), the expression (1) or its extended expression, the a-priori method, the in-situ method, or a processing method with temperature correction added thereto, or patent literature The pressure is converted using a pressure-sensitive paint calibration method such as the a-priori / in-situ hybrid method shown in FIG. 1 or the temperature-corrected in-situ method shown in Patent Document 2. At this time, both the reference image and the wind-on image are subjected to processing such as spatial filter processing, alignment, binning, reflection correction, mapping to a three-dimensional grid, lens distortion and chromatic aberration correction, and camera image flat field correction. .

  Next, statistical processing is performed on the pressure distribution data reflecting the time series and spatial distribution. For periodic unsteady fields, the most common method is to use FFT or DFT. If it is not possible to assume periodic unsteady conditions, it is necessary to use wavelet analysis. In the method using FFT or DFT, processing as time series data is performed in pixel units. At this time, a window function is applied so that the influence of the end of the statistical processing section is not mixed. Also, if the data has a sufficient data length in the time direction, dividing the whole into small sections that can obtain sufficient frequency resolution, and performing averaging processing using these will have a smoothing effect. can get. However, when performing such processing, the phase is not stored for each subsection, so when trying to obtain phase and phase velocity information, the phase of the representative point in the image or a separately measured reference It is necessary to perform a process for aligning the phases by a method such as using a signal from the sensor for reference as a reference.

  Power, energy, and phase velocity can be calculated based on amplitude information and phase information obtained by the FFT or DFT processing. This result is displayed for each frequency or frequency band, or using various frequency band display methods such as 1 octave, 1/3 octave, 1/8 octave, 1/12 octave, etc. An intensity distribution such as energy and state quantities such as phase distribution and phase velocity distribution are displayed as an image using a display method such as color, gray scale, and contour lines.

FIG. 1 shows a photograph of an unsteady pressure field generator using an unsteady standing wave of a resonance tube, and FIG. 2 shows the entire system. A large output speaker 2 is installed at one end of an acrylic cylinder 1 and a pressure sensitive paint sample 3 is installed at the other end. When the speaker 2 is set at a constant frequency, an antinode of a standing wave exists at the end where the pressure-sensitive paint is set, and an unsteady fluctuation pressure can be measured. The resonance tube was set at 108 Hz and tested. The pp value of the unsteady pressure fluctuation is about 500 Pa.
As the pressure-sensitive paint, a fast-response pressure-sensitive paint based on anodized aluminum and adsorbing bathophenanthroline / ruthenium (Ru (dpp)) was used. As an excitation light source, an excitation filter that transmits only blue light was used as a xenon light source. In FIG. 1, the black cylindrical objects visible on both sides of the pressure-sensitive paint sample are irradiation heads that irradiate the sample with xenon light transmitted through a light guide. As the image acquisition device, a CMOS type high-speed camera is used. To increase the number of images in the time direction, the basic spatial resolution, the number of images is 512 x 512 pixels, 4096 is 128 x 128 pixels, and the number of images is about Increased to 33,000. Of these, 2 ^ 15 images were used to process the FFT. The measurement frame rate of the high-speed camera is 1000 fps.

  In order to increase the data count value on this image, a binning operation is performed with 2 × 2 pixels as a new pixel, and this pixel is converted into 2 ^ 15 one-dimensional data in the time direction, and FFT processing is performed. As the frequency resolution of measurement, about 1/1000 of the frame rate of a high-speed camera is sufficient for practical use. Instead of applying FFT over the whole, an FFT of 2 ^ 10 = 1024 frames was performed. At that time, the area to be FFTed was shifted every 512 sheets, and averaging was performed on a total of 63 FFT results. Before FFT processing, a window function by Hanning is applied to remove the influence of both ends. Figure 3 shows the result of repeating the processing for one pixel over the entire screen and integrating the results of integration in the band of the center frequency ± 20 Hz for each frequency, and plotting the frequency direction data at one point on the measurement model. FIG. 4 (a). From FIG. 4A, it can be seen that there is a narrow-band unsteady component corresponding to the input frequency to the resonance tube in the vicinity of 108 Hz. In the image of FIG. 3, the amplitude of the FFT result is integrated with a width of ± 20 Hz around the center frequency and displayed in gray scale. It can be seen that only the frequency band of 100-140Hz has a fluctuation component. There is no component corresponding to the phenomenon in other frequency bands other than the 100-140 Hz band. In the image at 100-140 Hz, the signal intensity on the pressure-sensitive paint sample is almost uniform, and the result is consistent with the physical knowledge that the pressure fluctuation due to the resonance tube is uniform across the cross section in the tube.

  3 and 4 (a) show not only a series of multiple images in the wind-on state, but also stop the wind tunnel immediately after acquiring the wind-on image and acquire a series of wind-off images to obtain no wind. This is a result in which only the shot noise is generated, and the wind-off data at the time of no wind is subtracted from the wind-on data for each pixel to correct the variation of the offset component due to the shot noise. 4 (b) and 4 (c) are wind-on and wind-off data of the same part as in FIG. 4 (a), and the data obtained by subtracting (c) from FIG. 4 (b) is shown in FIG. (a). Since there is a distribution in the amount of excitation light, the offset amounts of FIGS. 4B and 4C differ depending on the magnitude of the amount of excitation light. However, by introducing the subtraction of the wind-off data, the offset amount can be made zero for any part, which contributes to quantitative evaluation of the pressure fluctuation due to the phenomenon.

5 and 6, a jet chopper device 7 comprising a jetting jet 6 and a rotating plate connected to a motor for intermittently turning it ON / OFF is used, and the periodic non-rotation generated thereby is used. The steady pressure field was measured. The rotating flat plate 7a has six openings. Each time the flat plate 7a makes one rotation, the jet can be turned on and off six times. The apparatus was set up to turn the jet on and off at approximately 100 Hz and tested.
As the pressure-sensitive paint, a fast-response pressure-sensitive paint in which bathophenanthroline / ruthenium (Ru (dpp)) was adsorbed on anodized aluminum was used. In addition, a high-speed response type pressure-sensitive paint in which bathophenanthroline / ruthenium was adsorbed on a TLC plate was also used. Here, the result in the case of using anodized aluminum is shown.
Two types of excitation light sources were used: a xenon light source using an excitation filter that transmits only blue light, and an all-line output mode Ar laser. Here, the result in the case of using a xenon light source is shown.

  A CMOS-type high-speed camera is used as the image acquisition device. To increase the number of images in the time direction, the basic spatial resolution, the number of images is 512 x 512 pixels, 4096 is 208 x 304 pixels, and the number of images is 16500. Increased to one. Of these, 2 ^ 14 images were used to process the FFT. The measurement frame rate of the high-speed camera is 1000 fps. In order to increase the data count value on this image, a binning operation is performed with 2 × 2 pixels as a new pixel, and this pixel is converted into 2 ^ 14 one-dimensional data in the time direction, and FFT processing is performed. At this time, if the frequency resolution of measurement is about 1/1000 of the frame rate of a high-speed camera, it is sufficient for practical use. Instead of applying FFT over the entire 2 ^ 14 images, 2 ^ 10 = 1024 images. The FFT area was shifted every 512 sheets, and averaging was performed on 31 FFT results. Before FFT processing, a window function by Hanning is applied to remove the influence of both ends. FIG. 7 shows the result of repeating the processing for one pixel over the entire screen and integrating the result of integration in the band of the center frequency ± 10 Hz for each frequency, and the wind tunnel is stopped from the wind-on data as in [Example 1]. Subsequent wind-off data is subtracted for each pixel, and offset component variations due to shot noise are corrected. FIG. 8 is a plot of frequency direction data at one point on the measurement model. It can be seen from FIG. 8 that a narrow-band signal corresponding to ON / OFF of the unsteady pressure field by the jet chopper exists near 100 Hz. In the image of FIG. 7, the amplitude of the FFT result is displayed in gray scale. It can be seen that the area where the white circular jet collided was measured only in the 90-110 Hz frequency band. In other frequency bands, there is no component corresponding to the phenomenon.

  Unsteady pressure fluctuations due to Karman vortices around a cylinder were measured using a small low turbulence wind tunnel of the JAXA Technical Research Headquarters. The small low turbulence wind tunnel has a height of 0.65m and a width of 0.55m. The two-dimensional cylindrical model described in the next section was installed here and tested. FIG. 9 shows a two-dimensional cylindrical model. It has a diameter of 0.04m and a length of 0.55m, and a high-speed response PSP is coated at the center of 8cm. Three Kulite-XCS-093-5G pressure sensors (full scale 5 psig) are installed on the PSP coating center line at 30 ° intervals for comparison with PSP measurement. The high-speed camera for PSP measurement is installed below the wind tunnel measurement unit in FIG. 2 and measures the region of q = 0 to 180 ° of the model from directly below. The shooting conditions of the high-speed camera are a measurement rate of 1 kfps and an exposure time of 990 ms. 2 ^ 14 × 5 = 81920 images were used for data processing.

The measurement results shown here are for U = 33 m / s, Re = 8.6 × 104, and the Karman vortex frequency is 165 Hz when the Strouhal number is 0.2.
FIG. 10 shows the Cp distribution around the cylinder by the pressure sensor and the RMS value of the fluctuation component. DCprms uses the standard deviation in time series data. It can be seen from the Cp distribution that the flow is separated at around 70 °. The pressure fluctuation is small in the vicinity of the stagnation point, takes a maximum value near 80 ° slightly downstream from the peeling point, and continues almost constant after that.

  FIG. 11 shows time-series unsteady pressure data measured by a pressure sensor at θ = 80 ° on a cylinder. As shown in the amplitude spectrum described later, the fluctuating pressure component due to the Karman vortex is not a single frequency but rather a wide band, and thus shows a complicated behavior with the time series data in FIG.

FIG. 12 is an unsteady pressure distribution image obtained by frequency analysis of fast response PSP data using FFT. The FFT results for each pixel are integrated into an image in the 40 Hz range. In the acoustic analysis, 1 octave or 1/3 octave analysis is used, but here it was processed at equal intervals. FIG. 13 is a comparison of the amplitude spectrum of one pixel PSP and the pressure sensor at θ = 80 °.
In FIG. 12, the non-stationary component exists only in the 140-180 Hz component near 160 Hz corresponding to the Strouhal number 0.2, and the non-stationary component is hardly seen in other frequency bands. This can be said to be a reasonable result because there is no significant component other than the peak near 160 Hz in the amplitude spectrum of the pressure sensor of FIG. Also, the distributions of the pressure sensor and the PSP amplitude spectrum of FIG. 13 are in good agreement.

  FIG. 14 shows an image displayed by integrating the FFT results in the range of 160 Hz ± 10 Hz for each pixel, and the spatial resolution is 100 pixels vertically and 160 pixels horizontally. FIG. 15 is a circumferential distribution obtained by averaging 80 pixels in the horizontal direction assuming two-dimensional flow in the center of FIG. The pressure sensor data in FIG. 15 is obtained by performing FFT processing on time series data in the same manner as the PSP result, and integrating the range of 150-170 Hz. From FIG. 14, the unsteady pressure component is distributed almost two-dimensionally, the peak of fluctuating pressure exists near θ = 80 °, the fluctuating component is almost flat on the downstream side, and the fluctuating component is small on the upstream side. I understand. FIG. 15 shows a good qualitative match between the measurement result of the unsteady pressure by the PSP and the data from the pressure sensor although there is a difference in detail.

FIG. 16 is an example of evaluation of S / N ratio improvement by superimposing frequency analysis results. The FFT analysis achieved a measurement frequency resolution of 1 Hz, so one set was set to 1024 because the measurement frame rate of the high-speed camera was 1 kHz. Actually, in order to guarantee the inefficiency of using the measurement data by the window function, the time-series image group is divided into 512 images as one set, and 1024 images are set by high-speed camera measurement while overlapping the 512 images in pairs. 31 sets are made for 1 unit 2 ^ 14 = 16384 sheets. In one unit of measurement results, 31 sets of FFT results for 1024 images are accumulated.
FIG. 16 shows the noise reduction behavior on the spectrum when 16384 units are further integrated by 1 unit, 3 units, and 5 units. The results shown so far are 5 unit results. 16 (a) to 16 (c), it can be seen that as the data integration is increased, the separability between the signal near 160 Hz corresponding to the phenomenon and the noise level is improved. This means that by using frequency analysis, if the noise level is reduced by increasing the number of data integrations, even smaller pressure fluctuations can be measured.

  In a wide range of fields such as aerospace, transportation equipment such as railways and automobiles, and fluid machinery such as air conditioners and fans, unsteady pressure fluctuation fields are flow fields that cause problems such as vibration, noise, and metal fatigue. It is available for. In these fields, the location of the unsteady pressure field and its intensity, frequency component, or phase component are measured to understand the behavior of the unsteady field existing on the surface of the object, improving machine design and efficiency, It may be used as a basic technology that can be used for various purposes such as noise and vibration reduction.

It is a photograph of the generator of the unsteady pressure field using the unsteady standing wave of a resonance tube. It is a schematic diagram which shows the whole system of the apparatus of FIG. The processing for one pixel is repeated over the entire screen, and the result of integration in a band with a center frequency of ± 20 Hz is displayed as an image for each frequency. (a) is a plot of frequency direction data at one point on the measurement model, (b) and (c) are wind-on data and wind-off data of the same part as (a), (b) − (c) is the data of (a). It is a photograph of a jet chopper device comprising a jetting jet and a rotating plate connected to a motor for intermittently turning it on and off. It is a schematic diagram of the jet chopper apparatus of FIG. The processing for one pixel is repeated over the entire screen, and the result of integration in a band with a center frequency of ± 10 Hz is displayed as an image for each frequency. This is a plot of frequency data at one point on the measurement model. It is the photograph of the two-dimensional cylinder model used for experiment. This is the RMS value of the Cp distribution and fluctuation component around the cylinder by the pressure sensor. It is time-series unsteady pressure data measured by a pressure sensor at θ = 80 ° on a cylinder. It is a distribution image of unsteady pressure obtained by frequency analysis of fast response PSP data by FFT. It is a comparison figure of the amplitude spectrum of a pressure sensor and PSP. For each pixel, the FFT results in the range of 160 Hz ± 10 Hz are integrated and displayed as an image. FIG. 14 is a circumferential distribution obtained by averaging 80 pixels in the lateral direction assuming a two-dimensional flow in the center of FIG. It is an example which shows evaluation of the S / N ratio improvement by superimposition of the frequency analysis result.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Acrylic cylinder 2 Large output speaker 3 Pressure-sensitive paint sample 4 Excitation light source 5 High-speed camera 6 Jetting jet 7 Chopper device 7a Rotating flat plate

Claims (7)

  Image measurement using pressure-sensitive paint or pressure-sensitive coating, and based on the obtained light-emission data from pressure-sensitive paint or pressure-sensitive coating, the distribution of pressure fluctuation components in the frequency dimension in the light emission data is imaged Image measurement method for unsteady pressure field using pressure-sensitive paint.   Multiple time-series images obtained by an image acquisition device that is light emission data from pressure-sensitive paint or pressure-sensitive coating are treated as a series of three-dimensional data consisting of two-dimensional axes and time axes on the image, and frequency analysis is performed. The image measurement method according to claim 1, wherein a state quantity related to unsteady fluctuating pressure is obtained by time frequency analysis.   The image measurement method according to claim 2, wherein one of FFT, DFT, wavelet, and short-time Fourier transform is adopted as frequency analysis or time frequency analysis applied to the time series image group.   Amplitude distribution, power distribution, amplitude spectral density distribution, power spectral density, phase difference distribution from reference point, homogenous component corresponding to each frequency component from the result of FFT, DFT, wavelet or short-time Fourier transform for time series images The image measurement method according to claim 3, wherein one of the correlation region distributions is obtained.   FFT, DFT, wavelet, or short-time Fourier transform for frequency analysis or time-frequency analysis applied to time-series images. Noise can be measured by averaging by integrating the data processing results over multiple times. The image measurement method according to claim 3, wherein the minimum fluctuating pressure level is improved.   Measurement results are displayed for each frequency, for each frequency band with an arbitrary frequency width, or for various frequency band display methods such as 1 octave, 1/3 octave, 1/8 octave, 1/12 octave, etc. The image measurement method according to claim 4, wherein an intensity distribution such as a power spectrum, an amplitude spectrum, and an energy spectrum and a state quantity such as a phase distribution and a phase velocity distribution are displayed as an image in one frame of an image or video.   The pressure-sensitive paint or pressure-sensitive coating contains any one of a porphyrin-based, transition metal complex, polycyclic aromatic compound, rare-earth complex, and phthalocyanine-based pressure-sensitive dye. The image measurement method described in 1.