• <tr id="yyy80"></tr>
  • <sup id="yyy80"></sup>
  • <tfoot id="yyy80"><noscript id="yyy80"></noscript></tfoot>
  • 99热精品在线国产_美女午夜性视频免费_国产精品国产高清国产av_av欧美777_自拍偷自拍亚洲精品老妇_亚洲熟女精品中文字幕_www日本黄色视频网_国产精品野战在线观看 ?

    An algorithm to separate wind tunnel background noise from turbulent boundary layer excitation

    2019-10-26 01:11:32XiojinZHAOMingsuiYANGJieZHOUJunminLEI
    CHINESE JOURNAL OF AERONAUTICS 2019年9期

    Xiojin ZHAO, Mingsui YANG, Jie ZHOU, Junmin LEI

    a School of Aerospace Engineering, Beijing Institute of Technology, Beijing 100081, China

    b AECC Shengyang Engine Research Institute, Shenyang 110015, China

    c Department of Mechanical and Aerospace Engineering, Hong Kong University of Science and Technology, Hong Kong, China

    KEYWORDS Background noise;Sound source separation;Turbulent boundary layer;Wall pressure fluctuation;Wind tunnel

    Abstract Wall pressure fluctuations generated by Turbulent Boundary Layers (TBL) provide a significant contribution in reducing the structural vibration and the aircraft cabin noise. However,it is difficult to evaluate these fluctuations accurately through a wind tunnel test because of the pollution caused by the background noise generated by the jet or the valve of the wind tunnel. In this study, a new technology named Subsection Approaching Method (SAM) is proposed to separate the wall pressure fluctuations from the background noise induced by the jet or the valve for a transonic wind tunnel test. The SAM demonstrates good performance on separating the background noise from the total pressure compared to the other method in this study.The investigation considers the effects of the sound intensity and the decay factor on the sound-source separation. The results show that the SAM can derive wall pressure fluctuations effectively even when the level of background noise is considerably higher than the level of the wall pressure fluctuations caused by the TBL. In addition, the computational precision is also analyzed based on the broad band noise tested in the wind tunnel. Two methods to improve the precision of the computation with the SAM are also suggested:decreasing the loop gain and increasing the sensors for the signal analysis.

    1. Introduction

    The continual development of jet-engine-powered, wellstreamlined aircraft draws extensive attention to the sound and the vibration induced by the turbulent boundary layer of high-speed airflow. The turbulent boundary layer is responsible for reducing the sound level inside the cabin during the cruise condition,1which is an important index of aviation requirements. To develop effective ways to reduce this sound level, the noise associated with the turbulent boundary layer needs to be reliably predicted.

    Reliable wall pressure fluctuations and noise generated by the turbulent boundary layer are usually obtained in wind tunnels.2-6The measured wall pressure fluctuations could be used to validate the CFD simulations7-12and correct wavenumberfrequency models.13-16However, wind tunnel tests of wall pressure fluctuations are not always perfect, and the results are usually contaminated by the background noise in the wind tunnel.

    To obtain the signal of interest, the background noise should be separated and deleted from the measured results.

    Generally, in a low-speed wind tunnel with the maximum speed not higher than 100 m/s (Mach number Ma ≈0.3),17,18the primary noise source is the fans used to accelerate the airflow. However, in a transonic or supersonic wind tunnel, the background noise is mainly composed of the jet and the valve noise,19-21and the noise generated by slots and the perforated walls of the test chamber.22-24

    Some investigators17,18,25-28tried to reduce the effect of background noise on wall pressure fluctuations and aeroacoustic measurement. However, these studies focused more on the measurement of radiated noise and most required a higher SNR (Signal to Noise Ratio) for the signal of interest to remove the background noise with the acoustic array technology.17Meanwhile, the flow speeds involved in these studies17,18,25-28did not exceed 100 m/s (Ma ≈0.3) and the suppression of background noise was mostly achieved with advanced signal processing methods.18,26-28

    Ref.23showed that strong background noise still exists in many transonic and supersonic wind tunnels and the pressure fluctuations measured in these wind tunnels could be seriously contaminated by the background noise caused by the jet or the perforated wall.To control the background noise generated by the perforated wall of the test chamber, the perforated wall could be covered with wired screens.23To reduce the effect of the valve noise on the aero-acoustic measurement,a special absorptive structure29could be installed around the test chamber of the wind tunnel. However, the experimental cost would certainly be increased. In addition, some optimization techniques have been proposed to reduce the effect of the background noise through phase array signal processing;however, the algorithm used for the phase-array test cannot directly separate the background noise from the wall pressure fluctuations owing to the special requirement of the array distribution (for example, spiral design) in the phase-array technique.

    Hence, this study attempts to develop an algorithm to subtract the component of background noise caused by the jet and the valve in a transonic wind tunnel without using any noise absorptive structure.This technique could be applied to higher flow speeds beyond the range of the applications of traditional phase-array methods. This investigation is inspired by the work of Goldstein30and Sinayoko et al.31To separate these two kinds of pressure fluctuations, Goldstein30suggested in an unbounded turbulent field, the aerodynamic noise be decomposed into non-radiating and radiating components.Then, Sinayoko et al.31separated the radiating acoustic components from the non-radiating components in the nonlinear flow equations using linear convolution filters. The wall pressure fluctuations can be treated as the non-radiating acoustic component and the jet and valve noise belong to the radiating acoustic component according to their forms of production.30,31The method developed here aims at separating the jet and valve noise from the pressure fluctuations in a transonic flow when strong pressure fluctuations are usually induced32over the structure.

    The remainder of the paper is organized as follows: Section 2 describes the theoretical model of different sound sources. Result comparisons are presented in Section 3. The final section, Section 4, provides conclusions.

    2. Theoretical model

    2.1. Sound source model

    To separate the background noise,the first step is to determine the mathematical expressions for the background noise and the wall pressure fluctuations generated by the turbulent boundary layer.Fig.1 shows a test model fixed in the wind tunnel in the wall pressure fluctuations test. According to the equation of the convection wave, the background noise caused by the jet exit vanes downstream or the valve used for controlling the airflow upstream can be expressed as follows:

    where pbω( ) indicates the amplitude of the sound pressure around the test model belonging to the background noise, ω the angular frequency, c0the sound speed in the wind tunnel,Usthe velocity of the local flow and ξxis the variation of the distance along the stream-wise direction. Eq. (1) presents the background noise generated by valves or jets, propagating to the test section of the wind tunnel.It is assumed that the amplitude of the background noise does not decay along the flow direction or around the test model because the coordinates of the sensors installed on the test model change only slightly.

    Generally, the wall pressure fluctuations under a TBL can be simulated by wavenumber-frequency spectrum models,such as the Corcos model,32-34Efimtsov model,35,36and Chase model.37These different TBL models are adapted to different situations mostly determined by the inflow speed and the configuration of the test model.The classical Corcos model is only suitable for describing TBLs with low mean flow velocities.However, the extended Corcos model32can cover low-speed,subsonic, and transonic flows passing through a typical covecylinder structure. Hence, to separate the background noise at a transonic flow, which regularly induces serious pressure fluctuations over the structure, the extended Corcos model32is adopted in this study to simulate wall pressure fluctuations.

    Fig. 1 Test model measured in a wind tunnel.

    In the frequency domain, the wall pressure fluctuations described by the classical Corcos model can be expressed as:33,34where Ucindicates the convection velocity, and it is suggested that 0.5U0≤Uc≤0.7U0,38U0is the mean flow velocity;x and y represent the coordinates along the stream-wise and spanwise directions in the model, respectively; ξxand ξyindicate the distances between the two arbitrary points on a plate in stream-wise and span-wise directions, respectively; αxand αyare the decay rates in x and y directions,respectively;Sppindicates the amplitude of the wall pressure fluctuations.

    Generally, Eq. (2) is only suitable for describing low-speed flows when pressure gradients are too small to be considered within the TBL. If the incoming flow is in high subsonic or transonic speeds,the extended Corcos model32should be used to simulate the attached flow, and the expression of the extended Corcos model is:

    In addition, if the extended Corcos model is used to simulate the separated flow,32Eq. (3) should be modified as:

    To describe the shock wave oscillation,32the Corcos model can be expressed in the form of:

    Eqs. (3)-(5) show that TBL models always consist of two components, the decay of the amplitude of wall pressure fluctuations and the phase change of the TBL excitations at different points. The expressions of pressure fluctuations under different flow conditions, including the attached flow, the separated flow and the shock wave oscillation can be integrated into a uniform form:

    Eq. (6) indicates that the cross-spectral density of the wall pressure fluctuations at two different points on the test model could be expressed in terms of the correlation between the pressures at two arbitrary points:

    where pTBLX,ω( ) indicates the variation of the wall pressure fluctuations induced by the TBL over the structure. Although Eq. (7) indicates the correlation in power and Eq. (6) represents the correlation in pressure,both equations present a consistent phase variation as Eq. (7), which can be derived from Eq. (6).

    Eqs. (1) and (7) require that the propagation speed of the background noise in the wind tunnel must be different from the speed of the flow convection.Otherwise,serious coherence between the background noise and the wall pressure fluctuations will be generated and the change of phase differences between the two disturbances is eliminated. In this case, the computed phase difference between any two continuous sensors remains constant and it will be impossible to separate the background noise from wall pressure fluctuations by Subsection Approaching Method (SAM). The condition of‘‘c0+Us≠Uc” is the assumption required by the technology used for sound-source separation in this study. However,‘‘c0+Us≠Uc” is always satisfied in this study because c0>Ucall the time for the transonic flow. In addition,although the proposed method aims at solving the background noise of the transonic flow, its application is not confined to transonic flows. If the condition ‘‘c0+Us≠Uc” is satisfied,sound source separation could be extended to supersonic flows or even hypersonic flows theoretically.

    2.2. Sound source separation

    Eqs.(1)-(7)show that the background noise and the wall pressure fluctuations generated by the TBL are not coherent when c0+Us≠Uc. Therefore, it is possible to subtract the background noise directly from the compound signal measured in the wind tunnel test.Based on the difference between the propagation velocity of the background noise and the convection velocity of the flow,the SAM is proposed to separate the background noise from wall pressure fluctuations in this study.The background noise propagates at the speed of sound, while the pressure fluctuations induced by the TBL are determined by the convective speed of the inflow.

    Wall pressure fluctuations induced by the TBL are usually measured by a sensor array. Fig. 2 shows the sensor array designed for the SAM,in which sensors are arranged in a line.Assume that the pressure signals are collected by the sensors of the array, and they could be expressed as:

    where piindicates the pressure-signal collected by the ith sensor,and N is the total number of the sensors used in the array.

    At each point,the sound pressure picontains both the pressure induced by the TBL and the background noise.According to Eqs.(1),(7)and(8),the pressure measured at each test point can be represented as:

    Fig. 2 Sketch of sensor array.

    Then the prince s servant brought in the three robes, and when the judges saw the mist-gray one which the princess usually wore, they said, Have this robe embroidered10 with gold and silver, and then it will be your wedding robe.

    The SAM is a method based on numerical calculation,and basically, the coherence of wall pressure fluctuations at different test points is used to separate the background noise.According to Eq. (7), though the amplitude of the wall pressure fluctuation at each test point is different,the phase differences between the wall pressure fluctuations at any two successive test points are the same if the sensors are distributed with equidistance. This is the fundamental principle used by the SAM for sound source separation.

    The computational flowchart of the SAM is shown in Fig. 3. In using the SAM to separate the noise, the first step is to define the initial variations used to store all the pressure data measured by the sensors and some important parameters required in the computation. The definitions of all the initial variations are listed as follows:

    Fig. 3 Subsection approaching method with n subsections.

    where ‖ ‖ indicates the modulo, temp0determines the ranges of the amplitudes of the background noise and the wall pressure fluctuations, anddefines the range of the propagating speed of the background noise in the wind tunnel relative to the sound speed.

    To avoid the divergence of computation, a loop gain φ is introduced to control the step of computation and‘‘0<φ<1” is suggested to buffer the computation. When n subsections are used for the computation, as presented in Fig.3,the loop gain could choose‘‘φ=1/n”.The step lengths of the amplitude and the velocity can be expressed as‘‘Δp=φ·temp0” and ‘‘ΔUs=φ·U0”, respectively. Then we address the first three sensors:

    Through the first step of the computation, the computed background noise and the wall pressure fluctuations are closer to the real values than the initial values defined in Eq.(13). In the second step of computation, the step lengths of the amplitude and the velocity are updated with new variations to improve the precision of the simulated background noise:

    In addition, increase one more sensor for the computation of the phase differences between two neighboring sensors:

    Define Max=Maximum(abs (ΔPha2,3-ΔPha1,2), abs(ΔPha3,4-ΔPha2,3), ... , abs(ΔPhaN-1,N-ΔPhaN-2,N-1)),ΔPhai,j=Phase()-Phase(),i,j=1,2,···,N, and Max indicates the maximum value when i and j are changed from 1 to n.

    Meanwhile, change k and l to find ‘‘Minimum (Max)”.When ‘‘Minimum (Max)” is found, update l and k with lminand kmin, respectively, and stop the computation. Then, the final background noise and the wall pressure fluctuations are the closest to the real values:

    where pb=lminΔp+amp, Us=kminΔUs+Pha, and Δp, ΔUsare computed by Eq. (17).

    According to the presentation above,the SAM is an asymptotic method based on the numerical computation. An overt merit of the SAM developed is that it could resolve wall pressure fluctuations without knowing the precise information of them,such as their intensity,the decay factor,and the convection velocity, thus it is easier to apply in engineering applications. In addition, although the SAM is derived based on the signal collected by an equi-spaced array, the instrument is not confined to equi-spaced array systems.The reason of using an equi-spaced array here is that it is also used to obtain the pressure distribution along the streamwise direction of the model.

    2.3. Precision analysis

    The precision of the computation is a vital factor affecting the separation of the wall pressure fluctuations from the background noise.According to the SAM introduced in Section 2.2,the precision of the signal separation is determined by the step lengths of the amplitude and the velocity used in the computation.The expressions of the step length are shown in Eq. (17),which shows that the step lengths of the amplitude and the velocity are both determined by the loop gain φ(0<φ<1)and the number of sensors used to record the raw signals.The larger the step length, the lower the computational precision. Hence, the precision of analyses could be improved in two ways: decreasing the loop gain, or increasing the number of sensors used to collect the raw signals. However, it is uneasy to determine the value N and the loop gain to obtain a best result of sound separation.The more the number of sensors and the smaller the loop gain, the higher the accuracy of the acoustic separation and the longer the iteration time.Sometimes, for small loop gains, the iterative computation of the sound separation is not even convergent. In most cases,the appropriate values N and the loop gain should be determined by the calculations based on the frequency range of interest and the required accuracy.

    3. Comparisons of results

    Generally,the separation of signals is influenced by many factors, such as the intensities of the wall pressure fluctuations and the background noise, the decay factors of the wall pressure fluctuations, and the analysis frequencies. In this section,these factors are discussed and the proposed method of background noise separation is validated.

    3.1. Effects of intensities of wall pressure fluctuations and background noise

    Fig. 4 demonstrates that the SAM can accurately separate the wall pressure fluctuations from the background noise when they are at the same level. Unfortunately, the wall pressurefluctuations are sometimes much lower than the background noise. In such cases, the wall pressure fluctuations are almost submerged in the background noise (Fig. 5), and the difficulty in separating this signal from the background noise is considerably increased.Fig.5 presents the results of the sound source separation when the background noise is much larger (ten times of the wall pressure fluctuations) than the wall pressure fluctuations, and pbω( )=10 Pa, pTBLω( )=1 Pa. The results show that the SAM can still separate the wall pressure fluctuations from the background noise even when the wall pressure fluctuations are much lower than the background noise.

    Table 1 Coordinates of sensors in array.

    Fig. 4 Comparison of results of sound-source separation with intensities of background noise and wall pressure fluctuations in same level.

    Fig. 5 Comparison of results of sound-source separation with intensities of background noise ten times of wall pressure fluctuations.

    3.2. Effects of decay factor

    Generally, the amplitude of wall pressure fluctuations varies along the flow direction, as described by the decay factor in Eqs. (4) and (5). Hence, it is necessary to consider the effect of the decay factor on sound-source separation. Fig. 6 shows the results of sound-source separation when the decay factor is considered, and the real parts of the background noise and the wall pressure fluctuations are presented. The decay factor chosen here is 0.5. The integrated sound-source is computed by the method mentioned in Section 3.1 with sound speeds c0=340 m/s, U0=225 m/s, and Uc=0.65 U0. The broken lines in Fig. 6 indicate the integrated pressure signals of the background noise and the wall pressure fluctuations.The solid lines represent the theoretical results of the wall pressure fluctuations,and the diamonds show the results obtained with the SAM.The decay of the wall pressure fluctuations decreases the weighting of the wall pressure fluctuations in the integrated signal (of the background noise and the TBL noise) along the flow direction. As shown in Fig. 6(b), the interested signal has almost decayed to zero after X=0.2 m when the decay coefficients are large. The integrated sound pressure is dominated by the background noise which is presented in the form of a standard sinusoidal wave (Fig. 6(b),) after the interested signal has decayed to zero (X>0.2 m). However, the SAM can still separate the wall pressure fluctuations from the background noise accurately (Fig. 6).

    3.3. Broadband signal separation analysis

    Fig. 6 Comparison of results of sound-source separation with decay factors considered.

    Aeroacoustics also deals with broadband noise.To investigate the sound-source separation for broadband noise, a pressure signal measured in the FD-12 wind tunnel(China)is used here to simulate the wall pressure fluctuations generated by the TBL. The pressure signal is measured by a pressure sensor at a Mach number of Ma=0.8 and the measured sound pressure spectrum is shown with a solid line named‘Original'in Fig.7,and SPL indicates sound pressure level. Here, this measured pressure signal is assumed to be an interested signal which needs to be separated from the background noise. To create a noisy environment for the sound-source separation, an artificial source of background noise is created to interfere with the interested signal, and the spectrum for the artificial source is shown by the dashed line in Fig. 7. The artificial sound is assumed to propagate in the form of Eq. (1). The integrated noise composed of the measured signal and the background noise is described by the dash-dot line in Fig. 7. As shown in Fig.7,the background noise increases the intensity of the measured result (interested signal) remarkably if the background noise is at the same level with the interested signal. The investigation here attempts to extract the interested sound in this artificial noisy environment to validate the reliability of the SAM in separating the TBL induced pressure fluctuations and the broadband noise.

    Fig.7 Integrated noise signals used for sound-source separation.

    Before the computation of the sound-source separation,the sound level at each sensor needs to be computed by superposing the results of Eqs. (1)and(7) which represent the artificial background noise and the interested pressure signal, respectively. The computation assumes that the flow speed at‘Ma=0.8' is 272 m/s. The interested signals separated from the integrated noise by the SAM are presented in Fig.8.In this figure, the ‘Original' described by the solid line indicates the interested sound. For the SAM computation, two different parameter settings are used. The dashed line expresses the results computed with N=10 and φ=0.1, and the dashdot line indicates the results computed with N=5 and φ=0.2.

    In addition,the method introduced in Ref.17is also used to remove the background noise and the results before and after the removal are compared. The solid-circle line in Fig. 8 indicates the results obtained by the method in Ref.17. The interested sound is computed by subtracting the background noise measured when the wind tunnel is hollow from the total pressure directly.The resolved expression of the interested sound is where Sppf( ) indicates the power spectral density of the interested sound pressure, e is the steering matrix, and G and Gbackgroundindicate the cross spectra of the fluctuation pressure measured on the model and in a hollow wind tunnel, respectively. For comparison, G and Gbackgroundare computed by the integrated noise and the background noise shown in Fig. 8, respectively.

    Fig. 8 Computational results with SAM.

    As shown in Fig.8,the SAM can separate the background noise from the interested signal accurately in a broad frequency range. However, more sensor elements with a smaller loop gain will achieve remarkably higher computational precision. Compared with the SAM, the method introduced in Ref.17is influenced by the SNR. The higher the SNR, the weaker the coherence presented between the two kinds of sound sources and the better the results computed with the method in Ref.17. Fig. 8 shows that the SNR decreases remarkably when the sound frequencies exceed 100 Hz, thus a palpable error is generated when using the method in Ref.17to obtain the interested sound over 100 Hz.

    As can be seen in Fig. 8, the upper limit of the interested frequency is 1 kHz. To extend the frequency range of interest,the half-length of a wave should be more than the minimum distance between any two sensors. Hence, the maximum frequency under investigation could be confined by the dimensions of the used sensors. The wave length corresponding to the maximum frequency should not be smaller than the diameters of the sensors.

    4. Conclusions

    Wall pressure fluctuations are mostly measured in a wind tunnel due to its reliability in simulating a turbulent boundary layer. However, the measurement results obtained by wind tunnel tests are usually contaminated by the background noise of wind tunnels.The jet and the valve noise are important contributors to the background noise in a transonic wind tunnel.In this study,a new method,the SAM,is developed to remove the background noise generated by the jet and the valve of wind tunnels.A salient merit of the SAM is that it can resolve the wall pressure fluctuations without knowing the exact information about the wall pressure fluctuations,such as the intensity, the decay factor, and the convection velocity.

    The results of the investigation demonstrate that the SAM can separate the wall pressure fluctuations from the background noise accurately when they are at the same level. In addition, even when the wall pressure fluctuations are sometimes much lower than the background noise,the wall pressure fluctuations can still be picked up from the signals almost dominated by the background noise. The decay of wall pressure fluctuations changes the weighting of wall pressure fluctuations in the integrated signal of the background noise and the TBL noise. However, it does not affect the results of soundsource separation. Meanwhile, the sound-source separation based on the signal collected in a wind tunnel also reveals the effectiveness of the SAM for broad-band noise.

    The precision of computation plays an important role in determining the wall pressure fluctuations separated from the background noise.Generally,the step length is inversely proportional to the precision of computation.Hence,two ways are suggested to increase the computational accuracy: (A) decreasing the loop gain or(B)increasing the sensors for signal collections.

    高清日韩中文字幕在线| 免费在线观看成人毛片| 在线播放国产精品三级| 内射极品少妇av片p| 三级国产精品欧美在线观看| 91麻豆精品激情在线观看国产| 两个人视频免费观看高清| 日本撒尿小便嘘嘘汇集6| 国产色婷婷99| 身体一侧抽搐| 九色成人免费人妻av| 国产精品久久视频播放| 中文字幕人妻熟人妻熟丝袜美| 人妻制服诱惑在线中文字幕| 亚洲精品一区av在线观看| 在线免费观看的www视频| 69人妻影院| a级毛色黄片| av专区在线播放| 成年女人看的毛片在线观看| 国产欧美日韩精品亚洲av| 国产精品久久视频播放| 真实男女啪啪啪动态图| 身体一侧抽搐| 男女边吃奶边做爰视频| 精品无人区乱码1区二区| 美女被艹到高潮喷水动态| 欧美+日韩+精品| 国产精品99久久久久久久久| 欧美日本视频| 欧美最黄视频在线播放免费| av在线老鸭窝| 国产aⅴ精品一区二区三区波| 久久精品国产亚洲av天美| 一本精品99久久精品77| 亚洲在线观看片| 99热这里只有是精品50| 国产aⅴ精品一区二区三区波| 亚洲在线自拍视频| 欧美潮喷喷水| 91狼人影院| 婷婷精品国产亚洲av| 国产成人精品久久久久久| 99久久精品热视频| 春色校园在线视频观看| 成人美女网站在线观看视频| 国内少妇人妻偷人精品xxx网站| 亚洲精品色激情综合| 俺也久久电影网| av黄色大香蕉| 插逼视频在线观看| 亚洲熟妇中文字幕五十中出| 少妇熟女欧美另类| 久久久欧美国产精品| 亚洲av中文av极速乱| 国产精品三级大全| 两个人视频免费观看高清| av在线天堂中文字幕| 成人特级av手机在线观看| 国产国拍精品亚洲av在线观看| 国产成年人精品一区二区| 国产一区二区在线av高清观看| 欧美一区二区国产精品久久精品| 日韩强制内射视频| 观看美女的网站| 亚洲五月天丁香| 午夜福利视频1000在线观看| 国产淫片久久久久久久久| 1024手机看黄色片| 国产单亲对白刺激| 亚洲av成人精品一区久久| 又黄又爽又免费观看的视频| 中文字幕av成人在线电影| 国产精品日韩av在线免费观看| 狂野欧美白嫩少妇大欣赏| 国产欧美日韩精品一区二区| 国产av在哪里看| 99九九线精品视频在线观看视频| 免费看日本二区| 日韩精品中文字幕看吧| 黑人高潮一二区| 日韩亚洲欧美综合| 午夜福利在线观看免费完整高清在 | 一进一出抽搐gif免费好疼| 变态另类丝袜制服| 午夜爱爱视频在线播放| 欧美激情国产日韩精品一区| 99久久精品热视频| 99久国产av精品国产电影| 欧美一级a爱片免费观看看| 亚洲中文字幕一区二区三区有码在线看| 国产成年人精品一区二区| 国产淫片久久久久久久久| 精品久久久久久久人妻蜜臀av| 国产三级中文精品| 狂野欧美激情性xxxx在线观看| 国产中年淑女户外野战色| 国产精品不卡视频一区二区| 午夜a级毛片| 一区二区三区免费毛片| 日韩大尺度精品在线看网址| 99九九线精品视频在线观看视频| 99国产精品一区二区蜜桃av| 欧美在线一区亚洲| 国产精品女同一区二区软件| 最近视频中文字幕2019在线8| 亚洲av第一区精品v没综合| 18禁在线无遮挡免费观看视频 | 国产视频一区二区在线看| 插逼视频在线观看| 日韩av不卡免费在线播放| 亚洲成人中文字幕在线播放| 免费电影在线观看免费观看| 国产高清视频在线观看网站| 精品久久久久久久久av| 狠狠狠狠99中文字幕| 亚洲最大成人手机在线| 成年免费大片在线观看| 国产又黄又爽又无遮挡在线| a级一级毛片免费在线观看| 国产探花极品一区二区| 国产精品久久久久久亚洲av鲁大| 亚州av有码| 亚洲性夜色夜夜综合| 综合色av麻豆| 国产成人a∨麻豆精品| 日韩亚洲欧美综合| av卡一久久| 性插视频无遮挡在线免费观看| 99热6这里只有精品| 国产综合懂色| 精品国内亚洲2022精品成人| 免费av观看视频| 在线观看66精品国产| 老熟妇仑乱视频hdxx| 成人国产麻豆网| a级毛片a级免费在线| 高清毛片免费观看视频网站| 精品午夜福利视频在线观看一区| 18+在线观看网站| 三级男女做爰猛烈吃奶摸视频| 亚洲专区国产一区二区| 自拍偷自拍亚洲精品老妇| 国产蜜桃级精品一区二区三区| 深爱激情五月婷婷| 麻豆成人午夜福利视频| 日韩欧美国产在线观看| 午夜日韩欧美国产| 亚洲欧美日韩东京热| 欧美三级亚洲精品| av天堂在线播放| 亚洲熟妇熟女久久| 免费人成视频x8x8入口观看| 国产真实乱freesex| 啦啦啦啦在线视频资源| 99久久成人亚洲精品观看| 熟妇人妻久久中文字幕3abv| 99视频精品全部免费 在线| 神马国产精品三级电影在线观看| 色综合亚洲欧美另类图片| 舔av片在线| 禁无遮挡网站| 欧美极品一区二区三区四区| 免费人成在线观看视频色| 国产精品国产高清国产av| 亚洲真实伦在线观看| 五月玫瑰六月丁香| 日韩欧美精品免费久久| 99热这里只有是精品50| 国产三级在线视频| 春色校园在线视频观看| 黄色配什么色好看| 狂野欧美激情性xxxx在线观看| 亚洲av电影不卡..在线观看| 日韩一本色道免费dvd| 久久人人爽人人爽人人片va| 99久国产av精品| 国产高清不卡午夜福利| 欧美日韩乱码在线| 人妻丰满熟妇av一区二区三区| 亚洲精品一区av在线观看| 久久午夜福利片| 看片在线看免费视频| 搡老岳熟女国产| 亚洲欧美清纯卡通| 亚洲欧美成人精品一区二区| 日韩欧美三级三区| 成人亚洲精品av一区二区| 国产午夜福利久久久久久| 亚洲精品国产成人久久av| 精品99又大又爽又粗少妇毛片| 在线天堂最新版资源| 五月玫瑰六月丁香| 又爽又黄无遮挡网站| 国产老妇女一区| 丰满乱子伦码专区| 级片在线观看| 午夜影院日韩av| 99国产精品一区二区蜜桃av| 亚洲第一电影网av| 欧美人与善性xxx| 国产老妇女一区| 日日啪夜夜撸| 国产一区二区在线观看日韩| 亚洲最大成人av| 最近2019中文字幕mv第一页| 午夜精品一区二区三区免费看| 国产激情偷乱视频一区二区| 日本爱情动作片www.在线观看 | 国产精品伦人一区二区| 级片在线观看| 伊人久久精品亚洲午夜| 亚洲国产高清在线一区二区三| 亚洲成人中文字幕在线播放| 国产精品三级大全| 欧美色视频一区免费| 精品人妻一区二区三区麻豆 | 欧美日韩精品成人综合77777| 激情 狠狠 欧美| 男人的好看免费观看在线视频| 国产日本99.免费观看| 99热精品在线国产| 尾随美女入室| 嫩草影视91久久| 哪里可以看免费的av片| 亚洲精品久久国产高清桃花| 人人妻人人看人人澡| 精品国产三级普通话版| 一级av片app| 1000部很黄的大片| 一级毛片我不卡| 国产成人一区二区在线| 亚洲无线在线观看| 人妻丰满熟妇av一区二区三区| 日韩欧美精品v在线| 国产亚洲精品久久久com| 精品国内亚洲2022精品成人| 午夜福利高清视频| 免费电影在线观看免费观看| 两个人视频免费观看高清| 日韩 亚洲 欧美在线| 国产av一区在线观看免费| 亚洲av一区综合| 久久久久九九精品影院| 国产亚洲欧美98| 特级一级黄色大片| 草草在线视频免费看| 国产欧美日韩一区二区精品| 成人永久免费在线观看视频| 日韩欧美三级三区| 亚洲一区高清亚洲精品| 亚洲成人av在线免费| 床上黄色一级片| 亚洲av电影不卡..在线观看| 国产高潮美女av| 亚洲aⅴ乱码一区二区在线播放| 久久人妻av系列| 国产三级在线视频| 男人的好看免费观看在线视频| 久久久久免费精品人妻一区二区| 色吧在线观看| 国产亚洲欧美98| 国产亚洲精品久久久久久毛片| 日韩中字成人| 中国国产av一级| 国产v大片淫在线免费观看| 特级一级黄色大片| 黄片wwwwww| 亚洲精品影视一区二区三区av| 久久久久精品国产欧美久久久| 国产一级毛片七仙女欲春2| 国产黄a三级三级三级人| 久久精品国产鲁丝片午夜精品| 三级国产精品欧美在线观看| 中国国产av一级| 久久人人爽人人片av| 国产一区二区在线av高清观看| 日韩av在线大香蕉| 久久99热6这里只有精品| 日韩成人伦理影院| 国产探花在线观看一区二区| 久久精品国产亚洲av天美| 熟妇人妻久久中文字幕3abv| 男人舔女人下体高潮全视频| 欧美xxxx黑人xx丫x性爽| 免费黄网站久久成人精品| 日本成人三级电影网站| 亚洲自偷自拍三级| 亚洲国产色片| 国产精品野战在线观看| 老熟妇仑乱视频hdxx| 亚洲精品久久国产高清桃花| 欧美不卡视频在线免费观看| 两个人视频免费观看高清| 国产亚洲精品av在线| 国产黄a三级三级三级人| 国产aⅴ精品一区二区三区波| 少妇被粗大猛烈的视频| 国产v大片淫在线免费观看| 亚洲精品在线观看二区| 美女 人体艺术 gogo| 国产免费一级a男人的天堂| 亚洲高清免费不卡视频| 亚洲精品一卡2卡三卡4卡5卡| 亚洲七黄色美女视频| 欧美日本视频| 美女xxoo啪啪120秒动态图| 中文字幕久久专区| 十八禁网站免费在线| 国产av一区在线观看免费| 看非洲黑人一级黄片| 日本黄色片子视频| 日韩欧美精品v在线| 在线免费观看不下载黄p国产| 亚洲成人久久性| 亚洲精品国产av成人精品 | 久久久久久大精品| 欧美另类亚洲清纯唯美| 尾随美女入室| 在线看三级毛片| 99精品在免费线老司机午夜| ponron亚洲| 97在线视频观看| 国产高潮美女av| 亚洲精品亚洲一区二区| 变态另类丝袜制服| 最近的中文字幕免费完整| 老女人水多毛片| 日本黄色片子视频| 日本熟妇午夜| 97人妻精品一区二区三区麻豆| 99riav亚洲国产免费| 99热这里只有是精品50| 亚洲欧美日韩无卡精品| 男人狂女人下面高潮的视频| av国产免费在线观看| 亚洲一区高清亚洲精品| 97人妻精品一区二区三区麻豆| 草草在线视频免费看| 亚洲成人久久爱视频| 欧美三级亚洲精品| 国产欧美日韩一区二区精品| 老熟妇仑乱视频hdxx| 给我免费播放毛片高清在线观看| 身体一侧抽搐| 能在线免费观看的黄片| 精品一区二区三区视频在线观看免费| 久久人人爽人人爽人人片va| 又爽又黄无遮挡网站| 悠悠久久av| 国产一区二区三区在线臀色熟女| 一级黄片播放器| 欧美最黄视频在线播放免费| 亚洲精品在线观看二区| 九色成人免费人妻av| 少妇人妻一区二区三区视频| 亚洲国产精品成人综合色| 在线播放国产精品三级| 欧美日本亚洲视频在线播放| 午夜老司机福利剧场| 久久久a久久爽久久v久久| 悠悠久久av| 亚洲av第一区精品v没综合| 国产aⅴ精品一区二区三区波| 99久久久亚洲精品蜜臀av| 观看美女的网站| 亚洲国产精品国产精品| 日韩av不卡免费在线播放| 国产人妻一区二区三区在| 国产女主播在线喷水免费视频网站 | 日韩欧美国产在线观看| 亚洲18禁久久av| 人妻丰满熟妇av一区二区三区| 五月玫瑰六月丁香| 99久久精品一区二区三区| 国产一区二区亚洲精品在线观看| 久久久久久久久中文| or卡值多少钱| 亚洲性久久影院| 我要搜黄色片| 国内精品一区二区在线观看| 国产精品久久视频播放| 永久网站在线| 国内精品宾馆在线| 看十八女毛片水多多多| 色综合站精品国产| 一级a爱片免费观看的视频| 综合色av麻豆| 成人二区视频| 插逼视频在线观看| 久久精品人妻少妇| 日韩av不卡免费在线播放| 成人特级av手机在线观看| av黄色大香蕉| 日韩精品有码人妻一区| 欧美色视频一区免费| 人人妻人人看人人澡| 色噜噜av男人的天堂激情| 亚洲四区av| 两性午夜刺激爽爽歪歪视频在线观看| 男人舔女人下体高潮全视频| 男人和女人高潮做爰伦理| 国产伦精品一区二区三区四那| 久久这里只有精品中国| 99热这里只有是精品50| 久久久久久国产a免费观看| 在线国产一区二区在线| а√天堂www在线а√下载| 免费无遮挡裸体视频| 欧美成人一区二区免费高清观看| 小说图片视频综合网站| 久久久国产成人免费| 又粗又爽又猛毛片免费看| 99热6这里只有精品| 午夜福利18| 色av中文字幕| 日日撸夜夜添| 精品久久久噜噜| 国产一区二区亚洲精品在线观看| 欧美人与善性xxx| 99国产精品一区二区蜜桃av| 夜夜看夜夜爽夜夜摸| 久久久国产成人免费| 久久欧美精品欧美久久欧美| 狂野欧美白嫩少妇大欣赏| 免费观看在线日韩| 亚洲精品影视一区二区三区av| 免费黄网站久久成人精品| 免费看av在线观看网站| 亚洲性久久影院| 国产成人91sexporn| 免费观看的影片在线观看| 丰满人妻一区二区三区视频av| 亚洲国产精品成人综合色| 免费人成在线观看视频色| 99久久精品热视频| 99久久精品国产国产毛片| 人妻制服诱惑在线中文字幕| 美女黄网站色视频| 99久国产av精品| 精品一区二区三区视频在线观看免费| 搡老妇女老女人老熟妇| 又爽又黄a免费视频| 精品午夜福利在线看| 波多野结衣巨乳人妻| 久久久精品94久久精品| 不卡视频在线观看欧美| 插阴视频在线观看视频| 天天躁夜夜躁狠狠久久av| 亚洲一区高清亚洲精品| 丰满的人妻完整版| 成人综合一区亚洲| 露出奶头的视频| 特级一级黄色大片| 日韩欧美精品免费久久| 天天躁夜夜躁狠狠久久av| 又爽又黄无遮挡网站| 欧美日韩精品成人综合77777| 精品人妻一区二区三区麻豆 | 日韩 亚洲 欧美在线| 欧美日韩在线观看h| 欧美国产日韩亚洲一区| 久久精品影院6| 中文资源天堂在线| 国产精品久久视频播放| 一进一出抽搐gif免费好疼| 夜夜夜夜夜久久久久| 久久精品久久久久久噜噜老黄 | 十八禁网站免费在线| 中文字幕人妻熟人妻熟丝袜美| 美女黄网站色视频| 97超碰精品成人国产| 欧美成人精品欧美一级黄| 精品久久久久久久久亚洲| 中文在线观看免费www的网站| 麻豆国产97在线/欧美| 国产午夜精品论理片| 久久天躁狠狠躁夜夜2o2o| 99精品在免费线老司机午夜| 久久久久久久久中文| 亚州av有码| 人妻久久中文字幕网| 婷婷精品国产亚洲av在线| 一区二区三区免费毛片| 97超级碰碰碰精品色视频在线观看| 欧美不卡视频在线免费观看| 精品免费久久久久久久清纯| 中文资源天堂在线| 乱人视频在线观看| 18+在线观看网站| 久久精品综合一区二区三区| 成人欧美大片| 国产av不卡久久| 啦啦啦观看免费观看视频高清| 亚洲18禁久久av| 99精品在免费线老司机午夜| 天堂动漫精品| 成人综合一区亚洲| 欧美性猛交╳xxx乱大交人| 亚洲精品一卡2卡三卡4卡5卡| 国产片特级美女逼逼视频| 国产精品福利在线免费观看| 国内久久婷婷六月综合欲色啪| 麻豆国产97在线/欧美| 伊人久久精品亚洲午夜| 一本一本综合久久| 97人妻精品一区二区三区麻豆| 又黄又爽又刺激的免费视频.| 中国国产av一级| 国产高清不卡午夜福利| 秋霞在线观看毛片| 国产蜜桃级精品一区二区三区| 免费一级毛片在线播放高清视频| 亚洲aⅴ乱码一区二区在线播放| 亚洲一区二区三区色噜噜| 性欧美人与动物交配| 色吧在线观看| 欧美日韩国产亚洲二区| 美女cb高潮喷水在线观看| 男人舔奶头视频| 国产久久久一区二区三区| 国产一区二区亚洲精品在线观看| 亚洲色图av天堂| 亚洲成人久久爱视频| 国内揄拍国产精品人妻在线| 看片在线看免费视频| 天美传媒精品一区二区| 黄色一级大片看看| 欧美成人一区二区免费高清观看| 精品少妇黑人巨大在线播放 | 九九热线精品视视频播放| 成人综合一区亚洲| 99在线人妻在线中文字幕| 久久久久精品国产欧美久久久| 亚洲av.av天堂| 国产综合懂色| 99热网站在线观看| 老女人水多毛片| 久久中文看片网| 亚洲在线自拍视频| 国产成人福利小说| 午夜日韩欧美国产| 免费大片18禁| 日韩在线高清观看一区二区三区| 亚洲成人久久性| 午夜爱爱视频在线播放| 成人亚洲精品av一区二区| 久久婷婷人人爽人人干人人爱| 亚洲av成人精品一区久久| 丰满乱子伦码专区| 欧美绝顶高潮抽搐喷水| 成人无遮挡网站| 国产免费一级a男人的天堂| 日本撒尿小便嘘嘘汇集6| 国产高清激情床上av| 国产综合懂色| 国产一区二区激情短视频| 欧美高清成人免费视频www| 亚洲av一区综合| 看十八女毛片水多多多| 天天躁日日操中文字幕| 亚洲最大成人中文| 少妇裸体淫交视频免费看高清| 日韩精品青青久久久久久| 又爽又黄无遮挡网站| 一区福利在线观看| 国产在视频线在精品| 国产国拍精品亚洲av在线观看| 少妇的逼水好多| 99九九线精品视频在线观看视频| 日韩精品中文字幕看吧| 精品免费久久久久久久清纯| 国产精品亚洲一级av第二区| 成人av在线播放网站| 久久人人精品亚洲av| 国产精品久久电影中文字幕| 深夜精品福利| 精品久久久久久久人妻蜜臀av| 深爱激情五月婷婷| 国产成人一区二区在线| 成人午夜高清在线视频| 欧美最新免费一区二区三区| 久99久视频精品免费| 日本五十路高清| 久久精品国产鲁丝片午夜精品| 国产高清视频在线观看网站| 97超碰精品成人国产| 卡戴珊不雅视频在线播放| 亚洲欧美精品自产自拍| 欧美xxxx性猛交bbbb| 人人妻人人澡欧美一区二区| 如何舔出高潮| 久久久久久伊人网av| av在线老鸭窝| 国产精品三级大全| 搞女人的毛片| 国产精品久久视频播放| 亚洲中文字幕日韩| 在线免费观看不下载黄p国产| www日本黄色视频网| 国产又黄又爽又无遮挡在线| 我要搜黄色片| 22中文网久久字幕| 尤物成人国产欧美一区二区三区| 一个人免费在线观看电影| 国产精品综合久久久久久久免费| 级片在线观看| 两个人视频免费观看高清| 国产精品久久久久久亚洲av鲁大| 精品久久久久久久久久免费视频| 国产成人a∨麻豆精品| 午夜影院日韩av| av专区在线播放| 男人的好看免费观看在线视频| 男人舔女人下体高潮全视频|