Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line PROPELLER WAKE ANALYSIS BY MEANS OF PIV 493 Propeller Wake Analysis by Means of PIV F.Di Felice (Istituto Nazionale per Studi ed Esperienze di Architettura Navale, Italy) G.Romano (Rome University, Italy) M.Elefante (Centro Esperienze Idrodinamiche Marina Militare, Italy) ABSTRACT An experimental investigation of the propeller wake in a cavitation tunnel has been performed using Particle Image Velocimetry (PIV). The hydrodynamic and geometrical investigation of the wake and its evolution has been pointed out. The blade viscous wake, developing from the blade surface boundary layers, the trailing vortex sheets, due to the radial gradient of the bound circulation, and the velocity fluctuation distributions are identified and discussed. The near wake geometry is described through the bending of the blade wake sheets, the slipstream contraction and the tip vortex trajectory and viscous interactions. In the near field the effects of turbulent diffusion and viscous dissipation, which cause a rapid space-broadening of the velocity gradients in the trailing edge wake, are also examined. In the far wake the development of the slipstream instability and the breakdown of the hub and tip vortices are outlined. INTRODUCTION The experimental investigation of the propeller wake holds an important role for the design and the performance analysis of ship propulsion. In modern design, to reduce propeller-induced hull vibrations, efficiency decay and noise generation, due to cavitation, there is a continuous trend towards an increased complexity of the blade geometry. This complexity is primarily due to the low aspect ratio and to the skew of marine propellers, which cause strong three- dimensional effects. Therefore, there is a rising interest on detailed data of the velocity flow field around the blades and in the wake. The knowledge of the velocity field in the wake enables to check locally the design requirements. For example, this can be done by comparing the measured blade section drag coefficients and bound circulation with those provided by design distributions (Kobayashi 1982, Koyama 1986 and Jessup 1989). Moreover, the knowledge of the position of the trailing vortex sheets is necessary to evaluate the actual wake-induced velocity field around the blades and to determine the propeller performances. Velocity measurements are also a tool for the development and the validation of numerical codes and flow modeling. Most current numerical methods for propeller investigations are based on potential flow theories and simplified wake models. More complex and refined models of viscous flows (Kobayashi 1981, Jessup 1989, Arndt and Maines 1994), of the hub effects (Wang 1985), of the trailing vortex sheets and of tip roll-up process are required for increasing the accuracy of numerical predictions. Measurements by Laser Doppler Velocimetry (LDV) have been a turning point in the analysis of the three- dimensional complex flow around rotors and propellers, providing among the others, quantitative information on the roll- up process, the slipstream contraction and the tip vortex evolution (Min 1978, Kobayashi 1982, Cenedese et al. 1985, Jessup 1989, Chesnack and Jessup 1998). LDV has the capability of velocity direction recognition, high spatial resolution, good frequency response and of a non intrusive probe. However, most of the previous LDV analyses are focused on cross-sections of the wake and a few examples are available concerning the evolution of the wake downstream of the slipstream contraction (Stella et al. 1998). This is mainly due to the fact that the LDV technique allows an efficient data acquisition at a point in time (different propeller revolution angles) and hence an easy reconstruction of the flow field in a cross-section, while the longitudinal survey requires a time consuming sweep over many points. In the present study, the analysis of the propeller wake is performed by using the 2D Particle Image Velocimetry technique (PIV), which allows a powerful investigation of the radial and axial velocity components in the longitudinal plane. PIV measurements, as previously done by Cotroni et al. the authoritative version for attribution.
lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line PROPELLER WAKE ANALYSIS BY MEANS OF PIV 494 (1999), are taken in phase with the blade angular positions in order to resolve the evolution of the wake during the propeller revolution. Wake characteristics and tip vortex spatial fluctuations, leading to the breakdown farther downstream, are pointed out considering three adjacent windows starting from the blade trailing edge to about 1.5 propeller diameters downstream. The propeller used in the present experiment is the same investigated by Cenedese et al. (1985) and by Stella et al. (1998). Although some wake detail is strictly dependent on both propeller geometry and loading conditions, the results of this investigation will be discussed with emphasis on those flow features of a general content. EXPERIMENTAL SET-UP The present PIV measurements were performed at the Italian Navy Cavitation Tunnel (C.E.I.M.M.). The test section is a square, closed jet type (0.6mÃ0.6mÃ2.6m). Perspex windows on the four walls enable the 90° optical access required for PIV. The nozzle contraction ratio is 5.96 and the maximum water speed is 12 m/s. The maximum free stream turbulence intensity in the test section is 2% in the regions behind the wake of the shaft supports, while it reduces to 0.6% in the propeller blade inflow at a radial position equal to 0.7 R (R being the propeller radius). The flow uniformity of the axial and the vertical components is within 1%. The sketch of the instrumentation set-up is shown in figure 1. The propeller model is mounted on a front dynamometer shaft. This arrangement of the propeller and the length of the test section, which is about 15 times the propeller diameter, allows the slipstream to develop freely in the downstream direction as in a real operative condition. An encoder, with a resolution of 0.1°, mounted on the dynamometer shaft, feeds a special signal processor which sends a trigger signal to a special synchronising device for each propeller angular position. The synchroniser provides a TTL trigger signal to a cross-correlation camera (1018x1018 pixel), and to a double cavity Nd-Yag laser (200 mJ per pulse at 12.5 Hz), to allow image acquisitions for each propeller angular position. The digital cross-correlation video camera, allows the recordings of two separate images (one for each laser pulse) within a few microseconds at a maximum frame rate of 15 Hz. By using cross-correlation, the directional ambiguity is completely removed. The instantaneous velocity fields were acquired from a distance up to 700 mm from the side window, using a 60 mm lens with 2.8 f-number and imaging an area of about 100 x 100 mm2. The tracer particles are one of the critical aspects of the PIV technique, especially in case of large facilities. Being the technique based on the measurement of the particle displacement, it is fundamental that the seeding accurately follows the water flow velocity (Hunter and Nichols 1985, Melling 1986, Mayers 1991). This requires particles having a diameter on the order of some µm. At the same time, it is mandatory to achieve a high uniform seeding density in the region of interest, at least 15 particle pairs per interrogation window (Keane and Adrian 1990), in order to accurately perform auto/ cross correlation analysis. To this purpose, the water in the tunnel was initially filtered, and then seeded with 10 µm silver coated hollow glass spherical particles with high diffraction index and density of Figure 2: Tested propeller model. Figure 1: Experimental set-up the authoritative version for attribution.
lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line PROPELLER WAKE ANALYSIS BY MEANS OF PIV 495 about 2 g/mm3 The four blade propeller model (figure 2) is skewed, with a uniform pitch (pitch/diameter=1.1) a forward rake angle of 4° 3â and a diameter of 227.2 mm. PIV measurements were carried out with a propeller angular velocity n=25 rps and an upstream water speed Uinf=4.25 m/s, corresponding to an advance ratio J=Uinf/(nD) of 0.748, in moderately high loading condition. In such conditions, the propeller shows tip vortex hub cavitation which was maintained outside of the measurement windows to avoid camera blooming. The blade Reynolds number Rn=(c0.7V0.7)/ν, where c0.7 and V0.7 are respectively the chord length and the velocity at r/R=0.7, was equal to 1.12·106. The cavitation number ÏV=(PâPV)/q, being P the absolute ambient pressure, PV the vapour pressure and q the stagnation pressure of the propeller upstream flow, was 9.3. The PIV system was arranged to measure, in the mid longitudinal plane of the propeller, the axial and vertical velocity components simultaneously in the tunnel frame. In view of the symmetry of the propeller inflow and of the steady conditions, when the light is located on the vertical radius (along the z-axis), the axial component of the velocity and the vertical one correspond respectively to the axial and the radial components in the propeller moving frame. To investigate the propeller wake at least 1 diameter farther downstream of the stream tube contraction, the measurements have been performed over 3 adjacent windows by traversing the camera (with an accuracy of about 0.1 mm) as shown in figure 3. The initial reference position is fixed with an accuracy of about 0.5 mm by imaging a special target device. Figure 3: Measurement planes IMAGE ANALYSIS The acquired images were analysed using an algorithm in which the window off-set correlation method has been implemented (Westerweel 1997). Furthermore a recursive processing method is used by implementing a hierarchical approach in which the sampling grid is continually refined and also the size of the interrogation windows is reduced during the iterations. In Figure 4 the iterative process starting from windows of 128 px2 to the final one 16 px2 is presented. In the last iteration the windows are also overlapped to obtain a better reconstruction of the whole flow field especially in the regions with strong gradients. This procedure has the added capability of applying interrogation windows with size smaller than the particle image displacement increasing both the dynamic range and the spatial resolution of the measurement technique. The effectiveness of this recursive algorithm, allows the analysis using sub-windows up to 16 px2, with a limited number of spurious determinations (less than 7% of the total number using sub windows of 16X16 pixel). To eliminate the remaining spurious vectors, always present and due to the lack of particles in the interrogation windows or noise in the background of the images, each data set is subjected to a validation procedure to detect and replace spurious displacement vectors. Four different kinds of validation techniques have been implemented: - a local median-filtering method, to identify displacement vectors that deviate by a prescribed amount in magnitude or direction, from adjacent vectors (Westerweel 1994); - a cross-correlation Signal to Noise Ratio (SNR) validation, where the highest correlation peak is compared with the second one, and validated if the ratio is greater than a predefined value d=1.2 (Keane and Adrian 1992) or even lower according to the seeding density; - a displacement range validation, which rejects vectors outside a certain velocity range; - a geometric validation, which rejects vectors within a certain predefined area, useful when solid surfaces are within the area of investigation. Different flags are associated with the spurious vectors from each one of the previous four different types of validation. Therefore, in the following statistical analysis the rejection criterion can be selected as a combination of some as well as all of them for filtering the data to be used for statistics computation. the authoritative version for attribution.
lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line PROPELLER WAKE ANALYSIS BY MEANS OF PIV 496 Figure 4: Multigrid adaptive window offset correlation method: displacement vector field at different iteration steps For the result presented in the following a final window size of 24 px2 has been adopted as the best compromise of spurious vector reduction and spatial resolution that in the present case is equivalent to 2.4X2.4 mm2. For a given propeller angle, 65 pairs of image have been acquired to evaluate statistical quantities. Angles from 0° to 85° have been considered with a step of 5° for a total amount of about 4000 pairs of images for the considered windows. Statistic evaluation has been performed considering only velocity vectors that fulfil simultaneously all the filter above specified. MEASUREMENT UNCERTAINTY A comprehensive discussion on the uncertainty and the accuracy of the PIV technique is out of the goal of the present work and this aspect is a complex topic with many open points. A detailed analysis of this aspect can be found in Raffel et al. (1997). In the following, the main assessments necessary to qualify the present results are reported. The uncertainty on velocity measurements by means of a PIV system is mainly due to the error on particle displacements evaluation which can be considered less than 1/10th of a pixel for the present image analysis algorithm. In terms of velocity is in the order of 1 cm/s. Peak locking errors, mainly due to particle image size, has been reduced, as much as possible, by using image defocusing techniques. Errors due to noise were important only in flow regions where light reflections from the cavitating hub or from the blade surface were present. Even if erroneous vectors are eliminated and replaced by interpolation during post-processing, sometimes spurious vectors are validated and affect the statistics. This effect is relevant especially for the second order statistics. the authoritative version for attribution.
lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line PROPELLER WAKE ANALYSIS BY MEANS OF PIV 497 The accuracy of the statistical estimations definitely depends on the number of acquired samples and on the shape of the velocity probability distribution function. The probability density function in the tip vortex core and in the blade wake markedly differs from the Gaussian. Furthermore, a lack of data and less samples available for statistics have been observed in these regions. In figure 5 the total number of samples used for statistical computations is plotted at a given angle for the first measurement window. PIV provides an explanation for this phenomenon: the strong centrifugal forces in the vortex core reduce dramatically the probability to have useful particles required for the measurement as can be seen in figure 6 where a single PIV image is shown. The tip vortex location can be easily identified as a black hole in the image of about 40 px in diameter. This problem has been noticed also for LDV measurements and the same explanation, as before, is suggested. However, only PIV provides a direct justification of such a problem. In the same figure, the dimension of the interrogation window used for the analysis is shown. The first outcome of such aspect is that the tip vortex velocities are underestimate. By using the t-Student distribution (for which the confidence interval is ±1.96*rms/â(Nâ1), with N=65), it is possible to estimate the uncertainty on a velocity component to be about 1/4th of the measured rms and hence equal to 0.025 m/s (at the measurement points far from the tip vortex and the blade wake). In the tip vortex core and near to the blade (where the highest velocity gradient are encountered), only a few data are available to compute statistics which result in low accuracy estimation especially for the second order statistics PROPELLER WAKE ANALYSIS An example of an instantaneous flow field obtained in the first measurement window for a revolution angle θ=0° is shown in figure 7. For graphical reasons the vectors have been skipped of a factor two and the upstream velocity has been subtracted in order to point out the flow perturbation induced by the propeller. Two tip vortices due to the actual blade and to the previous one, as well as the wake released by the blade, are recognised. The strong flow acceleration near the hub due to the presence of the hub vortex is also observed. The mean velocity obtained over 65 image pairs for the same angle shows similar features, thus indicating that the flow field is dominated by the propeller revolution. In figure 8 the whole measured flow field is given. Due to graphical reasons, the contour plots for the U and V components, nondimensional by the upstream velocity Uinf are shown in a mirrored layout. In the diagram the error, due to the camera positioning in the overlapping region of the three measurement windows near the hub, can be also Figure 6: PIV single esposure image. The tip vortex Figure 5: Samples distribution, θ=0° location is identified due to the lack of particles in the vortex core which is compared with the interrogation window the authoritative version for attribution.
lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line PROPELLER WAKE ANALYSIS BY MEANS OF PIV 498 noticed. Figure 7: Instantaneous perturbation velocity field of the propeller (upstream velocity Uinf removed). for θ=0 The traces of six tip vortices are evident in the measurement plane. The tip vortex velocity isocountours show a typical Rankine vortex pattern: the U distribution is similar to the V distribution after rotation by 90°. The viscous wake due to the boundary layer on the blade is represented by a defect in the velocity. The velocity defect is strong at the trailing edge of the blade and is rapidly smoothed and faded downstream. As shown by Stella et al. (1998), the velocity defect is stronger at the blade root due to the larger thickness of the blade profile near the hub. The blade wake almost disappears within one diameter downstream, whereas a strong deformation, due to the higher axial velocity at the inner radius bends the blade wake. The strong acceleration of the radial velocity component near the hub reveals the strong roll up process of the hub vortex (which is cavitating just outside the measurement area). It also contributes to the wake deformation convecting downstream the flow field. This type of information is also evident in figure 9 where the modulus of the in plane velocity ((U2+V2)1/2/Uinf), and the streamlines are shown. The strong deformation of the streamlines due to the effect of the tip vortices and the hub vortex is highlighted. In figure 10, the vorticity generated by the propeller, non-dimensional by the upstream velocity and the propeller diameter, is given. In such a figure the streamlines obtained by subtracting the upstream velocity are also shown: they highlight the effect of the tip vortex roll-up. In figure 11, the evolution of the vorticity field for the revolution angles 20°, 40°, 60° and 80° are shown. The following considerations can be done: - The trailing vorticity, shed from the blade trailing edge, consists of two layers of opposite sign which remain distinct in the wake. This is mainly due to the presence of the blade boundary layer which separately generates the two sheets. - The blade wake continuously spirals around the vortex core which enlarges downstream the authoritative version for attribution.
lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line PROPELLER WAKE ANALYSIS BY MEANS OF PIV 499 - The tip vortex vorticity diffusion is very strong up to x/R=1 as also shown in figure 12, where the mean vorticity of the tip vortex is plotted along the downstream distance. Such result is similar to those obtained by Jessup (1989) for the unskewed propeller 4119. - As soon as the blade wake is released and before it is dissipated, a strong deformation takes place. At about x/ R=1 the wake of the actual blade feels the action of the tip vortex of the previous one which starts to deform the wake. At x/Râ1.5 the wake of the actual blade interacts with the tip vortex of the previous one. For x/R>2 the wake of the actual blade looses the link with its tip vortex and is rolled up by the previous blade tip vortex. Figure 13 shows the distribution of the turbulence intensity ÏU and ÏV for the measured velocity components for θ=0°, while figure 14 shows the evolution of ÏV for θ=20°, 40°, 60°, 80°. Even if the confidence of the statistical estimator is limited, due to the fact that has been evaluated only over 65 samples, some important features of the wake can be recognised. The turbulent wake released by the blade is quickly dissipated and diffused downstream. The same process of wake deformation and broadening, due to the action of the tip vortices and of the hub vortex, observed in the vorticity plots, is also seen in the turbulence level distributions. Nevertheless, new information are obtained by this second order statistics: - The effect of hub vortex roll up in the wake is very important and at x/R=2 the blade wake, linking the tip vortex to the hub vortex, almost disappears especially for the ÏU distribution. - Turbulence diffusion from the hub vortex occurs in the longitudinal evolution especially for ÏV as expected due to the hub vortex orientation. - Some small scale turbulence, more evident in the ÏV distribution, generated probably at the leading edge of the blade, is quickly dissipated downstream within one propeller radius - The effect of noise in the images, due to the cavitating hub is pointed out by the intense spikes in the turbulence level distribution - Velocity fluctuations at the tip vortex core increase while this is convected downstream. This aspect is related to the vortex breakdown instability which affects the velocity fluctuations through spatial oscillations of the core. Furthermore, the pattern of the turbulence distribution suggests that the tip vortex oscillation occurs in some preferential directions. The tip vortex fluctuations, leading to the breakdown, can be pointed out also by evaluating the standard deviation of the spatial fluctuations of tip vortex core with respect to the mean at a given longitudinal position. The result, given in figure 15, shows that the amplitude of the spatial fluctuation is increasing downstream and that the amplitude of transversal fluctuations has a higher growth rate in respect to the longitudinal. This result is similar to those obtained by Cotroni et al. (1999) for a different propeller and seems to be a general feature of the helical vortex system behaviour. Another important feature of the tip vortex system instability, pointed out by PIV, is the separation of the four tip vortex trajectories after the section of maximum contraction. In figure 16 the evolution of the tip vortex trajectory for different revolution angles is shown. After the contraction, and approximately at the same location where the tip vortex interacts with the wake of the subsequent blade, there is a clear separation of the trajectory of the tip vortices due to the different blades. Furthermore, as pointed out in the same figure, the location of the tip vortices in the measurement plane at different angles points out the oscillation of the whole stream tube. This is loosing its axisymmetry and blade periodicity but in the first phase of the breakdown still maintains the phase with the propeller revolution. Such behaviour also provides an explanation of the blade sub harmonics pressure fluctuations sometime experienced by the hull. Such measurements are in perfect agreement with the flow visualisations obtained in incipient cavitation given in figure 17. The tip vortex trajectory separation is strictly related to the hub vortex deformation which moves from a line into a spiral. This behaviour is mainly due to mutual tip-hub vortex interaction as demonstrated by their simultaneous generation. The starting point of the instability could be the cross-blade wake interaction as proved by the fact that when reducing the loading conditions this interaction is shifted downstream as do the breakdown. CONCLUSIONS The PIV technique was used in a circulating water tunnel to investigate the spatial and temporal evolution of the wake of a four blade marine propeller in a uniform inflow. Both instantaneous and averaged velocity fields are achieved, the latter after phase sampling averaging over the same angular position of the propeller blade. The experimental results, in terms of velocity and vorticity fields, reveal some of the different contributions to the complex propeller flow field: the authoritative version for attribution.
lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line PROPELLER WAKE ANALYSIS BY MEANS OF PIV 500 1. The viscous part of the wake generated by the boundary layers on the blade surfaces. 2. The potential part of the wake deriving from the vortex sheet at the blade trailing edge. 3. The slipstream contraction of the wake along the downstream direction and the beginning of its broadening after the contraction 4. The cross-blade interaction which seems to be the starting point of the breakdown process. 5. The behaviour of the tip and hub vortex in the instability of the helical vortex system. The PIV technique, both for the careful control of the set-up adopted in the present experiment and for the image analysis algorithm implemented, allows a good spatial resolution (of the same order of magnitude as in LDV) to be obtained. PIV has proved to be a suitable means of investigating the complex flow field in the wake of a propeller giving additional and complementary information in comparison to the LDV technique. ACKNOWLEDGEMENTS. The authors are grateful to the CEIMM personnel and to Mr. Di Florio who supported the PIV measurements. This work was sponsored by Italian Ministero dei Trasporti e della Navigazione in the frame of INSEAN research plan 2000â 2002. REFERENCES Arndt, R., Maines, B. âViscous effects in tip vortex cavitation and nucleationâ, Proc. of the 20th Symposium on Naval Hydrodynamics, 1994. Biggers, J.C., Orloff, K.L., âMeasurements of the helicopter rotor-induced flow fieldâ, Journal of American Helicopter Society, Vol. 20, no. 1, 1975. Cenedese, A., Accardo, L., Milone, R. âPhase sampling techniques in the analysis of a propeller wakeâ, Proc. of the International Conference on Laser Anemometry Advances and Application, Manchester UK 1985. Chesnack C., Jessup S., (1998), Experimental characterisation of propeller tip flow, 22th Symposium on Naval Hydrodynamics, Washington D.C. Cotroni A., Di Felice, F., Romano, G.P., Elefante M. âPropeller Tip vortex Analysis by means PIVâ, 3rd International Workshop on PIV, Santa Barbara CA, (1999) Hunter, WW, Nichols, CE., âWind Tunnel Seeding Systems for Laser Velocimeters.â NASA Conference Publication 2393, Workshop, March 19â20, 1985, NASA Langley Research Jessup, S.D. âAn experimental investigation of viscous aspects of propeller blade flowâ. Ph.D. Thesis. 1989 The Catholic University of America, Washington D.C.. Keane R.D., Adrian R.J., âTheory of cross-correlation analysis of PIV imagesâ, Applied Scientific Research, Vol. 49, 1992, pp. 191â215. Keane R.D., Adrian, R.J., âOptimization of Particle Image Velocimeteresâ. Part 1: Double pulse system. Meas. Sci. Tech., 1, pp. 1202 1215, 1990 Kobayashi, S. âExperimental methods for the prediction of the effects of viscosity on propeller performanceâ. Dep. of Ocean Engineering, Rep. 81â7 MIT, 1981. Kobayashi, S., âPropeller wake survey by laser Doppler velocimeterâ. Proc. of the International Symposium on the Application of laser-Doppler Anemometry to Fluid mechanics, Lisbon 1982. Koyama, K., Kagugawa, A, Okamoto, M. âExperimental investigation of flow around a marine propeller and application of panel method to the propeller theoryâ, Proc. of the 16th Symposium on Naval Hydrodynamics, 1986. Landgrebe, A.J., Johnson, B.V., âMeasurement of model helicopter rotor flow velocities with a laser Doppler Velocimeterâ, Journal of American Helicopter Society, 1974. Vol. 19, no. 3. Melling A., âSeeding Gas Flows for Laser Anemometryâ. AGARD Conference on Advanced Instrumentation for Aero Engine Components, AGARD- CP 399, 1986. Min, K.S, âNumerical and experimental methods for prediction of field point velocities around propeller bladesâ. Dep. of Ocean Engineering, Report no. 78â12, MIT, 1978. Meyers, J.F., âGeneration of Particles and Seedingâ. Von Kárman Institute for Fluid Dynamics, Lecture Series 1991â05, Laser Velocimetry, Brussels, 1991 Raffel, M., Willert, C., Kompenhans, J., âParticle Image Velocimetryâ, Springer ISBN 3â540â63683â8, 1998 Serafini, J.S., Sullivan, J.P., Neumann, H.E. âLaser Doppler flow-field measurements of an advanced turbopropâ, 17th Joint Propulsion Conference, AIAA/SAE/ASME, Colorado Springs, Colorado, 1981. Stella, A., Guj, G., Di Felice, F., Elefante, M. âPropeller wake evolution analysis by LDVâ, Proc. of the 22nd Symposium on Naval Hydrodynamics, Washington, 1998. Wang, M.H. âHub effects in propeller design and analysisâ, Dep. of Ocean Engineering., Rep. 85â14, MIT, 1985. the authoritative version for attribution.
About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as the authoritative version for attribution. PROPELLER WAKE ANALYSIS BY MEANS OF PIV Figure 8: Longitudinal (U) and radial (V) component distribution for θ=0° Westerweel J., âFundamentals of Digital Particle Image Velocimetryâ, Meas. Science and Technology Vol. 8, pp. 1379â1392., 1997 Westerweel J., âEfficient detection of spurious vectors in particle image velocimetry dataâ, Experiments in Fluids, Vol. 16, pp. 236â247., 1994 501
About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as the authoritative version for attribution. PROPELLER WAKE ANALYSIS BY MEANS OF PIV Figure 9: Modulus of the inplane velocity with streamlines Figure 10: Non dimensional vorticity distribution and streamline in the frame moving with the upstream flow for θ=0° 502
About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as the authoritative version for attribution. PROPELLER WAKE ANALYSIS BY MEANS OF PIV Figure 11: Vorticity evolution for θ=20°.40°, 60°, 80° 503
About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as the authoritative version for attribution. PROPELLER WAKE ANALYSIS BY MEANS OF PIV Figure 13: Turbulence intensity distribution for =0° Figure 12: Mean tip vortex vorticity downstream evolution 504
About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as the authoritative version for attribution. PROPELLER WAKE ANALYSIS BY MEANS OF PIV Figure 14: ÏV distribution evolution at θ=20°, 40°, 60°, 80° 505
About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as the authoritative version for attribution. PROPELLER WAKE ANALYSIS BY MEANS OF PIV Figure 16: Tip vortex location at different revolution angle θ Figure 15: Tip vortex spatial fluctuation with respect the mean position 506
About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as the authoritative version for attribution. PROPELLER WAKE ANALYSIS BY MEANS OF PIV Figure 17: Flow visualitation in incipient cavitation of the propeller vortex system dispersion and hub vortex deformation 507
lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line PROPELLER WAKE ANALYSIS BY MEANS OF PIV 508 DISCUSSION R.Arndt University of Minnesota, USA I would assume that the propeller tip vortex interaction would be a function of advance coefficient. Is this true? Have you investigated vortex breakdown? AUTHOR'S REPLY The propeller tip vortex and wake blade interaction as well as the vortex breakdown are function of the advance coefficient as expected. In the present paper results regarding only the case of advance coefficient J=.74 has been presented but measurements for J=.88 and J=1.02 have also been performed. In figure is shown the comparison of the vorticity field for the above higher and lower advance coefficients for the same propeller revolution angle. At higher J due to the higher pitch of the wake, the interaction of the actual blade wake with the tip vortex of the previous one is shifted downstream. Furthermore at higher advance coefficients, a weaker wake and tip vortex, reduce the cross-interaction resulting in a more stable helical vortex system as observed in incipient cavitation flow visualization (Stella et al 2000) REFERENCES Stella, A., Guj, G:, Di Felice, F., Elefante, M. âPropeller wake analysis by LDV and Flow Visualitationsâ, September 2000, Journal of Ship Research, Vol 44, N 3, pp. 155â169 Figure: Comparison of the vorticity field at two different values of the advance coefficient J the authoritative version for attribution.
lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line PROPELLER WAKE ANALYSIS BY MEANS OF PIV 509 DISCUSSION G.Bark Chalmers University of Technology, Sweden First of all I would like to thank the authors for a very interesting paper demonstrating the potential of the PIV technique for the study of aspects of propeller wakes that can be of interest to understand propeller operation as well as cavitation. For understanding the cavitation process a thorough knowledge of the unsteady flow field of the non-cavitating flow is important also if the actual flow can be significantly influenced by the cavitation. The recording of the velocity as a function of time in a point is limited by the frame rate of the cross correlation camera, in this case 15 Hz. This frame rate is too small to generate a relevant time series at typical conditions. Thinking of the frame rates of high speed digital video cameras, being typically some thousands frames per second, my question is whether it is possible with present technology to increase the frame rate in the PIV-application and record a time series of significantly higher resolution in time? Or do we have to relay on the periodicity of the flow and construct the time series from measurements from different revolutions with different displacements relative the propeller? (Due to the tangentially induced velocity a material point will however move slowly out of the axial plane in which it was originally observed). For a steady homogeneous inflow the quasi steady method for tracing the vortex development downstream the blade is adequate. In the more or less unsteady wake of a ship, where the so called tip vortex bursting sometimes occurs in the cavitating vortex, it can however be of interest to follow the development of a particular vortex. Is a study of an unsteady motion of such a time scale possible with the needed requirements on the accuracy? With the present questions in mind, what is then the exact meaning of the term âinstantaneousâ in Figure 7? AUTHOR'S REPLY To reply to the interesting points raised by Prof. Bark, we would like to refer to the images given in the figure. They represent the near-wall region of a turbulent water channel flow acquired with a high-speed video camera (up to 2000 images/s) at the University âLa Sapienzaâ in Rome. A continuous sequence of up to 1000 images as those showed can be acquired and stored on a PC. At present, this device is used to acquire PIV images and to derive the time evolution of the whole PIV field in different flows (1). The information from these plots are used to detect flow structures (as that at x +â150÷220, y+â50 in the figure, where the+is used to indicate wall variables) and to follow their evolution and interactions. With a similar system it is possible to compute cross-correlation functions between successive frames and to obtain the velocity field with a rather high temporal resolution (similar to that obtained with an LDA system). Therefore, the answer is definitely yes, it is possible to achieve a resolution in time as high as 1 kHz. In this way, vortices can be followed also in unsteady and inhomogeneous flow conditions (as in the wake of a ship or propeller). It is not strictly necessary to average over different propeller revolutions, unless this was required by the user. Problems arise in respect to the spatial resolution of the system when high frame rates are required (as in the case of large flow velocity). It should also be noticed that, with the PIV measurement system employed in the paper, although the velocity time history is not derived, the velocity field is almost instantaneous. Indeed, the time interval between the two laser exposures (used to compute the cross-correlation function) is rather small (100 µs). The term âinstantaneousâ must be considered as an average over a time interval of such an order of magnitude. REFERENCES (1) Ciraolo G., Romano G.P. âInvestigation on wall structures in a turbulent channel flow using stereo PIVâ, Proceedings of the 8th European Turbulence Conference, Barcelona, 2000. the authoritative version for attribution.
About this PDF file: This new digital representation of the original work has been recomposed from XML files created from the original paper book, not from the original typesetting files. Page breaks are true to the original; line lengths, word breaks, heading styles, and other typesetting-specific formatting, however, cannot be retained, and some typographic errors may have been accidentally inserted. Please use the print version of this publication as the authoritative version for attribution. PROPELLER WAKE ANALYSIS BY MEANS OF PIV Figure. Velocity vectors in the near-wall region of a water channel recorded at time intervals equal to 1/250s. 510
