An experimental investigation of flow past a dual step cylinder

Flow development in the wake of a dual step cylinder has been investigated experimentally using Laser Doppler Velocimetry and flow visualization. The dual step cylinder model is comprised of a large diameter cylinder (D) mounted at the mid-span of a small diameter cylinder (d). The experiments have been performed for a Reynolds number (Re _D ) of 1,050, a diameter ratio (D/d) of 2, and a range of large cylinder aspect ratios (L/D). The results show that the flow development is highly dependent on L/D. The following four distinct flow regimes can be identified based on vortex dynamics in the wake of the large cylinder: (1) for L/D ≥ 15, three vortex shedding cells form in the wake of the large cylinder, one central cell bounded by two cells of lower frequency, (2) for 8 < L/D ≤ 14, a single vortex shedding cell forms in the wake of the large cylinder, (3) for 2 < L/D ≤ 6, vortex shedding from the large cylinder is highly three-dimensional. When spanwise vortices are shed, they deform substantially and attain a hairpin shape in the near wake, (4) for 0.2 ≤ L/D ≤ 1, the large cylinder induces vortex dislocations between small cylinder vortices. The results show that for Regimes I to III, on the average, the frequency of vortex shedding in the large cylinder wake decreases with L/D, which is accompanied by a decrease in coherence of the shed vortices. In Regime IV, small cylinder vortices connect across the large cylinder wake, but these connections are interrupted by vortex dislocations. With decreasing L/D, the frequency of dislocations decreases and the dominant frequency in the large cylinder wake increases toward the small cylinder shedding frequency.     

Visualisation and les simulation of cavitation cloud formation and collapse in an axisymmetric geometry

Visualisation and Large Eddy Simulations (LES) of cavitation inside the apparatus previously developed by Franc (2011) for surface erosion acceleration tests and material response monitoring are presented. The experimental flow configuration is a steady-state closed loop flow circuit where pressurised water, flowing through a cylindrical feed nozzle, is forced to turn 90° and then, move radially between two flat plates towards the exit of the device. High speed images show that cavitation is forming at the round exit of the feed nozzle. The cavitation cloud then grows in the radial direction until it reaches a maximum distance where it collapses. Due to the complexity of the flow field, direct observation of the flow structures was not possible, however vortex shedding is inferred from relevant simulations performed for the same conditions. Despite the axisymmetric geometry utilized, instantaneous pictures of cavitation indicate variations in the circumferential direction. Image post-processing has been used to characterize in more detail the phenomenon. In particular, the mean cavitation appearance and the cavity length have been estimated, showing good correlation with the erosion zone. This also coincides with the locations of the maximum values of the standard deviation of cavitation presence. The dominant frequency of the ‘large-scale’ cavitation clouds has been estimated through FFT. Cloud collapse frequencies vary almost linearly between 200 and 2000 Hz as function of the cavitation number and the downstream pressure. It seems that the increase of the Reynolds number leads to a reduction of the collapse frequency; it is believed that this effect is due to the agglomeration of vortex cavities, which causes a decrease of the apparent frequency. The results presented here can be utilized for validation of relevant cavitation erosion models which are currently under development.     

Statistical Properties of Turbulent Free Jets Issuing from Nine Differently-Shaped Nozzles

This paper reports the centerline evolutions of turbulent statistical properties in nine air jets issuing from differently-shaped nozzles into still air surroundings. All nozzles of investigation have nominally identical opening areas or equal equivalent diameters (D _e ) and their aspect ratio (AR) varies from AR = 1 (circle) to AR = 2.5 (isosceles triangle). Present measurements were made at the Reynolds number (based on D _e ) of approximately 15,000. Results show that the loss of jet-axisymmetry at the exit generally causes the mean velocity decaying faster, and the fluctuating intensity growing, in the near field, thus indicating the increased overall entrainment rate. It is also shown that a change of shape of the nozzle exit does not affect the asymptotic decay rate of the centreline velocity in the far field. The near-field structure of the isosceles-triangular jet is deduced to be most three-dimensional, compared with the circular counterpart from smooth contraction being least. These discrepancies, however, weaken as the downstream distance x is increased. Beyond x/D _e  = 20–30, the normalized velocity spectra for all jets of small AR collapse well, indicating similar statistical behaviors over a wide range of scales in the central region. Indeed, sufficiently downstream from the exit, insignificant differences occur in jets’ velocity probability density function (PDF), the related skewness and flatness factors, and also in their Taylor and Kolmogorov microscales. It is demonstrated that all the length scales grow approximately linearly with x at x/D _e  ≥ 20.     

Experimental investigation of added mass effects on a hydrofoil under cavitation conditions

The influence of leading edge sheet cavitation and supercavitation on the added mass effects experienced by a 2-D NACA0009 truncated hydrofoil has been experimentally investigated in a hydrodynamic tunnel. A non-intrusive excitation and measuring system based on piezoelectric patches mounted on the hydrofoil surface was used to determine the natural frequencies of the fluid–structure system. The appropriate hydrodynamic conditions were selected to generate a range of stable partial cavities of various sizes and also to minimize the effects of other sources of flow noise and vibrations. The main tests were performed for different sigma values under a constant flow velocity of 14 m/s and for incident angles of both 1° and 2°. Additionally, a series of experiments in which the hydrofoil was submerged in air, partially and completely submerged in still water and without cavitation at 7 and 14 m/s were also performed. The maximum added mass effect occurs with still water. When cavitation appears, the added mass decreases because the cavity length is increased, and the added mass is minimum for supercavitation. A linear correlation is found between the added mass coefficients and the entrained mass that accounts for the mean density of the cavity, its dimensions and its location relative to the specific mode shape deformation.     

Numerical research of diesel spray and atomization coupled cavitation by Large Eddy Simulation (LES) under high injection pressure

A special spray model is applied to study the spray behavior with high injection pressure and micro-hole nozzle. To reveal the cavitation in diesel nozzle and its influence on spray and atomization, the Large Eddy Simulation (LES) turbulence model is adopted to detect the cavitation, and then the special spray model coupling the cavitation is build. From research results, three important conclusions can be drawn. Firstly, the cavitation flow can raise the effective velocity at the nozzle exit and such effect become even more obvious with higher injection pressure, e.g.180 MPa. Secondly, the applied spray model is in good agreement with the spray characteristics and images obtained from the EFS8400 spray test platform. Thirdly, the cavitation with high injection pressure and micro-hole nozzle can increase the spray cone angle and reduce the spray penetration; the cavitation intensity has a great impact on the spray velocity field and vorticity intensity, especially at the initial spray field under the condition of high injection pressure.     

The effect of chamber length and Reynolds number on jet precession

The jet axial velocity field exiting from a nozzle/chamber configuration with an expansion ratio of 5 is investigated using Stereo-PIV for a range of chamber lengths and Reynolds (Re) numbers. The jet can exit the chamber in axial jet (AJ) mode with the maximum velocity near the chamber axis or precessing jet (PJ) mode with the maximum velocity near the chamber wall and rotating or precessing about the chamber axis. Algorithms were developed to determine the jet mode from exit conditions and allow conditional averaging of the velocity field in PJ mode. The probability of the jet in PJ mode was found to be a strong function of chamber length, L/D and only a mild function of Re for Re > 10,000. High precession probability was found for chambers of length in the range 2 < L/D < 2.75 for all cases for Re > 10,000. An abrupt reduction in precession probability occurred for chamber lengths L/D~3. For increasing chamber lengths, an increase in precession probability was observed. The ratio of entrainment-into-the-chamber of surrounding fluid to jet exit fluid was found not to be a function of Re or jet mode (AJ or PJ) but only a function of L/D. A maximum ratio entrainment-into-the-chamber was observed to occur in the range 2 < L/D < 2.5. Conditionally averaged velocity profiles also showed the exiting jet to be a strong function of L/D and with only a mild effect of Re for all cases of Re > 10,000.     

Aerodynamics of intermittent bounds in flying birds

Flap-bounding is a common flight style in small birds in which flapping phases alternate with flexed-wing bounds. Body lift is predicted to be essential to making this flight style an aerodynamically attractive flight strategy. To elucidate the contributions of the body and tail to lift and drag during the flexed-wing bound phase, we used particle image velocimetry (PIV) and measured properties of the wake of zebra finch (Taeniopygia guttata, N = 5), flying at 6–10 m s⁻¹ in a variable speed wind tunnel as well as flow around taxidermically prepared specimens (N = 4) mounted on a sting instrumented with force transducers. For the specimens, we varied air velocity from 2 to 12 m s⁻¹ and body angle from −15° to 50°. The wake of bounding birds and mounted specimens consisted of a pair of counter-rotating vortices shed into the wake from the tail, with induced downwash in the sagittal plane and upwash in parasagittal planes lateral to the bird. This wake structure was present even when the tail was entirely removed. We observed good agreement between force measures derived from PIV and force transducers over the range of body angles typically used by zebra finch during forward flight. Body lift:drag (L:D) ratios averaged 1.4 in live birds and varied between 1 and 1.5 in specimens at body angles from 10° to 30°. Peak (L:D) ratio was the same in live birds and specimens (1.5) and was exhibited in specimens at body angles of 15° or 20°, consistent with the lower end of body angles utilized during bounds. Increasing flight velocity in live birds caused a decrease in C _L and C _D from maximum values of 1.19 and 0.95 during flight at 6 m s⁻¹ to minimum values of 0.70 and 0.54 during flight at 10 m s⁻¹. Consistent with delta-wing theory as applied to birds with a graduated-tail shape, trimming the tail to 0 and 50% of normal length reduced L:D ratios and extending tail length to 150% of normal increased L:D ratio. As downward induced velocity is present in the sagittal plane during upstroke of flapping flight, we hypothesize that body lift is produced during flapping phases. Future efforts to model the mechanics of intermittent flight should take into account that flap-bounding birds may support up to 20% of their weight even with their wings fully flexed.     

Effect of self-issuing jets along the span on the near-wake of a square cylinder

The study herein focuses on the vortex shedding characteristics and near-wake vorticity patterns of a square cylinder having self-issuing jets through holes along its span. Three different values of spacing between the consecutive holes λ with respect to the cylinder diameter D, i.e., λ/D = 1.5, 3 and 4 are studied experimentally via Digital Particle Image Velocimetry for the Reynolds number range extending from 200 to 1,000. It has been observed that the three-dimensionality of the wake flow depends on the spacing between the holes and Re number. For sufficiently low Reynolds numbers, the jet flows issuing from the holes yield a non-uniform distribution of mean flow characteristics like the shedding frequency and the formation length of vortices along the span of the cylinder when the spacing between jets along centerline is close to wavelength of the naturally existing three-dimensional wake instability. Additionally, for Re number up to 500, the self-issuing jets emanating from the holes show an indirect interaction with shear layers originating from upper and lower separation lines of the cylinder. However, for higher Re numbers of 750 and 1,000, they directly interact with and modify the vortices forming from the cylinder.     

Interaction of circular cylinder wake with a short asymmetrically located downstream plate

This study reveals the interaction patterns of separated shear layers from a circular cylinder with a short downstream plate and their reflection on the frequency and the formation length of the vortices from the cylinder as a function of plate location relative to the cylinder. The effect of horizontal (G/D) and vertical (Z/D) distances between the cylinder and the plate on the near wake is studied via Digital Particle Image Velocimetry (DPIV) in a water channel for Reynolds numbers of 200, 400 and 750, based on the cylinder diameter D. It is shown that the interaction of wake with the plate of length D can be categorized depending on the horizontal and the vertical distances between the cylinder and the plate. For the vertical distance range of Z/D ≤ 0.7, there is a critical horizontal spacing before which the shear layers from the cylinder are inhibited to form vortices in front of the plate. Resulting elongated recirculation region between the plate and the cylinder suggests modification of the absolutely unstable near wake of free circular cylinder in favor of convective instability. Z/D = 0.9 provides a passage from Z/D ≤ 0.7 to ≥1.1 and is associated with a dominant effect on the near-wake characteristics of interaction of shear layers from the cylinder with those from the downstream plate. For Z/D ≥ 1.1, there is again, yet a smaller critical horizontal spacing after which vortices interact with decreased downstream plate interference. In this vertical separation distance range, a gap flow between the plate and the cylinder plays a determining role on the formation length and St number of vortices for small horizontal spacing values.     

Coefficient of discharge and spray cone angle of a pressure nozzle with combined axial and tangential entry of power-law fluids

Flows of incompressible, time-independent purely viscous power-law fluids through pressure nozzle with combined axial and tangential entry are analysed. Theoretical predictions of coefficient of discharge and spray cone angle are made through an approximate analytical solution of hydrodynamics of flow inside the nozzle. In the converging section of the nozzle, the boundary layer equations have been derived with modified order approximation [O(δ/R)≈1, O(δ ²/R ²)≪1] of Navier-Stokes equations for a better accuracy. Smoother attainment of the free-stream condition at the edge of the boundary layer is ensured by requiring the appropriate shear rate terms, compatible with the above order analysis, to be zero. The pertinent independent input parameters which govern the flow field are the generalized Reynolds number at inlet to the nozzle based on the tangential velocity of injection , the ratio of the axial-to-tangential velocity at the inlet to the nozzle V _R , the flow behaviour index of the fluid n, the length-to-diameter ratio of the swirl chamber L ₁/D ₁, the spin chamber angle 2α and the orifice-to-swirl-chamber-diameter ratio D ₂/D ₁. Experiments reported in the paper corroborate the qualitative trends of analytical results.     

Experimental study on turbulent mixing of spray droplets in crossflow

Centrifugal spray injected at various angles in gas crossflow has been studied experimentally using PIV visualization system and image-processing techniques. Experiments were carried out inside a rectangular duct (95 mm × 95 mm in cross-section) at ambient temperature and pressure, with different gas Reynolds numbers (vary from 12,900 to 45,000) and three injection angles (60°, 90° and 120°). The spray angle of the centrifugal nozzle is 80°, with D₃₂ of 80 μm. The instantaneous images of droplets distribution and the values of the line-averaged D₃₂ at different positions on the cross-sections along the flow field for each condition were obtained, and their flow field configurations were achieved. Quantitative assessments of mixing degree between two phases for different injection angles were determined using a spatial unmixedness parameter. It is found that the addition of droplets into the gas crossflow enhanced the turbulence intensity of the gas crossflow and caused different-scale vortices. The flow field structure, to a great extent, is dependent on the injection angle. The entrainment and centrifugal force of large vortex lead to uneven droplet distribution and moreover influence the mixing of droplets and gas crossflow. A better mixing result can be obtained with the injection angle of nozzles of 60°.     

Volumetric velocity measurements of vortex rings from inclined exits

Vortex rings were generated by driving pistons within circular cylinders of inner diameter D = 72.8 mm at a constant velocity U ₀ over a distance L = D. The Reynolds number, U ₀ L/(2ν), was 2500. The flow downstream of circular and inclined exits was examined using volumetric 3-component velocimetry (V3V). The circular exit yields a standard primary vortex ring that propagates downstream at a constant velocity and a lingering trailing ring of opposite sign associated with the stopping of the piston. By contrast, the inclined nozzle yields a much more complicated structure. The data suggest that a tilted primary vortex ring interacts with two trailing rings; one associated with the stopping of the piston, and the other associated with the asymmetry of the cylinder exit. The two trailing ring structures, which initially have circulation of opposite sign, intertwine and are distorted and drawn through the center of the primary ring. This behavior was observed for two inclination angles. Increased inclination was associated with stronger interactions between the primary and trailing vortices as well as earlier breakdown.     

Effect of an Exit-Wedge Angle on Pinch-off and Mass Entrainment of Vortex Rings in Air

The effects of exit-wedge angle on evolution, formation, pinch-off, propagation and diffusive mass entrainment of vortex rings in air were studied using digital particle image velocimetry. Vortex rings were generated by passing a solenoid-valve-controlled air jet through a cylindrical nozzle. Experiments were performed over a wide range of exit-wedge angles (10° ≤ α ≤ 90°) of the cylindrical nozzle, initial Reynolds numbers (450 ≤ Re ≤ 4,580) and length-to-diameter ratios (0.9 ≤ L/D ≤ 11) of the air jet. For sharp edges (α ≤ 10°), a secondary ring may emerge at high Reynolds numbers, which tended to distort the vortex ring if ingested into it. For blunt edges (α ≥ 45°), by contrast, stable vortex rings were produced. The formation phase of a vortex ring was found to be closely related to its evolution pattern. An exit-wedge angle of 45° was found to be optimal for rapid pinch-off and faster propagation and better stability of a vortex ring. Diffusive mass entrainment was found to be between 35% and 40% in the early stages of a vortex ring propagation and it gradually increased throughout the course of vortex ring propagation. Entrainment fraction was found to be sensitive to the L/D ratio of the initial jet and decreases when the L/D ratio is increased.     

High speed digital imaging of cavitating vortices

 Researchers at the Cavitation and Multiphase Flow Laboratory of the University of Michigan worked in conjunction with Princeton Scientific Instruments (PSI) engineers to employ a new digital imaging system in the study of partial attached cavitation. The new high speed solid state system, the Princeton Scientific Ultra Fast Framing Camera (UFFC), was designed for cavitation studies where framing rates of 10⁵–10⁶ frames/s are required to image the detailed mechanisms of cavitating flows. The UFFC, which uses a PSI patented Charge Coupled Device (CCD) array image sensor, was designed to capture 30 frames at a maximum framing rate of 1 million frames/second. In these experiments, a maximum framing rate of 125000 frames per second (8 μs/frame) was used to examine cavitating vortices in the closure region of a partial attached cavity. The vortical structures in the closure region of the attached cavity were imaged, and the evolution and collapse of these flow structures were examined. Relationships between the cavitating vortices size, strength, and collapse time were observed. Received: 15 November 1996/Accepted: 1 December 1997     

单液滴内空化气泡的生长及溃灭研究

超空化燃油射流使得喷雾中部分燃油分裂液滴内含有空化气泡;空化气泡的生长及溃灭对液滴的分裂与雾化具有重要影响. 基于VOF 方法首次对超空化条件下燃油液滴内空化气泡的生长及溃灭过程进行了数值模拟. 通过研究发现,单液滴内空化气泡的生长过程可以按控制机理划分为表面张力控制阶段、综合竞争阶段和惯性力控制阶段;在第I 阶段,空泡的生长主要受表面张力的控制作用,惯性力对空泡生长的促进作用及黏性力对空泡生长的抑制作用可以忽略;在第II 阶段,空泡的生长受表面张力、惯性力及黏性力三者的综合作用,空泡的生长速率是促进空泡生长的惯性力和抑制空泡生长的表面张力及黏性力相互竞争、共同作用的结果;在第III 阶段,空泡的生长主要受惯性力的控制作用,抑制空泡生长的表面张力及黏性力的作用基本可以忽略. 单液滴内空化气泡的溃灭过程由多个溃灭阶段和反弹阶段构成,类似于有阻尼弹簧振子的振动过程;根据每个溃灭周期结束时空泡半径随时间的变化历程,可以将空泡的溃灭分为快速溃灭期、缓慢溃灭期以及稳定期;溃灭初期空泡溃灭压力的变化非常剧烈,但空泡溃灭体积的变化则要相对平缓得多;空泡反弹压力随时间的变化与空泡反弹体积随时间的变化基本对应.      

收缩扩张管内液氮空化流动演化过程试验研究

本文基于低温空化试验平台研究了收缩扩张流道内液氮非定常空化流动的演化过程. 试验采用高时空分辨率的高速摄像机对77 K液氮在不同空化数σ下空穴结构的演变进行了精细化的分析和研究. 利用试验得到的空穴长度和面积等数据, 定量分析了液氮空化流动的非定常特性与时空演变规律. 研究结果表明: (1)在相似来流速度和温度条件下, 随着空化数的减小, 液氮空化流动呈现四种典型流型, 空穴长度在2.5 h以内为初生空化、空穴长度在2.5 h ~ 7.5 h之间为片状空化、空穴长度在7.5 h ~ 15 h之间为大尺度云状空化, 空穴长度超过15 h为双云状空化, 且在大尺度云状空化和双云状空化阶段均捕捉到了回射流现象; (2)液氮空化流动从初生空化到双云状空化, 脱落空穴的尺度逐渐增大, 空穴面积脉动的幅值和准周期均有所增加. 同时, 在大尺度云状空化与双云状空化阶段, 喉口处堵塞效应对空化流动的影响显著增强; (3)相比于初生空化, 片状空化、大尺度云状空化以及双云状空化中脱落空穴的移动距离依次增加了0.97倍、2.65倍与2.68倍, 溃灭时间依次增加了1.18倍、3.59倍与4.47倍, 但溃灭速度依次减小了0.10倍、0.20倍与0.30倍. 除此之外, 对于双云状空化阶段, 存在两种显著不同的脱落空穴演化过程.      

Measurements of heat transfer coefficients between turbulent liquid jets and a radially finned heated disk placed at the bottom of an open vertical circular cavity

 Experiments have been performed to assess the impact of an extended surface on the heat transfer enhancement for axisymmetric, turbulent liquid jet impingement on a heated round disk. The disk, with an array of integral radial fins mounted on its surface, is placed at the bottom of an open vertical circular cavity. Hydrodynamic and heat transfer data were obtained for a dielectric fluorocarbon liquid FC-77. For a fixed circular heater of diameter D=22.23 mm, several geometric parameters were tested: the nozzle diameter (4.42≤d≤9.27 mm), the confining wall diameter of the vertical cavity (22.23≤D _c≤30.16 mm), and the nozzle-to-heater spacing (0.5≤S/d≤5.0). The FC-77 flow rates varied from =0.2 to 11.0 l/min producing Reynolds numbers in the wide interval 700≤Re _d ≤44,000. For d=4.42 mm, the heat transfer response to the separation distance S/d was small but increased gradually with increasing nozzle diameter up to d=9.27 mm. The thermal resistance R _th increased with the confining wall diameter D _c and also with the nozzle diameter d. A minimum value of the thermal resistance of R _th,min=0.4 cm² K/W was attained for a combination of d=4.42 mm, D _c=22.23 mm, S/d=1, and =7.5 l/min. Based on a simplified heat transfer model, reasonable agreement was obtained between measured values of the thermal resistance and the R _th-predictions. The total fin effectiveness ɛ_f was shown to increase with increasing nozzle diameter, but was invariant with the flow rate (or the jet exit velocity). More than a three-fold heat transfer enhancement was realized through the addition of the array of integral radial fins on the heated round disk. Received on 30 August 2000 / Published online: 29 November 2001     

The spray characteristics of a pressure-swirl injector with various exit plane tilts

The gasoline spray characteristics of a pressure-swirl injector were investigated with various exit plane tilts. The analysis focused on the correlation between tilt angle and flow angle. Mie-scattering technique and phase Doppler anemometry were employed to analyze the macroscopic spray development and droplet size distribution of the spray. An analytical method for mass flux estimation was applied to understand the velocity distribution at the nozzle exit. The results showed that the spray shape and velocity distribution of the spray were more asymmetrical at high tilt angles. In particular, an opened hollow cone spray was formed when the tilt angle is greater than the complementary flow angle. The pressure drop inside the spray, one of the crucial factors for the swirl spray collapse at various surrounding conditions, was attenuated in this opened hollow cone spray since the pressure inside the spray was assimilated to the surrounding air pressure. The spray collapse at high fuel temperature and back pressure conditions did not appear when the tilt angle is larger than the complementary flow angle due to the reduced pressure drop inside the spray. However, tilt angle should be optimized to fulfill the requirements of spray robustness and avoid the locally rich area. The droplet size of 70° tilted nozzle spray shows a value similar to that of the original swirl spray in the plane that includes nozzle axis and the major axis of exit surface ellipse (Major axis plane) while it shows an increased value in the plane that includes nozzle axis and the minor axis of exit surface ellipse (Minor axis plane).     

Fluid-dynamic forces and wake frequencies on a tilted rectangular cylinder near a solid wall

We investigate the impact of different boundary conditions on the flow field developing around a tilted rectangular cylinder with two different values of the aspect ratio (l/s=3 and 4). We are mainly interested in analyzing the changes in force coefficients and in the vortex shedding Strouhal number when the cylinder is placed at various distances from a bottom wall and different values of attack angle. The angle of attack ranges between −30° and +30° and the cylinder elevation above the bottom wall is varied between almost zero and 5 times the thickness of the cylinder. A large body of experimental results is related to the small elevation conditions at different attack angles, where the presence of the wall has a non-negligible effect on the behavior of the force coefficients and Strouhal number of the vortex shedding.     

Numerical analysis of unsteady cavitating turbulent flow and shedding horse-shoe vortex structure around a twisted hydrofoil

Cavitating turbulent flow around hydrofoils was simulated using the Partially-Averaged Navier–Stokes (PANS) method and a mass transfer cavitation model with the maximum density ratio (ρ_l/ρ_v,clip) effect between the liquid and the vapor. The predicted cavity length and thickness of stable cavities as well as the pressure distribution along the suction surface of a NACA66(MOD) hydrofoil compare well with experimental data when using the actual maximum density ratio (ρ_l/ρ_v,clip = 43391) at room temperature. The unsteady cavitation patterns and their evolution around a Delft twisted hydrofoil were then simulated. The numerical results indicate that the cavity volume fluctuates dramatically as the cavitating flow develops with cavity growth, destabilization, and collapse. The predicted three dimensional cavity structures due to the variation of attack angle in the span-wise direction and the shedding cycle as well as its frequency agree fairly well with experimental observations. The distinct side-lobes of the attached cavity and the shedding U-shaped horse-shoe vortex are well captured. Furthermore, it is shown that the shedding horse-shoe vortex includes a primary U-shaped vapor cloud and two secondary U-shaped vapor clouds originating from the primary shedding at the cavity center and the secondary shedding at both cavity sides. The primary shedding is related to the collision of a radially-diverging re-entrant jet and the attached cavity surface, while the secondary shedding is due to the collision of side-entrant jets and the radially-diverging re-entrant jet. The local flow fields show that the interaction between the circulating flow and the shedding vapor cloud may be the main mechanism producing the cavitating horse-shoe vortex. Two side views described by iso-surfaces of the vapor volume fraction for a 10% vapor volume, and a non-dimensional Q-criterion equal to 200 are used to illustrate the formation, roll-up and transport of the shedding horse-shoe vortex. The predicted height of the shedding horse-shoe vortex increases as the vortex moves downstream. It is shown that the shape of the horse-shoe vortex for the non-dimensional Q-criterion is more complicated than that of the 10% vapor fraction iso-surface and is more consistent with the experiments. Further, though the time-averaged lift coefficient predicted by the PANS calculation is about 12% lower than the experimental value, it is better than other predictions based on RANS solvers.