URANS Simulations of Cavitation and Hull Pressure Fluctuation for Marine Propeller with Hull Interaction

Transcription

1 Third International Syposiu on Marine Propulsors sp 1, Launceston, Tasania, Australia, May 201 UANS Siulations of Caitation and Hull Pressure Fluctuation for Marine Propeller with Hull Interaction Kwang-Jun Paik, Hyung-Gil Park, Jongsoo Seo Sasung Ship Model Basin (SSMB), Sasung Heay Industries Co., Ltd., Daejeon, Korea ABSTACT Siulations of caitation flow and hull pressure fluctuation for a arine propeller operating behind a hull using the unsteady eynolds-aeraged Naier-Stockes (UANS) are presented. A full hull body suberged under the free surface is odeled in the coputational doain to siulate directly the wake field of the ship at the propeller plane. A syetry boundary condition is applied on the top of coputational doain instead of the free surface, which is siilar to a caitation tunnel. To obsere the caitation pattern during the reolution of propeller, the propeller is rotated with a constant rotation angle of 1.5 using a sliding esh technique. Siulations are perfored in design and ballast draught conditions to study the effect of caitation nuber. And two propellers with slightly different geoetry are siulated to alidate the detectability of the nuerical siulation. All siulations are perfored using a coercial CFD software FLUENT. Caitation patterns of the siulations are copared with the experiental results carried out in Sasung CAitation Tunnel (SCAT). Siulation results for the hull pressure fluctuation induced by a propeller are also copared with the experiental results with good agreeent. Keywords Propeller, Caitation, Hull pressure fluctuation, UANS, Propeller hull interaction 1 INTODUCTION The practical iportance of accurate nuerical siulation to predict the caitation pattern and hull pressure aplitude has been increased in the field of ship building industries. The researches to siulate the caitation flow using ANS soler were perfored actiely in the last seeral years because caitation odels for ANS soler and coputation power hae been rapidly deeloped in the last decade. Watanabe et al. (200) started the study on the caitation flow for a arine propeller using ANS. Salatore et al. (2009) copared seen coputational odels using ANS, LES, and BEM for the INSEAN E779A propeller in unifor flow and wake field. The caitation siulation for conentional and highly-skewed propellers in the behind-hull condition using an in-house ANS soler was perfored by Shin et al. (2011). Hasuike et al. (2010) siulated the caitation flow and showed a possibility to predict the caitation erosion risk with the indexes using tie differential of caity oid fraction or pressure. On the other hand, the usage of coercial CFD software was gradually increased to siulate the caitation flow for arine propellers. Boorsa & Whitworth (2011) used STA-CCM+ to study on the prediction of caitation and erosion for a propeller and rudder. Bertetta et al. (2011) copared the results of ANS soler using STA-CCM+ with those of potential soler for the caitation flow of a controllable pitch propeller. Morgut & Nobile (2011) studied on the perforance of three caitation odels for a arine propeller using ANSYS-CFX. Li & Terwisga (2011) used FLUENT to inestigate unsteady caitation phenoena for 2D and D foils. Kawaura (2010) assessed caitation erosion risk based on pressure ipacts and siulated the caitation flow and the agnitude of pressure aplitude for a arine propeller using FLUENT. Most of siulations except for Kawaura (2010) prescribed a elocity distribution at the inlet boundary of the coputational doain to achiee the correct wake field at the propeller plane. In that case, the adjustent of inlet boundary condition to get a desired elocity contour at propeller plane is of priary iportance and ery laborious work. Therefore, Kawaura (2010) used a full hull body to generate the wake field. Neertheless, the caity extent was under-predicted and the pressure aplitudes were about 70% of the experiental results. In this paper the results of nuerical siulations using UANS for the caitation pattern and hull pressure fluctuation of arine propellers are presented. A coercial CFD software FLUENT ersion 14.0 is used for the nuerical siulations. Schneer & Sauer (2001) caitation odel is applied to siulate the caitation flow. A coputational doain including a full hull body suberged under the free surface is used to siulate directly the wake field at the propeller plane. Two kinds of studies arying caitation nuber and propeller geoetry are executed in order to alidate the nuerical siulation ethod. And the nuerical siulations are 89

2 perfored at the sae as the test conditions in the caitation tunnel and copared with the experiental results. 2 NUMEICAL METHODS AND MODELS A coercial CFD software FLUENT ersion 14.0 was used for the nuerical siulations, in which the caitation flow was soled by a ixture odel based on a single-fluid ultiphase ixture flow approach. 2.1 Goerning Equations In the ixture odel, the continuity equation and the oentu equation becoe as t t 0 (1) p T g F. (2) The ixture density and iscosity coefficient are defined as (1 ) (1 ) where is the apor olue fraction. Subscripts,, and l represent ixture, apor, and liquid phase, respectiely. 2.2 Caitation Model Schneer & Sauer (2001) caitation odel applied in this research soles the apor olue fraction with following transport equation: t where the ters, e c l l () (4) and e c, account for the ass transfer between the liquid and apor phase in caitation denoting the eaporation and condensation of the apor bubbles. The fors of and are written as follows: when when p p, e l p p, c l e c 1 (5) B 2 p p 1. (6) B 2 l p p l where n b is the bubble nuber density. MODEL TEST Model tests to obsere caitation flow on a propeller blade and to easure pressure fluctuation on a hull surface induced by the propeller caitation were carried out in Sasung CAitation Tunnel (SCAT). The principal particulars of the test section of SCAT are suarized in Table 1 and the set-up of a odel ship and propeller is shown in Figure 1. The definition of propeller blade angle is depicted in Figure 2. The propeller blade angle begins fro the top position toward propeller rotation direction. Pressure transducers to easure the pressure fluctuation on a hull surface were installed on the top of propeller with the interals of 0 as illustrated in Figure. Table 1: Principal particulars of the test section of SCAT Ite Diension of Test Section (L B D) Maxiu Speed Value Contraction atio /s Figure 1: Set-up of odel ship and propeller in SCAT Here, the bubble radius, B, is expressed as B nb 1 (7) Figure 2: Definition of propeller blade angle 90

3 The length of the odel ship, a crude oil tanker, was about 7.6/s and the diaeter of the odel propeller was in this research. The blockage of the test section due to the odel ship was about 1.2%. The surface of the odel propeller was coated with a rough paint to stabilize the behaior of caitation flow. The inflow elocity of odel test was fixed as 5.5/s. The propeller rotation speed was set with the thrust identity ethod based on the results of self-propulsion test in a towing tank. Figure 4: Outline of grid syste and boundary conditions for coputational doain and propeller block Figure : Position of pressure transducers installed on a odel ship 4 NUMEICAL SIMULATION 4.1 Grid Syste To siulate the propeller caitation flow under siilar situation to the odel test, a full hull body suberged under the free surface was odeled in the coputational doain. The free surface was substituted with a syetry boundary condition because the ixture odel for the caitation flow and the Volue of Fluid (VOF) odel for the free surface could not be ipleented siultaneously in FLUENT. A sliding block surrounding the propeller, coposed of unstructured girds, was applied to ipleent the effect of propeller rotation. Pyraid cells were used for the surface of propeller and boundaries, and tetrahedral cells were filled in the block. On the other hand, structure girds were applied to the other doains expect for the sliding block. Figure 4 shows the outline of grid syste including boundary conditions. The grid sizes were 1.4M and 2.6M cells for the sliding block and the other blocks, respectiely. 4.2 Siulation Method As the nuerical odel for caitation flow, Schneer & Sauer (2001) was applied in this research. Bubble nuber density was set as 1e+15. To iproe the conergency of caitation flow and reduce the coputational tie, Multiple eference Frae (MF) was used at first and Sliding Mesh Model (SMM) was then applied to siulate the rotation of propeller. Propeller rotated with a constant tie step corresponding to the rotating angle of 1.5. Tie histories of fluctuating pressure on the hull surface were recorded to copare with the experiental results at the sae positions as the pressure transducers installed on the odel test. The siulation ethods applied in this research are suarized in Table 2. Table 2: Suary of siulation ethods Ite Goerning Equation Turbulence Model Caitation Model P-V Coupling Pressure Soler Moentu Soler Method ANS SM Mixture SIMPLEC Standard Second Order Upwind 91

4 A propeller coated with the rough paint tended to increase the caity extent and pressure pulse as copared with a sooth surface propeller in the experient. Howeer, since it was ipossible that the effect of the paint was ipleented in nuerical siulation, a saller caitation nuber was used for the siulation instead of the rough paint. The adjustent was approached with ignoring a wae height at stern which was considered in the odel test. 5 ESULTS AND DISCUSSION In this research two kinds of siulations were executed and copared with the experiental results. The first siulation was perfored at two draught conditions to inestigate the effect of caitation nuber. The other study was to alidate the detectability of the nuerical siulation about the differences of caitation pattern and pressure aplitude fro the sall odification of propeller geoetry. Prior to the caitation siulation, the bare hull without the propeller was siulated to copare the wake field at the propeller plane. Figure 5 shows the coparison of elocity contours and ectors between the experiental data and siulation result at the propeller plane. Inlet elocity of the nuerical siulation was 5.5/s, which was the sae elocity as the experient. The elocity contour and ector are noralized by the inlet elocity. The pattern of the wake fields is siilar to each other. Howeer, the axial elocity of the experient at outer radii of lower part is about 0.1 higher than the siulation because the elocity at lower part is probably accelerated due to the blockage effect at the experient. The hook shape ortex generated by bilge ortex is appeared around 20 in the experient, while it is not obsered in the siulation. 5.1 Draught Condition The siulations were perfored at design and ballast draught conditions, and the siulation results were copared with the experiental results of SCAT. The odel propeller, tested in the odel test, is a 4-blade propeller (Prop-A) designed for a crude oil tanker. Mean pitch ratio and expanded area ratio of the propeller are and 0.480, respectiely. The siulation conditions are suarized in Table. The caitation nuber is defined as p p (8) n D where p is a static pressure at the 70% of propeller radius aboe the propeller center and p is a apor pressure. n and D are a propeller rotation speed and a propeller diaeter, respectiely. The thrusts of the siulations at the design and ballast conditions were about 4% higher than those of the experients. The larger thrusts of the siulations result fro the slower axial elocity at the propeller plane as copared in Figure 5. Caitation patterns at the propeller blade angle of 0, 20, and 40 at the design and ballast draught conditions are copared with the experiental results in Figures 6 and 7. The caity extent in the siulation are expressed with the isosurface of =0.1. The siulation results hae soe liitations to capture the tip ortex caitation, but the caity patterns generally well agree with the experients. At the design draught condition, the leading edge caitation near 0.8 of the propeller blade is not appeared in the nuerical siulation, while it is obsered with ery unstable behaior in the experient. At the ballast draught condition, the caity extents at all blade angles well correspond with the experients. The sheet caitation was not appeared in the range between 90 and 0 in the experients, but it was obsered until about 240 in the siulations due to the slower axial elocity than the experient as shown in Figure 5. Figure 5: Velocity contours and ectors of experient (top) and siulation (botto) at propeller plane (dashed circle: outline of propeller disk) Table : Suary of siulation conditions to study the effect of draught condition Inflow Speed Propeller Diaeter Design 5.5 /s Ballast Propeller Speed 8.0 rps 9. rps Caitation Nuber

5 and ballast draught conditions are plotted in Figure 9. Hull pressure pulse is expressed with the pressure pulse coefficient defined as p K P n (9) 2 D 2 p is the pressure fluctuation aplitude fro the where ean pressure. The pressure fluctuation at the design draught condition has eight peaks during one reolution, while that at the ballast draught condition has four peaks. Experient Siulation Figure 6: Coparison of caitation patterns at design drauhgt condition for Prop-A (top:0, iddle:20, botto: 40) Figure 8: Instantaneous pressure distribution on the hull surface induced by the propeller caitation Figure 9: Tie history coparison of pressure fluctuation of P2 transducer at ballast and ballast draught conditions for Prop-A Experient Siulation Figure 7: Coparison of caitation patterns at ballast drauhgt condition for Prop-A (top:0, iddle:20, botto: 40) Figure 8 shows an instantaneous pressure distribution on the hull surface induced by the propeller caitation. Very high pressure concentrates aboe the propeller position, especially on starboard side rather than port side. Tie histories of pressure pulse on P2 transducer at the design The hull pressure aplitudes for Prop-A at the design draught condition are copared with the experiental results in Figure 10. Since the position of P7 was concealed beneath the rudder trunk, P7 could not be easured in this research. For the first blade frequency (1BF), the siulation results show good agreeent with the experiental results in the tendency and agnitude of pressure fluctuation, but the pressure aplitudes of P1 and P2 at starboard side are about 20% and 10% higher than the experients, respectiely. At the other pressure transducers, the pressure aplitudes are ery siilar to the experients. The second blade frequencies (2BF) of the siulation are alost half of the experiental results, 9

6 een though the tendency is siilar to the experients. As explained in Figure 5, the gradient of axial elocity due to the hook shape ortex probably affects the increase of 2BF aplitude in the experient. In contrast, the elocity gradient is ery sooth in the siulation, resulting in relatiely sall aplitude of 2BF. Oerall, the hull pressure aplitudes of starboard side are higher than port side, and the pressure aplitude of P6 located upstrea fro the propeller plane is higher than P located at the propeller plane. Figure 11 shows the coparison of the hull pressure aplitudes at the ballast draught condition with the experiental results. The aplitudes of 1BF are about 10% higher than the experients except for P1 and P6. The pressure aplitude of P1 is about 28% higher than the experient, while that of P6 is about 5% lower. Neertheless, the tendency of pressure aplitude is ery close to the experient. Howeer, the aplitudes of 2BF are significantly lower than the experients because the effect of axial elocity gradient in the siulation ay be diinished due to the relatiely large caity extent in the ballast draught condition. This phenoenon can be explained through Figure 9. The two peaks with about 40 interal at the crest of the signal obsered in the design load condition is not obsered in the ballast condition. 5.2 Propeller Geoetry Another propeller (Prop-B) was designed to study on the effect of propeller rake distribution. Prop-B has the backward rake of 0.014D at tip, while Prop-A has the forward rake of 0.007D at tip. The other paraeters for propeller geoetry such as pitch, caber, and chord including propeller diaeter are exactly sae. Due to the rake distribution, the rotation speed of Prop-B is slightly higher than that of Prop-A; as a result, the caitation nuber of Prop-B is relatiely lower than that of Prop-A. The siulation conditions for Prop-A and Prop-B are suarized in Table 4. Table 4: Suary of siulation conditions to study the effect of propeller geoetry Inflow Speed Propeller Diaeter Prop-A 5.5 /s Prop-B Propeller Speed 9. rps 40.6 rps Caitation Nuber Hull Pressure (Kp) EFD (1BF) EFD (2BF) CFD (1BF) CFD (2BF) P1 P2 P P4 P5 P6 Figure 10: Coparison of pressure aplitudes at design druaght condition for Prop-A Hull Pressure (Kp) EFD (1BF) EFD (2BF) CFD (1BF) CFD (2BF) Experient Siulation Figure 12: Coparison of caitation patterns at ballast drauhgt condition for Prop-B (top:0, iddle:20, botto: 40) P1 P2 P P4 P5 P6 Figure 11: Coparison of pressure aplitudes at ballast druaght condition for Prop-A The caitation patterns at the ballast draught condition for Prop-B are copared with the experiental results in Figure 12. The caity extents of Prop-B are not different fro Prop-A in both experient and siulation. Howeer, the effect of propeller rake is distinguished in the hull pressure aplitude as shown in Figure 1. Prop-B 94

7 with the backward rake reduces the pressure aplitude of about 5% at P1 to P and about 1% at P4 and P5 in the experient. Siilarly, the aounts of reduction are about 10% at P1 to P, about 5% at P4 and about 10% at P5 in the siulation. Fro the siulation results, it can be concluded that the nuerical siulation using UANS is useful to detect the difference of pressure aplitude due to the sall odification of propeller geoetry. Hull Pressure (Kp) EFD (Prop-A) EFD (Prop-B) CFD (Prop-A) CFD (Prop-B) P1 P2 P P4 P5 P6 Figure 1: Coparison of 1BF pressure aplitudes between Prop-A and Prop-B at ballast condition 6 CONCLUSIONS Nuerical siulation studies using UANS for caitation flow and hull pressure fluctuation of arine propellers were presented in this paper. A coputational doain including a full hull body suberged under the free surface was used to siulate directly the wake field at the propeller plane. And the siulation results were copared with the experiental results perfored in the caitation tunnel. Two kinds of studies were executed in this research. At first, the nuerical siulations were perfored in design and ballast draught conditions to study the effect of caitation nuber. The caitation patterns at both conditions agreed well with the experiental results. The first blade frequencies of the pressure fluctuations in the siulations were slightly larger than the experients, whereas the tendency of the pressure aplitudes according to the location of the transducers was ery siilar to the experients. Howeer, for the second blade frequencies the siulations showed soe liitations to predict the agnitude of pressure fluctuation due to the difference of wake field at the propeller plane. The other study was to alidate the detectability of the nuerical siulation for the caitation pattern and hull pressure aplitude. Two propellers with slightly different geoetry were siulated and copared with the experiental results. The caity extents of the siulation and experient were not different between two propellers. Howeer, the pressure aplitudes of two propellers had sall difference, and that was alidated in the nuerical siulation. In conclusion, the caitation siulation using UANS showed reliable perforance to predict the caitation pattern and hull pressure aplitude through this research. To iproe the accuracy of 2BF haronics in the siulation, howeer, nuerical ethods to achiee ore accurate wake field at the propeller plane need to be studied in detail. For the further work, arious ship types will be siulated and copared with experiental data to ealuate this nuerical siulation ethod. ACKNOWLEDGEMENTS This work was partially carried out in the research grant (No ) funded by the Korean Ministry of Knowledge Econoy. EFEENCES Bertetta, D., Brizzolara, S., Gaggero, S., Saio, L. & Viiani, M. (2011). Nuerical and Experiental Characterization of a CP Propeller Unsteady Caitation at Different Pitch Settings. Second International Syposiu on Marine Propulsors, Haburg, Gerany. Boorsa, A. & Whitworth, S. (2011). Understanding Details of Caitation. Second International Syposiu on Marine Propulsors, Haburg, Gerany. Hasuike, N., Yaasaki, S., Ando, J. & Okazaki, A. (2010). Nuerical Study on Caitation Erosion isk of Marine Propellers Operating in Wake Flow, International Propulsion Syposiu, Okayaa, Japan. Kawaura, T. (2010). Nuerical Siulation of Propulsion and Caitation Perforance of Marine Propeller. International Propulsion Syposiu, Okayaa, Japan. Li, Z. & Terwisga, T. (2011). On the Capability of Multiphase ANS Codes to Predict Caitation Erosion. Second International Syposiu on Marine Propulsors, Haburg, Gerany. Morgut, M. & Nobile, E. (2011). Influence of the Mass Transfer Model on the Nuerical Prediction of the Caitating Flow Around a Marine Propeller. Second International Syposiu on Marine Propulsors, Haburg, Gerany. Salatore, F., Streckwall, H. & an Terwisga, T. (2009). Propeller Caitation Modelling by CFD - esults fro the VITUE 2008 oe Workshop. First International Syposiu on Marine Propulsors, Trondhei, Norway. Shin, K. W., Anderson, P. & Mikkelsen,. (2011). Caitation Siulation on Conentional and High- Skewed Propellers in the Behind-Hull Condition. Second International Syposiu on Marine Propulsors, Haburg, Gerany. 95

8 Schneer, G. H. & Sauer, J. (2001). Physical and Nuerical Modeling of Unsteady Caitation Dynaics. 4th International Conference on Multiphase Flow, Orleans, United States. Watanabe, T., Kawaura, T., Takekoshi, Y., Maeda, M. & hee, S. H. (200). Siulation of Steady and Unsteady Caitation on a Marine Propeller Using a ANS CFD Code. Fifth International Syposiu on Caitation (CAV200), Osaka, Japan. 96