Scientific Bulletin of thePolitehnica University of Timisoara

Transactions on Mechanics Special issue

Workshop onVortex Dominated Flows –

Achievements and Open ProblemsTimisoara, Romania, June 10 - 11, 2005

APPLICATION OF CFD IN ANALYSIS OF CONVENTIONALPROPELLER AND AZIMUTH THRUSTER

Mihaela AMORĂRIŢEI, Lecturer, Naval Hydrodynamic Department

“Dunarea de Jos” University of Galati*Corresponding author: 47 Domnesca Street, Galati, Romania

Tel.: (+40) 236 495400, Email: mamor@ugal.ro

ABSTRACTThe paper presents a numerical procedure to inves-

tigate the hydrodynamics performances and flow fieldsaround conventional propellers and azimuth thrusters.The propeller running at azimuth thruster is influencedby many interactions between propeller and the gondolaand strut. The geometry of propeller, gondola and strutare known and the problem is solved using a 3D modeland a commercial code FLUENT.

KEYWORDSPropeller, azimuth thrusters, Pods, RANS

NOMENCLATUREJ [-] advance ratioQ [kN] torqueT [kN] thrustDpod/D [-] diameter ratio of pod to propellerkT,kQ [-] thrust, torque coefficientkFx [-] hydrodynamic force developed by

azimuth thruster on the propelleraxis direction

SUBSCRIPTS AND SUPERSCRIPTSp propeller at PODm computed valuec measured value

1. INTRODUCTIONShip propulsion at the required sea speed occurs

with the help of a propulsion device. The most commonpropulsor is the screw propeller, which converts enginetorque to ship thrust, accelerating the fluid in which itworks. The propeller plays an important role in theinteraction between ship, engine and propulsor. Conven-tional propellers are mounted on a shaft and operatebehind ship’s hull.

The last years have shown a remarkable interest inshipbuilding industry for rudder propellers (steerablepropulsion units), which perform both the propulsion

and steering functions. A steerable propulsion unit isable to actively deliver steering moment by rotatingthe thrust vector through the rotation of the thrusters[1]. The most renowned product in this range is the“oldest” azimuth thruster (steerable or rotatable thruster)using the Z-drive or L-drive concept. A new concept,is the Podded drive propulsor – POD – with outboardelectrical motor. In this propulsion system, an electricalpropulsion motor is located inside a “pod” bellow thebottom of the ship and the propeller is mountingdirectly to the motor shaft. The total unit can beazimuting through 360 degree around its vertical axisand this new generation of rudder propellers is namedAzimuthing Podded Drives or AZIPOD for short. Apodded drive has got clear advantages in manoeuvrabilityand hydrodynamic efficiency, in space usage, weightand production efficiency. Space saving is obvious;big propulsion motors are moved outside the ship andthe problem of positions of engine and shaft line doesnot exists. In manoeuvrability of a podded drive iswithout doubt superior to a conventional shaft line.Losses from reduction gears, long shaft line, ruddersand stern thrusters are eliminated. Hydrodynamicadvantages are obtained from the arrangement of thepropeller in relation to the angle of attack; the weakfield and the resulting lower propeller induced pressurepulses [2]. With pods, excellent inflow characteristicsand small cavitation extents on the propeller bladeshave been observed. Addressing the comfort issues, itcan be state that the minimisation of propeller inducedpressures fluctuations and bearing forces fluctuationsis of utmost importance.

The conventional propeller’s design based on standardseries (for initial design) and circulation theory: liftingline and lifting surface theory cannot be directly appliedto the propeller design of a Pod propeller. The pressurefields around the pod housing formed by a gondolaand a strut influence the propeller running at an azimuththruster. The pod house does not only influence thelocal inflow of the propeller but also it represents anadditional resistance [3].

Proceedings of the Workshop on VORTEX DOMINATED FLOWS. ACHIEVEMENTS AND OPEN PROBLEMS, Timisoara, Romania, June 10-11, 2005140

The paper presents a numerical procedure to investi-gate the open water hydrodynamic performances ofa conventional propeller and an azimuth thruster. Thegeometry of propeller, gondola and strut are knownand the problem is solved using a 3D turbulent flowand a commercial code FLUENT.

2. GENERAL ASPECTS REGARDSPROPELLER DESIGN AND ANALYSISA propeller is designed to absorb minimum power

and to give maximum efficiency, with minimumcavitation, noise and vibrations. The design of a shippropeller takes places in three stages: preliminarydesign, design and analysis. In the preliminary designthe problem is to determine main characteristics ofthe propeller to achieve the expected performancesusing standard series. The objective of the second step,known like “indirect” problem, is to find the propellergeometry, problem that can be done using the liftingline theory with correction factors based on liftingsurface theory.

Once the design is completed, the propeller is analysisin all operation conditions taking into account thecomplete wake distribution. The analysis of a propellercan be carried out experimentally and theoretically. Theexperimental tests in towing tanks and cavitation tunnelare time consuming and expensive and for this reason,modern techniques using Computational Fluid Dynamics(CFD) were been developed for investigation the flowaround propeller. CFD methods are concerned withthe solutions of equations of motions of fluid usingboundary conditions, equations, which describe therelations between forces and motions in the flow. Neworientations in analysis of propeller in steady andunsteady flow are: panel methods and RANS codes,methods, which require the knowledge of propellergeometry, computational grids, an iterative processto get solutions and a large computer capacity [4].

Panel methods are potential–flow codes, equationof motion is Laplace’s equation and the boundarycondition of tangential flow is satisfied on the panelson the surface of hub and blades. The panel methodsare useful for problems where pressure distribution onthe blade surface is desired. The panel methods canpredict the open water performance of a propeller, thepressure distribution on the blade in uniform flow, thepressure fluctuations on the blade surface of a propelleroperating in non-uniform flow behind ship and theywould be able to predict the occurrence of cavitationon the blade [5]. Panel methods have an advantageover circulation theory in that they allow the calculationof minimum pressure at the leading edge and handlethe root and the tip better. A very dense grid with smallpanels is necessary at the leading edge.

The RANS codes are useful for problems whereviscous flow is dominate, they are important for theinvestigation of the interaction between the wake field

of the ship and the propeller flow, the tip vortex, thehub vortex and separation along the leading edge. Thecalculation of the viscous flow around propeller is achallenge: the difficulties are not only raised due tothe complex geometry of ship propeller, but also thecomplicated operations conditions. RANS codesrequired computational grids in the entire fluid regionincluding the body surface. The pre-processing partof CFD, i.e. grid generation around a marine propeller,might be improved by having better–geometry fairingsoftware [4]. A turbulent method must be used and thesolution required a very large number of iterations.

CFD methods have become an indispensable toolin the design and analysis of propeller running at anazimuth thruster. The most significant difference forthe hydrodynamic design of a rudder propeller is thenecessity of taking into account the interaction betweenthe propeller and the thruster body. The geometry ofthe pod and strut may be optimised successfully bynumerical simulation. One of the advantages of CFDcodes is than a greater number of design alternatives(different main dimension for the pusher and pullingtype of thruster) can be investigated in a less time withreduced expenses. The flow visualization offered byCFD programs allows better comprehension of the flowphenomena.

CFD methods compliments experimental tests anddesign process. In conjunction with traditional towingtests and cavitation tests and with analytical methodsbased on circulation theory and standard series, CFDcodes represent a new capability to greatly improvethe propeller design and analysis process [6].

3. NUMERICAL PROCEDURESThe Reynolds Navier-Stokes equations (RANS) for

incompressible flow are applied to investigate thehydrodynamics performances and flow fields aroundconventional propeller and azimuth thrusters.

The governing equations for the incompressible andviscous flow are:

01 2

Re

VV V V p V F

t

∇⋅ =∂ + ⋅∇ + ∇ − ∇ −∇⋅τ =∂

(1)

When the equations of motion are solved in a rotatingframe of reference, the relative velocity is introduced,and Coriolis and centripetal terms must be includedin source term.

W V r= − ω× (2)A commercial code FLUENT is used to compute

the turbulent incompressible flow around the propellerblade in two applications: free propeller and propellerrunning at an azimuth thruster. The effects of turbulenceare modelled by RNG k-ε model, derived from theinstantaneous Navier Stokes equations using a mathe-matical technique called ”renormalization group”(RNG). The analytical derivations results in a model

Proceedings of the Workshop on VORTEX DOMINATED FLOWS. ACHIEVEMENTS AND OPEN PROBLEMS, Timisoara, Romania, June 10-11, 2005 141

with constant different from those in the standardk-ε model, and additional terms and function in thetransport equations for the turbulence kinetic energy(k) and its dissipations rate (ε) [7], [8].

The main characteristics of propeller chosen forcomputations are: number of blades z=5, diameterD = 0.21m, blade area ratio AE / A0 = 0.718, meanpitch ratio P / D = 1.03. The open water test of thispropeller, with and without azimuth thruster housing,where carried out at ICEPRONAV Galati and arepresented in [9].

First, open water conditions were simulated for theisolated (free) propeller. The detailed geometrical de-scription of the propeller and the operational condi-tions of the propeller are known. Only one blade ismodelled (flow is circumferentialy periodic) and therotational periodic boundary is used to reduce meshsize. Unstruc-tured tetrahedral cells are used todefine the volume control. A mesh refinement zonewas defined near the propeller blade, but the size ofcells is not enough to boundary layer analysis. Theboundary conditions for computing the flow withina solution domain are presented in Figure 1.

Second, the numerical method was applied to apulling type azimuth thruster. Two flow domains areanalysis: the propeller domain, which is rotating at theprescribed angular velocity, followed by the pod’shousing domain. By means of the CFD simulation,the flows around different main dimensions of the podare investigated.

Figure 1. Coordinate system and boundaryconditions around propeller blade

Figure 2. Boundary conditions around azimuth thruster

4. RESULTSBy means of the numerical simulation based on

CFD, the flow field around free propeller and thepod with a rotating propeller are investigated.

The computed open water characteristics of thepropeller alone and at the podded propulsion arepresented in Figure 3. The thrust and torque coefficientkT and kQ of the free running propeller increase ifthe propeller is running at a pod. The propeller thrustand torque are increased owing to the disturbancepotential wake caused by pod and strut. However, thetotal thrust of podded propulsion is reduced to the podresistance. The computed results have shown that itis possible that the kT increase is bigger that the kQincrease.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.2 0.4 0.6 0.8 1.0 1.2 J

KT,10 KQKtpc

Ktc

KFxc

10*KQp

10*KQ

Figure 3. Open water characteristics (computed)

A comparison between numerical calculation ofthrust coefficient and experimental results available in[7] are presented in Figure 4. The thrust coefficientresults were in good agreement with experiment.Furthermore, the author intends to realise a completevalidation of theoretical calculation method withexperiment in cooperation with HydrodynamicDepartment of ICEPRONAV Galati.

0.0

0.1

0.2

0.3

0.4

0.5

0.2 0.4 0.6 0.8 1.0 1.2 J

KT

Ktpm

Ktm

Ktpc

Ktc

KFxc

Figure 4. Open water results (thrust) computedand measured.

The pressure distributions on propeller blades andon pod are presented in Figures 5,6,7.

Proceedings of the Workshop on VORTEX DOMINATED FLOWS. ACHIEVEMENTS AND OPEN PROBLEMS, Timisoara, Romania, June 10-11, 2005142

Using numerical methods it is possible to optimisethe geometry of the pod and strut. From the compu-tation results: the smaller diameter ratio of pod topropeller Dpod / D decreases the resistance of the pod.From these considerations, according to the diameterof ratio around 0.35 is optimal. The values of diameterratio can vary dawn 0.46 by compromise with the motordimension [10].

Figure 5. Pressure distribution on conventionalpropeller blade (back)

Figure 6. Pressure distribution aroundconventional propeller blade

Figure 7. Pressure distribution on pod

5. CONCLUSIONSThe results presented in this paper are only the begin-

ning of a large research program including numericalinvestigations of flow around conventional propellersand azimuth thrusters. All these effects of interactionsbetween the propeller and gondola, between propellerand hull’s ship require further basic researches to explainthe interaction effects completely. Improvements indesign techniques of propeller behind ship and azimuththrusters may be easier found from the computationalside as from experimental sides. CFD codes complimentsexperimental tests and design process. The detailed flowfield solutions provided by CFD codes enable thedesigner to develop or modify design quickly, differentmain dimension for the propeller geometry and forthe pusher and pulling type of thruster can be investi-gated in a less time with reduced expenses. The flowvisualization offered by CFD programs allows bettercomprehension of the flow phenomena.

REFERENCES1. Van Terwisga T., Quadvlieg F. (2001) Steerable Propulsion

Units: Hydrodynamic issues and Design Consequences.80th anniversary of Schottel GmbH&Co.

2. *** (2000) Annual Report Germanischer Lloyd, Advan-tages of PODs over Conventional Drives.

3. Kaul S., (2000) Special Propulsors: POD new PropulsionSystem, Hydrodynamical Fundamentals and Aspects.34th WEGEMT School, Delft

4. Lungu A. (2001) CFD Modeling of Tip Vortex forOpen Water Marine Propellers, NuSEng’01, Galati

5. *** (1998) 22th ITTC Propulsion Committee Propeller,Conclusions of RANS Panel Method Workshop, Grenoble,France

6. Korpus R., Hubbard B, Jones P. (1998) HydrodynamicDesign of Integrared Propulsor/Stern Concepts byReynolds-Averaged Navier-Stokes Technique, ElsevierScience B.V.

7. Fluent Inc. (2001) FLUENT 6. User’s Guide, FluentIncorporated, Lebanon

8. Fluent Inc. (2001) Gambit 2. User’s Guide, FluentIncorporated, Lebanon

9. Amoraritei M., Micu A. D. (2002) Theoretical andExperimental Investigations of Hydrodynamic Charac-teristics of an Azimuth Thruster in Oblique Flow, SixthInternational Conference on Marine and ScienceTechnology, Varna, Bulgaria,

10. Funero I. (2004) Hydrodynamic Development and Pro-peller Design Method of Azimuthing Podded PropulsionSystem, 9th Symposium on Practical Design of Shipsand Other Floating Structures, Germany.

11. Blazek J. (2001) Computational Fluid Dynamics: Prin-ciples and Applications, Elsevier

12. Amoraritei M. (2002) Some Notes Concerning SteerablePropulsion Units, The Annals of “Dunarea de Jos”University” of Galati, Fascicle XI