Hydrodynamic Fine-Tuning of a Pentamaran for High-Speed Sea ...

International Conference on Fast Sea TransportationFAST’2005, June 2005, St.Petersburg, RussiaHydrodynamic Fine-Tuning of a Pentamaran forHigh-Speed Sea Transportation ServicesEd Dudson, * Stefan Harries *** BMT Nigel Gee and Associates Ltd, Floors 1-3, Building 14, Shamrock Quay,William Street, Southampton, Hampshire, SO14 5 QL, UKEDudson@ngal.co.uk** FRIENDSHIP-Systems GmbH, Benzstr. 2, D-14482 Potsdam, Germanyharries@FRIENDSHIP-Systems.comABSTRACTThis paper discusses the hydrodynamic optimization of the central hull of a 290m Pentamaran for SeaBridge inorder to maximize the speed of the vessel with a pre-determined machinery package. The optimization of thecentral hull was performed by combining the parametric CAD program FRIENDSHIP-Modeler with the CFDcode SHIPFLOW via the generic optimization tool FRIENDSHIP-Optimizer. A scale model of the optimizedcentral hull was made and a series of resistance tests undertaken to verify the accuracy of the CFD calculationsand to prove the validity of the optimization.Figure 1: Rendering of the SeaBridge PentamaranINTRODUCTIONThe vessel is a 290m, 6500t Dwt high speedPentamaran Ro-Pax vessel designed for SeaBridgeInternational Inc. The vessel is powered by a DieselElectric Propulsion plant in combination withwaterjets. A rendering of the vessel is shown inFigure 1.The SeaBridge approach uses the sea to complementexisting land-based travel on selected routes to attractfreight movements away from overcrowdedhighways, for at least part of the journey from shipperto receiver. By using the sea as a bridge, SeaBridgeintends to transport more goods, with lessenvironmental impact, at lower cost, and in a shortertime frame than any comparable transportationalternative available today.Sea transport at 40 knots and higher requires theutmost effort to reduce energy consumption. Thework presented in this paper is, therefore, dedicatedto the systematic fine-tuning of the Pentamaran’scentral hull. The volume, position and shape of thesidehulls were not considered as part of this study ,since they had been subject to previous investigationsin which a favourable method for designing thesidehull configuration had already been established.For the central hull, formal strategies to improve thecalm water hydrodynamic performance wereemployed. A sophisticated parametric definition ofthe hull shape was established within the ComputerAided Design (CAD) tool FRIENDSHIP-Modeler foran efficient and effective form optimization.The repeating stages of geometric modelling,hydrodynamic simulation, performance assessmentand systematic variation were integrated into a

worship, transforming ministry, food, fellowship and anointing! Robyn Vincent will be the mainspeaker, and Karen Vaughn, Emma Colbert and Elaine Priestly will all be ministering inworkshops. The gathering will begin at 10 AM with worship, and run through the afternoon andwill end with an anointing service! There is no cost to attend, so be sure to invite your friends tojoin you. So we can best prepare for lunch, please register with Lori Shelbourn at the back of theTabernacle this Sunday morning, or email her at wilori[at]live[dot]com. For those of you whoare not able to join us onsite, this gathering will be available by live webcast.Blessings,Chuck D. PierceIf you would like to give towards the ongoing work of this ministry, you can donate online (clickHERE), by calling 1 (888) 965-1099 or 1 (940) 382-7231, or sending a gift to PO Box 1601,Denton TX 76202.Key Resources to Help You Break Forth in a New Way!Be Revived Anointing Oil and CandleRevival is in earth's atmosphere! Because the anointing breaks the yoke, we sought the Lord onthe ingredients needed for a revival fire anointing oil and candle. The first aspect of Revived is"awakening". We use costmary, an original herb used by the colonists to stay alert andawake. The second ingredient we incorporate is pomegranate, representing healing, overcomingand prosperity. The third ingredient is "harvest". This is a mixture of the seven species of Israelwhich represent the fullness of unlocking the harvest fields. There is a boldness that God isbringing on His people to pray over others. We must come alive and be bold witnesses in thishour. Be revived and use this oil to "break the yoke" as you minister to others.Roll-on Anointing Oil $10; Candle: $12; Roll-on Anointing Oil + Candle: $20.The Portal!

Installing waterjets on a vessel requires a submergedtransom. A submerged transom on a vessel travellingat relatively low Froude numbers, as with theSeabridge, can have a significant impact on the totalresistance of the vessel. BMT Nigel Gee andAssociates Ltd have, therefore, spent a significantamount of time developing the optimum transomdesign for a waterjet propelled vessel operating at lowFroude numbers. The design is developed around thewaterjets and the requirements to avoid waterjetventilation in waves. A sketch of the transom designis shown in Figure 5.Figure 5: Gulley transom designFigure 3: General Arrangement of the SeaBridgePentamaran1.4. Machinery ArrangementThe vessel is propelled by a diesel electric propulsionsystem which provides for most economic andflexible power plant installation for this application.Medium speed diesel engines burning heavy fuel oilare situated around amidships.Five waterjets are driven by tandem electric motorspositioned close to the transom. The waterjets andelectric motors are configured such that with a singlejet out of operation, the four remaining waterjets canabsorb 100% of the generated power, resulting inminimal speed loss. A sketch of the machineryarrangement is shown in Figure 4.Figure 4: Machinery Arrangement2. PARAMETRIC MODELLING OF THECENTRAL HULL2.1. Parametric ApproachIn order to generate and vary functional surfaces ofcomplex shape such as ship hull forms with a degreeof freedom suitable for optimization, a parametricapproach ought to be followed. The essence of aparametric definition is that a context-dependent andsolution-oriented description is established, whichallows production of those shapes that are beneficialto performance while still acceptable with regard tothe many constraints.The FRIENDSHIP-Modeler grants a high-leveldefinition of hull shapes via form parameters such asbeam, deadrise, draft, entrance angles, sectional areasetc. and was therefore selected for the optimization ofthe Pentamaran’s central hull. For details see(Harries and Abt, 1999), (Harries et al, 2001) and(Harries and Heimann, 2003), further information anddownloads are available at www.FRIENDSHIP-Systems.com.2.2. Form ParametersThe shape of the Pentamaran’s central hull called fora specific parameterization which was developed byFRIENDSHIP-Systems on the basis of the baselinedesign by BMT Nigel Gee and Associates Ltd, seeFigure 6. The form features were closely examined,

for instance via curvature plots, and suitable formparameters were identified.Figure 8: Form parameters for definition of sectionFigure 6: Curvature plot of the baseline designFigure 7: Hull form and basic descriptors of the SeaBridgeFigure 7 depicts the central hull generated via theFRIENDSHIP-Modeler. The figure shows importantshape descriptors, namely the basic curves which aredetermined from small sets of parameters and which,in turn, provide input for surface generation. Selectedform parameters which define a typical section areillustrated in Figure 8. Note that data from the spacecurves presented in Figure 7 appear here in their planand profile components, respectively.Special features of the stern region are presented inFigure 9 – compare to Figure 5 – while the topologyof the surface patchwork is shown in Figure 10.Figure 9: Special features of stern region(marker shows area of influenceif parameter assume non-zero values)

Figure 10: Surface patches of parametric modelAn important task during an optimization is thecontrol and monitoring of design constraints, see (Abtet al, 2004). The design displacement for instancewas fixed at 30000 t. This constitutes an equalityconstraint which could be handled directly within theFRIENDSHIP-Modeler by utilizing the areainformation from the sectional area curve as input forshape generation. Since the aft body of thePentamaran’s central hull featured little freedom fordisplacement changes a special mechanism needed tobe developed: Firstly, the aft body was generatedwith respect to the dominating positional constraintsof the propulsion unit. Secondly, the actual sectionalarea curve was computed for the aft body and thencontinued forward in compliance with thedisplacement constraint. Finally, the sectional areacurve was utilized for the forward body as parametricinput. In this way, infeasible designs with regard todisplacement could be readily avoided2.3. Form VariationsFigure 11 illustrates a representative shape variationout of the many which were studied. Here thedeadrise was systematically changed from zero to 14degrees.Figure 12 gives an impression about the scope ofform variations for the aft region. Theparameterization allowed a smooth transition fromthe Gulley transom to a more conventionalconfiguration. The Gulley however was eventuallyfavoured for the final design (compare to section 1.4).Figure 11: Parametric variation via changes indeadrise

test results had been available for a similar hullreasonable confidence could be established withregard to the flow code’s ability to handle thedifficult flow phenomena aft of the Gulley transom.The CFD based prediction resulted in resistanceestimates below model test results, however, theresistance characteristics were well-captured by CFDfor the speed range in question.3.2. Optimization ApproachOften an optimization programme is beneficiallysubdivided into two phases of design exploration. Inthe former, a large set of designs are distributed in thesearch domain while in the latter promisingcandidates of the first phase are further improved in atargeted variation.Figure 12: Parametric variations of stern region3. OPTIMIZATION3.1. Hydrodynamic AnalysisThe calm water hydrodynamics was of prime concernin this study, therefore the total resistance was chosento be the objective of the optimization. The non-linearwave resistance was assessed by means of thepotential module of the well-known SHIPFLOWcode. The wave making resistance was determined bypressure int egration and wave cut analyses. Viscouseffects were accounted for via the ITTC friction lineand a suitably selected form factor.A systematic panelization study was performed inorder to select a good compromise between accuracy,computational demands and robustness. Since modelFigure 13 shows the outcome of a Design-of-Experiment (DoE) which represents the firstexploration phase. A fleet of 1000 hull forms weregenerated, 527 of which passed all constraints andwere successfully evaluated by the flow code. A setof 21 form parameters were chosen as free variablesfor the DoE. In Figure 13 all valid designs are rankedwith respect to wave pattern resistance calculatedfrom a transverse wave cut analysis. The dashed lineindicates the performance of the baseline. FromFigure 13 it can be clearly seen that a substantialnumber of designs have lower resistance than thebaseline. Naturally, many designs also are rankedinferior.Based on the DoE further optimization runs wereundertaken using the Tangent Search Method byHilleary. This deterministic strategy follows a Hookeand-Jeevesdirect search in connection with a feasibledirection approach as soon as inequality constraintsFigure 13: Resistance estimates sorted by wave pattern resistance(abscissa: accumulated number of designs, ordinate: various resistance estimates)

Figure 14: History of final optimization run(abscissa: sequence of design variants, ordinate: combined and normalized objective function)Figure 15: Wave pattern of baseline and optimized hull SeaFinal9.0074are violated.Figure 14 features the optimization history of thefinal run from which the best design was identified(SeaFinal9.0074).3.2. Selected ResultsFigure 15 compares the wave patterns of the baselineand the optimized central hull, respectively. (Thefigure corresponds to the before-and-after pictures ofa plastic surgery candidate.) The wave resistancefrom pressure integration is reduced by about 13.8 %which yields an improvement of about 5.6 % in totalresistance.4. EXPERIMENTAL VALIDATION4.1. Test ProgramEstimates from the CFD analysis suggested a vesselresistance lower than originally predicted (see alsosection 3.1). The optimized hull also indicated afurther reduction in wave resistance. In order toconfirm the resistance predictions from CFD a set ofcalm water model tests was conducted at

MARINTEK in Norway. Figure 16 shows aphotograph from the tests.Figure 16: Resistance tests at 40 knots4.2. ResultsFigure 17 presents wave cuts from both CFD andexperiment, 0 and 1 being the position of the forwardand the aft perpendicular, respectively. Goodagreement in the phase can be seen while theamplitudes from the CFD are smaller than those fromthe experiments. One possible explanation for thediscrepancies is that the wave cut was measured quitefar from the hull which may have caused numericaldamping in the numerical simulation.The model test results confirmed the improvement inresistance over the baseline. Figure 18 compares thefull scale resistance for the optimized hull asmeasured from model tests against the predictedresistance for the baseline hull at 30000 t.Figure 17: Wave cut from CFD and model testsFigure 18: Model test results for three displacements5. CONCLUSIONSThe process of parametric optimisation through CFDprovides the designer with a tool which significantlyenhances the usability of CFD and results in a trulyoptimized hull, rather than an improved hull.Naturally, the user’s knowledge and understanding ofCFD is vital as well as an appreciation of the limits ofthe flow code. A suitable parameterization of theproblem is essential to reduce the number of freevariables and, consequently, the overall number ofdesigns for which a CFD simulation is to beundertaken. The application of formal strategies tofine-tune the Pentamaran’s central hull for high-speedsea transportation services proved to be successful.ACKNOWLEDGMENTSThe authors would like to thank SeaBridgeInternational Inc. for their support in publishing thispaper. Further information on the vessel and conceptmay be found at www.seabridgeferries.com.REFERENCES1. Abt, C.; Harries, S.; Hochkirch, K., ConstraintManagement for Marine Design Applications, 9 thInternational Symposium on Practical Design ofShips and Other Floating Structures, Lübeck-Travemünde, September 2004.2. Dudson, E.; Gee, N., Optimisation of theSeakeeping and Performance of a 40 KnotPentamaran Container Vessel, 6 th InternationalConference on Fast Sea Transportation,Southampton, September 2001.3. Dudson, E.; Rambech, H.J.; Wu, M.K.,Determination of Wave Bending Loads on a 40 Knot,Long Slender Open Topped Containership throughModel Tests and Hydrodynamic Calculations withParticular Reference to the Effects of Hull Flexibilityon Fatigue Life, 6 th International Conference on FastSea Transportation, Southampton, September 2001.4. Harries, S.; Heimann, J., Optimization of theWave-making Characteristics of Fast Ferries, 7 thInternational Conference on Fast Sea Transportation,Ischia (Golf of Naples), October 2003.5. Harries, S.; Valdenazzi, F.; Abt, C.; Viviani, U.,Investigation on Optimization Strategies for theHydrodynamic Design of Fast Ferries, 6 thInternational Conference on Fast Sea Transportation,Southampton, September 2001.6. Harries, S.; Abt, C., Formal HydrodynamicOptimization of a Fast Monohull on the Basis ofParametric Hull Design, 5 th International Conferenceon Fast Sea Transportation, Seattle, August 1999

AUTHORS’ BIOGRAPHYEd Dudson graduated from the University ofSouthampton in 1990 and joined Nigel Gee andAssociates the same year where he has workedcontinuously with the exception of a year’s sabbaticalin MARINTEK. He is Director of Ship Design forBMT Nigel Gee and Associates Ltd. In October 2004he was elected to the board of Directors of BMTNigel Gee and Associates Ltd. Ed Dudson is aChartered Engineer and Member of the RoyalInstitute of Naval Architects.Stefan Harries is managing director of FRIENDSHIP-Systems, a design consultancy that he co-founded in2001. He studied mechanical engineering at theTechnical University of Darmstadt and navalarchitecture at both the University of Michigan (MSEin 1990) and the Technical University Berlin(Diplom-Ingenieur in 1992). In 1998 he received hisdoctoral degree from TU Berlin, following five yearsas scientist with Prof. Nowacki. He was head ofMarine Hydrodynamics at the Berlin Model BasinVWS (1998 to 2000) before he became head of TUBerlin's research unit Design and Operation ofMarine Systems (2001 to 2003). Since 2004 he isfull-time with FRIENDSHIP-Systems but keeps to beassociated with TU Berlin as lecturer for ComputerAided Engineering.