Abstract. 1 Introduction

Transcription

1 Effect of stern tunnel height on propulsive efficiency of inland vessels J. Kulczyk, P. Zaj^c Institute ofmachine Construction and Operation, Technical University of Wroclaw Wybrzeze Wyspianskiego 27, Wroclaw, Poland ikem.pwr.wroc.pl Abstract The paper presents results of an analysis of how the stern tunnel height affects the propulsive efficiency of an inland vessel. The study is concerned with a typical configuration of one pusher plus two dumb barges. For each tunnel form and predefined design requirements an optimal propeller type was selected. A mathematical model was used to account for hydrodynamic interactions within the propulsive system. The model is capable of determining service parameters of the propeller with account of the hull and the waterway bed interactions. In particular, the effect of the waterway depth on the propulsive efficiency was determined. 1 Introduction Stern tunnels are commonly used in inland vessels. The tunnels make for higher propulsive efficiency because larger propellers can then be used. They have some drawbacks, too. The complicated afterbody adds to the irregularity of the flow rate field at the propeller disk plane. This in turn has an unfavourable effect on the propeller operation by reducing its efficiency, augmenting hull vibrations and increasing noise level. With higher tunnels it is possible to use larger propellers and hence attain higher efficiency of the propulsor. There is, however, a limit for the tunnel height value. In too high a tunnel part of energy will have to be expended on raising water above the water line. The losses may then exceed the gains due to a larger propeller diameter. It is therefore reasonable to presume that for each stern tunnel form an optimum height value exists. The present paper is aimed at presenting a method for determining the effect of the tunnel height on propulsive efficiency. Both the computations and analysis are based on a mathematical model of hydrodynamic interactions taking place within the propulsion system. To each tunnel variant a unique optimal

2 540 Marine Technology II propeller design (a propeller in a nozzle) will be assigned and, finally, performance of the propulsor will be determined taking account of the hull and the waterway bottom interactions. The propeller itself will be designed using a bearing line model for the propeller and a surface vorticity distribution model for the nozzle (duct). The analysis will be performed for the cases involving both complete and partial annular ducts. The actual reference configuration is that consisting of a BizonlllP pusher and two BP-500 dumb barges. All computational results will be compared with those obtained for the reference vessel. 2 Principles of designing stern tunnels References [1], [2] give general recommendations concerning proper forms of stem tunnels. As an outcome of actual open-water tests and tank tests they offer useful though chiefly qualitative guidelines. With only these guidelines at their disposal a designer is never sure if a better water inflow pattern could be obtained with a different geometry. The principal characteristic of a tunnel is its roof line shape. The roof line is a line connecting the highest points of the tunnel surface. Figure 1 shows a schematic representation of such a line with all pertinent parameters. The basic parameters include: tunnel slope angle a/ and tunnel clearance height E (as measured from the water line upwards). It is recommended that angle a/ should lie within the range 15 to 25 and the clearance is smaller than 30% of the draught value. Most investigators believe that the smaller the slope the more uniform is water inflow into the propeller zone. If E is fixed, then a/ cannot be too small since an excessively long afterbody makes an arrangement of the engine room a difficult task. The tunnel clearance height adopted in this study is larger than the reference one. A larger slope angle a\ is then a simple consequence of the previous assumption. The reference configuration is presented in Figure 2. Table 1 lists all principal data concerning the pusher hull and propulsive system. The forthcoming analysis will use five different tunnel clearance values, four of them larger and one smaller than the reference one. 3 Propeller type selection For all tunnel forms the same principles were adopted for selecting propeller parameters. Computations were performed for fixed values of engine power, suction coefficient and wake fraction coefficient (Table 1). Propeller diameter values resulted from the adopted tunnel height value and the assumption of a nozzle integrated with the hull. Rotational speed of the propeller was then a direct result of earlier assumptions. Those sets of values were adopted that gave maximum propeller efficiency. Advance speed was assumed constant (V=2.78 m/s). The Wageningen 19a duct form was adopted throughout. Propeller blade geometry was determined using a bearing surface model accounting for the nozzle interaction effect. The nozzle effect was accounted for

3 Marine Technology II 541 Figure 1: Basic parameters of a stern tunnel Figure 2: Reference pusher profile

4 542 Marine Technology II by assuming a surface vorticity distribution. A complete computation procedure is presented in [3]. Additionally, for each propeller variant hydrodynamic characteristics were found for the propeller-nozzle assembly. The propeller was then modelled using a surface vorticity distribution. As for the nozzle, its modelling was the same as that used in our earlier design analysis. 4 Model of hydrodynamic interactions within the propulsor-hull system A theoretical presentation of the propeller-nozzle-hull model can be found in [4], [5]. For reasons of space we will recall only main features of the model. It is suitable for analysing hull forms of inland vessels operating on shallow waters. The propeller model is based on the bearing surface theory. The nozzle is modelled using a surface vorticity distribution and therefore asymmetric nozzles placed within a nonuniform velocityfieldcan be readily accounted for. The both models assume fluid to be ideal. Viscosity is accounted for by introducing coefficients of profile resistance. An iterative method is used for design calculations of the propeller-nozzle assembly. The model is capable of analysing assemblies placed within a circumferentially nonuniform field and those composing incomplete annular ducts. If the external (ie external with respect to the propeller-nozzle assembly) velocity field is assumed to be uniform then the model is able to cope with nonuniformity resulting from the action of a limited number of blades. The presented analysis of flow along the hull concerns the area between the ship outer bottom and waterway bed It is based on the modified Karman algorithm allowing for a varying distance between the two planes and can be effectively applied to a viscous turbulent flow. When it comes to determining pressure variation within the zone in question it was assumed that a resulting pressure is a sum of interactions between viscous and potential flows. The pressure distribution due to potential flow is determined using a model assuming surface vorticity distributions both for the ship outer bottom and the waterway bed plane. The model of flow between the ship bottom and the waterway bed accounting for an operating propeller enabled the authors to work out an algorithm for analysing interactions occurring between the propulsive system and the vessel hull at a limited waterway depth. In order to determine these interactions it is necessary to know geometrical parameters of a nozzle, propeller and ship's afterbody. Computations are performed for a given set of kinematic quantities (advance speed, rotational speed). An iterative model is used whose schematic representation can be seen in Figure 3. As far as an integrated nozzle is concerned, its part incorporated into the hull is modelled as belonging to it while the protruding part is taken as an incomplete annular aerofoil section.

5 Marine Technology II 543 Solution of a hull flow (Nominal wake) Ducted propeller performance calculation (iterative procedure) Calculation of vebcity induced by propeller Solution of a hull flow (effective wake) Convergence test coincidence of the effective wake fraction Final calculation of propeller performance Figure 3: Design algorithm

6 544 Marine Technology II With the presented model as a basis a computer program package was built that enables the following parameters to be determined: nominal velocity field; effective wake fraction coefficient; velocityfieldat a given plane with account of an operating propeller; actual operational parameters of a propeller positioned behind the hull (thrust, torque required, efficiency). The propeller efficiency 77 obtained from the analysis accounts for the effects caused by the hull and the waterway bed and can be meant to be a product of the open-water efficiency 770, hull efficiency rjn, and relative rotative efficiency Neglecting the shaft line efficiency the above efficiency value can be understood as the propulsive efficiency rjo ' _ PS _ fa where Ps=R -V- towing power, PB - output power as measured at the main engine clutch,?jtr - shaft line efficiency. If a hydrodynamic characteristic of the open propeller is given then it is possible to determine the suction coefficient from the following relationship: (3) where f - suction coefficient, Too-open water propeller thrust, T- propeller behind hull thrust. The open water propeller thrust T* can be found from the hydrodynamic characteristic provided that both diameter and rotational speed of the open water propeller and behind the hull propeller are equal and velocity of the first type propeller is equal to vessel advance speed. The effective wake fraction coefficient is found directly from calculations. If the hydrodynamic characteristic of an open water propeller is given then the coefficient can be determined from the equality of thrust or torque (as it is done in resistance-propulsion tests).

7 5 Results and conclusions Marine Technology II 545 The most significant results are presented in Table 2. Column 3 of the table refers to the tunnel whose height was left unchanged. The results show that the tunnel height can be safely increased by 0.3 m. The propeller diameter can be then increased by 25% with the efficiency being 18% higher. A further increase of the height causes the efficiency to drop. In reality the efficiency gain is probably smaller since the model does not take account of power losses due to lifting water above the water line. Even then, however, it is clearly seen that there is a single tunnel height value which can be regarded the best choice for a given stern tunnel form. In order to evaluate performance of the incomplete ring nozzle fitted to the reference vessel an assembly with a complete annular duct was modelled. For a fixed tunnel form to accommodate such a solution the propeller diameter had to be reduced to D= 1.14m. The nozzle length was also modified from the standard one Lo = 0.5D to Lo = 0.3D. This particular calculation was made for the DU-2 type duct which is similar to type 19a with the profile section being slightly simplified. The results are presented in Table 3. As can be seen the duct profile form has a negligible effect on the propulsive efficiency. It should be noted, however, that the full-ring nozzle (despite a smaller propeller) is better in this regard than the integrated one. Apart from the tunnel height the waterway depth was another variable investigated. Calculations were made for depths h =,, and m (see Table 3). Additionally, the calculated coefficients of suction and nominal effective coefficients of wake fraction are given. Irrespective of the tunnel form, larger waterway depth causes the propulsive efficiency to decrease. It is yet another confirmation that stem tunnels show their beneficial effect in shallowwater operation. With larger waterway depth the wake fraction and the suction coefficients become markedly higher. The smaller suction coefficient values for the low tunnel and that enlarged by O.lm may result from the fact that the propellers adopted for these cases had a large pitch coefficient. It has been found elsewhere that the propeller type and its geometry have a considerable effect on the suction coefficient [4]. 6 Concluding remarks The presented analysis shows the effect the stern tunnel height has on operational parameters of an inland vessel propulsive system. The analysis is based on a mathematical model of hydrodynamic interactions observed within the propulsive system. The results show that using this model it is possible to determine the effect of a selected tunnel form on propulsor performance and therefore to optimise the afterbody form with account of waterway environment interactions.

8 546 Marine Technology II Lcfml Mfm] Tfml Lwi[m] Engine output Pe[KW] Engine RSP Nominal wakefractionwn Suction coefficient Pushboat Bizon x Ou Dumb barge BP Table 1: Technical data for the reference pusher-barge assembly 1 Tunnel height increase E[m] Diameter D[m] Propeller RPS Pitch ratio at r/r=0.6; P/D Free running propeller efficiency Efficiency with hull Table 2: Tunnel forms investigated AE h(m] TI t Wn We Table 3: Effect of the waterway depth References 1. Heuser, H., Muller, E. Verdrangungsschiffe auf flachem Wasser. Neue Ergebnisse hydrodynamischer Forschung und ihre Anwendung bein Entwurf, Schiffstechnik, B. 33, Heft 1, April 1986 pp Nawrocki, S. Effect of stem tunnel form, puscher-barge on hydrodynamic characteristic of inland vessels, Institute of Ship Design, Technical University of Gdansk, Poland.

9 Marine Technology II Kulczyk, J., Tabaczek, T., Zaj^c, P. Application of surface vorticity distribution in design of ducted propellers, pp , First International Conference on Marine Technology. Odra' 95, Szczecin 1995, Computational Mechanics Publ. Southampton Kulczyk, J. Numerical modeling of hydrodynamical interaction in propulsion system of inland navigation vessel, Scientific Papers of the Institute of Machine Construction and Operation, Technical University of Wroclaw, Poland, No 66, Kulczyk, J. Propeller-hull interaction in inland navigation vessel, pp , First International Conference on Marine Technology, Odra' 95, Szczecin 95, Computational Mechanics Publ. Southampton 1995.