Future Voith Water Tractor Development Using Sophisticated Simulation Models

Southampton, UK Organised by ABR Company Ltd Day Paper No 1 7 Future Voith Water Tractor Development Using Sophisticated Simulation Models Dr Dirk Jürgens, Voith Turbo Marine, Germany and Captain Bo Caspersen, FORCE Technology, Denmark SYNOPSIS: The latest developments on Voith Water Tractor (VWT) will be shown. Voith Turbo Marine, partly in co-operation with partners, has been executing full-scale trials, model tests and simulations. The results will be shown for calm water but also for exposed sea conditions. The Voith Turbo Fin (VTF) is a new development. It has been successfully operating for two years. Its use improves escort capability of a VWT and increases steering forces by about 20 per cent. Model test results and full-scale experiences will be presented. The Voith Schneider Propeller (VSP) is key element VWT. New developments on VSP and experiences with slotted guard will be shown. The slotted guard increases bollard pull. 1 ABSTRACT Ship motion simulators have been used for training of tugboat crews for a long time. Nowadays y can effectively be used to enhance ship designs because permanent improvement accuracy of simulator models. FORCE Technology and Voith Turbo Marine are working toger to improve quality simulator models of vessels driven by VSP. Full-scale trials, tank test results and results of Computational Fluid Dynamics (CFD) are being applied and will be presented in paper. Voith has recently installed a tug bridge simulator, SimFlex Navigator, delivered by FORCE, which will be used for improving VWT design. The company will co-operate with FORCE to enhance accuracy simulator models for both tugs and or VSPdriven vessels including incorporation of new developments VWT in simulator technology. This will enable design of an optimised VWT for escort work, harbour services and offshore terminal operations. 2 SIMULATOR TECHNOLOGY AT FORCE TECHNOLOGY The Division for Maritime Industry (DMI) at FORCE has been working with simulation of ship manoeuvring for more than 40 years. The mamatical model named DEN-Mark1 (abbreviation for: DMI Eclectic Nomenclature Mark 1) is 5 th and current generation simulation software and has been under constant development and expansion since its introduction in 1993. identifi cation); 2 Eclectic ie combines specialist knowledge from different fi elds; 3 Modular and separable structure in so-called devices eg rudder-propeller or hawser; 4 Extensive control through input data; 5 Table look-up of data instead of traditional hydrodynamic derivatives; 6 Non-dimensional data based on physical dimensions eg underwater lateral area instead of Lpp; 7 Use of so-called shape coeffi cients in nondimensionalisation. Each above points has played an important role in developing a physical and fl exible model of a VSP propulsion device within DEN-Mark1 framework. Based on this, FORCE Technology and Voith have developed mamatical models for a number of vessels with Voith Schneider Propellers, mainly VWTs. A detailed description mamatical model can be found in Development of mamatical model of a Voith Schneider Tug and experiences from its application in an offshore simulation study 1. Some unique features that make this software state-of--art are: 1 Forces based on physical models (as opposed to a purely mamatical model form eg system Figure 1: VSP Bridge at FORCE Technology. 1

3 THE TECHNICAL PRINCIPAL OF THE VSP: A SHORT DESCRIPTION The VSP bridge in full mission simulator at FORCE Technology is shown in Figure 1, while Figure 2 shows a real VWT in escort operation. Figure 3 shows bridge assisted ship in simulator at FORCE Technology and Figure 4 a simulation of a departure manoeuvre with an LNG carrier assisted by three VWTs. The VSP is a unique propulsion system, offering precise manoeuvrability, because it is a controllable pitch propeller with no preferred pitch direction. The thrust can be steered fast applying x-y-logic. Figure 5 shows sectional drawing of a VSP and Figure 6 installation of two VSPs in a VWT. Figure 2: VWT in indirect mode. Figure 5: Sectional drawing of a VSP. Figure 3: Bridge assisted ship at FORCE Technology. Figure 6: Two VSP installed in a VWT. The thrust is created by vertically mounted blades in a rotor casing. While rotor casing is rotating, blades are oscillating. The blades oscillation is steered by law of intersecting normals. On page three, Figure 7 shows mechanical principle VSP and Figure 8 corresponding hydrodynamic thrust creation. Figure 4: Departure manoeuvre of an LNG carrier from an offshore terminal assisted by three VWTs in pulling mode. A descriptive simulation program showing technology VSP and hydrodynamics can be downloaded at www.voithturbo.com/marine. 2

Figure 9: VWT before launching. Figure 7: Mechanical principle VSP. Figure 10: Modern VWT, equipped with two VSPs and Voith Turbo Fin. The towing equipment and propellers are arranged in a way such that a stable equilibrium is guaranteed, attack point towline force being located directly above fin, just above deck height. 4.1 THE VWT IN WAVES Figure 8: Hydrodynamic principle VSP. To analyse behaviour of a VWT in waves Voith uses following methods: 4 THE VWT, NEW HYDRODYNAMIC ASPECTS 1 Calculation of motions using seakeeping program developed by Technical University of HamburgHarburg (Prof H Söding); 2 Measurements in Voith circulation tank with small models, having installed a wave generator in 2006; 3 Measurements in international model basins, eg participation in SAFETUG programme at MARIN; 4 Full-scale trials, measurements of motions. The VWT as shown in Figures 9 and 10 is well known in tugboat industry because of its safe and reliable technology. Main advantages are: precise manoeuvring, high escorting forces, good seakeeping behaviour no thrust losses in waves, low roll and pitch motion and fast and flexible force generation in all directions. In following principle VWT and some of its important characteristics, such as behaviour in waves and escorting capability will be explained. The VWT has good seakeeping behaviour with low roll and pitch motion. The reason for this is damping effect created by following: 1 The VSP itself with its vertical axis has a strong stabilising effect. Damping coefficients have been calculated by R. Dallinga2 supporting actual results. VSP blades have a high lever relative to rolling centre tug and refore y are efficient even without active anti-roll steering; 2 The large fin and large guard plate with struts as shown in Figures 6, 9 and 10 acts like stabilising fins. From hydrodynamic point of view VWT consists following parts: 1 Bare hull 2 Two VSPs 3 Protection guard with struts 4 Fin 3

The forces that y create are lift forces and forces due The advantage of an active Voith Roll Stabilisation to added mass in water, which give an important system is fact that it can be activated for vessel 2contribution; The large fin and large guard plate with struts as shown in Figures 6, 9 and acts speed like both at10zero and at speeds up to about 85 stabilising fins. The forces that y create are lift forces and forces due to added mass in water, 3 Bilge (if installed). per cent maximum. Figure 13 shows test which givekeels an important contribution; equipment for a model of a VWT ready for tests in 3 Bilge keels (if installed). Figure 11 shows as an example effect Voith circulation tank. Figure 14 shows effect with guard11on a VWT = 36m) at 10ofknots in beam seas active Figure shows as an (L example effect guard on a VWT (L = 36m) and at 10 without knots in beam seasroll stabilisation measured at SVA calculated withwith and without propeller guard. The guard. increase The roll motion is about 35 per cent calculated and without propeller Potsdam. at resonance frequency if VSP guard is35 removed. increase roll motion is about per cent at resonance frequency if VSP guard is removed. Irregular sea from Starboard (90 ) Hs = 1.5 m, V = 10 kn with Bilge Keels 6,0 Voith Water Tractor without VSP - Guard Significant Roll Angle [ ] 5,5 5,0 4,5 4,0 3,5 3,0 2,5 Voith Water Tractor with VSP - Guard 2,0 1,5 1 2 3 4 5 6 7 8 9 10 Wave Period [s] Figure 13: Model tests for a VWT in circulation tank of Voith Turbo Marine. Figure 11: Effect VWT guard on roll motion. Figure 11: Effect VWT guard on roll motion. The company has developed a Voith Roll Stabilisation The company has developed a Voith Roll Stabilisation (VRS) system which is already installed on (VRS) system which is already on some some Platform Supply Vessels. Because installed possibility VSP to change thrust quickly, it isplatform possible tosupply counteract exciting roll moment waves efficiently. Tests have been carried Vessels. Because possibility out in full scaleto with a buoy layer vessel3 quickly, and with models at MARIN and at SVA Potsdam. The roll VSP change thrust it is possible reduction was in range of 65 per cent to 93 per cent in waves up to 3m in height. The roll to counteract exciting moment values waves reduction depends on ship hullroll design, where lower are found for hard chine hulls. There effidamping ciently.oftests have beenhigh carried in full scale with at knuckles. hull is already becauseout vortex generation a buoy layer vessel3 and with models at MARIN and at The mechanism of roll stabilisation for VWTs is explained in Figure 12. Because high lever of SVA Potsdam. The roll reduction was in range of 65 VSP relative to roll centre tug VSP can be used efficiently as an active roll per centdevice. to 93 The perroll cent in waves up to 3m height. The stabilising centre can be calculated by in considering displacement ship and mass in depends water. The roll is normally slightly where beneath centre of gravity. rolladded reduction on centre ship hull design, lower values are found for hard chine hulls. There damping hull is already high because vortex generation at knuckles. The mechanism of roll stabilisation for VWTs is explained in Figure 12. Because high lever VSP relative to roll centre tug VSP can be used efficiently as an active roll stabilising device. The roll centre can be calculated by considering displacement ship and added mass in water. The roll centre is normally slightly beneath centre of gravity. Figure 14: Results of measurements with and without Voith Roll Stabilisation carried out by SVA Potsdam4. 5 GENERATION OF A MATHEMATICAL MODEL OF A VWT In order to generate a mamatical model of a VWT for simulator, FORCE Technology has carried out extensive PMM tests with a model of a typical VSP tug, Kurtama 3. These tests were carried out with hull without propellers but including propeller guard, both with and without fin. Voith provided data describing open water performance propellers as well as interaction between propellers and hull/guard. The wind loads were estimated using database of wind tunnel test data at FORCE, while wave loads were determined numerically. A new VSP propulsion module was designed and implemented in DENMark1 mamatical model that forms basis for simulators at FORCE. The principles and structure of this module are described1. Figure 12: Mechanism of Roll stabilisation for VWT. 4

5.1 VALIDATION OF THE SIMULATOR MODEL The mamatical model was validated by having an external tug master sail tug model in full mission simulator at FORCE both in still water and in waves. An extensive series of manoeuvres including turning circle, crash stop and bollard pull tests were carried out and checked against full-scale trial data in order to tune model. As an illustration validation procedure, Table 1 gives a summary manoeuvres that were carried out for validation simulator model of Kurtama 3. The table contains results from simulations as well as full-scale data from trials in Istanbul on April 2005. The validation manoeuvres were laid out to cover each relevant operational modes tug during offshore berthing operations: free-sailing (nos 1-4), connected to land point (no 5), assisting LNG carrier unconnected (no 6), connected to LNG carrier (no 7) and working in waves (no 8). It should be noted that tug s escort performance was not included in mentioned validation session, but was tuned later against full-scale measurements. Figure 16: Wave pattern of a VWT at MARIN, drift angle about 30 degrees. Figure 15: Wave pattern of a VWT calculated by CFD, drift angle = 30 degrees. Figure 16: Wave pattern of a VWT at MARI drift angle about 30 degrees. The CFD results for a speed of 10 knots and a drift The CFDof results for a speed of fi10 a drift angle of = 30 0 degrees, fin first is sailing, heel angle 30 degrees, n knots firstand sailing, heel degrees, degrees, is shown in following table: shown in following table: Full scale Model scale 1 Model scale 2 D_VSP [m] 3.20 0.20 0.08 LWL [m] 37.00 2.31 0.93 Draught [m] V [kn] 3.30 10,00 0.21 2.50 0.08 1.58 cl 0.0181 0.0037-0.0023 cq 0.501 0.500 0.487 c_lift 0.442 0.436 0.421 Table 2: CFD scale ratios. Table 2: CFD results results for various for scalevarious ratios. Table 1: Summary of manoeuvres carried out for validation simulator model of Kurtama 3 (length = 34m, power = 5000kW) at FORCE. Refer to end of paper for this table. The surge coefficient (cl) is defined in a way that a negative cl gives a force against fin ie The surge coefficient is defined in remarkably a way that resistance. It is interesting that (cl) cl-coefficient changes with a higher Reynolds number. Thereclisgives even a change in against force direction. is low pressure negative a force The fin reason ie resistance. It field at fin tip that is significantly influenced by Reynolds number. The predicted achievable is interesting cl-coeffi changes escorting forces arethat too low if large model cient scales (small models) remarkably are used. The same effect was shown in Reynolds SAFETUG project. For There small model (corresponding to Model scale 2) with higher number. is even a change escort forces for a 37m VWT were about 150t while for a bigger model (Model scale 1) in direction. reason is low pressure escortforce forces were in rangethe of 165t. 5.2 SCALE EFFECTS The development of a mamatical model of a VWT by FORCE is an example usual procedure, where hydrodynamic forces for simulator models are determined by model tests at a certain scale ratio ie at Reynolds numbers somewhat lower than at full scale. Recently Voith has carried out CFD calculations that give insight into Reynolds number effect. The commercial CFD code COMET has been used. The calculations have been done including free surface effect. Figure 15 shows calculated wave pattern for a 2.3m long model of a tug in an escort situation sailing fin first with a drift angle of 30 degrees. It is clear to see lowering water surface at suction side of fin corresponding to a high steering force. The same effect can be seen in Figure 16 measured by MARIN in SAFETUG project. field at fin tip that is significantly influenced by Reynolds number. The predicted achievable escorting 5.3 NEW DEVELOPMENTS forces are toolike low large model scales (smallinmodels) New developments, ifslotted guard, will also be incorporated simulation model. The slotted guard, as shown in figures 17was and 18,shown offers an in improvement bollard pull of are used. The same effect SAFETUG 3 per cent to 6 per cent depending on propeller load. project. For small model (corresponding to Model scale 2) escort forces for a 37m VWT were about 150t while for a bigger model (Model scale 1) escort forces were in range of 165t. 5.3 NEW DEVELOPMENTS New developments, like slotted guard, will also be incorporated in simulation model. The slotted guard, as shown in figures 17 and 18, offers an improvement of bollard pull of 3 per cent to 6 per cent depending on propeller load. Figure 15: Wave pattern of a VWT calculated by CFD, drift angle = 30 degrees. Figure 17: CFD calculation mesh for slotted guard. 5

Figure 18: Stream lines for a VWT with slotted guard. It is intention that above-mentioned hydrodynamic effects as well as various or effects (see 5,6,7 ), will be integrated into simulator model by Voith and FORCE Technology. 6 VWT SIMULATOR TECHNOLOGY, TASKS FOR THE FUTURE For a long time now, simulator technology has been used for training of tugboat crews and pilots and re will be still furr growth of this application. The simulator offers an additional option. It can be used as a design tool for future tugs. Therefore Voith has installed a SimFlex simulator at ir premises to accelerate and enhance design of new vessels. The SimFlex simulator offers simulation of tug operation with and without waves, with and without an assisted ship and also with one or more tugs of different types. Although SimFlex simulator has already achieved a high accuracy comparing real ship and simulated ship, re is still a potential for improvement. This is one area where Voith and FORCE are working toger. The following table shows individual tasks two organisations in coming development of simulation models: Measurements Calculation Simulator model Design variation Voith Model scale in own circulating tunnel (small models), static measurements Full scale trials CFD Strip calculations Improvement VSP model, delivering of new VSP features and new vessel types. Modification ship parameters and test in SimFlex FORCE Model scale measurements in towing tank (larger models) incl. dynamic tests (PMM) CFD Time domain calculation Furr development physical and mamatical models Test in full mission simulator Table 3: Tasks in future development of VSP simulation models. In case VWT, following questions for example could be answered by using simulation technology: 1 What are right dimensions for a VWT to assist an LNG of a certain size in a specifi c area? Which forces have to be created by tug? Which sea conditions are important for VWT design? 2 What is right size of a VWT for a dedicated harbour (eg L = 30m or L = 25m)? A comparison can be made by modelling both vessel sizes and testing m in required manoeuvres and harbour conditions. 3 In general: which power should be installed? This is directly connected to bollard pull and towing (steering and braking) forces at speed. 4 What is optimal fi n size? A bigger fi n offers higher indirect escort forces, but vessel turns slower on spot. 5 What is optimal GM-value? A higher GM is necessary for high steering forces. A lower GM could be better for seakeeping behaviour. 6 Which combination of pitch and rpm changes are most effi cient? An optimum for fuel effi ciency and response can be developed by using simulator. 7 CONCLUSIONS FORCE Technology and Voith are working toger to enhance quality simulator models for vessels propelled by VSPs. Voith has installed in its premises a SimFlex simulator. Model tests, CFD calculation and full scale trials are used for furr development mamatical models ship motions. Simulator technology will be used by Voith to improve VWT designs for dedicated tasks. The VWT can for instance be optimised for LNG terminals. The VWT has a good seakeeping performance due to vertical axis propellers, fi n and VSPguard including struts, and this gives an important benefi t in offshore environment. 8 REFERENCES 1 Agdrup, K; Olsen, A; Jürgens, D, Development of mamatical model of a Voith Schneider Tug and experiences from its application in an offshore simulation study. PROCEEDINGS MARSIM 2006, 25 30 June 2006. 2 Dallinga, R P; Keukens, H J, Seakeeping Tests for a 150 FT Voith Tug. Report No. 010220-1-ZT, December 1990. 3 Jürgens, D, New hydrodynamic aspects of Voith Schneider Propulsion. Hydrodynamic Symposium - Voith Schneider Propeller, Heidenheim, 23-24 March 2006. 4 Roll Tests with Model of a Motor Yacht with an active Controller of Voith-Schneider Propellers, Model No M1247Z000, Potsdam, November 2006. 5 Heinke, H, Model tests with Voith Schneider Propeller at high thrust coeffi cient. Hydrodynamic Symposium - Voith Schneider Propeller, Heidenheim, 23-24 March 2006. 6 Bartels, J-E; Jürgens, D, Latest Developments in Voith Schneider Propulsion Systems. Miami, USA, ITS 2004. 7 Jürgens, D, Vessel with Voith Schneider Propeller new results on seakeeping and creation of escort forces. Schiffbautechnisches Kolloquium, TU-Hamburg- Harburg. 6