Joint Industry Project (JIP) Shaft Dynamic Loads and Responses at Extreme Maneuvering and Ventilation of Mechanical Azimuthing Thrusters Page 1
Contents 1.0 Background... 3 2.0 SHARES JIP Objectives... 5 3.0 Scope of Work... 6 WP1 Operation studies and investigations... 10 WP2 Shaft dynamic response and modeling... 11 WP3 Dynamic loads tests during extreme maneuvering and interactions... 15 Dynamic loads during extreme maneuvering... 15 Dynamic loads during thruster interactions... 18 WP4 Dynamic loads tests at ventilation... 21 Working principle of thruster ventilation... 21 Dynamic loads measurement of propeller ventilation... 25 WP5 Full scale measurement and monitoring... 27 Loads measurement... 28 Vibration measurement... 30 Pressure pulses / underwater noise measurement... 32 Video observations... 33 Operational parameters measurements... 33 Short term dedicated full scale measurements... 34 Long term monitoring during service... 35 4.0 Deliverables... 35 5.0 Finance... 36 Costs... 36 Participation & Contributions... 37 6.0 Time schedule... 38 Page 2
1.0 Background In the last years during the boom of the global shipbuilding industry, a considerable amount of mechanical azimuthing thrusters (typically pushing arrangements with ducted propellers and pulling arrangements with open propellers) have been manufactured and delivered for various applications of different types of vessels, covering wide ranges of operational profiles. Recently, some gear and bearing failures were reported after the thrusters being used in service, for instance, only half or one year, irrespective of the thruster manufacturers or the ship operators. Obviously the operation of those mechanical azimuthing thrusters has exceeded the design constraints and limits, which are based on the present understanding of hydrodynamic loads on thrusters and their shafting systems, including gears and bearings. At least two possible causes have been identified and blamed to be responsible for those mechanical failures. One of them is the extreme maneuvering of azimuthing thrusters, including interactions, typically for offshore structures used for transit or dynamic positioning (DP). The other is the thruster ventilation, which occurs both in DP and also in high speed sailing conditions when fitted to the vessel at a location close to the free surface. In both cases, large variations of hydrodynamic loads and dynamic responses of the shafting system occur, leading to high level transient dynamic loads on the propeller blades, which further transmit through the propeller hub and shaft to the underwater gears, to the pinion shaft and to its bearings. Typical damages found in thrusters are the broken teeth on the bevel gears and the worn bearings of the pinion shafts, although a bunch of safety factors have often been applied in the design stage. These safety factors include the safety factor against surface pitting damage, the safety factor against sub surface fatigue, the safety factor against tooth root damage, the safety factor against the breakage of lubrication film and the safety factor against TIFF (Tooth Interior Fatigue Fracture). It is understood nowadays that the TIFF is the major failure mode that causes the majority of the teeth damages. The following is a picture showing the typical damages. An example of TIFF failure (source: http://coral corp.com/main/bulletins/bulletin%2051_%20interior%20fatigue%20fracture%20_tiff.pdf) Page 3
The scale of the damage is large that worries not only the thruster manufacturers, the gear makers and the ship operators, but also the classification society who defines the rules for designing, manufacturing and operating of safe, reliable and durable thrusters. Currently most thrusters are designed for a single (static) design condition. In reality however, thrusters hardly ever operate exactly at the design point. The impact of the most important cause for the mechanical failures (continuously and/or suddenly changing loads) cannot be assessed in this way. Off design conditions may result in damages and overload the thrusters which are the large risks for the operators. Recent observations have shown that the numbers of failures have been rising due to larger external loads and uncertainties in the failure strength of components. Without deep understanding of the off design conditions and possible dynamic loads on the power train of the thrusters, the shafting system can be either over or underdimensioned. Systems which are over dimensioned, are too expensive to purchase and to maintain while consuming too much fuel and causing unnecessary emissions since they operate often away from their optimal design condition. But for systems that are under dimensioned, damages may occur due to un expected high loads from, such as, propeller ventilation and thruster thruster interactions at off design conditions. In serving the shipbuilding, shipping and offshore industries by thorough studies and research to gain insights on the dynamic loads, especially at off design conditions, MARIN proposes this Joint Industry Project (JIP) on hydrodynamic loads and SHAft RESponses of mechanical thrusters SHARES. By participating in this JIP, participants share their mutual knowledge, their expertise and their experience in solving the mechanical damage problems of azimuthing thrusters. Page 4
2.0 SHARES JIP Objectives In order to enable better designs of safer and more reliable thruster systems onboard special ships, the thruster propulsion systems have to be improved in design, hence tuning the system to actual service conditions. This will result in increased lifetime, lower capital cost, lower maintenance cost, lower emissions and lower fuel consumption. Research into the shaft dynamic response of thruster system is needed to reach this goal. The current tools and methods that are able to simulate the dynamic behavior of a thruster drive train in actual service conditions are partly available, mainly at the classification societies. However, these have limited abilities to predict actual loading during off design and dynamical conditions. Off design conditions, like ventilation of thrusters, are known recently to cause damages but are not well understood. SHARES aims at a better understanding of loads and responses following from such conditions, making the project beneficial for better design of the complex thruster mechanical system. This reduces the risk of failure of systems, which is a requirement to enable special vessels to operate safely. The objective of the SHARES JIP is to achieve the thorough, complete and deep understanding of the present problems. It will do so by: Investigating the vessel operations with thrusters; Measuring and monitoring the full scale thrusters; Analyzing and simulating the shaft dynamic responses; Testing and measuring the shaft dynamic loads in model scale at off design conditions. It is expected that this will lead to sound solutions for the gear and bearing problems of the mechanical azimuthing thrusters in future designs. Page 5
3.0 Scope of Work Shafting systems for marine applications differ very much from vessel to vessel. For instance, the system of a low speed merchant single screw vessel differs very much from that of a twin screw high speed cruise ship, the system of azimuthing thrusters differs from fixed shafting system, and even the shafting system for the thrusters of tug boats, where very long and thin shafts between the upper gearbox and the main engine are often seen, differs very much from the L drive system with electric motors, where very short vertical shafts connect the pinions to the electric motors with large inertia of the rotors. On top of that, azimuthing thrusters have a very short and thick propeller shaft connecting the propeller to the gearwheel. The elasticity of the propeller shaft is so low that any shock load and dynamic peak load on the propeller will be directly transmitted to the gearwheel without any relief. It is assumed that the period of the shock load is so short and high that the whole peak load is taken by only one single tooth on the gearwheel when it is in contact with the pinion. The load peak is so high that it is enough to initiate the TIFF failure in the root center of the tooth. The crack in the center of the tooth will propagate within the root during normal operation with fluctuating loads. In the end, one tooth will break off from the gearwheel. At the same time, the peak load generates high side forces on the bearing of the pinion shaft, resulting in damage to the bearing. This is the present the state of arts on the understanding of the gear damage for thrusters, explaining why most mechanical damages are found in one of the tooth on the gearwheel and at the same time in the pinion bearing. When the thrusters are operating close to the free surface of the water, the probability that the thrusters are subjected to air ventilation becomes high. Sea going vessels with roll and pitch motions in waves are exposed to higher risk of air ventilation, see the picture on the next page. When ventilation starts and recovers, very high peak load on the propeller shaft has been observed. It is so high that the loads are comparable to the peak load due to propeller hitting an ice floe. Pessimistic people would think that there is no way to design thrusters that are strong enough to withstand the ventilation peak loads. The only way to prevent damage is to prevent ventilation. But this is not the whole story of the problem. Gear damages are also found on large thrusters used by drill rigs mainly for dynamic positioning (low power operation) and low speed transit (high power operation), where the thrusters are typically installed on the bottom of the pontoons with at least 10 meters submergence and no air ventilation to be expected. Obviously, the understanding of the gear damage of mechanical thrusters is not complete and thorough research is needed. Page 6
Example of thruster ventilation both in a sea way and in calm water, both for thruster with ducted propellers and also for thrusters with open propellers In order to have focus for the present project, it is decided to study only the dynamic loads at off design condition of the thrusters for the following two operations, which are thought to be the type of operations when gear damages take place. Extreme maneuvering where thruster interactions play an important role on the loads on thrusters. Propeller ventilation where shock type loads on the propeller may directly transmit to the gearwheel and the pinion. The study will incorporate the operational point of view of the vessel operators to prevent mistakes in operating the thrusters and the vessel, and also from the manufacturing/structural point of view to understand the loads and the responses of the total mechanical system in order to design a reliable and durable balanced system. Operational studies and investigations in order to derive operation guidelines. Loads and response studies in order to understand loads and responses and to make reliable and durable designs. Since the damages are found both for thrusters with ducted propellers for low speed and DP operations (typically 0 to 8 knots), and also found for thrusters with open propellers for relatively high speed operations (typically 15 to 20 knots), we propose to focus on the following two types of thrusters. Pushing thruster with ducted propeller Pulling thruster with open propeller The computational fluid dynamics (CFD) is not yet developed good enough to simulate unsteady transient flows for a thruster at off design condition, and is at its infancy for thruster ventilations, Page 7
it is proposed to carry out the dynamic loads study for the present phase of the project mainly by model testing (and full scale), and leave the CFD as future option in the next phase of the project. Mainly model tests will be used to identify the dynamic load levels and responses. Dynamic loads can never be split from the structural response. It is essentially important to simulate the response during the model tests in order to understand the dynamic loads. On the other hand, the propeller shaft of the thrusters is rather thick and short, as we discussed previously, so that the loads may be transmitted direct to the gearwheel without relief. It is proposed to measure the dynamic loads with very stiff system, but at the same time try to simulate the elasticity of the real thruster shaft train. A stiff shafting system with limited elasticity for most of the studies in the present scope An elastic shafting system simulating the real thruster system pilot study The stiff test system can be realized by using stiff sensors with electric motor in the underwater housing of the thruster with large rotor inertia, as shown in the following figure. The propeller will be made of light material, e.g. aluminum. The test setup with stiff shafting system (electric motor in the underwater housing) For an elastic system, it is proposed to make the shafting system with right angle gears so that the elasticity of the vertical shaft and the inertia of the fly wheel or rotor can be adjusted. This system is illustrated in the following figure. The system should be able to simulate the elasticity of the real thruster shafting system. A correct simulation law should be enforced. The simulation law is also part of the study in the present project. Page 8
The test setup with elastic shafting system (right gears in the underwater housing) In order to achieve a structured study, it is proposed to sub divide the project into the following 5 work packages. WP1 Operation studies and investigations WP2 Shaft dynamic response and modeling WP3 Dynamic loads tests during extreme maneuvering and interaction WP4 Dynamic loads tests during ventilation WP5 Full scale measurement and monitoring The WP ordering number is not necessarily the execution order. WP 1 will be addressed first and also WP 5 due to its time consuming aspect and importance. The detailed scope of work for each of the work package will be discussed in the following sections. Page 9
WP1 Operation studies and investigations In this work package, we will choose a few vessels with/without damaged thrusters to carry out operational studies and investigations by interviews, meetings and seminars with the captains and personnel on board in order to understand the way the thrusters were used in practice for DP, transit and high speed operations. This WP will try to map the operational experience and perceptions of risk on damaging the mechanical thrusters among officers, operators and owners who operate vessels with azimuthing thrusters, and to identify the limits related to extreme maneuvering and thruster ventilation, aiming at describing and closing the gaps between how the thrusters are used in practice and what the assumptions and limits are, set by manufacturers and classification societies. This study will identify which dynamic loads are avoidable and which are not. For avoidable loads, together with the results and the studies from other WPs of this JIP, the findings will at the end translated into the guidelines on designing and operating azimuthing thrusters in extreme maneuvering and ventilation conditions for the operators, the owners, the designers and the manufacturers, in order to avoid unnecessary high dynamic loads. For un avoidable dynamic loads, the results of this WP will be used to detail the model testing and investigation programs in order to define the magnitude of those un avoidable dynamic loads and the related load characteristics for dimensioning the mechanical shafting systems. In order to have focus, it is proposed to study only two types of vessels, A typical drilling rig/ship (or heavy transport or pipe layer) with multiple thrusters fitted to the bottom of the hull (6 to 8 thrusters pushing type with ducted propellers), used both for dynamic positioning and also for transit with speed between 6 to 8 knots. A typical fast transporter with light draught (potential of ventilation) equipped with twin mechanical POD pulling type azimuthing thrusters with open propellers, sailing speed around 15 to 18 knots. For each of the above mentioned types of vessels, both vessels with thruster gear damage and vessels without gear damages will be investigated. Due to the yet limited budget, currently only these vessels will be taken into account in the present studies. Possible expansions could be considered if participation (budget) allows or a next phase of the project emerges. The owner, the user and the contractors of the these types of vessels who take part in the present JIP are invited to suggest vessels for investigations/interviews. It is proposed that one of the partners from the classification societies takes the lead for this work package. Page 10
WP2 Shaft dynamic response and modeling When investigating dynamic forces and moments on thrusters and their shafting systems, dynamic loads can never be separated from the response of the systems. Simulating correctly the responses of the mechanical systems in order to understand the dynamic loads is extremely important to get insight into the problem area. This WP aims at investigating the responses of the shafting systems of typical mechanical thrusters, including shaft torsion loads, bending loads, gear meshing and hammering characteristics; identifying and deriving the similarity law on shafting responses simulation in model scale; building and testing the model shafting systems with different dynamic responses in order to simulate the responses in full scale. During the investigations, existing and well validated computer software on thruster shafting systems (typically from classifications as partners of this JIP) will be used to analyze the dynamic responses of both typical full scale thrusters and also the model scale thrusters with modeled shafting systems. The results of this study will be used as instructions on building the model scale shafting systems and on analyzing and interpreting the model test results. The existing 6 component shaft sensor with high accuracy, as shown in the following figure, will be used throughout most of the model tests in the present JIP. The dynamic characteristics of the sensor will be studied and tested within this WP. In order to obtain a higher response frequency of the sensor, another 6 component shaft sensor with higher stiffness and high accuracy will be made and used in the situation where short and single shock load occurs, such as during thruster ventilation. The existing propeller 6 component shaft sensor (depending on the inertia of the shaft and the propeller model, previous tests show a natural frequency of above 300Hz in model scale) The dynamic characteristics of the sensor itself is very important for the dynamic measurement of the shafting system of a thruster. In principle, the sensor should be as stiff as possible, simulating the short and thick stiff propeller shaft of most mechanical azimuthing thrusters. But on the other hand, the sensor should be elastic enough in order to be able to generate enough strain for the strain gauges so that the forces and moments can be accurately measured. A typical response of a sensor to the forces can be explained in the following figure. Page 11
An example of sensor response to the impact force In order to make the dynamic forces and moments measurements, the natural frequency f n of the sensor should be as high as possible. The valid data is only in the frequency regime lower than the natural frequency of the sensor. The following figure shows a typical dynamic force sampled during a model test and its spectrum by a sensor. The measurement is only valid for the frequency range lower than the natural frequency of the sensor. The signal should be filtered by a low pass filter. An example of sampled force and its spectrum For the analysis of the propeller shaft gearwheel system, finite element method should be used to analyze the dynamic characteristics of the total system, including also the added mass effect of the water on the propeller and hub, in addition to the inertia of the mechanical system. The following figure shows a propeller shaft system with simulated added mass of the water and the calculated displacement at its natural frequency of the first mode. Page 12
An example of the analysis of the natural frequency of a propeller shaft system (left: the simulation of the added mass, right: the displacement of the blade due to the thrust force) For the hydrodynamic forces, the model tests should follow the similarity of Strouhal law, as given in the following equation, which describes the ratio between the fluid unsteady force and the inertia force. As part of the study, the similarity law for the mechanical response will be investigated within this work package. According to the above discussions and the goal of this work package, the following scope of work is defined. Analyzing the natural frequency of the model propeller, the sensor, the gearwheel and the propeller shaft system. Determine the valid range of the dynamic measurements, for both the thruster with pushing ducted propeller configuration and the thruster with pulling open propeller configuration. Study and propose the similarity law for the simulation of the mechanical shafting system for the present dynamic response model tests. By selecting two full scale cases with gear damages, one pushing ducted propeller and one pulling open propeller, and using existing class calculation programs to analyze the shafting characteristics and provide guidance for the modeling. Design and build two thruster configurations, one pushing with ducted propeller and one pulling with open propeller, with correct elasticity so that the dynamic response of the Page 13
system simulates the full scale thruster shafting system. This systems should be able to adjust its characteristics so that the dynamic response can be changed. Pilot study and tests to try out this system at selected thruster operational conditions. Due to the complexity of the dynamic response modeling, this wok package remains as a pilot study. The results of this study will lead to a clear picture on modeling dynamic response in model scale in a correct way and further detailed study can be carried out in the next phase of the present project. Class is invited to participate through numerical analysis of the determined and agreed test setup. Page 14
WP3 Dynamic loads tests during extreme maneuvering and interactions Until now, there is limited knowledge on the possible transient impact loads on the propeller blades and the shaft of an azimuthing thruster when the thruster is set to arbitrary angles in any possible advance speed and movement ahead, astern, drift aside, etc. The only practice at this moment for using multiple thrusters is the forbidden zone, which is pre defined for each thruster. Those forbidden zones are defined neither based on the dynamic loads on the thrusters nor on the possible damages to their mechanical parts and shafting systems, but mainly with a view to the degradation of the effective thrust due to interactions. The interaction effect on shaft loads remains unknown and un explored. In addition to the interactions, steering the azimuthing thrusters to a certain large angle during maneuvering at low or high vessel speed may result in unexpected high level dynamic loads variations, with high load spikes and hysteresis dynamic characteristics, as an example shown in the following figure. An example of dynamic loads on the propeller shaft of a thruster (where is the steering angle and the two colors show the two steering directions) According to the previous similar tests carried out at MARIN, the natural frequency of the existing 6 component propeller shaft sensor is high enough to capture the dynamic loads for this WP. It is proposed to use the existing shaft sensor throughout all the test scope below. This work package shall consist of the following two sub packages, one is on the dynamic loads of thrusters during maneuvering and the other is on the dynamic loads of thrusters during thruster interactions. Dynamic loads during extreme maneuvering Within this sub work package, it is proposed to measure the 6 component dynamics forces and moments both on the propeller shaft and also on the total thruster unit on the thruster Page 15
foundation, as shown in the following figure with the coordinate defined for all the forces and moments. Although the coordinates are shown with a thruster with pulling open propeller, these tests should be also carried out for pushing thruster with ducted propellers. The thrust of the duct will be measured simultaneously. The coordinate systems on the propeller shaft and on the thruster foundation During the dynamic force measurements, the following 6 component forces and moments will be simultaneously measured and delivered in non dimensionalized coefficients as given in the following formula. The advance coefficient is defined as, On the thruster coordinate system o xyz, the following 6 components shall be measured,,, Page 16
On the ship coordinate system O XYZ, the following 6 components shall be measured o the foundation of the thruster, The sampled signal will be filtered by low pass filter at a frequency lower than the natural frequency of the propeller sensor shaft gearwheel system. The following scope of work for this sub work package is defined: The tests will be carried out for two thruster configurations, one of pushing type thruster with ducted propeller and one of pulling thruster with open propeller. For each of the configurations, three propellers at three pitch settings (P/D=0.8, P/D=1.0 and P/D=1.2) will be manufactured and tested. The tests will be carried out for the following advance coefficients from J=0.0, 0.2, 0.4, 0.6, 0.8, 1.0 to 1.2 (only for open propeller, pulling thruster). The test will be carried out at both steady and unsteady conditions o In the steady condition, the thrusters will be set to a fixed steering angle and the dynamic forces and moments will be measured. The results will be presented as the mean values together with 95% occurrence intervals around each mean Page 17
showing the dynamic fluctuation of the forces and moments. The range of the steering angles will be chosen as, 0 ±30 o ±60 o ±80 o ±90 o ±100 o ±110 o ±120 o ±130 o ±150 o ±180 o o In the unsteady condition, tests will be carried out at two difference steering rate of the thruster, one will be representative for the maximum steering rate of the thrusters at bollard condition (typically 1/30 Hz in full scale) and the other will be the maximum steering rate of the thrusters at free sailing condition (typically 1/120 Hz in full scale). Dynamic loads during thruster interactions Within this sub work package, it is proposed to measure the 6 component dynamics forces and moments both on the propeller shaft and also on the total thruster unit at the thruster foundation during thruster interactions. We will make use of a stock thruster which propels a jet flow into the instrumented thrusters with 6 component sensors and measure the dynamic forces and moments at both steady and unsteady conditions. The following shows the concept of the interactions test setup. L Test setup for thruster interactions The same definition of the forces and moments as provided in the last sub sections will be used during the interaction tests and the presentation of the results. In addition to the dynamic forces and moments measurement during the tests, it is also essentially important to measure the flow during the test in order to understand the slipstream of the thruster which blows into the other thruster. Therefore, it is proposed to make use of MARIN s Particle Image Velocimetry (PIV) to investigate the flow details for selected interaction conditions. The following figure shows the advanced PIV system used daily at the deep water towing tank of MARIN. Page 18
The advanced PIV flow measurement system at MARIN An example of the measured flow field for An example of the measured wake development a single screw vessel along an open shaft of a twin screw vessel The model test for this sub work package should be carried out in such a way that the instrumented thruster will be set to the prescribed setting angle. The tests should start as unsteady tests. The thruster, which propels the flow into the instrumented thruster, will sweep over an angle range, as shown in the sketch at two different steering rate and at difference advance speed defined as the advance coefficient J. During the tests, the dynamic forces and moments on the propeller shaft as well as on the total unit will be measured. The result will be analyzed in the same way as the previous sub section. After analysis of the unsteady tests results, the critical setting angles will be found when the most severe interactions and peak forces and moments will be recorded and measured. For the critical conditions, steady measurement will be carried out at those setting angles both for instrumented thruster as and also for the thruster blowing flow as. Besides the tests discussed above based on the open water setup of thrusters, model test and dynamic forces and moments measurements will be also measured in behind conditions. Two typical vessels will be selected for this behind model tests, one shall be a low speed drill rig and the other shall be a fast mechanical PODs propelled transporter. For the drill rig, the tests shall Page 19
be mainly tested with the consideration of the influence of the flat bottom on the thruster interaction and therefore the dynamics loads on the thrusters. For the fast transporter, the influence of the ship hull and stern skeg will have strong influence on the dynamic loads on the thruster during extreme maneuvering. The following is the detailed scope of work for this sub work package. The dynamic forces and moments measurements with open water thruster setups will be carried out by unsteady sweeping tests. These should be carried out for the following conditions. o Three relative distance should be tested, from L/D=5, 10 and 20, representing the typical distance between thrusters for drill rigs. o Two sweeping steering rate will be tested, one will be representative for the maximum steering rate of the thrusters at bollard condition (typically 1/30 Hz in full scale) and the other will be the maximum steering rate of the thrusters at free sailing condition (typically 1/120 Hz in full scale). o Three advance conditions should be tested J=0.0 (dynamic positioning and bollard pull), J=0.4 (intermediate condition) and J=~0.8 (free sailing condition). o Four setting angles of the instrumented thrusters shall be tested, =0 o, =30 o, =60 o and =90 o. The dynamic forces and moments measurements with open water thruster setups will be also carried out at steady test conditions at the most severe condition for the forces and moments. o The steady tests will be carried out at the combinations of the selected parameters for, and L. In total, 6 combinations will be tested and the dynamic forces and moments will be analyzed and presented. For the 6 combinations selected in the steady test conditions with the open water thruster setups, PIV measurement will be carried out for each of the selected combination. It is proposed to make 5 flow cuts for each of the 6 combinations. In total, 30 flow cuts will be carried out. The behind test conditions, 3 test conditions for the drilling rig and 3 test conditions for the fast transporter will be investigated and the dynamic loads will be recorded and analyzed. Page 20
WP4 Dynamic loads tests at ventilation Thruster ventilation is an important issue nowadays, not only due to the sudden loss of the thrust and torque during e.g. dynamic positioning, but also the high dynamic loads fluctuations on the blades and shafts. It is well known that the propulsor looses suddenly both thrust and torque within one single propeller revolution when ventilation starts, followed by a few high impact load peaks in the coming revolutions during the recovery phase, back to the fully wetted situation. It is known from literature and some fundamental studies, that the impact loads during the propeller ventilation is as high as those of a propeller hitting an ice floe. The loads can be much higher than people would expect in open water operations. The objective of this WP is to understand thruster ventilation mechanism, to investigate the ventilation inception (onset) criteria, and to measure the dynamic loads on the blades, in the shaft and for the total thruster unit during ventilation, by means of model testing in atmospheric as well as in vacuum conditions with cavitation on the propeller blades. The ventilation event is a very short impact event. In order to capture the peak load, it is proposed to built a new 6 component sensor for the propeller shaft with higher stiffness and higher natural frequency. Since sensor will be used throughout the whole tests within this WP. The ventilation work package consists of the following two sub work packages, one is on the working principle and the other is on the dynamic loads measurements. Working principle of thruster ventilation For the working principle study of ventilation inception (onset), we will make use of an oscillating open water setup for different submergence of the thruster at different amplitudes and frequencies. High speed underwater video camera will be used during the tests to view the flow and the inception of the ventilation. In addition, the videos will be synchronized with the dynamic loads measured on the propeller blades and the shaft. Selected tests will also be carried out in the waves in the renovated depressurized towing tank of MARIN in vacuum conditions with waves, in order to simulate ventilation in a more realistic condition with cavitating propellers. The finding of this study will lead to the better understanding of thruster ventilation and the criterion of ventilation inception, and will result in the guidelines for preventing ventilation from occurring. The photos on the next page show the high speed video observation of ventilation event, together with the measured dynamic loads on the blades which was synchronized with the video record. Propeller ventilation is strongly affected by many factors, including ambient conditions, propeller loading conditions, water quality, cavitation, waves, etc. Primarily, the submergence ratio played an important role, which is defined as, Page 21
where H is the submergence of the propeller shaft under the free surface and R is the propeller radius. Propeller ventilation recorded by high speed video with synchronized dynamic loads In addition, the propeller loading plays also an important role. This could be described by the thrust load coefficient C T which changes with the advance coefficient J. The propeller load will eventually determines the pressure distribution on the blade surface. It is widely accepted that propeller ventilation depends on the relative pressure difference between the ambient pressure on the free surface and the suction pressure on the blade surfaces. That is why many ventilation tests were carried out in the towing tank at atmospheric pressure. But experiments in the depressurized towing tank of MARIN show that ventilation does be affected by the ambient pressure, directly or indirectly. At the same time, cavitation on the propeller blade plays also an important role in the inception of the ventilation. In order to correctly simulate ventilation, it is proposed to carry out the test at pressure condition which is scaled by cavitation number as given in the following formula. When the cavity is ventilated, P v = P a, the cavitation number becomes ventilation number defined as, Froude number, which ensures the similarity of the gravitation, should be also satisfied during the model tests. It affects both the surface waves and also the gradient of the pressure increase with the submergence. The other factors that are believed to have strong influence on propeller ventilation are the surface tension and viscosity of the fluid. Both of these two factors are not able to be simulated during the model tests, but they are all known to have a certain critical values. When the values Page 22
are higher than the critical values during the model test, the influence of the surface tension (described by Weber Number) and the viscous effect (described by Reynolds Number) on propeller ventilation will be limited. At this moment, the Depressurized Towing Tank of MARIN is undergoing a major renovation. Besides the renovation on the instrumentation, the major change that will be made to the tank is adding wave generators on two sides of this tank. The renovation will be finished at the end of Year 2011. From the beginning of Year 2012, the fully renovated new tank will be in operation with the ability for model tests in depressurized condition with waves. The new name of the tank will be DWB (Depressurized Wave Basin). It will become the ideal facility for ventilation studies. The following figure shows the concept of the renovation of the DWB. The Depressurized Wave Basin (DWB) of MARIN Different from a conventional cavitation tunnel where the Froude number cannot be simultaneously simulated as the cavitation number and the advance coefficient, all the similarities can be satisfied in the DWB (including Froude number), except for Webber and Reynolds numbers. Since the starting of propeller blade ventilation is a very short event which takes place within one propeller revolution and starts from one single blade, It is important to measure the shock impact on the single blade, besides the dynamic measurements with the 6 component sensors on the propeller shaft, as discussed and used in the last section. This provides also additional information on blade dynamic loads that will be very interesting for the propeller manufacturer in order to dimension the hub size. It is proposed hence also to measure the blade spindle torque during the propeller ventilation by instruct one blade with sensor, as shown in the following figure. Page 23
Propeller model instrumented with shaft and blade sensors It is also known that propellers can have different stages of ventilations. When a propeller partially ventilated, the dynamic loads on the propeller is rather similar to that of cavitating propeller. When a propeller is fully ventilated, the mean loads on the propeller will be dramatically reduced and at the same time the fluctuation of forces and moments will be limited too. The most dangerous ventilation is in between. People often called critical ventilation. According to the discussions above, we propose the following scope of work for the present sub work package. For the thrusters with pushing configuration with ducted propeller, the study will be focused on dynamic positioning and bollard pull operations, at zero or low speed conditions. We will make use of the oscillating open water system to simulate the roll and pitch motions of the vessel in a sea way. The shaft 6 component sensors will be used to record the dynamic loads on both the propeller shaft and the duct, and also the total unit loads. Only one propeller (P/D=1.0) will be used in this study. During the test, we will make synchronized high speed videos. The advance coefficients, the Froude number, the cavitation number (ventilation number) will be simulated, while the Reynolds number and Webber number will be higher than the critical numbers. o The tests will be carried out at three propeller shaft submergences, being h=0.8, 1.0, 1.2. o The tests will be carried out at three amplitude of oscillations, 0.3R, 0.6R and 0.9R. o The test will be carried out at two advance ratio, representing bollard pull at zero speed and 3 knots current flow. o Only one oscillating frequency will be used during the tests, representing an envisaged full scale roll period of typical DP vessel. For the thruster with pulling configuration with open propeller, we propose to carry out the similar model test as that for the ducted propeller, but focusing at sailing condition. We will make use of the oscillating open water system to simulate the roll and pitch motions of the vessel in a sea way. The shaft 6 component sensors will be used to record the dynamic loads on both the propeller shaft and the duct, and also the total unit loads. Only one propeller (P/D=1.0) will be used in this study. During the test, we will make synchronized high speed videos. The advance coefficients, the Froude number, the Page 24
cavitation number (ventilation number) will be simulated, while the Reynolds number and Webber number will be higher than the critical numbers. o The tests should be carried out at three propeller shaft submergences, being h=0.8, 1.0, 1.2. o The tests should be carried out at three amplitude of oscillations, 0.3R, 0.6R and 0.9R. o The test should be carried out at one advance ratio close to the design sailing speed of the vessel in service. o Only one oscillating frequency will be used during the tests, representing an envisaged full scale roll period of typical fast sailing vessel. In addition to the above tests, for the open propeller fitted to pulling thruster, it is proposed also to investigate the blade dynamic spindle torque during ventilation. (Most ducted propellers used for pushing thrusters have fixed pitch propeller. No investigation will be carried out for the blade spindle torque for ducted propellers during ventilation) Only two cases selected from the above tests with critical ventilation should be carried out. o Propeller blade spindle torque tests at two selected critical open propeller ventilation conditions by using open water test setup without thruster housing. Optional (depending on the availability of the ship model and the willingness of the ship owner to test a real ship in operation), for the selected fast transporter fitted with pulling thrusters where the gear damages have been observed, one behind test of the vessel fitted with the instrumented thrusters will be carried out in a sea way in the DWB. During the tests, the dynamic forces and moments will be measured. High speed videos will be taken simultaneously too. Dynamic loads measurement of propeller ventilation After studies and investigations carried out in the last section, the ventilation mechanism of the investigated thrusters will become clear and the most severe ventilation critical conditions will be found. It is then proposed to carry out systematic model tests as we do for the extreme maneuvering tests, but in ventilation conditions. For the pushing thruster fitted with ducted propellers, we propose to carry out the dynamic forces and moments measurements with open water thruster setup at steady conditions. These should be carried out at the following conditions. o Three propellers at three pitch settings (P/D=0.8, P/D=1.0 and P/D=1.2) will be tested. o The tests will be carried out at two advance coefficient representing the envisaged full scale DP and in current (3 knots) conditions, for each of the propeller. o Only one oscillating frequency will be used during the tests, representing an envisaged full scale roll period of typical fast sailing vessel. o Only one steering setting angle at zero degree will be tested. For the pushing thruster fitted with ducted propellers, we propose also to carry out the ventilation tests at the 6 severe interaction cases selected in WP3, combining with the Page 25
most severe ventilation combination of shaft submergence and oscillating amplitude. The tests will be carried out at one oscillating frequency. For the pulling thruster fitted with open propellers, we propose to carry out the following tests for sailing conditions at the most severe combination of the shaft submergence and the oscillating amplitude found in the studies in the last section, o Three propellers at three pitch settings (P/D=0.8, P/D=1.0 and P/D=1.2) will be tested. o The tests will be carried out at one advance coefficient representing the envisaged full scale sailing condition, for each of the propeller. o Only one oscillating frequency will be used during the tests, representing an envisaged full scale roll period of typical fast sailing vessel. o The test will be carried out at steady condition. The thruster will be set to fixed steering angles and the dynamic forces and moments will be measured. The results will be presented as plots of time signals. The range of the steering angles will be chosen as, 0 ±30 o ±60 o ±80 o ±90 o ±100 o ±110 o ±120 o ±130 o ±150 o ±180 o Page 26
WP5 Full scale measurement and monitoring This work package is aimed at full scale measurements on large thrusters. The objectives of the measurements will be To identify: o Physical phenomena that trigger excessive loads in gears and bearings; o Operational circumstances that provoke these effects; o Amplitudes of load variations in the thruster system; To validate results of model tests and design tools The investigations will be conducted along a stepwise path starting with extensive measurements and observations in a dedicated short term trial, followed by a longer term monitoring campaign where the operational profile of usage and thruster response is captured under normal service conditions. This WP is strongly linked, not only, with WP3 and WP4 for the model tests, but also with WP1 on the operation studies. The results of this WP should contribute to the setup of the operational guidelines too. Scope It is understood that the type of damages that occur in practice, cannot be explained by the loading profile that is assumed in the design. The design does not differ substantially (except for scale) from smaller designs that do not exhibit these problems. It is thus feared that unknown loading phenomena are playing a part in the problem. To identify and quantify the actual loads under service conditions following scope of work will be addressed: 1. Capture the response and loads of the thruster in order to estimate the loads due to: o static loads (propulsion power), o propeller harmonic load variations (due to wake, propeller geometry and cavitation, and o thruster internal dynamics (due to gearing ratio, number of teeth, and bearing dynamics. From these observations it is assumed to be possible to identify the events when the loads exceed the design assumptions. 2. Observe the flow behaviour along the thruster during the dedicated trials by means of high speed video observations in order to compare thruster performance, cavitation and ventilation pattern with design assumptions. 3. Capture the operational conditions that determine the actual loading onto the drive system during the events determined from the response and load observations. These conditions are then to be the basis for further in depth investigations in the model basin or numerical calculations. These may then be expected to reproduce the identified loads and responses. The model (test or simulation) can then in turn be used to evaluate design modifications or operational guidelines to avoid the exceptional peak loads. Page 27
For the dedicated sea trials and the monitoring, the following means will be used. Loads measurement Measuring the loads throughout the thruster structure during various operations would be the preferred way to capture what actually happens. Unfortunately thrusters are huge casted assemblies of gears, bearings and seals that are for the majority filled with gearbox oil which makes it difficult to fit measurement sensors at the locations where one would be specifically interested. These are the propeller shaft, the vertical thruster shaft and the engine drive shaft. drive shaft upper gear engine vertical shaft lower gear propeller propeller shaft Schematic thrusters drive train It is considered that measuring the loads in the drive shaft will be possible; measuring loads in the vertical shaft will not be possible due to accessibility measuring loads in the propeller shaft will be a big challenge. In addition to the principal loads in the shafts, the structure will also exhibit vibrations under the acting loads. Although these vibrations do not provide direct indication of loads they do point out where measured response is coming from by means of typical frequencies related to the number of teeth in the gears, the dimensions of the bearings etc. Measurement of accelerations of the thruster unit in 6 degrees of freedom is expected to provide qualitative information of loads in the thruster. When the thruster is operated in extreme conditions for instance while maneuvering, under shallow draft, or at bollard pull conditions, the inflow patterns and propeller loads may be affected by cavitation and ventilation. These may be possible to observe using video and or high speed video. Drive shaft loads The drive shaft is principally the only shaft that runs inside the hull and is accessible for data collection. If sufficient length of shaft is available (approx 50 cm) then the drive torque and RPM can be captured without too much problem. The main challenge in capturing drive shaft loads is in the required sampling frequency. Since it may be that singular gear tooth loads trigger the first crack initiation, the propeller teeth Page 28
frequencies are of interest. These are determined by the number of teeth on the pinion wheels and the RPM of the shafts. It is recommended to oversample the load variations in order to determine more than one harmonic and check for harmonic load variations from tooth to tooth. It is proposed to capture the load variations at 1000 Hz that allows capturing the third harmonic of the upper gear tooth frequencies. MARIN s shaft torsion load monitoring system is based on strain gauge technology. Data is transmitted from the turning shaft to a receiver as a frequency modulated signal. The centre frequency is 10 khz with variations of 5 khz in plus or minus direction relating to 2000 micro strain. Normal data logging is done with 10 Hz to have maximized resolution of shaft torque. By changing signal collection parameters data can be captured at 1000 Hz at resolution of 2 micro strain. Since the objective is to look for large load variations, this is considered to be a reasonable compromise. In particular when assuming that tooth and shaft designs are most likely designed for similar max working loads related to strains order of magnitude 500 1000 micro strain. Further details with the shaft diameters and nominal loads may be used to assess resolution in terms of actual loads. It is assumed that normal operating conditions will provide a reference performance / response of the loads in the drive shaft. By comparison of in service data versus this reference performance, it should be possible to identify off design events. RPM will be measured along with the recordings to allow assessment of power along with torque and synchronize measurements with shaft angle. The technical challenges for these observations lie mainly in the powering of the shaft mounted equipment. For short term trials this does not pose a problem. For long term trials MARIN would like to consider options in communication with thruster manufacturers and operator of the vessel to be instrumented. Propeller shaft torque (optional) It is considered whether it is possible to fit a propeller shaft torque measurement even though the environment close to the hub of the propeller is particularly harsh with pressure variations, vibrations, and eventual wires/weeds getting caught around the shaft. Supposedly a small length of shaft should be available for measurements between the outer seal/bearing and the propeller hub. (Figure 3). Depending on actual size it might be possible to fit a sensor to the shaft and capture local shaft load. Page 29
Propeller shaft Apart from the challenge to fit the sensor the following complications have to be addressed: Installation of data collection conditioner / logger Waterproofing of sensor and conditioner / logger for prolonged time Power supply of the data logger / conditioner Data retrieval from the data logger Powering and data retrieval from the sensors may end up being the most relevant show stoppers. Two variations can be considered for implementation. The first a standalone setup with battery pack and data logger. The second with a data link for power and continuous data transmission. The latter option may prove to be impossible due to the number of turning axes that need to be crossed prior to arriving inside the hull. The first option however provide challenges due to limited battery life, and practical complications when a diver or a ROV has to retrieve the data. Not all vessels (e.g offshore vessels) will regularly go out of operation. Because of the complications, the option to capture propeller shaft loads will be evaluated and considered but not necessarily be implemented if complications turn out to be too many. Vibration measurement The effects of the various loads are assumed to induce vibrations in the thruster that should be large enough to capture by using accelerometers. As such the actual load is not captured but the effect is. Deviations from the normal loads patterns should be clearly sticking out of the operational profile and allow recognizing off design conditions. In addition to the loads by the teeth and propeller, the accelerations are likely to represent frequencies originating from the bearings as shaft vibrations in the bearings and dynamics of the bearings in the bearing races. The characteristic frequencies from teeth and bearings in upper and lower gears, are expected to allow the identification of the origin of the loads, even when measuring from the top of the thruster unit. It is thus proposed to measure accelerations by capturing: Longitudinal, lateral and vertical accelerations at the height of the upper bearing. Page 30
Horizontal accelerations at thruster azimuthing bearing in order to capture accelerations in pitch and roll directions by combination with sensors upper bearing. Acceleration sensors should have: sufficient range to include extreme design conditions (crash stop, hard over tests) where high vibrations are expected, sufficient resolution to distinguish smaller effects of for instance bearing dynamics over effects by ship motions and engine vibrations data collection characteristics to accommodate the frequencies of the anticipated phenomena. o tooth frequencies and hammering should be captured in excess of 1000 Hz. o bearing dynamics to be captured depending on the bearing geometry but expected in similar range as tooth frequencies. Accelerometers along thruster Typical specifications considered: Range 0 10g, Bandwidth 0.3 10 khz, Sampling rate 10 khz base frequency with filtering for aliasing effects Reduced to 1000 Hz with digital filtering Fc=400 Hz for high resolution and low noise. Effective frequency bandwidth: 0.3 400 Hz. Page 31