Elektric Thruster. Preliminary Design and Evaluation + Appendices

Electric Thruster

Preliminary Design and Evaluation

J.J.M. van Eijnatten

Delft University

of

Technology ov,s'98/07

This study describes the preliminary design and evaluation ofa new type of electric thruster, in the market known as Azipod'. The work is the second part ofa study for such a thruster, carried out for Lips WV., in Drunen the Netherlands, and is the concluding part of a

graduation assignment in Marine Technology.

This study is supervised by prof. J. Klein Woud and H.T. Grimmelius for the chair in Marine Engineering of the Ship Design section of the faculty of Mechanical Engineering and Marine Technology at Delft University of Technology and T. van Beek and J.Ju for Lips B.V. in Drunen.

For those not familiar with thrusters in general and this electric thruster in particular, a small overview is given in Chapter

Chapter 2 describes the actual preliminary design, with the emphasis on the problematic areas of the design, being the cooling of the electric motor, the structural design and the assembly of the thruster.

Chapter 3 deals with the evaluation of the design, the criteria being costs, weight, performance and manoeuvring capabilities.

The conclusions are drawn in chapter 4, as well as the recommendations for future research. The project that is described in this_{report, is not theoretical, but will be built in the near} future. This practical aspect of this study had the effect that not all of the choices in this report are based on thorough theoretical research. Some of the investigations are superficial, partly because of the wide range of subjects that had to be covered and partly because ofthe little amount of data that was available on the subject. However, a great effort was put in the creation of conscientious and well balanced design, independant of the source or the amount of the data. 1.1501s-,, Fy-f-rUL

<r

Ecl Vtct, 19i

Preface

Joost van Eijnatten Marine Engineering student faculty of Mechanical Engineering and Marine Technology Delft University of Technology June 24th, 1998

Preface

,. ,. .,

....

,

.

.

ne N. V.. ' V V .J e

I

Summary

Overview

_;

Preliminary Design

... ,

2.1 Introduction ₃

2.2 User, classification and technical requirements ₅

2.2.1 User requirements. Lt4 5

2.2.2 Classification requirements

e ,.,

6

2.2.3 Technical requirements

,

₆

2.3 Motor selection _....,,,91

2.4 Geometry.. SS. ,.. Si. _IlIl

2.4.1 Cabling e I I

2.4.2 Cooling , .. q y

,

₁₃

24.3 Strut and pod geometry

_{, -}

₂₃

2.5 Selection of propeller f- 26 - 25.1 Pushing or pulling? 26 2.5.2 Propulsive efficiency . .27 2.6 Pod Components

,

₂₉ 26.1 Bearings ₂₉

2.6.2 Shaft holding brake. _{, 32}

2.6.3 Shaft Seal ₃₃

2.6.4 Miscellaneous equipment_ ....

-

₃₄

2.7 Steering mechanism

,

..-, ₃₅

27.1 Bearings ,

' ' -

₃₅

2.7.2 Fluids transfer unit...

-.36

2.7.3 Slipring mechanism...,-_.,

4...

'

₃₇ _ 2.7.4 Hydraulic systems -4-"- ₃₈ 27.5 Seals.

-38 ' ... ' '. '.: V. 38

2.8.1 Bilge and oil pumps.

,

2.9 Structural Design., ... ...., ... , ₃₉

2.9.1 Dimensioning the parts of the structured... -.. ... ,...,...,... .... . ₃₉ 2.9.2 E-Thruster Assembly

Contents

1 3 2.8 Auxiliaries 38 47

3.. Concept Evaluation

w

49

3.1 Introduction Wol e _... 3.2 Size' _49' 3.3 Weight ₄₉ 3.4 Cost ₅₁ 3.4.1 Steel structure ₅₁ 3.4.2 Motor ₅₁ 3.4.3 Bearings ₅₂ 3.4.4 Seals ₅₂ 3.4.5 Auxiliary equipment _N.

3.4.6 Overview of costs and comparison ₅₄

3.5 Performance ₅₆

3.5.1 Efficiency ₅₆

3.5.2 Manoeuvring' ₅₇

3.6 Reliability %fa-% 63

4. Conclusions and recommendations

_...

Literature'

,

Nt. _%.."4 65

iv 49

53

Summary

An electric thruster, or an E-thruster as it is named in this report, is a new ship propulsion unit. It is a combination of a propeller directly driven by an electric motor installed in a hydrodynamically streamlined construction. This construction can be turned around a vertical axis, providing thrustforce in all directions. A Lips competitor, ABB Azipod, already sells such thrusters and Lips wishes to enter the same market with a similar product, since it is the strategy of Lips to be an integral provider of ship propulsion solutions. This report describes the preliminary design and evaluation of an E-thruster with an installed power of 7 MW.

The preliminary design uses as a

starti4oint

the assumption that if the E-motor fits in the thruster pod, the rest of the components that have to be installed in the pod will fit also. Therefore, the E-motor was selected first, using the data from a preliminary study and data that became available after contacts with various E-thruster motor manufacturers.

Since the most detailed data for the E-motor is coming from manufacturer Ansaldo, their motor was used in this preliminary design, although it is wished for that the outer diameter of the motor will be less for future designs. This is due to the optimal diameter that is thought to be half of the propeller diameter, which is in this case 4 m. Except for influence on the shape of the pod, all currently 'available' motors are identical in output and technical demands. This means that parts of the design have to be adapted when another motor is selected, but concepts used in the design remain the same.

The survey into the cooling demands shows, as can be expected for motors of this size, that large quantities are needed if the motor is to be cooled by air. Since all the current options for the motor are of the air-cooled type, some effort has been put in the design of a set of cooling channels that will accommodate the demands of a 7MW motor that is designed for this E-thruster. However, the cooling of a motor in the confined space of an E-thruster remains difficult, and a close co-operation between the designer of the pod and the designer of the motor is necessary to achieve a satisfactory solution. In the future, the concept of electric motors cooled directly by water will make the design easier and allow for a more elegant design of pod and strut.

The influence of the pod diameter_{on the hydrodynamic properties is significant, but no} further qualitative or quantitative analysis has been performed carried out here to validate this, since this would require model testing. This future analysis includes the making ofa detailed propeller design that considers the shape of the pod and its influenceon thrust and wake fraction. The pod shape and propeller design_{can then be used as the input for model tests.} The data given in the publications of ABB Azipod indicate that the performance gain, dueto less appendage drag and a better incoming flow, will be between six and eight percent. The same or higher values for performance improvement are found in tests in modelbasins with electric thrusters other than Azipods, suchas the Siemens-Schottel-SSP, which makes the electric thruster feasible from an operational point of view.

The preliminary design has been carried out in more detail for the parts of the thruster that were expected to be problematic. The cooling of the motor was already mentioned, the bearings, the structure and the assembly method are other areas with difficulties.

To get an idea of the bearings that can be used, the forces that acton and determine the size of the bearings in the pod, have been estimated using_{very simple static equilibrium equations. It} was proved that 'normal' bearings can be chosen from standard catalogues, for the bearings in the pod, carrying the motor and the propeller, but the bearings in the steering mechanism, carrying the pod and the thrustforce are too large and have to be custom-made.

It has become clear that much attention has to be given to the assembly scheme and method of the thruster, since most of the parts are too large to be machined once welded together.

Because it was needed for the weight and cost calculation to gain some knowledge about the construction of pod, strut and steering mechanism, an initial structural design was made. Welded steel is chosen for the construction of the thruster, with machined and cast steel parts to support the bearings and seals. Since the pod construction is related to the assembly scheme, this is considered together with the structural design, resulting in a more realistic construction. The result is a construction drawing, but is should not be seen as a production-ready drawing, as the dimensioning of the steel parts is based on superficial strength

calculations and a limited experience.

An evaluation has been performed, starting with a weight that is calculated with the use of the drawings made during the preliminary design of the structural design and several values from an earlier Lips project. The weight is, with 162 tonnes, high when comparedto an Azipod unit of twice the power, 14 MW, that is said to weigh 100 tonnes. However, it is not clear if the Azipod figure is for the pod alone_{or for the pod and installation block together.}

The costs of the unit are related to the weight and are thus high. Comparison with the Azipod is difficult, because the known price of the ABB Azipod equipment is including the complete power generation and distribution, which is unlike the cost calculation carried out in this report, where only de thruster and its auxiliaries are considered. Comparison with the

conventional steerable thrusters of Lips shows that the costs have to be reconsidered carefully to accomplish the connection with the existing thruster_{range, that stops at 6500 kW.}

Manoeuvring at service speed was not really expected to be a problem, but model and full scale tests with Azipods show that, the _{course stability at service speed is less good. A small} investigation in steering forces shows that there is _{a difference between steering with}

conventional rudders and steering with E-thrusters,_{but the measures to be taken for} improvement are not yet clear.

1. Overview

This report describes the preliminary design and evaluation of new ship propulsion unit, similar to azimuthing thrusters or rudderpropellers. The new aspect of this thruster is its size, with the purpose to extend the (mechanical) azimuthing thruster range, that stops at

approximately 7.5 MW. In the conventional, mechanical thrusters, the single or double bevel gear is the limiting factor to the power that can be installed. In the new thruster, the gear is eliminated by placing a large electric motor in the under water housing of the thruster, the pod, that is directly connected to the propeller. This solution implies a diesel electric (DE) propulsion system, but as the larger versions of mechanical thrusters were already driven by an electric motor, this is no drawback.

The thrusters, known as Azipod, are already successfully marketed by ABB Azipod in Finland and they started their research in 1990. In 1994 and 1996 Kamewa-Cegelec and Schottel-Siemens followed, but to date, only the units of ABB Azipod are operational. Their units are at this moment installed in two small icebreakers, two tankers with icebreaking capacity and two cruise ships. A large number is on order, almost all for application on cruise ships.

In the near future, two offshore support vessels will also be fitted with Azipods as their main propulsion units. In general, it can be said that almost every ship that benefits from DE propulsion is a possible candidate for this new propulsion unit. Since Azipod is a registered name that belongs to ABB Azipod, the designation E-thruster will be used throughout this report, except where original Azipods are meant.

The goal of Lips is to offer the complete_{range of ship propulsion systems and with this in} mind, this development could not be excluded. In 1995, Lips starteda co-operation with Fincantieri and Ansaldo to develop an electric thruster, but this project was cancelled. A year later, Lips decided to take up the project again and this study was started in the beginning of 1997 with an investigation into electric drives and motors, which is described in [Eijnatten

1997].

Several suppliers and manufacturers were consulted to obtain their views on the motor and the accompanying power system. Aim of the study was to provide Lips with more knowledge of the limitations and possibilities of electric drives.

The study and reactions show that it is possible_{to fit an electric motor in the confined space} of the new thruster. However, several problems have to be overcome, on the electrical and the mechanical field and on the co-operation between_{an electrical system and a ship propulsion} supplier.

The main difficulties for the electric motor are its dimensions, that differ from what is more or less standard in the industry, and its heat production and extraction.

The dimensions, which mainly_{are determined by the low speed of the motor, were covered} extensively in the previous report [Eijnatten 1997]. The cooling properties will be described more in detail here.

The mechanical properties for the E-thruster_{cannot be copied from the existing Lips thruster} range. For most parts of the steering mechanism, the _{same principles can be used, but the size}

for the way the thruster is going to be assembled, again due to the unusual size and weight of the unit.

An advantage of making a new design is that the data for the evaluation based on price, weight, reliability and performance, do not have to be extrapolated, but can be directly drawn from the design. However, in the remainder of this report, it will become clear that much is still unknown and will remain unknown until the first data from an operational unit can be extracted.

The result of this second study is not a perfect design, but an improved overview of what is needed for the engineering of an E-thruster. At the same time, background information is provided on how to estimate and calculate the various aspects of the E-thruster.

It is hoped that this approach will provide a firm funding for a quick redesign or even a more detailed design, at the time that an E-thruster will have to be engineered for an actual

2. Preliminary Design

2.1 Introduction

In this chapter the preliminary design ofa 7 MW E-thruster is described. The consecutive steps in the preliminary design, with cross-references to the concerning sections, are given below.

A sketch that describes the various parts and components of the electric thruster (E-thruster) can be found in figure 2.1 on the next page.

3

User and Technical requirements _2.2

Geometry of Strut and Pod _2.4

Pod Components _2.6

Steering Mechanism Components _2.7

Structural Design and Assembling _2.9

Electric Motor _2.3

Propeller _2.5

Auxiliaries _2.8

I

plIMIIIMIIIIIIfiumit...g ,-ammon...?___ , I lieILLI1 4,4t

i

"11

0

!... I fial.111111.1p1.1111114=--Steering mechanism Thruster pod Strut

Electric motor (synchronous) Propeller

Forward motor and propeller bearing Rear motor and propeller bearing Thrust bearing

Reverse thrust bearing Shaft seal

Shaft brake

Weight and steering force bearing Pod thrust bearing

Steering mechanism seal Cooling air in and output unit Slipring unit

Fluids transfer unit

Hydraulic steering motors with reduction

Figure 2.1 7 MW E-thruster overview drawing

4

117)

2.2 User, classification and technical requirements

At this time, no actual requirements are known. Making a preliminary design is difficult without a specific design target. It has been concluded in the previous study [Eijnatten 1997] that the primary goal will be to design an E-thruster for application on a 'small' cruise ship.

The requirements that where given in the last chapter of that study are extended below and will be used during the preliminary design where applicable.

2.2.1 User requirements

The initial user requirements can be divided in the following categories. Efficiency

Overall efficiency should preferably be better than a comparable, meaning electric, twin propeller propulsion system.

Manoeuvring

The ship should maintain its manoeuvrability at all times, i.e. during harbour_manoeuvres, accelerating and decelerating.

Service and full speed

The manoeuvrability requirements at high speed are not different from those for ships with 'normal' propulsion. They can be summarised_{as follows:}

Straight line stability: if the rudder or thruster is fixed, the ship should be able to follow straight line course not necessarily with the_{same heading, after an external} disturbance

Course stability: when manually or automatically steered, the ship should be able to return to the same heading, after an external disturbance;

Path stability: when manually or automatically steered and position information is available, the ship should return to the same path.

The above mentioned corrections of the course should be possible without the need for excessive (thruster) rudder angles and the overshoot as result of large course

corrections should be s m Low speed (harbour)

Requirements can be summarised _as: turning rate in degrees per minute slow manoeuvre capabilities

Both are mostly depending_{on the thrust force that can be developed. Bow-thruster} power is to be matched to the installed power of the Azipods.

Noise and vibration

One of the reasons, besides manoeuvrability,_{to apply E-thruster propulsion is the expected} reduction in noise and vibrations levels,_{so the levels should be at least equal to those} found for similar cruise ships, but preferably_{much lower. The levels can be specified for} the different modes of operation:

Different speeds between zero and maximum speed with emphasis on the (normal) cruise speed.

Transition manoeuvres, i.e. accelerating _{or decelerating of the ship.} Special manoeuvres, i.e. crash stops_{or manoeuvring in small circles.}

Reliability

Units should be designed for five year survey intervals, as this is the time between, main dockings for cruise ships.

Repairability

Should be such that the small parts of the thruster can be exchanged relatively easy, larger parts are designed to be more robust and require more radical repair work (see also below at the technical requirements).

Installation on board

To minimise building time, the E-thruster should be suppliedto, the shipyard ,as, a complete assembled unit that has been tested. Insertion into the ship will be as late in the building process as possible, so, the additional installation onboard should be limited to the connection of power and data leads, cooling air ducts and hydraulics.

Size and weight

Actually, only the height of the inboard part of the E-thruster is ljmitéd, since the position where the thruster will be installed is in the aft part of the ship, where the deck height is low. There are no limits known or given at this moment for the maxium weight of the thruster.

Again, all requirements above assume comparison with a 'conventional twin screw, diesel-electric propelled cruise ship. The cruise ship that wilt be providing the starting values that are, needed for the design and dimensioning process, is a ship that is currently build for

Renaissance cruises at Chantiers de FAtlantique_{in France. Detailed data can be found in} Appendix B.

2.2.2 Classification requirements

The main categories of classification requirements are strength, manoeuvrability and electrical system. Partly, the common requirements for electric drives and thrusters can be used, but since the E-thruster is a_{new concept, close co-operation with the society under whose} classification the thruster is built, is_necessary.

2.2.3 Technical requirements

Combining the technical requirements from_{several previous propeller designs for ABB} Azipod provides a fair impression of the_{desired capacities of the E-thruster.}

It was mentioned in the earlier study_{that in some cases the torque or power of the E-motor} should be limited. This is to prevent_{damage of the propeller or its connection to the shaft,} damage or overload of the azimuth _{mechanism and damage to the bearings. The following} data were copied from a request by ABB _{Azipod for a propeller design.}

?Propeller, steering mechanism and mechanical strength

In case of a crash stop by reversing the rotational direction of the propellers, torque is limited to 85% of the nominal torque. Reversing of the direction is supposed to takea minimum of 30 seconds from nominal _{speed to zero speed and again 30 seconds to}

4.1r0,100'.

4 6

speed up from zero speed to nominal speed. Astern speed limitation of the propeller is to be taken 20% above nominal speed.

Rudder (azimuth) angles of up to 35 degrees should be possible while maintaining full power. Above angles of _{35°, considered as an off-design condition happening in}

case of a hydraulic failure, power is limited with the power going to zero foa

± 180 °. The system Should be able to withstand this last condition without damage, so. that it can be used as another means of performing a crash stop.

While manoeuvring (in harbours or narrow waters), power is limited to 70% of nominal power. The accompanying torque is limited to 95% in the ahead condition and 85% in the astern condition (of the propeller) for all steering angles. Quick rotation of the propulsor over 180° should be possible, with a maximum azimuthing speed of 9 degrees per second (9 °/s). Speed is limited to 15 Kri for these manoeuvres. In the above project, the reversing time of the propellerseems to be chosen

conservatively long For the mechanical thrusters that Lips manufactures, a reversing time of a few seconds is stated.

Propeller strength

The propeller strength requirements indi

electrical thrusters are operated in

_{a different way, of which the cra9p manoeuvre with the}

thrusters both e outsille-Ns the mos vious example.

Therequirements of ABB Azipod for propeller strength, interms of steering angle, propeller speed and torque, are collected below. The requirements are a summary from several requests for propeller designs for (future) cruise ships to build by Kvaemer_{Masa Yards. Between} brackets, the requirements for the first Azipods (Carnival Elation) are given. The difference shows that the requirements are gradually become less extreme.

Conditon _{1.. Full power ahead, 100% rpm, 100% speed}

Full power ahead, 97% rpm, ship speed reduced for full power consumption. (condition to be matched with engine capabilities)

Astern condition as represented by the backing bollard condition at a reversed rate of revolutions of minus 70% rpm, Azipod angle iszero degrees.

Full power at 100% rpm atan Atmod angle of ±35 (0-180) degrees. Crash stop manoeuvre, Azipods both turning 35 (90) degrees in 'opposite direction

Failure condition, Azipod turning 180 degreesat full speed with zero motor power.

The data in the list can be sometimes _{very questionable, especially for condition 4, but} indicates that a propeller for this type of thruster has to meet strict strength requirements. Details of the design

A few specific design difficulties exist for the_{E-thruster itself. One of them, the size the} motor, is extensively described in [Eijnatten

0-e,-13ct'

e that ships that are propelle these large

(Aire"

ra-r ;'(.4°

The most obvious problem for_{an E-thruster is the accessibility. Although it would be possible} to provide space for monitoring and maintenance_{purposes, it would increase the dimensions}

±

97].

of the pod enormously. This would have a significant influence on the hydrodynamic properties of the thruster. For larger installed powers. the accessibility problem becomes smaller, since the propeller size allows a larger pod body and the motor size does not increase as much. The 14 MW units that ABB Azipod has build for the Carnival Elation are said to have some access to the pod [The Motor Ship 4-1998].

For the smaller version of the E-thruster, that is considered here, it is chosen to have no access to the pod from within the ship. This means that when the ship is in operation, there is no possibility other than by (electronic) sensors to monitor the pod and its contents.

It is then logical to make the contents of the pod as robust as possible. This can be carried out as follows:

Keep the number of (non-accessible) components low; Apply long life components;

Apply an appropriate safety margin for critical components; Install redundancy.

8 ad I

Components with moving parts that do not necessarily have to be placed in the pod, / are installed in the ships hull. Examples of such components are the pumps for bilge water and lubrication oil. An example of a component that has to find its place in the pod is the mechanfeal shaft brake.

One of the few items that can be made reliable by redundancy is the shaft seal and this is performed by adding additional seal stages, that do normally not run on the shaft, but can be deployed when necessary. This is already a normal procedure when designing conventional shaft seals, and the E-thruster does not necessarily have to be

constructed differently.

Making the pod robust also has an advantage in the engineering stage. When the pod is engineered to last for a certain time without large overhauls, the construction does not have _to be adapted to regular disassembly and assembly of parts that have to be inspected. This allows for a more effective construction. On the other hand, with this method of engineering, if for

any reason an interim inspection should be necessary, a disassembly of certain parts may result in disassembly of large parts of the pod.

2.3 Motor selection

In the preceding study, the ABB, motor in the Azipods of the tankers Uilcku and Ltinni was chosen as a reference. This provided guidelines for the size of the 7 MW E-motor. The responses of E-motor manufacturers Ansaldo, Brush and LDW (Lloyd Dynamo Werke,

Bremen), show that a very small 7 MW is indeed feasible, although the estimated 1700nun for the diameter, found in the previous study, turns out to be too optimistic.

Although there might be still room for improvement, the diameters of the Brush and LDW motors are sufficiently small.

A small remark with the Brush motors should be made concerning the brushes: "Brush maintenance is expected to be on a 1 to 2 year basis"; "It may be prudent to fit redundant, brushgear". This short maintenance interval of 2 years will not be acceptable.

The favourable option is the brushless motor, with excitation by a rotating rectifier as proposed by both Ansaldo and LDW.

If brushes are to be used, efforts should be put in the selection of brushes with a prolonged lifetime, for example by overdimensioning, or by installing redundant brushgear, as was as suggested by Brush

table 2.1 Electric motor data

Rotor current

(excitation) 1! A

In appendix A, the original data can be found. An uted

_{table with motor data and}

corresponding graphs, equal to those in the previou study, is_{also included in the appendix, to} allow comparison with other motors.

9'

-14 ups ia alti, 9.e.t.

rt

C`Cc I

Brush

LDW I Desired DP - r . i S,B !, S S(B) 1 1 7000 , 35100I 7000 ! 70001 7000 ' H 393

/

393 \ 393 1 393 393 0-230 185 I 170 I 170 170 12

12\

14 12 23 18.5 I 19.8 17-1945 1700 \ , 1850 1850 1 1700 1 3855 I 2410 1 3670

3370

3750 I t. Do <51i 1355 25 '

10

_20.5

₂₀ , I I 1 18.5 ₁₁₂ 45 i

22

'

39

32 I

95

I 93 H 93.3 95.8 H 650 2.2000 I _{2-4000 2-2900} 24355 640 1 637 770 Si1 0.95 0L85 1 0.85 0.93 i 1 1

200/

1 325

531

i 600 1 i

Power

kW

Torque

IcNm Speed (range) rpm 1Poles

Frequency

Hz

Diameter

Length Rotor weight 1 t

Stator weight

Total weight

t

Efficiency _IVoi

Stator voltage

V

Stator current

A

Power factor

Rotor voltage (excitation)

E-motor data

Application

Type

S = synchronous B = brushless 1 -/ mm mm t [ ]

During the remainder of this preliminary design, the data for the 7 MW 170 rpm Ansaldo motor will be used for calculations and drawings. The data from Ansaldo are the only

available detailed motor data that includes drawings. Solutions for cooling and support for the Ansaldo motors were already known from [X40200]. The data from Brush became available in February 1998, the LDW data in April 1998, and are referred to in this report when

applicable.

2.4 Geometry

The rotating part of the E-thruster can be split in three subsections being the housing for the E-motor, the pod, its connection to the ship, the strut and the circle shaped steering pipe, located inside the steering mechanism. The pod and the strut together form the underwater part. The geometry of these three parts of the E-thruster is mainly determined by the following aspects, which are given in order of importance;

size and shape of the E-motor (for the pod)

space needed for cabling and cooling of the E-motor (pod, strut and steering pipe) hydrodynamic properties (pod and strut)

Although it seems strange to make the hydrodynamics subordinate to the motor and the cooling, it shows the importance of the latter and is characteristic for the way these thrusters are designed.

The aspects will be described in the following sections.. The size and shape of the electric motor are fixed properties and will not be further detailed. The cabling will be described first, followed by the cooling. The results will be used at the end of this section to propose shapes for pod and strut.

2.4.1 Cabling

'Three groups of cables are supposed to enter the pod through the steering pipe and the strut. Main power supply

From the previous Lips 16.5 MW electric thruster project [x4o2o0b Ansaldo figures that a set of 20 cables of 3*120 mm2 each is needed to carry the current of 4000A. Although the accompanying high voltage determines the thickness of the insulation, the current that the cables have to carry determines the diameter of the copper core. The latter has the most important influence on cable diameters for this application.

When looking at 7 MW motors, Jeumont provides _{a figure of 1150A (with a voltage of} 5.5kV) and Ansaldo uses 1900A (with a voltage of 2.6kV), which gives_{a range from 1000'} to 2000A. This corresponds with 5 to 10 cables of 3*120 mm2each for a three phase system. Considering that the current for the Ansaldo motors is set at 1355A, seven cables (1400A) should be sufficient. Assuming that the space needed for packing and additional insula

of

the cables is equal to the cross section of a single set of three phase cables, the amount of space consumed by the cables becomes 2.7.3.120 _{= 5040 nun'. The diameter needed to lead} the cables through then is approximately 160_{mm, which appears to be a low figure. When}

considering normal household cabling, 2.5 mm2 of copper for 10A, a similar evaluation for a 1600A, three phase cable set, with another factor two to account for the individual cable isolation ,would lead to the dimensions 2.2.3-(l 600-(2.5/10)) = 2400 mm2 with an outer diameter of 78 mm. This more or less validates the previous value.

Brush (appendix A) specifies that six cables of_{150 mm2 each are needed to feed their design} for a 7 MW motor, with 600A. An _{additional earthing cable is needed, as well as two cables}

/

n

for the excitation of the rotor, also with a capacity of 600 A, bringing the total number of cables to eight.

Both the Ansaldo and the Brush specifications are sketched in figure 2.2. Data transport

Because distortion by the strong magnetic fields is likely, fibre optics seem to be the appropriate medium for the transport of data. The data is coming from sensors for temperature, shaftspeed, shaftseal (leakage) and bearing vibration, as well as voltage and current measurements in the rotor and stator windings.

ABB Azipod PSG 6-1997] refitted a system of fibre optics for this purpose on the Carnival Elation and Paradise and Ansaldo indicated the same in [Lips X40200].

Reserved space for data cables

In the strut design, a cross section of 300x300 mm is reserved for the previously named cables.

All the other cables, such as the cables needed for data transmission, are supposed to fit in this space also. In figure 2.2 it can be seen that the cable duct is somewhat over-dimensioned, with the chosen cross section, but the figure from the calculation above might be too low and a

relatively small amount of space is involved.

figure 2.2 Power cables and power/data duct

12

2.4.2 Cooling General

All the electrical experts that were consulted by Lips see the cooling of the electric motor as one of the largest problems of the E-thruster. Electric motors of this size have a more or less fixed level of losses, but the larger the motor, the outside surface area becomes relatively smaller, making it more difficult to remove the heat. The confined space of the pod makes it impossible to apply a standard motor construction type with fins or ribs on the outside surface of the stator package, to enlarge the cooling surface. In addition the forced air-cooling that is usually performed by a fan installed on the rotor shaft will not do for this application, as the rate of revolutions of the rotor is simply too low, especially when the motor is running at reduced speed.

The E-motors that were proposed by Ansaldo and Brush are air-cooled machines, similar to

the motors that are supplied by ABB for ABB Azipod, with the air being provided by a separate, powerful fan. The first proposals of STN/LDW also assume cooling by air, so the air cooled option is described in detail in this section . Atthe end of the section, a small overview

of possible future developments will be given. Cooling by air

The required flow of cooling air is high and, as has been mentioned,--grenerated by a separately driven fan. The fan(s) and optional heat exchanger(s) are_placeil in the hull of the ship, for instance in the room that also houses the hydraulicOciisteering the thruster. It might be possible to use the air from the engine room tatool the thruster, but it is expected that a (sea)water cooling system is needed 'topre ool the air, so this is the solution that is worked out in this report.

All this means that large ducts have to be incorporated in the pod, the strut and the steering (azimuthing) mechanism.

Heat production and transfer

With efficiencies ranging from 93 to 95 percent, the heat losses are between 350 and 500 kW. Part of the heat, 50 percent according to peumont 0011, could be dissipated by using the thermal connection between the stator and the outside of the pod, with the surrounding seawater, acting as a cooling medium. Both Ansaldo and Brush supply valuesindicating that the full load is to be transferred by means of the airflow. The estimated requirements of both

manufacturers are collected in table 2.2 on the next page, and the requirements themselves can be found in appendix A. LDW did not yet provide cooling requirements.

A rough estimation of the cooling properties can be carried out as follows:

11motor 0.945 (Ansaldo 7MW for cruise application) = 390 kW Cooling by air Ti = 20 °C Tout =45 °C

p= 1.2

kg/m3 cP 0 kJ/kgK 390

130m

(13v'air AT - c

p

25 -1.0 .1.2

Another, more empirical, formula [STN/LDW 4-19,98] provides _{an even higher figure, for this} motor:

Pp

. -2.5 (-1)V, air = 60

3902.5

rri3

=16.3-60

The figures above and the data from the _{motor manufacturers in table 2.2 show that the} removal of the heat losses from the motor requires large_{quantities of cooling air, represented} by the worst case value of the Ansaldo motor, 17 m3/s.

If a percentage of the heat could be cooled by_{means of the connection of the stator to the} outer hull of the E-thruster, the amount of air would be considerably less.

/P),ijak\-'

PA-14

Cooling requirements

Ansaldo Brush

Motor designed for

Cruise-ship thruster , Dynamic/ Positioning thruster Cruise ship thruster-Power KW 7000 7000 3500 7000 Losses to exhaust KW 390 350

Air

Flow rate m3/s 17 15 2-6.3 2.12

Outlet air temperature

°C 45 45

Pressure drop across motor

da Pa 220 250 280 280

Fan power

KW 30 60

Water (heat exchanger)

Inlet/outlet water temperature

°CA

( kg/cm2) 35/42 < 5 35/42 < 5

Water intake pressure during

normal operation

Air to water heat exchanger

test pressure

kg/cm2 10 10

Water flow rate (approx.)

1/min 750 750

Water pressure drop

kg/cm2 - .0. 6 ._.0.6 table 2.2 Cooling recjuirements

-'

/

/

' '

The amount of air together with the velocity that can be reached by the fans installed in the engine room, is responsible for the cross sectional area of the ducts.

The velocity of the air is to be kept as low as possible because the pressure losses vary quadratic with the speed.

For the following, it is assumed that only 75 %, midway between the estimations of Jeumont (50%) and Ansaldo (100% = 17 m3/s), of the losses has to be transported by the cooling air. This results in an air flow of approximately 13 m3/s. The motor that will eventually be used, will have a better efficiency than the Ansaldo motor, so even when the full 100% of the losses has to be transported, this figure of 13 m3/s, will be sufficient. In appendix A, some data can be found for the combination of a fan and a heat-exchanger, originally meant for the X40200

16.5 MW project. Since the flow and capacity of this combination, 12 m3/s and 300 kW respectively, are close to the requirements above and other data is not available at this moment, the data will be used for this preliminary design.

The fan is driven by an electric motor with an installed total power of 55 kW. This electrical power for the fan has to be seen as a loss that has to be added to the losses in the propulsion chain and it lowers a possible efficiency gain for propulsion with E-thrusters.

Flow path

The flow of the air in the Ansaldo motor isa combination of an axial an a radial flow, as can be seen in figure 2.3, below. The air is going in through the ends of the motor, going out through holes in the stator surface and is collected in the space between the stator and the pod

structure. According to [STN/LDW 4-1998] this would require longitudinal holes in the stator, that would be impossible to make in their opinion,_{as they would disturb the winding pattern.} However, when looking at the relatively large diameter of the Ansaldo motors, there might be some room between the stator windings reserved for the transport of cooling air. Furthermore, Ansaldo states that a distance of 300 mm [X40200] is needed between the surfaces of the motor and the pod for cooling purposes.

tIO1 AIR OUILL I

I

figure 2.3 Flow path of cooling air in

the Ansaldo E-motor 15

When looking at the, very sparse, information ABB provides in literature on its Azipod motors (see figure 2.4), it would seem as if only an axial cooling flow exists. This would greatly simplify the cooling channels in the motor and eliminate those surrounding the motor. However, in the Azipod the flow of air is going mainly around the stator instead of through the rotor, a system that is patented by ABB and can be seen in figure 2.5 on the next page. The steel support structure for the stator may well be combined with the cooling duct demands, as ABB Azipod has done for their electric thruster. The pod frames and stator winding support are combined and form cooling ducts together with the outer shell. A real comparison of technologies cannot be made as the flow path in the Brush and LDW motors is unknown at this moment. The Ansaldo motor will remain the example, although not much information is available on the exact path of the air through the motor and the feasibility of their design.

figure 2.4 Expected flow path of cooling air in the ABB Azipod E-motor

For this preliminary design, the following path through the motor is chosen, in accordance with the Ansaldo data. Going from the strut to the pod, the incoming flow makes a sharp turn to enter the motor axially at the ends. The flow passes the ends of the windings of the_stator, enters the rotor, passes the rotor poles, leaves the motor in a radial way through the stator surface and is collected in the space between stator and pod hull. The entrance to the motor and the transition from pod to strutare large enough and not likely to be hindering the flow. The ducts will be described in more detail further in this section but first the future

possibilities for the cooling will be mentioned.

figure 2.5 Actual flow path

of

cooling air in the ABB Azipod E-motor

Future developments on the cooling of the E-motor

In the previous part of this paragraph, it became clear_{that the transport of large quantities of} air is very inefficient and in fact unwanted. It is _{the opinion of [STNI/LDW 4-1998] that cooling} by water, which is much more effective, will be the preferred technology for future motor designs and they expect a similar development with their competitors, such as ABB and Cegelec.

Cooling with water or another liquid would allow _{for a more elegant hydrodynamic design of} the E-thruster, since no room for large ducts_{has to be reserved in the strut and the pod. The} cooling air in- and output unit_{can be replaced by a much smaller fluid transfer unit and this is} only when the cooling liquid is to be transferred_{to a heat-exchanger outside the pod.}

The other possibility with liquid cooling is _{to keep the complete cooling system inside the} thruster pod and to use the pod surface _{as a heat-exchanger. If the surface area is not enough,} it can be equipped with fins _{or seawater ducts.}

The problem that remains is that the heat that is produced in the rotor-body, has to be transferred to the stator and that the ends of the stator winding, that are not embedded in the stator body, are to be cooled extensively. This cannot be easily achieved with liquid cooling, unless the motor is adapted to work in a "flooded" environment. Holec in the Netherlands designs motors for dredging applications that use this technique, but none of the

manufacturers consulted for this electric thruster does.

Therefore, the water cooling option will, unfortunately, not be further detailed in this study, because only data of air-cooled machines are available. The liquid cooling is promising, but will have to be developed from scratch, according_{to [STN/LDW 4-1998] and is thus not} applicable at this moment.

The above makes it clear that numerous E-thruster concepts can be generated for the cooling alone. The basic requirement for each concept is that it has to be reliable, meaning as less mechanical components installed in the thruster pod as possible. With that in mind, the air cooled Ansaldo solution is a good starting point. The dimensions of the cooling ducts for this motor will be detailed in the following sections

Cooling ducts

Since the value for the necessary airflow is ambiguous, the approach will be to create as much room as possible for the cooling ducts.

From a few quick calculations and estimations, the duct area to be aimed at is set at 1 m2 for each duct, i.e. in- and output. The previously calculated value for the flow of 12 m3/s thus yields a flow rate of 12 m/s, which is large but considered_feasible.

The ducts are situated in the cooling air in- and output unit, the steering pipe or stem section, the strut and around the electric motor in the pod. They will be described in that order. The in- and outgoing flows are supposed to be delivered in two separate channels, that enter the in- and output unit, see figure 2.6_{on the next page. This unit transforms the rectangular} channels into two concentric channels that go through the stem section. The concentricity is needed to provide the transition from the static ship structure to the rotating thruster, which for the rest is a matter of applying seals to prevent air leakage. The resulting concentric channels of the steering pipe are sketched in figure 2.7 on page 21, cross section AA. The total area of the ducts is 2.54 m2, corresponding with an outer diameter of 1800 mm. This diameter can be chosen less, for instance to lower the costs of bearings and construction. However, when the losses due to steelwork (15%)_{and the cable duct (0.1m2) are considered in} the area calculation, the area reduces_{to 2.06 m2, so this is a good starting value for the}

diameter of the steering pipe to meet the duct _{area demand.}

CAE Doc

r

(VO SLitDG- i/v6- UNIT)

IN s-I

;-/ I

twos

, MANSFER, uN1T DU'r TO HEMT EXCHIWGEk

14Tori

ENGINE Roof`'l

figure 2.6 Cooling air in and output unit

The transition from the steering pipe (inside the_{ship) to the strut is the point where the} concentric channels have to change into three separate channels, two for the flow in and one for the flow out. It's the point where the first_{passage difficulties arise, but in order to be able} to further define the cooling ducts, a profile for the strut cross section has to be chosen. This is in the following intermezzo, anticipating on the hydrodynamic properties.

Intermezzo

A thick profile is required for the _{strut section to provide the space for the channels. A thick} rudder profile with a sufficient chord length seems suitable for this purpose. In the table on the next page, the cross sectional areas are given for different chord lengths of appropriate rudder wing sections. In addition, _{an effective area is given, that stretches over 50% of the} chord length around the maximum thickness. This area, minus 15 % to account for the construction parts and cabling channel, is an indication for the area that has to be available to make the transition fromstem section _{to the strut.}

(the complete set of data can be found in appendix C)

It' can be seen from the table, that_{a large chord length combined with the thickest wing} sections, 0021 or 0024, is necessary to accommodate_{the chosen duct area.}

Only two sections provide enough space in the midsection two accommodate the ducts, these are the 0021/c5000 and the 0024/c5000. However, the chord length offive_{meters is too long,} when compared to the dimensions of the pod, so the 0024/c4500 wing section is used for the geometry of the strut. The available area will actually be larger than the _{sketched 50%} midsection.

The proposed transition from steering pipe_{or stem section to strut is 'sketched in figure' 2.7 on} the next page. Special attention _{was given to the geometry of the transition, but the 'solution}

should be verified using a 3D model in_{a further design step, to see if the narrow passage in} the general direction of the flow is not_{causing high pressure losses.}

2ti NA CA

ie

Wing Sections NACA 00244:

r s

s

/Al ra

r -

ter"

AREA EFFECTIVE AREA

_

i 1 1 I ----.___

CI

-

i -

-MAXIMUM NET AREA 0018 0021 ₀₀₂₄

max.

eff

net. max. eff

net

max. eff. net.

1 m2 m2 m2

mi

m2 m2

mi

m2 m2

'chord

I 3000 1.11 [ 0.74 0.53 1.29 0.86 0.63 1 1.47 0.98 0.71 3500 i 1.50 1.00 0.75 1.76 1.17 0.89 2.01 1.34 1.04 4000 1.97 1.31

I.0]

229

1.53 1.20 2.62 1.75 1.38 4500 2.49 1.66 1.31 , 2.90 1.93 1.54 ,' 3.32 2.21

L78

5000

107

2.05 1.64

158

239

1, 1.93 , 4.09 2.73 2.22 1

table 2.3 NACA wing sections Abbott 19581

1800

72 730

30

fiANSALDO

f MOTOR! 2100.04005)

COOLING AIR OUT COOLING AIRIN POO MOO,/

IMI11111r,

kn.

CROSS SECTION A -A S TURING PIPE

IS

CROSS SEC TION 9-8 CONNECTIONSTEERINGPIPE STRUT

'eh\

ka ITO kj 46,

\tir-V

r

A..

CROSS SECTION C-C OTN' 4500 754 in 750 ISO /50 vior 107'1%. 40N 111N '','W110200` e'llb

CONNECTION S TEE R ING PIPE - STRUT

10 N,

7197

713?

LAYOUT OF COOLING DUCTS

PRELIMINARY E- THRUSTER DESIGN 03-10-1998

figure 2.7 _{Cooling ducts in the steering pipe and strut their transitions}

21

-7004---7350

CROSS SEC TION 0-0 STRUT

ANS ALDO

92

0

-The solution that was sketched on the previous page, was created with the use of a large number of sketches and concept ideas and the drawings are to some extend similar to the ones that ABB Azipod provides in literature[ABB 004]. However, they are the result of the

described design process and NOT copied.

Although it seems premature to give so much attention to the cooling system, it is the common opinion that it will be one of the major design restrictions of the E-thruster. Therefore, some more time was spent to create a basic solution that hopefully can be easily adapted to other requirements in flow.

An overview drawing of the cooling ducts in the steering pipe and the strut can be found in appendix F, where larger versions the drawings are collected..

Cooling channels around the motor

In [X40200], a previous study of Lips and Ansaldo figure of 300 mm was stated for the free space around the motor. Since this motor has less than half the power and the 600 mm that is added by the cooling channels would be 20% of the pod diameter, a small approximation' will be made in the following to see if the provided figure is feasible.

figure 2.8 Space around the motor

As has been stated before, the cooling air enters the motor at both ends, goes through the rotor and emerges at the outside surface of the_{stator. The space that is needed between the stator} and the surrounding construction is determined by the narrowest passage 'a', that can be found at the top of the construction, where the pod housing is connected to the strut, as can be seen in figure 2.8.

Effective 'outside dimension of the stator, the area where cooling air emerges:

Ad/. = 85Vo * *L stator* Dstator

Part of the cooling air flow that leaves the top of the_{motor and therefore does not have to go} through the space between pod housing and motor:

°passage = V,arr -_{°direct = °V.air * (I -A50%Naca/Aelf)}

22

From appendix C and the relative position of the motor housing to the strut, it follows that approximately 50%, 2.21 m2, of the Naca 0024/4500 profile is above the effective length of the stator package (A50,4Nacj.

table 2.4 Required space around the motor

Remarks V,air.LDW Picas.LDW7MW 2.5

2942.5

12.3

m3 (1) 60 60 M3

eV,air.Ansaldot= 17.0

are, = Areql-'_stator

The required area, (1),geNak, for the Ansaldo solution is equalto the overall passage that was chosen earlier, because the required airflow was used for the calculation, and not the dower value that was estimated on page 14.

This shows that the value that was provided by Ansaldo, 300 mm, is a little too high, but it leaves some space for the construction parts. For this study, a value of 220 mm is chosen. Several guiding vanes will be needed to minimise the losses and to get the cooling air where necessary. Not described above, but extremely important is the flow over the ends of the stator windings. Since they are not embedded in the_{stator core material and cannot give off heat} through conduction, the cooling air has to take care of the heat removal.

2.4.3 Strut and pod geometry

Strut geometry

The strut geometry is fixed with the NACA 0024/4500 that_{was chosen during the} dimensioning of the cooling channels, in the strut..

Pod geometry

In the preceding study, ,a maximum diameter of the pod was chosen, based on a selection of a propeller. Since the actual dimensions of the electric_{motor and its cooling requirements are} now known, the pod diameter can be estimated _{more accurate. However, support structure of}

23

Ansaldo LOW

Aeff m2 14.3 11.1

I AMP/ NAOS _TI12 2.21 m2

2.21 m2

[ mi3/ s , 17.0 12.3

d),

m3/ I s 12.0 9.8 vor m/s 12 12 Are, m2 1.0 0.82

aq

mm 200

216 OV,air

the motor is still unknown. It will vary a little but not much with the motor diameter, as it is depending on the motor torque and not on the size, when only one power level is considered. The motor diameter still has the largest influence on the diameter of the pod, which in its turn influences the hydrodynamic properties of the E-thruster. This preliminary design will be continued with largest option for the motor. Later, using drag calculations or model tests, it can be tested if the influence of a much smaller motor diameter, as is shown in figure 2.9, is significant or not. Although the Ansaldo drawings in appendix A show that the actual

'footprint' is somewhat smaller than the maximum dimensions, the motor will be considered to be a cylindrical over the full length. This makes the geometry of this preliminary design suitable for any motor that fits within the dimensions of the chosen cylinder.

The 'nose' and 'tail' of the pod are simple transitions from propeller hub to maximum pod diameter and back again, with the tail ending abruptly to prevent swirling.

Adding 2*220 mm, for cooling purposes and 2*90 mm for the stiffeners inside the pod to the motor diameters of Ansaldo, Brush and LDW, results in the values for the diameter of the pod in the sketch below. Measured is on the inside of the pod, as is common in the shipbuilding industry. The optimal value for the Brush and LDW motors is achieved by placing the motor below the centre of the pod, as is shown in the rightmost cross section in figure 2.8.

1 I 2720 ANSALDO 1 2420 BRuSH/LOw 2750 - 8000 15 - _25' 8RuSP-VLOw OPTIMAL 4000

figure 2.9 Global pod geometry

It might be possible to use more elaborate shapes, to create a better pressure pattern around the pod body and to avoid separation, but the frontal area of these shapes is larger, generating more losses due to drag resistance. The friction losses are a function of the wetted surface, S, and the velocity of the flow over the pod. S can be kept as small by choosing a basic shape for the pod body. The velocity is the result of the_{propeller working in front of the pod and cannot} be influenced. The shape of the pod on the previous page is not yet optimal, as can be seen from the sharp transitions or 'shoulders'.

The Schiffbau Versuchsanstalt Potsdam has carried out systematic research for the pod bodies of azimuthing thrusters, and found the following for (mechanical) azimuth thrusters. See figure 2.10 on the next page for the parameter designations.

The total efficiency of the azimuthing thruster will strongly decrease for pod (or gondola as it is named in [Heinke 1998]) diameters that _{are more than % of the propeller diameter. dG/D in}

figure 2.10 should be less then 0.75. In [Eijnatten 19971a value of 0.5 was chosen and the optimum motor diameter was derived from the resulting pod diameter.

This diameter that was calculated above for the pod with the Ansaldo motor results in dc/D being 0.68, with a propeller diameter of 4000 mm, which still would allow for a satisfying hydrodynamic efficiency. The other diameter that is known for the motor, the 1850 mm of the Brush and LDW motors, yields a parameter of dc/D = 0.62 in the worst case. The parameter will more likely have a value of 0.55, which is very acceptable.

According to the same source [Heinke 1998], y should not exceed 25°. For the current pod shape, this value is 15° and thus well under this limit.

The gondola's that were tested at SVA, are shorter than the E-thruster pod. The 1c/13 is 1 or less for mechanical azimuth thrusters and in the order of 1.5 to 2 for E-thrusters (values are estimations based on Azipod and SSP sketches). Due to the long thruster body, the results of model tests for conventional thrusters cannot be used directly for E-thrusters.

figure 2.10 Parameters

of

the pod (gondola) [Heinke 1998]

2.5 Selection of propeller

A propeller for use with this E-thruster could not be designed in time for use in this study. For comparision, an existing propulsion installation, consisting of two FP propeller, shaftlines and bow/stemthrusters, has been chosen. Lips designed and delivered two identical systems to Chantiers de l'Atlantique in the end of 1997 and the beginning of 1998, for cruise ships operated by Renaissance Cruises.

With the aid of the data of this project, the main characteristics and geometrical data of the propeller are fixed. The complete set of data can be found in appendix B.

table 2.5 Renaissance data

2.5.1 Pushing or pulling?

Although the choice for a pulling E-thruster has been made in an earlier stage, in the

following some more attention is given to the choice_{between a pulling and a pushing version} of the E-thruster.

Pulling

- _{The oblique inflow of the propeller when steering the E-thruster will increase the steering}

torque and affect the propeller efficiency.

26

Name unknown Classification

society

Bureau Veritas

Owner

Renaissance Cruises

Class Notation +100A1

Yard

Chantiers de l'Atlantique # H31,131 Lo. _(m)

_Propulsion

plant

Lpp (m) 157.85 Pmcm

[raj

_2*6750 Lw, (m) 161.15 Vivicil [Kn] 20.4

B,

_(m) 25.46 [rpm] 170

Tr

_{(m) 5.83 PCSR [MW] 2*5738} A (ire) 15327 Val:, [Kn] Cb (Lpp) 0.654 [rpm] 161 Cp (Lpp) 0.671

Propeller

# blades _[-] 4

diameter

[mm] 4000 speed (MCR) [rpm] 170 wake _[-I 0.09±0.01 ' ' I

Improved propeller efficiency, since the propeller is operating in a more uniform wakefield.

Wake fraction is lower, thrust deduction is higher, so the hull efficiency will be lower, which has to be compensated by the improved efficiency.

Propeller induced vibrations and noise are lower because of smooth flow in which the propeller operates.

Pushing

The steering torque will not increase as much as for the pulling version of the E-thruster. The pod and strut body will align the flow to some extent, so that the inflow of the propeller is not as oblique as with a pulling propeller.

The hull efficiency for this mode will be better, as the propeller operates in the wake of the pod. Friction losses over the pod will be smaller, but the wakefield will show a strong dip due to the thick strut.

2.5.2 Propulsive efficiency

A comparison between a conventional diesel electric twin screw vessel and an E-thruster driven vessel could be made with the use of the basic equations for the propulsion efficiency. The result would be deceitful, since many parameters are unknown and would have to be guessed. Therefore, the following data is for information purposes only.

Differences

The resistance R of the ship, which is lower for the E-thruster propelled option because

_of

the lack of appendage losses for rudders, stern thruster openings, outboard shafts and brackets. Omitting these appendages will result in a lower resistance of_{up to 10%, this is} the largest part of the efficiency gain of E-thruster propulsion.

The values for wake fraction and thrust deduction. The propellers are working under different flow conditions, as is indicated in the table below. However, the differences in hull efficiency Tim do not seem large.

table 2.6 _{Ship/propeller/pod interaction data}

27 2*7MW conventional DE cruise vessel (Renaissance) (Appendix Al 2*7MW

&thruster

cruise vessel (estimate) 2*14MW Azipods

(Carnival

Elation) IKMY/ABB 0021 Wake fraction w _[-] 0.10 0.08 0.043

Thrust deduction t

and relative rotative

efficiencyllR [-] HI 0.15 0.12 0.082 1.064 Hull efficiency nu (1-t)/(1-w) _{[-] 0.944 0.957 0.959}

Approximations for forward and reverse thrust

The only values needed at this moment for the mechanical design of the E-thruster are approximations for the forward and reverse thrust, and this is described below. The data is coming from the Lips steerable thruster folder for the bollard pull (zero speed) values and from appendix B, the Renaissance data for the service speed values. A small verification carried out by adding values from the manoeuvring handbook of [Brix 1993] and from the Lips X40200 project.

table 2. 7 Thrust/power ratio

The values in the lower part of the table indicate that a proper ratio for the 7 MW unit in this report would be 70 N/kW (0.07 1(1\1/kW), at nominal speed Vs = 20.5 Kn.

Although the ship will be cruising nominal or service speed most of the time, in practice manoeuvring and accelerating are also part of the mission profile. The bollard pull (Vs= 0) is, for 'open' thrusters, considerably larger, with a ratio of 135 N/IcW:

= 7000 * 0.135 = 9451cN

Since the motorspeed, and thus the thrust, can be limited and a large bollard pull is not needed for cruise ships. the maximum thrust is chosen to be:

T= 0.07 * 7000 =490 IcN 500 IcN

Reverse thrust is taken as 75% of the forward thrust. Trcerse = 0.75*500 = 375 IN

This assumption cannot be made for DP systems or large tugs, where a maximum bollard pull is a design requirement.

28

Zero speed Lips Brix

Steerable thruster folder

X40200 Page 63

Vo

Installed power _{Pb [kWi} 6000 16500 -

-ducted open ducted open

Thrust T

_[IN

1080 2640

Thrust/power

Co [N/kW] 180 160 185 135

Nominal speed Renaissance [Brix 19931

Appendix B Page 55 Vs [Kri] 20.5 22.0 23.5 Installed power Pb [kW] 2-6750 9700 22320 Ship resistance RTS [kN] 800 Trust deduction t [-] 0.155 - -Thrust T [IcN 947 698 1483 Thrust / power _[N/kW] 70 72 - 66 I ' ,

2.6 Pod Components

The expectation is, that if the motor with its cooling requirements fits within the pod geometry, all the auxiliary equipment will fit also, because much smaller diameters are concerned.

2.6.1 Bearings

The set of bearings that can be found in the thruster pod is a "normal" shaft bearing system, consisting of a support and a thrustbearing for the combined E-motor-propeller shaft. The difference with a standard shaftline is that the bearings do not only have to carry the weight of the shaft and the forces of the propeller, but also the same items of the electric motor.

- -14- -1+114-1-1 -11 -1-14-- - - -t II-1 II 1_ 1:4- Ljj -- _{_ _} r-.-waIlit -0-4 - H-1-4 I t-t. ',I. U BEARING CONFIGURATIONS

(DLIENsoNs ARE NOT TO SCALE)

I -41-,=F-= +1H-011 -I 1F1-1 " - _a_ IL. -1 tts t , U _r _ -4- n I

f

- 4--- - u_ - 1 III as Mar- -11.-=,-k 71:-:q1:717 71.14ji- -I n --rV"-rr V-r*-1- - - w - - kg V 1B) (0)

--figure 2.12 Bearing configurations 29

:re H-FFI -1 if- 41-.4- -II

I , _ - _

I

U

Both roller and journal (plain) bearings can theoretically be applied in the E-thruster.

Journal bearings can be produced cost effective and large shaft diameters do not influence the availability. In this case, the combination of an electric motor and the frequently changing loads due to the steering capabilities of the pod would cause the bearings to wear out faster than the bearings of conventional shaftlines. This can be avoided by the use of hydrostatic bearings, or at least bearings that are hydrostatic for low propeller speeds. This would add to the complexity of the E-thruster, due to the pressurised oil system that would have to be incorporated in the pod.

Spherical roller bearings have low friction coefficients, even at low speeds, and offer a very precise alignment, necessary to maintain a proper airgap, while still allowing for a low level of misalignment, to compensate for shaft deflections. For diameters of approximately 500 _mm to 1600/1800mm, roller bearings form part of the standard delivery program of several

manufacturers, but only on special demand and not from stock.

For the 7 MW E-thruster with its shaft diameter of about 400 mm, both bearing types or even combinations are possible, but the less complex and thus favourable solution with roller bearings is chosen for this preliminary design.

In figure 2.12 on the previous page, five bearing configurations are displayed.

The configurations C and especially D provide a good access to the (thrust) bearings, with the rear part of the thruster pod being removable. The optional shaft brake, see figure 2.1 on page 4, would have to move to the front of the pod to make room for the bearing block and is less accessible in this position. The propeller thrust is working on the full length of the shaft and this results in less movements of the rotor in relation to the stator, both axially and radially, although the effect of the bending of the shaft will be greater.

The symmetric solution E is more theoretical and in practice unwanted because of the difficulties that arise for the axial alignment, due_{to temperature differences of the shaft.} If sufficient space is available, the configurations A or B would be most the favourable, because the thrustforce is transferred from the propeller shaft to the pod, before the shaft runs through the motor. The accessibility of the bearing block for_{maintenance has to be ensured,} but this can be considered during the engineering of the pod steel work.

Bearing calculations

In Appendix D, a rough estimate has been made_{to determine the forces that act on the} bearings, to see if the requirements_{can be covered with 'standard bearings. A summary of} the calculations is given at the end of the appendix and repeated in table 2.12 on the next page of this report.

Because the bearings will be installed in the, relatively thin, pod housing, a spherical roller bearings seems to be the appropriate type of roller bearing, as this bearing type can

accommodate some misalignment.

The data in table 2.8 on the next_{page, for the both the shaft and thrust bearings. partly can be} found in the SKF "Large Bearings" catalogue.

The bearing selection process provides rough estimates only, and depends strongly on several assumptions for load, lubrication and operating environment.

Continuous oil lubrication is foreseen to cool the bearings that are sealed from the rest of the pod to create an optimal operating environment.

The bearing configuration and the way the bearings are installed do also influence the bearing life. In this case, the bearing capacities strongly increase with the diameter, because of the necessary diameters, as the thrust bearings are supposed to be installed on the shaft first and then the forward motor bearing.

When more is known of the E-motor and the propeller, the bearing selection will have to be performed again, so that a more accurate selection can be made. Even with a careful selection, almost using no assumptions, the result should be compared to a bearing selection based on experience.

The table above however, in combination with the catalogue, shows that suitable bearings can be found for this application without encountering_{technical impossibilities.}

31

Bearing position

front rear thrust reverse thrust

Type 23080 CAC/W33 23980 CAC/W33 29384 29288

Inner

diameter

D m m 400 400 420 440 Mass M kg 150 75.5 170 78

Rating

C kN 2880 1730 3740 2070

CT

[-] 9.863 8.693 8.113 6.900 Bearing life _Y 16.1 10.5 8.4 4.8

Lubrication

oil oil oil oil

table 2.8 Bearing type and data for SKF large bearings

I

II

226.2 Shaft holding brake

Depending on the type of electric motor that is installed, a shaft holding brake has to be added to prevent the rotor and propeller from turning in specific operating conditions. The brake is

not meant for ship deceleration, as this is very well feasible by controlling the input frequency of the motor. As has been described in [Eijnatten 1997],the synchronous motor with external

excitation is the only motor of which the rotor can be blocked by applying DC current to both stator and rotor. All other types, i.e. synchronous PM, brushless synchronous and

asynchronous motors have rotors that cannot be electrically fixed and thus require a means of stopping the shaft from rotating..

The special cases in which a shaft brake is needed (a twin propeller cruise ship is considered):, During towing operations

When manoeuvring with only one thruster in operation When using only one unit because of failure of the other The reasons for blocking of the propeller are:

Avoiding damage to the bearing system, although this is limited because the (spherical roller) propeller/shaft bearings are always properly loaded with the rotor and propeller weight. The (spherical roller) thrustbearings are also pre= loaded.

Avoiding the feedback of power to the ship, for instance when maintenance of any part of the propulsion system takes place (slipring brushgear, converter components, control electronics)

It has to be determined to what (ship)speed limit a blocked propeller would be the most feasible and at what speed a trailing propeller is inevitable or acceptable, which can be performed best with the use of a four-quadrant diagram for ship and propeller, which is not 'available at this moment.

Inquiry with the electric motor supplier or manufacturer will determine if_{a separate shaft} holding brake is necessary Fixation of the propeller is possible when keeping the E-motor _at zero or very low speed with the frequency controller,, but this is of course not possible in case of a motor failure..

'The space for the shaft holding brake is reserved, but there is no actual brake selected. This would' require a four quadrant diagram of ship and propeller to be available, which is not the case, as the detailed propeller design still has to be made.

2.6.3 Shaft Seal

The short shaft line of the E-thruster allows the application of a conventional lip sealing method, that will be even more effective, due to the small deformations. The ABB Azipod is equiped with a four rubber lip outer seal. For this preliminary design, an outer seal with four lips and an inner seal with two lips has been chosen, similar to a John Crane Marine Lips (JCM-L) 4BL shaft seal, see appendix C.3. JCM already engineered such a seal as an alternative for the current Azipod seals, but did not sell one yet. Shaft seals that are build up like this are not pollution free, so if very strict regulations are to be met, another type of seal has to be applied.

Although the technical solution is available, there's no Azipod with a pollution free shaft seal at this moment.

figure 2.13 Possible 4 + 2 lip shaft seal arrangement for the E-thruster