Dimensioning of marine propulsion shafts

Whirling and torsional vibrations

Master’s thesis in the international master’s programme Naval Architecture and Ocean
Engineering

GUÐNI PÁLL GUNNARSSON

SINDRI PÁLL SIGURÐSSON

DEPARTMENT OF MECHANICS AND MARITIME SCIENCES

C HALMERS U NIVERSITY OF T ECHNOLOGY
Gothenburg, Sweden 2021
www.chalmers.se
 Master’s thesis 2021

Dimensioning of marine propulsion shafts

Investigation on marine propulsions shaft diameter, from

classification societies. To see if the diameter can be decreased,
compared with current calculation methods.

GUNNARSSON & SIGURÐSSON

Department of Mechanics and Maritime Sciences

Division of Marine Technology
Chalmers University of Technology
Gothenburg, Sweden 2021
Master’s Thesis 2021:29
Investigation on marine propulsions shaft diameter, from different classification so-
ciety. To see if the diameter can be decreased, compared with current calculation
methods.
GUÐNI PÁLL GUNNARSSON
SINDRI PÁLL SIGURÐSSON

© GUÐNI PÁLL GUNNARSSON & SINDRI PÁLL SIGURÐSSON, 2021.

Supervisor: Tobias Huuva, Berg Propulsion Production AB

Examiner: Per Hogström, Mechanics and Maritime Sciences

Master’s Thesis 2021

Department of Mechanics and Maritime Sciences
Division of Marine Technology
Chalmers University of Technology
SE-412 96 Gothenburg
Telephone +46 31 772 1000

Cover figure:
Shaft line setup based on one of the case study shaft line of the report, simulated
in Nauticus Machinery Shaft alignment from DNV.

Typeset in LATEX, template by Magnus Gustaver

Printed by Chalmers Reproservice
Gothenburg, Sweden 2021

iv
Dimensioning of marine propulsion shafts
Investigation on marine propulsions shaft diameter, from different classification so-
cieties. To see if the diameter can be decreased, compared with current calculation
methods.
GUÐNI PÁLL GUNNARSSON & SINDIR PÁLL SIGURÐSSON
Department of Mechanics and Maritime Sciences
Chalmers University of Technology

Abstract
Dimensioning of marine propulsion shafts for new builds is done early in the design
phase of a vessel. Thus, often the designer only has the main particulars of the vessel
including the general size and power requirements, as well as the intended purpose
of the vessel. With the limited information, a shaft line needs to be dimensioned
which is fit for purpose, fulfils class requirements as well as showing good vibrational
behavior. Until now this has been done safely and successfully using old principles
and fundamentally relying on a single formula that is recognized by all the major
classification societies for giving a safe design. This method has been shown to
give very reliable designs but is however believed to result in overly dimensioned
components. By utilizing expertise, access to good data, and a systematic approach
in behavior analysis, it is believed to be possible to step away from current methods
and provide slimmer and more power-dense shaft lines.
This thesis describes the underlying theory of rotating machinery often summarized
as rotor dynamics. A comprehensive study of classification rules and requirements
governing the design of marine propulsion shafts. Case studies of typical shaft lines
for different types of vessels, for those a software tool from DNV-GL is used to
simulate the shaft behavior. The simulation software Nauticus Machinery available
from DNV-GL is a specialized marine shaft line analysing software, containing few
different tools. The tools used during this thesis work are Nauticus Shaft Alignment,
Nauticus Torsional Vibration, and approved formulas and equations from classifica-
tion societies. Using this combination of software and formulas the writers evaluate
an improved method to define the minimum shaft diameter.

Keywords: propulsion shaft, whirling, torsional vibration, shaft alignment, shaft

dimensioning, classification societies.

v
Acknowledgements
This thesis constitutes part of the requirements for the master’s degree in Naval
Architecture and Ocean Engineering at Chalmers University of Technology, Gothen-
burg, and has been carried out with Berg Propulsion Production AB, Öckerö, be-
tween January and June of 2021.
We would like to express our gratitude to our supervisor Tobias Huuva, Ph.D., and
Samuel Ottosson for all their support and guidance throughout the project. We
would like to thank our examiner Per Mottram Hogström, Director of the Master
program in Naval Architecture and Ocean Engineering, for his guidance and mo-
tivating enthusiasm. Special thanks go to Deniz Yagiz and Nancy Phang for their
expert support and feedback on the details of rotor dynamics analysis.

Many thanks to all the people at Berg Propulsion for a warm welcome and for
showing interest in our project.

Guðni Gunnarsson & Sindri Sigurðsson, Gothenburg, May 2021

vii
 Contents

List of Figures xiii

List of Tables xvii

Abbreviations xix

1 Introduction 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Research question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 Theory 3
2.1 Shaft line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.1 Low-speed shaft . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.2 High-speed shaft . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1.3 Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1.4 Sterntube & Propeller shaft . . . . . . . . . . . . . . . . . . . 7
2.1.5 Intermediate shaft . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.6 Propeller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.7 Bearings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Static-loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.1 Breakdown of the IACS M68.4 formula . . . . . . . . . . . . . 11
2.2.2 Stress concentration . . . . . . . . . . . . . . . . . . . . . . . 12
2.3 Dynamic-Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.3.1 Lateral vibrations - Whirling . . . . . . . . . . . . . . . . . . . 13
2.3.2 Torsional vibrations . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3.2.1 Damping in TVC . . . . . . . . . . . . . . . . . . . . 18
2.3.2.2 Application factor KA . . . . . . . . . . . . . . . . . 19
2.3.2.3 Estimating the application factor . . . . . . . . . . . 21
2.3.3 Axial vibration . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3.4 Fatiuge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.3.4.1 Fatigue Stress Concentration Factor . . . . . . . . . 25
2.4 Classification rules & guidelines . . . . . . . . . . . . . . . . . . . . . 25
2.4.1 IACS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4.2 DNV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.4.2.1 ICE-class application . . . . . . . . . . . . . . . . . . 28

ix
Contents

2.4.2.2 DNVGL-CG-0038 . . . . . . . . . . . . . . . . . . . . 28
2.4.3 Classification rule requirements for shafting alignment. . . . . 31
2.4.3.1 DNV - Det Norske Veritas . . . . . . . . . . . . . . . 31
2.4.3.2 LR - Lloyd’s Register . . . . . . . . . . . . . . . . . 32
2.4.3.3 BV - Bureau Veritas . . . . . . . . . . . . . . . . . . 33
2.4.3.4 ABS - American Bureau of Shipping . . . . . . . . . 33

3 Methods 35
3.1 Study on the IACS M68 formula . . . . . . . . . . . . . . . . . . . . . 35
3.1.1 Static calculation comparison . . . . . . . . . . . . . . . . . . 35
3.1.2 Fatigue approach on IACS M68 . . . . . . . . . . . . . . . . . 37
3.1.3 IACS M68.4 Concentration factor . . . . . . . . . . . . . . . . 38
3.2 Simulation analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.2.1 Nauticus Shaft Alignment . . . . . . . . . . . . . . . . . . . . 43
3.2.1.1 Building the shaftline model . . . . . . . . . . . . . . 43
3.2.2 Nauticus Torsional Vibration . . . . . . . . . . . . . . . . . . 46
3.2.2.1 Inertia and stiffness . . . . . . . . . . . . . . . . . . 46
3.2.2.2 Damping . . . . . . . . . . . . . . . . . . . . . . . . 48
3.2.2.3 Propeller . . . . . . . . . . . . . . . . . . . . . . . . 50
3.3 The improved diameter approach . . . . . . . . . . . . . . . . . . . . 51
3.3.1 The method . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.3.2 Case study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.3.3 Additional input values . . . . . . . . . . . . . . . . . . . . . . 55
3.3.3.1 Dimensions of each component . . . . . . . . . . . . 55
3.3.3.2 Surface roughness of shaft . . . . . . . . . . . . . . . 55
3.3.3.3 Propeller and ice-applications . . . . . . . . . . . . . 55
3.3.3.4 Material . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.3.4 Calculator guide . . . . . . . . . . . . . . . . . . . . . . . . . 56

4 Results 57
4.1 IACS M68 compared with the mimic-formula . . . . . . . . . . . . . . 57
4.2 Optimized dimensioning of propulsion shafting . . . . . . . . . . . . . 59
4.2.1 Comparison on finished Berg orders . . . . . . . . . . . . . . . 59
4.2.1.1 The Tanker shaft . . . . . . . . . . . . . . . . . . . . 61
4.2.1.2 The Dredger shaft . . . . . . . . . . . . . . . . . . . 62
4.2.1.3 The RoPax ferry shaft . . . . . . . . . . . . . . . . . 63
4.2.2 Weight reduction of raw material . . . . . . . . . . . . . . . . 64
4.2.2.1 Sensitivity-DNV CG 0038 calculations . . . . . . . . 65
4.2.3 Shaft alignment - whirling analysis . . . . . . . . . . . . . . . 66
4.2.3.1 Ro-Pax . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.2.3.2 Bunker tanker . . . . . . . . . . . . . . . . . . . . . 67
4.2.3.3 Dredger . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.2.3.4 Result for the case study . . . . . . . . . . . . . . . . 69
4.2.3.5 Alignment Tanker without ICE-class . . . . . . . . . 70
4.2.4 Torsional vibration calculation . . . . . . . . . . . . . . . . . . 71
4.2.4.1 Ro-Pax . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.2.4.2 Bunker tanker . . . . . . . . . . . . . . . . . . . . . 77

x
 Contents

4.2.4.3 Dredger . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.2.4.3.1 Free running empty condition . . . . . . . . 82
4.2.4.3.2 Dredging condition . . . . . . . . . . . . . . 85

5 Future work 89
5.1 More thorough alignment investigation . . . . . . . . . . . . . . . . . 89
5.2 Propeller failure load criteria . . . . . . . . . . . . . . . . . . . . . . . 89
5.3 More precise investigation of the case study . . . . . . . . . . . . . . 90

6 Conclusion 91

References 95

A Appendix-Screenshots I
A.1 DNVGL-CG-0038 flowchart . . . . . . . . . . . . . . . . . . . . . . . I
A.2 Main instruction manual script . . . . . . . . . . . . . . . . . . . . . V
A.3 DNVGL-CG-0038 document calculator . . . . . . . . . . . . . . . . . VIII
A.4 Shaft lines (without components) dimensions . . . . . . . . . . . . . . X

B Appendix step-by-step deduction XIII

B.1 Torsional-static step-by-step deduction making mimic formula . . . . XIII
B.2 Mimic formula- step-by-step deduction with fatigue limit criteria ap-
plied . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XIV
B.3 KAice Estimate ice hit torque on shaft . . . . . . . . . . . . . . . . . XV
B.4 Failure load of propeller . . . . . . . . . . . . . . . . . . . . . . . . . XVI

C Appendix- Data XVII

C.1 Mimic formula compared with IACS M68.4 . . . . . . . . . . . . . . . XVIII
C.2 Shaft diameter investigation input and output . . . . . . . . . . . . . XXI
C.3 Case study result data from Nauticus Alignment . . . . . . . . . . . . XXII
C.4 Case study - Effects of propeller inertia change on flexible coupling
alternating torque. . . . . . . . . . . . . . . . . . . . . . . . . . . . . XXVI
C.5 Input data for Nauticus Machinery Analysis . . . . . . . . . . . . . . XXXIII
C.5.1 Shaft Alignment model inputs . . . . . . . . . . . . . . . . . . XXXIII
C.5.2 Torsional Vibration Calculation model inputs . . . . . . . . . XXXIII
C.5.3 Propeller curves for the dredger . . . . . . . . . . . . . . . . . XLII

xi
Contents

xii
 List of Figures

2.1 The figure describes the setup of a low speed shaft line, directly coupled
to a low speed engine. In a shaft alignment calculation the crankshaft
in this case is modeled with the propulsion shafting . . . . . . . . . . 4
2.2 The figure descripes the setup of a low speed shaft line with reduction
gear. Simplified beam schema is presented above. In a shaft alignment
calculation the output shaft and gear of the reduction gear is included 5
2.3 Typical high speed shaft line, directly coupled to a high speed diesel en-
gine. MTA thruster unit coupled to a CAT 3516 engine with an engine
mounted clutch. Gears located in the top housing of the thruster unit
by the input shaft and also in the leg of the thruster by the propeller
shaft. Curtesy of Berg Propulsion Production AB. . . . . . . . . . . . 5
2.4 3D rendered image of a propeller shaft and a stern tube. The stern
tube is somewhat transparent. Stern tube bearings highlighted blue.
Courtesy of Berg Propulsion AB. . . . . . . . . . . . . . . . . . . . . 7
2.5 3D rendered image of two intermediate shafts (gray-green). Courtesy
of Berg Propulsion AB. . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.6 A common combination of load cases on a circular body, With on-
plane distortion theory [5]. Figure edited with image from [2], . . . . 10
2.7 Geometry of rotating disk on a shaft, supported by bearings. The
angular velocity of the shaft is ω, the center of mass G is outside
of the neutral axis. The thin dotted lines denote the deflection of
the shaft, during whirling vibration [22, p. 405]. Drawing inspired by
graphics in Inman (2001). . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.8 Notation for whirling calculations. [22, p. 405] Drawing inspired by
graphics in Inman (2001). . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.9 Normalized deflection amplitude versus frequency ratio for different
amount of damping. The resonance behaviour is clearly demonstrated
close to frequency ratio of one. The steady state amplitude becomes
equal to the offset of the center of mass [22, p.407]. . . . . . . . . . . 16
2.10 The figure describes the twist of a shaft transmitting torque and re-
lation between, torque, twist angle and position [22, p. 452]. Drawing
inspired by graphics in Inman (2001). . . . . . . . . . . . . . . . . . . . 17
2.11 Torsional vibratory stress for different load scenario and its relation
with mean stress tau0 and KA, in time domain. Based on direct
coupled plant. Image gotten from DNVGL-CG-0038 [10] . . . . . . . 21

xiii
List of Figures

2.12 Example SN- diagram for a steel material with LCF range up to 103
and enurance limit at 106 . The x-axis is number of cycles in log scale
and y-axis is the fatigue stress range [17] [5]. . . . . . . . . . . . . . . 22
2.13 Simplified stress amplitude for two different loadcases . . . . . . . . . 24
2.14 Torsional stress based on torsional vibration (τv ) and the torsional
stress at specific speed (τ ).Image was taken from DNVGL-CG-0038
[10] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

3.1 Von Mises FEM analysis of the RoPax vessel. . . . . . . . . . . . . . 37

3.2 Table of k and Ck factors from IACSM68 document [21] . . . . . . . 39
3.3 The three SCF approaches done in this study . . . . . . . . . . . . . . 40
3.4 Visual comparison of each calculated K (SCF/1.45), both as the crit-
ical factor using the three approaches previously described. And the
Berg (finished orders) vessels. . . . . . . . . . . . . . . . . . . . . . . 40
3.5 Model of the shaft arrangement in the Nauticus - Shaft Alignment
tool. The model represents the shaftline for the bunker tanker. . . . . 44
3.6 Model of the shaft arrangement in the Nauticus - Shaft Alignment
tool, 3D view. The model represents the shaftline for the bunker tanker. 45
3.7 Simple view of Nauticus torsional vibration model. The mass elastic
model of the Tanker. . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.8 Detailed view of Nauticus torsional vibration model. Showing the dif-
ferent modeling elements along with values for inerta, stiffness and
damping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.9 The propeller torque curves for the two conditions after measurements
in AutoCad. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.10 The study cases on the three-shaft lines under investigation. Note,
those three shaft lines are not in 1:1 ratio of each other. . . . . . . . . 54
3.11 Few screenshots from DNVGL-CG-0038 document calculator. . . . . . 56

4.1 Comparison of the IACS M68.4 formula and the writer’s mimic for-
mula of that one. The mimic formula is labeled ’New DV’ (new Dang
Van). With all design parameters the same and having input power
the variable. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.2 Minimum diameter for IACS M68.4 and writers mimic formula, using
random generated input data. Data values (input and output) can be
found in appendix C section C.1 . . . . . . . . . . . . . . . . . . . . . 59
4.3 Minimum allowed diameter from each approach, for all case study
shaft locations, for comparison. Blue cross marks the final design
from BERG, red bar the final design in this thesis investigation(our
improved method). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.4 Dimensions for the Berg design of the shaft and the currently im-
proved dimensions. The material that can be removed/also represent
the initial design is highlighted yellow on the shafts on the image. . . 62
4.5 Dimensions for the Berg design of the shaft and the currently im-
proved dimensions. The material that can be removed/also represent
the initial design is highlighted yellow on the shafts on the image. . . 63

xiv
 List of Figures

4.6 Dimensions for the Berg design of the shaft and the currently im-
proved dimensions. The material that can be removed/also represent
the initial design is highlighted yellow on the shafts on the image. . . 64
4.7 Visual comparison of the weight/volume of raw steel that could be
saved from the three orders . . . . . . . . . . . . . . . . . . . . . . . . 65
4.8 Unitless graph showing the effect of the inner diameter bore in relation
to the outer diameter on resulted safety of a shaft design . . . . . . . 66
4.9 Shaft deflection for all operating condition, side by side comparison
for the RoPax vessel. . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.10 Whirling modes in forwarding condition, side by side comparison for
the RoPax vessel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.11 Shaft deflection for all operating condition, side by side comparison
for the Bunker tanker. . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.12 Whirling modes in forwarding condition, side by side comparison for
the Bunker tanker vessel. . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.13 Shaft deflection for all operating condition, side by side comparison
for the Dredger vessel. . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.14 Whirling modes in forwarding condition, side by side comparison for
the Dredger vessel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.15 Reaction for Tanker shaft line when The propeller failure ICE criteria
is not considered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.16 Simple schema representing the RoPax shaftline arrangement. Con-
nections marked with C represent clutches. . . . . . . . . . . . . . . . 72
4.17 Heat dissipation from the engine damper, indicating the work done by
the damper. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.18 Engine damper alternating torque, the thick blue line shows synthesis
of vibratory torque amplitudes measured at the damper element. The
thinner red and cyan represent the most prominent excitation modes. 74
4.19 Main flexible coupling heat dissipation. . . . . . . . . . . . . . . . . . 75
4.20 Main flexible coupling alternating torque. . . . . . . . . . . . . . . . . 75
4.21 Main gear alternating torque. . . . . . . . . . . . . . . . . . . . . . . 76
4.22 Propeller shaft torsional stress amplitudes for normal running engine
and misfiring engine. . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.23 Propeller shaft nominal torque and upper and lower torque synthesis,
where the upper and lower synthesis diverge from the nominal indi-
cates fluctuating shaft torque. Comparison of original diameter to
-10mm, -20mm and -30mm decrease. . . . . . . . . . . . . . . . . . . 77
4.24 Propeller shaft torsional stress amplitudes, comparison of original di-
ameter to -10mm, -20mm and -30mm decrease. IACS M68 torsional
stress limits plotted also, as can be seen the stress levels are well below
the limits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.25 Vibratory torque amplitudes in motor coupling, comparison of original
diameter to -10mm, -20mm and -30mm decrease. Limits for vibratory
torque is set at 4.1 kNm. . . . . . . . . . . . . . . . . . . . . . . . . . 80

xv
List of Figures

4.26 Heat dissipated in motor coupling, comparison of original diameter to

-10mm, -20mm and -30mm decrease. Limits for heat dissipation or
power loss is set at 0.480 kW. . . . . . . . . . . . . . . . . . . . . . . 81
4.27 Vibratory torque amplitudes in motor coupling, limits for vibratory
torque is set at 10 kNm. . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.28 Heat dissipated in motor coupling, limits for heat dissipation or power
loss is set at 0.780 kW. . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.29 Nominal torque in main gear with upper and lower torque synthesis.
The alternating torque demonstrated by the upper and lower synthesis
are showed to be small for the whole speed range. . . . . . . . . . . . . 83
4.30 Torque amplitudes in the main gear meshing. Excessive alternating
torque in the messing indicates gear hammering, this would result in
high forces on the gears and shorter life. The results show torque well
within limits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
4.31 Intermediate shaft torsional stress amplitudes, IACS M68 torsional
stress limits plotted also, as can be seen the stress levels are well below
the limits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.32 Propeller shaft nominal torsional stress with upper and lower alter-
nating stress synthesis. . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.33 Propeller shaft torsional stress amplitudes, IACS M68 torsional stress
limits plotted also, as can be seen the stress levels are well below the
limits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.34 Vibratory torque amplitudes in motor coupling, limits for vibratory
torque is set at 10 kNm. . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.35 Heat dissipated in motor coupling, limits for heat dissipation or power
loss is set at 0.780 kW. . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.36 Nominal torque in main gear with upper and lower torque synthesis.
The alternating torque demonstrated by the upper and lower synthesis
are showed to be small for the whole speed range. . . . . . . . . . . . . 86
4.37 Torque amplitudes in the main gear meshing. Excessive alternating
torque in the messing indicates gear hammering, this would result in
high forces on the gears and shorter life. The results show torque well
within limits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.38 Intermediate shaft torsional stress amplitudes, IACS M68 torsional
stress limits plotted also, as can be seen the stress levels are well below
the limits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.39 Propeller shaft nominal torsional stresses plotted with upper and lower
alternating stress synthesis. . . . . . . . . . . . . . . . . . . . . . . . 88
4.40 Propeller shaft torsional stress amplitudes, IACS M68 torsional stress
limits plotted also, as can be seen the stress levels are well below the
limits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

6.1 Flow chart of the thesis process . . . . . . . . . . . . . . . . . . . . . 92

xvi
 List of Tables

2.1 Notation for TVC formulation. . . . . . . . . . . . . . . . . . . . . . 18

2.2 Notation and definitions for propulsion system calculations. . . . . . . 19
2.3 DNV definition for estimated application factor for different system
setup[11] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3.1 Input parameters used for the calculation comparison. Operating RoPax
ferry, from a finished order from BERG database . . . . . . . . . . . 36
3.2 Calculation results from the IACS M68.4 formula, the statical torsion
approach. Response statical stress from torsion and bending on the
shaft, based on Von Mises and FEM method . . . . . . . . . . . . . . 36
3.3 Summary of Nauticus Machinery calculation packages . . . . . . . . . 42
3.4 The table summarises drawings and specifications needed for building
the shaft analysis model in lateral. . . . . . . . . . . . . . . . . . . . . 44
3.5 List of the basic modeling elements availible in the DNV Nauticus
Machinery - Torsional Vibration tool. [13, p. 16] . . . . . . . . . . . . 47
3.6 Comparison of how damping values and coefficients are defined for
the programs Nauticus Machinery and Shaft Designer [13] [4]. . . . . 49
3.7 Conversion table for different damping values, relevant for Nauticus
Machinery [13].The damping magnifier M as used by Nauticus is more
often presented as Q. More comprehensive conversion tables available
in the Nauticus and ShaftDesigner documentation. . . . . . . . . . . . 49
3.8 The different possibilities for defining the propeller damping. . . . . . 50
3.9 Notation for propeller torque and power for the propeller curve con-
version. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.10 Each study location label of the shaft lines . . . . . . . . . . . . . . . 53

4.1 Description of abbreviations used in figure 4.3 . . . . . . . . . . . . . 61

4.2 Steel weight reduction on each shaft line . . . . . . . . . . . . . . . . 64
4.3 Critical speed RPM of the shaft line, before diameters, were decreased
and after optimisation. . . . . . . . . . . . . . . . . . . . . . . . . . . 70

C.1 Propeller torque values from given curves and calculated power for
given propeller speed. . . . . . . . . . . . . . . . . . . . . . . . . . . . XLII

xvii
List of Tables

xviii
 Abbreviations

ABS American Bureau of Shipping

CAT Caterpillar Inc
COO Chief operating officer
CPP(CP) Controllable pitch propeller
Ch Chapter
DNV(DNVGL) Det Norske Veritas
DV Dang Van
Est Estimated
FEM Finite Element Method
FPP(FP) Fixed Pitch Propeller
GB Gear box
HCF High cycle fatigue
HSS High Speed Shaft
IACS International Association of Classification Societies
ICE Ice-classification
KR Korean Register of Shipping
LCF Low cycle fatigue
LSS Low Speed Shaft
MCR Maximum Continuous Rating
ME Main engine
MTA Marine Thruster Azimuth
PRS Polish Register of Shipping
Pt Part (as in: section)
RS Russian Maritime Register of Shipping
RoPax Roll-on and roll-off Passenger vessel
PTO Power take off (connection on gearbox)
RPM (rpm) Round per minute
SCF Stress consentration factor
SN(-diagram) Plot of stress (S) against the number of cycles (N)
TVC(TVA) Torsional vibration calculations (- analysis)
List of Tables

xx
 1
Introduction

Marine propulsion systems are designed in accordance with industry standards from
classification societies [3], which in turn are developed by modeling and experience
from the field. Increasing accuracy in numerical models allow for more detailed
studies on the phenomena of the operation of propulsion systems and, in turn,
optimization of the design.

1.1 Background
Berg Propulsion is one of the world’s leading designer and manufacturer of Control-
lable Pitch Propulsion systems, Azimuthing Thrusters, and Transverse Thrusters.
It is a Swedish company with main office and manufacturing on the Swedish west
coast as well as sales and service offices located in Shanghai, Dubai and Singapore.
The company was founded in 1912 and started to build propellers in 1929. In 2013,
Johan Walter Berg AB and the brand name Berg Propulsion was acquired by Cater-
pillar Inc, also known as CAT. Caterpillar is one of the worlds largest manufacturer
of heavy equipment and machinery. For the marine industry Caterpillar is a sup-
plier of both generators and main engines under the brands, Caterpillar and MaK.
With the acquisition, Berg Propulsion became part of Caterpillar Marine under the
brand of CAT Propulsion Systems. With this Caterpillar was now able to offer com-
plete power and propulsion package. June 30th 2020, the Berg Propulsion brand
is brought back to the scene when Swedish investors with Stefan Sedersten, former
COO of Berg Propulsion, bought back the company from Caterpillar. [6] [7] [30]
[29]
The main propulsion systems usually consist of blades, hub, shafting, control sys-
tem, and some integration components based upon customer need. Analyses of the
behaviour of the propulsion shaft lines are carried out in an increasingly systematic
way giving a good insight in vibration behaviours and resonances. Having access
to good data and experties at the company Berg Propulsion, gives the potential of
developing new procedures for dimensioning of shaftlines.

1.2 Aim
The project aim is to develop current calculation methods that Berg Propulsion
uses to calculate the shaft diameter for low-speed shaft. The current method of
calculating the low-speed shaft at Berg Propulsion is to use the DNV version of
the IACS (International Association of Classification Societies )formula. That is a

1
1. Introduction

conservative old, simple formula that describes the minimum allowed diameter and
maximum hollow bore. Though the formula states that the equation defines the
minimum diameter, there are some rumors in the industry that you can have the
diameter lower and still get it approved by the classification society. By having the
shaft diameter smaller, not only are savings on expenses on material weight, but also
savings on the mechanical equipment on the shaft. Smaller flanges, smaller bearing,
and so on
• The aim is to look into the the basics and background of classification rules
with the intention of establishing a refined calculation method for the dimen-
sioning of shaft lines.

1.3 Research question

Can you make shaft diameter smaller than current calculations methods and get
approval from the classification society ?

- How much room is there for weight/design optimisation in the propulsion shaft
line ?
- How much cost reduction is within reasonable reach by optimising the shaft line ?
- Can the new improved approach be used for preliminary design, or is it too exten-
sive ?

The thesis goes over the path and the surrounding environment that was undergone,
in the development of the new improved minimum shaft diameter method.

1.4 Limitations
• Definition of a success in the project is to decrease the diameter of the propul-
sion shaft and get it approved by the classification society.
• Calculation/simulation results are limited to classification society approval.
• The project is only focusing on the shaft itself (only low speed shaft), though
other components do influence the behaviour of the shaft. Foremost compo-
nents like the propeller, engine, and bearings.
• Rules and regulations in respect of the scope of the project. That is, low-speed
shaft, ductile material, and maximum ultimate strength of the material(some
equations are only applicable for a certain range of ultimate strength).
• Software license limitation. Because of license limitations, simulation of a
2-stroke engine TVC could not be performed.

2
 2
Theory

This chapter provides the theoretical basis for the thesis work. It starts by describ-
ing different mechanical components of a shaft line. Following is an introduction of
the theory behind the statical and the dynamic loads that are present during an op-
eration. Important dynamic design phenomena theories, the vibrational behaviour,
and fatigue. Lastly, the foremost rules and classifications that were used from IACS
and DNV societies.

2.1 Shaft line

A shaft line is the collection of machinery components needed to propel a ship,
more specifically the components that serve the purpose of transmitting the power
from the prime mover to the propeller. The propulsion shafting is the main concern
of this project but it is crucial to recognise the different components making up a
complete shaft line. The purpose of the propulsion shaft is to transmit torque from
the engine or motor for rotating the propeller. In the case of a typical propeller
shaft, the shaft is also a structural member that supports all the weight and forces
from the propeller. This is described as a low-speed shaft below. In the case of
high-speed shafts, the shaft is more similar to a driveshaft in a heavy-duty truck.
Its only purpose is to transfer the mechanical power or torque from the engine to
the propeller [16].

2.1.1 Low-speed shaft

Low-speed shaft - LSS, like the name implies, is a shaft that rotates at a low rota-
tional speed. But without any reference, a low speed does not mean anything. For
a shaft line setup, the low speed would be the rotational speed of the propeller. The
engines and electric motors often operate at higher rotational speed than the de-
signed speed of the propeller. In this case reduction gears are required [16], and are
typically positioned right by the engine. Connected to the output of the reduction
gear is the shaft which the propeller is then attached to.
The speed range for a propeller in a commercial vessel could be from 70 to 250
revolutions per minute. The limit has been getting lower with an increased size of
propellers on the very large cargo vessels at the same time as the low-speed two-
stroke propulsion engines have gotten larger, more powerful and more fuel-efficient.
Low speed 2 stroke engines operate at the same speed as the propeller and are thus
directly connected, as seen in figure 2.1, where the engine crankshaft is considered

3
2. Theory

an indirect continuum of the propulsion shafting. Looking at the large MAN B&W
engines they operate at 70 - 80 rpm [24] and similar to the WinGD 2 stroke diesel
engines. WinGD building on the tradition of Sulzer diesel engines and Wärtsilä
two-stroke [35]. One example, the record setting WinGD 12X92DF engine deliver-
ing power output rated at 63,840 kW at 80 rpm and its performance measures a
thermal efficiency of 51-53% [36].

Figure 2.1: The figure describes the setup of a low speed shaft line, directly coupled
to a low speed engine. In a shaft alignment calculation the crankshaft in this case is
modeled with the propulsion shafting

In figure 2.2 an LSS shaft line is demonstrated, incorporating a reduction gear, con-
trollable pitch propeller, CPP, and typically a medium-speed engine. In comparison
with the directly coupled shaft, this setup is more common in ferries, fishing vessels
and different types of specialised service and research type vessels. The setup gives
more flexible operation and quicker response. The control mechanics for rotating
the blades on a CPP propeller are often located on the reduction gearbox and the
mechanical actuation is done by hydraulics or shaft running through the middle
of the propeller shaft. By having the possibility to rotate the blades, effectively
changing the pitch of the propeller, changing from forward to reverse thrust can be
done seamlessly as well as controlling the load on the engine according to operation
modes.

4
 2. Theory

Figure 2.2: The figure descripes the setup of a low speed shaft line with reduction
gear. Simplified beam schema is presented above. In a shaft alignment calculation
the output shaft and gear of the reduction gear is included

2.1.2 High-speed shaft

In contrast to the earlier discussion about LSS, high-speed shafts are usually directly
coupled to medium or high-speed diesel engines, hence the high speed notation,
they operate at the speeds of the propulsion engine. High-speed shafts, HSS, are
combined with azimuth thrusters the reason being that the reduction gearing is
located in the angle drives of the thrusters. In the thruster housing, there are two
ring and pinion drives that transfer the mechanical power through a 90-degree turn,
hence the nickname Z-drives. Having a ring gear that usually has more teeth than
the pinion gear means a reduction in rotational speeds, to suit the propeller and
engine combination [16].
(Note: HSS are not considered in this thesis investigation).

Figure 2.3: Typical high speed shaft line, directly coupled to a high speed diesel
engine. MTA thruster unit coupled to a CAT 3516 engine with an engine mounted
clutch. Gears located in the top housing of the thruster unit by the input shaft and
also in the leg of the thruster by the propeller shaft. Curtesy of Berg Propulsion
Production AB.

5
2. Theory

2.1.3 Engine
In the case of Berg Propulsion, most commonly propulsion power comes from medium
or high-speed marine diesel engines. Reciprocating internal combustion engines will
always cause some torsional vibrations, especially if the engine suffers from misfir-
ing. Another way of delivering the propulsion power needed is to have diesel engines
connected to generators. The generating set, engine and generator, then delivers
just electric power. Electric motors are used to turn the propeller, giving in many
cases more flexibility in operation and fewer vibrations.
In the book Ship Knowledge, the range of marine engine speeds is categorised in the
following manner. [16]:

• High-speed four-stroke engines, RPMs above 960.

• Medium-speed four-stroke engines, RPMs ranging from 240-960.
• Low-speed two-stroke engines, RPMs below 240.

The large slow 2 stroke engines are efficient but very tall and heavy, usually accom-
panied with fixed pitch propeller, FPP, and not flexible in manoeuvring. The shaft
line shown in figure 2.1 is a typical setup using a directly coupled low-speed engine
with an FPP propeller. Having an FPP propeller and direct connection to the en-
gine, doing a reverse manoeuvre, requires that a 2 stroke engine will be stopped and
then started again, turning in the opposite direction to reverse the rotation of the
propeller. This is typical for dry bulk, tanker and container vessels. The two-stroke
engines are typically in-line engines with 5 to 9 cylinders in a row, the slow running,
long stroke and huge torque from each cylinder makes torsional vibrations critical
in the shaft. The firing pulses from the engine along with fluctuating moment from
the propeller can match natural twisting and bending modes of the shaft. The most
critical phenomena being the engine’s main excitation order coinciding with the first
torsional natural frequency of the shafting system [10]. This phenomenon is due to
the fact that the crankshaft of the engine and the propeller shaft are rotating at
the same speed and the torque input is pulsating very much. The power of each
cylinder can easily be 3500 kW at 80 - 200 rpm. This results in much higher torque
values per cylinder at the same time the strokes per minutes are fewer [24].
Compared for example V12 and V16 engines as common examples of medium and
high-speed engines and more importantly the power output of each cylinder of about
1000 kW at 500 rpm giving much lower torque values per power stroke. This can
be demonstrated by a simple example, power is in essence the ability to do certain
amount of work within a time frame. It does not really matter how the work is
proceeded (ideal simplification), it can be demonstrated by cycling up a steep hill
within 5 minutes. The task requires the same amount of work whether one has
the cycle in high gear or low gear. How ever it is much harder to step the pedals
and turn the crank in high gear, this results in slow rotation but high torque on
the crank, the other way is to be in lower gear, much easier to step the pedals but
it needs many more revolutions. Hence, low torque but high rpm’s on the crank,
delivering the same amount of work. It is torque that snaps a shaft not power, then
fatigue complicates things further [16][17][28].

6
 2. Theory

2.1.4 Sterntube & Propeller shaft

From the Wärtsilä encyclopedia of marine technology we get the description as
following:

Stern tube: The watertight tube enclosing and supporting the propeller
shaft. It consists of a cast-iron or casted steel cylinder fitted with a
bearing surface within which the propeller shaft, enclosed in a sleeve,
rotates. The stern tube is installed from aft and bolted to the stern
frame boss. It can be press-fitted or installed with epoxy resin [32].

For a typical single propeller merchant or fishing vessel, the stern tube is a tubular
section running through the aft structure of a ship, below the waterline. In essence,
it is an open-ended tube between typically the machinery room and the open wa-
ter. The stern tube makes a way for the propeller shaft to run through the hull,
mechanically connecting the propeller to the propulsion engine or motor, as well
as supporting the propeller. For this purpose, the stern tube is fitted with bearing
surfaces and seals on the aft and forward end [16].
The bearings are usually of white metal and lubricated with oil. To prevent the sur-
rounding water to enter the stern tube and most importantly not flood the machine
room there are circular lip seals on the aft end of the stern tube that seal against
the propeller shaft.

Figure 2.4: 3D rendered image of a propeller shaft and a stern tube. The stern
tube is somewhat transparent. Stern tube bearings highlighted blue. Courtesy of Berg
Propulsion AB.

There are different types of setups, with single, two or more stern tube bearings, oil
lubricated or seawater lubricated bearings. Here we will describe the setup of the 3
examples studied.

7
2. Theory

• Typical RoPax setup with twin screws having very long stern tubes extended
outside the hull on either side supported by struts. The length demands a
mid bearing resulting in a stern tube having aft, mid and forward bearing. Oil
lubricated white metal bearings with seals at aft and forward end.

• Typical single screw tanker vessel having stern tube straight through the mid-
dle of the stern, aft bearing and forward bearing, aft and forward seals.

• Twin screw dredger vessel with ducted propellers having stern tubes running
through fins or narrow skegs on either side. Forward and aft bearings with
seals on forward and aft end.

The propeller shaft is the part of shafting the propeller is attached to. The shaft
extends outside of the hull such that it acts as a counter lever, supporting the weight
of the propeller, therefore there acts a large moment on the propeller shaft which
is supported by the aft and forward stern tube bearings. The shaft forward of the
stern tube is not affected much by this moment.

Propeller shaft or tail shaft: The aftermost section of the propulsion

shafting in the stern tube in single screw ships and in the struts of
multiple screw ships to which the propeller is fitted [33].

2.1.5 Intermediate shaft

The intermediate shaft is an extension of the propeller shaft for applications where
the shafting between propeller and gearbox or engine is long. Typical for ferries,
Ro-Pax vessels, offshore service vessels for example. Those types of vessels have in
common twin-screw arrangements where the machinery space is rather far forward
from the propellers, thus requiring longer shafting than for example single screw
tanker where the engine would be located just forward of the stern structure. The
intermediate shafts are often forged as one piece having flanges on both ends. The
intermediate shafts transmit the torque from the engine but they do not support
any weight but their own. Therefore much smaller moment acts on them than the
propeller shaft. See pair of a typical intermediate shaft in figure 2.5 [11].

Intermediate shafting The lengths of shafting between the propeller shaft

and the engine or the gearbox [34].

8
 2. Theory

Figure 2.5: 3D rendered image of two intermediate shafts (gray-green). Courtesy

of Berg Propulsion AB.

2.1.6 Propeller
The propeller has the purpose of converting the mechanical torque into thrust to
propel the ship. In the context of this thesis, the important factors are the weight
of the propeller and forces acting on the propeller shaft when in operation. It is
typical for a controllable pitch propeller to have 4 or 5 blades. “The number of blades
affects the vibration excitation or "beats" caused by the propeller as it operates in
a non-uniform wake field” [28, page 228].

2.1.7 Bearings
The bearings are carefully placed along the shaft according to shaft alignment. Both
as standard beam support, but also dynamic support for lateral vibrations. Thus
the number and location of bearings have to be calculated to fulfil the standards of
the classification [11]. Typically two or more bearings are placed in the stern tube
to support the bending moment from the propeller. After that, it depends on the
length and circumstances of the intermediate if bearings are needed there. Berg
does not produce their bearings, but they buy them from a third partner.
There is also a thrust bearing. The thrust bearing purpose is to fix the axial move-
ment of the thrust from the propeller pushing the ship and axial vibrations [11]. This
bearing is either placed as a separate bearing at the shaft or inside the gearbox. See
example of stern tube bearings in figure 2.4.

2.2 Static-loads
By looking into the usual stresses that a propeller shaft is exposed to, it can be listed
out as the torsional shear stress from the torque applied to the shaft, Normal stress
from the axial load, normal- and shear stress from the transverse bending loads [5].
When multiple loads are applied to the body, see image 3.1.

9
2. Theory

Figure 2.6: A common combination of load cases on a circular body, With on-plane
distortion theory [5]. Figure edited with image from [2],

Because the shearing stress due to transverse load is maximum at the centre of the
shaft, decreasing moving to the outer body of the shaft (the first moment of the
portion of the area Q is equal to zero) and the opposite is applied for the bending
moment from the transverse load and the torsion, where the maximum load is at the
outer part of the shaft. Because of that (also the vertical shear load relative to other
loads is minor) the shearing stress due to transverse load is neglected. According to
that statement area marked by “B” in image 3.1, is considered the critical design
criteria. With that definition and based on distortion energy theory and Von Mises
criterion the maximum allowed stress for the (ductile) shaft can be estimated in the
statical environment. The Von Mises criterion is defined like so, see equation 2.1[5].

1 q
σ 0 = √ (σx − σy )2 + (σy − σz )2 + (σz − σx )2 + 6(τxy
2 + τ2 + τ2 )
yz zx (2.1)
2
But in a simpler approach as it has been defined here above, the equation for on-
plane stress can be calculated with formula 2.2[5].
q
σ0 = σx2 − σx σy + σy2 + 3τxy
2 (2.2)
and with final simplification, as it can be assured that the critical design criterion
when the shaft is under unidirectional transverse load (often gravity), axial load,
and torsional load the maximum allowed stress can be estimated with formula 2.3.
(for more information see section 5-5 in Shigley’s[5]).

10
 2. Theory

q
σ0 = σx2 + 3τzx
2 (2.3)
Though this design model is simple, the fact in reality, it is not. There the shaft
is under more load phenomena, oscillations in the dynamic spectrum, other un-
predictable loads (new load vectors) like irregularities from the propeller in water,
manoeuvring, and ice. The Classification societies present methods and simple for-
mulas that can be trusted for the shaft design [21] [10] [11].

2.2.1 Breakdown of the IACS M68.4 formula

The IACS M68.4 formula is the method that Berg Propulsion uses to define their
shaft diameters, which is further discussed in chapter 2.4.1. By looking into the
formula and trying to solve and understand from what it is defined. From the start,
it can be assumed that the formula is based on empirical data. The two factors ’K’
and ’F’ are both multipliers to influence the data result, but also to adjust the unit
dimensions. Within the cubed square root it can be seen how the formula is divided
into these three sections, where we have (in this order) the load, the inertia, and the
material [21]. (as it has been said, the IACS formula is further discussed in chapter
2.4.1. For the convenience of side by side comparison, the ICAS M68.4 formula for
minimum shaft diameter can be seen here in equation 2.4.

v
u p 1 560
u
dmin = F ∗ ku
3 ∗ d4i
∗ (2.4)
t n 0 1− σB + 160
d40

To conclude where it might originate from few different solutions were set up, with
one standing out. By setting up a statical torsional load case for a hollow cylindrical
section and solve it for the diameter [5], the equation looks fairly similar to the IACS
M68.4. See equation solution here below (equation 2.5).

T ·D 60P · D · 32
τ= →τ = →
2·J 2πn · 2 · π(D4 − d4 )
skipping few steps (full step by step in appendix B section B.1) (2.5)
v
P 1 480 P 1 480
u
→ D3 = · 4 · →D= · 4 ·
u
d
3
t = d
n 1 − D4 τ π 2 n 1 − D4 τ π 2

As it can be seen, in the final solution when the diameter has been isolated from
the statical torsional load of the shaft, the equation resembles the IACS M68.4. The
only difference is the material element of each equation and, as previously mentioned,
the multiplier factors (which influence the result based on the empirical study of the
formula)
Typically in a statical approach for ductile materials, the yield strength is regarded
as the point of failure [5], however, it can be seen in the IACS M68.4 formula the
material factor is in respect of the ultimate strength. It is rather reasonable that the
critical design criteria of a propulsion shaft are based on fatigue and the endurance

11
2. Theory

limit for the dynamic load of the system [5] [17]. Therefore the material part of the
IACS M68.4 formula is limited to the fatigue strength of the design. The fatigue
investigation of the formula is done in the dynamic chapter of the report, see section
3.1.2.
In addition, since the IACS M68.4 formula is based on an empirical study from
IACS, so deviation is expected.

2.2.2 Stress concentration

A critical design factor for a mechanical component such as the shaft is the stress
concentration factors. The stress concentration factor is a result influence of dif-
ferent components on the mechanical structure, or irregularity and defect. In the
shaft case, components like change in diameter, shrink-fit coupling, flanges, notches,
keyways, and more. It is critical to consider these factors when designing the shaft
[10] [11].
The simple definition is that when the geometry is subjected to stress, and typi-
cally the stress is considered uniform throughout the cross-section of the geometry
under the load. When the loading is uniform and the geometry as well, the stress
distribution can be assumed to be homogeneous as well. But when the load or the
geometry is not uniform the stress distribution varies along the part. The peak
maximum stress occurs at these specific locations, as here previously mentioned [5].
The stress concentration factor (Kt and Kts for shear load) is the ratio between the
peak maximum stress and the average mean stress (the regular stress, for instance
for normal stress σ = F/A ), see equation 2.6 [5].
σmax
= Kt =⇒ σmax = Kt ∗ σ (2.6)
σavrage
Though the stress consideration factor is typically focused on a cross-section that is
smaller in the area (material loss due to notches, keyways, shrinkage of diameter)
the factor is not a function of the size, but rather the shape or the specific part.
Flange fillet for example, the factor mainly based on the fillet radius, meaning if the
radius is small the stress concentration is high [5].
To calculate the factor, one can simply use charts of Theoretical Stress-Concentration
factors, which can be found in mechanical engineering books, or simply on the in-
ternet.
Have in mind that these Theoretical Stress-Concentration factors are only applicable
for static loading. In the case of a ship’s propulsion shaft, failure due to fatigue
is the critical design criterion [17] [5]. Though these factors are not valid in the
dynamic environment, a specific factor is used for the fatigue stress concentration
which is derived from the Stress-Concentration factor Kt . Called the Fatigue Stress-
Concentration Factor Kf , which is a reduced factor of Kt with a scale relation called
notch sensitivity q. Similar to the other maximum stress, see equation 2.7 [17] [5].
σmax
= Kf =⇒ σmax = Kf ∗ σ (2.7)
σavrage
The reason for the maximum dynamic stress affecting the fatigue life is not the
same maximum stress at the notch (Kt for irregularities in the material), but it is

12
 2. Theory

the mean stress over the volume around the maximum stress area. More on this will
be addressed in the fatigue section of the thesis, see section 2.3.4.1 [5].

2.3 Dynamic-Loads
The dynamic loads on a propulsion shaft in operation are mainly divided into three
cases. The lateral vibrations, torsional vibration, and axial vibration. These dy-
namic loads (mainly lateral- and torsional vibrations [10] [11]) are critical to know
when designing a shaft and its fatigue limit.

2.3.1 Lateral vibrations - Whirling

Rotor dynamics describes the dynamic lateral vibration behaviour of rotating ma-
chinery. Especially flexural vibrations that can occur at the so called critical speed.
For establishing the theory describing rotor dynamics, the most basic setup of rotat-
ing machinery is considered, see figure 2.7. There is a shaft of some length supported
by two bearings, the shaft holds a disk that has a mass m [22].

Figure 2.7: Geometry of rotating disk on a shaft, supported by bearings. The

angular velocity of the shaft is ω, the center of mass G is outside of the neutral axis.
The thin dotted lines denote the deflection of the shaft, during whirling vibration
[22, p. 405]. Drawing inspired by graphics in Inman (2001).

This disk in the simple model can represent a stator in an electric motor or fan
blades in a gas turbine. Because of manufacturing precision less than perfect, and
possible residual mass like paint or some kind of dirt, the disk’s centre of mass is
not perfectly centred at the longitudinal neutral axis of the shaft [22].

13
2. Theory

Figure 2.8: Notation for whirling calculations. [22, p. 405] Drawing inspired by
graphics in Inman (2001).

The dynamic force balance of the rotating shaft and disk system is described in
vector form by equation 2.8,

mr̈ = −kxî − ky ĵ − cẋî − cẏ ĵ (2.8)

where, the inertial force is at the left side and at the right side the stiffness and
damping components. Here, m is the mass of the disk, r̈ is the second derivative
of the position vector r giving the acceleration of the mass center. The stiffness
coefficient k is related to the displacements x and y, similarly the damping coefficient
c is related to the velocities ẋ and ẏ [22].
From the geometry we get equation 2.9 for the transverse position of the center of
mass from the neutral axis, r.

r = (x + a cos ωt)î + (y + a sin ωt)ĵ (2.9)

Here, the angular velocity of the shaft is given by ω. The angular velocity has the
units of radians per second, rad
s
, as can be seen on figure 2.8 the rotation of the shaft
is defined as ω · t resulting in the angular displacement in radians. Worth noting
is that usual practice is to give engine speeds in rotations per minutes, rpm, then
normally noted with n0 [31]. Figure 2.8 describes also the geometry of the whirling
motion and relates that to equation 2.9 and 2.10 [22].

r̈ = (ẍ − aω 2 cos ωt)î + (ÿ − aω 2 sin ωt)ĵ (2.10)

(mẍ − maω 2 cos ωt + cẋ + kx)î + (mÿ − maω 2 sin ωt + cẏ + ky)ĵ = 0 (2.11)

14
 2. Theory

mẍ + cẋ + kx = maω 2 cos ωt (2.12)

mÿ + cẏ + ky = maω 2 sin ωt (2.13)

Steady state solution of equation 2.12

!
ar2 2rζ
x(t) = q cos ωt − arctan (2.14)
(1 − r2 )2 + (2rζ)2 1 − r2

similarly for equation 2.13

!
ar2 2rζ
y(t) = q sin ωt − arctan (2.15)
(1 − r2 )2 + (2rζ)2 1 − r2

where, r = √ ω and ζ = c/(2mωn ). The phase angle, φ, describing the angular

k/m
shift between the rotation of the shaft about its neutral axis and the rotation of the
2rζ
disk about the shaft, is given by φ = arctan 1−r 2 [22].

y sin (ωt − φ)
tan θ = = = tan (ωt − φ) (2.16)
x cos (ωt − φ)
This gives the relationship θ = ωt − φ which differentiated in respect to t gives
the velocity of whirling θ̇ = ω. It is called syncronous whirling when the speed of
whirling is the same as the angular velocity. Whirling is the rotation of the disk
around the nautral axis of the shaft. The amount of deflection is described by r
the radius from the nautral axis to the circular path the shaft follows when whirling
about the nautral axis [22].

r = xî + y ĵ (2.17)

q q
|r(t)| = x2 + y 2 = X sin2 (ωt − φ) + cos2 ωt − φ = X (2.18)

ar2
X=q (2.19)
(1 − r2 )2 + (2rζ)2

When r ∼ = 1, the system hits a resonance and the diflections of the shaft-disk system
or rotor, can start to experience very large amplitudes. At r = 1 the angular
q velocity
of the disk equals the natural bending frequency of the shaft, ωr = k/m, this is
called the critical speed of the rotor [10][22]. Since the marine shafting systems have
many vibration modes, each having a natural frequency, the propulsion shaft will
have whirling speeds for each mode of vibration. Those are the critical operating
speeds later estimated by shaft alignment procedures.

15
2. Theory

Figure 2.9: Normalized deflection amplitude versus frequency ratio for different
amount of damping. The resonance behaviour is clearly demonstrated close to fre-
quency ratio of one. The steady state amplitude becomes equal to the offset of the
center of mass [22, p.407].

For further reading about marine shafting lateral and axial vibration theory, the
interested reader can access the open access article, Vibration Characteristics and
Power Flow Analysis of a Ship Propulsion Shafting System with General Support
and Thrust Loading by Deshui Xu, Jingtao Du, and Chuan Tian at the College
of Power and Energy Engineering, Harbin Engineering University, Harbin 150001,
China. Published 13 June 2020, Hindawi - Shock and Vibration.

2.3.2 Torsional vibrations

Torsional vibration calculation - TVC is an important step in the design of a propul-
sion system for ships. It is essential to carry out these calculations early in the design
phase when the shafting diameters have been selected [31]. Typically the propulsion
shafting is dimensioned according to the classification requirements very early on
in the design phase, with minimal amount of inputs using the simple IACS M68.4
rules, then when preliminary shaft arrangement has been drawn the shaft design
is checked for torsional vibrations problems in a TVC analysis. Usually this would
be done at the same time as first shaft alignment and whirling analysis are made.
There are different stakeholders in a ship building project that could do a TVC
analysis, the engine or gearbox manufacturer, supplier of flexible couplings as well
as the propeller and shafting manufacturer.
The reason for doing those calculations early in the design process of a ship is to
locate possible problems before the actual physical ship building is underway. It is
much easier and less expensive to do modifications, move or add support bearings,
change shaft dimensions while still in the design phase rather than needing to scrap
or change something already built and manufactured, like stern tube and bossing or
large propeller shafts [31]. According to rotor dynamics experts at Berg Propulsion,

16
 2. Theory

the experience is that most often only small adjustments are needed and those can
be done by choosing the correct flexible coupling, tuned according to TVC analysis,
that dampens out unwanted harmonics [19].
Torsional vibration is a vibration phenomena related to a unsteady torque applied
to a shaft. The vibration occurs in an angular direction around the center axis of the
shaft in a plane perpendicular to the cross section of the shaft. [22, page 452] The
rotation of the shaft fluctuates, hence, it is time dependent. It is also dependent on
position along the shaft because the shaft is twisted by the torque it is transmitting.
Hence, the rotation is given by θ(x, t), where x is the position along the length of
the shaft and t denotes time.
To establish the general equation of motion for a shaft transmitting torque, a short
section of a shaft is considered which has a moment force and reaction at each
end [22].

Figure 2.10: The figure describes the twist of a shaft transmitting torque and
relation between, torque, twist angle and position [22, p. 452]. Drawing inspired by
graphics in Inman (2001).

∂θ(x, t)
τ = GJ (2.20)
∂x

∂τ ∂ 2θ
τ+ dx − τ = J0 2 dx (2.21)
∂x ∂t
!
∂ ∂θ ∂ 2θ
GJ = ρJ (2.22)
∂x ∂x ∂t2
When the stiffness of the shaft GJ can be regarded as constant, the result is equation
2.23, describing twisting vibration of a shaft.
!
∂ 2 θ(x, t) G ∂ 2 θ(x, t)
= (2.23)
∂t2 ρ ∂x2

17
2. Theory

Table 2.1: Notation for TVC formulation.

τ : Torque
θ(x, t) : Rotation
G : Shear modulus
J : Polar moment of area
GJ : Torsional stiffness

2.3.2.1 Damping in TVC

Damping is not as clearly defined as the inertia and stiffness properties of shafts
and engine components. The inertia of a shaft or a wheel can be calculated directly
from its geometry and density, similarly, the stiffness is a property of the geometry
and material properties.
The calculation results depend upon inertial moments of actual masses,
stiffness of shafting components, as well as the excitation forces and
moments exerted by the propulsion engine(s) and the propeller. Inertial
moments and stiffness can be determined with no ambiguities. However,
this is not the case with either the damping or the engine excitation.
Actually, during validation onboard the calculation supposed damping
is the main influential factor to be verified. [31, p. S-156]
Propeller damping factors used to define the propeller absolute damping bAp , are the
Frahm number DF , see equation 2.25 and Archer number DA , see equation 2.26.
The relationship between those two factors can be demonstrated by [31, p. S-168]:

30
DA = DF · (2.24)
π
The application of the Frahm number:
2 3
30 Tp 30 P0 30 P0
 
bAp = DF · · = DF · · 3 · np = D F · · · ωp (2.25)
π np π np0 π n3p0
The application of the Archer number:
2
Tp 30 P0 30 P0
  
bAp = DA · = DA · · 3 · np = D A · · · ωp (2.26)
np π np0 π n3p0
The practical values of the damping factors are for Frahm, between 2.9 to 3.7 and
for Archer, 25 to 35 [31].
In Nauticus Machinery TVC where the propeller properties are defined, a typical
value for Frahm is suggested as 2.8. For the Archer number, the lowest value possible
is 20, the range of typical values is stated 25-35 [13]. A good rule-of-thumb value
according to Berg rotor dynamic experts is Archer number equal to 27.
Another model for calculating the propeller damping is the Schwaneke damping.
With this model the damping is calculated through elaborate formulation using a
collection of factors, all describing the geometrical properties of the propeller [13].
The propeller damping C is given by [13, p. 41]:

18
 2. Theory

2
DB H AE
  
5
C=f· ·D · · ·n (2.27)
D D Ao

q q q q 
D D D D

DB

DB
−1 DB
−1 DB
−1 arctan DB
− 1
f· = 4.271· + − + (2.28)

D 2 3
D 8· 8
  
D D
DB 12 · DB
3· DB

Table 2.2 summarises notation and definitions used in propulsion system calculations
[31, pp. S-168 - S-169].

Table 2.2: Notation and definitions for propulsion system calculations.

P0 : Engine nominal rating power, MCR [kW ]

n0 : Engine nominal speed at MCR [rpm]
i : Gearbox transmission ratio
np0 = ni0 : Propeller nominal speed [rpm]
ωp = π·n
30
p
: Propeller angular velocity [rad/s]
np : Propeller speed [rpm]
T0 = 30
π
· Pn00 : Engine nominal torque [kN m]
Tp0 = i · T0 : Propeller nominal torque [kN m]
np 3
 
P = P0 · : Propeller power curve [kW ]
 np0 2
np
Tp = T0 · np0
: Propeller torque curve [kN m]

2.3.2.2 Application factor KA

In calculations of the dynamic torsional load (simply torsional vibration) on a shaft

the application factor KA is used. The application factor is the factor that describes
the range between the nominal torsional load and the amplitude of the torsional
vibration, that is the increase in rated torque affected by superimposed dynamic
or impact loads. For each different load case, a separate application factor is used.
These application factors can either be acquired with Torsional Vibration Analysis
(also called TVC) or in cases of lack of data they can be estimated from provided
tables [10] [11].
For the continuous normal operations of the ship the factor KAnorm is utilized (it is
a KA factor, but labeled KAnorm to identify what amplitude is for normal operating
conditions). The KAnorm is a key design criterion in the dynamic response since it is
the factors that have an important influence on the high cycle fatigue assessment of
the shafts. The KAnorm can be defined like previously said with TVC or estimated
from a table like DNV provides, where the given KA is based on different system
types [10] [11], see table 2.3.

19
2. Theory

Table 2.3: DNV definition for estimated application factor for different system
setup[11]

System type KAnorm KA

Turbines and electric drives 1.1
Diesel engine drive system with fluid clutch between engine and gears 1.1
Diesel engine systems with highly flexible coupling between engine and gears 1.3
Diesel engine system with no flexible coupling between engine and gears 1.5
Generator drives 1.5

A more precise approach is to define the KA based on simulated TVC data. The
KA is calculated by certain formula, see equation 2.29 (only applicable for geared
plants). Where τv is the torsional amplitude and τ0 the nomial torsional load [10]
[11].

Tv τv
KA = 1 + =1+ (2.29)
T0 τ0

KAP defines the amplitude of the highest torsional loads, which define the lowest
accumulative fatigue criteria (lowest of the low-cycle fatigue). That is frequent
loads like extreme maneuvering, clutching in shock load, and if the vessel is ice-class
grated the torsional load from ice shock [11]. The KAP is calculated with the given
formula, see equation 2.30 (only applicable for geared plants)[10] [11].

Tpeak τpeak
KAP = = (2.30)
T0 τ0

δKA is the factor for the torque range for reversible torsion, that is the torsional
range is applied in both directions (minus and positive loads). This is applicable
for reversing plants. The load case is regarded as low cycle stress and can simply
be defined as the KA amplitude times two. But with more precision and if one has
the data the ∆KA can be calculated as so, see equation 2.31 (only applicable for
geared plants)[10] [11].

τ0 KAP + |τmaxreversed |
∆KA = or → ∆KA = 2KAP or KAice + KAP (2.31)
τ0

The application factor is important in this study and is used in the calculations of
the minimum allowed shaft diameter, in this project [10] [11].

An example how the application factor is displayd can be seen on figure 2.11 (figure
is presented in DNVGL-CG-0038 [10] for their definition of the application factor)

20
 2. Theory

Figure 2.11: Torsional vibratory stress for different load scenario and its relation
with mean stress tau0 and KA, in time domain. Based on direct coupled plant.
Image gotten from DNVGL-CG-0038 [10]

2.3.2.3 Estimating the application factor

If the application factor is needed for a shaft or other propulsion calculations, and
a TVC analysis has not been done for that specific propulsion system, an estimated
application factor can be acquired.
The application factor for a normal operating condition (KAnorm or simply KA
(sometimes KA ) has been mentioned here in the previous section and can be ob-
tained from table 2.3.
A low cycle application factor KAP (Application factor, temporary occasional peak
torques, can simply be estimated as 1.4 [12]

The ice-class application factor is more difficult to estimate than the other two
factors here previously. The torque is based on peak torque when the propeller hits
the ice. An empirical method to estimate the factor is described in DNVGL-RU-
SHIP Pt.6 Ch.6. Part 15.5.3. An extensive step-by-step solution will be found in
Appendix B sectionB.3. The calculation output will provide the peak torque, which
can then be used with equation 2.30. The estimation is based on few variables, that
is if the propeller is in a duct, the different ice-class, the mass moment of inertia
ratio between the propulsion side and engine side, the pitch ratio at radius 0.7, and
others. Note the ice-class should not be less than the KAP , and if that happens,
the calculations will always depend on the critical amplitude, which would then be
KAP [12].

2.3.3 Axial vibration

Acts in the axial direction, along the shaft, resulting in alternating compression and
tension.
The axial vibration is not considered in this thesis study, since it is a minor dynamic
load and does not affect the fatigue-life or vibrations of a marine shaft. As it is stated

21
2. Theory

in DNV CG 0038 [10, p. 6], "stresses are considered negligible for marine shafting
systems as they are dominated by torsional- and bending stresses", and in terms
of fatigue strength "the combined vibratory torsional and rotating bending stresses
(axial stresses disregarded) relative to the respective component fatigue strengths."
[10].

2.3.4 Fatiuge
When a part is subjected to its critical yield strength the material enters the perma-
nent deformation, and in many mechanical design cases that is considered a failure
[5]. The ultimate strength, or the tensile strength of the material, is the maximum
stress the material can be subjected to before a complete failure of the part. The
observation of this kind of failure is in the respect of stress/strain. In the Berg
Propulsion case for a shaft line, the most common material chose for the shafts
is EN 1.1170 (28Mn6) steel, which has requirement standards: tensile strength of
σB = 600N/mm2 , and yield strength: σy = 350N/mm2 . Fatigue failure due to
stress/cycle is another design criterion important in dynamic load cases such as the
rotating shaft of a vessel. That is the breakdown of the mechanical component fails
way below the ultimate strength of the material due to accumulative repeated cycle
load on the part. Fatigue failure predicts the number of cycles a component can
withstand before a failure, by comparing the operation of the part to experimental
empirical data or do a real-life experiment on the design.
A brief explanation of fatigue strength is that the estimation for the life for the part
or the design criteria for the part can be explained and plotted up in an SN-graph
(or Wöhler curve) [17] [5]. The y-axis is the fatigue strength or the stress range and
the x-axis the number of cycles, see an example of the graph in figure 2.12.

Figure 2.12: Example SN- diagram for a steel material with LCF range up to 103
and enurance limit at 106 . The x-axis is number of cycles in log scale and y-axis is
the fatigue stress range [17] [5].

22
 2. Theory

Where each point on the graph describes the number of cycles the material can with-
stand with the given stress. Depending on the material, the shape, type of load,
surface finish, and more the curve changes. It can be expensive and time-consuming
to perform real-life tests for particular designs, so a conservative SN curve can be
plotted for different cases. Note that typical SN diagrams are only applicable for
completely reverse loading [17] [5].
The SN diagram is divided into two cases, the low cycle fatigue and the high cycle
fatigue. The low cycle is typically rated for cycles up to 103 or 104 number of cycles
(note that DNV rates LCF for up to 104 , this is all based on experimental data).
Low cycle fatigue are high stresses that can only accumulate few cycles before failing,
they are accounted for stresses that can both be plastic deformation and elastic-.
The high cycle fatigue accounts for more than the given number of cycles for LCF
and up [10] [11].
An explanation along with figure 2.12, the diagram on the figure represents a con-
servative SN diagram with steel material and LCF at 103 . The diagram is divided
into three slopes, begins at the ultimate tensile strength, and with a fixed slope (in
log scale) it reaches the HCF zone. The slope is defined by the constant f at the
103 cycle, the Fatigue Strength Factor, which is a conservative factor based on the
material ultimate strength. From this point is another fixed slope (in log scale) to
the endurance limit, Se at 106 . The endurance stress limit is defined like so, see
equation 2.32 [5].

Se = ka kb kc kd ke kf ∗ Se0 (2.32)

Where the Se0 strength is conservatively 0.5Sut (if the Sut < 1400MPa) , but can
also be acquired from an experimental data graph, where the ratio between Se0
and Sut ranges from 0.4 to 0.6. The k factors account for different set-ups, surface
condition, size factor, load factor, temperature-, reliability-, miscellaneous effect-
and more. The slope between f ∗ Sut and Se can be described as Sf = aN b , where
the a and b factors can be solved for, see equation 2.33

f ∗ Sut = a(103 )b & Se = a(106 )b −→ ...

(2.33)
... a = (f ∗ Sut )2 /Se & b = −[log(f ∗ Sut /Se )]/3

From the endurance limit (below it) the part should be able to withstand an infi-
nite amount of cycles without failure [5].(remember this statement, because that will
be the main design criteria when trying to unravel the IACS M68 formula (mimic it))

This explanation of the fatigue life of a designed component is simplified and in

real life there are other input variables to account for, and based on the mechanical
component different approaches to the SN-diagram are supported by empirical data.
For instance, when looking at the behaviour of the stress cycle, see image 2.13.

23
2. Theory

Figure 2.13: Simplified stress amplitude for two different loadcases

For the SN diagram (figure 2.12) to be valid the load must be complete reverse
stress amplitude, that is when the mean stress is equal 0 (σmax = |σmin |). In the
case of a vessels shaft in operation, the torque load is constant with addition to
vibrations, Meaning that the mean torque (called Nominal torsional stress at max-
imum continuous power τ0 ) is not zero, and the shaft operation condition is more
like the orange line in the graph(figure 2.13), having the mean stress under tension
will affect the fatigue life of the component. To address this, either the SN graph
for the specific mean value must be acquired, or use methods to check the criteria
of failure, approaches like the Soderberg, Goodman, Gerber, Dang Van and more
[18] [17] [5].
In addition to this complication, in real-life cases like the ship shaft, will not be
submitted to simple dynamic load cases as has been visualized here. Instead, many
different factors influence the operating condition of the environment. In statistical
calculations, the Rainflow Counting Method can be used to "count" the peaks and
divide the stress cycles into separate simpler constant stress amplitude. Each sep-
arate cycle can be accumulated with The Palmers-Miner Cycle Ratio Summation
Rule, or simply the Miners rule, to examine if the component will reach fatigue
failure [9] [18] [5], see the Miner rule in equation 2.34.
X ni
D= (2.34)
Ni

If the component is under different loads, bending-, axial and/or torsional stress,
combining them with von-Mieses is an acceptable way to acquire the stress ampli-
tude (σa ) and mean stress (σm ) [5] (see section 6-14 in Shigley’s)

DNV Classification society uses the torsional load amplitude (according to the theory
here previously) to estimate the fatigue life of the shaft. To estimate the load cycle

24
 2. Theory

amplitude a factor is used that represents the ratio between the mean torsional load
and the maximum load for each loading scenario (load scenario meaning, normal
operating conditions, an encounter with ice-, barred speed range, and so on). The
factor is called application factor KA , has been further explained in section 2.3.2.2.

2.3.4.1 Fatigue Stress Concentration Factor

Critical design criteria for a mechanical component like the shaft is the Fatigue Stress
Concentration Factor (Kf and Kf s for torsion). The Fatigue Stress Concentration
Factor is a reduced value of the Stress Concentration factor (previously talked about
in section 2.2.2), defined by the notch sensitivity q, see equation 2.35.

Kf − 1 Kf s − 1
q= or q= (2.35)
Kt Kts
The notch sensitivity can either be acquired from notch sensitivity charts or by using
Neuber equation [5] (see section 6-10 in Shigley’s).

2.4 Classification rules & guidelines

Classification society provides rules and regulations that apply under their flag state,
for the ships registered under that specific country. In this chapter the relevant rules
and regulations that are applied in this study is presented. All the classification rules
from each society are freely available on their websites.

2.4.1 IACS
Looking into a few of the major classification society rules, it can be seen how each
standard defines the minimum allowed diameter of the propulsion shafting. By look-
ing into the foundation of the rules, the International Association of Classification
Societies (or IACS), which is a non-governmental organization that consists of the
twelve major classification societies. 90% of shipping tonnage in the world market
falls under the IACS classification, so it is safe to say that the standard is respected
in the world of shipping [3]. The minimum allowed propulsion shaft diameter is
defined in the IACS classification rules (rather a conservative one), which provides
the 12 member societies a foundation for their shaft rules and guidelines. By looking
into some of the member societies it can be seen that every member defines their
minimum allowed shaft diameter with formulas that resemble the IACS formula.
Either by just having it completely the same, maybe small changes, or simply just
the same but they break it apart into separate equations and also change the letters
of the variables. This conservative formula from IACS is defined in their M68 docu-
ment [21]. The classifications from IACS that are relevant for this thesis study can
be obtained from that specific document. Formulas, definitions, and other rules such
as material limitations for the calculations, Permissible torsional vibration stresses
(used to find barred speed range), influence from different shaft design features, and
more [21].

25
2. Theory

The minimum allowed diameter formula that has been referred to here and will be
referred to as the IACS M68.4 formula is defined like so, see equation 2.36 [21].
v
u p 1 560
u
dmin = F ∗ ku
3 ∗ d4i
∗ (2.36)
t n0 1− σB + 160
d40

Not many input values are needed to perform the calculation since the method is
considered simple and conservative. Though the classification societies display this
method of designing the propulsion shaft(section 2.4 in PRS [26], 5.2 in RS [27], Pt
5, Ch 3 section 203 in KR [23] and many more) some also provide a more complex
alternative way of defining the shaft and still be approved by the class [10].
Limitations are defined so the calculation method for the M68.4 formula (applies to
all equations and requirements in the m.68 document). Where the selected material
is required to be:

- For carbon and carbon manganese steels, a minimum specified tensile strength not
exceeding 600 N/mm2 for use in M68.5 and not exceeding 760 N/mm2 in M68.4.
- For alloy steels, a minimum specified tensile strength not exceeding 800 N/mm2.
- For propeller shafts in general a minimum specified tensile strength not exceeding
600 N/mm2 (for carbon, carbon manganese and alloy steels).

Other requirements must also oblige and are listed in the document (one can look
it up in m.68 document [21].) and since they do not fall under the study case, it is
unnecessary to define them here [21].

The IACS UR M68.5 formula for allowed torsional vibration based on defined mate-
rial thickness for continuous operation. Torsional vibration analysis of a given shaft
with a scenario when one cylinder is misfiring is performed to compare with the
Permissible torsional vibration stresses M68.5. This calculation can be used both
for coupled and geared plants. The output result is the allowed vibrational range,
from no rotation up to the full speed of the shaft. The results can give the barred
speed range, that is the shaft rotational speed range where the torsional vibration
stress for the given speed exceeds the allowed torsional vibrational stress [21]. See
equations 2.37 and 2.38.

σB + 160
± τC = ∗ CK ∗ CD ∗ (3 − 2 ∗ λ2 ) f or : λ < 0.9 (2.37)
18
σB + 160
± τC = ∗ CK ∗ CD ∗ 1.38 f or : 0.9 < λ < 1.05 (2.38)
18
As it has been mentioned, the M68.4 formula is simple and conservative and thus
there should be room for improvement. Hence defined by some of the IACS mem-
bers are more complex and detailed methods for calculating the minimum shaft
diameter. IACS classification does mention this, therefore the alternative methods
are approved. One method strikes out and was the main method used in the thesis
study, it is considered a guideline rule from DNV. The reason for the thesis study

26
 2. Theory

to be so dependant on DNV is because of their influence and relevancy in the indus-

try, at least here in northern Europe. The method is defined in a document called
DNVGL-CG-0038, see later in section 2.4.2.2.

2.4.2 DNV
The document DNVGL-RU-SHIP Pt.4 Ch.4 goes over the rule requirements, and
the calculation methods for the propulsion shaft, hence the name of the document
"Rotating machinery – power transmission"[11].
DNV has few variable ways to calculate the critical diameter for the shaft from
the classification rules. The first design formula is the IACS M68.4 formula (previ-
ously revived in chapter 2.4.1). Were the minimum allowed diameter is defined with
the pre-defined formula 2.36 (see page 26). The minimum diameter formula (IACS
M68.4) is rather conservative, arguably too safe, and the one that Berg Propulsion
uses today to calculate its shaft diameter. Equation can be found on page.18 in
DNVGL-RU-SHIP Pt.4 Ch.4. named "2.2.7.2 IACS UR M68.4 Shaft diameter" [11].
In addition, the shaft design should be checked with torsional vibrations, if the shaft
can withstand the torsional vibration stresses, see equations 2.37 and 2.38 on page
26. It is under the assumption that the barred speed range is passed rapidly both
upwards and downwards with acceptance of 5 seconds, and can not be within 20%
of the MCR (λ < 0.8) [11].

Then DNV provides another method is described in the document, which DNV
calles a ’simplified method’ , yet it´s more complex than the IACS M68.4 formula.
DNV tries to keep it simple and accessible to use, the formula is aimed at geared
plants with low torsional vibrations, with shaft material as steel. For the formula
to be relevant few different criteria have to be met (same goes for stainless steel
shafts):

σy limited to 0.7 σB (for calculation purpose only)

Application factors KAP ≤ 1.4
Vibratory torque Tv ≤ 0.35 T0 in all driving conditions
Application factor, torque range KA ≤ 2.7
Inner diameters di ≤ 0.5 d0 except for the oil distribution shaft with
longitudinal slot where di ≤ 0.77 ≤ d0
Protection against corrosion (through oil, oil based coating,material selection
or dry atmosphere).

It is distinguished into two separate calculations, either for low cycle fatigue (LCF)
and high cycle fatigue (HCF). Where the low cycle is defined as the shaft subjected
to a specific load case accumulation of not more than 104 loads. The case consists
of loads on the shaft such as a stop to full operational speed and high repetitive
yield in compression followed tension (relevant in both directions), for example, a
reverse plant, referred as "crash stop". The High cycle fatigue has its scope on
normal operation conditions, and accumulates to more than 3 ∗ 106 and typically
up to 109 and 1010 . Typical high cycle fatigue loads can be forces from each firing

27
2. Theory

pulse of the engine and different loads from the propeller. The minimum diameter
from low cycle criteria as it is defined by DNV. See equation 2.39, and page 15 in
DNVGL-RU-SHIP Pt.4 Ch.4 [11], and for parameter for different variables.
s
T0
dmin = 28 ∗ k1 ∗ 3
(2.39)
σy
The minimum diameter from high cycle fatiguesee. See equation 2.40, and page 15
in DNVGL-RU-SHIP Pt.4 Ch.4 [11], and for parameter for different variables.
s 2 ! 16
T0 Mb

dmin = 17.5 ∗ k2 ∗ 3
1 + k3 (2.40)
0.32 ∗ σy + 70 T0
These formulas are recommended by DNV for the diameter of the shaft, the two
formulas can only be applied for steel shafts, not stainless steel shafts. Separate
equation, with the same shaft requirements as the steel shafts (the list in the same
paragraph, above). Required values and further explanation can be acquired from
DNVGL-RU-SHIP Pt.4 Ch.4 page 17 [11].

The class offers these ’simplified calculations’ for the minimum diameter of the
propulsion shaft. For a more detailed method, DNV offers a class guideline document
on how to assess the safety factor for LCF and HCF for a shaft setup, the document
is called DNVGL-CG-0038. To acquire more detail and push the design to the limit
the detailed DNV method can be used, which was done in this thesis study. In the
next section(2.4.2.2), a summary of the theory and output of the document is done.

2.4.2.1 ICE-class application

If the vessel under consideration has ice-class the propulsion system must fulfil spe-
cific ice-class standardisation. The criteria taken under consideration in this project,
the first one is the application factor KAice (previously talked about in section
2.3.2.3). The other one is that the propulsion shaft should be able to withstand
the blade failure load. If the propeller hits an ice block with too much force, the
propeller should fail before the shaft. This criterion is concerned for ice crashes but
is also applied for the failure of the propeller when it grounds [12]. The calculation
method and formula can be seen in appendix B.4 (based on DNVGL-RU-SHIP Pt.6
Ch.6 12.5.2 [12])

2.4.2.2 DNVGL-CG-0038
The document DNVGL-CG-0038 from the DNV classification society describes their
method how to calculate the approved minimum diameter of a propulsion shaft,
including the influence factor from the dimensions of the mechanical components
on the shaft. The method takes into consideration the plants set up, if it is geared,
or directly coupled, and others. The calculation method is vast and it can change
a lot based on small changes (many different paths). The fatigue case is the same
as described in the section above (2.4.2), that is low cycle fatigue and high cycle
fatigue. If the shaft properties do not fulfil the requirements criteria for IACS M68.4

28
 2. Theory

or the ’simplified method’ (mentioned in section 2.4.2) DNVGL-CG-0038 calculation

method can be used, it has a slightly larger property requirements margin [10].

Material of forged or hot rolled steels with minimum tensile strength of 400 MPa.
Material tensile strength, σB up to 950 MPa and yield strength (0.2% proof stress),
σy up to 700 MPa.
No surface hardening.
No chrome plating, metal spraying, welds etc. (which will require special
considerations).
Protection against corrosion (through oil, oil based coating, paint, material
selection or dry atmosphere).

The calculations are divided into two separate fatigue criteria, one where low cycle
fatigue is considered (LCF) and the other when the high cycle fatigue (HCF). The
LCF load case is the same as described in the section above (2.4.2), that is loads on
the shaft that accumulates less than 104 cycles. Loads like, from start to full load,
from start to peak loads (like ice shock or clutch-ins) [10].

σy
τmax ≤ (2.41)
2SKL

The main equation for the LCF, see equation 2.41. Where the τmax is the peak
nominal torsional stress. That is the τ0 , the nominal torsional stress at a selected
speed, or in this case the maximum since it is the most critical. To get the most
probable highest torsional stress, additional loads, such as clutch-in and ice shocks
have to be taken into consideration. For geared plants, the τmax (peak nominal
torsional stress) is computed using a KA factor (see equation 2.42), the appropriate
factor is selected based on the shaft design criteria (like ice class). In figure 2.11
(see again on page 21) the maximum torsional stress is visually described with the
combination of the KA factor and τ0 [10].

τmax = τ0 ∗ KA (2.42)

The process of acquiring the peak nominal torsional stress for direct-coupled plants
is a little bit different. The equation is like so, τmax = τ (n) ∗ τv (n). The torsional
vibration, τv has to be acquired with TVC analysis. The highest nominal torsional
stress is then based on a specific speed [10], see image 2.14.

29
2. Theory

Figure 2.14: Torsional stress based on torsional vibration (τv ) and the torsional
stress at specific speed (τ ).Image was taken from DNVGL-CG-0038 [10]

In general, the whole calculation method for LCF is based on equation (2.41). Where
the diameter of the shaft, the engine load is addressed in the τmax . The component
influence factor KL considers the shaft component and the surface finish [10]. The
Safety for LCF can not be lower than 1.25 for LCF calculations[11].
The HCF load case is the same as described in the section above (2.4.2). The load
conditions will accumulate not less than 3 ∗ 106 and typically up to 109 and 101 0. It
includes loads in normal operating conditions, like pulses from the engine and loads
from gears, based on formula, see equation 2.43.
!2 !2
1 τv σb
2
≥ + (2.43)
s τf σf
Similar to the calculation method for LCF, each variable in the equation addresses
different factors. The τv , the nominal torsional vibratory stress for continuous opera-
tion. That is based on τ0 and the KA for normal operating conditions. The nominal
reversed bending stress amplitude σb . Based on the diameter of the shaft and the
rotating bending moment, Mb . According to DNV the rotating bending moment can
either be calculated from the shaft alignment or simply be approximated as 40% of
the torque at maximum continuous power [10]. T0 . The other two variables in the
HCF equation (eq.2.43) is the τf high cycle fatigue strength and the σf high cycle
bending fatigue strength. the determination of those two values can be vast and
more complicated, they represent the component influence factors KHτ and KHσ .
As the name suggests they take into consideration the mechanical component on the
shaft and its dimensions in the relation to the material properties and the vibrating
stresses [10]. With the dimension of the shaft and the components defined, as well
as the material, the safety factor can be determined. Where the minimum safety
factor for HCF according to DNV is 1.6 [11].

30
 2. Theory

To see the whole calculation process for both LCF and HCF of the calculation
method, including reverse calculation method and direct coupled plant in HCF
analysis, one can look up the document DNVGL-CG-0038 or a flowchart created
by the writers, based of the document, see in the appendix A.1.

2.4.3 Classification rule requirements for shafting alignment.

The rules specifically for alignment approval from a few different societies. Note,
that the rules and instructions are taken straight from each classification document.
The exact wording is vital for the definition of the regulations not to be altered. All
the classification rules from each society are freely available on their websites.

2.4.3.1 DNV - Det Norske Veritas

“For geared plants, the alignment calculations are only applicable for the low speed
part of the shaft line, including the output gear shaft with radial bearings. Vertical
shaft alignment is always applicable, while horizontal alignment is applicable upon
request. The rule requirements applies to fully submerged propellers. [14, p. 32]”
DNVGL-RU-SHIP Pt.4 Ch.2 Rotating machinery, general, Section 4, 1.1.1 Ed.2020-
07 Revised.
Criteria for systems requiring shaft alignment calculation or only requiring shaft
alignment specification. Shaft alignment calculation report shall be submitted for
approval for propulsion plants with one out of the following criteria:

• Minimum shaft diameters (low speed side) of 400 mm or greater for single
screw and 300 mm for twin screw
• Gear transmissions with more than one pinion driving the output gear wheel,
even if there is only one single input shaft as for dual split paths
• Shaft generator or electrical motor as an integral part of the low speed shaft
in diesel engine propulsion
• Single stern tube bearing arrangement.
DNVGL-RU-SHIP Pt.4 Ch.2 Rotating machinery, general, Section 4, 1.3.2 Ed.2020-
07 Revised
For all propulsion plants other than those listed above, a shaft alignment specifi-
cation shall be submitted for information. The shaft alignment specification shall
include the following items:
• Bearing offsets from the defined reference line
• Bearing slope relative to the defined reference line if different from zero
• Installation procedure and verification data with tolerances e.g. gap and sag
and jacking loads (including jack correction factors and jack positions) and
verification conditions (cold or hot, propeller submersion, etc.).
DNVGL-RU-SHIP Pt.4 Ch.2 Rotating machinery, general, Section 4, 1.3.3 Ed.2020-
07 Revised
In addition to the requirements listed above
Aft most bearing acceptance criteria and modelling of aft most bearing are depen-
dent of risk:

31
2. Theory

— White metal lined aft stern tube bearing which is either double sloped, or has a
journal diameter 500 mm or greater, criteria is not considered in this study
— Stern tube arrangements incorporating a single stern tube bearing only, are not
considered in this study.
— Other propulsion plants where alignment calculation is required shall fulfil re-
quirements in the acceptance criteria listed below.
Guidance note:
Aft most bearing is in most cases to be understood as aft stern tube bearing, but can
also be other designs e.g. strut mounted bearings which are common in twin screw
designs without skegs. DNVGL-RU-SHIP Pt.4 Ch.2 Rotating machinery, general,
Section 4, 1.1.3 Ed.2020-07 Revised

Survey Installation of propulsion plants requiring a shaft alignment calculation shall

be verified by a surveyor DNVGL-RU-SHIP Pt.4 Ch.2 Rotating machinery, general,
Section 4, 1.1.4 Ed.2020-07 Revised

Acceptance criteria:
The shaft alignment shall fulfil the following acceptance criteria for all relevant
operating conditions.
• Acceptance criteria defined by manufacturer of the prime mover, e.g. limits
for bearing loads, bending moment and shear force at flange
• Acceptance criteria defined by the manufacturer of the reduction gear, e.g.
limits for output shaft bearing loads and load distribution between bearings
• Bearing load limits as defined by bearing manufacturer and Ch.4 Sec.1
• Zero or very low bearing loads are only acceptable if these have no adverse
influence on whirling vibration Tolerances for gap and sag less than 5/100 mm
are not accepted. Acceptance criteria for aft most tail shaft bearing:
• In hot static and hot running conditions the relative nominal slope between
shaft and aft most propeller shaft bearing should not exceed 3·10-4 rad (0.3
mm/m) and 50% of minimum diametrical bearing clearance divided by the
bearing length, whichever is less. For definition of relative nominal slope, see
Operating conditions:
The shaft alignment calculations shall include the following conditions:
— Alignment condition (during erection of shafting)
— Cold, static, afloat, fully submerged propeller
— Hot, static, afloat, fully submerged propeller
— Hot, running with hydrodynamic propeller loads.
For geared shafting systems:
— Running conditions as required to verify gear acceptance criteria
— All relevant combinations of prime mover operation
— Horizontal alignment is upon request.

2.4.3.2 LR - Lloyd’s Register

Few appropriate shaft alignment requirements from Lloyds Register

32
 2. Theory

• Relative slope between the propeller shaft and the aftermost sterntube bearing
is, in general, not to exceed 3 × 10-4 rad.

• Intermediate shaft bearings’ loads are not to exceed 80 per cent of the bearing
manufacturer’s allowable maximum load

• Flexible couplings can not exceed the radial, axial and angular alignment re-
quirement from the manufacturer.

• he main gear load is to be within the gearbox maker limit.

Part 13 Shaft Vibration and Alignment - Chapter 4 Shaft Vibration and Alignment
- Section 1 Shaft alignment Ed.2020 July

2.4.3.3 BV - Bureau Veritas

Some relevant requirements from Bureau Veritas.
• Aft most stern tube bearing must not have higher slope than 0.3mm/m.
• All bearing must remain loaded.
• Intermediate shaft bearings must not exceed 80% of the provided maximum
load from the bearing manufacturer.
• Minimum stern tube bearing length must be not less than two times the di-
ameter of the shaft. But it may be less if the bearing pressure is less than
0.7N/mm2 , then the minimum length of the bearing is to be 1..5 times the
diameter.
Steelships,Pt C, Ch 1, Sec 7 3.4.2 Ed.2020.Jan

2.4.3.4 ABS - American Bureau of Shipping

From ABS guidance notes on propulsion shafting alignment In general for all ships,
shaft alignment calculations and a shaft alignment procedure should be submitted
to ABS for reference [1, p. 7]. However, for propulsion shafting of diameters equal
to or larger than 300 mm, and propulsion shafting with no forward stern tube
bearing, they are to be submitted for review by ABS for compliance with ABS Rule
requirements.
The alignment calculations should include bearing loads and bearing reactions, shear
forces and bending moments along the shafting, slope boring details as applicable,
and a detailed description of the alignment procedure.
The alignment calculations are carried out for theoretically aligned cold and hot
conditions of the shafting and for the maximum allowable alignment tolerances.
They are to show that:
• Bearing loads under all operating conditions are within the acceptable limits
specified by the bearing manufacturer.
• Bearing reactions are always positive (i.e., supporting the shaft), except as
determined acceptable in accordance with current ABS Rule requirements.
• Shear forces and bending moments at the crankshaft flange are in accordance
with the engine manufacturer’s limits.

33
2. Theory

• The designed relative misalignment slope between the shaft and the aft stern
tube bearing is to be positive and not exceeding 0.3 ∗ 103 [rad].
ABS Guidance notes on - Propulsion shafting alignment, September 2019. American
Bureau of Shipping, New York. 2019

34
 3
Methods

The work process during the thesis work is to study the classification rules and uti-
lize the calculation methods stated in those for dimensioning the shaft and checking
safety factors. Those are sets of simple calculations where one chooses the correct
method for the specific type of application. The DNV class guidelines for the cal-
culation of shafts in marine applications are summarized in an interactive Excel
sheet where the designer is guided through the different steps of the calculations.
Further detailed analyses are done using the software package Nauticus Machinery
from DNV. The goal is to show that it is possible to decrease the dimensions of the
shaft and still achieve a safe design with good vibration behaviour.

3.1 Study on the IACS M68 formula

With the definitions made in the theory chapter, an analysis of the IACS M68.4
formula was done. With a statical approach, then implemented with a fatigue en-
durance limit. The goal is to try and mimic the IACS m68.4 equation from scratch
and come up with similar results.

An interesting factor of the IACS M68.4 formula is the stress concentration factor.
An investigation of this factor (KA) was done to understand how it is applied in the
initial formula. As well as to get a better instinct when emulating the approach for
the mimic equation.

3.1.1 Static calculation comparison

Based on the static torsional approach (equation 2.5), where the diameter is solved
out for hollow cylinder to compare it to the IACS formula, the marginal difference
can be better visualized. A pre-designed RoPax ferry (currently in operation) was
used for this comparison case, where the main input values can be seen in table 3.1.
According to the parameters for that vessel, the minimum allowed outer diameter
based on the ICAS M68.4 formula is 396mm. From the static torsional approach
286mm, giving that the maximum shear stress(τ ) in the formula (equation 2.5,
page 11) is the maximum allowed yield shear stress for the material. According to
Maximum shear stress theory, 0.5 ∗ σy (Tresca [5]). Note no safety is applied to the
286mm shaft, meaning it is the critical diameter, see side by side comparison in table
3.2. It is safe to say that the margin between each method is large. Note that the
diameter is in relation to area, so the ratio between them has to be taken squared,

35
3. Methods

3962 /2862 = 92%, meaning the IACS formula for minimum diameter provides 92%
larger shaft than the statical approach. Results can be seen in table 3.2.
Note, defining the shaft diameter based on statical load case is though inaccurate and
should not be considered, since the correct design should be based on the dynamic
perspective and the fatigue limit of the design. This statical analysis is solely to
just examine how the statical loading is comparing with the IACS M68.4 equation.

Table 3.1: Input parameters used for the calculation comparison. Operating RoPax
ferry, from a finished order from BERG database

The loadcase (RoPax ferry)

D 410 mm Designed outer diameter
d 80 mm Designed inner diameter
P 12600 kW Power
n 150 rpm Speed
σy 350 Mpa Yield strength of the material
σB 600 Mpa Ultimate strength of the material

Table 3.2: Calculation results from the IACS M68.4 formula, the statical torsion
approach. Response statical stress from torsion and bending on the shaft, based on
Von Mises and FEM method

Approach Min allowed Diameter [mm] Maximum stress [Mpa]

IACS M68.4 396 -
Static Torsion 286 -
Von Mieses 113
Based on 410mm
FEM 107

In addition, a statical safety analysis was done on the shaft with the given load case.
Where Von Mieses calculations were done on the shaft and a computerised method
of finite element analysis or simply FEM. According to DNV a conservative method
to estimate the bending loads on a propulsion shaft is to take 40% of the torque
load [11] and apply it as transverse bending on the shaft, which was done in this
case. Since FEM is based on the same theory as Von Mises, the same result was
expected, as it did. The FEM result can be seen in figure 3.1. The stress result for
both approaches is listed in the same table as the minimum allowed diameters were,
in table 3.2.

36
 3. Methods

Figure 3.1: Von Mises FEM analysis of the RoPax vessel.

This shows, though there is a resemblance between IACS M68.4 and the statical
torsional approach equation (equation 2.5), the result is not the same.

3.1.2 Fatigue approach on IACS M68

A dynamic, vibrating rotating shaft has to be designed in terms of fatigue limit. The
statical approach has been defined in section 2.2.1 described with equation 2.5. But
the torsional load variable in that formula has to be interpreted for a fatigue limit
(endurance limit) of a given shaft design, material, and load case. It is not known
what fatigue criterion was used by the people who developed the ICAS M68.4. In
this study, the fatigue limit method that was used was the Dang van criterion. Using
reduction factors for the fatigue limit (Se ) from mechanical coursebook [5].
Note, in this investigation, considering the IACS M68.4 formula, specifies that
σB < 600M pa [21], and Berg Propulsion default shaft material selection is 600Mpa
ultimate strength, thus this investigation will use σB = 600M pa.
Calculation steps and values can be found in the appendix B section B.2.
The Dang van criterion is a fairly new criterion that was issued in 1971. It gained
popularity since it addresses rotating principal stresses and is a multi-axial fatigue
analysis. By using this criterion for such a general application, shaft designs, load-
and material design, assumptions had to be made to perform the approach. The
equations that were used to define the endurance limit for high cycle/finite life, see
equation 3.1 (See page 28 Anders Ekberg, Multiaxial Fatigue [18])
σ1 − σ3 σ1 + σ2 + σ3 σf l σf l
σeq,dv = + Cdv −→ σedv = + Cdv ∗ (3.1)
2 3 2 3
Where Cdv is a material parameter defined with equation 3.2
This might look familiar, since it is similar to how in regular (not fatigue) failure
theory where the tri-axial stress is computed as T riAxial stress = Deviatoric
stress + Hydrostatic stress.

37
3. Methods

3 σf l − σf lp
Cdv = ∗ (3.2)
2 2σf lp − σf l

Where the σf l is the fluctuating loading and σf lp is the pulsation loading. Accord-
ing to Anders Ekberg [18] for a general estimation, the definition of these two loads
can be presumed the fatigue limit in uniaxial pulsating loading is 80% of that in
alternating loading. Meaning to estimate the critical fatigue limit for this case σf l
is defined as the reduced fatigue limit. From the previous definition the σf lp would
be 80% off that. [18]

Reducing the endurance limit (Se ) in respect of:

• Actual test data for endurance limit based on tensile strength. Estimated as
50% of ultimate tensile strength.
• Surface finish, machined ≈ 83%
• Torsional load case, 0.59%
• Stress concentration factor (SCF). It states in the IACS formula that their
equation is based on a SCF of 1.45. Therefore use the same concentration in
this approach.
Results and comparison on the mimic equation and the IACS M68.4 formula are
done in the Results chapter.

3.1.3 IACS M68.4 Concentration factor

A question emerged from curiosity, why IACS has defined the k in the M68.4 equa-
tion 2.36 and the Ck in the allowed torsional vibration formula M68.5 2.37, the way
they have done. Where do these multipliers come from. See each factors´ value for
different locations and components in the table on figure 3.3. Both equations have
an initial condition for the stress concentration factor (SCF) as a shaft component
with a step up with a radius fillet of 0.08d, resulting in a SCF=1.45 [21].
Applying a different designed component to the formula (for instance, keyway or
shrink fitted) the multiplying factor K and Ck maintains the required change for
different stress concentrations. The IACS formulas are developed and defined with
a radius fillet component as the default state of the equation(with a filler radius
of 0.08d). In the case of a different stress concentration component, the K and Ck
factors are in fact just ratios between the default stress concentration factor and the
current stress concentration, see figure 3.2. To get a better perception of how the
factors affect the result and where they come from, a brief investigation was done
on these factors.

38
 3. Methods

Figure 3.2: Table of k and Ck factors from IACSM68 document [21]

By selecting the factors for flange mount, for the intermediate shaft, and also for the
propeller shaft, Kint = 1.00 and Kprop = 1.22. Note that the factor for intermediate
is just 1 since the default stage of the formula is based on the stress concentration
of a fillet radius [21].
Stress concentration factor as previously went over in section 2.2.2, is multiplied by
the load or the strength of the material.
It can be assumed that the only significant load case for the intermediate shaft is
torsional load, and for the propeller shaft it undergoes torsional load and bending
as well . According to IACS the smallest fillet radius allowed is a radius of 0.08d,
which was then the critical design form in this investigation. Critical stress concen-
tration factor under torsional load and a bending load was acquired using charts for
theoretical stress concentration factors. The K factor provided by IACS (figure 3.2)
can be compared with the acquired stress concentration factors, see equation 3.3.

SCF
K= (3.3)
1.45
But since there are two different loads on the propeller shafts, it can be too simple
to expect that the stress concentration factor for the propeller shaft is only based on
bending load. So three methods were tested to compare the calculated K factor to
the IACS provided factor. One method (labeled approach A) where the calculated
K factor for propeller shaft is only based on bending SCF. For the next method,
since DNV reports a conservative way to define the bending moment, as 0.4 of the
torsional load. For example, for every ten units of torsion, we have four units of
bending, indicating for every 14 units of load 2/7 (4/14) is bending, and 5/7 (10/14)
is torsion. By that means the second method (labeled approach B) was by defining
the SCF as this ratio from each load case. And the third one (labeled approach
C) was defined the same way but according to Von Mises, where torsional shear-
load and normal-load from bending do not result in the same failure criteria [5], see
equations on image 3.3.

39
3. Methods

Figure 3.3: The three SCF approaches done in this study

With these three approaches, the resulted factor for the stress concentration K can
be compared with the provided one from the table. A visual comparison can be seen
in figure 3.4. It can be seen on the figure that the blue axis represents the provided
value of K from the IACS table (on image 3.2). Meaning that everything below the
x-line and left of the y-line is within the design criteria(or over designed) defined by
IACS.

Figure 3.4: Visual comparison of each calculated K (SCF/1.45), both as the critical
factor using the three approaches previously described. And the Berg (finished orders)
vessels.

From the plot (figure 3.4), it can be seen that the three methods are all above (left of

40
 3. Methods

the y-axis) the K for intermediate-shaft by a small margin. The three methods are
all based on the same formula for this K intermediate, that is why they do all stay
on the same horizontal line. One could argue that a 0.03 difference is neglectable.
For the propeller shaft K-factors, the three methods line up at different positions.
A and B are below the IACS K-factor, and method C is above. It is difficult to
justify one method from another, but if one would stick to method C, that would
mean that the IACS K factor for the propeller shaft is under-designed (which is
hard to believe, coming from such an institution). It is arguably too innocent to
estimate that the SCF for the propeller shaft is only based on bending. But out of
these three approaches, method-A is the closest one to the provided K-values. It
is convincing that IACS had possibly added some extra safety how they approach
and defined their provided stress concentration factor K, or done the interpolation
of bending SCF and torsion SCF differently than these three approaches done here.
Based on the closest approach (method A, propeller shaft SCF only based on bend-
ing load) three operating ship shaft designs were put under the test to see where
their SCF is on the IACS K scale. On these three ships, a fillet and its parameters
were used to define an SCF and based (like before) on equation 3.3 the K was cal-
culated and compared. See again the plot on figure 3.4. They all show a positive
result, since the IACS formula should be the critical benchmark for the STC, and
they all result in a lower one.

Ck is defined the same way as K. But since it is used in a different formula, M68.5
Permissible torsional vibration stresses there the material factor (ultimate tensile
strength) is inverted and thus is Ck defined the same way as K, but inverted, Ck =
1.45/SCF [21].

3.2 Simulation analysis

Nauticus Machinery is one of many software packages DNV Digital Solutions offers,
the program is specialised in marine propulsion design and analysis. The main
program provides the working environment, file manager and license server. Then
there are 8 different calculation tools all related to the rotating propulsion machinery.
To be able to do any kind of calculation with the program one needs to open a
calculation tool, each of them has a separate license such that one needs to choose
which calculations are needed and buy the appropriate license.
To make this thesis project possible, an academic software license was established
from DNV through a license server at the Chalmers University of Technology. The
license gives access to all of the calculation tools.

41
3. Methods

Table 3.3: Summary of Nauticus Machinery calculation packages

DNV Nauticus Machinery

Calculation Tool Background Used
Shaft Alignment Continuum Mechanics X
Torsional Vibrations Continuum Mechanics X
Shaft Fatigue Rule Based
Propeller Blade Rule Based
Gear Rating Rule Based
Gear Faceload Rule Based
Crankshaft Fatigue Rule Based
Controllable Pitch Mechanism Rule Based

However the tools were limited to some extent, at first, the Shaft Alignment tool was
limited to eight bearing elements, this was later updated and caused no limitation.
One important limitation remained for the Torsional Vibration tool, which was to
calculate excitation’s resulting from a 2 stroke engine.
This capability would have been preferred for comparison reasons. Of the differ-
ent types of propulsion system arrangements possible for a ship, the direct-coupled
low-speed shaft lines propelled by 2 stroke diesel engines are the most susceptible
to torsional vibration problems [25][28]. As discussed on page 6 “The engine’s main
excitation order coincides with the first torsional natural frequency of the shafting
system” [10, page 7].

Looking at Berg Propulsion portfolio to find suitable shaft line setups for simulation
analysis, the limitation of not being able to simulate 2 stroke propulsion is seen to be
of no real concern. The power plants are rarely of a 2 stroke type. The fact is that
BERG Propulsion specialises in designing and producing CP propellers which most
often are connected to a medium or high-speed diesel engine through a reduction
gear. The benefit with having a CP propeller is that it is easier to accelerate fast
through any barred speed ranges utilising the adjustable pitch ratio, compared to
a fixed-pitch propeller, “running through the barred speed range with zero or low
pitch” [10, p. 7]. Even better in regards to torsional vibration problems, electric
motors are getting more common for primary propulsion.
A typical application for BERG products are vessels that need to have flexible opera-
tion profiles and good manoeuvrability, this could be ferries, fishing vessels, offshore
service vessels and tug boats.
With this reasoning, it is suggested that the classification shaft dimensioning meth-
ods are overly conservative for many Berg shaft line applications. To be able to
analyse the effect of decreased diameters of existing shaft designs, examples of ref-
erence shaft lines were simulated in Nauticus Machinery. Analysis and calculations
were proceeded according to class requirements. In the process of setting up the
simulation models for the three reference shaft lines authors had access to a large
collection of specifications and drawings. Those drawings and technical specifica-
tions can not be included in this report.

42
 3. Methods

3.2.1 Nauticus Shaft Alignment

The shaft alignment tool provides a way to proceed with the lateral shaft analysis,
lateral covering the vertical deflection, and whirling vibration calculations. Deflec-
tion analysis in the horizontal plain are also possible but usually only needed if
high moments in the horizontal plain are realised. The tool provides detailed class
required bearing lubrication analysis.

Description of the tool from a DNV Nauticus Machinery brouchure:

Nauticus Machinery - Shaft Alignment provides efficient methods to
build a shaft model and analyze bearing loads, bending moments, de-
flections and more. It also provides customised verification results such
as hydraulic jacking curves and gap/sag values in flanged connections.
Whirling and axial vibration analysis also covered. It provides the after-
stern tube bearing lubrication analysis from two aspects: calculating
the required minimum shaft rotational speed to ensure sufficient lubri-
cation; calculating the oil film pressure and thickness distribution inside
the after stern tube bearing to find the maximum pressure and minimum
thickness. [15, p. 2]

3.2.1.1 Building the shaftline model

When beginning building a model in the Shaft Alignment tool detailed drawings
and specifications of the whole low speed part is needed. For direct coupled 2-stroke
propulsion the the model includes the propeller, propulsion shafting and the engine
crankshaft. For a geared propulsion system the model includes the propeller, propul-
sion shafting towards the reduction gear and the output shaft of the reduction gear
including the main gear wheel. Properties of the propeller, location and properties
of all bearings and all relevant loads for different operation conditions need to be
defined. For a full analysis extra conditions on top of the operation conditions need
to be added, those are GAP-SAG conditions for alignment procedures purposes.
The first step in shaft alignment analysis is to recreate the shaft elements as accu-
rately as possible using detailed dimensions from the Shaft Arrangement drawing
as well as extra specifications for couplings, bearings and gearbox. The shafting
is modelled with shaft elements, a shaft element can be cylinder or a cone, with
outer diameter, length and hollow bore. The shaft elements have weight, stiffness
and more settings. There are few more specific model elements, those are; flange,
spacer ring, shrink fit, bearing, propeller and joint. The main objective of the shaft
alignment analysis is to check how well the propulsion shafting is aligned with the
supporting bearings. The bearings need to be adjusted by offsetting them in a way
for the shaft to be supported evenly and the lubrication pressure to be within a safe
range. This is done to stay safe against abnormal or excessive wear on the shaft
and bearings. The model gives also the ability to check the natural frequencies in
lateral and analyse the mode shapes. This is called whirling vibration analysis, axial
vibrations are generally thought to give much lower stress values, therefore the axial
analysis is skipped, see section 2.3.3.

43
3. Methods

Table 3.4: The table summarises drawings and specifications needed for building
the shaft analysis model in lateral.

Information needed for shaft alignment

1 Shaft arrangement drawing
- Detailed dimensions of all shafts and different components
2 Sterntube bearing drawing
- Length of bearings and bearing clearance
3 Intermediate bearing drawing
- Length of bearings and bearing clearance
4 Couplings drawing
- Length, diameter, flange or not
5 GB / ME drawing and alignment criteria
- Gearbox or Main Engine, bearing clearings and loads
6 Propeller data
- Weight, Inertia, Thrust eccentricity

Figure 3.5: Model of the shaft arrangement in the Nauticus - Shaft Alignment tool.
The model represents the shaftline for the bunker tanker.

In figures 3.5 and 3.6 the shaft elements, modeling the actual shaft, can be seen
as light green, blue and gray. The green color represents submersion in water, the
blue represent submersion in oil and the gray elements are in air. This definition
is done to account for buoyancy effects. From left there the propeller hub is gray,
modeled as in air, and actually the weight is also modified such that the specific
kg
weight is 1 m 3 , this is done so that the calculated weight is essentially nil. This is

done because the load from the propeller, hub and blades, are defined specifically
for different operation modes. The arrow on top of the middle of the hub denotes
this load definition. The green shaft section is the propeller flange at the end of the
propeller shaft, where the propeller is attached using bolts. The propeller flange is
outside the hull submerged in water, just forward of the hub is the bossing and aft
stern tube seal, not shown. The blue colored part indicates the part inside the stern
tube, the cyan colored elements are the bearings. The aft stern tube seal keeps

44
 3. Methods

the lubricating oil inside the stern tube and water outside. The first blue shaft
element is the surface where the seal is located, next is the aft stern tube bearing.
Rule-of-thumb value for the aft stern tube bearing length ranging from 1.5D to 2D
depending on application and needs to be verified during the analysis. Forward stern
tube bearing is much smaller than the aft stern tube bearing, reason being the large
loads on the aft bearing, both have two contact points defined at either end, noted
with a red line. The stern tube bearings are fixed solid and usually have zero slope,
in some cases the aft bearing is manufactured with small positive slope, fraction of
a millimeter, to account for relative slope difference between the shaft and bearing.
Positive slope being the bearing is higher at forward end.

Figure 3.6: Model of the shaft arrangement in the Nauticus - Shaft Alignment tool,
3D view. The model represents the shaftline for the bunker tanker.

Where the blue color stops and gray takes over the forward stern tube seal is po-
sitioned and keeps the lubricating oil inside the tube. The gray part of the shaft
is inside the engine room. Following the shaft to the right we have a thin section
with a large diameter, this is a brake disk for being able to stop the shaft, the
brake disk is bolted to a shrink fit flange coupling. The flange is in brown color
and bolts to the gearbox flange in yellow color. The large wheel is the main gear
output wheel supported by the output shaft and journal bearings on either side.
The arrows represent the forces acting on the wheel when the pinion wheel transfers
torque from the propulsion engine/motor. Furthest to the right is a thin wheel, this
is a thrust-bearing that supports the axial force from the propeller.
In shaft alignment the shaft is really acting like a beam, which is rotating resulting
in possible whirling vibrations. The fact that the shaft needs to be able to rotate
means it is supported by bearings, typically white metal bearings with lubrication.
Water lubricated composite stern tube bearings are also available. The principle

45
3. Methods

is the same, the lubrication needs to be able to form a thin film supporting the
shaft without the surface scraping. To align the bearings in such a way that the
lubricating pressure in the bearings stays within suitable range , is one of the main
objectives of the shaft alignment task. The fact that the shaft has just certain
amount of stiffness and is actually curved can make this challenging. This means
the shaft supports can not be aligned in a straight line, rather they need to have a
vertical offset. To complicate this further the shaft alignment analysis is done for
different conditions, with the shafting components cold and static, static meaning
the shaft not turning. Another warm static check with the shafting component at
running temperatures but stopped, not turning. Two running conditions are checked
also, the MCR condition and 0 pitch condition, for CPP propellers.

3.2.2 Nauticus Torsional Vibration

Nauticus - Torsional Vibration tool is a highly specialised tool for calculations of

vibratory excitation’s and responses in marine type of rotating machinery. The
tool includes different types of prime movers, specific shafting coupling connection
elements, damper and gear elements found in marine propulsion shafting as well as
the propeller definition. The tool has been found to offer a broad range of predefined
plotting and result printout definitions which facilitate standard reporting for marine
application but could as well be used for analysing rotating shaft systems in general.
The calculation are made in the frequency domain, excitations and reactions related
to engine speed, all but the ice loads calculations. The ice loads calculations are
specific for ice classification and simulate load amplitudes in time domain resulting
from the propeller blades milling through ice [15].

3.2.2.1 Inertia and stiffness

The building blocks of the torsional vibration mass elastic model are nodes that
define the inertia masses and shaft stiffness elements which connect the inertia nodes.
The inertia, noted with two green circles connected vertically by narrow line. A
stiffness is an arrow connecting two inertia’s, the arrow points in the direction of
power, from a power source to a power user. On figure 3.7 the nodes no. 1 and
5 are electric motors, a power source. The power sink, the user in this system is
the propeller, inertia element no. 14. The electric motors and the propeller are
inertia elements with specific extra features defining the physical properties of those
components, the rest of the inertias in this system are regular lump-mass elements.
The list of inertia and stiffness elements can be seen in table 3.5 [13].

46
 3. Methods

Figure 3.7: Simple view of Nauticus torsional vibration model. The mass elastic
model of the Tanker.

Different compination of inertia and stiffness elements can then be grouped in specific
groups to form a higher level machinery component. The dotted boxes in figure 3.7
can be seen representing elastic coupling, see lump mass nodes 3 and 4 connected
by a coupling stiffness, similar with notes 7 and 8.
Notes 9 and 10 are regular lump masses connected by mess stiffness, when adding
the gear group joining those, creates a reduction gear that can have a gearing ratio
and more properties. The electric motors and propelle are just a single inertia
objects. However looking at inerta no.2 and the stiffness connecting no. 1 and 2,
those represent the inertia and the stiffness of the output shaft of the electric motor.

Table 3.5: List of the basic modeling elements availible in the DNV Nauticus
Machinery - Torsional Vibration tool. [13, p. 16]

Modeling elements
Inertia Stiffness
Lump mass Shaft stiffness
Cylinder Crank throw stiffness
Electric motor Coupling stiffness
Gas turbine Mesh stiffness
Steam turbine Damper stiffness
Propeller
Generator

More detailed view of the Tanker model is presented in figure 3.8, where the stiffness
is represented by spring connections and damping by damper icons. The inertia
values can be seen in black straight above the lump mass elements, moment of
inertia given in kg ∗ m2 . The stiffness values can be seen in purple above the shaft
stiffness connections, represented in the units of N ∗ m/rad. The damping is given
by red values below the stiffness connections, here they are represented as damping
magnifier values.

47
3. Methods

Figure 3.8: Detailed view of Nauticus torsional vibration model. Showing the
different modeling elements along with values for inerta, stiffness and damping.

3.2.2.2 Damping

Damping for a multi-dimensional systems can be very difficult to determine. This

is different from stiffness and inertia. Therefor when the physical propulsion system
has been installed in a vessel the vibration behaviour needs to be measured to check
if estimated damping values have been acceptable [31]. For this reason the damping
values used are often general values that have been shown to fit the application. For
the torsional vibration models damping magnifiers were used. The values M = 180
for steel shafts and M = 100 for gear meshing are rule-of-thumb values that rotor
dynamics experts have seen to give good results when using Nauticus Machinery.
The damping magnifiers for flexible couplings were computed from damping spec-
ification values, for a typical flexible coupling it is good to have a value of about
M = 7. Similar rule-of-thumb estimates for damping values can be seen in [31,
p. S-167] for the TVC analysis program SimulationX.

48
 3. Methods

Table 3.6: Comparison of how damping values and coefficients are defined for the
programs Nauticus Machinery and Shaft Designer [13] [4].

Damping values
Nauticus Machinery Shaft Designer
Linear viscous damping Linear viscous damping
k C
[N m ∗ s/rad] [N m ∗ s/rad]
κ Undimensioned damping factor κ Undimensioned damping factor
ψ Ratio of damping energy ψ Ratio of damping energy
M (Dynamic) Magnifier Q Vibration magnifier
- Not defined  Percent of critical damping [%]
- Not defined M Dynamic magnifier
Phase velocity of vibration Phase velocity of vibration
ω ω
[rad/s] [rad/s]
C Dynamic stiffness [N m/rad] K Stiffness [Nm/rad]

The undimensioned damping factor κ has also be seen with the notation µ. The
damping for flexible couplings is often given by this undimensioned factor. Very
important to notice is the difference between the damping magnifier values used
by Nauticus and Shaft Designer. Since the project work was done using Nauticus
Machinery, M will represent the damping magnifier used by Nauticus Machinery,
also called dynamic magnifier. When setting up the torsional vibration analysis
models it gave best results to convert the damping values stated for couplings and
more to this damping magnifier M. Doing this conversion was also really good for a
verification check to see if one is using the correct values. It was for example in some
cases confusing if specifications were stating damping as undimensional damping
factor κ or ratio of damping energy ψ. But by checking the dynamic magnifier it
quickly reveals if the specifications have been misinterpreted.

Table 3.7: Conversion table for different damping values, relevant for Nauticus
Machinery [13].The damping magnifier M as used by Nauticus is more often pre-
sented as Q. More comprehensive conversion tables available in the Nauticus and
ShaftDesigner documentation.

Damping conversion table

ψ κ or µ k
√ √ √
4·π 2 +ψ 2 1+κ2 C 2 +k2 +ω 2
M= ψ κ k·ω

The damping of the propeller is defined with the Frahm’s propeller damping factor
DF or Archer’s propeller factor DA , discussed in theory chapter 2.3.2.1. Other ways
of defining the propeller damping are available also. If propeller damping is not
given on specifications. Typical, rule-of-thumb value for the propeller damping used
is Archer number = 27.

49
3. Methods

Table 3.8: The different possibilities for defining the propeller damping.

Damping Model Input

Manual Dynamic magnifier or dynamic magnifier percent
Schwaneke Propeller geometric properties
Archer Typical range 25-30
Frahm Typical value 2.8
Variable Knuckle-Point of propeller variable damping curve

3.2.2.3 Propeller
The propeller is the power sink in the system and needs considerable amounts of
inputs, depending on if it is a FPP or CPP, ICE classed or not.
The program computes theoretical propeller curve as default, this defines the rela-
tion between the speed of the propeller and the power it consumes. This is a good
enough estimate to be used in many cases. For both of the systems having CPP
propellers the RoPax and the Bunker Tanker this was used.
The Dredger on the other hand has FPP propeller and the two operation profiles
resulting in different load on the propeller. On one hand free running with empty
holds and the other in dredging condition, loaded with sand and silt, dragging the
draghead along the bottom.
Propeller torque curves were provided, those needed to be converted to power curves.
The pdf files with the torque curves showing torque in kNm versus rpm were im-
ported in AutoCad to measure few torque values. The figures were scaled such that
one unit length is equal to one kNm, by this the length of the vertical measuring
lines represent the torque value for the respective propeller speed. The torque values
were then converted to power values. The list of calculated power versus propeller
speed was written in a Nauticus type of text file which can be imported as a propeller
curve.

(a) Free running condition (b) Dredging condition

Figure 3.9: The propeller torque curves for the two conditions after measurements
in AutoCad.

The conversion of propeller torque in kilo-Newton-meters kN m to power in kilo-

50
 3. Methods

Watts kW was done using:

np
Pp = Qp · 2 · π · (3.4)
60

Table 3.9: Notation for propeller torque and power for the propeller curve conver-
sion.

Qp : Propeller torque [kN m]

n : Propeller speed [rpm]
Pp : Propeller needed power [kW ]

In appendix C.5.3 the reader can find table C.1 listing the torque and power values
as well as Nauticus standard format text files for the propeller power curves.

3.3 The improved diameter approach

The method to design a shaft line using DNV-CG-0038 document integrated with
Nauticus machinery TVC and Alignment, is "in theory" the whole methodology
approach of the thesis project. A guide to the method itself and the case study are
reflected in this section here.

3.3.1 The method

Defining an improved step-by-step solution method for the minimum diameter of the
propulsion shaft line was the plan. By integrating computational power, empirical
methods, and equation approved by the classification society, the improved method
was established. Though the aim was to push the rules and regulations to their
boundary and try to defy the classification societies, the final calculation route that
was come up with should all stay within regulations and rules (has not been tested
by a class expert when this is written).
The procedure consists of different steps and there is no argument that this improved
method takes a lot longer time to perform compared with the current approach, the
ICAS 68.4 formula . Since this method approach is specifically outlined for Berg
Propulsion, a conventional user will experience trouble following the step-by-step
approach because Berg Propulsion has its own preliminary shaft system software
(called ProjCalc). It provides (preliminary) design parameters, both for shaft com-
ponents, engine, and propeller. Meaning this project calculation approach for the
minimum shaft diameter relies on some of the output parameters from ProjCalc. Of
course, conventional users can follow this approach, but one has to have the input
to do so.
note! in this section writer refer to the procedure as an explanation to a user in a
third person. Though the solution method was the same for them.
A user starts by having a preliminary design of a shaft and uses the improved
method recommended by DNV, defined in the document DNVGL-CG-0038, pre-
viously talked about section 2.4.2.2. The method defines the shaft in respect of
fatigue. To perform this method one must have the preliminary design data, such

51
3. Methods

as what kind of a component, bore of the shaft, surface finish, and other parame-
ters of the design. The user must also have torsional vibration data of the whole
propulsion system (TVC), simply the application factor (KA). The data can ei-
ther be gathered from a TVC report or more conservatively estimated with tables
and statements from DNV. In addition, if the vessel has ICE classification the esti-
mated application factor for ice crash vibration must be calculated (section 2.4.2.1).
Defining the application factor is not as easy as saying it, the user must have knowl-
edge and access to a TVC software (such as DNV Nauticus Machinery TVC, talked
about in previous section 3.2.2). Now to continue with the DNV improved method
(DNVGL-CG-0038) the user must also have the rotating bending moment. A sim-
ple and conservative way is to define the bending moment as 40% of the nominal
torque on the shaft [10] [11]. A more precise bending moment can be acquired using
alignment software (such as DNV Nauticus Machinery Alignment, talked about in
previous section 3.2.1). The same goes for that procedure, knowledge, and skills
of how to set up lateral alignment cases in such software. The user then computes
the safety factor for high cycle fatigue (HCF) and low- (LCF). Each safety must be
within the defined minimum safety of the method, defined by DNV, 1.25 for LCF
and 1.60 for HCF. With that begin said, the user can decrease the diameter (or
alter component dimensions, e.g. increase the radius of a flange, switch component)
until the desired safety is reached. In this project, the diameter was decreased until
either one of the LCF or HCF minimum safety was reached.

At this point in the process, the user has either a shaft design based of a TVC report
or using preliminary application factors. The next measure the user must take is to
check the propulsion shaft for its minimum shaft diameter based on yield load from
the propeller failure load. That is the check if the propeller shaft is stronger than
the blades of the propeller because you want the shaft to survive the maximum load
of the propeller blades before they break off. According to class, this is only required
when the hip has an ice-class [12]. But at Berg Propulsion, they do this for every
shaft line, since grounding the propeller can happen to any vessel, ice-classed or
not. The user can then take its shaft line design that has been defined by the steps
here previously, and check it with the propeller failure load, method mentioned in
section 2.4.2.1. It can be that the aft-most propeller shaft must be increased from
the previous study to fulfill this criteria. At this stage in the progress the shaft
is designed in respect of fatigue limit, it fulfills the yield of one load crash of the
propeller, if the user used TVC simulation then shaft is also designed within the
natural frequency limit and classification torsional vibration regulations. The last
step is to use alignment software to check for lateral vibration frequency, because
the previous assessments have not taken into consideration the lateral displacement
and its supports (bearings).

If all these design applications are met, the vessel shaft-line should be smaller,
lighter, and thus cheaper than the previous conservative method.

52
 3. Methods

3.3.2 Case study

This method for improved shaft diameter is considered to be used when designing
a shaft, but it can also be applied to previously designed shaft lines. Either im-
prove the design or to compare how much could have been saved. Hence the task
was to look into few old/finished Berg Propulsion orders and see how over-designed
the shaft lines are, and how much material could have been saved using this new
approach.

Three orders were looked into. For the sake of confidential information of these shaft
lines, the vessel name, owner, or other specific information will not be published.
The three study cases were:
-RoPax ferry, 12600 KW (each engine) dual prop vessel, 1B ICE-class.
-Dredger vessel, 4000KW, no ICE-class
-Bunker tanker, 2500KW, 1C ICE-class

Each vessel’s shaft line was investigated, and the conclusion was to select few dif-
ferent locations of each shaft line and place that under investigation. To keep the
diversity up, various components were selected. Based on the length and complexity
of the shaft, two locations were selected on the Bunker and Dredger and three on
the RoPax shaft line.

Table 3.10: Each study location label of the shaft lines

Vessel Shaft part no Component

Bunker tanker A1 (Prop) Flange
Bunker tanker A2 (Interm) Plain shaft
Dredger vessel B1 (Prop) Shirnk fit
Dredger vessel B2 (Interm) Flange
RoPax SB C1 (Prop) Flange
RoPax SB C2 (Interm) Shrink fit
RoPax SB C3 (Interm) Flange

See each shaft location in figure 3.10. Where each location is marked with a shaft
part no. consistent with labelling in table 3.10.

53
3. Methods

Figure 3.10: The study cases on the three-shaft lines under investigation. Note,
those three shaft lines are not in 1:1 ratio of each other.

By using the procedure, here previously mentioned in section 3.3.1 each part location
is calculated again with that approach. Few variables were applied to each approach
in the investigation, resulting in a few different outputs for each re-calculated shaft
location. Based on estimated and simulated torsional vibrations (KA factor) and
same set for rotating bending moment(M b).
Approaches used:
-Bending moment equals 40% of nominal torque and KA estimated from class.
-Bending moment simulated from alignment software, and KA estimated from class.
-Bending moment =40% of nominal torque and KA simulated from TVC software.
-Bending moment simulated from alignment software, and KA simulated from TVC
software.

Using these input variables (two for each input parameter), four different outputs
were provided. A comparison from each approach could then be investigated to
monitor how much it affects the final result. In addition for the propeller shaft,
there the minimum shaft diameter must fulfill the ice-class criteria required by the
classification. This ice requirement is mentioned in section 2.4.2.1 based of DNVGL-
RU-SHIP Pt.6 Ch.6 12.5.2. This additional application is based on that the propeller
shaft is stronger and could withstand a failure load on the propeller-blades. This
requirement was also calculated and applied to the propeller shaft locations (mean-
ing that now five different output criteria are for each investigation)
The final conclusion of a minimum allowed shaft diameter was defined by the critical
diameter of these five approaches.
The "new" shaft diameters were simulated in shaft alignment software to investigate
if the current smaller shaft line would fulfill the whirling/alignment requirement by
classification.
The finalised approved diameters could then be compared with the previously de-
signed parameters defined by Berg Propulsion in the orders.

54
 3. Methods

3.3.3 Additional input values

The methodology in the previous section requires many other input parameters and
values than has been mentioned.

3.3.3.1 Dimensions of each component

Each different shaft component result in a different stress concentration. In con-

sideration of each component, the investigation did only have the shaft diameter
as a variable, and the dimension of each component was fixed from the initial de-
sign. Meaning, for example, though the shaft diameter from the initial design had
decreased by some percentage the radius of a flange component stayed the same.

3.3.3.2 Surface roughness of shaft

In the method the Ry , which is the surface roughness, the maximum height of the
profile. Also the surface roughness Ra , arithmetical mean deviation of the profile,
should normally not exceed 1.6µm. Ry can not be less than 1µm and can also be
estimated as six times Ra . In this study, the estimated critical surface is used. That
is Ra = 1.6µm and Ry = 6Ra = 9.6µm.

3.3.3.3 Propeller and ice-applications

Previously talked about in section 2.3.2.3. Where the torque excitation from ice
crash can be calculated with empirical formulas. The main input values that are
needed to compute the formulas are the Propeller and hub diameter, Pitch ratio
at radius 0.7 at bollard condition (which can be acquired from Bergs ProjCalc
software), and the mass moment of inertia of all members of the shaft up to the part
under investigation ratio against the mass moment of inertia of complete propulsion
system. (Note! Alteration of the shaft diameter by a few percentages does not have
a large impact on this ratio. meaning that the ratio was considered the same for
shaft calculations with variable diameter)
The rules regarding the shaft to withstand the failure of the propeller, mentioned
in section 2.4.2.1. The main input parameters needed to compute this minimum
propeller diameter are the chord length at a relative radius where the investigation
is taking place and max blade thickness at that particular location(which is provided
by ProjCals, Berg Propulsion software).

3.3.3.4 Material

The standard shaft material chosen by Berg is ’28mn6 steel’. It limits the IACS
formula having the ultimate tensile-strength:600MPa and its yield-strength:350MPa.
The same goes with the propeller material having it as a standard choice, using the
failure stress 0.6*yield-strength+0.4*tensile-strength of the bronze blade material,
395MPa.

55
3. Methods

3.3.4 Calculator guide

This whole calculation method is vast and spans many rule and guideline documents
from DNV, not to mention the significance of performing the TVC and alignment
simulations. It can be difficult to grasp the steps taken here in this thesis investi-
gation and in which order to follow them. An interactive script was made (based in
excel) where the user is lead through each step of the process. It can be said that
the calculation can be divided into three platforms, the main instruction script, a
DNV-CG-0038 document calculator, and a TVC/alignment software of choice.
Though these scripts were made in favor of other users to manage, the writers even-
tually used these scripts to perform the investigation, because of it´s practicality.
Images and screenshots from one of the Excel scripts can be seen in figure 3.11,
and with larger and more images in appendix A.3 for the DNV-CG-0038 document
calculator and appendix A.2 for the the main instruction script. A flow chart of the
DNV-CG-0038 document can be seen in appendix A.1.

Figure 3.11: Few screenshots from DNVGL-CG-0038 document calculator.

56
 4
Results

In this chapter, the results from the procedures previously talked in the report are
revealed. The investigation on the IACS formula (IACS M68.4) is developed with a
mimic formula, trying to simulate equivalent output. Both formulas are presented
with the same input to study if they provide the same result, which could be used
to argue the validity of the mimic formula.
The main investigation, developing the improved shaft diameter method, using TVC,
and alignment simulations on the case study shaft lines. With a physical comparison
on the initial- and improved shaft lines to see how much weight can be saved.
The case studies TVC behaviour is investigated for incremental dimensional changes.

4.1 IACS M68 compared with the mimic-formula

Many different approaches were done to mimic the IACS M68.4 formula, and one
stood out comparing to the others. Establishing a formula solving for the diameter
of a circular hollow shaft based on torsion, and defining the maximum allowed stress,
based on finite-life fatigue endurance limit stress obtained from Dang Van failure
criterion. The IACS M68.4 formula (the one that Berg uses for their min shaft
diameter) can be seen in equation 4.1
v
u p 1 560
u
dmin = F ∗ ku
3 ∗ d4i
∗ (4.1)
t n 0 1− σB + 160
d40

The solved (and simplified) Mimic formula can be seen in equation 4.2
v
u p 1 442.13
u
dmin = 3
u ∗ d4i
∗ (4.2)
t n
0 1− σB
d40

The finalised simplified edition of the mimic formula is defined like so (see equation
4.2). But if one would change the stress-concentration factor (SFC) this version
of the formula is then not valid. Because the fatigue endurance and the effect of
the SCF are "deep" within the 442.13 number, A full step-by-step deduction of the
formula can be found in appendix B section B.2.

Comparing the equations can be a little tricky since they both require many vari-
ables. A comparison was done on a finished Berg order for a RoPax ferry (see
designed data on table 3.1, on page 36) for a simple comparison. For this specific

57
4. Results

load case setup, the minimum required diameter based on variable power can be
seen in figure 4.1.

Figure 4.1: Comparison of the IACS M68.4 formula and the writer’s mimic for-
mula of that one. The mimic formula is labeled ’New DV’ (new Dang Van). With
all design parameters the same and having input power the variable.

The results were surprisingly similar for the writers. The mimic equation and the
IACS one had been compared with random input variables for comparison over the
developed time, but when this RoPax comparison was done it was not expected to
get so similar result.
Another comparison was done on the two equations. Where different input variables
had a randomly generated value.
- Inner diameter = 0
- Power = random generated variable from 800-10,000 KW
- Cycles (RPM) = random generated variable from 50-200 rpm
- Material ultimate tensile strength = random generated variable from 300-600 Mpa
(Each range was solely selected based on the writers’ awareness of the expected size
of the input values). Each equation received the same sequence of random inputs
100 times. The response graph for the output can be seen in figure 4.2.

58
 4. Results

Figure 4.2: Minimum diameter for IACS M68.4 and writers mimic formula, us-
ing random generated input data. Data values (input and output) can be found in
appendix C section C.1

From the graph (figure 4.2) it can be seen how some sequence of input values result in
very similar minimum allowed diameter result, while others have a margin between
them. The average margin between them is about 8.4mm. It can also be seen that
the mimic formula is more conservative since the minimum allowed diameter is in
most cases higher than the IACS M68.4 formula defines. The random generated
input data values, and each calculated output for the equations can be found in
appendix C section C.1.

4.2 Optimized dimensioning of propulsion shaft-

ing
Following the procedure defined in the Methods chapter, the shaft lines in the case
study were assessed.

4.2.1 Comparison on finished Berg orders

The seven shaft line locations (see image 3.10 on page 54) were put under the
scope. Every input parameter was gathered, some could be acquired from dimen-
sions from a technical drawing, some from ProjCal, and others from Nauticus (TVC
and alignment). A comparison was done to six different approaches to the diameter.
Approaches used:
-DNVGL CG0038 document method based on, bending moment equals 40% of nom-
inal torque and KA estimated from class.

59
4. Results

-DNVGL CG0038 document method based on, bending moment simulated from align-
ment software, and KA estimated from class.
-DNVGL CG0038 document method based on, bending moment =40% of nominal
torque and KA simulated from TVC software.
-DNVGL CG0038 document method based on, bending moment simulated from align-
ment software, and KA simulated from TVC software.
-Minium allowed propeller shaft diameter, based on propeller failure load.
-The IACS M68.4 formula.
-And of course compared with the initial diameter.

The results for each approach can be seen in image 4.3, where one can see a compar-
ison of each method in a column chart. The initial diameter and finalised diameter
for each shaft location are labeled with blue and red icon tags. Where the final
diameter lines up with the largest (most critical) diameter, excluding the IACS for-
mula column.
Each approach is labeled with an abbreviation of the method descriptions which
was listed here up in this section. The definition of each abbreviation can be seen
on table 4.1, and should be used if one is not sure what each approach (column)
represents

Figure 4.3: Minimum allowed diameter from each approach, for all case study
shaft locations, for comparison. Blue cross marks the final design from BERG, red
bar the final design in this thesis investigation(our improved method).

The results show by using this method for calculating the shaft diameter the min-

60
 4. Results

Table 4.1: Description of abbreviations used in figure 4.3

Approach labels Description

Using estimated application factor(KA) values.
Est KA
Estimating the bending moment (Mb) to be 40% off the
0.4 Mb
torque.
Simulating the torsional vibration of the system to more
TVC KA
precisely estimate KA.
Simulating the lateral alignment of the shaft to more
Align.M. Mb
precisely estimate the bending Mb.

Prop.S. Propeller-shaft
Intermediate.S. Intermediate-shaft

imum diameter can be smaller than the IACS formula states. In every case, the
previous diameter (from the old orders) could be decreased, except for A1.
When the shaft location is the aftmost part of the propeller-shaft the design criteria
’minimum diameter from propeller failure load’ applies, and each time it passes all
of the other four DNV CG 0038 methods. Meaning that the minimum diameter
limit that this study is trying to reach with the comprehensive calculations methods
(using DNV CG 0038, TVC, and Alignment) is outmatched when this additional
criterion is added for that shaft location. (The blade failure load, based on DNVGL-
RU-SHIP Pt.6 Ch.6 12.5.2 [12]). That is why the minimum diameter and the blue
column for A1, B1, and C1 on figure4.3

The other four approaches yield a similar result for some (graph, figure 4.3). The
main reason that some approaches (column) have the exact same result is that each
method could share the same critical input parameter. Let’s set up an example
case, it does not matter for a certain design if one estimates the rotating bending
moment (40% of torque) or acquires it from alignment because for this hypothetical
situation the torsional dynamic shock from ice crash (KAi ce) defines the minimum
diameter. Or if another hypothetical scenario is looked at, the KAP (which is always
inputted as 1.4) is the most critical input parameter of this hypothetical example,
it outmatches the other application factors and the effect of the rotating bending
moment is lesser than the KAP. This means in this case it does not matter how one
acquires the application factors or the rotating bending moment. Because of this
effect, the results for the shafts (in figure 4.3) do show few diameter outputs as the
same.

4.2.1.1 The Tanker shaft

The investigation for the Tanker shaft line showed that using the improved method
yields the same minimum diameter for the aftmost propeller shaft, location A1.
The plain shaft in location A2 can be decreased from 260mm to 230mm.All current

61
4. Results

and initial parameters for the Tanker shaft line can be seen in figure 4.4, the yellow
highlighted volume represents material that could be cut off.
Larger scale of figure 4.4 can be found in appendix A.4

Figure 4.4: Dimensions for the Berg design of the shaft and the currently improved
dimensions. The material that can be removed/also represent the initial design is
highlighted yellow on the shafts on the image.

4.2.1.2 The Dredger shaft

The investigation of the Dredger shaft line was done on two locations. Based on
the current improved method both of these shaft locations can be decreased from
the Berg order design. Where the aftmost propeller diameter, B1, could be changed
from 315mm to 300mm. Note, for the sake of total volume change the tapered parts
and especially on this shaft where the propeller attachment is a tapered cone-thread,
the outer diameter was just linearly decreased. The other location, the connector
flange between both shafts, B2, can be decreased from 255 to 225mm. All current
and initial parameters for the Dredger shaft line can be seen in figure 4.5, the yellow
highlighted volume represents material that could be cut off.
Larger scale of figure 4.5 can be found in appendix A.4

62
 4. Results

Figure 4.5: Dimensions for the Berg design of the shaft and the currently improved
dimensions. The material that can be removed/also represent the initial design is
highlighted yellow on the shafts on the image.

4.2.1.3 The RoPax ferry shaft

Based on each location the shaft could be decreased based on the result (in figure
4.3). C1 could be decreased from 520mm to 465mm. C2 and C3 had a slightly
different diameter in the initial design, they resulted in the same diameter in the
current result. C2 from 420mm to 370mm and C3 405mm to 370mm.All current
and initial parameters for the RoPax ferry shaft line can be seen in figure 4.6, the
yellow highlighted volume represents material that could be cut off.
Larger scale of figure 4.6 can be found in appendix A.4

63
4. Results

Figure 4.6: Dimensions for the Berg design of the shaft and the currently improved
dimensions. The material that can be removed/also represent the initial design is
highlighted yellow on the shafts on the image.

4.2.2 Weight reduction of raw material

The total steel weight of the shafts that can be removed resulted from the current
method and the final design in the three Berg order was calculated, and can be seen
in table 4.2.

Table 4.2: Steel weight reduction on each shaft line

Before[kg] After[kg] Difference [kg]

Tanker 3118 2889 229 7,35%
Dredger 7805 7112 693 8,88%
Ropax 47972 43379 4592 9,57%

The total weight reduction is largely different from each shaft line, seeing that the
RoPax shaft line weight reduction is more than ten times larger than the Tanker.
The overall volume of each shaft line has to be considered. When the weight opti-
mization is compared in relation to the proportion of initial volume, it can be seen
that each shaft line can reduce weight by a similar margin, 7.4-9.6% (see table 4.2)
Just for an interesting visual comparison, a rendering of the volumetric size com-
parison of each mass of steel that can be decreased from each shaft line was done,
see image 4.7.

64
 4. Results

Figure 4.7: Visual comparison of the weight/volume of raw steel that could be saved
from the three orders

4.2.2.1 Sensitivity-DNV CG 0038 calculations

By performing manual inputs into the calculations the writers developed an aware-
ness of which input parameters are important, and who have a minor effect on the
output result.

From the input parameters, a user needs to operate (calculate) the DNV CG 0038
calculator (section 3.3.4) some have more influence than others.
The inner bore diameter of the shaft in Bergs case is defined by the hub size, regard-
ing the hydraulic equipment. If one would like to optimize the weight of the shaft
the inner bore can be increased, since the outermost part of the shaft is the body
that reacts most in torsion and bending (talked about in section 2.2) [5]. It can be
seen in the graph in figure 4.8. where the inner bore was increased in the ratio of the
outer diameter. The inner bore effect does not influence the result much until the
inner bore is about 60 to 70% of the outer bore (IACS M68 document states if the
inner bore is less than 40% of the diameter, the inner diameter can be disregarded
[21])

65
4. Results

Figure 4.8: Unitless graph showing the effect of the inner diameter bore in relation
to the outer diameter on resulted safety of a shaft design

The effect from the surface finish is included in the DVN CG 0038 document, but
with inspection, the size value of the surface finish does not affect much the fatigue
life of the shaft. It states in DNV that Ra (Surface roughness-mean deviation of
the profile) should not be more than 1.6 µm, and Ry (Surface roughness, peak to
valley) can be estimated Ry ≈ 6Ra [12]. So with that relation, running a vector up
to Ra=1.6 it can be seen that for every component the output of the calculations
does not shift much( based on DNV CG 0038 [10]).
Crucial parameters are the outer diameter, material, the KA (application) factors,
and the reaction loads (bending and torsion). But as noticed in the diameter result in
the sections here above (see section 4.2.1 and graph in figure 4.3), it can be seen how
the result vary little from how the KA (application) factor or the bending moment
is acquired. With a grain of salt, the sensitivity of the result is small depending on
how the user defines these two inputs.

4.2.3 Shaft alignment - whirling analysis

This section will summarise findings from the Shaft Alignment analysis, the results
focused on are the shaft vertical deflections, bearing slope and lubrication pressure.
Another important thing is the shaft rotational speed where the shaft has natural
frequencies. The program gives the revolutions per minutes and plots the mode
shapes for the whirling vibration modes, only the first 3 modes are printed out. The
program calculates much more, there can be set a limit to how high frequencies the
program computes. The default setting is quite high and the program usually gave
first 10 to 20 of the natural frequencies and mode shapes.

Each case study shaft line is compared with the initial (approved) alignment reaction
result to the optimised diameter shaft line alignment reaction result. Each diameter
modification can be seen, for the RoPax ferry in section 4.2.1.3, for the Bunker
tanker in section 4.2.1.1, and Dredger vessel in 4.2.1.2

66
 4. Results

4.2.3.1 Ro-Pax
The RoPax had few changes done on the shaft line, as it can be seen in section
4.2.1.3. The behaviour of the deflection does not change much for the naked eye,
reflecting from the graphs, see deflection graphs in figures 4.9a and 4.9b. But pres-
sures and slopes do change. Small increments on the vertical offset of the bearings
were changed to meet regulations on the slopes. (All the bearing reaction, before
and after shafts optimisations can be seen in appendix C.3). The main noticeable
reaction after the adjustment was the stern tube slope in 0-pitch condition, where
it was a little bit over 3mm/m which is the standard for full pitch operation [11]
The behaviour of whirling does shift from locations on the shaft, as well as the crit-
ical speed (natural frequency of the system) of each vibrating mode shifts down in
rotations, see graphs in figures 4.10a and 4.10b.

(a) Deflection original (b) Deflection decreased diameter

Figure 4.9: Shaft deflection for all operating condition, side by side comparison
for the RoPax vessel.

(a) Whirling modes original (b) Whirling modes decreased diameter

Figure 4.10: Whirling modes in forwarding condition, side by side comparison for
the RoPax vessel.

All graphs in larger scale and bearing reactions can bee seen in appendix C.3.

4.2.3.2 Bunker tanker

For the Bunker tanker, the propeller shaft is could not be decreased from the previous
design due to propeller blade failure load criteria. Thus the result from alignment

67
4. Results

software for both design was almost the same. Both in deflection, reaction on bear-
ings and in whirling, see deflection graphs for the Tanker in figures 4.11a and 4.11b.
The only noticeable difference in whirling is seeing a shift in critical speed of few
rpm´s for a different mode, see whirling graphs in figure 4.12a and 4.12b.

Because of this "uninteresting" result, it was decided to ignore the propeller blade
failure load for the tanker vessel, just to see how much the alignment could be
pushed. See that investigation in section 4.2.3.5.

(a) Deflection original (b) Deflection decreased diameter

Figure 4.11: Shaft deflection for all operating condition, side by side comparison
for the Bunker tanker.

(a) Whirling modes original (b) Whirling modes decreased diameter

Figure 4.12: Whirling modes in forwarding condition, side by side comparison for
the Bunker tanker vessel.

All graphs in larger scale and bearing reactions can bee seen in appendix C.3.

4.2.3.3 Dredger
Since the propeller shaft was not much changed (from 315 to 300mm) the result
from alignment was quite similar to the initial one (similar as for the Tanker vessel).
The slope and pressure of the stern tube bearing did not change a lot and is within
the regulations, see the deflection of the shaft line in figures, 4.13a and 4.13b. The
Dredger is a difficult shaft since the first whirling mode critical speed is close to the
operating speed. In this case, the MCR is 197rpm and the initial shaft design first
mode critical speed is 108rpm, but for the optimised shaft the critical speed has
shifted down to 103rpm, see whirling result in graphs in figures 4.14a and 4.14b .

68
 4. Results

(a) Deflection original (b) Deflection decreased diameter

Figure 4.13: Shaft deflection for all operating condition, side by side comparison
for the Dredger vessel.

(a) Whirling modes original (b) Whirling modes decreased diameter

Figure 4.14: Whirling modes in forwarding condition, side by side comparison for
the Dredger vessel.

All graphs in larger scale and bearing reactions can bee seen in appendix C.3.

4.2.3.4 Result for the case study

Each case study was simulated in Nauticus Alignment software to check if the new
design (new diameters) would be approved in terms of lateral vibrations (has been
talked about in the previous sections). Though diameters were not altered much,
the critical RPM of the shaft line can be shifted to a more severe region.

There were two benchmarks that the writers looked at to assess if whirling criteria
are approved. That is, based on class requirements, and based on the fact that the
initial designs of the BERG orders (design before it was changed) are approved by
the class, some having some exception approvement from the class.

Note. definition of success was mainly in respect of the shaft whirling. Other
measures were roughly taken into consideration: Stern tube slope not more than
0.3mm/m, bearing pressure not less than 1bar and with an eye on other rules and
regulations (mentioned in 2.4.3). Meaning some bearings could be slightly miss ad-
justed (which is just a result of lack of knowledge from writers using alignment
software), which was not the main study and could be solved with extent knowledge

69
4. Results

in alignment software to adjust everything correctly. In addition, the shaft line was
correctly aligned before the diameters were decreased. The writers only modified the
shaft diameter, stern tube bearing effective length, and small increments in vertical
offset of bearings, in alignment software from the initial design.

Results from alignment whirling investigation can be seen in table 4.3.

Table 4.3: Critical speed RPM of the shaft line, before diameters, were decreased
and after optimisation.

Inital design-Freq mode[rpm] Optimised design-Freq mode[rpm]

Vessel 1st critical mode 2nd - 1st critical mode 2nd -
Bunker 259 556 258 538
Dredger 208 352 203 339
RoPax 129 189 120 178

For comparison, Bunker tanker vessel has its MCR at 142 rpm, Dredger at 197 rpm,
and RoPax at 151 rpm. Beginning with the Bunker, it can be seen that the critical
speed does not change much after applying some changes to the shaft diameters.
Therefore the lateral vibrations on the bunker shift is not an issue after diameters
have been optimised. The Dredger is a difficult case. Both knowing some informa-
tion from BERG that the initial design of that shaft line is not within an approved
range(± 20% of MCR [11]), but the rotordynamics engineers at BERG did get an
exception from the class, with some conditions of serious precision when building in
the shipyard. Now resulting in even more critical natural frequency with a margin
of 6 rpm from the MCR(previous 11 rpm), it is difficult to say if that will be ap-
proved. The RoPax can also not fulfil the critical speed demand. The initial design
is within the 20%. The same circumstances is for the RoPax as for the Dredger, that
is BERG got a special approvement from class. Having that in mind and comparing
the critical speed after optimisation, it can be seen that the critical speed has shifted
away from the MCR resulting in a more reliable shaft against whirling(initial design,
mode one is 14.6% from MCR, in the optimised diameter design mode 2 178rpm is
17.8% from MCR).

4.2.3.5 Alignment Tanker without ICE-class

Having in mind that the optimised diameter is dependant on the four different
approaches using the DNV-CG-0038 documents and the additional propeller blade
failure (meant for ICE class). To see how much the whirling could be pushed, it
was decided to ignore the ICE-class criterion on the Bunker tanker propeller shaft,
and see how small the propeller shaft can be until the alignment rules will not
approve. With manual trial and errors, the smallest diameter that could be reached
was 300mm. See alignment result graphs in figures 4.15a and 4.15b. The whirling
happens aft of the aft-most stern tube bearing so additional bearings could not be
added. The limiting factor was reaching the maximum allowed slope in stern tube

70
 4. Results

bearing, and the foremost stern bearing had already gone below the pressure limit.
All output data from the alignment can be seen appendix C.3.

(a) Deflection Tanker without ICE-class (b) Whirling modes Tanker without ice

Figure 4.15: Reaction for Tanker shaft line when The propeller failure ICE criteria
is not considered

4.2.4 Torsional vibration calculation

Calculation results are compared in the form of plots, usually having engine or
propeller speed on the horizontal axis and different types of physical response on
the vertical axis. Most of the calculations are done in frequency domain, the program
simulates excitation’s and responses in the system for a range of revolution speed.
With this natural frequencies and harmonics are established.

4.2.4.1 Ro-Pax

The Ro-Pax reference shaftline represents the most complicated system of the three
systems analysed in this project work, see simplified arrangement schema. What
makes the system more complicated is first and foremost the fact that it has a
4 stroke diesel engine delivering the propulsion power. Compered to the electric
motors used for the Tanker and Dredger shaft lines the diesel engine requires much
more inputs and is comprised of many different modeling components. The electric
motor is in essence just one inertia node with a definition of power and MCR speed,
on top of that one stiffness element and one inertia node represent the stiffness and
inertia of the output shaft.
Another important thing is that the shaftline has a reduction gear with PTO, power-
take-off, where an shaft generator is connected. There are clutches for both the
generator output and propulsion shaft output, resulting in many possible operation
profiles for the shafting system.

71
4. Results

Figure 4.16: Simple schema representing the RoPax shaftline arrangement. Con-
nections marked with C represent clutches.

The RoPax shaft line was the first system to be modeled which took a lot of research
to get running without errors. But after all the check for natural frequencies showed
too low values to be regarded correct. The results could be compared to a bench-
mark report done by industry experts, analysed in similar type of software. It was
understood that most likely there was a mistake somewhere in the definition of an
inertia or stiffness, possibly damping. Damping how ever normally does not influ-
ence the natural frequency of the system much. “In case of natural vibrations there
is no excitation, (ME = 0). Furthermore the influence of damping is considered to
be negligible” [28]. The natural frequencies are calculated as free vibrations, the
effect of the damping can be seen on figure 2.9. Damping dissipates high amplitudes
but does not shift the frequency of the harmonics of any considerable amount.
Below are printouts from the Nauticus TVC tool showing natural frequencies of the
first 3 vibration modes, for two separate models of the RoPax. This check is a good
indication of correctly defined model.

NATURAL FREQUENCIES AND MODE SHAPES

Mode number 1 2 3

Natural frequency (rad/s) 0.34 19.32 31.97

Natural frequency (vibs/min) 3.27 184.48 305.29
Natural frequency (Hz) 0.05 3.07 5.09

NATURAL FREQUENCIES AND MODE SHAPES

Mode number 1 2 3

Natural frequency (rad/s) 0.22 20.91 44.62

Natural frequency (vibs/min) 2.12 199.72 426.10
Natural frequency (Hz) 0.04 3.33 7.10

Looking at the frequency values given as minvibs

, vibrations per minutes, the first nat-
vibs
ural frequency is 3.27 and 2.12 min respectively. This is a very low frequency, the
rotational speed of the shaft at MCR is 150.8 rpm for comparison. This indicated to
authors that most likely a shaft stiffness somewhere in the system was defined with
an error of 106 . Having an effect on the system just as on of the steel shafts where
replaced with a rubber one. The reason for this suspicious is because the same kind
of error had occurred while building the shaft alignment model. The effect there
was whirling deflections of about a meter instead of millimeters. The reason for such
type of an error was found being that specifications for stiffness are often given as

72
 4. Results

M N ∗m
rad
,mega newton meters per radian, while the setting in Nauticus was in Nrad ∗m

and needed therefor multiplication by 106 . The trick with computer simulations like
those is that the packages just deliver the numbers however out of reality they can
be, of course in both of those cases the shaft would fail immediately. After build-
ing the model from scratch two times, going through all inputs multiple times and
studying the function behind different elements in a quest for understanding what
part could cause such a large error in the frequency of the system.
Finally it was found as suspected that stiffness for the input shaft for the reduc-
∗m
tion gear had been defined as 108.63 Nrad resulting in value inside Nauticus as
1.086E+002 N*m/rad instead of the correct one being 1.086+008 N*m/rad.
The main cause of error was by this eliminated but in the process another mistake
was discovered, the damping of the engine crankshaft damper had been defined twice.
Both as absolute damping in inertia node one, representing the outside portion of
the engine damper, and in the damper stiffness element also as relative damping.
The damper stiffness being shaft element no. 1. All input data can be found in
appendix C.
Below frequency check shows values of mode number 1 and 3 within 0.2% and mode
number 2 within 3%.

NATURAL FREQUENCIES AND MODE SHAPES

Mode number 1 2 3

Natural frequency (rad/s) 16.80 23.47 50.73

Natural frequency (vibs/min) 160.40 224.15 484.46
Natural frequency (Hz) 2.67 3.74 8.07

While the RoPax model still had errors the study proceeded with the two electric
driven shaft lines, the Tanker and Dredger. It was believed that part of the problem
could be related to the definition of the diesel engine. For this reason the main part
of the study was proceeded on those two models.

The results presented here for the RoPax shaft focus mainly on representing vibra-
tion behaviour related to the diesel engine in comparison with the electric motor.
The comparison is between the engine running normal and misfiring. The TVC
calculation for a diesel engine in a misfiring condition is an important and required
step, here cylinder no. 12 is compressing but there is no fuel and therefore no power
stroke. The operation mode is both clutches engaged, such that both propulsion
shaft and shaft generator is spinning. The propeller is at full pitch and the engine
on full power, how ever the generator is spinning passive. The propeller is consuming
all the engine power.

73
4. Results

(a) Normal (b) Misfire

Figure 4.17: Heat dissipation from the engine damper, indicating the work done
by the damper.

The engine damper is of a steel spring design, which has two circular steel sections
with steel springs in between, voids between the springs are filled with oil. This type
of vibration damper can be specifically tuned for each application [19]. The exact
type of the spring is not known, interested reader can find information about this
type of dampers from company called Geislinger. According to figure 4.17 there is
just slightly elevated heat build up in the damper, well within limits. Figure 4.18
shows the alternating torque experienced by the damper.

(a) Normal (b) Misfire

Figure 4.18: Engine damper alternating torque, the thick blue line shows synthesis
of vibratory torque amplitudes measured at the damper element. The thinner red and
cyan represent the most prominent excitation modes.

Between the engine and the gearbox it is normal to have a flexible coupling, this
is in large part an elastic vibration damper, but can also accommodate for some
misalignment.

74
 4. Results

(a) Normal (b) Misfire

Figure 4.19: Main flexible coupling heat dissipation.

By investigating figure 4.19b it becomes clear what effects it has to run the engine
under misfiring condition. Uneven forces will certainly affect the engine itself but
what can happen before is failure of the flexible coupling. The heat builds up in the
rubber elements, causing the coupling to loose stiffness, elevating the risk of failure.

(a) Normal (b) Misfire

Figure 4.20: Main flexible coupling alternating torque.

When studying the responses for the coupling in figures 4.19 and 4.20 it is important
to realize the operating range of the propulsion engine, the MCR speed is 514 rpm,
the clutch in speed is between 60% and 70% of the maximum speed. This means the
engine and shaft will never operate on lower speed than about 300 rpm [8]. Looking
at the high alternating torque to the left, the shaft will never dwell at this speed,
it is however worth noting the higher torque amplitudes at operating speed for the
misfiring condition.
Worth noting also is the number of excitation modes contributing to the total syn-
thesis, the thick blue line. The order plotted are just the most prominent. This
shows the effect of the reciprocating engine, when studying the Tanker and Dredger
shaft lines, propelled by electric motors. Usually the only excitation’s seen is P1
and P2, the first and second propeller orders.

75
4. Results

(a) Normal (b) Misfire

Figure 4.21: Main gear alternating torque.

Alternating torque over gears can cause gear hammering, in figure 4.21 the alter-
nating torque amplitudes for the main reduction gear are presented. As before the
rotational speed below 300 rpm are not of real interest. A clear increase can be seen
in alternating torque for the misfiring condition compared to the normal condition,
however the torque amplitudes are still small and well below limits of 81.9 kN*m.
Final calculation for the RoPax shaft line is torsional stress amplitude synthesis. In
figure 4.22 the torsional stress amplitudes are compared to alternating stress limits
in regards to fatigue failure of the propeller shaft. It can be seen that alternating
torsional stresses are of no concern for either the normal nor the misfiring engine.

(a) Normal (b) Misfire

Figure 4.22: Propeller shaft torsional stress amplitudes for normal running engine
and misfiring engine.

In figures 4.22a and 4.22b the cyan colored line at the top represents a transient
fatigue stress limit. The stress is not allowed to get higher than the transient limit.
The red colored line in the middle is the continuous running fatigue stress limit,
stresses during normal operation are not allowed to be higher than the continuous
limit. The zone between the continuous and transient represent stress level that
should be avoided, the shaft should not dwell on this speed. Speeds having harmonic

76
 4. Results

spikes causing shaft stresses to rise above the continuous limit should be regarded
as barred speed ranges and ship operators must be aware of those speeds. See
discussion about IACS M68 in chapter 2.4.1. The simulation results indicate that
flexible couplings and gearbox should be checked carefully for alternating torque but
the propulsion shafting is not affected much.

4.2.4.2 Bunker tanker

The reference shaftline for the Tanker type of a vessel is much shorter, has fewer
components and results in a considerably less complicated model than the RoPax,
especially for the TVC analysis. It has similar setup as for 2 stroke shaft but using
two electric motors through a reduction gearbox instead of a diesel engine. Hence
skipping the trouble some vibration behaviour of the 2 stroke engine. The operating
condition considered in this investigation is 2 Motor Mode - Normal Operation. The
torsional vibration response is calculated in frequency domain from zero to MCR
speed.

(a) Original diameter (b) Decreased 10 mm

(c) Decreased 20 mm (d) Decreased 30 mm

Figure 4.23: Propeller shaft nominal torque and upper and lower torque synthesis,
where the upper and lower synthesis diverge from the nominal indicates fluctuat-
ing shaft torque. Comparison of original diameter to -10mm, -20mm and -30mm
decrease.

Figure 4.23 demonstrates that alternating torque is not a problem in the Tanker’s

77
4. Results

propeller shaft, further it demonstrates that with decreased diameter the response
frequency moves to lower speeds and the stress amplitudes become smaller.

78
 4. Results

Figure 4.24 compares the propeller shaft torsional stress amplitudes, for the different
diameters. The curve at top is the transient curve, next below is the continuous
line according to IACS M68.5, see equations 2.37 and 2.38. The fat blue line is
the torsional stress amplitude synthesis, P1 and P2 represent the first and second
propeller excitation.

(a) Original diameter (b) Decreased 10 mm

(c) Decreased 20 mm (d) Decreased 30 mm

Figure 4.24: Propeller shaft torsional stress amplitudes, comparison of original

diameter to -10mm, -20mm and -30mm decrease. IACS M68 torsional stress limits
plotted also, as can be seen the stress levels are well below the limits.

“Deflection is not affected by strength, but rather by stiffness as represented by

the modulus of elasticity, which is essentially constant for all steels. For that rea-
son rigidity cannot be contolled by material decisions, but only by geometric deci-
sions” [5, p. 348].
From the behaviour of the shaft with decreased diameter it is suggested that the
stiffness of the shaft influences mostly the torsional frequencies but not the actual
strength of the shaft. Figure 4.24 shows no increase in stresses, but the harmonics
occur at lower speeds. It was also noted that the natural frequencie of the whole
system got lower.

79
4. Results

Figure 4.25 compares the vibratory torque amplitudes in the flexible coupling for
motor 1. The horizontal line at the top indicates the maximum elastic torque, 4.1
kNm. The limit is a manufacturers specification, from a torsional vibratory report
authors had access to at BERG Propulsion.

(a) Original diameter (b) Decreased 10 mm

(c) Decreased 20 mm (d) Decreased 30 mm

Figure 4.25: Vibratory torque amplitudes in motor coupling, comparison of original

diameter to -10mm, -20mm and -30mm decrease. Limits for vibratory torque is set
at 4.1 kNm.

Figure 4.26 compares the heat dissipated, representing power loss in the flexible
coupling for motor 1. The horizontal line at the top indicates the maximum heat
dissipation of 0.480 kW. The limit is a manufacturers specification, from a torsional
vibratory report authors had access to at BERG Propulsion. The heat builds in the
coupling when the rubber elements flex under vibratory torque. The flexible coupling
thus dampens the torsional vibrations but if the load becomes to much and the
rubber element start to heat up, the rubber loses stiffness, gets softer and eventually
can not support the torque and could fail. This is why for combustion engines it is
important to investigate the load on the coupling under misfire condition. Misfiring
is obviously not relevant for electric motors. This is the real cause of worry, not a
power loss of 0.48 kW when transmitting 1250 kW.

80
 4. Results

(a) Original diameter (b) Decreased 10 mm

(c) Decreased 20 mm (d) Decreased 30 mm

Figure 4.26: Heat dissipated in motor coupling, comparison of original diameter

to -10mm, -20mm and -30mm decrease. Limits for heat dissipation or power loss is
set at 0.480 kW.

The results indicate that torsional vibrations are not so much a problem of strenght,
rather a design challenge on a system level. The main thing is to know where har-
monics occur and then the system can be modified in a way such that the shafting
is not operating at bad speeds. Example from DNV class guidelines suggest for a 2
stroke system with vibration problems to decrease the diameter of the shaft to move
the vibration excitation to lower speeds away from the operating speed [10, p. 65].
This result is also supporting why the natural frequency of the RoPax shaftline was
so low, because the stiffness value defined for a mistake represented a input shaft
having diameter of about 13mm instead of 310 mm. The above study has demon-
strated the effects of decreased diameter of the shafting, which mostly affects the
stiffness of the propulsion system. The shaft does not have high inertia compared to
the propeller, in this case the propeller shaft has inertia of about 40 kgm2 compared
to the propeller’s 4000 kgm2 with the added entrained water. In appendix C.4 a
comparison of how different propeller inertia affects the flexible coupling loads can
be seen.

81
4. Results

4.2.4.3 Dredger
The Dredger is a twin screw vessel with fixed pitch shrouded propellers. For the
analysis only on of the shafting is considered, the starboard and portside are iden-
tical. Two running conditions are considered, free running and dredging.
This is due to the fact that the vessel has fixed pitch propellers and the propeller
curves, hence the power curve are different between the operation profiles.

4.2.4.3.1 Free running empty condition The free running condition would
be when the dredger is sailing with empty dredging material cargo holds.

(a) Original diameter (b) Decreased

Figure 4.27: Vibratory torque amplitudes in motor coupling, limits for vibratory
torque is set at 10 kNm.

(a) Original diameter (b) Decreased

Figure 4.28: Heat dissipated in motor coupling, limits for heat dissipation or power
loss is set at 0.780 kW.

82
 4. Results

(a) Original diameter (b) Decreased

Figure 4.29: Nominal torque in main gear with upper and lower torque synthesis.
The alternating torque demonstrated by the upper and lower synthesis are showed to
be small for the whole speed range.

(a) Original diameter (b) Decreased

Figure 4.30: Torque amplitudes in the main gear meshing. Excessive alternating
torque in the messing indicates gear hammering, this would result in high forces on
the gears and shorter life. The results show torque well within limits.

83
4. Results

(a) Original diameter (b) Decreased

Figure 4.31: Intermediate shaft torsional stress amplitudes, IACS M68 torsional
stress limits plotted also, as can be seen the stress levels are well below the limits.

(a) Original diameter (b) Decreased

Figure 4.32: Propeller shaft nominal torsional stress with upper and lower alter-
nating stress synthesis.

84
 4. Results

(a) Original diameter (b) Decreased

Figure 4.33: Propeller shaft torsional stress amplitudes, IACS M68 torsional stress
limits plotted also, as can be seen the stress levels are well below the limits.

4.2.4.3.2 Dredging condition Dredging condition is when the dredger has

sand and mud in the cargo hold and has the dredging head and pumping pipe
down. Dragging the dredging digger head along the bottom. More of a bollard pull
condition.

(a) Original diameter (b) Decreased

Figure 4.34: Vibratory torque amplitudes in motor coupling, limits for vibratory
torque is set at 10 kNm.

85
4. Results

(a) Original diameter (b) Decreased

Figure 4.35: Heat dissipated in motor coupling, limits for heat dissipation or power
loss is set at 0.780 kW.

(a) Original diameter (b) Decreased

Figure 4.36: Nominal torque in main gear with upper and lower torque synthesis.
The alternating torque demonstrated by the upper and lower synthesis are showed to
be small for the whole speed range.

86
 4. Results

(a) Original diameter (b) Decreased

Figure 4.37: Torque amplitudes in the main gear meshing. Excessive alternating
torque in the messing indicates gear hammering, this would result in high forces on
the gears and shorter life. The results show torque well within limits.

(a) Original diameter (b) Decreased

Figure 4.38: Intermediate shaft torsional stress amplitudes, IACS M68 torsional
stress limits plotted also, as can be seen the stress levels are well below the limits.

87
4. Results

(a) Original diameter (b) Decreased

Figure 4.39: Propeller shaft nominal torsional stresses plotted with upper and lower
alternating stress synthesis.

This shows the scale of the alternating torsional stresses in comparison to the nom-
inal torque. The alternating stresses appear at low speed, but are mild. This could
be something to have in mind if the vessel had diesel engine coupled to the propeller
because this speed range could be the idling speed of the engine. Also because this
vessel has fixed pitch propeller it would be worse to dwell on this speed. With
CP propeller one usually would rev quickly through any bad speed ranges with low
pitch. Since the dredger uses electric motor there is no idling speed, but it could be
advisable to limit sailing time on this speed. Overall though the stress amplitudes
are very low and well under recommended limits therefor this would not be consid-
ered a reason for any actions. There are almost no alternating stresses at operating
speeds.

(a) Original diameter (b) Decreased

Figure 4.40: Propeller shaft torsional stress amplitudes, IACS M68 torsional stress
limits plotted also, as can be seen the stress levels are well below the limits.

88
 5
Future work

When work commenced for this project, the literature study suggested the torsional
vibration behaviour of the shafting would be the critical case to study. During
briefing meetings and discussions with propulsion experts at Berg Propulsion this
was also suggested to be the limiting factor, especially in the case of the IACS M68
unified requirements. It was however pointed out by Berg Propulsion rotor dynam-
ics experts that in majority of cases the shaft line solutions included medium or
high speed engines connected to a reduction gear. For a geared plants the torsional
vibrations are realized in the flexible couplings and can affect the gear with gear
hammering. Since flexible couplings can be ordered with specific damping stiffness
behavior, the systems can most often be tuned in accordance to torsional vibration
calculation results.

5.1 More thorough alignment investigation

During the process of this work the shaft alignment analysis was found to be a more
limiting factor for the dimensioning of the shafts. For torsional vibrations decreasing
the diameter of the shafting caused no problem, except when considering ice loads.
However considering the bearing loads and whirling in alignment analysis the first
whirling frequency was just above the operating speed of the shaft for two cases.
Thus making even a small decrease in diameter problematic. Since this first mode
of vibration was related to the overhang bending load of the propeller weight it
is not an easy fix to just add an extra intermediate bearing to tune the vibration
mode, as could have been the case for whirling in intermediate shafts. As an extra
challenge the bearing slope in the aft stern tube bearing reaches classification limits
and lubrication becomes a challenge. All above suggest further detailed study for
the shaft alignment as a limiting factor in the dimensioning of a propeller shaft.

5.2 Propeller failure load criteria

The project was to go around the IACS M68.4 formula and push the classification
rules to the limit, which in theory was done with a typical shaft. But applying the
propeller failure load criterion (section 2.4.2.1 and method in 3.3.1) on top of the
improved method results that the comprehensive approach that had been done is in
vain, it is useless. Because the propeller failure load criteria will always result in a
larger diameter than going through all the hassle of defining the diameter based on

89
5. Future work

DNV CG 0038 document (the improved method [10]). The chain is never stronger
than the weakest link.
So there has to be more investigation done on this propeller failure load. The
classification method that was followed in this study is simply more conservative
than the DNV CG 0038 document.

5.3 More precise investigation of the case study

For a higher certainty an investigation on more parts and sections of the case study
(the three shafts). In this thesis study, only two or three locations were examined
and taken through the analysis steps. Though the most critical ones were chosen
to rule out allowed diameter in the shaft sections around it, it can be that the next
component next to the investigated section has a more critical design. Or the op-
posite, that a further investigation would result in a smaller shaft at some sections
and lead to a more complicated optimal design.
If the whole shaft is designed in respect of this improved method this minor uncer-
tainty will not apply.
With regards to the TVC study in Nauticus Machinery further study of the misfiring
condition of a 4 stroke diesel engine would be beneficial to support further claims
of torsional vibrations not to be a prominent problem in propulsion shafting using
geared medium and high speed propulsion engines.
Shaft fatigue study using a multi-physics FEM software setting up load cycle history
according to Section 5 in DNV guidelines, document DNVGL-CG-0038 would be a
good starting point.

90
 6
Conclusion

This report presents a generalized methodology for determining the diameter of low
speed marine propulsion shafts, which aims to reduce weight in relation to current
designs. With limited information, a shaft line needs to be dimensioned which is fit
for purpose, fulfils class requirements as well as showing good vibrational behaviour.
It is believed to be possible to step away from current methods and provide slimmer
and more power-dense shaft lines. The writers evaluate an improved method to
define the minimum shaft diameter.

The path and the goal to proceed from the initial method, using the IACS formula
to an improved method is what defined the thesis work. Early on, the writers moved
into two directions from the research of the initial background theory. One where
the writers tried to improve and boldly trying to find a way around it. That lead
to the determination of the "mimic formula". Which is an emulated mimic formula
of the IACS formula, made for further understanding of the IACS M68 formula.
The other path from the initial research was to come up with a new definition of a
method to define the minimum shaft diameter. The "improved method" is mainly
defined by three definitions, The DNV improved method (DNV CG-0038), and us-
ing computer software to define the shaft alignment (lateral vibrations) and TVC
(torsional vibrations). Old finished Berg Propulsion orders were acquired to test
the effect of the newly defined approach (case study investigation), to compare and
see how much could have been saved in shaft material. In addition, an extra TVC
investigation was performed on the case study shaft lines to analyse the vibrational
behaviour of the shaft lines with incremental diametric changes. Finally, when the
new shaft diameter approach was achieved, showing promising results, the approach
was defined in a step by step guide for a user to utilise. See in figure 6.1, a flow-cart
of the work from the initial formula to the final method.

The results of the TVC simulation analysis suggest that for the simulated shaft lines
the dimensions are on the safer side and quite reservative.

Though the mimic formula of the IACS M68.4 formula looks right and are good
results in the study, it does not in fact change anything in the whole investigation
of the improved method. It just gave the writers an inside into the current formula
that BERG uses today(or at the time this was written). This separate study was
done in request from the company. There are great possibilities that the writers
approach of the formula is not like the approach that IACS experts did when they
formulated theirs, but it is a trial of an approach to that rule.

91
6. Conclusion

The same goes with the investigation on the stress concentration factor of the IACS
formula. It was a separate investigation done on the IACS M68.4 to try to under-
stand it better.

Figure 6.1: Flow chart of the thesis process

The improved method approach, which is mainly defined by the DNV CG-0038 doc-
ument follows four different routes (as has been explained in the report, depending
on est KA and bending moment). Though the three different vessels can not define
a theory confirmation of a conclusion, it does though indicate that all of the four
approaches do result in a similar outcome. It seems regarding the thesis result that
the most conservative approach is to define the bending as 40% of the shaft torque
and KA based on estimations (can be applied in the future use of the method).
If Berg Propulsion decides to integrate this method into their regular procedure
and every time save around 8% in material weight, it is questionable how much is
actually saved in material cost, due to the fact that whole axles are bought and cut
down. Meaning, that one shaft piece that has different diameters and slopes is al-
ways dependant on the largest diameter of that piece as the initial bought raw steel
axle. Answering the research question, how much room is there for weight/design
optimisation. The simple answer is a weight optimization around 8%. But as said,
it can not be equal in material cost saving.

The current method, using the IACS m68.4 formula is quick and simple, especially
compared with the writers improved method. If Berg Propulsion decided to in-
tegrate it into their routine default procedure of designing a shaft line, the most
probable solution would be to use (as stated here previously in conclusion) use esti-
mated KA and bending moment (Mb). Since the method requires input values for
the shaft component, the procedure could be as it is today, defining the shaft and
other values using the ICAS formula, to support and acquire the input values to
perform the "new improved method". Followed by TVC and alignment simulations,
as it is already done today.

For the case study optimised dimensions, and the approval in terms of alignment and
whirling. Since the classification expert has not looked over the data (at the time)

92
 6. Conclusion

it is not know for certain if all three shaft lines will be approved. To begin with,
the Tanker vessel will most surely be approved, since it has gotten approval before,
and the improved shaft has little to no whirling changes from the initial result. The
RoPax result looks better than the initial one, for critical whirling. But for the
initial (approved) one the critical speed is 14.6% from the MCR in mode shape 1,
and for the improved shaft, the critical speed is 17.9% from MCR in mode 2, see the
whirling graphs in appendix C.3. The first mode shape appears behind the propeller
and the second mode shape between the bearings on the shaft, which is a problem.
The regulations say the critical speed must be more than 20% from the MCR, but
if it is within that and the whirling happens on the area aft of the stern tube (at
propeller), the classification society does consider how much the propeller dampens
the vibrations, and thus give exceptions for that. This statement is solely based
on experience from BERG expert. Based on this statement the improved RoPax
critical speed 17.9% from MCR can result in a problem because that is within the
20% and is not at the propeller, thus it must be considered that it will not get
approval from the class. But with more experience using the alignment software,
better adjustments and possibly moving and adding some bearing could solve this.
The dredger is a difficult one to predict if the class will approve it, hence it must be
consider and have in mind that it will not be approved.

In general, the improved method will most surely be approved by class since all
of the calculations and computation manoeuvres are all based on classification rules.
Nothing is really defying or challenging the regulations, everything is in fact provided
within different classification documents and softwares.

93
6. Conclusion

94
 References

[1] ABS (2019). ABS Guidence notes on - Propulsion shafting alignment (Septem-
ber 2019). American Bureau of Shipping, New York.
[2] Adnan, M. (2019). Retrieved 2021-03-15 from <https://www.researchgate.
net/publication/333747137_A_Design_Approach_Against_Predictable_
Fatigue_Failure_of_Airplane_Hydraulic_Pump_Shaft_in_Combined_
Rotating_Bending_and_Torsional_Loading_Problem_Statement>
[3] The International Association of Classification Societies – IACS (2020). Classi-
fication societies – what, why and how?. Westminster, London, United Kingdom
[4] Batrak, Y. (2017). Torsional vibration calculation issues with propulsion sys-
tems. ShaftDesigner, c/o Machine Support B.V.
[5] Budynas, R. G., Nisbett, J. K. (2008). Shigley’s Mechanical Engineering Design
(8. ed.). McGraw-Hill.
[6] Caterpillar Inc. (2013, July 1). CATERPILLAR TO ACQUIRE BERG
PROPULSION, SIGNIFICANTLY EXPANDING CUSTOMER OPTIONS
FOR ADVANCED MARINE POWER SOLUTIONS Retrieved 2021-05-06
from
https://www.caterpillar.com/en/news/corporate-press-releases/h/caterpillar-
to-acquire-berg-propulsion-significantly-expanding-customer-options-for-
advanced-marine-power-solutions.html
[7] Caterpillar Inc. (2013, September 1). CATERPILLAR UNVEILS CAT®
MARINE PROPULSION SYSTEMS, SUCCESSFULLY COMPLETES
ACQUISITION OF JOHAN WALTER BERG AB Retrieved 2021-05-
06 from <https://www.cat.com/en_ZA/news/engine-press-releases/
caterpillar-unveilscatsup174supmarinepropulsionsystemssuccessful.
html>
[8] Caterpillar (2008). MaK - VM43C Propulsion (Issue September 2008). Cater-
pillar Motoren GmbH & Co. KG, Kiel.
[9] Department of Shipping and Marine Technology (2014) Reliability Analysis of
Marine Structuress SJO750 Division of Marine Design, CHALMERS UNIVER-
SITY OF TECHNOLOGY, Göteborg, Sweden 2014
[10] DNV-GL (2019). Class Guidelines DNVGL-CG-0038, Calculation of shafts in
marine applications (Edition July 2019). DNV GL AS.
[11] DNV-GL (2018). Rules for classification: DNVGL-RU-SHIP Pt.4,Systems and
components (Edition January 2018). DNV GL AS.
[12] DNV-GL (2020). Rules for classification: DNVGL-RU-SHIP Pt.6,Additional
class notations Chapter 6 Cold climate (Edition July 2020). DNV GL AS.

95
References

[13] DNV-GL (2020). User manual - Nauticus Torsional Vibration (Version 14.3.0
- November 2020). DNV GL - Digital Solutions.
[14] DNV-GL (2020). Rules for classification: DNVGL-RU-SHIP-Pt4Ch2 - Rotating
machinery, general (Edition July 2020). DNV GL AS.
[15] DNV-GL Digital Solutions (2018). Nauticus Machinery - For marine propulsion
design and analysis. DNV GL - Digital Solutions.
[16] Dokkum, K. V, et.al (2016). Ship Knowledge - Ship dsign, construction and
operation (9. ed). Dokmar Maritime Publishers B.V.
[17] Dowling, N. E. (2013)Mechanical Behavior of Materials - Engineering Methods
for Deformation, Fracture, and Fatigue (4th ed). Pearson Education. Courier
Westford, USA.
[18] EKBERG, A. (2018), Multiaxial Fatigue Edition 5.3, Department of Ap-
plied Mechanics, CHALMERS UNIVERSITY OF TECHNOLOGY Gothen-
burg, Sweden, 2018
[19] Geislinger GmbH (2021). Geislinger damper - Tuned, torsional vibra-
tion damper. Retrieved 2021-05-28 from <https://www.geislinger.com/en/
products/product/damper>
[20] Gustaver, M. (2020) A Chalmers University of Technology Master’s thesis tem-
plate for LATEX. Unpublished.
[21] IACS (2015). UR M68 Dimensions of propulsion shafts and their permissible
torsional vibration stresses - Rev.2 Apr 2015. IACS - the International Associ-
ation of Classification Societies and International Association of Classification
Societies Limited.
[22] Inman, D. J. (2001). Engineering vibration (2. ed.). Prentice Hall.
[23] Korean Register (2020), Rules for the Classification of Steel Ships Part 5 Ma-
chinery Installations, 2020
[24] MAN Energy Solutions (2021). A powerful portfolio: MAN B&W two-stroke
propulsion engines. Retrieved 2021-03-09 from <https://www.man-es.com/
marine/products/two-stroke-engines>
[25] Murawski, L., & Charchalis, A. (2014). Simplified method of torsional vibration
calculation of marine power transmission system. Marine structures, 39:335-349.
ISSN 0951-8339, https://doi.org/10.1016/j.marstruc.2014.10.004.
[26] Polski Rejestr Statków (2021), RULES FOR THE CLASSIFICATION AND
CONSTRUCTION OF SEA-GOING SHIPS PART VI MACHINERY IN-
STALLATIONS AND REFRIGERATING PLANTS, Gdansk 2021
[27] Russian Maritime Register of Shipping (2020), FOR THE CLASSIFICATION
AND CONSTRUCTION OF SEA-GOING SHIPS PART VII MACHINERY
INSTALLATIONS, Saint-Petersburg 2020
[28] Senjanović, I., Hadžić, N., Murawski, L., Vladimir, N., Alujević, N., & Cho,
D.S. (2019). Analytical procedures for torsional vibration analysis of ship power
transmission system. Engineering Structures, 178:227-244. ISSN 0141-0296,
https://doi.org/10.1016/j.engstruct.2018.10.035.
[29] Sjöström, P.H. (2020, June 12). Berg Propulsion är tillbaka: Caterpillar Propul-
sion AB blir åter svenskägt och återfår namnet Berg Propulsion AB. Sjöfartstid-
ningen. Retrieved 2021-05-06 from <https://www.sjofartstidningen.se/
caterpillar-propulsion-far-ny-agare/>

96
 References

[30] Sold by Cat, Berg Propulsion Returns to its Original Name.

(2020, June 9). THE MARITIME EXECUTIVE. Retrieved 2021-
05-06 from <https://www.maritime-executive.com/article/
sold-by-cat-berg-propulsion-returns-to-its-original-name>
[31] Vulić, N., Dobrota, Ð., & Komar, I. (2016). Damping and excitation in the
torsional vibrations calculation of ship propulsion systems. In Conference Pro-
ceedings, Contemporary Issues in Economy & Technology (CIET 2016), June
16th-18th 2016 (pp. S-165 - S-174). University of Split, University Department
of Professional Studies. Split, Croatia.
[32] Wärtsilä (2021). Wärtsilä Encyclopedia of marine technology. Retrived on 2021-
05-08 from <https://www.wartsila.com/encyclopedia/term/sterntube>
[33] Wärtsilä (2021). Wärtsilä Encyclopedia of marine technology. Retrived
on 2021-05-08 from <https://www.wartsila.com/encyclopedia/term/
propeller-shaft-tail-shaft>
[34] Wärtsilä (2021). Wärtsilä Encyclopedia of marine technology. Retrieved
on 2021-05-08 from <https://www.wartsila.com/encyclopedia/term/
intermediate-shafting>
[35] Winterthur Gas & Diesel Ltd (2021). WinGD Company Overview Re-
trieved 2021-03-09 from <https://www.wingd.com/en/about-wingd/
wingd-company-overview/>
[36] Winterthur Gas & Diesel Ltd (2021). World Record 12X92DF. Re-
trieved 2021-03-09 from <https://www.wingd.com/en/engines/
world-record-12x92df-en/>

97
References

98
 A
Appendix-Screenshots

A.1 DNVGL-CG-0038 flowchart

See next full page.

I
 Extra equations If the shaft is also under reverse
Flow chart for Low cycle fatigue torque
Is it directly coupled or geard?

Coupled Geard
Can be used ONLY for gear plant Is it directly coupled or geard?
For direct coupled, use highest τv for Direct coupled Geard
the rpm range(see left graph at
bottom). To do that

Or τmax =τ0·KAice if that is greater

KA is the factor for
different scenarios,
ice, missfire and so
on. Also check
„extra equation in
Ordinary steel or is it Component Influence Factor left corner“ To find out what αt to use look at tables on the
stainless- in salt water? (follow both paths, right and left) right.

Other than that . The safety can be calculated

with this formula, with selected parameters
Ordinary Stainless
To find out what αt to use
look at tables on the right.

Now for the given parameters the safety can be

checked with given formula. Note! If yield is gotten
from 0.2% proofstress, σy = 0.7 σb

Allowed stress factors can be seen in a

table on flow chart „Flow chart direct
coupled plant, high cycle fatigue” Typical torsion stress in shaft in direct Typical geard shaft torsional stress
couple plant amplitude with the relation to Ka
 Flow chart for geard plant, high cycle fatigue
Safety is given with this formula. Follow all three
paths for each variable(use τ max instructions from(
Flow chart for Low cycle fatigue)
Nominal vibratory torsional Note. Mb is the rotating bending
τv σb moment and is the moment
stress for continuous
created by alignment. The value
operation. This formula, can be zero (resulting in σb=0) I
unless a higher value applies believe that it can be acquired
below the full speed range from alignment software.
(~90% to 100% engine
speed).

If the component under inspection is a shrink-

fit (no notch) go to the right, else continue
down.

r = notch radius, use r

described for different KHτ and KHσ formulas have to
component from tables be gotten from the table on the
on the left or shaft left (not the blue one).
radius whichever is
smaller. If r >100, 100
shall be used.

αt and αb can be
found for different
components in table
s on the left. If the
component is just a
plain shaft, mt and
mb = 1.

At this stage, all variables ar known

τ is typically used as τ0..
to calculate the safety for the given
Probably because that is the
pararmeters and set-up.
highest torsion experieced
on nominal rpm for HCF
yield strength or 0.2% proof Allowed stress factors can be seen in a
stress limited to 0.7 σB. table on flow chart „Flow chart direct
coupled plant, high cycle fatigue”
 Flow chart direct coupled plant, high cycle fatigue
Similar procedure to geared. But the major difference is that
the ´τv´ Nominal vibratory torsional stress amplitude can not To analyze how the τvHC Permissible
be calculated, thus the safety calculation goes the other high cycle torsional vibration stress
amplitude, is through the rpm range,
direction. For a specific parameter setup and a certain safety we have to look at τv as a function.
factor the allowed vibratory stress can be calculated, called
τvHC Permissible high cycle torsional vibration stress
amplitude. For the rpm range.
(I guess in software, like Nauticus machinery, the vibrational
behavior of the system can be plotted (blue graph) through
the rpm range (0 to Shaft rotational speed at maximum For the given parameters and safety the limit for
continuous power, r.p.m (n0)) continuous operation can be calculated (orange
graph) to the Torsional vibration calculations (blue
Same steps to calculate KHτ and KHσ, graph). The barred speed range section above the
as in „Flow chart for geard plant, high permissible torsional vibration.
cycle fatigue“

Use highest τv for the rpm range (n)

(Just the same as in LCF)

τv(n)

Torsional vibration stress from analysis of the shaft (blue) compared with the permissable torsional stress
(orange) in respect of the parameters and specific safety for HCF. Continous opperation can not be above the
permissable stress (orange). So the rpm width were the torsional vibration passes the permissable stress is the
e barred speed range
 A. Appendix-Screenshots

A.2 Main instruction manual script

V
 A. Appendix-Screenshots

A.3 DNVGL-CG-0038 document calculator

The user navigates to correct location (sheet) by clicking the blue buttons on each page.

VIII
Geared – Step with undercut (with some input values)

Geared - flange/fillet component (with some input values)

A. Appendix-Screenshots

A.4 Shaft lines (without components) dimensions

X
A. Appendix-Screenshots

XI
A. Appendix-Screenshots

XII
 B
Appendix step-by-step deduction

B.1 Torsional-static step-by-step deduction mak-

ing mimic formula
τ : Shear stress (from torsion)
D: External diameter
d: Internal diameter
J: Polar moment of inertia
T: Torque
P: Power
n: rpm

Begin with equation for torsional shear stress from torque.

T ·r T ·D
τ= −→ (B.1)
J 2·J
Polar moment of inertia for hollow cylinder.

J = π/32 · (D4 − d4 )

60
T = · P/n
2π
Merge:

60 · P · D · 32 1 π2 · n
τ= −→ = · (D4 − d4 )
2 · π · n · 2 · π · (D4 − d4 ) τ D · 480 · P
!
1 D4 · π 2 · n d4 · π 2 · n 480 · P d4
−→ = − −→ = 1− 4
τ D · 480 · P D · 480 · P τ · D3 · π 2 · n D
3 2
τ ·D ·π ·n 1 3 P 1 480
−→ = d4 −→ D = · d4 ·
480 · P 1 − D4 n 1 − D4 τ ∗ π 2

v
uP 1 480
u
D= 3
t · d4
· (B.2)
n 1 − D4 τ ∗ π 2

XIII
B. Appendix step-by-step deduction

B.2 Mimic formula- step-by-step deduction with

fatigue limit criteria applied
D: External diameter
d: Internal diameter
P: Power
n: rpm
σB or Sut : Ultimate tensile strength
σe or Se: Endurance limit (σedv : withDangV ancriterion
Ka : surface reduction factor (will use machined surface)
Kf : stress concentration factor
kc : Load factor
σf l : Fluctuating loading
σf lp : Pulsation loading.

We shall begin by defining the reduced fatigue limit for a typical rotating shaft.

All reduction factors are based on Shigleys [5].

Se’ = 0.5*σB
Ka = 4.51 · σB− 0.265
kc = 0.59
Kf = 1.45

Se = Se0 · ka · kc /Kf = 0.166 · σB M P a

Now applying the Dang Van criterion.

σf l = 0.166 · σB σf lP = σf l ∗ 0.8

σf l σf l
σedv = + Cdv ∗ = 0.5
2 3
σ1 − σ3 σ1 + σ2 + σ3
σedv = + Cdv
2 3
σf l σf l
σedv = + Cdv ∗ = 0.111σB
2 3
Insert the newly defined Dang Van reduced endurance limit into the statical statical
torsion mimic equation.
v
uP 1 442.13
u
D= 3
t · d4
· (B.3)
n 1 − D4 σB

XIV
 B. Appendix step-by-step deduction

B.3 KAice Estimate ice hit torque on shaft

If the propeller is in duct:
k1 = 7.7 and k2 = 14.6

If the propeller is open:

k1 = 10.9 and k2 = 20.7

D: Propeller diameter
d: Hub diameter
P: Power
n: cycle speed [rpm] (nn : rotational speed bollard condition, see table)
Ie /It : The mass moment of inertia ratio between the propulsion side and engine
side,
P0.7 /D: the pitch ratio at radius 0.7
Q: Torque

Define Hice :

nn taken from table(of not known):

If D < 1.8 ∗ Hice

Qmax = k1 ∗ (1 − d/D) ∗ (P0.7 /D)0.16 ∗ (nn ∗ D)0.17 ∗ D3 (B.4)

If D > 1.8 ∗ Hice

Qmax = k1 ∗ (1 − d/D) ∗ (P0.7 /D)0.16 ∗ (nn ∗ D)0.17 ∗ D1.9 ∗ Hice (B.5)

Qemax taken from table:

XV
B. Appendix step-by-step deduction

Qpeak = Qemax + Qmax ∗ Ie /It (B.6)

Now everything is known to calculate the application factor KAice :

KAice = Qpeak /Q0 (B.7)

Based on DNVGL-RU-SHIP Pt.6 Ch.6 15.5.3 [12]

B.4 Failure load of propeller

v
Fex ∗ D
u
u
d= 160t
3
4 (B.8)
σys haf t ∗ 1( did4 )
D: The Propeller diameter.
Fex : Blade Failure Load.

Where:

300 · c · t2 · σref
Fex = (B.9)
0.8 · D · 2 · r
σref :
c: Chord length at r/R.
t: Max blade thickness at r/R.

Where:
D ∗ r_R
r= (B.10)
2
r_R= relative radius of the stress scope.

Based on DNVGL-RU-SHIP Pt.6 Ch.6 12.5.2 [12]

XVI
 C
Appendix- Data

XVII
 C. Appendix- Data

C.1 Mimic formula compared with IACS M68.4

800-
50-800 10000 50-200 300-600
# D P n σB IACS Mimic
1 345 9715 74 320 535.0396 564.3666
2 221 5744 110 345 386.8769 404.7702
3 402 8332 134 337 412.2481 432.3935
4 737 3192 149 363 284.1325 295.7088
5 127 1066 126 496 193.2963 195.5131
6 444 2947 133 355 288.8261 301.2828
7 775 8389 195 459 338.8891 345.0075
8 113 7096 140 458 358.1198 364.654
9 599 2788 173 405 251.8423 259.3014
10 65 7822 56 494 492.6971 498.5115
11 406 1011 103 588 194.4166 194.1123
12 749 3477 51 425 402.614 412.6889
13 288 7247 189 555 310.8323 311.656
14 557 8879 51 380 565.1921 585.5235
15 793 6490 65 578 423.1435 423.0017
16 275 3854 64 389 394.5625 407.8198
17 503 7809 92 379 445.1116 461.2437
18 574 1425 190 433 192.0429 196.5164
19 342 3215 103 337 327.6361 343.6467
20 624 4909 108 303 380.2408 403.5678
21 140 9895 153 566 368.1221 368.5617
22 50 4118 150 404 300.9548 309.9409
23 618 3116 113 443 294.7564 301.0089
24 384 4343 71 508 371.519 375.0566
25 213 6244 181 518 305.4318 307.8642
26 279 2069 62 446 313.5894 320.0508
27 186 7057 52 316 542.4994 573.039
28 581 3275 64 588 337.1107 336.5831
29 357 2040 127 519 236.6182 238.4665
30 346 6168 71 465 426.9706 434.1955
31 523 4236 198 511 261.3678 263.7328
32 237 4610 181 471 282.7442 287.2156
33 266 1424 184 589 179.5282 179.2255
34 713 9538 122 547 395.6295 397.1102
35 550 3650 71 590 337.3291 336.7196
36 547 3313 172 591 243.0788 242.6104
37 668 4398 189 548 264.0276 264.9793
38 773 9734 75 316 534.4869 564.5754
39 166 3961 141 370 309.6226 321.6145
40 172 3824 61 332 414.7665 435.7368
41 139 8571 176 379 369.8617 383.2665

XVIII
42 89 6177 127 585 331.8959 331.4975
43 271 3153 172 349 272.2003 284.4446
44 257 5781 147 457 329.2528 335.3236
45 469 7391 81 385 454.2935 470.0313
46 450 1049 125 383 205.3137 212.5347
47 612 1137 117 449 207.5145 211.6656
48 210 7768 57 517 483.0713 486.9924
49 66 1888 198 486 202.1921 204.8524
50 485 3808 152 443 285.4793 291.5351
51 717 4290 148 537 285.5729 287.0437
52 239 3340 107 371 320.4931 332.8157
53 56 5118 67 596 383.911 382.9429
54 790 3106 64 470 350.7166 356.3271
55 269 7950 100 363 439.8972 457.8197
56 633 1321 192 551 175.6467 176.2075
57 346 4706 110 491 332.6254 336.7184
58 618 2464 179 577 218.6936 218.6477
59 678 5702 73 549 395.1379 396.5077
60 273 1943 182 429 216.5167 221.7459
61 413 1450 109 501 224.2037 226.5908
62 197 9273 92 315 491.6344 519.4952
63 548 8294 192 494 333.1893 337.1213
64 374 7731 173 442 346.4126 353.8317
65 510 7601 111 373 415.9133 431.672
66 150 9926 56 442 548.3518 560.0957
67 438 2170 109 365 276.9198 288.0411
68 448 7420 136 432 372.3149 381.067
69 56 7973 73 350 493.1423 515.1708
70 536 1864 91 376 277.6361 287.9253
71 163 4561 147 471 301.9723 306.7479
72 371 7028 165 320 367.6474 387.7992
73 324 4996 126 323 358.2319 377.4772
74 605 9595 121 468 413.5195 420.2866
75 284 4434 132 585 293.3714 293.0193
76 477 7996 194 427 340.0481 348.4084
77 780 1269 114 371 227.2742 236.0126
78 546 8977 65 432 507.4128 519.3406
79 74 3746 92 439 336.3931 343.8054
80 414 5179 118 460 340.9731 347.064
81 468 3270 117 552 280.1281 280.9842
82 293 1483 95 564 229.4153 229.7491
83 266 5210 109 499 343.7462 347.5184
84 767 7709 87 349 460.2386 480.9413
85 772 4835 133 366 338.2588 351.7457
86 298 3546 161 414 278.0117 285.6572
 87 540 9926 102 574 420.2961 420.3667
88 394 3458 90 357 346.5453 361.2815
89 592 9817 58 465 533.2771 542.3008
90 646 6164 141 355 362.2498 377.8731
91 796 4737 111 372 355.4831 369.0513
92 630 9319 66 589 472.6183 471.8214
93 56 9875 73 537 477.225 479.6829
94 520 9763 137 575 378.6644 378.6802
95 659 2476 150 598 230.1757 229.541
96 749 8717 79 536 446.1076 448.4692
97 529 7990 195 589 312.9004 312.3728
98 569 7899 200 302 363.1026 385.5251
99 302 3564 71 537 342.9348 344.7011
100 381 6594 159 540 321.327 322.8447
 C. Appendix- Data

C.2 Shaft diameter investigation input and out-

put

XXI
 C. Appendix- Data

C.3 Case study result data from Nauticus Align-

ment

On left-Dredger initial shaft design (BERG order design) On right- Derdger improved diameter

XXII
On left-RoPax initial shaft design (BERG order design) On right- RoPax improved diameter
On left-Bunker tanker initial shaft design (BERG order design) On right- Bunker tanker improved dia
Bunker tanker without propeller blade failure load criteria (no ICE)- Aftmost shaft from 350 to 300mm
 C.4

XXVI
Propeller Mass Inertia test previously done on the tanker shaft line

Only change is the propeller inertia J was first increased 25%, 50%, 100%
Then the prop inertia was dropped 25% and 50%
Entrained water was not changed
Larger plots below
C. Appendix- Data

100% Higher inertia 50% Higher inertia 25% Higher inertia

Original values 25% Lower inertia 50% Lower inertia

on flexible coupling alternating torque.
Case study - Effects of propeller inertia change
100% Higher inertia
50% Higher inertia
25% Higher inerta
Original values
25% Lower inertia
50% Lower inertia
 C. Appendix- Data

C.5 Input data for Nauticus Machinery Analysis

C.5.1 Shaft Alignment model inputs
C.5.2 Torsional Vibration Calculation model inputs
RoPax

DATA FOR MASSES AND DAMPERS

Mass DESCRIPTION RPM MASS MOMENT DAMPING FIRING FIRING TYPE

No RATIO OF INERTIA see note ANGLE A ANGLE B
[-] [kg*m^2] [N*s*m/rad] [deg] [deg]
--------------------------------------------------------------------------------------------
1 OPDamper 1.000 3.7700E+002 0.000E+000 Lump mass
2 IPDamper 1.000 1.2150E+002 0.000E+000 Lump mass
3 Cyl-A6B6 1.000 5.3670E+002 5.400E+002 0 50 Cylinder
4 Cyl-A5B5 1.000 5.3670E+002 5.400E+002 120 170 Cylinder
5 Cyl-A4B4 1.000 5.3670E+002 5.400E+002 600 650 Cylinder
6 Cyl-A3B3 1.000 5.3670E+002 5.400E+002 240 290 Cylinder
7 Cyl-A2B2 1.000 5.3670E+002 5.400E+002 480 530 Cylinder
8 Cyl-A1B1 1.000 5.3670E+002 5.400E+002 360 410 Cylinder
9 Flywheel 1.000 1.8360E+003 0.000E+000 Lump mass
10 PPCoup 1.000 1.4800E+002 0.000E+000 Lump mass
11 MPCoup 1.000 2.2800E+002 0.000E+000 Lump mass
12 SPCoup 1.000 3.9000E+002 0.000E+000 Lump mass
13 PTOWheel-J5-T95 1.000 9.4730E+001 0.000E+000 Lump mass
14 ClutchOP-J4 1.000 1.2636E+002 0.000E+000 Lump mass
15 ClutchIP-J3 1.000 7.6850E+001 0.000E+000 Lump mass
16 Pinion-J2-T22 1.000 4.3680E+001 0.000E+000 Lump mass
17 Gear-J1-T75 0.293 2.2622E+003 0.000E+000 Lump mass
18 GBFlange 0.293 1.3352E+002 0.000E+000 Lump mass
19 INT2frw 0.293 1.3352E+002 0.000E+000 Lump mass
20 INT2aft 0.293 1.3352E+002 0.000E+000 Lump mass
21 SpacerRing 0.293 3.0000E+001 0.000E+000 Lump mass
22 BrakeDisc 0.293 1.1992E+002 0.000E+000 Lump mass
23 INT1frw 0.293 1.3669E+002 0.000E+000 Lump mass
24 INT1aft 0.293 1.3669E+002 0.000E+000 Lump mass
25 OKCX 0.293 4.7800E+001 0.000E+000 Lump mass
26 PropShaft-frw 0.293 4.1140E+002 0.000E+000 Lump mass
27 PropShaft-aft 0.293 4.1140E+002 0.000E+000 Lump mass
28 Propeller1 0.293 1.9150E+004 0.000E+000 Propeller
29 HubCap 0.293 1.7300E+002 0.000E+000 Lump mass
30 PTOPinion 3.519 2.9800E+000 0.000E+000 Lump mass
31 PTOClutchIn 3.519 1.8000E+000 0.000E+000 Lump mass
32 PTOClutchOut 3.519 1.5200E+000 0.000E+000 Lump mass
33 PTOCoupPP 3.519 3.7000E+000 0.000E+000 Lump mass
34 PTOCoupMP 3.519 4.7000E+000 0.000E+000 Lump mass
35 PTOCoupMP1 3.519 4.7000E+000 0.000E+000 Lump mass
36 PTOCoupSP 3.519 1.0400E+001 0.000E+000 Lump mass
37 Generator1 3.519 5.5000E+001 0.000E+000 Generator

Note: Negative sign on mass damping means dynamic magnifier

DATA FOR SHAFT ELEMENTS

SHAFT NODES RPM TORSIONAL DAMPING DIAMETER TYPE

No RATIO STIFFNESS see note OUTER INNER
[-] [N*m/rad] [N*s*m/rad] [mm] [mm]
---------------------------------------------------------------------------------------
1 1 2 1.000 1.350E+007 3.200E+004 0.0 0.0 DamperStiffness
2 2 3 1.000 3.106E+008 -180.0 380.0 0.0 Shaft
3 3 4 1.000 2.012E+008 2.012E+004 380.0 0.0 Crankthrow
4 4 5 1.000 2.012E+008 2.012E+004 380.0 0.0 Crankthrow
5 5 6 1.000 2.012E+008 2.012E+004 380.0 0.0 Crankthrow
6 6 7 1.000 2.012E+008 2.012E+004 380.0 0.0 Crankthrow

XXXIII
C. Appendix- Data

7 7 8 1.000 2.012E+008 2.012E+004 380.0 0.0 Crankthrow

8 8 9 1.000 2.551E+008 -180.0 0.0 0.0 Shaft
9 9 10 1.000 1.000E+012 -180.0 0.0 0.0 Shaft
10 10 11 1.000 3.038E+006 -7.1 0.0 0.0 Coupling
11 11 12 1.000 3.038E+006 -7.1 0.0 0.0 Coupling
12 12 13 1.000 1.086E+008 -180.0 310.0 0.0 Shaft
13 13 14 1.000 2.495E+009 -180.0 855.0 0.0 Shaft
14 14 15 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft
15 15 16 1.000 1.200E+008 -180.0 295.0 0.0 Shaft
16 16 17 1.000 1.000E+012 -100.0 0.0 0.0 Mesh
17 17 18 0.293 4.126E+008 -180.0 450.0 0.0 Shaft
18 18 19 0.293 1.000E+012 0.000E+000 0.0 0.0 Shaft
19 19 20 0.293 1.990E+007 -180.0 405.0 0.0 Shaft
20 20 21 0.293 1.000E+012 0.000E+000 0.0 0.0 Shaft
21 21 22 0.293 1.000E+012 0.000E+000 0.0 0.0 Shaft
22 22 23 0.293 1.000E+012 0.000E+000 0.0 0.0 Shaft
23 23 24 0.293 1.790E+007 -180.0 405.0 0.0 Shaft
24 24 25 0.293 1.000E+012 0.000E+000 0.0 0.0 Shaft
25 25 26 0.293 1.000E+012 0.000E+000 0.0 0.0 Shaft
26 26 27 0.293 1.600E+007 -180.0 420.0 0.0 Shaft
27 27 28 0.293 1.000E+012 0.000E+000 0.0 0.0 Shaft
28 28 29 0.293 1.000E+012 0.000E+000 0.0 0.0 Shaft
29 30 31 3.519 9.680E+006 -180.0 140.0 0.0 Shaft
30 13 30 3.519 1.000E+012 -100.0 0.0 0.0 Mesh
31 31 32 3.519 7.508E+007 0.000E+000 210.0 0.0 Shaft
32 32 33 3.519 4.110E+006 -180.0 130.0 0.0 Shaft
33 33 34 3.519 1.350E+005 -6.1 0.0 0.0 Coupling
34 34 35 3.519 1.350E+005 -6.1 0.0 0.0 Coupling
35 35 36 3.519 1.350E+005 -6.1 0.0 0.0 Coupling
36 36 37 3.519 1.406E+007 -180.0 160.0 0.0 Shaft

Note: Negative sign on shaft damping means dynamic magnifier

ENGINE DATA

Index: #2
Manufacturer: MaK
Type: 12VM43C
Crank throw:

Mass numbers: 3,4,5,6,7,8

RPM (@MCR): 514.0 rpm
Total power (@MCR): 12600.0 kW
Firing order: A1-B1-A2-B2-A4-B4-A6-B6-A5-B5-A3-B3
Firing interval: Uneven
Firing type: Consecutive
Number of strokes: 4
Firing angle between banks: 50.00 deg
Cylinder diameter: 430.0 mm
Length of stroke: 610.0 mm
Connecting rod ratio: 0.236
Compression ratio: 16.8
Reciprocating mass (per cylinder): 467.4 kg
Number of cylinders: 12
Angle ignition: -5.0 deg
Angle end of combustion: 55.0 deg
Angle exhaust starts: 150.0 deg
Angle inlet starts: 0.0 deg
Angle Inlet/exhaust close: 0.0 deg
Compression Ratio: 16.8
Thermal efficiency: 0.47
Compression exponent: 1.39
Expansion exponent: 1.33
Form Factor: 1.00
Pressure charge air: 3.40 bar
Pressure compression: 175 bar
Max cylinder pressure at MCR: 210 bar
Temperature charge air: 45 DegC

XXXIV
 C. Appendix- Data

DAMPER DATA

Damper type: Steel

Node numbers: 1-2

Inertia outer member: 377.00000 kg*m^2
Inertia inner member: 121.50000 kg*m^2
Dynamic magnifier: 0.000
Damping coefficient: 32000.000 N*s*m/rad
Torsional stiffness: 1.350E+007 N*m/rad

GEAR DATA

Manufacturer: RENK
Type: RSHL-1120

Mesh index: 16
Node numbers: 16-17
RPM ratio: 1.000
Ratio reference: Ingoing

GEAR DATA

Manufacturer:
Type:

Mesh index: 30
Node numbers: 13-30
RPM ratio: 3.519
Ratio reference: Outgoing

PROPELLER DATA

Index: #28
Manufacturer:
Type:

Ice class notation: ICE-1C

MCR Power: 12600.0 kW

MCR Speed: 150.8 rpm
Bollard Speed: 150.8 rpm
Ducted propeller: No
Pitch type: Controllable pitch
Number of propeller blades: 4
Damping method: Frahm (Frahm number 2.9)

Boss diameter: 965.0 mm

Blade diameter: 4800.0 mm
Pitch ratio: 1.000
Expanded area ratio (Ae/Ao): 0.000

Inertia, in air (@MCR): 1.327E+004 kg*m^2

Inertia, entrained water (@MCR): 5.880E+003 kg*m^2
Inertia, entrained water 0 Pitch (@MCR): 9.700E+002 kg*m^2

Propeller excitation in % of shaft torque (Nm): 6.0%

Propeller excitation (*2) in % of shaft torque (Nm): 1.5%

Bunker tanker - Original diameter

DATA FOR MASSES AND DAMPERS

Mass DESCRIPTION RPM MASS MOMENT DAMPING FIRING TYPE

No RATIO OF INERTIA see note ANGLE

XXXV
C. Appendix- Data

[-] [kg*m^2] [N*s*m/rad] [deg]

---------------------------------------------------------------------------------------
1 ElectricMotor1 1.000 8.5000E+001 0.000E+000 Electric motor
2 ShaftEnd 1.000 1.5000E-001 0.000E+000 Lump mass
3 CouplingPrimary1 1.000 2.6500E+000 0.000E+000 Lump mass
4 CouplingSecondar 1.000 4.2200E+000 0.000E+000 Lump mass
5 ElectricMotor2 1.000 8.5000E+001 0.000E+000 Electric motor
6 ShaftEnd 1.000 1.5000E-001 0.000E+000 Lump mass
7 CouplingPrimary2 1.000 2.6500E+000 0.000E+000 Lump mass
8 CouplingSecondar 1.000 4.2200E+000 0.000E+000 Lump mass
9 InputGear 1.000 2.1900E+000 0.000E+000 Lump mass
10 OutputGear 0.118 1.7610E+003 0.000E+000 Lump mass
11 OMC-Coupling 0.118 2.1730E+001 0.000E+000 Lump mass
12 BreakDisk+HalfSh 0.118 4.0000E+001 0.000E+000 Lump mass
13 HalfPropShaft 0.118 1.9925E+001 0.000E+000 Lump mass
14 Propeller1 0.118 3.9938E+003 0.000E+000 Propeller

Note: Negative sign on mass damping means dynamic magnifier

DATA FOR SHAFT ELEMENTS

SHAFT NODES RPM TORSIONAL DAMPING DIAMETER TYPE

No RATIO STIFFNESS see note OUTER INNER
[-] [N*m/rad] [N*s*m/rad] [mm] [mm]
---------------------------------------------------------------------------------------
1 1 2 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft
2 3 4 1.000 9.676E+004 -7.0 0.0 0.0 Coupling
3 2 3 1.000 5.460E+006 -180.0 140.0 0.0 Shaft
4 5 6 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft
5 6 7 1.000 5.460E+006 -180.0 140.0 0.0 Shaft
6 7 8 1.000 9.676E+004 -7.0 0.0 0.0 Coupling
7 9 10 1.000 1.000E+012 -100.0 0.0 0.0 Mesh
8 4 9 1.000 1.594E+007 -180.0 150.0 0.0 Shaft
9 8 9 1.000 1.613E+007 -180.0 150.0 0.0 Shaft
10 10 11 0.118 6.575E+007 -180.0 0.0 0.0 Shaft
11 11 12 0.118 1.000E+012 0.000E+000 0.0 0.0 Shaft
12 12 13 0.118 6.800E+006 -180.0 268.0 90.0 Shaft
13 13 14 0.118 1.000E+012 0.000E+000 0.0 0.0 Shaft

Note: Negative sign on shaft damping means dynamic magnifier

Input data for system components: propulsion motor/engine, reduction gear,

flexible couplings and propeller.
This information is given for the original system and not repeated. Changes are not
done on those components, just the shafts.

ELECTRIC MOTOR DATA

Manufacturer: Siemens 1NA1504

Type: A5E46380263A
Node number: 1
RPM (@MCR): 1200.0 rpm
Power (@ MCR): 1250.0 kW

ELECTRIC MOTOR DATA

Manufacturer: Siemens 1NA1504

Type: A5E46380263A
Node number: 5
RPM (@MCR): 1200.0 rpm
Power (@ MCR): 1250.0 kW

GEAR DATA

Manufacturer: RENK B747507

XXXVI
 C. Appendix- Data

Type: RSV-1060HR i=8.45

Mesh index: 7
Node numbers: 9-10
RPM ratio: 1.000
Ratio reference: Ingoing

ELASTIC COUPLING DATA

Manufacturer:
Type: CX-72-GFS1-300-60-0637
Stiffness index: 2
Node number: 3-4
Stiffness type: Linear
Nominal coupling torque: 0.000 kN*m
Norminal torsional stiffness: 9.676E+004 N*m/rad

ELASTIC COUPLING DATA

Manufacturer:
Type: CX-72-GFS1-300-60-0637
Stiffness index: 6
Node number: 7-8
Stiffness type: Linear
Nominal coupling torque: 0.000 kN*m
Norminal torsional stiffness: 9.676E+004 N*m/rad

PROPELLER DATA

Index: #14
Manufacturer:
Type:
Ice class notation: ICE-1B
MCR Power: 2500.0 kW
MCR Speed: 142.0 rpm
Bollard Speed: 142.0 rpm
Ducted propeller: No
Pitch type: Controllable pitch
Number of propeller blades: 4
Damping method: Archer (Archer number 27.0)
Boss diameter: 760.0 mm
Blade diameter: 3600.0 mm
Pitch ratio: 1.000
Expanded area ratio (Ae/Ao): 0.000

Inertia, in air (@MCR): 2.647E+003 kg*m^2

Inertia, entrained water (@MCR): 1.347E+003 kg*m^2
Inertia, entrained water 0 Pitch (@MCR): 3.367E+002 kg*m^2

Propeller excitation in % of shaft torque (Nm): 6.0%

Propeller excitation (*2) in % of shaft torque (Nm): 2.0%

Bunker tanker - Diameter decreased by 10 mm

DATA FOR MASSES AND DAMPERS

Mass DESCRIPTION RPM MASS MOMENT DAMPING FIRING TYPE

No RATIO OF INERTIA see note ANGLE
[-] [kg*m^2] [N*s*m/rad] [deg]
---------------------------------------------------------------------------------------
1 ElectricMotor1 1.000 8.5000E+001 0.000E+000 Electric motor
2 ShaftEnd 1.000 1.5000E-001 0.000E+000 Lump mass
3 CouplingPrimary1 1.000 2.6500E+000 0.000E+000 Lump mass
4 CouplingSecondar 1.000 4.2200E+000 0.000E+000 Lump mass
5 ElectricMotor2 1.000 8.5000E+001 0.000E+000 Electric motor
6 ShaftEnd 1.000 1.5000E-001 0.000E+000 Lump mass
7 CouplingPrimary2 1.000 2.6500E+000 0.000E+000 Lump mass
8 CouplingSecondar 1.000 4.2200E+000 0.000E+000 Lump mass

XXXVII
C. Appendix- Data

9 InputGear 1.000 2.1900E+000 0.000E+000 Lump mass

10 OutputGear 0.118 1.7610E+003 0.000E+000 Lump mass
11 OMC-Coupling 0.118 2.0110E+001 0.000E+000 Lump mass
12 BreakDisk+HalfSh 0.118 3.7460E+001 0.000E+000 Lump mass
13 HalfPropShaft 0.118 1.7350E+001 0.000E+000 Lump mass
14 Propeller1 0.118 3.9938E+003 0.000E+000 Propeller

Note: Negative sign on mass damping means dynamic magnifier

DATA FOR SHAFT ELEMENTS

SHAFT NODES RPM TORSIONAL DAMPING DIAMETER TYPE

No RATIO STIFFNESS see note OUTER INNER
[-] [N*m/rad] [N*s*m/rad] [mm] [mm]
---------------------------------------------------------------------------------------
1 1 2 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft
2 3 4 1.000 9.676E+004 -7.0 0.0 0.0 Coupling
3 2 3 1.000 5.460E+006 -180.0 140.0 0.0 Shaft
4 5 6 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft
5 6 7 1.000 5.460E+006 -180.0 140.0 0.0 Shaft
6 7 8 1.000 9.676E+004 -7.0 0.0 0.0 Coupling
7 9 10 1.000 1.000E+012 -100.0 0.0 0.0 Mesh
8 4 9 1.000 1.594E+007 -180.0 150.0 0.0 Shaft
9 8 9 1.000 1.613E+007 -180.0 150.0 0.0 Shaft
10 10 11 0.118 6.575E+007 -180.0 0.0 0.0 Shaft
11 11 12 0.118 1.000E+012 0.000E+000 0.0 0.0 Shaft
12 12 13 0.118 5.860E+006 -180.0 258.0 90.0 Shaft
13 13 14 0.118 1.000E+012 0.000E+000 0.0 0.0 Shaft

Note: Negative sign on shaft damping means dynamic magnifier

Bunker tanker - Diameter decreased by 20 mm

DATA FOR MASSES AND DAMPERS

Mass DESCRIPTION RPM MASS MOMENT DAMPING FIRING TYPE

No RATIO OF INERTIA see note ANGLE
[-] [kg*m^2] [N*s*m/rad] [deg]
---------------------------------------------------------------------------------------
1 ElectricMotor1 1.000 8.5000E+001 0.000E+000 Electric motor
2 ShaftEnd 1.000 1.5000E-001 0.000E+000 Lump mass
3 CouplingPrimary1 1.000 2.6500E+000 0.000E+000 Lump mass
4 CouplingSecondar 1.000 4.2200E+000 0.000E+000 Lump mass
5 ElectricMotor2 1.000 8.5000E+001 0.000E+000 Electric motor
6 ShaftEnd 1.000 1.5000E-001 0.000E+000 Lump mass
7 CouplingPrimary2 1.000 2.6500E+000 0.000E+000 Lump mass
8 CouplingSecondar 1.000 4.2200E+000 0.000E+000 Lump mass
9 InputGear 1.000 2.1900E+000 0.000E+000 Lump mass
10 OutputGear 0.118 1.7610E+003 0.000E+000 Lump mass
11 OMC-Coupling 0.118 2.1730E+001 0.000E+000 Lump mass
12 BreakDisk+HalfSh 0.118 3.5115E+001 0.000E+000 Lump mass
13 HalfPropShaft 0.118 1.5035E+001 0.000E+000 Lump mass
14 Propeller1 0.118 3.9938E+003 0.000E+000 Propeller

Note: Negative sign on mass damping means dynamic magnifier

DATA FOR SHAFT ELEMENTS

SHAFT NODES RPM TORSIONAL DAMPING DIAMETER TYPE

No RATIO STIFFNESS see note OUTER INNER
[-] [N*m/rad] [N*s*m/rad] [mm] [mm]
---------------------------------------------------------------------------------------
1 1 2 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft
2 3 4 1.000 9.676E+004 -7.0 0.0 0.0 Coupling
3 2 3 1.000 5.460E+006 -180.0 140.0 0.0 Shaft
4 5 6 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft

XXXVIII
 C. Appendix- Data

5 6 7 1.000 5.460E+006 -180.0 140.0 0.0 Shaft

6 7 8 1.000 9.676E+004 -7.0 0.0 0.0 Coupling
7 9 10 1.000 1.000E+012 -100.0 0.0 0.0 Mesh
8 4 9 1.000 1.594E+007 -180.0 150.0 0.0 Shaft
9 8 9 1.000 1.613E+007 -180.0 150.0 0.0 Shaft
10 10 11 0.118 6.575E+007 -180.0 0.0 0.0 Shaft
11 11 12 0.118 1.000E+012 0.000E+000 0.0 0.0 Shaft
12 12 13 0.118 5.010E+006 -180.0 248.0 90.0 Shaft
13 13 14 0.118 1.000E+012 0.000E+000 0.0 0.0 Shaft

Note: Negative sign on shaft damping means dynamic magnifier

Bunker tanker - Diameter decreased by 30 mm

DATA FOR MASSES AND DAMPERS

Mass DESCRIPTION RPM MASS MOMENT DAMPING FIRING TYPE

No RATIO OF INERTIA see note ANGLE
[-] [kg*m^2] [N*s*m/rad] [deg]
---------------------------------------------------------------------------------------
1 ElectricMotor1 1.000 8.5000E+001 0.000E+000 Electric motor
2 ShaftEnd 1.000 1.5000E-001 0.000E+000 Lump mass
3 CouplingPrimary1 1.000 2.6500E+000 0.000E+000 Lump mass
4 CouplingSecondar 1.000 4.2200E+000 0.000E+000 Lump mass
5 ElectricMotor2 1.000 8.5000E+001 0.000E+000 Electric motor
6 ShaftEnd 1.000 1.5000E-001 0.000E+000 Lump mass
7 CouplingPrimary2 1.000 2.6500E+000 0.000E+000 Lump mass
8 CouplingSecondar 1.000 4.2200E+000 0.000E+000 Lump mass
9 InputGear 1.000 2.1900E+000 0.000E+000 Lump mass
10 OutputGear 0.118 1.7610E+003 0.000E+000 Lump mass
11 OMC-Coupling 0.118 2.1730E+001 0.000E+000 Lump mass
12 BreakDisk+HalfSh 0.118 3.3040E+001 0.000E+000 Lump mass
13 HalfPropShaft 0.118 1.2960E+001 0.000E+000 Lump mass
14 Propeller1 0.118 3.9938E+003 0.000E+000 Propeller

Note: Negative sign on mass damping means dynamic magnifier

DATA FOR SHAFT ELEMENTS

SHAFT NODES RPM TORSIONAL DAMPING DIAMETER TYPE

No RATIO STIFFNESS see note OUTER INNER
[-] [N*m/rad] [N*s*m/rad] [mm] [mm]
---------------------------------------------------------------------------------------
1 1 2 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft
2 3 4 1.000 9.676E+004 -7.0 0.0 0.0 Coupling
3 2 3 1.000 5.460E+006 -180.0 140.0 0.0 Shaft
4 5 6 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft
5 6 7 1.000 5.460E+006 -180.0 140.0 0.0 Shaft
6 7 8 1.000 9.676E+004 -7.0 0.0 0.0 Coupling
7 9 10 1.000 1.000E+012 -100.0 0.0 0.0 Mesh
8 4 9 1.000 1.594E+007 -180.0 150.0 0.0 Shaft
9 8 9 1.000 1.613E+007 -180.0 150.0 0.0 Shaft
10 10 11 0.118 6.575E+007 -180.0 0.0 0.0 Shaft
11 11 12 0.118 1.000E+012 0.000E+000 0.0 0.0 Shaft
12 12 13 0.118 4.260E+006 -180.0 238.0 90.0 Shaft
13 13 14 0.118 1.000E+012 0.000E+000 0.0 0.0 Shaft

Note: Negative sign on shaft damping means dynamic magnifier

Dredger - Original diameter

DATA FOR MASSES AND DAMPERS

Mass DESCRIPTION RPM MASS MOMENT DAMPING FIRING TYPE

No RATIO OF INERTIA see note ANGLE

XXXIX
C. Appendix- Data

[-] [kg*m^2] [N*s*m/rad] [deg]

---------------------------------------------------------------------------------------
1 MotorRotor 1.000 1.7490E+002 0.000E+000 Electric motor
2 MotorOutput 1.000 8.5000E+000 0.000E+000 Lump mass
3 CoupSP 1.000 1.6950E+001 0.000E+000 Lump mass
4 CoupPP-B 1.000 8.9400E+000 0.000E+000 Lump mass
5 GB-Pinion-20T 1.000 2.9700E+000 0.000E+000 Lump mass
6 GB-Wheel-122T 0.164 8.2596E+002 0.000E+000 Lump mass
7 BreakDisc 0.164 2.7630E+001 0.000E+000 Lump mass
8 INTSFlange 0.164 1.4705E+001 0.000E+000 Lump mass
9 OMC 0.164 5.0740E+001 0.000E+000 Lump mass
10 PropFlange 0.164 2.9010E+001 0.000E+000 Lump mass
11 Propeller1 0.164 2.4000E+003 0.000E+000 Propeller

Note: Negative sign on mass damping means dynamic magnifier

DATA FOR SHAFT ELEMENTS

SHAFT NODES RPM TORSIONAL DAMPING DIAMETER TYPE

No RATIO STIFFNESS see note OUTER INNER
[-] [N*m/rad] [N*s*m/rad] [mm] [mm]
---------------------------------------------------------------------------------------
1 1 2 1.000 9.270E+006 -180.0 180.0 0.0 Shaft
2 2 3 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft
3 3 4 1.000 2.465E+005 -7.1 0.0 0.0 Coupling
4 4 5 1.000 2.613E+007 -180.0 200.0 0.0 Shaft
5 5 6 1.000 1.000E+012 -100.0 0.0 0.0 Mesh
6 6 7 0.164 6.694E+007 -180.0 249.0 0.0 Shaft
7 7 8 0.164 6.640E+006 -180.0 260.0 0.0 Shaft
8 8 9 0.164 1.000E+012 0.000E+000 0.0 0.0 Shaft
9 9 10 0.164 4.740E+006 -180.0 255.0 0.0 Shaft
10 10 11 0.164 1.000E+012 0.000E+000 0.0 0.0 Shaft

Note: Negative sign on shaft damping means dynamic magnifier

ELECTRIC MOTOR DATA

Manufacturer: AMI
Type: 560L6L BSFTMS

Node number: 1
RPM (@MCR): 1200.0 rpm
Power (@ MCR): 4000.0 kW

GEAR DATA

Manufacturer: RENK
Type: RSV-900 HR Ratio 20/122

Mesh index: 5
Node numbers: 5-6
RPM ratio: 1.000
Ratio reference: Ingoing

ELASTIC COUPLING DATA

Manufacturer: CENTA
Type: CX-179-GSS1-50Sh

Stiffness index: 3
Node number: 3-4
Stiffness type: Linear
Nominal coupling torque: 0.000 kN*m
Norminal torsional stiffness: 2.465E+005 N*m/rad

XL
 C. Appendix- Data

PROPELLER DATA

Index: #11
Manufacturer: BERG Propulsion
Type: MPP 950 Fixed

Ice class notation: No ice class

MCR Power: 4000.0 kW

MCR Speed: 197.0 rpm
Ducted propeller: Yes
Pitch type: Fixed pitch
Number of propeller blades: 4
Damping method: Archer (Archer number 27.0)

Boss diameter: 0.0 mm

Blade diameter: 0.0 mm
Pitch ratio: 1.000
Expanded area ratio (Ae/Ao): 0.000

Inertia, in air (@MCR): 1.360E+003 kg*m^2

Inertia, entrained water (@MCR): 1.040E+003 kg*m^2

Propeller excitation in % of shaft torque (Nm): 6.0%

Propeller excitation (*2) in % of shaft torque (Nm): 2.0%

Dredger - Diameter decreased, 230mm propshaft and 225mm intermedi-

ate
DATA FOR MASSES AND DAMPERS

Mass DESCRIPTION RPM MASS MOMENT DAMPING FIRING TYPE

No RATIO OF INERTIA see note ANGLE
[-] [kg*m^2] [N*s*m/rad] [deg]
---------------------------------------------------------------------------------------
1 MotorRotor 1.000 1.7490E+002 0.000E+000 Electric motor
2 MotorOutput 1.000 8.5000E+000 0.000E+000 Lump mass
3 CoupSP 1.000 1.6950E+001 0.000E+000 Lump mass
4 CoupPP-B 1.000 8.9400E+000 0.000E+000 Lump mass
5 GB-Pinion-20T 1.000 2.9700E+000 0.000E+000 Lump mass
6 GB-Wheel-122T 0.164 8.2596E+002 0.000E+000 Lump mass
7 BreakDisc+HalfINT0.164 3.8845E+001 0.000E+000 Lump mass
8 INTSFlange 0.164 1.1225E+001 0.000E+000 Lump mass
9 OMC+HalfProp 0.164 3.3675E+001 0.000E+000 Lump mass
10 PropFlange 0.164 1.1945E+001 0.000E+000 Lump mass
11 Propeller1 0.164 2.4000E+003 0.000E+000 Propeller

Note: Negative sign on mass damping means dynamic magnifier

DATA FOR SHAFT ELEMENTS

SHAFT NODES RPM TORSIONAL DAMPING DIAMETER TYPE

No RATIO STIFFNESS see note OUTER INNER
[-] [N*m/rad] [N*s*m/rad] [mm] [mm]
---------------------------------------------------------------------------------------
1 1 2 1.000 9.270E+006 -180.0 180.0 0.0 Shaft
2 2 3 1.000 1.000E+012 0.000E+000 0.0 0.0 Shaft
3 3 4 1.000 2.465E+005 -7.1 0.0 0.0 Coupling
4 4 5 1.000 2.613E+007 -180.0 200.0 0.0 Shaft
5 5 6 1.000 1.000E+012 -100.0 0.0 0.0 Mesh
6 6 7 0.164 6.694E+007 -180.0 249.0 0.0 Shaft
7 7 8 0.164 3.920E+006 -180.0 225.0 0.0 Shaft
8 8 9 0.164 1.000E+012 0.000E+000 0.0 0.0 Shaft
9 9 10 0.164 2.040E+006 -180.0 230.0 0.0 Shaft

XLI
C. Appendix- Data

10 10 11 0.164 1.000E+012 0.000E+000 0.0 0.0 Shaft

Note: Negative sign on shaft damping means dynamic magnifier

C.5.3 Propeller curves for the dredger

The table below lists the propeller torque and propeller power consumption versus
speed, for the two operating conditions.

Free running empty condition Dredging condition

Prop Speed Prop Torque Power Prop Speed Prop Torque Power
[RPM] [kNm] [kW] [RPM] [kNm] [kW]
20 2.0 4.1 20 2.7 5.6
30 4.4 13.8 30 6.0 18.9
40 7.8 32.6 40 10.7 44.8
50 12.2 63.7 50 16.7 87.4
60 17.5 110.0 60 24.0 151.1
70 23.9 174.9 70 32.7 240.0
80 31.2 261.1 80 42.7 358.1
90 39.4 371.8 90 54.1 510.1
100 48.7 509.9 100 66.8 699.8
110 58.9 678.7 110 80.8 931.3
120 70.1 881.1 120 96.2 1209.0
130 82.3 1120.2 130 112.9 1537.2
140 95.4 1399.0 140 131.0 1920.1
150 109.5 1720.6 150 150.3 2361.6
160 124.6 2088.1 160 171.0 2865.6
170 140.7 2504.5 170 193.1 3437.8
180 157.7 2972.9 177 209.3 3879.9
190 175.7 3496.6
196.7 188.4 3880.0

Table C.1: Propeller torque values from given curves and calculated power for given
propeller speed.

XLII
 C. Appendix- Data

Below is a .txt file defining the propeller power curve for the free running condition.
This file is loaded in the propeller load definition.
// Comments starting with two backslash are allowed and will not
be read
// Blank lines shall also be allowed

//’Speed’ and ’Power’ columns MUST be included in custom load curve

//’Speed’ column should be with ’Speed’ word on top, it should be the
first column
//’Power’ column should be with ’Power’ word on top and this column
should be on right side of ’Speed’ column
//’PitchRatio’ column CAN be included. If it is, it should be the
third column and with ’PitchRatio’ word on top
//’EntrainedWater’ column CAN be included. If it is, it should be the
last column and with ’EntrainedWater’ word on top
//If Custom load curve contains only the ’Speed’ and ’Power’ columns,
it means the user has defined the power changes to each speed
//If ’PitchRatio’ column is defined, it means power changes according
to PitchRatio.
//’EntrainedWater’ column could be added if user wants to define the
Inertia of Entrained Water of Propeller changes to Speed (or
PitchRatio).
//’EntrainedWater’ column could be empty in some of the cells but the
other columns should be fully specified at each step
//If ’PitchRatio’ column is not defined, no duplicated values are
allowed in the "Speed" column .
//If ’PitchRatio’ column is defined, no duplicated values are allowed
in the ’PitchRatio’ column.But duplicated values are allowed for
"Speed"
column to simulate like ’constant speed and variable pitch’
condition.

//User can freely define the speed step (or PitchRatio steps)
//The unit of each column is fixed as this example

// Dredger
// Free running empty condition

Speed[Rpm] Power[KW]
20 4.1
30 13.8
40 32.6
50 63.7
60 110.0
70 174.9
80 261.1
90 371.8
100 509.9
110 678.7
120 881.1
130 1120.2
140 1399.0
150 1720.6
160 2088.1
170 2504.5
180 2972.9
190 3496.62
196.7 3880.0

XLIII
C. Appendix- Data

Below is a .txt file defining the propeller power curve for the dredging condition.
This file is loaded in the propeller load definition.
// Comments starting with two backslash are allowed and will not be
read
// Blank lines shall also be allowed

//’Speed’ and ’Power’ columns MUST be included in custom load curve

//’Speed’ column should be with ’Speed’ word on top, it should be the
first column
//’Power’ column should be with ’Power’ word on top and this column
should be on right side of ’Speed’ column
//’PitchRatio’ column CAN be included. If it is, it should be the
third column and with ’PitchRatio’ word on top
//’EntrainedWater’ column CAN be included. If it is, it should be the
last column and with ’EntrainedWater’ word on top
//If Custom load curve contains only the ’Speed’ and ’Power’ columns,
it means the user has defined the power changes to each speed
//If ’PitchRatio’ column is defined, it means power changes according
to PitchRatio.
//’EntrainedWater’ column could be added if user wants to define the
Inertia of Entrained Water of Propeller changes to Speed (or
PitchRatio).
//’EntrainedWater’ column could be empty in some of the cells but the
other columns should be fully specified at each step
//If ’PitchRatio’ column is not defined, no duplicated values are
allowed in the "Speed" column .
//If ’PitchRatio’ column is defined, no duplicated values are allowed
in the ’PitchRatio’ column.But duplicated values are allowed for
"Speed"
column to simulate like ’constant speed and variable pitch’
condition.

//User can freely define the speed step (or PitchRatio steps)
//The unit of each column is fixed as this example

// Dredger
// Dredging condition

Speed[Rpm] Power[KW]
20 5.6
30 18.9
40 44.8
50 87.43
60 151.1
70 240.0
80 358.1
90 510.1
100 700.0
110 931.3
120 1209.0
130 1537.2
140 1920.1
150 2361.6
160 2865.6
170 3437.8
177 3880.0

XLIV
DEPARTMENT OF SOME SUBJECT OR TECHNOLOGY
CHALMERS UNIVERSITY OF TECHNOLOGY
Gothenburg, Sweden
www.chalmers.se