Attachment 0

Ref.
Ares(2019)781524 - 11/02/2019
Holistic Optimisation of Ship Design and Operation for Life Cycle
Horizon 2020 - 689074
Structural and Functional Simulation

Tools for Platform I1
Deliverable D4.2
February 2019
Authors: Dr. Lars Molter (CMT); Abbas Bayatfar, Ph Rigo (Université de Liège)
ABSTRACT
This deliverable shall give an overview about public available information of the work
and tools developed in the scope of the HOLISHIP project. The work presented is
focused on structural and functional tools and their implementation to the CAESES®
platform. For more detailed information, please contact the author.
The research leading to these results has received funding from the European
Commission. The information contained in this report is subject to change without
notice and should not be construed as a commitment by any members of the
HOLISHIP Consortium. In the event of any software or algorithms being described in
this report, the aforementioned Consortium assumes no responsibility for the use or
inability to use any of its software or algorithms. The information is provided without
any warranty of any kind and the Holiship-Consortium expressly disclaims all implied
warranties, including but not limited to the implied warranties of merchantability and
fitness for a particular use.

COPYRIGHT 2017 The HOLISHIP Consortium
This document may not be copied, reproduced, or modified in whole or in part for
any purpose without written permission from the HOLISHIP Consortium. In addition, to
such written permission to copy, acknowledgement of the authors of the document
and all applicable portions of the copyright notice must be clearly referenced.
All rights reserved.
2
Document data
Document Title: Structural and Functional Simulation Tools for Platform I1
Work Package(s): 4
Dissemination level: Public Confidential
Deliverable nature: Deliverable
Lead beneficiary: CMT
Responsible author: Dr. Lars Molter [CMT]
Co-authors: Abbas Bayatfar [ULG], Philippe Rigo [ULG]
Reviewer: Apostolos Papanikolaou [HSVA]
Date of delivery: 11.2.2019

WP Partners Quality Assurance Group
Circulation:
Steering Group EC
Document history
Version Date Description
1 30.10.2018 First version
2 5.11.2018 Updated version
4 6.02.2019 Version for QC
5 10.2.2019 Revised final version
3
Contents
ABSTRACT.................................................................................................................................................... 2
1. INTRODUCTION .................................................................................................................................. 6
2. THE STRUCTURAL ASSESSMENT TOOL ............................................................................................... 7
3. PRESENTATION OF THE TOOLS, DEVELOPED AND USED IN TASK 4.2 (CONCEPT STAGE) ........... 8
3.1 GENERAL METHODOLOGY ...................................................................................................................... 8
3.2 LIST OF TOOLS ..................................................................................................................................... 10
3.3 MARS2000® BV SOFTWARE ............................................................................................................... 10
3.4 STEEL® BV TOOL ................................................................................................................................. 12
3.5 MODEFRONTIER® .............................................................................................................................. 14
3.6 COUPLING OF THE WP4 TOOLS AND THE WP7 INTEGRATION PLATFORM (CONCEPT DESIGN) ............... 16
3.7 SCANTLING-SPACING UPDATER TOOL .................................................................................................. 19
3.8 RULE VIOLATION INDICATOR TOOL ....................................................................................................... 21
3.9 STRESS ASSESSMENT TOOL ...................................................................................................................... 22
3.10 THICKNESS CORROSION UPDATER TOOL ............................................................................................... 22
3.11 STEEL FRAME WEIGHT ASSESSMENT TOOL .............................................................................................. 23
3.12 DESIGN VARIABLES AND PARAMETERS .................................................................................................. 23
3.13 CONSTRAINT TOOL ................................................................................................................................ 24
3.14 WEIGHT/CG CALCULATOR TOOL ........................................................................................................ 25
3.15 GENERATOR OF PRODUCIBILITY INDEXES ............................................................................................... 26
3.16 DEMO VIEWER TOOLS ........................................................................................................................... 28
4. APPLICATION: “THE ULJANIK ROPAX”........................................................................................... 34
4.1 INTRODUCTION ..................................................................................................................................... 34
4.2 PRESENTATION OF THE ULJ ROPAX ...................................................................................................... 35

4.2.1. Reference design model .............................................................................................................. 35
4.2.2. Considered Loads ........................................................................................................................ 37
4.2.3. Selected design variables and their upper/lower limits .............................................................. 40

4.2.4. Selection of the constraints and their limits................................................................................ 43
4.3 MARS MODEL ..................................................................................................................................... 47
4.4 STEEL MODEL ...................................................................................................................................... 61
4.5 ASSESSMENT-OPTIMISATION RESULTS ...................................................................................................... 65

4.5.1. Optimisation under center of grativity location .......................................................................... 72
4.6 THE DERIVED RESPONSE CURVES FOR WEIGHT AND CG OF THE ULJ ROPAX ...................................... 78
5. CONCLUSIONS ABOUT THE STRUCTURAL OPTIMISATION TOOLS DEVELOPED IN WP4 .......................... 80
4
6. THE FUNCTIONAL TOOLS ................................................................................................................. 80
6.1 COST ASSESSMENT TOOL FOR STRUCTURAL DESIGN .............................................................................. 80

6.1.1. Tool abstract ............................................................................................................................... 80
6.1.2. Tool description........................................................................................................................... 81
6.1.3. Tool status................................................................................................................................... 83

6.1.4. Tool Integration .......................................................................................................................... 83
6.1.5. Tool in Application ...................................................................................................................... 83
6.2 PRODUCIBILITY ASSESSMENT TOOL FOR STRUCTURAL DESIGN ................................................................ 83

6.2.1. Tool abstract ............................................................................................................................... 83
6.2.3. Tool status................................................................................................................................... 85

6.3 COMPOSITE ASSESSMENT TOOL FOR CRUISE VESSEL SUPERSTRUCTURE ..................................................... 85

6.3.1. Tool abstract ............................................................................................................................... 85
6.3.3. Tool status................................................................................................................................... 87
6.4 INSTALLATION OF ARTIC OFFSHORE PLATFORM ....................................................................................... 88

6.4.1. Tool abstract ............................................................................................................................... 88
6.4.3. Tool status................................................................................................................................... 89

7. CONCLUSIONS ................................................................................................................................. 92
8. INDEXES ............................................................................................................................................ 93
8.1 INDEX OF FIGURES ................................................................................................................................. 93
8.2 INDEX OF TABLES ................................................................................................................................... 96
9. REFERENCES ...................................................................................................................................... 96
5
1. Introduction
Holistic ship design comprises the tasks of all needed disciplines from the general
arrangement, the hydrodynamic, machinery, etc. to the assessment of cost and
costumer requirements. The unique issue in HOLISHIP’s holistic approach is to leave
the traditional series connection of the above-mentioned design tasks, the well-
known design spiral, towards the use of a connection in parallel. This approach is
used in the very early design stage, the concept design, as well as in a more
sophisticated stage, the contract design phase.
Structural design tasks have to fulfil in this manner several functions: from giving rule
accordant scantlings and layout to the structural weight estimation. The more
advanced the design is, the more information are needed as well from the structural
analyses, touching new materials, noise vibration, productions issues, etc.
Furthermore, by using optimisation processes, the structural assessments need to be
done in a parameterised way allowing for batch-mode operation. But not only ship
design is triggered in the scope of the HOLISHIP project. Also, the structural analyses
of specialised offshore platforms are part of the scenarios.
This public deliverable gives an overview of the structural tools and their use in the
industrial use scenarios, the so-called application cases during the concept design
phase. Here, the level of information is low and based on experience and owner
requirements as well as specifications. The main challenge in this design phase is the
limited time to create feasible designs. The aim of HOLISHIP is to use the new design
methodology to enable a faster design procedure leading to more accurate as well
as more variants of designs.
The tools developed can be grouped into structural and functional tools. Structural
tools serve the basic tasks of determining scantlings, spacings, etc. of main structural
members of a ship. Here, they are connected to a parameterised model as well as
to an optimisation algorithm finding a feasible structural design according to the
rules of the classification society (here Bureau Veritas [BV]). The latter group
comprises tools made to assist or add information to the design of ships and
structures. This includes information about the production complexity of current
designs, the use of composite material as well as the structural concept and erection
of specialised offshore platforms.
The toolsets comprise functional and structural assessment and optimisation tools. This
deliverable will focus on the structural tools as this is the major part of the work done
in HOLISHIP. Here, an integrated structural optimisation workflow has been
developed in order to conduct the actual design of relevant application cases (i.e.
RoPAX and Double-Ended Ferry) for effective least steel weight, through HOLISHIP
concept design integration platform, CAESES®.
6
2. The Structural Assessment Tool
An integrated structural assessment-optimisation workflow has been developed in
order to perform automatic scantling assessment of the HOLISHIP application cases
(i.e. RoPAX and Double-Ended Ferry), targeting a reduced weight of the hull (steel).
Figure 2.1 presents the integrated workflow, and its interface with HOLISHIP concept
design integration platform (with CAESES® environment).
Figure 2.1. Integrated structural design assessment-optimisation workflow and its

interface with HOLISHIP concept design integration platform.
The prime objective is to perform a rule based structural design assessment-

optimisation of ship for effective least steel weight, considering 2D midship cross
section. For the contract platform (Task 4.3), the methodology will be extended to a
series of sections distributed along the ship.
The calculation of the steel weight (as objective function) is based on the weight
(tonne/m) of a series of transverse sections (including the mid ship section) and then
integrated along ship’s length. On this subject, more details can be seen later in
Section 3.14.
7
Considering this objective, a number of existing tools along with their new
script/batch-mode developments (namely MARS2000® and STEEL® BV tools and
modeFRONTIER®) as well as new in-house tools/modules developments (e.g.
Scantling-Spacing Updater, Rule Violation Indicator, Constraint Module and
Weight/CG Calculator) have been integrated under an automated iterative routine.
More details on the methodology and on the individual tool’s description are given
in the Section 3.
3. PRESENTATION OF THE TOOLS, DEVELOPED AND USED IN TASK

4.2 (CONCEPT STAGE)
3.1 General methodology
In concept design phase, a rule-based methodology is considered to perform the
structural design assessment-optimisation of HOLISHIP application cases.
This assessment-optimisation is carried out on a series of 2D ship sections covering the
entire ship length. In concept phase, the assessment is based on the midship section
(using MARS) but the weight calculator used a series of cross sections (see Section
3.14) provided by the Yard/Designer, the midship section is modelled with the
MARS2000® and STEEL® BV tools, which respectively include plating, longitudinal
stiffeners, and transverse frames (longitudinally stiffened structure). These two models
are considered as the reference midship structural designs for the future structural
assessment-optimisation process.
Then, MARS2000® is called to assess the hull girder, plating and longitudinal stiffeners,
considering the appropriate BV rule-based loads (e.g. hull bending moment,
pressures on hull - see Section 4.2.2). On one side, the assessment output results are
used by the in-house “Rule Violation Indicator” tool (Section 3.8) to check the
feasibility of the structural design with respect to BV rule requirements (global and
local). On the other side, the structural weight of the considered transverse section
(tonne/m) is transferred to the “Weight/CG Calculator” in-house tool (Section 3.14).
This calculator uses the different transverse sections along ship’s length in order to
assess the steel weight of ship (based on the weight assessed for the midship
section).
For each new configuration (for instance with modified ship length, beam, frame
spacing, deck height, …), but also for any changes of the stiffener spacing (which is
design variable of the scantling), the “Scantling-Spacing Updater” in-house tool
(section 3.7) updates the midship section structural layout and exports it into
MARS2000® tool for next iteration (and assessment).
The above-described iterative process is steered by modeFRONTIER® together with
some new script/batch-mode developments under a seamless automated routine.
In the automated assessment-optimisation loop, the steel weight is the objective
function to be minimised (eventually constrained by the vertical position of the
gravity centre for stability reasons). The optimisation is of course restricted by the
requirements/constraints imposed by the BV Class Rules and Yard/Designer. The
8
values of the design objective function and the constraints are used by
modeFRONTIER® at each iteration to generate new sets of design variables (plate
thickness, longitudinal stiffener spacing and sizing, and frame scantling; the
transverse frame spacing being considered as constant). The stiffener scantlings are
taken for the catalogue of profiles of the yard. This iterative process continues until a
set of suitable alternative midship structural designs are obtained.
Depending on the expected level of accuracy (linked to the convergence of the
optimisation process) and the available resource (hours), a set of best
feasible/optimal solutions are selected. Then, for each of these best feasible
solutions, the STEEL® BV tool-based assessment-optimisation workflow is called to
determine the adequate frame scantling (again corresponding to a feasible design
and a reduced weight). The purpose of this process is to perform assessment-
optimisation on the midship transverse frames, considering only the frame scantling
as design variables (web and flange dimensions).
The iterative process within STEEL®-based workflow is similar to the MARS2000®-based
workflow, which was earlier described. With STEEL®-based workflow, the targeted
structural weight concerns the weight of the transverse frames.
By launching sequentially MARS® and STEEL®-based workflows (for various sets of
fixed design parameters such as Lpp, B, T, Cp, transverse frame spacing, … which are
given by the HOLISHIP design platform - called here after a configuration), a series of
feasible-optimal designs are obtained with respect to BV Class Rules and
Yard/Designer requirements.
From these results, we derive the relations between the weight (and the gravity
centre) and a series of parameters (frame spacing, ship beam, ship length and deck
height...). These relationships are approximated models of the real behaviour, but
enough accurate to perform further optimisation. They are “response surfaces - RSM”
or “meta models” - see Section 3.6 and Section 4.6).
Finally, these approximated models (RSM) of the steel weight and centre of gravity of
the ship hull are transferred/implemented on the HOLISHIP platform (WP7) to be used
for the future HOLISHIP application cases.
For further use and optimisation, the final set of feasible-optimal midship structural
designs are used by the “Generator of Producibility Indexes” in-house tool to
generate producibility measures or indexes for Producibility Assessment (see section
3.15).
9
3.2 List of TOOLS
Existing tools/software along with its new script/batch-mode developments are:
• MARS2000® BV software - section 3.3

• STEEL® BV tool - section 3.4
• modeFRONTIER® - section 3.5
Major in-house developments are the followings:

• Coupling of the WP4 tools
and the WP7 integration Platform - section 3.6
• Scantling-Spacing Updater Tool - section 3.7
• Rule Violation Indicator Tool - section 3.8
• Stress Assessment Tool - section 3.9
• Thickness Corrosion Updater Tool - section 3.10
• Steel Frame Weight Assessment Tool - section 3.11
• Constraint tool - section 3.13
• Weight/CG Calculator Tool - section 3.14
• Generator of Producibility Indexes - section 3.15
• Demo Viewer tools - section 3.16
3.3 MARS2000® BV software

MARS 2000® was chosen to be used in the HOLISHIP project for the assessment of the
steel weight and in parallel the compliance with the rules. Mars2000® is a BV
structural analysis software focusing on the longitudinally scantling assessment
plating and longitudinal stiffeners. This software provides rule based calculation and
checking of the scantlings of any cross-sections according to Bureau Veritas Rules
and IACS Common Structural Rules. The frame spacing is considered by the
assessment of the frames is not included in MARS.
• For a transverse section of the ship (midship section), MARS2000® calculates:
o The geometric properties
o The hull girder strength criteria
o The hull girder ultimate strength
Check the respect of the local scantling rule (plates and stiffeners)
For more details, reference may be made to [2].
Figure 3.1 shows a typical MARS model - “scantling check”
10
Figure 3.1. Mars2000®, View of a typical MARS model - “scantling check”
The characteristics that allow the integration of MARS2000® into the optimisation
iterative process of HOLISHIP are:
• its ability to execute strength assessment (based on rules) and provide results
in Batch Mode;
• the existence of a Neutral Transfer file making possible the modification of the
values of the design variables and the automatic generation of the
corresponding MARS database.
Neutral Transfer file
It is an XML extension file that follows a specific XSD format . It contains the
information of a MARS database and it is used as a means to transfer database
information between BV structural software, but also other commercial software (ex.
NAPA Steel). This function facilitates the fast production of MARS models when
imported from other design software. The Transfer File had to be improved in the
scope of the HOLISHIP project in order to contain the full set of information needed
to become an automatically computable MARS2000® database.
Batch Mode
Mars2000 can execute automatic calculations without the user’s intervention by
reading directly a Mars2000 database (.ma2) or by accessing directly the XML file
(converting it automatically to .ma2) and it can provide results in a text format
covering BV Rules or CSRH criteria.
11
This mode was primarily used only by BV for internal purposes for the testing of the
software, so it had to be adapted for the Holiship needs.
This tool is ready to be used in the scope of HOLISHIP WP4-concept phase.
3.4 STEEL® BV tool

BV STEEL® [3] is a BV structural software that performs 3D Beam structure calculations
used for the direct strength assessment of primary supporting members (frames)
according to BV Rules requirements.
As for MARS2000®, the STEEL database can be directly translated in TXT format and
modified outside of the STEEL software (tool). The STEEL tool has now the capability to
be launched in batch mode (this was developed in the framework of HOLISHIP) to
generate the STEEL database and produce the results of the assessment.
The results show the respect (or not) of the stresses with regards to the allowable
stresses defined in the BV Rules. The BV Rule loads are not included in the STEEL tool,
so the loads need to be applied manually in the model (hopefully to be done once
for a configuration).
In the framework of HOLISHIP, BV STEEL® completes MARS® 2000, which does not
consider the frame (transverse beam) assessment. So to check the compliance of
the frames to the BV rules, we need to use BV STEEL®.
The principle is to model a 2D structure (even if BV STEEL® allows to model 3D)
composed of all the transversal elements of the midship section, including the
plating (which is the associated plates to the frames (see Figure 3.2.a and Figure
3.2.b).
Figure 3.2.a. Typical 2D model of the frames at midship section, using BV STEEL®
12
Once the model is created, one can define different types of beams such as
H, T, I sections, rectangular box girder, u section, cylindrical pipe & bar, bulb
plate, corrugation as well as composite sections.
In order to model the boundary conditions of the structure, the degrees of
freedom for each node have be defined, such as free node, fixed node,
forced deformation and elastic support. In addition to the weight of structure,
beam loads, hydrostatic loads, thermal loads and loading cases can be
defined by the user.
Figure 3.2.b. shows the view of a typical STEEL model.
Figure 3.2.b. STEEL®, View of a typical STEEL model – “beam types”
Once the computation is performed, deformations and rotations for each node and
deformation, local shear forces, bending moments, normal and shear stresses for
each beam are obtained.
13
The output file generated by STEEL contains for each beam the results in a TXT file.
Beam bending stresses and shear stresses are given at predefined points on a cross
section of the beam in this output file.
3.5 modeFRONTIER®
modeFRONTIER is a Multidisciplinary Design Optimisation (MDO) platform developed
by ESTECO SpA [4]. Its workflow based environment, and multi-objective optimisation
algorithms are used for streamlining the engineering design process to cut time and
cost while obtaining improved results.
ModeFRONTIER encompasses tools for process automation and integration, design
space exploration, multidisciplinary optimisation, RSM modelling and virtual
optimisation, robustness and reliability. modeFRONTIER includes a set of advanced
optimisation algorithms covering deterministic, stochastic and heuristic methods for
both single and multi-objective problems. The platform comes with analytics and
visualization tools that facilitate decision making.
ModeFRONTIER has three environments, each managing one step of the optimisation
process. They are called Workflow Editor, Run Analysis and Design Space.
As an example, the integrated MARS-based assessment-optimisation development,
steered by modeFRONTIER, is shown Fig 3.3 in the Workflow Editor environment.
Figure 3.3. View of the MARS-based development integrated in the Workflow Editor
environment of modeFRONTIER®
Once, the workflow is launched, the following process is launched:
1. modeFRONTIER sends input values (e.g. plate thickness, stiffener
spacing/sizing) to the WP4 integrated structural simulation software;
2. The simulation software uses these input values to compute the outcomes of
the system (e.g. structural weight);
3. modeFRONTIER extracts these output values and saves them in its “Designs
Table” (Design Space);
4. The entire optimisation process is driven by an optimisation algorithm that
searches for the optimal combination of input values to reach the objective(s)
(structural weight). Here, the chosen algorithm is the “Non-Dominated Sorting
14
Genetic Algorithm II (NSGA-II)” selected from the library of optimisation
algorithms of modeFRONTIER®.
At each iteration, modeFRONTIER generates a new set of design variable
values by using the selected algorithm.;
These steps are repeated for each iteration.
The obtained results can be visualized/analysed within design space environment of
modeFRONTIER (see Section 4.5, Figure 4.26).
There are many tools in the Design Space environment of modeFRONTIER as some
RSM tools. If the user has limited computational resources, he/she can use these RSM
tools to train, compare and validate metamodels in order to speed up the
optimisation process.
For more details concerning modeFRONTIER are available in [4].
3.5.1 RSM (Surrogate model or metamodel)

Surrogate model or metamodel is an engineering method that is used to
approximate multivariate input-output behaviour of complex systems based on a
limited set of computational simulations. Surrogate models represent the underlying
simulation system as closely as possible while being computationally cheaper to
evaluate. They can be used for design optimisation, design space exploration,
parametric studies, etc. One of the most popular surrogate models is polynomial
response surfaces.
Response surface methodology (RSM) can be considered as a collection of
mathematical and statistical techniques for empirical model building. In the
beginning the surface was developed in order to model experimental responses [5].
The method was used to model the numerical experiments.
The purpose of using RSM is to reduce the cost of expensive analysing methods such
as CFD analysis, finite element analysis and holistic optimisation (as for HOLISHIP
project).
Using response surfaces in structural optimisation helps to deal with software
integration problems related to multidisciplinary optimisation approach. Using the
Response Surface Method, the optimisation process is performed in an economical
way with optimum solutions and it gives the best relationship between design and
target variables using polynomial regression techniques.
RSM is usually used for determining objective function values for some design data
range where the experimental or actual optimisation results are not available.
Furthermore, in addition to the traditional response surface approach, kriging
method was analysed to generate surrogate models. Kriging interpolation model
requires the following steps respectively: data analysis, choice of a trend function,
examining the standard error in order to choose a correlation function and
estimation of correlation parameters.
15
In the case of WP4, as the used dataset was simple (only 1 parameter), the curve
fitting feature in Excel was enough to get the weight and gravity centre elevation
response surfaces.
3.6 Coupling of the WP4 tools and the WP7 Integration Platform (concept
design)
Figure 3.4a shortly describes how the WP4 structural assessment-optimisation model is
coupled with HOLISHIP concept design platform (WP7), in both directions, i.e. from
WP7 towards WP4 and opposite. The aim is to determine automatically a feasible
structural design associated to, as much as possible, a reduced steel weight and
keeping the CG in the acceptable range.
Data flow from WP7 Integration Platform towards WP4 tools
The WP4 structural assessment-optimisation workflow needs to obtain from the WP7
Platform the required input data for each configuration, this mean for each hull
form/GA design (keeping the topology fixed compared to the reference design
given by Yard).
These data are:
• Midship internal layout, including the coordinates of the section nodes (y=
horizontal position of node, z = vertical of node position) and also the main
ship characteristics. Figure 3.4.a shows an example of internal layout including
nodes as well as a text file containing the node coordinates and the main
characteristics.
• A text file (Figure 3.4.b) containing sets of nodes with coordinates (x, y, z), i.e.
x= longitudinal position of section, y= horizontal position of node, z = vertical of
node position. There are typically 21 sections including the midship section.
Once HOLISHIP design platform (CAESES®) exports these data, the longitudinal
stiffeners are automatically generated on the 2D internal layout. Then, this updated
layout is incorporated into the MARS® workflow for the structural assessment-
optimisation of plating and longitudinal stiffeners, and later in the STEEL®-based
workflow for the structural assessment-optimisation of transverse frames.
16
Figure 3.4.a. Template for WP7-WP4 coupling
17
Figure 3.4.b. Cross section nodes’ coordinate along ship length
Data flow from WP4 tools towards WP7 Holiship Platform

For each configuration (hull form/GA, internal layout, L, B, T, c b as well as frame
spacing …,) generated in the HOLISHIP platform CAESES®), the WP4 structural
assessment-optimisation workflow is launched and a series of feasible-optimal
solutions are generated. Then, the steel weight and the centre of gravity are
exported to an output database.
This procedure is done for each configuration of the hull form/GA design.
In this way, for each new configuration, the WP4 output database is enriched in
order to later derive a surrogate models (RSM) of the steel weight and the position of
the centre of gravity, which will be used later for the Holiship application cases.
18
3.7 Scantling-Spacing Updater Tool
The Scantling-Spacing Updater tool is capable of portraying the changes in the

structural design parameters (i.e. plate thickness, stiffener scantling and stiffener
spacing). Figure 3.5 shows the automatically generation of stiffeners on the internal
layout of the structure (RORO).
The tool uses the 2D cross-section scantling design - plating and longitudinal stiffeners
(XML) as an input and updates it based on the changes in the structural design
parameters.
The updated file is read by MARS2000® BV tool for further process during the
structural assessment-optimisation loop.
The tool has been successfully developed and tested and is ready to use.
Figure 3.5. Automatically generation of stiffeners on the internal layout (ROPAX)
Figure 3.6.a shows the change of stiffener profile (taken for the Yard catalogue) for
deck 5 (Figure 4.30) of the ULJ RoPAX during the iterations of the fully automatic
Examples concerning the changes of plate thickness and stiffener spacing of the
deck 5 (Figure 4.30) are given in Figure 3.6.b and Figure 3.6.c.
19
Figure 3.6.a. History of stiffener size (Deck 5) during the iterative process (MARS2000®)
Figure 3.6b. History of plate thickness (Deck 5) during the iterative process
(MARS2000®)
20
Figure 3.6.c. History of stiffener spacing (Deck 5) during the iterative process
(MARS2000®),
3.8 Rule Violation Indicator Tool
The Rule Violation Indicator tool indicates whether BV rule requirements are fulfilled
or not. This means the global hull girder strength criteria such as section modulus at
deck and bottom, and the local strength criteria such as yielding and buckling of
plates and stiffeners.
The tool takes as input the output file (TXT) produced by MARS2000® BV and
indicates -through text files- the full list of the BV rule requirements which are violated.
Figure 3.7 shows the number of violations of the BV rule at each iteration. For
instance, the reference design (provided by ULJ), used as initial point (ID000), is
infeasible and is associated to 26 violations of the BV rule.
On contrary, the designs ID1600 and ID2959 are feasible and respect the BV rule
requirements (see Figure 4.27 to Figure 4.29).
21
Figure 3.7. History of MARS2000® rule violation by Rule Violation Indicator
The module has been successfully developed and tested and is ready to use.
3.9 Stress assessment tool

The Stress assessment in-house tool is capable of capture the maximum Von Mises
stresses (given by STEEL® BV) for a given load case. This maximum stress is then
compared to the allowable stress defined by the BV Rules.
As example, the outcomes of the stress assessment tool are shown Figure 3.8 (see
also Figure 4.27 to Figure 4.29).
This operation takes approximately 0.6 seconds per iteration.
Figure 3.8. Evaluated and violated rules output TXT file

The tool has been successfully developed and tested, and is ready to be used.
3.10 Thickness Corrosion Updater Tool

The Thickness Corrosion Updater in-house module uses the corrosion margin in
order to re-calculate the net scantling (or gross scantling). In BV STEEL, the stress
assessment is performed using the net scantling approach. However, the plate
thickness values provided by MARS2000® include the corrosion margin. Therefore, to
22
run the stress assessment tools with net scantling, corrosion margins should be
removed.
Corrosion margins are considered to obtain the total weight of the model.
This operation takes approximately 0.5 seconds per iteration.
The module has been successfully developed and tested, and validated version is in
place. The module might need to be adjusted due to a possible future amendment
in BV class rules (as the values of the corrosion margin imposed by BV).
3.11 Steel Frame Weight Assessment Tool

The purpose of the Steel Frame Weight Assessment in-house tool is to calculate
the gross weight once the net scantling is updated with corrosion margins. Previously
it was calculated by STEEL® BV batch mode; however, in order to save 3 seconds per
iteration this tool was developed and it takes approximately 0.2 seconds per
iteration. The input file related to the frames and the results are delivered in a TXT file.
The module has been successfully developed and tested, and is ready to use.
3.12 Design Variables and Parameters

In concept design phase, the design variables and the parameters used in the
optimisation (automatic scantling procedure) are the followings:
• The global design variables (main ship’s characteristics), which are length
between particulars (Lpp), breadth (B), draft (T), free board deck height, deck
clearance, block coefficient (c b ), bending moment, max service speed and
main frame spacing. These are parameters and are provided by the HOLISHIP
design platform (CAESES®). These parameters, for each configuration (hull
form/GA design) provided by WP7 platform, are kept constant during
structural assessment-optimisation process in WP4. Of course, we can change
them between different configurations. Each configuration corresponds to a
different feasible design, associated to a weight (T/m) and a gravity centre
position.
• The local design variables, considered by the WP4 structural assessment-
optimisation workflow are: plate thickness, stiffener spacing, stiffener sizing
(scantling), frame web height, frame web thickness, frame flange width and
frame flange thickness. They are uses as design variables of the optimisation
process.
23
3.13 Constraint tool
This constraint tool considers the various constraints (technological, geometrical) and
equality constraints (for standardisation of the structure as an equal spacing of
stiffeners in various decks) to be considered by the structural assessment-optimisation
loop.
The 2D cross-section scantling design - plating and longitudinal stiffeners (modelled
with MARS) as well as frames and their attached plates (used in BV-STEEL) - are
considered as inputs and associated with the constraints linked to these modelled
structural elements.
As an example, Figure 3.9 shows the output of “geometrical constraint tool”.
Based on the inputs given in the middle part (Blue frame in Figure 3.9), the evaluation
is done for two typical geometrical constraints, i.e. flange thickness of stiffener must
be less than two times of web thickness of stiffener, and plate thickness must be less
than two times of web thickness of stiffener.
The results can be seen in the two first lines (Black frame in Figure 3.9) that the
constraint is satisfied ([1.000000]) while a latter evaluation is failed ([0.000000]). In this
example one constraint is not respected.
The data in the Red frame shows the evaluation for the entire midship section
structure. It should be noted that “Rules” here is referring to geometrical rules which
may come from production for instance.
The module has been successfully developed and tested, and is ready to use.
Figure 3.9. Typical output of the geometrical constraint tool (for test purpose)
24
3.14 Weight/CG Calculator Tool
The Weight/CG Calculator tool estimates the steel weight and the centre of gravity
of the RoPAX, using the data provided by the shipyard and the HOLISHIP concept
design platform.
This tool needs hull form data (e.g. extracted from NAPA, Maxsurf,..), the weight
(tonne/m) corresponding to midship section, and the weight of the bulkheads and
superstructure.
The input data has been taken from the initial design provided by Uljanik.
Subsequently, the tool utilises the same data with a factor accounting for the
change in dimensions (section area of the different sections compared to the
midship section). Then, the tool generates the hull steel weight of the entire ship. This
tool is used by the optimisation routine to assess the weight at each iteration.
The hull form data is provided by a file containing the nodes and their coordinates (x,
y, z). The weight estimation tool uses the 21 section areas and the midship section
weight, along with bulkhead and superstructure weight, to evaluate the total hull
weight. This is accomplished by evaluating the weight of each section as a
percentage of the weight of the midship section (this percentage corresponds to the
ratio of the considered sectional area to the midship sectional area). After
evaluating the weight (tonne/m) of all sections (and the bulkheads and
superstructure), the total weight of the hull is evaluated through an integration along
the ship length. The schematic of the algorithm is shown in Figure 3.10.
Figure 3.10. Schematic overview of the weight estimation algorithm, where; 𝑨𝒎 =area
of midship section, W m = Weight of midship section, A FX =area of x section fore of
midship, W FX =weight of x section fore of midship, w(b) = weight of bulkheads.
The “weight/CG calculator” tool has been used for the benchmark performed on
the ULJ ROPAX and has evaluated the weight, Xcg and Zcg of the ship with margins
of 0.5% for weight, 0.2% for Xcg and 5% for Zcg.
25
3.15 Generator of Producibility Indexes
The Generator of Producibility Indexes in-house tool provides the required input
data to determine the producibility indexes (e.g. longitudinal stiffener spacing to
plate thickness, transverse frame spacing to plate thickness, …). These producibility
indexes are important for the shipyard to assess the production complexity, cost and
time.
The tool uses as inputs the complete scantling designs obtained through the
structural assessment-optimisation process. It generates the expected producibility
measures.
Figure 3.11 shows an example of the output generated by the tool for ULJ RoPAX for
determining the producibility index. Parameters are according to the sketches
below.
Figure 3.11. Outcomes of the Generator of producibility indexes
26
27
3.16 Demo Viewer tools
The Demo viewer shows the results obtained during optimisation process. It takes as
inputs the results (as excel file) stored by modeFRONTIER® and images of the results
(BMP) generated by MARS2000® BV tool and it shows visually the structural
Figure 3.12 presents the GUI of the Demo viewer with its sub-tools, i.e. the Graph
Viewer, the 2D Viewer and the Picture Viewer.
There are screens (Figure 3.12.b, c and d) taken from MARS on which the user can
choose the results to be displayed. They show the results of the hull girder and the
local strength of stiffeners and plates (Picture Viewer).
In Figure 3.12, the optimisation history (Graph Viewer) is also shown at top right (zoom
is given in Figure 3.12 a), and the display of the 2D Viewer (at bottom left – see zoom
at Figure 3.13 or Figure 3.14).
The layout/arrangement of the Demo viewer can be changed by the user to meet
their needs.
The user can use the timeline slider (in the bottom part of the screen Figure 3.12) in
order to see the optimisation process (iterations).
The 2D Viewer (Figure 3.13 and Figure 3.14) was developed in order to display the
geometrical characteristics of the transverse cross section during the optimisation
process. The 2D viewer has the capability to read a MARS XML file or a STEEL STW File.
The 2D Viewer is based on a "common model (MARS and STEEL)" of the transverse
cross section of the ship. This common model includes the lists of nodes, segments
(plate between 2 nodes) and stiffeners that are considered.
Figure 3.13 shows plate properties and Figure 3.14 the plate thicknesses with colour
and line style.
The user of the 2D Viewer can choose and filter the information he wants to display.
For instance, if the user selects a segment of the MARS model, the thickness and its
associated characteristic (plate thickness, minimal/maximal/ stiffener spacing, web
thickness, …) are displayed.
The DEMO Viewer tool is fully operational.
28
Figure 3.12. Demo Viewer
Figure 3.12.a. Graph Viewer
29
Figure 3.12.b. Check ratio - Hull girder strength
Figure 3.12.c. Check ratio – Local strength - Stiffeners
30
Figure 3.12.d. Check ratio – Local strength - Strakes
31
Figure 3.13. 2D Viewer showing plate properties
32
Figure 3.14. 2D Viewer showing plate thickness with colours and line styles
33
4. APPLICATION: “THE ULJANIK ROPAX”
4.1 Introduction
In the framework of HOLISHIP, the WP4-structure team had the opportunity to work
on several study cases lead by shipyards (i.e. RoRo and RoPAX).
On one side, although that required high effort, it helps the structural assessment-
optimisation workflow to be developed towards a more generic approach.
On the other side, due to the lack of support for the Uljanik [ULJ] shipyard (not more
active since June 2018), which proposed the RoPAX study case (WP16), the WP4-
structure team did not have the opportunity to enough interact with the ULJ RoPAX
designers and get benefit of their experience and of a critical view of the results,
neither advises and expectations for the integration within the HOLISHIP concept
design platform.
Nevertheless, in this report, the focus is done on the ULJ RoPAX, which was used as
internal benchmarking for validation purpose; validation performed for the
development of the optimisation procedure.
34
4.2 Presentation of the ULJ RoPAX
4.2.1. Reference design model

In order to present and validate the integrated structural assessment-optimisation
tool (concept phase), the ULJ RoPAX has been taken as support.
The considered ULJ RoPAX (reference case) is based on the 2D midship section
drawing shown on Figure 4.1.
Figure 4.1. Midship section of the ULJ RoPAX , used as reference design
The characteristics of the ROPAX are given in Table 4.1.

The RoPAX is operating with unrestricted navigation, and subjected to a maximum
still water bending moment of 1500 MN.m.
35
Table 4.1. Characteristics of ULJ RoPAX, used as reference design
Design parameters Values
Main Characteristics Overall length (Lpp) 217 m
Breadth 32.2 m
Depth to main deck 9.65 m
Speed 24 Knots
Displacement 28372 m3
Scantling and spacing Plate thickness 11.5 mm

(for Deck 5 as an
Stiffener spacing 600 mm
example)
Stiffener profile HP200x10
(mmxmm)
Frame web height 950 mm
Frame web thickness 11 mm
Frame flange width 400 mm
Frame flange thickness 19 mm
Steel weight 12043 tonne
Vertical centre of gravity 13.15 m
Note that the 4 upper decks (Figure 4.5) as considered by ULJ as the
superstructure. ULJ asked not to touch at the superstructure and to limit the
optimisation to the main hull (all decks below the superstructure - Figure 4.5).
36
4.2.2. Considered Loads
As mentioned before, the WP4 structural assessment-optimisation workflow is BV-rules
based and is planned to be used by the HOLISHIP Platform (WP7) for the application
cases (as WP16- RoPAX).
In this respect, the loads to be applied on the ULJ RoPAX vessel are those defined in
the BV rules for steel ships, based on a combination of ship loading conditions (still
water level) and sea load cases (wave loads). The ship loading conditions used in
this example are:
• the full load condition for the assessment of all structural elements;
• the ballast condition for the assessment of ballast tanks.
For these loading conditions, the global hull girder loads and the local loads and
pressures (sea and cargo) are calculated for the following sea load cases:
• load case “a” (positive and negative) for maximum sea pressure in upright
condition combined with maximum vertical hull girder loads;
• load case “b” for maximum longitudinal and vertical acceleration in upright
condition combined with maximum vertical hull girder loads;
• load case “c” for maximum sea pressure in inclined ship condition combined
with horizontal and torsional hull girder loads;
• load case “d” for maximum transverse and vertical accelerations in inclined
ship conditions combined with horizontal hull girder loads;
More details on this subject can be found through [1].
In the STEEL® BV tool, the structural analysis of the frames requires the load model
(not automatically defined in the tool). Sea pressure loads as well as deck loads are
based on the BV rules.
Sea pressure loads consists of loads due to still water pressure and save pressure.
As mentioned in BV Class rules NR 467, Pt B, Chapter 5, Section 5, still water
distribution is given at Figure 4.2.
Figure 4.2. Still water pressure distribution
37
Wave pressure loads are calculated according to BV Class rules NR 467, Pt B,
Chapter 5, Section 5 for each load case.
The considered wave pressure loads are divided into two main categories, which are
load cases at upright conditions (Load case a & Load case b) and load cases at
inclined conditions (Load case c & Load case d).
Load case a is considered when the ship is at rest and encountered with a wave
which produces a relative motion of the sea waterline with a height of h1 from the
still waterline (both positive and negative) which is symmetric on both sides. Wave
pressures are calculated by the given empirical formulation as per defined BV rules.
For instance, the illustration of Load Case a is as below Figure 4.3.
Figure 4.3. Wave Pressure at Load Case a

Wave pressure for the load cases c and d are calculated according to the
applicable BV rules and Figure 4.4 shows the wave pressure distribution for load case
c and d respectively.
Figure 4.4. Wave pressure distribution for load case c and d respectively
The deck loads for the wheeled cargo are received from MARS2000® model and
distributed as a uniform load in STEEL® BV model.
38
In addition to the deck loads for the wheeled cargo, local loads for the
accommodation part is defined as per applicable BV rules and modelled as shown
in Figure 4.5.
Figure 4.5. View of the deck loads at the upright ship condition loading case b+
39
4.2.3. Selected design variables and their upper/lower limits
The selected design variables together with its upper/lower limits are given in Table
4.2.
These design variables concern the main hull and not the superstructures composed
of the 4 upper desks.
The ranges of variation of the design variables correspond to the ULJ Yard practice
and experience. For instance, for the web height of the frames, ULJ proposed to
consider 200mm (plus/minus) compared to the original design values.
Table 4.2. Upper and lower limits of design variables (*), imposed by the Shipyard
(unit is in mm).
Design variables Lower limit Upper limit
Plate thickness 6 17
MARS-based workflow Stiffener spacing 400 700
Stiffener profile HP 80x5 HP430X21
Frame web height 750 1150
Frame web thickness 9 13

STEEL-based workflow
Frame flange width 300 500
Frame flange thickness 14 24
(*) Based on the agreement with ULJ yard, the design variables only concern to the hull part
(not to the superstructure - upper decks).
In this respect, 92 design variables are considered during the WP4 assessment-
optimisation process, in which 36 and 56 design variables are respectively used in
MARS2000-based and in STEEL-based workflows.
The frame design variables are composed of the scantlings of the beam sections in
order to be optimised with respect of the constraints. The STEEL optimisation loop is
established with 56 design variables, which are the scantlings of all the deck and side
frames. The range of design input variables are selected in accordance with the
initial data (reference design – Figure 4.1). As example Table 4.3 gives a series of
design variables with their lower and upper bounds.
40
Table 4.3. Upper/lower bounds of the Design variables associated to the frames (BV
STEEL)
Figure 4.6 shows the variations of the web thickness (for Deck 5) during the automatic
iterative process. We see that these design variables remain between the upper and
lower limits (9 mm and 13mm).
Figure 4.6. Changes of web thickness of Deck 5 within a fully automatic iterative
process
Once the structural strength assessment of the plating is performed by MARS®-

modeFRONTIER® workflow, the optimum thickness of these plates is frozen and used
in STEEL® modeFRONTIER® optimisation workflow to define the frame scantling.
41
Therefore, the attached plate thickness values are incorporated as constant values
in STEEL.
As examples, Table 4.4 gives the optimised plate thicknesses in millimetres (i.e. deck
plate thickness, side shell thickness etc.) given by MARS2000®/modeFRONTIER® and
used by STEEL® modeFRONTIER® workflow as constants.
Table 4.4. Plate thicknesses used in STEEL® modeFRONTIER® workflow
42
4.2.4. Selection of the constraints and their limits
In addition to the technological constraints received from ULJ yard (see Section 4.2.3
- Table 4.2), the following requirements/constraints from the BV-rules are taken into
account during the structural assessment-optimisation process.
In Mars2000®;
• For the Hull girder strength,
- Minimum cross section inertia
- Minimum cross section modulus at deck and at bottom
- Allowable normal stress at any point of the cross section contributing to
the longitudinal strength (plating and stiffeners)
- Allowable shear stress at any point of the cross section (plating)
• For the plating elements,
- Yield thickness criteria under local load;
- Yield thickness criteria under testing pressure (if applicable);
- Minimum technological thickness criteria (5 mm gross);
- Allowable buckling stress;
- Allowable shear buckling stress.
• For the stiffeners (longitudinals),

- Yield section modulus under local load;
- Yield section modulus under testing pressure (if applicable);
- Yield shear section under local load;
- Yield shear section under testing pressure (if applicable);
- Allowable buckling stress;
- Allowable ultimate strength stress;
- Minimum technological web thickness (5 mm gross);
- Proportional criteria for web and flange (geometrical constraints).
43
In STEEL® BV tool;
The constraints concern:

- The allowable Von Mises stress at any point of the analysed frames
(beams),
- The geometrical constraints;
- The frame deflection.
The WP4 optimisation workflow considers geometrical constraints and the
upper/lower bounds of each design variable.
The geometrical constraints relate to frames scantlings and attached plate
thickness (defined in BV Rules NR 467, Pt B, Ch4, Sec3) and they are
implemented in the STEEL® modeFRONTIER® optimisation workflow.
As examples, the 3 following formulations are considered, and each design
variable generated in modeFRONTIER® takes into account of these
𝐴𝑎 ≥ 𝑡𝑓 𝑏𝑓
ℎ𝑤
≥ 10
𝑡𝑝
ℎ𝑤
≥ 10
𝑡𝑓
with:
A a : Net sectional area of the attached plating in mm2
h w : Web height, in mm, of an ordinary stiffener or a primary supporting
member
t p : Net thickness, in mm, of the plating attached to an ordinary stiffener or a
primary supporting member
t f : Net face plate thickness, in mm, of an ordinary stiffener or a primary
supporting member
b f : Face plate width, in mm, of an ordinary stiffener or a primary supporting
member
Before starting the iterative procedure, geometrical constraints are verified by
modeFRONTIER®.
Once the iterative procedure starts, the deflection constraint will be applied by
modeFRONTIER®. The allowable deflection constraint is applied to the cargo decks
as 5 millimetre per metre as per applicable BV rules. Therefore, depending on the
requirements from client, user can modify the deflection constraint in the
modeFRONTIER® optimisation workflow.
44
PRODUCTION CONSTRAINTS
As example, Figure 4.7 shows the values of a geometrical constraint (ratio of web
height to the flange thickness) during the iterative process.
We see that the expected ratio (Hw/Tf > 10) is always verified in the case.
Figure 4.7. History of a geometrical constraint (ratio of web height to the flange
thickness > 10)
45
VON MISES STRESS
The WP4 structural optimisation workflow has as objective to minimise the weight of
entire model but also to respect all the constraints.
Therefore, one condition is to respect the allowable Von Mises stresses in all the
structural elements (plates, stiffeners and frames) considering the material type (MS,
HTS, ...).
As example, considering a yield stress of 355 MPa (AH36) and taking into account of
the material factor and the resistance partial safety factor defined in the BV-rules,
the allowable maximum Von Mises stress is 290 MPa. Therefore (Figure 4.8), any
design having at least one element with a Von Mises stress larger than 290 MPa will
be discarded and the design corresponding to this iteration will be considered as an
unfeasible.
Figure 4.8 shows the optimisation history for a AH36 steel structures (the values
correspond to the maximum stress of all frames of the considered midship)
Figure 4.8. The maximum Von Mises stress history for AH36 high tensile steel (290 MPA
allowable stress)
46
4.3 MARS Model
The ROPAX structural model built in MARS2000® BV tool, based on the data received
from ULJ yard, is shown Figure 4.9. This model refers to the ULJ RoPAX reference
design, used as initial design for the optimisation.
Figure 4.9. MARS model built based on 2D drawing received from ULJ yard (Ref
design)
Details concerning the initial scantling of the structural model of the reference design
(RoPAX) are given in Figure 4.10 to Figure 4.13:
- Figure 4.10 gives the plate thickness;
- Figure 4.11 gives the stiffeners layout and profile types;
- Figure 4.12 gives the material properties. The midship structure is made of mild
steel (ST235) and high tensile strength steel (ST355). The Poisson ratio and
Young’s modulus are 0.3 and 206 GPa respectively.
- Figure 4.13 shows the distributed cargo/wheel loads acting on the cargo
decks (140 kN/m2) and the accommodation decks are loaded as defined in
the BV Class Rules.
47
MARS model (Reference design – ID000)
Figure 4.10. Plate thickness Original MARS model (reference design) -
48
Figure 4.11. Stiffener profile (Bulb) and layout (spacing) - Original MARS model
(reference design) -
49
Figure 4.12. Material type - Original MARS model (reference design)
50
Figure 4.13. Loads on Deck5 - Original MARS model (reference design) -
51
MARS2000 RESULTS for the Reference design (ID000)
Figure 4.14 shows the results of the MARS hull girder strength assessment.
Figure 4.15 gives the stress ratio for local strength of strakes (ratio between the actual
stress and the allowable stress).
Figure 4.16 gives the stress ratio for stiffener strength.
Figure 4.14. Ratio for the Hull girder strength - MARS - reference design (ID000)
52
(max value of 0.8 and average < 0.6)
Figure 4.15. Stress ratio for local strength of strakes - MARS - reference design (ID000)
(max value of 1.0 and average 0.8)
53
Figure 4.16. Stress ratio for local strength of stiffeners - MARS - reference design
(ID000)
(max value of 1.0 and average 0.8)
54
MARS feasible and optimised design (iteration ID 1600)
The following figures give the scantlings of the selected feasible design
corresponding to iteration ID1600, and after the corresponding strength assessment
(stress ratio).
These figures are:

Figure 4.17 gives the plate thickness of the MARS feasible and optimised design, to
be compared with the reference design (Figure 4.10). Table 4.5 presents a
comparison between the two designs
Figure 4.18 gives the stiffeners layout and types of the MARS feasible and optimised
design, to be compared with the reference design (
Figure 4.11). Table 4.5 presents a comparison between the two designs
and
Figure 4.19 shows hull girder strength assessment of the MARS feasible and optimised
design, to be compared with the reference design (Figure 4.14).
Figure 4.20 gives local strength of strakes of the MARS feasible and optimised design,
to be compared with the reference design (Figure 4.15). Table 4.5 presents a
comparison between the two designs
Figure 4.21 presents the stiffener strength of the MARS feasible and optimised design,
to be compared with the reference design (Figure 4.16). There are no major changes
55
Figure 4.17. MARS feasible/optimal design (ID1600) - Plate thickness
(see Table 4.6 and Fig 4.10 for comparison with initial design)
56
Figure 4.18. MARS feasible/optimal design (ID1600) - Stiffener size (bulb profile)
(see Table 4.6 and Fig 4.11 for comparison with initial design)
57
Figure 4.19. MARS feasible/optimal design (ID1600) - Ratio for hull girder strength
(see Fig 4.14 for comparison with initial design)
58
Figure 4.20. MARS feasible/optimal design (ID1600) - Stress ratio for local strength of
strakes
59
stiffeners
60
4.4 STEEL Model
STEEL® has been used for the structural strength assessment of the primary supporting
members (called frames).
This tool used 2D cross section scantling model and produced the results through text
files. Using the graphical interface, main nodes and interconnected beams are
defined manually.
Once the frames are modelled, the loads can be included either as node loads or
beam loads.
As only the half section is modelled, the symmetry boundary conditions are applied
to all nodes belonging to the symmetry plan.
Figure 4.22 shows the distribution of material type used for the frames and
the loads used for the STEEL® model of the Uljanik RoPAX vessel.
Figure 4.22. Material types (left) and loads (right) in STEEL® model
61
Figure 4.23. Beam types in STEEL® model
62
Figure 4.24. Von Mises Stresses in STEEL® model for load case a+
(Initial design ID000)
The maximum Von Mises stress occurs in the flange of Deck3 is 140 MPa.
63
Figure 4.25. Von Mises Stresses in the optimized in STEEL® model for load case a+
(Final design ID 1600)
The maximum Von Mises stress occurs in the flange of Deck3 is 249 MPa.
64
4.5 Assessment-Optimisation results
Based on the previous sections (methodology, tools, reference design and basic
results provided by MARS and BV STEEL), the structural assessment-optimisation
procedure was run for the ULJ RoPAX within the modeFRONTIER® environment, in an
automated-smooth manner. This section presents the results of the optimisation
procedure.
To handle the assessment-optimisation, the Non-dominated Sorting Genetic
Algorithm II (NSGA-II) was chosen (from the library of optimisation algorithms included
in modeFRONTIER®).
First, the MARS2000®-based workflow coupled with modeFRONTIER® was run to
assess and optimise plating and longitudinal stiffeners included in the ULJ RoPAX
midship section.
The iterative process shown in Figure 4.26 has been obtained, through a full
automated process, without any manual intervention using the MARS graphical user
interface.
The total computation time per iteration, using a machine with Bi-Xeon ES-2630 v2
@2.60GHz and RAM 64 GB, is about 8 seconds. This means 3.5 hours to obtain a
feasible-optimal design after 1600 iterations (ID1600). To reach the convergence
(after 3000 iterations), we would need 6.5 hours, but this is not necessary at the
concept stage.
Figure 4.26. History of assessment-optimisation results for MARS-based workflow
Figure 4.26 shows that the reference design ID000) is not a feasible design as it is
associated to 26 violations with respect to BV rules (see details in Figure 4.28).
65
Among the list of violations (Figure 4.28), there is one major violation, which is
associated with global hull girder strength criteria (Figure 4.27).
Figure 4.27. The violation against hull girder modulus for the reference design (from
the MARS GUI)
Figure 4.28. List of violations, identified by the “Rule Violation Indicator” tool for the
reference design (ID000)
66
Figure 4.29.b shows that the MARS-based assessment-optimisation workflow proposes
a feasible structural design (after about 1600 iterations) with respect to BV rule
requirements; design which corresponds also to a least weight solution.
Figure 4.29.a also shows that the main violation obtained for the reference design
(Figure 4.27) does not occur anymore for the proposed feasible solution (ID1600).
Figure 4.29.a. Violation against BV rule requirements, on the feasible-optimal design

(iteration ID1600) - The violation against modulus requirement in MARS GUI.
Figure 4.29.b. Violation against BV rule requirements, on the feasible-optimal design

(iteration ID1600) - List of all violations, produced by the in-house tool “Rule Violation
Indicator”
67
Table 4.5 gives a detailed comparison between the scantling of the reference
design (ID000) and the proposed feasible alternative design (ID 1600), for 3 structural
members:
- Deck 1,
- Side shell,
- Deck 5.
The locations of these 3 members are shown on Figure 4.30.
Table 4.5. Comparison between MARS-based assessment-optimisation results for 3

members of the ULJ RoPAX (dimension in mm) – see Figure 4.30
Design ID ID0 (Ref design) ID1600 Solution
Deck 1 Plate thickness 12 6.5
Stiffener spacing 600 400
Stiffener scantling HP200x10 HP180x9
Lower side shell Plate thickness 10.5 7
Deck 5 Plate thickness 11.5 15.5
Mode details concerning the history of optimisation results for Deck 5 (as an
example) can be seen in Figure 4.35 to Figure 4.37.
68
Figure 4.30. MARS model with three members marked
69
Based on this feasible design (ID1600) obtained with the MARS 2000®workflow, the
STEEL-based workflow is called to perform the assessment-optimisation of the midship
framing associated to its feasible design ID1600.
The total calculation time per iteration in the STEEL-based workflow, using a machine
with Bi-Xeon ES-2630 v2 @2.60GHz and RAM 64 GB, is averagely 5.5 seconds. This
means less than 1.5 hours to obtain the feasible-converged design after 900
iterations.
Figure 4.31 and Figure 4.32 show the iterative process of the optimisation (BV-
STEEL/modeFRONTIER), respectively for the weight and the position of the gravity
centre.
Figure 4.31. STEEL®-modeFRONTIER®; weight history
70
Figure 4.32. STEEL®-modeFRONTIER®; gravity centre history
From all the feasible designs, we selected 2 feasible designs from the STEEL® BV
workflow corresponding to 2 different durations (computation time) - see Figure 4.31.
The first optimized and feasible design is chosen after 253 iterations (ID253). The
weight is 12521 tonnes. The second point (ID number 882) corresponds to a more
optimized design, with smaller weight (12304 tonnes).
The difference between the 2 designs is about 2% of the weight. This difference (2%)
is small for a concept stage and does not justify multiplying the computation time by
3 (882 iterations instead of 253).
71
4.5.1. Optimisation under centre of gravity location
Figure 4.33 shows the optimisation history of the MARS-based workflow, which
includes a constraint on the centre of gravity (CG) elevation. The constraint was that
the CG must be lower than 12.71 m (12.71 m was taken from B 32.2, ID1600).
The figure shows that the optimisation process converges exactly to a CG at 12.71 m
above the keel.
Figure 4.33. STEEL®-modeFRONTIER®; gravity centre history
72
Table 4.6 compares the scantling of the reference design (ID000), which is not
feasible, to the final feasible and less weight design. The comparison concerns 3
locations (see Figure 4.34).
Table 4.6. STEEL-based assessment-optimisation results in detail for several members.

(Unit; weight in tonne and dimension in mm)
Design ID ID-000 ID-253 ID-882
Deck 1 Frame web height 450 420 300
Frame web thickness 7 8.75 4.75
Frame flange width 150 200 100
Frame flange thickness 9 6.75 9.25
Lower side Frame web height 1166 966 966

shell
Frame web thickness 12 13.5 12.5
Frame flange thickness 11 11 11
Deck 5 Frame web height 950 850 750
Frame web thickness 11 10 9
Frame flange thickness 19 15 14.5
Weight 12999 12521 12304
CG 13002 13070 13110
73
Figure 4.34. Location of the 3 areas used for comparison - STEEL®-model of RoPAX
midship
The evolution of the web height and flange (Deck5) during the optimisation process
are shown Figure 4.35 and Figure 4.36, respectively.
74
Figure 4.35. History of the Deck 5 web height of frames - STEEL®-modeFRONTIER®;
Figure 4.36. History of the Deck 5 flange width of frames - STEEL®-modeFRONTIER®;
75
The procedure of transferring the feasible solutions from MARS 2000 ®workflow to
STEEL® BV workflow has been repeated for the optimisation process of each
configuration.
Indeed, we have repeated the procedure for a series of configurations. The
difference between the configurations concerns the ship beam. We considered 3
different beams 29.2 m; 32.2m and 35.2), keeping all the other parameters constant.
The results for the 3 ship beams are plotted in Figure 4.37 (weight) and Figure 4.38
(gravity centre).
Figure 4.37. Response surface of the Hull steel weight versus the Ship beam (B)
76
Figure 4.38. Response surface of the gravity centre elevation versus the Ship beam
(B)
77
4.6 The Derived Response Curves for Weight and CG of the ULJ RoPAX
As explained before, the total steel weight and vertical position of the centre of
gravity assessed by the structural assessment-optimisation workflow are transferred to
CAESES® through approximated models (response surfaces).
In addition of the ULJ RoPAX reference case available on the HOLISHIP design
platform it was necessary to have other ship configurations in order to derive
Response Surfaces for this RoPAX.
For benchmarking purpose, we decided to use the ship beam (B) as the parameter
to establish response surface of the weight and the CG as function of the ship beam
(B).
So two additional ship designs have been created, one with B=29.2 m and the other
with B=35.2m (the beam of the reference design being 32.2 m).
Figure 4.39 shows the internal layout with B=35.2 m (to be compared with the layout
for B: 32.2 m in Figure 4.22).
Figure 4.39. Internal layout of ULJ RoPAX with a beam of 35.2 m
78
For each new layout, the longitudinal stiffeners have been generated and then the
structural assessment-optimisation workflow has been launched (for MARS and then
for BV STEEL) as described for reference design (B= 32.2m) in the previous sections.
Table 4.7 shows the obtained results for the steel weight and centre of gravity.
Table 4.7. Weight and gravity centre for the 3 configurations (B)
(Unit; W in tonne, B and Zcg in m)
Ship Beam ID (iteration number) Steel Weight Gravity centre

(m)
in tons (in m)
29.2 262 11713 13.02
29.2 885 11480 13.06
32.2 253 12521 13.07
32.2 882 12304 13.11
32.2 212 12550 13.07
32.2 883 12242 13.12
35.2 244 13359 13.15
35.2 894 13054 13.19
The derived response surfaces (which are in fact curves as we consider only 1
parameter: the beam) are presented in Figure 4.37 and Figure 4.38.
The RSM (curves) have been transferred to the HOLISHIP platform (concept design) -
CAESES® - for future holistic optimisation including ship structure optimisation (of the
various application cases)
79
5. CONCLUSIONS ABOUT THE STRUCTURAL OPTIMISATION TOOLS
DEVELOPED in WP4
Sections 2 to 4 of this report concerns the development and the integration of tools
to perform an automatic scantling of a midship section. The structural design has to
be feasible (with regards to BV rules) and correspond to a least weight design.
The integrated tools allow to derive response surfaces of weight and gravity centre
elevation, which can be used by the Holistic optimisation of the HOLISHIP platform 1
(concept design) for coupling with other optimisation (hull form, …).
The computation time to use of these response surfaces is very short (msec) and fits
with the objectives.
On the contrary, the time to run the combined MARS2000® and BV-STEEL analysis
(due to the required iterative process) is rather large (several hours). That explains
why response surfaces have to be derived.
All the developments have been validated through a continuous benchmarking
using the ULJ ROPAX, originally planned to be used in HOLISHIP WP16.
Room of future improvement (outside of the project) concerns the computation time
of the MARS2000® and BV STEEL, and maybe the integration of these BV software
within CAESES® (instead of ModeFRONTIER) to derive RSM.
For the Contract design platform (Task 4.3) of HOLISHIP (WP4), the use of a series of
ship sections (4 or 5) will be performed, instead of only the midship section.
6. The Functional Tools

As mentioned in the introduction, beside the larger developments of structural
optimisation and determination of scantlings for the concept design, functional tool
kits are developed to serve the need of e.g. producibility, cost, new materials and
early design configuration of artic offshore platforms. The following sections give a
short insight in the tools, their related industry scenarios.
6.1 Cost Assessment Tool for Structural Design
6.1.1. Tool abstract

Cost assessment tool for structural design is a Microsoft (MS) Excel workbook with an
integrated data model which calculates a dimensionless cost index for each given
design alternative. Calculation of cost index relies on Cost Estimation Relations (CERs)
which are developed upon data analysis of previously built ships by Uljanik Shipyard.
The ships considered for the development of the cost assessment are RoPAX and
RoRo. The cost assessment is focused but not limited to the cost of production for the
structures. However, also the cost for outfitting, etc. was exemplarily performed. The
dimensionless cost indexes are meant to rank the design alternatives from cost point
of view which will then be integrated in the design optimisation process.
80
6.1.2. Tool description
Cost assessment tool is developed based on parametric cost modelling and the Cost
Estimation Relationships (CERs). To assess the structural cost at early design stages for
optimisation purposes, it is sufficient to rank the alternative designs from cost point of
view. However, it is important to select and develop the CERs based on design
variables that have the highest impact on cost.
Figure 6.1.: Data Analysis for CER development

From this perspective, the most important components of structural cost are listed as
follows:
• Materials: all steel material used for the structure
• Material adjustment: market price, complementary material (primer, welding
consumables, etc.)
• Effort: all production related labour man-hours
• Effort adjustment: outsourced/sub-contracted work, productivity, etc.
All data needs to be comparable. This means that the source, in this case the ships,
need to be of a similar kind and size. The here used type of ships are RoRo and
RoPAX. The data from these previously built ships at Uljanik Shipyard were analysed
by the help of MS Excel together with data analysis related Add-ins (Analysis
Toolpack, Power Query and Power Pivot). Based on these analyses CERs were
developed to estimate / calculate a cost index for a given design alternative. For
instance, one of the CER estimates the required productivity (kg/manhours) based
on the length Overall (m), total net weight (tons) and ratio of weight of profiles to
weight of panels (%).
81
Figure 6.2.: CER for total net weight of steel structure vs. LOA
Weight of the steel structure is an important design variable. Therefore, it is common
to get the material related data from design department at contract stage. Similarly,
the material data could also be gathered from steel structure optimisation tools.
Therefore, material related cost estimation / assessment normally does not require
any CERs. The data analysis showed a strong relation between the total net weight
of steel structure (in tons) to the overall length (LOA) in meters (Figure 12).
Figure 6.3.: CER for total effort vs. total net weight of steel structure
For the estimation of production labour hours, the result of data analysis could be
used based on commonly used measures like manhours/ton or kg/manhours. The
estimation could be based on the main dimensions of the ship, total steel weight to
be built, or other variables. Some of the selected results of data analysis are given in
the figure 13, which could be used as CERs for the estimation of effort. The relation
between structural weight, but also between LOA and cost demonstrates a very
strong relation. In case the overall length of the ship and the total steel weight are
known, the required effort could be estimated based on built ships and the CERs.
82
The tool is flexible to allow the end user to freely choose and apply the CERs which
should be used for calculating the cost index. The input data related to the new
alternative designs are received as an Excel workbook including the total net weight,
weight of profiles, weight of plates, etc. Output is a part of the tool itself and gives a
dimensionless cost index for each design alternative.
6.1.3. Tool status

Data preparation, data modelling and performed analyses for deriving CERs (cost
estimation relations) related with the structural steel work were completed for the
Uljanik Shipyard application case. The tool is ready for use as a standalone MS Excel
workbook including the data analysis for Uljanik Shipyard and manual.
6.1.4. Tool Integration

A preliminary input file with over 200 design alternatives was received from ULG and
tested with the tool. Within this input file, each design alternative has several strakes
with different structural characteristics material and dimensions. The tool can handle
this data to aggregate a single value for each design. Depending on the final
optimisation process, the tool could also be integrated into CAESES platform with a
COM feature.
6.1.5. Tool in Application

Trainings for the end-users in Uljanik Shipyard were completed on how to use the tool.
Participants can assess the CERs, make alterations, perform further analysis and
develop their own CERs. The data connection to input file from ULG is also ready for
use.
6.2 Producibility Assessment Tool for Structural Design

Producibility assessment tool for structural design is a Microsoft (MS) Excel workbook
with an integrated data model which calculates a dimensionless producibility score
based on a system which is developed by Rigterink, et al., 2013. The producibility
scores are calculated based on utility functions which were also given by Rigterink,
et al., 2013. This dimensionless producibility score is meant to rank the design
alternatives from producibility point of view which will then be integrated in the
design optimisation process.

The producibility score system provides a dimensionless utility value between 0-1
which is calculated based on utility functions for each producibility element.
Rigterink, et al., 2013 suggested (7) producibility elements and corresponding utility
functions based on design variables of a structural design of a panel. An example is
the producibility element of X Direction Access based on flange to flange dimension
between longitudinal stiffeners. This measure is used for assessing whether there is
sufficient space for workers between the structural elements. For instance, designs
83
having this dimension of 1500mm and above will get the highest score (1) since the
access will be easier which also eases the production. Once all utility values are
obtained for each of the producibility element, they are averaged to get a final
producibility score for each panel. Please see the complete article from Rigterink, et
al., 2013 for further details.
The Producibility Assessment Tool for Design is arranged as an MS Excel workbook by
use of Power Query and Power Pivot Add-ins. This workbook includes the utility
functions suggested by Rigterink, et al., 2013 and a data model with measures which
calculates the required dimension and the corresponding score for each design. The
input data related to the new alternative designs are received as an Excel workbook
including the design variables such as stiffener spacing, plate thickness, profile
thickness, etc.
Figure 6.4.: Input scheme for producability index

Output is a part of the tool itself and gives a dimensionless producibility score for
each design alternative. In addition, a lower limit is defined to get a “Not Producible”
warning in case the score is below the requested value.
Figure 6.5.: Output for producability index – customised for Uljanik Shipyards (utility
value 0.2)
84
6.2.3. Tool status
Utility functions suggested by Rigterink, et al., 2013 are implemented in the tool as
separate worksheets in MS Excel. This allows the end-user to make changes on the
utility functions and the lower acceptable limit for a producible score. Similarly, other
utility functions could be added depending on the requirements of the end-user. The
tool is ready for use as a standalone MS Excel workbook.

Same as the cost tool, a preliminary input file with over 200 design alternatives was
received from ULG and tested with the tool. Within this input file, each design
alternative has several strakes with different structural characteristics material and
dimensions. The tool can handle this data to aggregate a single value for each
design. Depending on the final optimisation process, the tool could also be
integrated into CAESES platform with a COM feature.

Trainings for the end-users in Uljanik Shipyard were completed on how to use the tool.
Participants can assess the utility functions, make alterations and develop their own
utility functions. The data connection to input file from ULG is also ready for use.
6.3 Composite assessment tool for cruise vessel superstructure

The tool provides the engineer suitable composite designs for given design tasks.
Knowing the loads, limiting parameters such as dimensions, spacings, deflections,
etc. the tools provide a conclusion about the feasibility of composite structures as
well as restrictions or constraints.

The tool is divided into two parts. One for the general feasibility of composite
material as well as principle composite designs, weight, etc. It is called composite
pre-tool. The other one to take up the material and assess the structural part in terms
of strength in detail as well as deflection, vibration, etc. This is called composite-
structure tool. The composite pre-tool is based an analytical calculation of the
properties of the composite programmed as MS Excel workbook.
85
Figure 6.6.: composite pre-tool for FRP structure assessment
Therefore, characteristics of fibres, core material and matrix systems need to be
implemented to a database. This data can be retrieved from different suppliers and
easily be changed. Many information about fibres, core material and matrix systems
are publicly available. While the tool provides the option to configure your
composite panel, also pre-configures panels can be selected. Assessing the
feasibility, a first check on the strength and failure of the panel under known loading
conditions will be assessed.
Here, the beam or panel support, the load direction and its values can be selected.
The algorithm checks back against the given failure criteria. This allows the designer
easily to check if given load scenarios from a design are feasible for certain
composite materials. Weight and other criteria are displayed as well.
The second part, the composite structure-tool assesses the structural option of
building with composite material. The method is decided upfront, e.g. beam and
panel structure.
86
Figure 6.7.: composite structure tool: beam-panel built up for composite strucuture
With loading and the material pre-selection, the feasibility and the variety of these
structures can be assessed. The basis is a FEM model running several design variants
giving output to the key parameters in terms of response surfaces.
Figure 6.8.: composite structure tool: result of stiffener spacing on weight and
deflection for three structural options
6.3.3. Tool status

The composite pre-tool and the composite structural-tool are ready and need now
data from, validation and final adjustment within the application case.
87
Both tools are stand-alone kits designed for the use of design engineers. As they are
based on MS Excel, an exchange of data in terms of loads, etc. and results, here
weight, can be done within the CAESES platform.

Trainings for the end-users Meyer Shipyard started. The tools need now be adjusted
towards the application case, the scenarios and the shipyard specifics. As shown in
the graphic below, the first attempt to use the tools is for internal walls and a
composite structure in top of the weather deck of a cruise ship.
Figure 4.9.: application case for the use of the composite tools
6.4 Installation of artic offshore platform

The toolset comprises from two sets to help the designer in the initial design: first the
structure-soil interaction tool; second the installation tool. The design of offshore
structures is embedded in a complex life-cycle, as shown below.
88
Figure 6.10. Life Cycle of offshore oil fields

The structure-soil interaction tool is based on a FEM simulation taking the foundation
method (here: gravity foundation supported by piles), the top-load as well as the
water-load and the soil specifics into account. Having a defined set of designs,
several soil, foundation and load scenarios are assessed. The result will be
summarised in response surfaced to give the designer the needed information about
the feasibility of his design in an early stage. The same applies with the installation
tool. This kind of platform is often built in a multi-platform concept. Due to arctic
conditions, the installation time is limited and a crucial point in the planning. The tool
uses a discrete simulation offshore installation kit for wind turbines, redefined to the
time and conditions for the arctic platform. Both, response surfaces towards specific
parameters (weight, number of item, etc.) can be performed by the tool.
6.4.3. Tool status

Both tool methodologies are ready in general but still in their simulation environment.
Due to their specific use for shallow water artic platforms, data and processes need
to be determined for the application case and the tools updated towards an early
use in design.
The FEM analyses is based on building up a method to predict the soil-structure
Interaction between the seabed and a gravity base foundation. Loads are simplified
into horizontal and vertical ones together with a momentum. The figure below
demonstrates the load assumption as well as a run in the interaction with the soil.
89
Figure 6.11Load model (left) and structure-soil model (right)
The installation of the offshore platform considers transport and assembly in shallow
water. An Excel file will be used to calculate the cost and a Plant Simulation model
will simulate the transportation and installation process to consider the process time.
The cost model is based on algorithm which calculate daily charter rates of the main
transportation facilities like crane vessel, tugs, barges etc. as well the process time to
install the platform in shallow water. Due to various influences like weather
conditions, availability of main resources like the transport and installations vessels a
simulation model is used to create and validate the calculation of the process time.
Plant simulation is an event driven workflow simulation which is used to simulate
material workflow especially processes in production, like assembly or manufacturing
processes in different industries. In respect of HOLISHIP the an add-on (toolset) for
shipbuilding, the simulation toolkit for shipbuilding and offshore installation, is applied.
In this use case the experience from offshore installation of offshore wind turbine is
used.
90
Figure 4.12 Offshore Installation of a wind turbine with Jack-Up-Vessel [FSG]

Integration with the CAESES platform will be foreseen once the overall processes of
the application case are determined.

As mentioned before, the application case is a specialised design for shallow water
and artic conditions of an offshore platform.
Figure 6.13: Application case for the offshore tools

Both tools will interact in the overall design process for such a platform with other
tools, etc. Such a design comprises of a set of data and information. The application
will integrate the developed tools to accelerate two of the main design tasks for
such constructions.
91
7. Conclusions
This document gives an overview about the structural and functional tools
developed for initial / concept design for marine structures and ships. While major
parts of the development served the HOLISHIP RoPAX case in terms of providing initial
and optimised scantlings for the main structures as well as a structural weight
indication, other tools are specifically developed to certain critical design tasks of
new materials, cost and producibility as well as offshore installation guidance. Each
of the tools gives the industrial partner in their application case the insight of a new
way of holistic design using integrated and connected tools and routines, here via
CAESES®. This ensures, that within the same timeframe and low information level,
more design variants can be assessed.
Contact:
Dr. Lars Molter / Michael Hübler
CENTER OF MARITIME TECHNOLOGIES e.V.
Bramfelder Str. 164
D-22305 Hamburg
Germany
molter@cmt-net.org / huebler@cmt-net.org
92
8. Indexes
8.1 Index of figures

Figure 2.1. Integrated structural design assessment-optimisation workflow and its
interface with HOLISHIP concept design integration platform. ............................................ 7
Figure 3.1. Mars2000®, View of a typical MARS model - “scantling check” ................... 11
Figure 3.2.a. Typical 2D model of the frames at midship section, using BV STEEL® ...... 12
Figure 3.3. View of the MARS-based development integrated in the Workflow Editor
environment of modeFRONTIER® ................................................................................................ 14
Figure 3.4.a. Template for WP7-WP4 coupling ......................................................................... 17
Figure 3.5. Automatically generation of stiffeners on the internal layout (ROPAX) ...... 19
Figure 3.6.a. History of stiffener size (Deck 5) during the iterative process (MARS2000®)
............................................................................................................................................................... 20
Figure 3.7. History of MARS2000® rule violation by Rule Violation Indicator .................... 22
Figure 3.8. Evaluated and violated rules output TXT file ........................................................ 22
Figure 3.9. Typical output of the geometrical constraint tool (for test purpose) ........... 24
Figure 3.10. Schematic overview of the weight estimation algorithm, where; 𝑨𝑨=area
of midship section, W m = Weight of midship section, A FX =area of x section fore of
midship, W FX =weight of x section fore of midship, w(b) = weight of bulkheads. ......... 25
Figure 3.11. Outcomes of the Generator of producibility indexes ..................................... 26
Figure 3.12. Demo Viewer .............................................................................................................. 29
Figure 3.13. 2D Viewer showing plate properties .................................................................... 32
Figure 3.14. 2D Viewer showing plate thickness with colours and line styles .................. 33
Figure 4.1. Midship section of the ULJ RoPAX , used as reference design ....................... 35
Figure 4.2. Still water pressure distribution.................................................................................. 37
Figure 4.3. Wave Pressure at Load Case a ............................................................................... 38
Figure 4.4. Wave pressure distribution for load case c and d respectively ..................... 38
Figure 4.5. View of the deck loads at the upright ship condition loading case b+ ...... 39
Figure 4.6. Changes of web thickness of Deck 5 within a fully automatic iterative
process ................................................................................................................................................ 41
Figure 4.7. History of a geometrical constraint (ratio of web height to the flange
thickness > 10) ................................................................................................................................... 45
Figure 4.8. The maximum Von Mises stress history for AH36 high tensile steel (290 MPA
allowable stress) ............................................................................................................................... 46
93
Figure 4.9. MARS model built based on 2D drawing received from ULJ yard (Ref
design)................................................................................................................................................. 47
Figure 4.10. Plate thickness Original MARS model (reference design) - ........................... 48
Figure 4.11. Stiffener profile (Bulb) and layout (spacing) - Original MARS model
(reference design) -......................................................................................................................... 49
Figure 4.12. Material type - Original MARS model (reference design) ............................. 50
Figure 4.13. Loads on Deck5 - Original MARS model (reference design) - ...................... 51
Figure 4.14. Ratio for the Hull girder strength - MARS - reference design (ID000) .......... 52
Figure 4.15. Stress ratio for local strength of strakes - MARS - reference design (ID000)
............................................................................................................................................................... 53
Figure 4.16. Stress ratio for local strength of stiffeners - MARS - reference design (ID000)
............................................................................................................................................................... 54
Figure 4.17. MARS feasible/optimal design (ID1600) - Plate thickness .............................. 56
Figure 4.18. MARS feasible/optimal design (ID1600) - Stiffener size (bulb profile) ......... 57
Figure 4.19. MARS feasible/optimal design (ID1600) - Ratio for hull girder strength ...... 58
strakes .................................................................................................................................................. 59
stiffeners .............................................................................................................................................. 60
Figure 4.22. Material types (left) and loads (right) in STEEL® model .................................. 61
Figure 4.23. Beam types in STEEL® model.................................................................................. 62
Figure 4.24. Von Mises Stresses in STEEL® model for load case a+ ..................................... 63
Figure 4.25. Von Mises Stresses in the optimized in STEEL® model for load case a+ ..... 64
Figure 4.26. History of assessment-optimisation results for MARS-based workflow......... 65
Figure 4.27. The violation against hull girder modulus for the reference design (from
the MARS GUI) ................................................................................................................................... 66
Figure 4.28. List of violations, identified by the “Rule Violation Indicator” tool for the
reference design (ID000) ............................................................................................................... 66
Figure 4.29.a. Violation against BV rule requirements, on the feasible-optimal design

(iteration ID1600) - The violation against modulus requirement in MARS GUI. ............... 67
Figure 4.30. MARS model with three members marked ........................................................ 69
Figure 4.31. STEEL®-modeFRONTIER®; weight history ............................................................. 70
Figure 4.32. STEEL®-modeFRONTIER®; gravity centre history ............................................... 71
Figure 4.33. STEEL®-modeFRONTIER®; gravity centre history ............................................... 72
94
Figure 4.34. Location of the 3 areas used for comparison - STEEL®-model of RoPAX
midship ................................................................................................................................................ 74
Figure 4.35. History of the Deck 5 web height of frames - STEEL®-modeFRONTIER®; ... 75
Figure 4.36. History of the Deck 5 flange width of frames - STEEL®-modeFRONTIER®; . 75
Figure 4.37. Response surface of the Hull steel weight versus the Ship beam (B) ......... 76
Figure 4.38. Response surface of the gravity centre elevation versus the Ship beam
(B).......................................................................................................................................................... 77
Figure 4.39. Internal layout of ULJ RoPAX with a beam of 35.2 m ...................................... 78
Figure 6.1.: Data Analysis for CER development ..................................................................... 81
Figure 6.2.: CER for total net weight of steel structure vs. LOA........................................... 82
Figure 6.3.: CER for total effort vs. total net weight of steel structure ................................ 82
Figure 6.4.: Input scheme for producability index .................................................................. 84
Figure 6.5.: Output for producability index – customised for Uljanik Shipyards (utility
value 0.2) ............................................................................................................................................ 84
Figure 6.6.: composite pre-tool for FRP structure assessment .............................................. 86
Figure 6.7.: composite structure tool: beam-panel built up for composite strucuture 87
Figure 6.8.: composite structure tool: result of stiffener spacing on weight and
deflection for three structural options ........................................................................................ 87
Figure 4.39.: application case for the use of the composite tools ..................................... 88
Figure 6.10. Life Cycle of offshore oil fields................................................................................ 89
Figure 6.11Load model (left) and structure-soil model (right) ............................................. 90
Figure 4.39 Offshore Installation of a wind turbine with Jack-Up-Vessel [FSG] ............... 91
Figure 6.13: Application case for the offshore tools ............................................................... 91
95
8.2 Index of tables
Table 4.1. Characteristics of ULJ RoPAX, used as reference design .................................. 36
Table 4.2. Upper and lower limits of design variables (*), imposed by the Shipyard
(unit is in mm). ................................................................................................................................... 40
Table 4.3. Upper/lower bounds of the Design variables associated to the frames (BV
STEEL).................................................................................................................................................... 41
Table 4.4. Plate thicknesses used in STEEL® modeFRONTIER® workflow ........................... 42
Table 4.5. Comparison between MARS-based assessment-optimisation results for 3

members of the ULJ RoPAX (dimension in mm) – see Figure 4.30 ...................................... 68
Table 4.6. STEEL-based assessment-optimisation results in detail for several members.

(Unit; weight in tonne and dimension in mm) .......................................................................... 73
Table 4.7. Weight and gravity centre for the 3 configurations (B) (Unit; W in tonne, B
and Zcg in m) .................................................................................................................................... 79
9. References
[1] BV. 2018. Rules for the classification of steel ships (NR467). Part B: Hull and
stability. http://erules.veristar.com/dy/app/bootstrap.html
[2] MARS2000®. https://www.veristar.com/portal/veristarinfo/detail/software/rulessof
tware/
mars2000/MARS2000
[3] BV
STEEL®. https://www.veristar.com/portal/veristarinfo/detail/software/rulessoftwar
e/
STEEL/STEEL
[4] modeFRONTIER®. https://www.esteco.com/modefrontier
[5] Box, G.E.P. and N.R. Draper, 1987. Empirical model-building and response
surfaces. John Wiley & Sons, New York.
96

Attachment 0

Uploaded by

Copyright:

Available Formats

Attachment 0

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Attachment 0

Uploaded by

Copyright:

Available Formats

Ref.

Holistic Optimisation of Ship Design and Operation for Life Cycle

Horizon 2020 - 689074

Structural and Functional Simulation

Document Title: Structural and Functional Simulation Tools for Platform I1

Dissemination level: Public Confidential

Deliverable nature: Deliverable

Lead beneficiary: CMT

Responsible author: Dr. Lars Molter [CMT]

Co-authors: Abbas Bayatfar [ULG], Philippe Rigo [ULG]

Reviewer: Apostolos Papanikolaou [HSVA]

Date of delivery: 11.2.2019

Version Date Description

1 30.10.2018 First version

2 5.11.2018 Updated version

4 6.02.2019 Version for QC

5 10.2.2019 Revised final version

2. THE STRUCTURAL ASSESSMENT TOOL ............................................................................................... 7

3.1 GENERAL METHODOLOGY ...................................................................................................................... 8

3.2 LIST OF TOOLS ..................................................................................................................................... 10

3.3 MARS2000® BV SOFTWARE ............................................................................................................... 10

3.4 STEEL® BV TOOL ................................................................................................................................. 12

3.5 MODEFRONTIER® .............................................................................................................................. 14

3.7 SCANTLING-SPACING UPDATER TOOL .................................................................................................. 19

3.8 RULE VIOLATION INDICATOR TOOL ....................................................................................................... 21

3.9 STRESS ASSESSMENT TOOL ...................................................................................................................... 22

3.10 THICKNESS CORROSION UPDATER TOOL ............................................................................................... 22

3.11 STEEL FRAME WEIGHT ASSESSMENT TOOL .............................................................................................. 23

3.12 DESIGN VARIABLES AND PARAMETERS .................................................................................................. 23

3.13 CONSTRAINT TOOL ................................................................................................................................ 24

3.14 WEIGHT/CG CALCULATOR TOOL ........................................................................................................ 25

3.15 GENERATOR OF PRODUCIBILITY INDEXES ............................................................................................... 26

3.16 DEMO VIEWER TOOLS ........................................................................................................................... 28

4. APPLICATION: “THE ULJANIK ROPAX”........................................................................................... 34

4.1 INTRODUCTION ..................................................................................................................................... 34

4.2 PRESENTATION OF THE ULJ ROPAX ...................................................................................................... 35

4.2.3. Selected design variables and their upper/lower limits .............................................................. 40

4.3 MARS MODEL ..................................................................................................................................... 47

4.4 STEEL MODEL ...................................................................................................................................... 61

4.5 ASSESSMENT-OPTIMISATION RESULTS ...................................................................................................... 65

6.1 COST ASSESSMENT TOOL FOR STRUCTURAL DESIGN .............................................................................. 80

6.1.3. Tool status................................................................................................................................... 83

6.2 PRODUCIBILITY ASSESSMENT TOOL FOR STRUCTURAL DESIGN ................................................................ 83

6.2.4. Tool Integration .......................................................................................................................... 85

6.3 COMPOSITE ASSESSMENT TOOL FOR CRUISE VESSEL SUPERSTRUCTURE ..................................................... 85

6.3.3. Tool status................................................................................................................................... 87

6.3.4. Tool Integration .......................................................................................................................... 88

6.3.5. Tool in Application ...................................................................................................................... 88

6.4 INSTALLATION OF ARTIC OFFSHORE PLATFORM ....................................................................................... 88

6.4.2. Tool description........................................................................................................................... 89

6.4.3. Tool status................................................................................................................................... 89

6.4.4. Tool Integration .......................................................................................................................... 91

8.1 INDEX OF FIGURES ................................................................................................................................. 93

8.2 INDEX OF TABLES ................................................................................................................................... 96

Figure 2.1. Integrated structural design assessment-optimisation workflow and its

The prime objective is to perform a rule based structural design assessment-

3. PRESENTATION OF THE TOOLS, DEVELOPED AND USED IN TASK

Existing tools/software along with its new script/batch-mode developments are:

• MARS2000® BV software - section 3.3