See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/324908959

A MODEL BASED APPROVAL PROCESS FOR BASIC HULL DESIGN

Conference Paper · September 2017

CITATIONS READS

0 760

1 author:

Ole Christian Astrup

Det Norske Veritas
10 PUBLICATIONS 49 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Model based approval for ship design View project

All content following this page was uploaded by Ole Christian Astrup on 05 November 2018.

The user has requested enhancement of the downloaded file.

International Conference on Computer Applications in Shipbuilding 2017, 26-28 September 2017, Singapore

A MODEL BASED APPROVAL PROCESS FOR BASIC HULL DESIGN

O C Astrup, DNV GL, Norway and C Cabos, DNV GL, Germany

SUMMARY

This paper explores both the barriers to overcome and the requirements needed to establish a fully digital workflow
providing a seamless and transparent digital model exchange between stakeholders during the ship newbuilding basic
design stage. A 3D digital model established early in the design stage and shared among stakeholders will eliminate the
need for producing 2D drawings for the verification of the design by Class. The digital 3D model is the basis for verifying
the design against Class Rules and functional goals. A design centric work process involving the various stakeholders is
the result of linking the Class rules directly to the Computer Aided Design (CAD) early design tools used by the designer.

1. INTRODUCTION 3D models of airplanes, cars, bridges, digital cameras ...

and of course, ships, submarines and floating structures.
Traditionally, drawings are used for communicating
Many shipyards, designers and at least the larger suppliers
designs in industry because they are the clearest way to
already use 3D CAD systems and PDM/PLM including
tell a human what to make and how to make it. Technical
various 3D-based simulation and engineering tools such
drawings provide a means to communicate product
as FEM or CFD. A large amount of knowledge about the
complexity in a comprehensible and effective manner
ship is collected in this digital representation: Not only the
thanks to visual abstraction. In shipbuilding, it is still an
geometry and details about the production process but also
explicit requirement that the designer/yard shall provide
various aspects of the behaviour and the operation
the Classification society with Class drawings
procedures. This intensive use of 3D technology in the
documenting the design to be approved to the Society’s
engineering phase is an important enabler for increasing
Rules.
productivity and managing complexity.
At the same time Computer Aided Design (CAD) models
are now displacing technical drawings and documentation 2.4 USE OF 3D IN NEWBUILDING PLAN
as the main product definition in several major industries. APPROVAL BY CLASS
Within the last ten years or so, the engineering industry in
automotive, aerospace and construction has gradually The role of the Classification society during the new-
converted to using CAD models directly for building phase is to verify and approve that the design
communicating designs to manufacturers, builders, fulfils Class and Statutory requirements [1] (i.e. the
maintenance crews and regulators. This switch to creating Rules). Class carries out the verification through a
the engineering record in models, however, presents technical review of the design plans and related
problems not only for its long-term maintenance and documents for a new vessel to verify compliance with the
accessibility – due in part to the rapid obsolescence of the applicable Rules. Current practices in classification is to
hardware, software and file formats involved – but also for base the verification job on traditional documents such as
recording the evolution of designs, artefacts and products. 2D General Arrangement drawings and other design
information in paper/pdf formats made by the
One of the major problems facing long-term users of designer/yard. For the hull design verification, 2D
engineering data is a lack of compatibility between drawings have been the single most important design
software systems, specifically between competing document exchanged between yard/designer and Class.
systems and between different generations of the same The yard/design office will prepare 2D drawings as the
system. This can partly be explained by market forces, and basis for documenting the design. The society must
the consequent tactics used by vendors to encourage manually build up their own digital verification model
customer loyalty, and partly by genuine conceptual based on these documents. This process is time consuming
differences between systems. In consequence, data created and error prone. For every design revision, the society may
using one software is in danger of becoming inaccessible need to do a re-verification adding time to an already
even to its creators once that piece of software is retired or inefficient process.
replaced as part of ongoing modernization.
A case study [2] involving the exchange between NAPA
Steel and the DNV GL hull calculation tool POSEIDON
2. INTEROPERABILITY OF ENGINEERING
clearly demonstrated the potential of interoperability. The
DATA IN THE SHIPBUILDING INDUSTRY
study concluded:
2.3 USE OF 3D IN THE LIFECYCLE OF A SHIP
• Modelling activity for scantling check by the yard can
Computer-Aided Design (CAD) Systems help engineers be reduced by more than 50%, saving time for the
and designers in various industries, designing and building yard in early design.

© 2016: The Royal Institution of Naval Architects

 International Conference on Computer Applications in Shipbuilding 2017, 26-28 September 2017, Singapore

• The POSEIDON model can be used for prescriptive Enterprise Resource Planning (ERP), Customer
Rules Checks and for the complete finite-element Relationship Management (CRM) and Supply Chain
buckling analysis based on global FEA results. Management (SCM). Thus, there is an increasingly greater
reliance on CAD models which are now being used as the
• Loads for global FEA can be generated based on input method for recording product definition.
directly transferred from NAPA.
Traditionally, drawings are used for communication in
• Calculated scantlings are passed back to NAPA Steel industry because they are the clearest way to tell human
for update by the designer. what to make and how to make it. They are considered as
a graphic universal language. The fundamental purpose of
From the same study, SHI claimed to achieve the an engineering drawing is to carry, control and maintain a
following savings in time [3]: product’s definition in a precise and clear way with no risk
of misinterpretation or assumption. Technical drawings
provide a means to communicate product complexity in a
• Prescriptive rule check calculation 50%
comprehensible and effective manner to humans thanks to
• Global FEM calculation 25%
visual abstraction.
• FEM calculation for hold analyses 40%
• Local fine mesh FEM check for details 60%
Instead of shaping exchanged information towards human
perception, digitalization calls for exchanging data sets
The above study shows the potential of improved
directly consumable by algorithms, i.e. move from
interoperability through exchanging design models
drawings to models with semantical mark-up (as opposed
between yard and Class during the basic design stage.
to graphical mark-up). This results in opportunities for
automated processing but can lead to challenges for
2.5 THE INTEROPERABILITY CHALLENGE
human interpretation.
The variety of engineering tools used to support design,
Both the aviation and automotive industries are moving
procurement, manufacturing, and support of shipbuilding
towards a drawing free product lifecycle. We can see
products has never been greater. From company to
several drivers in this development [4]:
company, tools and processes range from manual capture
in 2D drawings to sophisticated 3D models that are tightly
• One master product definition
integrated with other enterprise systems. The challenge is
• Virtual prototyping and simulation
further compounded by the growing need to provide
• Computer-aided manufacturing (CAM)
engineering information for support extending beyond the
• Assembly automation
life span of individual applications. This heterogeneity has
• Production of maintenance documentation
created both technical and business challenges. Data
• Earlier coordination of design activities
integrity across the applications and systems that author
and consume engineering data is problematic. Point-to-
The above list is not exhaustive, but some common
point integration between systems is often so complex and
denominators are reduced time to market, reduced rework,
costly that organizations opt for manual data re-entry
reduced cost and improved transparency for stakeholders
when faced with program budget and schedule constraints.
throughout the product lifecycle.
This leads to high coordination effort between design
disciplines and consequently to expensive errors detected
Therefore, it is appropriate to ask the following questions:
only later, e.g. in production planning.
What are the main barriers that must be overcome to
transfer all the different functions of engineering drawings
3. CAN A MODEL-BASED DEFINITION
to digital models within the product lifecycle? What are
REPLACE ENGINEERING DRAWINGS?
the legal requirements that must be fulfilled? What are the
required model characteristics and what specific tools
3.1 A DIGITAL WORKFLOW
need to be put in place to move to a model based approval?
CAD models have traditionally been used in the design,
3.2 WHAT IS A MODEL?
evaluation and manufacturing phases. Up until the turn of
the millennium, engineering software was used to support
A model is a representation or idealization of the structure,
a document-based workflow – CAD packages were used
behaviour, operation, or other characteristics of a real-
to create virtual models of designs, from which drawings
world system. A model is used to convey design
and other design documentation could be produced. The
information, simulate real world behaviour, or specify a
manufacture or construction process was based on the
process. ASME [5] and ISO [6] gives the following
resulting documentation. However, current digital
definition of the term model:
environments allow a model-based flow of information
between heterogeneous systems for Computer Aided
A model is a combination of design model, annotation
Design (CAD), Computer Aided Engineering (CAE) and
and attributes that describes a product.
Computer Aided Manufacture (CAM) as well as

© 2016: The Royal Institution of Naval Architects

International Conference on Computer Applications in Shipbuilding 2017, 26-28 September 2017, Singapore

Engineers use models to convey product definition or • Data management: An appropriate method must be
otherwise define a product’s form, fit and function, see put in place to manage and record revisions of MBD
Figure 1. According to the Model Based Enterprise [7], datasets;
models can be applicable to a wide range of domains
(systems, software, electronics, mechanics, human • Data security: A mechanism that incorporates
behaviour, logistics, and manufacturing). Models can be security features (confidentiality, authentication,
either computational or descriptive. Computational integrity and non-repudiation) when accessing,
models are meant for computer interpretation and have a exchanging and interacting with MBD datasets will
machine-readable format and syntax. Descriptive models be required; and
are human interpretable and meant for human
consumption (symbolic representation and presentation). • Data retention: Today typically only 2D data is stored
Core to MBE is the integration of descriptive models with and used for legal purposes.
computational models.

4. EXISTING INTEROPERABILITY
STANDARDS FOR PRODUCT DATA
REPRESENTATION

4.1 ISO AND STEP

ISO and STEP has led to improvements in exchange and

sharing of simple CAD information, product models and
complete product structures. This is particularly true for
application protocols AP203, AP204 and AP214 [10],
[11], [12] which are of wide use in the automotive and
aviation industries. The recent ISO AP242 [11]
consolidates the previous two AP204 and AP214 into one
protocol for Managed model-based 3D engineering.
Figure 1: Model data elements, ASME [5] Furthermore, STEP has improved communications within
the extended enterprise (including suppliers, business
3.3 MODEL BASED DEFINITION partners and customers) and helped to support global
collaborations. However, due to the bulk of its
According to [8] a model-based definition(MBD) is a documentation and its complexity, there are still some
digital representation (artefact) of an object or system. It issues that hinder the application of STEP in an industry
is representative of the physical object or system and all with many specialized computer codes.
its attributes, and is used to communicate information
within various MBx activities in a model-based enterprise. 4.2 APPLICATION TO SHIPBUILDING
The MBD is rich in information – shape, behaviour, and
context – and it conveys the information architecture Considerable effort has been spent in developing standard
within an enterprise. protocols for product definitions both in the shipbuilding
and other industries. Due to its broad scope, the
Based on a study from the aerospace industry, the application protocol AP218 is principally capable to
following concerns of moving to a full MBD environment support the need for data exchange during the early stages
were identified, [9] [4]: of the design in a consistent manner. Due to the generally
• Data accessibility and visualization: Most limited support by the software vendors, the shipbuilding
downstream users (including suppliers and protocols are not used in actual commercial ship design
customers) do not have access to CAD software, and projects today [13]
therefore a visualization tool that allows them to read
and use MBD datasets must be adopted; The lack of a ship specific standard is of major problem to
the shipbuilding industry. This has left a void that needed
• Data content: Downstream users need to be confident to be filled and resulted in a continuation of the
that the MBD datasets will carry the drawing’s core development of “point-to-point” solutions based on a
and the required management elements; variety of ad hoc application programming interfaces
(API’s) and XML-based interfaces. While these have been
• Data presentation: Engineering drawings follow effective in terms of achieving the result, they have not
international standards in terms of how the data is been cost effective from an industry perspective as the
organized and structured and so must MBDs; number of interfaces grows at a combinatorial rate as the
numbers of applications increase.

© 2016: The Royal Institution of Naval Architects

 International Conference on Computer Applications in Shipbuilding 2017, 26-28 September 2017, Singapore

The problem of access to data across software and system • Context-specific information can be extracted into
revisions in the shipbuilding industry is therefore serious, several separate files to provide multi-layered mark-
and it is unlikely that a wholly satisfactory solution will be up, which can be passed around only as required,
forthcoming soon. reducing the amount of data that need to be
transferred and allowing the CAD model to remain
5. A LIGHTWEIGHT FORMAT APPROACH smaller in size.
TO MODEL BASED DEFINITION
• The same mark-up can be applied to different
The initial objective of a product model is to represent the representations of the same model, granting the mark-
design and engineering aspects of a product during the up information some independence of the CAD
configuration design, assembly design or detail design format used.
stages.
• Downstream processes (e.g. finite element analysis
As described in the previous section, interoperability of and manufacturing processes) become capable of
engineering data is still a major concern to the independence from the full CAD model.
shipbuilding industry. CAD models are currently
experiencing two conflicting demands. On the one hand, The ability to transfer information layers between
they must increasingly support the whole product different representations of an object makes this approach
lifecycle, not just the design stage. On the other hand, particularly valuable. E.g. computational dimensioning
there is a need for small files that can easily be shared with results can be transferred back to a CAD model, FE
partners whilst at the same time withholding Intellectual stresses can be used to highlight critical spots in inspection
Property (IP) rights. Compared to full CAD model models, etc. [15]
formats, lightweight geometry representations have some
advantages, for example, they are typically generated with Ding et.al, [16] presents the following advantages of
approximate geometry using various compression multi-layered annotations:
methods, to protect sensitive information. The small file
sizes speed the communication of the models. • Mark-up information is stored outside of CAD
Furthermore, lightweight geometry representations can be models by multiple XML-based documents per
viewed without the aid of expensive CAD packages. different viewpoints and security levels. The
advantages of XML include the possibility of
Patel [14] proposes the use of annotation to augment the manipulation of the mark-up information into
geometric model of the product with additional different viewpoints, the re-organization of the
information supporting the product lifecycle at the same information for various purposes and applications,
time enabling sharing of 3D models between multiple and the tailoring of shared information per recipient
stakeholders needs and IP considerations etc.

5.1 MULTI-LAYERED ANNOTATIONS • It enables partners that only receive lightweight

representations of the product to apply mark-up and
Annotation can be simply defined as adding any extra such mark-up can be linked directly to the original
information for various purposes, such as dimensioning, CAD model.
feature parameters, further explanations, viewpoint
interpretation, extra descriptions of or comments on an 5.2 CHALLENGES USING MULTI-LAYERED
existing entity. ANNOTATIONS
Patel [14] identifies two methods of applying annotation The design process involves complicated interactions
or mark-up: ‘inline’ and ‘stand-off’. Inline mark-up among tasks and people. Thus, it is presenting
involves adding tags directly into the text of a document unavoidable inclusion of iterations and rework, which
or representation of a model, whereas the stand-off results in the constant insertion, elimination and
(external or reference) method allows mark-up modification of annotations (e.g. constraints) through the
information to be stored separately, being linked back to whole process. Such evolutions not only reflect the design
the document or model using references or pointers. The process, but also possibly embed the rationale behind the
latter is the more appropriate for use with CAD models for decisions made. Meanwhile, the design process is also
several reasons: accompanied by the evolution of the product model, and
therefore there is a high risk of loss of annotations, if the
• Stand-off annotation allows the 3D geometric corresponding entity associated with the annotations
representation of a product to be progressively disappears. To address these issues, a strategy of
expanded to include additional metadata without management and preservation of annotation evolution
changing the representation method used for the should be implemented, including:
geometry of the product.

© 2016: The Royal Institution of Naval Architects

International Conference on Computer Applications in Shipbuilding 2017, 26-28 September 2017, Singapore

• Evolution of metadata: The evolution of annotations to achieve the ambition of model based approval by the
can be retraced if and only if evolving histories are Classification society, see Figure 2.
clearly recorded in an organized way. Normally, the
modifications of the associated annotations can 6.1 THE DIGITAL EXCHANGE SPECIFICATION
happen at two levels: model-level or element-level.
The lack of a ship specific standard is of major problem to
• Reference point: A core limitation identified by the shipbuilding industry. The problem of access to data
previous research [16] is that the annotation may not across software and system revisions is therefore serious,
be retraced when the CAD model changes, e.g. if the and it is unlikely that a wholly satisfactory solution will be
entity to which it is anchored is deleted. forthcoming soon. The authors therefore propose to
develop a DEX supporting the necessary exchange
While the use of a lightweight geometry model addresses capabilities between designer/yard and Class, as there is
many of the challenges of PLM, there remains a problem currently no single working and established standard for
in selecting a format which is the most appropriate not the exchange of engineering data within the shipbuilding
only for a use and view of the product, but also for long- community for this purpose.
term retention.
Based on research carried [17] out we want the DEX to
6. ENABLING A MODEL BASED APPROVAL have the following capabilities:
PROCESS FOR HULL BASIC DESIGN
• A lightweight and CAD neutral representation of the
The key to achieve interoperability across software 3D geometry.
applications is open standards, i.e., those developed by • The ability to carry both precise and approximated
consensus either within a standards development geometry representations depending on the intended
organization or a consortium of stakeholders. No single use.
software tool can perform all the engineering tasks needed • The ability to grow the product definition information
to design and manufacture a product. No single software throughout the design lifecycle.
product can do it all well. Users will mix and match • The ability to describe function and process related
software products matching their business objectives. information.
Standards define an agreed upon syntax and semantics of • The ability to reference function and process
3D modelling constructs and annotation so that users can information to the 3D model using multi-layered
understand one another’s models and design intents. annotations.
• A referencing mechanism which is robust to 3D
A model based definition provides the core foundation of model design changes.
the product lifecycle. It is one of the first pieces of • An easily extendable scheme to cater for new function
information that is created during product development. It and process related data.
then grows as the product matures evolving it into the
single authoritative source of product definition. These design assumptions lead us to logically structure the
DEX content into three groups, see Figure 3.
Because of obvious IPR considerations, the full MBD
created by the designer will not be shared with other
industry stakeholders. This is why the approach of
lightweight models with multi-layered annotations is
proposed [15]. The information passed to downstream
users shall be delivered in a format that is CAD neutral
and only contains those elements required for the intended
purpose.

Figure 3: DEX information structure

What: The vessel hull form described by its geometry,

Figure 2: Model Based Definition and lightweight features, materials, topology, …
exchange.
Why: The functions of the vessel represented by
The Digital Exchange Specification (DEX) which annotation layers describing requirements, constraints,
specifies the lightweight exchange format is instrumental functions or relationships.

© 2016: The Royal Institution of Naval Architects

 International Conference on Computer Applications in Shipbuilding 2017, 26-28 September 2017, Singapore

How: Process and business related data represented by and actions taken to resolve it if necessary. The first
layered annotations describing process information for mechanism requires authoring systems to keep IDs stable
e.g. approval, change management, inspection, testing etc. in case of model changes.

In this way, it is possible to separate the digital product As features and topology also need to reference the
form definition (the ship itself) from process and life cycle geometry, they are ideally based on the same referencing
specific data (annotations). mechanisms.

We envisage that this separation will allow us to establish Having a rich toolset for viewing and mark-up of 3D
a “conversation” around a 3D model using layered models will be important for preserving efficiency when
annotations to resolve any issues during the approval moving from working with a 2D drawings to a fully 3D
process. Figure 4 below illustrates how a designer and an based approval process.
approval engineer can use this mechanism to resolve
comments during the plan approval. 6.2 DEX: FORM

There is no single standard covering the complete

information scope. ISO and STEP protocols exist for the
exchange of 3D data (the 3D Form). These standards can
represent both exact and approximate geometries, but
have limitations with respect to representing global
geometry topologies. Adding to the picture the complexity
of a STEP implementation and the fact that current
standards do not support the shipbuilding industry, there
is a need to define a lightweight schema for the ship form.

AVEVA and Intergraph [18] neutral XML formats meets

our requirements stated earlier, and to some extent
complement each other. The neutral formats can represent
exact B-REP geometries with global topology in native
XML, and have mechanisms for referencing external
geometry files. With external geometry references several
formats can easily be supported (e.g. IGES, JT, STEP,
Figure 4: Envisioning model based approval utilising PRC). Through such external referencing of standard
lightweight model exchange and layered annotations. geometry formats, the application of commercially
available 3D review tools would be facilitated.
6.2 DEX: ANNOTATION REFERENCE
MECHANISM

It will be beneficial to separate function and process

specific metadata from the 3D form data. This will be
achieved by introducing the concept of the layered
annotations described earlier. Each annotation will have a
reference to the 3D geometry form. Depending on the
annotation type, it can reference the whole geometry or a
specific geometry part. As stated earlier, the challenge is
to construct a reference mechanism which is robust to
design changes.

For annotations referencing a geometry entity in the 3D

model, the authors propose to create a reference
mechanism which consists of two elements:

• Reference to a unique geometry entity ID in the 3D

model
• A reference to an object type and a coordinate
position, area or volume in the ship coordinate system
Figure 5: OpenHCM Test Case 4 [19]
This will introduce a resilient mechanism which can be
resolved when the model changes and the entity which is
referenced is deleted. The annotation can still be displayed

© 2016: The Royal Institution of Naval Architects

International Conference on Computer Applications in Shipbuilding 2017, 26-28 September 2017, Singapore

STEP AVEVA JT Intergraph specific strength rule requirements and acceptance

4 344 kB 63 kB 284 kB 120 kB criteria. By classifying the different structure functions
Table 1: Comparison of file sizes for test case and link these to the ship 3D model parts, it will be
possible to expose the Class requirements to the designer
A simple investigation on file size was carried out using and automate hull rule calculations.
the OpenHCM [19] test case depicted in Figure 5. The
investigation demonstrates the compactness of the vendor This is easiest explained by an example:
formats. The results are given in Table 1, and show that Minimum plate thickness requirements are defined in the
both the AVEVA and Intergraph neutral XML files DNV GL Rules for Classification: Ships, Part 3, Chapter
including Siemens JT are compact in size. It is worthwhile 6, Section 3:
to note the large size of the corresponding STEP file,
which is almost 100 times the size of the smallest file size. 𝑡𝑡 = 𝑎𝑎 + 𝑏𝑏 𝐿𝐿2 √𝑘𝑘
where:
6.3 DEX: FUNCTION a, b are coefficients defined by the plate type
𝐿𝐿2 is he Rule length
There is no available standard supporting Class k is a material factor based on the material type
requirements and constraints (the functional part) as per
today. There have been attempts to establish such a From the above, we see that the plate type defines the
standard in the past [20], but nothing has materialised. coefficient a and b. The Rules defines the different
applicable structure functions for the plating, see Figure 6.
The functional and regulatory requirements (i.e. the Rules)
Features
by Class are today published in paper form, e.g. the DNV
Function
GL Rules for Classification: Ships [21]. Such rules must Topology
Form
Structure Functions

be interpreted by a designer before it can be applied to a Materials

Catalogues End Connections

Cross-sections
new design. Such interpretations may lead to ambiguities Geometry
Materials

which need to be resolved by the designer and Class Loadings

during the plan approval process, adding time to an DEX

already constrained schedule. Schema

version Metadata Comments

Application
Process
Approval state
What if the Class requirements could be presented in a Date&Time

User
Measurements
precise and digital form through a DEX? We believe it is Units

possible to expose Class requirements through a set of Results

standardised Class catalogues. Such catalogues can cover ….

different functional requirements. The following

catalogues will be provided in the first version of the DEX: Figure 6: Taxonomy for Structure Function: Plating

• Vessel types The Structure Function catalogue provides the designer

• Structure Functions with the taxonomy for assigning the structure functions to
• End connections the 3D design model. We believe that this assignment can
• Cut outs be well supported by CAD systems.
• Materials catalogue When the structure function is assigned to a structure part
• Cross section catalogue in the 3D model, most of the minimum requirements can
easily be determined. The more complex requirements
Vessel Types: A high level characterization of vessels is like cross section ultimate strength (yield and buckling)
fundamental to most technical and commercial topics will need more comprehensive information to be supplied
within the maritime industry. Nevertheless, no universally to determine both the loadings and the relevant capacity
adopted standard for this exists. Most organisations models.
working in the maritime industry relate to one or more
propriety defined lists of ‘ship types’. This situation The same principles can be used to establish structure
prevents an effective exchange of information. The vessel function catalogues for stiffeners and primary support
type catalogue is based on a high-level characterisation of members.
vessels replacing the typical ambiguous vessel type lists
currently in use. A unique vessel type is defined through End connections and cut-outs: The Rules define a set of
the use of 5 basic variables [22]. The vessel type is the standardised end connections and cut outs for different
basis for selecting applicable rule sets and Class notations. ship details. These types will be provided in the form of a
catalogue, and the designer must map the relevant
Structure Functions: Class structure Rules refer to standard detail to his own design. For most details, this is
structural elements of a ship as girders, frames, a onetime exercise as the designers already are using yard
longitudinal bulkheads, plates and stiffeners. Most ship or company specific standard details.
structural elements have a load bearing function with

© 2016: The Royal Institution of Naval Architects

 International Conference on Computer Applications in Shipbuilding 2017, 26-28 September 2017, Singapore

Material catalogue: A catalogue defining the applicable • The industry will benefit from a single neutral
material types defined by the Rules. exchange format for ship specific 3D shape and
topology information avoiding costly point-to-point
Cross Section catalogue: A catalogue defining the integrations.
standard stiffener cross section types recognised by the
society. This can easily be mapped to the yard/designer • Engineering models can consistently be derived from
standards. one single 3D model source,

6.4 DEX: PROCESS • CAD idealizations supporting several engineering

analysis purposes can be covered: whole ship models
There are evolving standards targeting the process part, to detailed fatigue models
both ISO 10303-239 PLCS (Product Life Cycle Support)
[23] and industry standards (e.g. PLM XML [24]). They • Multi-layered annotations are independent of 3D
will be explored to see how the standards, or parts of the model exchange and CAD formats
standards, can support the plan approval process and
change management. • Annotations are loosely coupled to the correct entity
in the model and can therefore be transferred between
Comments and comments management are central to the discipline specific model representations
plan approval process carried out by Class. For hull plan
approval, typically the approval engineer today issues a • Annotations can be edited, circulated, and processed
comment in the form of a mark-up directly on the 2D plans independently of the model, whilst the 3D model
or drawing. In a fully model based approval process, the remains unchanged.
comments must be issued directly on the 3D model and
transferred back to the designer to be resolved, see also • Multiple independent annotation files can be safely
Figure 4. applied to the same 3D model.

The comment annotations will use the referencing • Lightweight 3D shape representation protects
mechanism described previously. In addition, the owner’s IPR
comments must carry the metadata supporting the
comment management process. • Annotations are easily extendable and offer support
for business specific needs
One important aspect when attaching a comment to a 3D
model, is to also record the camera position and view
direction as part of the comment metadata. In this way, it 8. FURTHER WORK
will be possible for each recipient to have the same view
as the one that originally issued the comment. It can also Further research must be carried out to complete a DEX
be valuable to store the display state of the model as part capable of fully supporting a model based approval
of the comment annotation meta data, i.e. any model process. This work needs to solve the following
clipping, colour coding, transparencies etc. challenges:

Any additional mark-up data (i.e. redlining, measurements Change management: Change management need to
etc.) will also be carried on as comment metadata. address both 3D model revisions and the state of any
annotation referencing the 3D model.
6.5 DEX: FORMAT
Model maturity: Model or design maturity will need to
A specific format (syntax) of a DEX is less important than be communicated between stakeholders. We want to
the actual information content (semantics). We have explore the layered annotation principle to see if it can be
chosen to start with a native XML format. This has some used to express design maturity.
obvious advantages when there is a need for quick
prototyping and testing of new concepts. XML is also Approval scope: Class must document what has been
human readable and text based, and comes with standard approved. Today this is documented by the stamping of
toolsets supporting an implementation. the 2D drawings. This mechanism needs to be transferred
to the 3D model.
7. CONCLUSIONS
Long term storage and retrieval: Class needs to retain
The authors have described a way forward for establishing the approval documentation for the lifecycle of the vessel,
a model based approval process for hull plan approval. which typically is 25 years. If the approval is based on a
The proposed solution has several benefits: 3D model and not drawings, we need to make sure that we
can still retrieve this information in a 25-year perspective.

© 2016: The Royal Institution of Naval Architects

 International Conference on Computer Applications in Shipbuilding 2017, 26-28 September 2017, Singapore

Rule and strength calculations: A natural use of the ‘A Collaborative Platform for Ship Design’, in
DEX is as input to rule and strength calculations. This will ICCAS, 2005.
facilitate recreating models from 2D drawings and fully [14] M. Patel, A. Ball, and L. Ding, ‘Strategies for the
leverage the potential demonstrated in the use case from Curation of CAD Engineering Models’, Int. J.
section 2.4 Digit. Curation, vol. 4, no. 1, pp. 84–97, Jun.
2009.
The project participants aim to carry out a proof of concept [15] C. Cabos, ‘Towards a DEX Exchange Standard -
of a fully model based approval when the DEX has been A Proposed Approach’. DNV GL, Malmø, 2017.
developed. [16] L. Ding, D. Davies, and C. McMahon, ‘Sharing
Information Throughout a Product Lifecycle via
9. ACKNOWLEDGEMENTS Markup of Product Models’, pp. 1–9, 2008.
[17] O. C. Astrup, ‘Report 2017-0413: A Model
This work has been supported by the Norwegian Research Based Definition for Shipbuilding’, DNV GL,
Council project APPROVED (Project No. 256432) with 2017.
participations from Rolls Royce Marine, ULSTEIN [18] M. Polini, ‘A NEUTRAL XML SCHEMA FOR
Design and Solution, STX France, AVEVA, DIGITREAD BASIC DESIGN STAGE INTERFACE TO
(Siemens) and Intergraph Corp. (Hexagon). CLASS SOCIETIES’, in ICCAS, 2011, vol. 2.
[19] OpenHCM, ‘OpenHCM Standard Description’.
10. REFERENCES [Online]. Available:
https://openhcmstandard.github.io/. [Accessed:
[1] IACS, ‘CLASSIFICATION SOCIETIES - their 06-Jul-2017].
key role’, 2011. [Online]. Available: [20] U. Haenisch, J. and Langbecker, ‘Exchange of
http://www.iacs.org.uk/document/public/explaine Structural Design Data for Hull Cross Section
d/CLASS_KEY_ROLE.pdf. Approval, EMSA Protocol’, 1999.
[2] C. Cabos, ‘3D Ship Design from the Start - An [21] DNV GL, ‘Rules for Classification: Ships Part 3
Industry Case Study’, in COMPIT, 2015, pp. Hull Chapter 1 General principles’, July 2016.,
257–268. DNV GL, 2016.
[3] T. Hulkkonen, B. S. Kang, C. Cabos, and B. [22] V. Vindøy and K. Dalhaug, ‘High level vessel
Tietgen, ‘PROCESS INNOVATION FOR characterization’, in Compit’ 2016, 2016, pp. 42–
HULL DESIGN AT SAMSUNG HEAVY 50.
INDUSTRIES’, in ICCAS, 2015. [23] ISO, ‘ISO 10303-239:2005 Product life cycle
[4] V. Quintana, L. Rivest, R. Pellerin, F. Venne, support’, iso, 2005.
and F. Kheddouci, ‘Will Model-based Definition [24] Siemens, ‘PLM XML: Siemens PLM Software’,
replace engineering drawings throughout the White Paper, 2011. [Online]. Available:
product lifecycle? A global perspective from https://www.plm.automation.siemens.com/en_us/
aerospace industry’, Comput. Ind., vol. 61, no. 5, products/open/plmxml/. [Accessed: 17-Nov-
pp. 497–508, 2010. 2016].
[5] ASME/Y14.41, ‘Digital Product Definition Data
Practices’, ASME International, 2012. 9. AUTHORS BIOGRAPHY
[6] ISO, ‘ISO 16792:2015 Technical product
documentation’, International Organization for O.C. Astrup joined DNV GL in 1991 (former DNV) and
Standardization, 2015. currently holds the position as Senior Principal
[7] MBE, ‘Model Based Enterprise Overview’, Researcher in DNV GL Group Technology and Research.
2016. [Online]. Available: http://model-based- He is currently working on research activities within
enterprise.org/model-based-enterprise- digitalization of shipping to improve safety and
overview.html. [Accessed: 16-Nov-2016]. efficiency with focus on dynamic barrier management and
[8] N. Hartman, ‘Defining the Minimum Information model based approval.
Model for the Model-based Definition’, in
Model-Based Enterprise Summit, 2017. C. Cabos joined DNV GL in 1989 (former GL). He is
[9] AIA, ‘Aerospace Industry Guidelines for responsible for Information Management Technologies in
Implementing Interoperability Standards for Maritime Class Development. This includes the
Engineering Data’, 2014. development of new computer-based methods in the fields
[10] ISO, ‘ISO 10303-1:1994 Industrial automation of maintenance, data exchange, PLM and computational
systems and integration -- Product data mechanics. His professional background is in numerical
representation and exchange’, 1994. mathematics and physics.
[11] ISO, ‘ISO 10303-242:2014 Managed model-
based 3D engineering’, 2014.
[12] ISO, ‘ISO 10303-204:2002 Mechanical design
using boundary representation’, ISO, 2002.
[13] R. Bronsart, U. Cantow, W. Grafe, and T. Koch,

© 2016: The Royal Institution of Naval Architects

View publication stats