EUROPEAN COMMITTEE FOR STANDARDIZATION
C O M I T É E U R O P É E N D E N O R M A LI S A T I O N
EUR OP ÄIS C HES KOM ITEE FÜR NOR M UNG


Management Centre: Avenue Marnix 17, B-1000 Brussels
© 2010 CEN All rights of exploitation in any form and by any means reserved worldwide for CEN national Members.


Ref. No.:CWA 16200:2010 E
CEN
WORKSHOP
AGREEMENT


CWA 16200

September 2010


ICS 35.040
English version

A Guide to the Development and Use of Standards Compliant
Data Formats for Engineering Materials Test Data

This CEN Workshop Agreement has been drafted and approved by a Workshop of representatives of interested parties, the constitution of
which is indicated in the foreword of this Workshop Agreement.

The formal process followed by the Workshop in the development of this Workshop Agreement has been endorsed by the National
Members of CEN but neither the National Members of CEN nor the CEN Management Centre can be held accountable for the technical
content of this CEN Workshop Agreement or possible conflicts with standards or legislation.

This CEN Workshop Agreement can in no way be held as being an official standard developed by CEN and its Members.

This CEN Workshop Agreement is publicly available as a reference document from the CEN Members National Standard Bodies.

CEN members are the national standards bodies of Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia,
Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland,
Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Switzerland and United Kingdom.

CWA 16200:2010 (E)

2
Contents


Foreword .......................................................................................................................................................... 3
0. Introduction ............................................................................................................................................... 5
1. Scope ....................................................................................................................................................... 10
2. Normative References ........................................................................................................................... 11
3. Abbreviations and Definitions .............................................................................................................. 12
4 Technology Review ................................................................................................................................ 14
5 Standards-compliant Schemas ............................................................................................................. 19
6 Standards compliant Vocabularies and Ontologies ........................................................................... 26
7 Mappings Between Existing Schemas and Ontologies...................................................................... 67
8 Business Analysis .................................................................................................................................. 88
9 Standards Development ........................................................................................................................ 98
10 Conclusions ...................................................................................................................................... 105
11 Recommendations ............................................................................................................................ 108
Annex A Proposed Informative Annex for EN ISO 6892-1:2009 (informative) ................................ 112
Annex B Review of Modelling and Schema Technologies (informative) ........................................ 119
Annex C ISO 10303 (STEP) Technology Review (informative) ........................................................ 120
Annex D ISO 10303-235 Example (informative) .................................................................................. 125
Annex E Schema RI—XSD-compliant Example Data Set (informative) .......................................... 133
Annex F ECCC—Industry Experience in Data Collation and Exchange (Informative) .................. 136
Annex G Ontology RI—Ontology-compliant Example Data Set (informative) ................................ 138
Annex H Future SC4 Architecture (informative) ................................................................................ 146
Annex I Ontology Review (informative) .............................................................................................. 148
Annex J Web Architecture (informative)............................................................................................ 155
Annex K Business Cases in Other Domains (informative)............................................................... 157
Annex L Business Survey Protocol (informative) ............................................................................. 159
Annex M Business Survey Results (informative) .............................................................................. 163
Annex N Review of the ISO Process for Electronic Inserts (informative) ....................................... 166
Annex O Future Development of Materials Data Standards (informative) ...................................... 169
Annex P Dissemination Activities and Informal Liaisons (informative) ......................................... 170
Bibliography ................................................................................................................................................. 177

CWA 16200:2010 (E)

3
Foreword

The production of this CWA (CEN Workshop Agreement), Economics and Logistics of Standards-Compliant
Schemas for Interoperability of Engineering Materials Data, was formally accepted at the Workshop's kick-off
meeting on 12th May 2009, held at CEN, Brussels.

The document was developed through the collaboration of a number of contributing partners in the CEN WS
ELSSI-EMD, including universities, digital curation centres, industry, consultants, software houses. The
CEN Workshop was active from May 2009 until June 2010.

A period of public comment was held during March-April 2010. The final text of the CWA was formally
endorsed at the Workshop final meeting, 27
th
May 2010, and following an electronic round of comments in
June 2010. The name of companies/organizations, which endorse the CWA, is listed hereunder.

Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. CEN [and/or CENELEC] shall not be held responsible for identifying any or all such patent rights.


Project team (PT)

Dr. C. Bullough, Alstom Power (chair)
Dr. T. Austin, SOASYS ltd (project leader)
Dr. D. Gagliardi, Manchester Institute of innovation Research at the University of Manchester
Mr. M. Loveday, Beta Technology
Mr. D.Leal, Caeser Systems

The CEN WS ELSSI Project Team (PT) expresses its thanks to Dr Chris Bullough for his active participation
in the work and his expert guidance.

Finally, the contributions from the registered participants, experts in the domains of engineering materials,
data curation, and standardization, to the work of CEN WE ELSSI are gratefully acknowledged.

NOTE While a decision on CEN hosting the data formats that CEN WS ELSSI-EMD delivers is pending, the
links to the resources at HTTP URIs beginning http://www.cen.eu/cen/cwa/elssi-emd/ will be unavailable. Until such
time as the data formats are published as HTTP URIs, the CWA includes an electronic copy (ZIP file) of the ontology,
the schema, and the examples.

Companies supporting the current CWA

Registered participants:

Airbus France SAS
Alenia Aeronautica S.p.A
Alstom Power
ASD-STAN
ASTM International
CAESAR Systems
CESI (China Electronics Standardization Institute)
Digital curation center University of Edinburgh
Doosan Babcock Energy Limited
EC JRC Institute for Energy
Exova
Federal Institute for Materials Research and Testing (BAM)
Granta Design Limited
High Temperature Mechanical Testing Committee
Hydro Aluminium Deutschland GmbH
Imperial College
IncoTest
Instron – A Division of ITW Ltd
MSC.Software Corporation

CWA 16200:2010 (E)

4
MDAO Technologies Ltd
National Physical Laboratory
SOASYS Ltd
The University of Queensland, School of ITEE
Toyo University, Faculty of Regional Development Studies
University of Manchester
University of Southampton
Volvo Aero Corporation
VZLÚ a.s. (Aeronautical Research and Test Institute)

This CEN Workshop Agreement is publicly available as a reference document from the National Members of
CEN: AENOR, AFNOR, BSI, CSNI, CYS, DIN, DS, ELOT, EVS, IBN, IPQ, IST, HZN, LVS, LST, MSA,
MSZT, NEN, NSAI, ON, PKN, SEE, SIS, SIST, SFS, SN, SNV, SUTN and UNI.

Comments or suggestions from the users of the CEN Workshop Agreement are welcome and should be
addressed to the CEN-CENELEC Management Centre.

CWA 16200:2010 (E)

5
0. Introduction

0.1 Overview

The engineering community invests significantly in generating materials test data of a high inherent value.
Very often, the data sets are richly structured and amenable to reuse. The materials community has
however, largely failed to address the issue of data capture and preservation. Although technologies for the
automated capture and preservation of test data exist, on the rare occasions that data are conserved, they
are invariably inaccessible to the wider materials community. This inevitably acts as an obstacle to the
research process, and hinders business activities in the engineering sector. In recognition of these issues,
CEN (the European Committee for Standardization) sponsored the 12-month ELSSI-EMD Workshop to
develop schemas and ontologies derived from procedural materials testing standards. With the emergence
of a Semantic Web of data, the project aimed to emulate other branches of the sciences that are developing
and leveraging Web technologies to their advantage, and that of associated business sectors. Beyond
developing schemas and ontologies for a chosen test type that complement existing specifications, the
Workshop investigated their role in promoting the capture and conservation of experimental data in the
materials sector, the opportunities that may arise for new and improved business, and the viability of
appending the schemas and ontologies to their corresponding materials testing standards. This CWA reports
the work performed and the findings of the ELSSI-EMD Workshop.

0.2 Requirements for Engineering Materials Data

Documentary Standards are an essential tool for underpinning virtually all aspects of society. In the
engineering sector, material testing standards play a vital role in ensuring that the design of structures,
monitoring of safety critical components, and the certification of materials for product release are all based
on an agreed and validated method for determining material properties. So far, paper-based standards have
been used to control the testing procedure, and paper certificates or reports are commonly used for reporting
and storing such data. Various stakeholders using the documentary Standards directly, or the data so
obtained, identify the need for greater interoperability of the data, not least in consistency of testing method,
storage of tests results, and usage of those results in a wide engineering context.

It is contended that there are significant benefits that may be accrued from developing documentary testing
and calibration Standards used for the determination of materials properties into formats that allow direct
interoperability with computers and computer-controlled facilities. This would allow such Standards to be
used to set-up mechanical testing machines and allow the measured output to be transferred directly to
material property databases or data processing tools, or enable the material properties to be directly
uploaded to product release certificates.

To understand the full benefits that would accrue from improved interoperability, it is important to consider
both the mechanics of use of the Documentary standards, and the subsequent usage of the data by key
stakeholders. One of the most ubiquitous materials tests is the tensile test, typically applied using the
documentary standard EN ISO 6892-1:2009. Often it is applied to qualify a particular product against a
material specification. A material test certificate produced in that way can have several uses, as illustrated
by a simplified ship-building example in Figure 1 - The use of a material test certificate in ship building
and operation
Here it is used not only to provide confidence that the ship plate has been produced to the required
specification, but it may also be used to demonstrate the quality processes employed during ship building. In
each of those transactions information (that is, data) is taken from the store of the sender, transmitted and
re-stored by the recipient.

CWA 16200:2010 (E)

6
Order plate
Create ship plate
Test product
Deliver plate
Fabricate
ship hull
Deliver ship Insure ship
Operate ship
Inspect &
Maintain
Order ship
Material Test Certificate
Legal document
Qualifies product
is a
has role
Define plate
specification
Determine
release
properties
Confirm
plate
properties
Confirm ship-
building quality

Figure 1 - The use of a material test certificate in ship building and operation
In general, there are three groups of stakeholders using data obtained from a mechanical test on a material

Within the supply chain - to demonstrate product quality assurance
During R&D and Design data evaluation – to develop and characterise materials
Structural Analysis / Life Cycle Analysis – use of assessed materials data calculate component life.

This is illustrated in Figure 2 - Stakeholders and their use of engineering materials data
:
• Who is involved & why?
• Quality Assurance, focus:
− Specifications
− Test certification
− Material Identification
• R&D Community, focus:
− Alloy development
− Characterisation
− Design data publishing
− Remanent life techniques
• Analytical, focus:
− Elastic, inelastic analysis
− Remanent life
− Life cycle analysis
• Manufacturing, focus:
− Cost, interoperability
Stakeholder
Quality
Assurance
R&D / Data
Asses't Analytical
R&D (Univ, Collaborations)
Publishers
- Print
- atab se
Alloy Producers
Materials Test L b
Stockholder
Sub-Contract HT/Manf
Manufacturer (OEM)
- Purchase Dept
- Stores
- Factory
- Materials
- Design
- Service ??
Customer
- Purch se Dept
- Service
3rd Party Service ??
Life Cycle Analysis
Sub-Community

Figure 2 - Stakeholders and their use of engineering materials data

CWA 16200:2010 (E)

7
The same documentary procedural standard for the mechanical test is used by each group of stakeholders
but they require the data for different reasons, often in different formats and to differing levels of detail,
abstraction and complexity.

One potential barrier to the interoperability of materials data is that the vocabulary and processes involved
with computer technology are unfamiliar to those in the materials testing domain. Terms such as ‗schema‘
and ‗ontology‘ are used representing the process outlined in the documentary standard in a computer
compatible format either in a schematic form (schema) or in a computer mark-up language (ontology). Thus
whilst present users of the procedural standards will recognise the benefits that will accrue, it is important to
recognise that technologies outside of their normal experience must be explained clearly, and without
recourse to jargon.

0.3 CEN Support and General Approach

Under the auspices of a CEN (European Committee for Standardization) Workshop, funding has been
forthcoming from the European Commission to evaluate the benefits that may be accrued from developing
documentary testing Standards used for the determination of materials properties into formats that allow
direct interoperability with computers and computer-controlled facilities. The CEN Workshop is entitled
‗Economics and Logistics of Standards-Compliant Schemas for Interoperability of Engineering Materials
Data‘ (ELSSI-EMD), and this CWA (CEN Workshop Agreement) reports its findings and recommendations.

The work reported in this document is motivated by the opportunity to deliver improved data management in
the engineering materials sector. Although the facilities used to generate and process test data have
evolved to be predominantly computer-controlled, there has not been an accompanying evolution of
standard data formats. Consequently the processes for generating, processing, and storing information are
not at all well integrated, and instead of a digital infrastructure, the materials community is simply left with a
collection of largely stand-alone, isolated systems. Although aggregating data from different sources is
possible, it is far from straightforward simply because of differences in data formats. Standard data formats
resolve this issue and allow an opportunity to be seized that has so far eluded the materials community,
namely the realization of a seamless infrastructure of computer-controlled facilities in anticipation of
improved business processes and more effective research.

In recent years, advances in information technology and increasing pressure for more effective business and
research practices has made the requirement for effective data management more a necessity than an
option. CEN WS ELSSI-EMD delivers a new and perhaps radical approach, but one that offers prospects for
a solution that has long evaded the engineering materials community. By developing computer-readable
versions of existing technical standards for mechanical testing, ELSSI-EMD aims to deliver data formats that
mitigate against obsolescence, that are maintainable, and that find widespread adoption. All these criteria
are met exactly because the technical standards to which the data formats comply are themselves already
maintained and accepted by the engineering materials community.

This document is a guide to using and developing computer-readable data formats that comply with technical
standards for mechanical testing. It focuses specifically on the EN ISO 6892-1:2009 ambient temperature
tensile testing standard. The scope of the work is defined entirely by this written testing standard, so that not
only the results of performing the test are computer-readable, but also all the metadata that are defined in
the written standard, including conditions, specimen and machine configurations, and constraints, all of
which are required to put the results of applying the standards into context.

The expectation is that technical standards in the engineering sciences can be translated to computer
readable reference implementations in the same way that specifications in the ICT find realization as
reference implementations. From another perspective, the work can simply be seen as delivering a
computer-readable equivalent to a natural language translation of the technical standard.

As with any translation, the primary objective is to remain true to the original meaning, and this is the
governing factor when implementing a computer-readable version. Beyond this requirement, there are other
considerations driven by contemporary IT paradigms, namely object-orientation and reuse. From the
perspective of object-orientation, the implications for developing a format that is compliant with a technical
standard for mechanical testing are that common entities need to be identified and form a base data model
that can then be specialized for information that is characteristic to an individual test. For tests on
engineering materials, properties data tend to be unique to individual test types, such as fracture toughness
for fracture mechanics tests, yield and proof strengths for tensile tests, and time to rupture for creep, while
ancillary information, such as production route and specimen type are common.

CWA 16200:2010 (E)

8

In terms of reuse, there is a requirement to align the work to existing data models and modelling standards.
Those relevant to the work reported in this document include ISO 10303 Parts 41, 45, and 235, MatML,
JRC MatDB, and NMC MatDB.

While the work mirrors the development of natural language editions of the technical standard, the
implications of delivering computer readable versions of technical standards are far more profound,
especially with respect to the impact on business models in both the industrial and standardization sectors.
This CWA anticipates the possible impacts, and offers guidance on how to proceed to promote the adoption
of standards-compliant data formats.

0.4 Motivations for Improved Data Management

0.4.1 Business Sector

From finance to health care, the electronic transfer of information, commonly termed electronic data
interchange (EDI), has had a significant impact on business practice
(http://www.aberdeen.com/c/report/sector_insights/5097-SI-supplier-enablement-enterprise.pdf ). In EDI, the
computer-to-computer exchange of business data relies on information being organized according to a
standard format recognized both parties, allowing computer transactions that require no human intervention.
All information contained in an EDI transaction typically the same as that of a conventionally printed
document. Some of the recognized benefits of EDI include reduced cycle time, increased productivity,
reduced costs, improved accuracy, improved business relationships, and minimized paper use and storage.

Standards-compliant data formats offer an opportunity to realize the benefits of EDI in the engineering
sector. Besides facilitating improved productivity, a move away from paper-based transactions and archiving
can reasonably be expected to counter accidental data loss, which can be a serious issue considering the
high inherent worth of test data. It can also be expected that standards-compliant data formats will lead to
new business opportunities, improved auditing, and greater traceability. As demands on the industrial sector
for greater accountability become ever more stringent, these are important considerations. Examples
include the long-term preservation of conformance certificates and increased data quality control, as
manifested in the ISO 8000 family of standards. The first ISO 8000 standard was published in 2008 as a
Technical Specification (TS) entitled Data quality -- Part 110: Master data: Exchange of characteristic data:
Syntax, semantic encoding, and conformance to data specification. ISO/TS 8000-110:2008 specifies
general, syntax, semantic encoding and data specification requirements for master data messages between
organizations and systems. The focus is on requirements that can be checked by computer. Further details
available from http://www.iso.org/iso/catalogue_detail.htm?csnumber=50800. Other Technical Specifications
published in 2009 include part 100 (introduction), part 120 (provenance), part 130 (accuracy) and part 140
(completeness). If ISO 8000 follows a trend similar to ISO 9000 in terms of industry adoption, then it can
reasonably be expected that in the coming years, data management will become a primary concern for
organizations in the industrial sector.

0.4.2 Research Sector

Although the motivations are fundamentally different to those of business, the research sector now
recognizes the fundamental role that access to research data in a viable research infrastructure. Under the
auspices of Knowledge Exchange partnership (http://www.knowledge-exchange.info), agencies from the UK
(www.jisc.ac.uk and www.dcc.ac.uk ), Germany (www.dfg.de), and Denmark (www.deff.dk), and The
Netherlands (www.surf.nl) are leading their efforts on the conservation and reuse of data. In the UK, BBSRC
has a stated data sharing policy that makes effective data management a prerequisite for obtaining funding
[11]. EPSRC has adopted a similar, if less stringent policy. Details of these and other data management
policies are available from the DCC at http://www.dcc.ac.uk/resources/policy-and-legal/data-management-
plans.

At the European level, the European Commission invests significantly in research that generates materials
data of an high inherent worth, little of which remains available to the broader engineering community. With
the introduction of an Open Access Pilot initiative for selected FP7 projects [14], it is clear that the European
Commission is also favouring the improved management of research output.

In the context of improved data management, the research sector also acknowledges the fundamental role
of standards to codify the boring, so that the exciting can happen on top of them [26,29]. Again the natural

CWA 16200:2010 (E)

9
and life sciences are developing a robust foundation, leveraging standards and technologies
(http://www.biosharing.org/ and http://esw.w3.org/topic/HCLSIG).

A recent DCC SCARP Synthesis Report [17] provides a critical review of the disciplinary differences in
research data sharing and data management. While the natural and life sciences have embraced and
benefited from efforts to conserve and share data, the engineering disciplines have a very poor record [10].
This inevitably hinders research in engineering sciences. Although it is something of an anomaly that the
engineering sciences have such a poor record in research data management, when its students are highly IT
literate and generate significant volumes of data, with projects such as MDC (
http://www.materialsdatacentre.com ) and EP2DC ( http://wiki.eprints.org/w/EP2DCOverview ), there are
indications, the engineering sciences are beginning to demonstrate an appreciation of the importance of
effective data management.

0.4.3 Standardization Sector

The European Commission White paper on ICT Standardization [13] supports the adoption of the more agile
standardization development process required in the ICT sector. In anticipation of the demands for standard
data formats, CEN WS ELSSI-EMD is pioneering the introduction of processes that will facilitate the
conventional Standards bodies contributing their expertise to the needs of the ICT and related sectors.

0.5 The Role of Ontologies and Schemas

There are a number of aspects whereby it would be beneficial if testing standards were to have compatible
interoperability with the computer systems that control the test machine and the output from the testing
machine software used to determine the material properties. In addition, it would be possible to transform
the data directly into product release certificates and other types of reports. In the case of EN ISO 6892-
1:2009 and other mechanical testing standards, the potential benefits include:
Improved interoperability—data formats compliant with testing standards will allow the seamless
transfer of test data between computer-controlled facilities, such as test machines and materials
databases.
Improved data conservation—data formats compliant with testing standards will allow a complete
records to be conserved, an important consideration in light the ever more stringent auditing and
traceability demands placed on the manufacturing sector.
New and improved business processes—data formats compliant with testing standards allow for
greater ease of use and manipulation of data, such as transformation into electronic material product
release certificates.
Opportunities for new and improved research—ever increasing volumes of high quality test data
will facilitate data mining and pattern discovery, and allow model validation and development.
Greater rigour in the development of testing standards—the breadth of knowledge of materials
engineers in combination the precise definitions demanded by ICT specialists combines to deliver a
combination of competencies that is particularly well suited to the process of developing robust
mechanical testing standards.

CWA 16200:2010 (E)

13
WDSL Web Services Description Language
XML Extensible Markup Language
XSD XML Schema Definition
XSLT XML Stylesheet Language Transformation

3.2 Definitions

Data format—implementable specification for the representation of data

NOTE 1 A conceptual schema is not a data format because it is not intended to be implementable.

NOTE 2 An EXPRESS model and the ISO 10303-21 encoding specification taken together define a data
format.

NOTE 3 An XML Schema is a data format.

NOTE 4 An ontology on its own is not a data format. However, an RDF vocabulary and the XML RDF
serialization taken together define a data format. An ontology can be also an RDF vocabulary.

Dereferenceable URI—resource retrieval mechanism that uses any of the internet protocols to obtain a
copy or representation of the resource it identifies.

NOTE 1 The commonly used resource retrieval mechanism is HTTP.

NOTE 2 Where the resource is a document, a copy can be obtained. Alternatively the representation can
be information about the document, such as where it can be purchased.

NOTE 3 The representation of a resource can be computer interpretable statements about a resource in
RDF.

Ontology—a semantically precise and computer processable definition of things and their relationships.

Refactor (information technology)—improve existing software by changing its method of implementation

Schema—complete description of the structure of a data base pertaining for a specific level of consideration
(from ISO/IEC 2382-17 Information technology – Vocabulary – Databases).

NOTE Schema describes structure, relationships, and rules, and can be represented using natural
language text, diagrams, or a computer-readable format.

CWA 16200:2010 (E)

14
4 Technology Review

4.1 Overview

Together with large collections of reference data hosted by international standards and measurements
bodies, there are a number of materials and other engineering properties schemas available that to a greater
or lesser extent provide the opportunity to deliver a computer-readable representation of a tensile test. The
existing schemas include ISO 10303-235, MatML, and JRC MatDB, and each has a different scope and
objectives. The choice of the most appropriate technology will depend on a range of considerations,
including whether the data is test or design data, the extent of community adoption, the capacity of the
technology to describe accurately the data and associated metadata, and the capacity of the technology to
describe the provenance of the data.

4.2 Standard Materials Reference Data

As evidenced by the large collections of data curated by different standards and measurements bodies
worldwide, the importance of standard engineering materials reference data is well-recognized. Examples
include:

NPL—the UK National Physical Laboratory hosts collections of materials data, including engineering
materials test data (retrieved March 16, 2010 from http://www.npl.co.uk/advanced-materials/measurement-
techniques/mechanical/tensile-testing-standards-and-tenstand).

EC-JRC-IE—the European Commission Institute for Energy hosts a wide range of engineering materials
data at its ODIN online data information network (retrieved March 16, 2010 from
https://odin.jrc.ec.europa.eu).

MMPDS (formerly MIL-HDBK-5)—the pre-eminent source for aerospace component design allowables. At
the time of writing, the latest release is MMPDS-04 (retrieved March 16, 2010 from
http://projects.battelle.org/mmpds).

ASM—the American Society for Metals hosts a wide variety of engineering materials data at ASM Materials
Information (retrieved March 16, 2010 from http://products.asminternational.org/matinfo/index.jsp).

NIMS—the Japanese National Institute for Materials Science hosts a range of engineering materials
databases and publishes engineering materials data sheets at its MatNav site (retrieved March 16, 2010
from http://mits.nims.go.jp/db_top_eng.htm).

4.3 ISO 10303 (STEP)

4.3.1 Major components of the ISO 10303 series of standards

ISO 10303 is an ISO standard for the computer-interpretable representation and exchange of product
manufacturing information. It is a suite of standards that includes:

The EXPRESS information modelling language (ISO 10303-11)—this is a data modelling language
which is particularly suitable for engineering data management applications. EXPRESS has good
support for complicated mathematical data structures found in engineering, and has a powerful
capability for defining constraints which ensure the completeness and consistency of data.
Implementation methods (20 series parts)—these parts define a text representation for an instance
of an EXPRESS schema (part 21 file); an API for a data base implementation of an EXPRESS
schema; and how an XML Schema can be generated from an EXPRESS schema.
Generic and integrated resources (IRs – 40 and 100 series parts)—these are small information
models which can be used in Application Models or Application Protocols to satisfy an industrial data
requirement.
Application protocols (APs)—an AP defines the information requirements for an industrial data
application area and defines a EXPRESS schema, created from the IRs which satisfies the
requirements. An AP can define conformance classes which define the precise content of an

CWA 16200:2010 (E)

15
information flow within an industrial application area. An AP can specify XML schemas for its
conformance classes.
Application modules (AMs)—an AM defines the information requirement within a very small
industrial data application area, and an EXPRESS schema which satisfies them. An AP can be
formed as a assembly of AMs.

See Annex B for more information on the ISO 10303 modelling and implementation languages.

4.3.2 ISO 10303 specifications relevant to mechanical testing

ISO 10303-41—this IR contains information models for fundamentals of product description and support. It
defines a collection of base entities that find widespread application within AMs and APs.

ISO 10303-45—this IR contains an information model for material and other engineering properties. It
defines the entities found in the engineering materials sector that can be used to develop APs for specific
engineering materials domains.

ISO 10303-235—this is an AP for the representation of engineering properties for product design and
verification. It contains a model for the representation for any property of a product and its value, specifying
the collection of processes by which a value of a product property is obtained [37,38].

The scope of ISO 10303-235 is described in detail in Annex D.

4.3.3 Implementation of an ISO 10303 AP

An ISO 10303 AP specifies and EXPRESS schema for the information that is shared or exchanged for within
an industrial data application area. This schema defines directly a representation of the information as a part
21 file.

An example of a Part 21 file for an instance of the ISO 10303-235 schema is shown in Annex D.

Where an AP defines a conformance class for a particular information flow within an industrial application
area, a commercial software package which manages data for that industrial data application area can have
an ISO 10303 import and export capability. This capability will enable data exchange between different
commercial software packages that implement the same conformance class.

Reliable automatic exchange requires great care in the definition of conformance classes to remove any
ambiguity in the way in which information is recorded. The programme of testing necessary to achieve
reliability is described in Annex C.6.

4.3.4 Merits and Limitations of STEP

STEP provides a powerful integrated approach to designing a file format for the exchange of data between
engineering activities. Many of the data design techniques, which are now commonly used in industry, were
pioneered by the STEP community.

Unfortunately, STEP is also paying the price for being a pioneer. The EXPRESS and part 21 file
technologies are not widely used outside of STEP (in spite of their effectiveness). The equivalent UML and
XML technologies are far more widely used.

In some areas, the STEP approach to semantics relies upon:
entities in the integrated resources which are either pure data structures or which have only loose
semantics, but which have attributes that can be used to make the semantics precise; and
a mapping table in an AP which specifies how concepts within an engineering domain are
represented by the entities.
This approach works, but is complicated.

A more common approach today would be based upon a single ontology or data model with layers. At the
top would be generic concepts, such as ―material object‖ – something you can kick, and ―activity‖ –
something that happens. Below this would be layers of specialisation, such as:

CWA 16200:2010 (E)

16
a layer for testing in general with specialisations, such as ―test specimen‖ or ―tensile test‖;
a layer for EN ISO 6892-1:2009 with further specialisations, such as ―tensile test to EN ISO 6892-1‖.

4.3.5 The future of STEP

STEP is the key standard in many areas of engineering design. This is especially true for those areas in
which product shape information is essential. Most CAD systems have a ―STEP button‖, which exports
design data in STEP format.

However some new projects have been reluctant to embrace the STEP methodology because it is old
fashioned in some respects and complicated. ISO TC 184/SC4, the parent committee for STEP, has
launched a project to develop a new architecture which will:
complement the existing integrated resources with ontologies;
allow the creation of a single data model instead of an ARM and AIM;
allow the use of UML as well as EXPRESS.

See Annex C for a more in depth review of ISO 10303 technologies and Annex H for an overview of the
future SC4 architecture.

4.4 MatML

4.4.1 General
MatML (Materials Markup Language) is the result of a NIST initiative to develop a data format specifically for
the interchange of materials information [42].

MatML is the only materials schema that is proposed as a standard. Initially presented as a materials
property markup language, the Public Review Draft 01, 06 June 2006 (retrieved March 16, 2010 from
http://docs.oasis-open.org/materials) of MatML is for a Materials Markup Language.

4.4.2 MatML Scope, Evaluation and Adoption

MatML is intended to allow existing information sets, available for example as publications and data sheets,
to be translated easily into a machine-readable data format. In addition to materials properties, its scope of
data coverage extends to data provenance and materials metadata, including material composition, heat
treatment, and production, and its possible scope of application extends from the aerospace industry to
emissions monitoring [42]. Various initiatives to use and trial MatML indicate that there is a demand for a
standard materials schema:
Several commercial applications, including Materiality (retrieved March 16, 2010 from
http://www.matereality.com/MMPDS-02.htm) and Granta MI (retrieved March 16, 2010 from
http://www.grantadesign.com/products/mi/exporters.htm) support an MMPDS (formerly MIL-HDBK-5)
MatML export format.
The ANSYS Workbench Engineering Data module (e.g., Workbench Simulation, Workbench FE
Modeler) can import/export material data as XML files. These XML files utilize the MatML format, and
ANSYS provides an example (retrieved March 16, 2010 from http://ansys.net/tips/Material_XML.zip).
This can be useful for users wishing to generate their own material libraries in XML format.
The ESI Group posting at the OASIS Materials Mailing List Archive (retrieved March 16, 2010 from
http://lists.oasis-open.org/archives/materials) indicates that the business sector anticipates a standard
materials schema adding value to materials business processes.

Some of the conclusions and recommendations from MatML trials include:
Storing curves as equations
Data type called equation
Store multiple parameters with property data
An attribute is required to a PropertyDetail ref Glossary ID
Support to attach a drawing to an XML document
Nesting instead of referencing of classes

CWA 16200:2010 (E)

18
4.6 NMC MatDB

4.6.1 General

NMC MatDB (retrieved March 16, 2010 from http://www.nims.go.jp/vamas_twa10/AMM_DB/XML-
Schema.LZH) was developed in the context of VAMAS TWA 10 by the New Materials Centre (NMC) as an
extension to MatML version 3.0. It is interesting because it includes a collection of schemas for common test
types, as follows:
Bend Test
Compression Test
Crack Growth
Creep Test
Fatigue
Fracture Toughness
Hardness
Impact Test
Shear Test
Tensile Test
Torsion Test

Where there is some ambiguity in the schema design concerns the integration of the test type schemas with
MatML. The test type schemas replicate entities that are defined in the MatML schema, the implication
being that the test type entities override those of the MatML schema.

4.6.2 NMC MatDB Evaluation and Adoption

There is no evidence of NMC MatDB having been used other than in the context of materials data
management research.

4.6.3 Merits and Limitations of NMC MatDB

Merits:
Good test coverage

Limitations:
No evidence of adoption
Based on an obsolete version of the MatML schema
VAMAS TWA 10 is closed and it appears that work to develop and maintain NMC MatDB has ceased

CWA 16200:2010 (E)

21
The two approaches are complementary, and it is reasonable to use both - perhaps carrying out stages (1)
and (2) of one approach and then validating the result using the other.

The architecture for ISO 10303 specifications requires three stages, which are very similar to those defined
above using UML:
The definition of the things of interest to users of the Application Protocol in natural language;
The representation of these things as an Application Reference Model (ARM);
The mapping of the ARM to the Application Interpreted Model (AIM) written in EXPRESS.

The final stage in this process is similar to mapping a UML class model to the AIM, and so the final stages in
all described cases a very similar, and so a UML class diagram can provide a common construct for
establishing consistency across the technologies relevant to the work of CEN WS ELSSI-EMD.

NOTE The ISO 10303 approach to schema development is described in more detail in Annex C. ISO
10303 uses the EXPRESS data modelling language, which is particularly suitable for data exchange applications.
The EXPRESS language has a powerful capability for applying rules which can ensure that data are complete
and consistent.

5.3.3 Modelling Mechanical Testing Standards

To a certain extent, the structure of a mechanical testing standard is already defined by its table of contents
(TOC). To develop a standards-compliant schema, the following procedures is recommended:
Define a schema that mirrors that of the TOC, using its clauses and subclauses of the technical
standard to identify classes and their associations
For all classes, identify the associated attributes
For all classes, identify constraints and additional parameters
Consolidate the schema:
Remove redundancy (meaning duplication of the same parameters in different classes)
Refine the class hierarchy
Identify the multiplicity of associations

To assist future development and maintenance, the schema is to be properly documented, so that all
classes, attributes, and constraints are accompanied by references to the technical standard, preferably by
reference to specific paragraphs.

Validation and verification of a proposed normative schema specification will necessarily involve materials
partners undertaking data collection and exchange exercises to ensure the schema meets their
requirements.

5.4 UML Class Diagrams for Uniaxial Tests

5.4.1 Classification

The abstract class diagram depicted in Figure 3 is proposed for uniaxial tests on engineering materials.


Figure 3 - Abstract class diagram for a uniaxial test on an engineering material

CWA 16200:2010 (E)

22
According to the methodology described in Clause 5.3.3, Figure 4 depicts the class diagram proposed for EN
ISO 6892-1:2009. This diagram extends the abstract model in Figure 3 with classes identified from close
inspection of EN ISO 6892-1:2009.


Figure 4 - Class diagram for EN ISO 6892-1:2009 tensile test

Although not shown in Figure 4, each class also includes properties that are characteristic of EN ISO 6892-
1:2009. It is expected that the same procedure of extending the abstract model with characterizing classes
and properties could be applied to any uniaxial testing standard.

Based on the above class diagram and the documentary EN ISO 6892-1:2009 standard, an XML Schema
(XSD) reference implementation (RI) is available at http://www.cen.eu/cen/cwa/elssi-emd/schema/iso-6892-
1.xsd. The RI is annotated using entries from EN ISO 6892-1:2009 Clause 3 (Terms and Definitions).

NOTE While a decision on CEN hosting the data formats that CEN WS ELSSI-EMD delivers is pending,
the links to the resources at HTTP URIs beginning http://www.cen.eu/cen/cwa/elssi-emd/ will be unavailable.
Until such time as the data formats are published as HTTP URIs, the CWA includes an electronic copy (ZIP file)
of the ontology, the schema, and the examples.

See Annex E for an example data set that complies with the XSD reference implementation.

5.4.2 Constraints (Business Rules)

From the perspective of developing robust and scalable software solutions, a thorough knowledge of
business rules is equally, if not more, important than a knowledge of the structure and relationships that
characterize the problem domain. In a software application, business rules, which are typically coded
directly into the business tier or accessible through a rules engine, allow logical decision making. This is
important in the context of standards-compliant formats for mechanical testing because by identifying and
encoding the business rules, there is the opportunity to validate test data automatically and thereby establish
to what extent they comply with the procedural standard.

NOTE It is too simplistic to expect that a validated data set will either comply or not comply with EN ISO
6892-1:2009. Instead, it should be expected that a degree of compliance will be determined from which the
quality and reliability of the data set can be established.

CWA 16200:2010 (E)

29

6.3.3 Methodology

The activities necessary to create an implementable ontology from a material test standard are shown in
Figure 6.

ISO 23718:2007
―Metallic materials –
Mechanical testing –
Vocabulary‖
framework
of
engineering
concepts
ISO 10303 (STEP)
SysML
ISO 18629 (PSL)
ISO 15926
NIST Product information
modelling framework
material
testing
ontology
ISO 10303-235
MatML
JRC MatDB
ISO 6892-1:2009
―Metallic materials --
Tensile testing --
Method of test at room
temperature‖
tensile
testing
ontology
ISO 6892-1:2009
―Metallic materials --
Tensile testing --
Method of test at room
temperature‖
tensile test
theasurus
material
testing
thesaurus
input from the data modellers and ontologists an extension to existing standardization practice
creating an implementable domain ontology
A new type of standardization task

Figure 6 - Activities necessary to create an implementable ontology

Here, the phrase ―implementable ontology‖ means an ontology that is sufficient to support the representation
of all the data about a material test required by a business activity.

Details of the different activities within this methodology are as follows:

input from the data modellers and ontologists: This is a framework of generic concepts which can be
used in many different areas of engineering and science. The framework will include:
Physical objects
Activities
Roles played by physical objects in activities
Composition and connection of physical objects
States and changes to physical objects with time
Physical quantities and units of measure

NOTE 1 ISO TC184/SC4 is moving slowly towards creating such a framework. At present there is the basis of
a framework hidden within existing standards such as ISO 10303, ISO 15926 and ISO 18629.

CWA 16200:2010 (E)

30
NOTE 2 When work on data standards began in the 1980s it was assumed that the schemas created by the
data modeller could be expanded to included many different domains, and that the work on the development of these
schemas would be managed by data modellers. This has proved difficult. The ―Future ISO TC184/SC4 Architecture‖
envisages the creation of a framework which is used by others, as shown in this methodology.

an extension to existing standardization practice: Existing standards are formal within the rules of paper-
based standard development. A minor extension to existing practices will enable the creation of
computer processable thesauruses: as follows:
the extension of the ―Terms and definitions‖ to include all concepts, and not only those for which the
natural language term is ambiguous;
the assignment of URIs to the concepts;
the provision of a computer processable representation of the concept identifiers, terms, definitions
and relationships.

EXAMPLE 1 The representation of the concept corresponding to the term ―tensile strength‖ defined in EN ISO 6892-
1:2009, using the SKOS (Simple Knowledge Organization System) vocabulary, would be as follows:

iso16892-1:tensile_strength
a skos:Concept ;
skos:prefLabel "tensile strength"@en ;
skos:notation "R_m" ;
skos:definition
"stress corresponding to the maximum force, F_m"@en ;
skos:broader iso16892-1:stress .

In this example, the prefix ―iso16892-1‖ stands for the URI namespace assigned by EN ISO 6892-1:2009.
A possible prefix is:

http://standards.iso.org/iso/6892/-1/tech/ ,

so that the full URI of tensile strength is:

http://standards.iso.org/iso/6892/-1/tech/tensile_strength .


creating an implementable domain ontology: This is a new task which requires a hybrid skill – some
knowledge of data modelling and ontologies, and some knowledge of the domain. It is probable that the
domain experts in material testing can perform most of this task, but some contribution from the data
modellers and ontologists will also be required.

The task also has two parts:

creating a simple ontology by relating terms in the thesaurus to the framework of engineering
concepts. This part is straightforward and is sufficient to create an implementable ontology.

analysing the constraints specified in natural language within a test standard and expressing those
constraints formally with respect to the ontology. This task is much more difficult, and has longer
term benefits, which include:
- the validation of the content of the test standard;
- the ability to deduce whether or not a test complies with the test standard from the information
recorded about the test.

These two tasks are discusses in more detail in the sections below.

6.3.4 From Thesaurus to Ontology

The concepts in a thesaurus can be turned into an ontology by the following steps:
identifying the concepts in the ontology that are classes;
defining a hierarchy of classes where the top is placed within a framework of generic engineering
classes:
identifying properties, which are relationships between classes.

CWA 16200:2010 (E)

32
rdfs:domain iso6892-1:Tensile_test_to_ISO_6892-1 ;
rdfs:domain test:Specimen_crosshead_extensometer_assembly_during_test ;
rdfs:range quomos:Stress .


where:
the prefix ―iso6892-1‖ specifies the namespace for an EN ISO 6892-1:2009 ontology;
the prefix ―test‖ specifies the namespace for a generic test ontology;
the prefix ―quomos‖ specifies the namespace for a generic ontology for quantities, units of measure
and scale.

The property EN ISO 6892-1:2009 tensile strength is an owl:FunctionalProperty because for any tensile
test it has no more than one value.

An ontology can be used to make statements about an individual.

EXAMPLE 2 Using the ontology in example 1, the following statement can be made about the individual
activity ―T_101‖ as follows::
T_101 is classified as a Tensile test to EN ISO 6892-1
T_101 gave a tensile strength value of 531 MPa.

These statements can be represented using N3 as follows:

http://www.MyLaboratory.de/tests/T_101
a iso6892-1:Tensile_test_to_ISO_6892-1 ;
iso6892-1:tensileStrength [ quomos:MPa 531 ] .

An equivalent representation using XML is as follows:

<iso6892-1:Tensile_test_to_ISO_6892-1
rdf:about="http://www.MyLaboratory.de/tests/T_101">
<iso6892-1:tensile_strength>
<quomos:Stress><quomos:MPa>531</quomos:MPa></quomos:Stress>
</iso6892-1:tensileStrength>
</iso6892-1:Tensile_test_to_ISO_6892-1>

6.3.5 Deriving Constraints from a Technical Standard

This is a much more difficult task than that outlined in section 6.3.4. The task requires analysing the natural
language text of the standard, and writing down the equivalent formal statements.

EXAMPLE Consider clause 5 ―Principle‖ of EN ISO 6892-1:2009. The full text of this clause is as follows:

―The test involves straining a test piece by tensile force, generally to fracture, for the determination of one or
more of the mechanical properties defined in Clause 3.

The test is carried out at room temperature between 10 °C and 35 °C, unless otherwise specified. Tests
carried out under controlled conditions shall be made at a temperature of 23 °C ± 5 °C.‖

This clause defines four subclasses of tensile test, as follows:
tensile test according to EN ISO 6892-1
tensile test according to EN ISO 6892-1:2009 at room temperature between 10 °C and 35 °C
tensile test according to EN ISO 6892-1:2009 at ―room temperature‖ outside the range 10 °C and 35
°C
tensile test according to EN ISO 6892-1:2009 under controlled conditions

NOTE the possibility of a tensile test to EN ISO 6892-1:2009 outside the temperature range 10 °C to 35 °C
appears to be permitted by the text "unless otherwise specified". This text can be interpreted as allowing a test to
be performed outside the usual temperature range, provided that a reason is given. A rule applied to a schema
for recording data about a tensile test to EN ISO 6892-1:2009 could require a reason to be given whenever the
recorded temperature is outside the range 10 °C to 35 °C.

CWA 16200:2010 (E)

33
The relationship between these classes is shown in Figure 7.
Tensile_test_to_ISO_6892-1
Tensile_test_to_ISO_6892-1_outside_10_to_35_degrees_C
Tensile_test
Tensile_test_to_ISO_6892-1_within_10_to_35_degrees_C
Tensile_test_to_ISO_6892-1_under_controlled_conditions

Figure 7 - Classes of tensile test defined in EN ISO 6892-1:2009

The relationship between these classes can be recorded formally using the RDF and OWL vocabularies, as
follows:

iso6892-1:Tensile_test_to_ISO_6892-1
a owl:Class ;
rdfs:subclassOf test:Tensile_test ;
owl:equivalentClass
[ owl:unionOf (
iso6892-1:Tensile_test_to_ISO_6892-1_within_10_to_35_degrees_C
iso6892-1:Tensile_test_to_ISO_6892-1_outside_10_to_35_degrees_C )
] .

iso6892-1:Tensile_test_to_ISO_6892-1_within_10_to_35_degrees_C
a owl:Class .

iso6892-1:Tensile_test_to_ISO_6892-1_outside_10_to_35_degrees_C
a owl:Class ;
owl:disjointWith
iso6892-1:Tensile_test_to_ISO_6892-1_within_10_to_35_degrees_C .

iso6892-1:Tensile_test_to_ISO_6892-1_under_controlled_conditions
a owl:Class ;
rdfs:subclassOf
iso6892-1:Tensile_test_to_ISO_6892-1_within_10_to_35_degrees_C .

The class Tensile test is in a generic testing ontology, upon which the EN ISO 6892-1:2009 ontology relies.
The subclasses for particular temperature ranges can be defined by using the property ambient
temperature from a generic testing ontology, as follows:

iso6892-1:Tensile_test_to_ISO_6892-1_within_10_to_35_degrees_C
owl:equivalentClass [ owl:intersectionOf (
iso6892-1:Tensile_test_to_ISO_6892-1
[ owl:onProperty test:ambient_temperature ;
owl:allValuesFrom :temperature_range_10_to_35_degrees_C ]
) ] .

CWA 16200:2010 (E)

34
The object temperature range 10 to 35 degrees C can be defined using a quantities and units of measure
ontology.

The creation of an OWL ontology from a natural language text requires some experience of ontology
development

6.4 Formal basis for a representation of information about a material test

6.4.1 Approach

The creation of an ontology, and a formal data model, relies upon the following steps:
1) identifying principal objects in the domain;
2) classifying the objects;
3) recording the relationships between them;
4) recording their physical properties (i.e. their relationships with physical quantities).

NOTE an alternative approach is to list a set of possible queries to a data repository and to define a data structure
with sufficient relationships to support them. This approach can lead to data structures which can be accessed
conveniently and efficiently. This approach can also lead to:
data structures within which the concepts have no relationship to reality;
data structures which can only support queries within the original list.

There are a many different types of test for which the principal objects are as shown in Figure 8.
T_101
specimen test machine
assembly
test laboratory
t
crosshead force at t
crosshead displacement at t
extensometer displacement at t
time
space
crosshead
S_101
exten om er

Figure 8 - Principal objects in a test

Figure 8 shows the following:

1 Test T_101 is an activity within which different things participate. These things include:
- the laboratory that performs the test;
- the test specimen S_101;
- the test machine and some of its parts.

NOTE 1 In order to keep the diagram simple, other parts of the test machine, such as the load cell are not
shown.

CWA 16200:2010 (E)

38
764 MPa
tensile strength to ISO 6892-1
specimen for test
source of specimen
specification for batch
tensile strength to ISO 6892-1
tensile strength to ISO 6892-1
T_101
B_101
P_101
S_101

Figure 10 - A naïve approach to batch and product properties
The different types of statement may be distinguished informally in a data model by ―context‖. Alternatively
the difference between the statements may be implicit, because it is assumed that a property of a test is a
measured value, whereas a property of a product is a prediction.

NOTE 1 This example has not addressed the issue of uncertainty. A value obtained for a test will be
uncertain because of limits to measurement accuracy. A value predicted for a batch or product has uncertainty
because of measurement accuracy, variability from test to test, variability from specimen to specimen within a
batch, and from batch to batch of a product type.

NOTE 2 The ontology only considers a single test. The testing of multiple specimens taken from the same
batch, and the assimilation of the different results, is not considered within the ontology.

6.5.2 Distinguishing between the types of statement using OWL

The use of an ontology enables the statements to be clearly distinguished because the use of an ontology
enables the text within statements 2 and 3 to be represented explicitly.

Consider statement 2. This statement defines the following classes:
the class S_B_101 that consists of all specimens taken from batch B_101;
the class T_S_B_101 that consists of all tests to EN ISO 6892-1:2009 on a member of class S_B_101.
There is then the assertion that each member of class T_S_B_101 has a tensile strength of 764 MPa.

The classes, and the assertion can be expressed using OWL as shown in Figure 11.

CWA 16200:2010 (E)

40
764 MPa
P_101
source of
specimen
onProperty
B
onProperty
allValuesFrom
specimen
for test
C
tensile
strength
onProperty
hasValue
Tensile test to
ISO 6892-1
intersectionOf
D
T_S_B_P_101
subclassOf
specification
for batch
onProperty
A
allValuesFrom
allValuesFrom

Figure 12 - A formal approach to a product type properties
It may be convenient to give this link, or property, a name such as ―tensile strength for product type‖.

The three properties have different domains, as follows:

property domain
tensile strength an individual tensile test
tensile strength for batch an individual batch of material
tensile strength for product type a class that is a material product type

It is probably useful to have a shorthand notation which can derived properties ―xx for batch‖ and ―xx for
product type‖ from a property ―xx‖ of a tensile test.

6.5.3 Distinguishing between the types of statement using ISO 15926

ISO 15926 has a formal approach which is close to that of OWL. This standard too can distinguish between
the types of statement. The ISO 15926 equivalent to Figure 11 is shown in Figure 13.

CWA 16200:2010 (E)

41
stress
765 MPa
batchspecimen
B_101S_B_101
source of
specimen
tensile
strength
T_S_B_101
tensile test
specimen
for test
z

Figure 13 - A formal approach to a batch properties using ISO 15926

6.6 Representation of the ontology

6.6.1 Separation of the ontology into layers

The ontology necessary to make statements about a test carried out to EN ISO 6892-1:2009 can be divided
into layers:
an extract from a core ontology for industrial data, which is concerned with material object, activities and
the relationships between them;
an ontology for testing in general, which is concerned with test activities, test specimens, and states of
the specimen, crosshead and extensometer during a test;
an ontology containing the concepts defined within EN ISO 6892-1:2009, especially properties such as
tensile strength, upper yield strength, etc., and control methods.

There is also a requirement for an ontology which can record physical quantities, units of measure, scales,
dates and times. The relationships between the ontology layers are shown in Figure 14.

CWA 16200:2010 (E)

42
ontology from ISO 6892-1
ontology for testing
ontology for
physical
quantities,
units of
measure,
scale, date
and time
ontology for industrial data


Figure 14 - Relationship between ontology layers

This document does not consider the ontology for physical quantities, units of measure, scale, date and time.
Different standards, such as ISO 15926 and ISO 10303, have their own approaches to these things. OASIS
Technical Committee ―Quantities and Units of Measure Ontology Standard‖ (QUOMOS) (http://www.oasis-
open.org/committees/tc_home.php?wg_abbrev=quomos) is developing an ontology in this area.

The ontology for testing,the ontology for EN ISO 6892-1:2009, and the extract from the ontology for industrial
data, are documented in clauses 6.7, 6.8 and 6.9.

The ontologies are shown as UML diagrams in clause 6.10.

6.6.2 Namespaces

The namespaces for the different ontologies defined in this document are as follow:

core: core ontology for industrial data
test: test ontology
iso6892-1: EN ISO 6892-1:2009 ontology

The following namespace is referenced but not defined:

quomos: ontology for quantities, units of measure, scales, data and time

The following namespaces are referenced to provide a link to ISO product data standards:

iso15926: ISO 15926 Integration of life-cycle data for process plants including oil and gas
production facilities
iso10303-1047 ISO 10303-1047 Product data representation and exchange – Activity
iso10303-1049 ISO 10303-1049 Product data representation and exchange – Activity method
iso10303-1164 ISO 10303-1164 Product data representation and exchange – Product as individual

The following namespaces are the RDF/OWL methodology:

rdf: Resource Description Framework
rdfs: Resource Description Framework schema
owl: Web Ontology Language

The ontology is represented using Notation 3 (http://www.w3.org/DesignIssues/Notation3.html).

CWA 16200:2010 (E)

43
Ontology specification:

@prefix core: <http://www.cen.eu/cwa/elssi-emd/ontology/core/> .
@prefix test: <http://www.cen.eu/cwa/elssi-emd/ontology/test/> .
@prefix iso6892-1: <http://www.cen.eu/cwa/elssi-emd/ontology/iso6892-1/> .

@prefix quomos: <http://www.example.org/> .

@prefix iso15926: <http://standards.iso.org/iso/15926/tech/> .
@prefix iso10303-1047:
<http://standards.iso.org/iso/10303/-1047/tech/express/Activity_arm.> .
@prefix iso10303-1049:
<http://standards.iso.org/iso/10303/-1047/tech/express/Activity_method_arm.> .
@prefix iso10303-1049:
<http://standards.iso.org/iso/10303/-1164/tech/express/
Product_as_individual_arm.> .

@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix owl: <http://www.w3.org/2002/07/owl#> .

6.6.3 Naming conventions

The following conventions are used within the CWA ontology:

form of a URI

Each class or property is identified by an HTTP URI that consists of a stem followed by a identifier which
is unique within the ontology. Fragment IDs are not used.

NOTE 1 When a fragment ID is used, each item in an ontology is associated with an identified ―whole‖.
This is appropriate when the items in an ontology are issued and revised as a single whole. However in
engineering ontologies, each item is usually issued and revised individually.

form of an identifier

An identifier is derived from an English language term for the class or property, with each space replaced
by an underscore.

NOTE 2 The URI of an item in an ontology, and hence the identifier of an item which is used to construct
a URI, is arbitrary. It is chiefly intended to be processed by a computer, and could be just a number.

Person readable terms in different languages can be assigned by the SKOS annotation property
preflabel. However the use of terms as identifiers is convenient for English language speaking ontology
developers.

case of initial letter in an identifier

- Each class has an identifier beginning with an upper case letter; and
- each property has an identifier beginning with a lower case letter.

NOTE 3 This is a common convention within RDF and OWL ontologies.

reference to classes and properties in natural language text definitions

In natural language text definitions, a class or property is referred to by its identifier in bold font.

CWA 16200:2010 (E)

44
6.7 Ontology for testing

6.7.1 Specimen

A Specimen is a Material_object that is separated from another Material_object so that it can be subjected
to a Test.

Ontology specification:

test:Specimen
rdfs:subclassOf core:Material_object .

NOTE programming language class names typically start with an uppercase letter, while properties and attributes
are typically lowercase, and this convention is adopted here.

6.7.2 Specimen_at_instant

A Specimen_at_instant is a Specimen that is also a Material_object_at_instant.

Ontology specification:

test:Specimen_at_instant
rdfs:subclassOf test:Specimen ;
rdfs:subclassOf core:Material_object_at_instant .

6.7.3 Test

A Test is an Activity that is performed upon a Material_object in order to find out something about it.

Ontology specification:

test:Test
rdfs:subclassOf core:Activity .

6.7.4 Test_control_method

A Test_control_method is a subclass of Test that has the control method for a Test as its criterion for
membership.

Ontology specification:

test:Test_control_method
rdfs:subclassOf core:Method ;
rdfs:subclassOf [ owl:onProperty rdfs:subclassOf ;
owl:someValuesFrom test:Test ] .

6.7.5 chosen_test_control_method

A chosen_test_control_method is a property of a Test that specifies the chosen Test_control_method.

This property specifies the intension of the performer of the Test.

Ontology specification:

test:chosen_test_control_method
rdfs:subPropertyOf core:chosen_method ;
a owl:ObjectProperty ;
rdfs:domain test:Test ;
rdfs:range test:Test_control_method .

CWA 16200:2010 (E)

45
6.7.6 Tensile_test

A Tensile_test is a Test that subjects a Specimen to tensile force.

Ontology specification:

test:Tensile_test
rdfs:subclassOf test:Test .

6.7.7 source_of_specimen

A source_of_specimen is a property of a Specimen that is the Material_object from which it was taken.

Ontology specification:

test:source_of_specimen
a owl:FunctionalProperty ;
rdfs:domain test:Specimen ;
rdfs:range test:Material_object .

6.7.8 specimen_for_test

A specimen_for_test is a property of a Test that is the Specimen of which a temporal part is tested.

Ontology specification:

test:specimen_for_test
a owl:FunctionalProperty ;
rdfs:domain test:Test ;
rdfs:range test:Specimen .

6.7.9 Specimen_test_machine_assembly

A Specimen_test_machine_assembly is a Material_object that is the assembly of a Specimen,
Crosshead, Extensometer, Load_cell, Control_system and other parts of a test machine such as its
frame.

Ontology specification:

test:Specimen_test_machine_assembly
rdfs:subclassOf core:Material_object .

6.7.10 Specimen_test_machine_assembly_for_period

A Specimen_test_machine_assembly_for_period is a Specimen_test_machine_assembly that is also
a Material_object_for_period.

Ontology specification:

test:Specimen_test_machine_assembly_for_period
rdfs:subclassOf test:Specimen_test_machine_assembly ;
rdfs:subclassOf core:Material_object_for_period .

6.7.11 Specimen_test_machine_assembly_at_instant

A Specimen_test_machine_assembly_at_instant is a Specimen_test_machine_assembly that is also a
Material_object_at_instant.

NOTE The relationship between a Specimen_test_machine_assembly_at_instant and a
Specimen_test_machine_assembly_for_period is temporal_part_of.

CWA 16200:2010 (E)

47

Ontology specification:

test:Load_cell
rdfs:subclassOf core:Material_object .

6.7.18 Load_cell_at_instant

A Load_Cell_at_instant is a Load_cell that is a Material_object_at_instant.

Ontology specification:

test:Load_cell_at_instant
rdfs:subclassOf test:Load_cell ;
rdfs:subclassOf core:Material_object_at_instant .

6.7.19 Extensometer

An Extensometer is a Material_object that measures the extension within a Specimen.

Ontology specification:

test:Extensometer
rdfs:subclassOf core:Material_object .

6.7.20 Extensometer_at_instant

An Extensometer_at_instant is an Extensometer that is a Material_object_at_instant.

Ontology specification:

test:Extensometer_at_instant
rdfs:subclassOf test:Extensometer ;
rdfs:subclassOf core:Material_object_at_instant .

6.7.21 Strain_rate_range

A Strain_rate_range is a part of Strain_rate that is a 1D manifold.

Ontology specification:

test:Strain_rate_range
rdfs:subclassOf [ owl:onProperty rdfs:subclassOf ;
owl:someValuesFrom quomos:Strain_rate ] .

6.7.22 specified_strain_rate

A specified_strain_rate is a property of a Test_machine_control_system_for_period that is the
specified strain rate.

The property can also be specified for the Specimen_test_machine_assembly_for_period that contains
the control system.

This property specifies the intension of the performer of the Test.

Ontology specification:

test:specified_strain_rate
a owl:FunctionalProperty ;
rdfs:domain test:Test_machine_control_system_for_period ;
rdfs:domain test:Specimen_test_machine_assembly_for_period ;
rdfs:range quomos:Strain_rate .

CWA 16200:2010 (E)

48

6.7.23 specified_strain_rate_range

A specified_strain_rate_range is a property of a Test_machine_control_system_for_period that is the
specified range of strain rates.

The property can also be specified for the Specimen_test_machine_assembly_for_period that contains
the control system.

This property specifies the intension of the performer of the Test.

Ontology specification:

test:specified_strain_rate_range
a owl:FunctionalProperty ;
rdfs:domain test:Test_machine_control_system_for_period ;
rdfs:domain test:Specimen_test_machine_assembly_for_period ;
rdfs:range test:Strain_rate_range .

6.7.24 specified_stress_rate

A specified_stress_rate is a property of a Test_machine_control_system_for_period that is the
specified stress rate.

The property can also be specified for the Specimen_test_machine_assembly_for_period that contains
the control system.

This property specifies the intension of the performer of the Test.

Ontology specification:

test:specified_stress_rate
a owl:FunctionalProperty ;
rdfs:domain test:Test_machine_control_system_for_period ;
rdfs:domain test:Specimen_test_machine_assembly_for_period ;
rdfs:range quomos:Stress_rate .

6.7.25 time_of_test_state

A time_of_test_state is a property of a Specimen_test_machine_assembly_at_instant that is the
duration of time since the reference instance at the beginning of the Test.

Ontology specification:

test:time_of_test_state
a owl:FunctionalProperty ;
rdfs:domain test:Specimen_test_machine_assembly_at_instant ;
rdfs:range quomos:Time_duration .

6.7.26 applied_force

An applied_force is a property of a Load_cell_at_instant that is the force applied by the test machine at
that instant.

The property can also be specified for the Specimen_test_machine_assembly_at_instant that contains
the Load_cell.

Ontology specification:

test:applied_force
a owl:FunctionalProperty ;
rdfs:domain test:Load_cell_at_instant ;

CWA 16200:2010 (E)

49
rdfs:domain test:Specimen_test_machine_assembly_at_instant ;
rdfs:range quomos:Force .

6.7.27 computed_stress_in_parallel_part

A computed_stress_in_parallel_part is a property of a Specimen_at_instant that is the force applied by
the test machine divided by the original_parallel_cross_section_area of the Specimen.

The property can also be specified for the Specimen_test_machine_assembly_at_instant that contains
the Specimen.

Ontology specification:

test:computed_stress_in_parallel_part
a owl:FunctionalProperty ;
rdfs:domain test:Specimen_at_instant ;
rdfs:domain test:Specimen_test_machine_assembly_at_instant ;
rdfs:range quomos:Stress .

6.7.28 percentage_plastic_extension

A percentage_plastic_extension is a property of a Specimen_at_instant that is the percentage plastic
extension of the Specimen.

The property can also be specified for the Specimen_test_machine_assembly_at_instant that contains
the Specimen.

Ontology specification:

test:percentage_plastic_extension
a owl:FunctionalProperty ;
rdfs:domain test:Specimen_at_instant ;
rdfs:domain test:Specimen_test_machine_assembly_at_instant ;
rdfs:range quomos:Percentage .

6.7.29 crosshead_displacement

A crosshead_displacement is a property of a Crosshead_at_instant that is the displacement of the
crosshead from a reference position for the Test.

The property can also be specified for the Specimen_test_machine_assembly_at_instant that contains
the Crosshead.

Ontology specification:

test:crosshead_displacement
a owl:FunctionalProperty ;
rdfs:domain test:Crosshead_at_instant ;
rdfs:domain test:Specimen_test_machine_assembly_at_instant ;
rdfs:range quomos:Length .

6.7.30 extensometer_displacement

An extensometer_displacement is a property of a Extensometer_at_instant that is the displacement
measured by the extensometer from a reference position for the Test.

The property can also be specified for the Specimen_test_machine_assembly_at_instant that contains
the Extensometer.

Ontology specification:

test:extensometer_displacement

CWA 16200:2010 (E)

54

6.8.11 percentage elongation after fracture

A percentage_elongation_after_fracture is a property of a Tensile_test_to_ISO_6892-1 that is
―percentage elongation after fracture‖ as defined in EN ISO 6892-1:2009.

Ontology specification:

iso6892-1:percentage_elongation_after_fracture
a owl:FunctionalProperty ;
rdfs:domain iso6892-1:Tensile_test_to_ISO_6892-1 ;
rdfs:range quomos:Percentage .

6.8.12 percentage reduction of area

A percentage_reduction_of_area is a property of a Tensile_test_to_ISO_6892-1 that is ―percentage
reduction of area‖ as defined in EN ISO 6892-1:2009.

Ontology specification:

iso6892-1:percentage_reduction_of_area
a owl:FunctionalProperty ;
rdfs:domain iso6892-1:Tensile_test_to_ISO_6892-1 ;
rdfs:range quomos:Percentage .

6.8.13 Proof_strength_plastic_extension_property

A Proof_strength_plastic_extension_property is the subclass of owl:FunctionalProperty that contains
the proof strength properties for different values of plastic extension.

EXAMPLE The property proof_strength_0.1_percent_plastic_extension is defined as follows:

iso6892-1:proof_strength_0.1_percent_plastic_extension
specified_percentage_plastic_extension 0.1 ;
a iso6892-1:Proof_strength_plastic_extension_property ;
rdfs:domain iso6892-1:Tensile_test_to_ISO_6892-1 ;
rdfs:range quomos:Stress .

Ontology specification:

iso6892-1:Proof_strength_plastic_extension_property
rdfs:subclassOf owl:FunctionalProperty ;
rdfs:subclassOf [ owl:onProperty rdfs:domain ;
owl:hasValue iso6892-1:Tensile_test_to_ISO_6892-1 ] .
rdfs:subclassOf [ owl:onProperty rdfs:range ;
owl:hasValue quomos:Stress ] .

6.8.14 specified_percentage_plastic_extension

A specified_percentage_plastic_extension is a property of a
Proof_strength_plastic_extension_property that is the specified plastic extension.

Ontology specification:

iso6892-1:specified_percentage_plastic_extension
a owl:FunctionalProperty ;
rdfs:domain iso6892-1:Proof_strength_plastic_extension_property ;
rdfs:range quomos:Percentage .

6.8.15 Strain_rate_range_1

Strain_rate_range_1 is the strain rates that are in the range 0.000 07 0.000 014 s
-1
.

CWA 16200:2010 (E)

55

Ontology specification:

iso6892-1:Strain_rate_range_1
rdfs:subclassOf quomos:Strain_rate .

6.8.16 Strain_rate_range_2

Strain_rate_range_2 is the strain rates that are in the range 0.000 25 0.000 05 s
-1
.

Ontology specification:

iso6892-1:Strain_rate_range_2
rdfs:subclassOf quomos:Strain_rate .

6.8.17 Strain_rate_range_3

Strain_rate_range_3 is the strain rates that are in the range 0.002 0.000 4 s
-1
.

Ontology specification:

iso6892-1:Strain_rate_range_3
rdfs:subclassOf quomos:Strain_rate .

6.8.18 Strain_rate_range_4

Strain_rate_range_4 is the strain rates that are in the range 0.006 7 0.001 33 s
-1
.

Ontology specification:

iso6892-1:Strain_rate_range_4
rdfs:subclassOf quomos:Strain_rate .

6.8.19 Test_phase_1

A Test_phase_1 is a Specimen_crosshead_extensometer_assembly_for_period from the beginning of a
test up to the determination of the upper_yield_strength, proof_strength_plastic_extension or
proof_strength_total_extension.

Ontology specification:

iso6892-1:Test_phase_1
rdfs:subclassOf test:Specimen_crosshead_extensometer_assembly_for_period .

6.8.20 Test_phase_2

A Test_phase_2 is a Specimen_crosshead_extensometer_assembly_for_period from the end of
Test_phase_1 up to the determination of the lower_yield_strength.

Ontology specification:

iso6892-1:Test_phase_2
rdfs:subclassOf test:Specimen_crosshead_extensometer_assembly_for_period .

6.8.21 Test_phase_3

A Test_phase_3 is a Specimen_crosshead_extensometer_assembly_for_period from the end of
Test_phase_2 to the end of the Test.

Ontology specification:

iso6892-1:Test_phase_3

CWA 16200:2010 (E)

59
Physical individual
whole part
has part
part of
temporal whole part
has temporal part
temporal part of
Method
Day
Thermodynamic
Temperature
chosen method
day of activity
ambient temperature
* *
* 0, 1
*
0, 1
Activity
Material object
Material object for
period
Material object at
instant


Figure 15 - Extract from an ontology for industrial data 1 of 1

ActivityMaterial object
chosen
method
*
*
Test control
method
chosen test
control method
Tensile test
* *
specimen for test
Specimen
source of specimen
*
0, 1
initial extensometer gauge length
Length
Test

Figure 16 - Ontology for testing 1 of 9

CWA 16200:2010 (E)

60
Material object
Specimen test
machine
assembly
Test machine control system
Load cell
Extensometer
Crosshead
Specimen
1
1
1
1
1

Figure 17 - Ontology for testing 2 of 9

Material object for
period
Specimen test
machine
assembly
0 1 *
specified strain rate range
Strain rate range
0 1
*
specified strain rate
\ \ \ \
Strain rate
0 1
*
specified stress rate
\ \ \ \
Stress rate
Test machine
control system
Specimen test
machine
assembly for
period
Test machine
control system
for period
Specimen test
machine
assembly during
test
1 1

Figure 18 - Ontology for testing 3 of 9

CWA 16200:2010 (E)

61
Material object for
period
Specimen test
machine
assembly
0 1 *
specified strain rate range
Strain rate range
0 1
*
specified strain rate
\ \ \ \
Strain rate
0 1
*
specified stress rate
\ \ \ \
Stress rate
Specimen test
machine
assembly for
period
Specimen test
machine
assembly during
test

Figure 19 - Ontology for testing 4 of 9 – simplification of Figure 18

Force
Length
Time duration
1 *
time of test state
1 *
applied force
1
*
crosshead displacement
1
extensometer displacement
Material object
at instant
Specimen test
machine
assembly
Specimen test
machine
assembly at
instant
1 1
Load cell
Load cell at
instant
*
1 1
Crosshead
Crosshead at
instant
1 1
Extensometer
Extensometer at
instant

Figure 20 - Ontology for testing 5 of 9

CWA 16200:2010 (E)

64
Tensile test
*
0, 1
tensile strength
*
0, 1
upper yield strength
*
0, 1
lower yield strength
*0, 1
proof strength yield point extension
*
0, 1
percentage yield point extension
*
0, 1
percentage plastic extension at maximum force
*
0, 1
percentage total extension at maximum force
*
0, 1
percentage total extension at fracture
*
0, 1
percentage elongation after fracture
*
0, 1
percentage reduction of area
Percentage
maximum force
*
0, 1
Force
Tensile test to
ISO 6892-1
Stress
1 1
Specimen test
machine
assembly during
test

Figure 25 - Ontology for EN ISO 6892-1:2009 1 of 4

Tensile test
*
0, 1
tensile strength
*
0, 1
upper yield strength
*
0, 1
lower yield strength
*0, 1
proof strength yield point extension
*
0, 1
percentage yield point extension
*
0, 1
percentage plastic extension at maximum force
*
0, 1
percentage total extension at maximum force
*
0, 1
percentage total extension at fracture
*
0, 1
percentage elongation after fracture
*
0, 1
percentage reduction of area
Percentage
maximum force
*
0, 1
Force
Tensile test to
ISO 6892-1
Stress

Figure 26 - Ontology for EN ISO 6892-1:2009 2 of 4 – simplification of Figure 25

CWA 16200:2010 (E)

67
7 Mappings Between Existing Schemas and Ontologies

7.1 Generic and specific concepts in ontologies and schemas

7.1.1 Making information precise

The purpose of an ontology or schema is to record information. The information should be precise and
unambiguous. The information should also not rely upon context, i.e. other information recorded elsewhere,
because this context may not be available when the information is read.

EXAMPLE 1 Information can be recorded without specifying the units of measure. This will work within an
organization or computer system that has rules for the choice of units of measure. However, the recorded
information will be useless outside the context provided by the organization or computer system.

In this case of an ontology, information consists of statements which are recorded using formal logic. In the
case of a schema, information is encoded within a data structure. In either case it is necessary to go beyond
generic concepts to ensure that the recorded information is precise and unambiguous.

EXAMPLE 2 The class ―tensile test‖ may be defined in a schema or an ontology. In order to specify that test
T_101 is carried out to EN ISO 6892-1, it is also necessary to define the subclass ―tensile test to EN ISO
6892-1‖.

The writers of a schema are free to choose:
which classes correspond to entities within the schema;
which classes represented by data values.
There is a trade-off between clarity which is provided by the use of entities, and simplicity which is provided
by the use of data values.

EXAMPLE 3 There are different possible approaches to constructing a schema to support Example 2, as
shown in Figure 29.

Test
chosen method ID : string
type ID : string
ID : string




The string ―T_101‖ is specified for ID.
The string ―tensile test‖ is specified for type ID.
The string ―tensile test to EN ISO 6892-1‖ is
specified for chosen method ID.
Test
chosen method ID : string
ID : string
Tensile test

The string ―T_101‖ is specified for ID.
The string ―tensile test to EN ISO 6892-1‖ is
specified for chosen method ID.

CWA 16200:2010 (E)

71

Tensile test
Tensile test to
ISO 6892-1
Quantity
Tensile test
property
possessor
*
1
possessed
*
1
StressPercentage Force
Class of tensile
test property
ID : string
*
*
type

Figure 34 - Property and class of property
The string ―tensile strength‖ can be specified for the ID of entity Class of tensile test property. Hence the
instance of this entity represents the ―tensile strength‖ property. This approach is used within ISO 15926,
and is a step closer to the creation of an ontology.

Within an ontology there is no distinction between data values and things in the schema.

EXAMPLE 3 The representation of the property tensile test is as shown in Figure 35
Tensile test to
ISO 6892-1
stress
tensile
strength
domain range

Figure 35 - Property in an ontology
A property, which defined in an ontology, can be referenced by a data value in a schema.

7.1.3 Generic property definitions

When a property is defined in an ontology it is necessary to:
provide a natural language definition;
specify the things which can have the property – its domain; and
specify the nature of a value for the property – its range.

7.2 Different roles for different technologies

The management of materials data over a broad scope is a difficult task. It is necessary to:

define the activities which will be supported, both at a strategic and at a detailed level;

Activities range from the testing of input materials against routine acceptance criteria, to testing
campaigns investigating behaviours of material classes, to data reduction activities leading to the
deeming of design values.

An activity model defines activities, shows sequences and nesting of activities, and shows the
information flows between them.

CWA 16200:2010 (E)

74
TENSTAND is a particularly efficient format for the representation of the measured values of force,
crosshead displacement, elongation and time during a test. This format has only limited support for
meta-data about the test.

NOTE 3 ISO 10303-235 identifies some key materials data activities and data flows, as shown in Figure 38.

I1
Stock material and
component parts *#
C1
System test
requirements
O3
Managed
materials
data
Procure and
test material
object
A211
P. 5
Reduce and
evaluate data
A212
P. 6
Maintain
material object
records
A213
Materials
informatio
Test specimen
definition
Test condition
records
Test results
Batch or
material object
definition
Material object specifications
I2
Supplier records *#
O2
Recycled or
waste product *
I3
Delivered product *#
O1
Tested material
object

Figure 38 - Activities involved with materials data defined in ISO 10303-235.
However other activities related to materials acceptance are not defined in ISO 10303-235, as shown in
Figure 39.

I1
Stock material and
component parts *#
C1
System test
requirements
O3
Managed
materials
data
Procure and
test material
object
A211
P. 5
Reduce and
evaluate data
A212
P. 6
Maintain
material object
records
A213
Materials
informatio
Test specimen
definition
Test condition
records
Test results
Batch or
material object
efinition
Material object specifications
I2
Supplier records *#
O2
Recycled or
waste product *
I3
Delivered product *#
O1
Tested material
object
issue test
certificate
accept
supply
input to
manufacturing
test certificate
to material
quality records


Figure 39 - Activities involved with materials acceptance.

CWA 16200:2010 (E)

75
NOTE 4 Although the data flow for issuing a certificate is not identified explicitly in the AAM of ISO 10303
Part 235, the ARM and AIM model provide for this requirement.

Different technologies are used for different parts of a framework for the management of materials data.
These include:
activity models;
computer processable thesauri for definitions
ontologies;
conceptual schemas using UML and EXPRESS
implementable schemas using XML

NOTE 5 The domains of application of these technologies overlap because:
ontologies and EXPRESS schemas can be implemented directly without using XML schemas;
XML schemas can be used as conceptual models.

The remainder of this clause contains a discussion of the role of mappings between the different
technologies, and the actual mappings between:
the thesaurus and ontology derived from EN ISO 6892-1:2009;
ISO 10303-235;
ISO 15926;
MatML; and
JRC:MatDB.

7.3 Precision about semantics

Both schemas and ontologies are devised to represent statements in a way which can be understood reliably
by a computer. Statements made in a natural language can be understood by people, but because of the
complexity of natural language they cannot be understood reliably by a computer.

Schemas and ontologies automatically give precision about the representation of data, i.e. ―syntactic
precision‖. They are analogous to forms with fields. All the data is in particular fields, so that the data can be
stored in a computer. However precision about the representation of data does not imply precision about the
meaning of data, i.e. ―semantic precision‖.

EXAMPLE Consider the following statements:

A. Test T_101 was carried out according to procedure P on specimen S_101 of material M and gave a
tensile strength of 345 MPa.

B. Any test according to procedure P on a specimen of material M would give a tensile strength of 345
MPa. We know this because of Test T_101 which was carried out on specimen S_101.

NOTE 1 Statement B is actually untrue because there exist slight variations between samples, so that the
values are unlikely ever to be replicated exactly. It may be correct to say "give a tensile strength close to 345
MPa". However, the representation of the concept of "close to" within a schema or ontology is not within the
scope of this CWA, and the untrue Statement B serves only as an example of relationships that can be defined by
the ontology.

A person will understand that statement A is different to statement B. However the same syntax could be
used to represent either statement. In XML we could have:

<Test id=”T_101”>
<Procedure id=”P”/>
<Specimen id=”S_101”>
<Material id=”M”/>
</Specimen>
<tensile-strength unit=”MPa”>345</tensile-strength>
</Test>

or:

<Material id=”M”/>

CWA 16200:2010 (E)

79
XML schema EXPRESS schema

The definition of the semantics of an XML schema by reference to an EXPRESS schema is common
practice within ISO TC184/SC4. ISO 10303-28 ―Implementation methods: XML representations of
EXPRESS schemas and data, using XML schemas‖ defines a method of generating XML schemas from
EXPRESS schemas, and thereby defining their semantics.

7.5 ISO 10303 — JRC:MatDB — MatML — NMC:MatDB — ISO 15926 Mappings

7.5.1 Methodology

The different standards have broad scopes so it is not an objective of this document to provide complete
mappings. Instead:
the data necessary to record a tensile test to EN ISO 6892-1:2009 will be separated into a set of
statements;
each statement will be represented according to each of the standards;
the mapping will be derived from the representation of the statements.

7.5.2 Test specimen for a testing activity

statement Test T_101 is carried out on specimen S_101.
ontology
specimen for test
S_101T_101

ISO 15926
T_101 S_101
whole part
specimen for
test
class of participation

JRC:MatDB
<Test testID=”T_101”>
< child element specifying test type >
<Specimen>
<SpecimenGeometryAndProduction specimenName="S_101"/>
</Specimen>
</child element specifying test type >
</Test>

CWA 16200:2010 (E)

80
MatML
<Material>
<PropertyData test=”testRef” specimen=”specimenRef”/>
</Material>

<MetaData>
<SpecimenDetails id=”specimenRef”>
<Name>S_101</Name>
<SpecimenDetails>
<TestConditionDetails id=”testRef”/>
</MetaData>

[NOTE: MatML does not appear to be able to assign an identifier to
a test.]
NMC:MatDB
< element specifying test type >
<Specimen/>
</element specifying test type >

[NOTE: NMC:MatML does not appear to have attributes which can
assign identifiers.]
ISO
10303-235
#5=PRODUCT_DEFINITION_FORMATION
('T_101', label, type of test and reference to material );
#8=PRODUCT_DEFINITION_WITH_ASSOCIATED_DOCUMENTS
('S_101', description , #5,
reference to keywords which specify that the PRODUCT DEFINITION
is a test specimen, and specify the type of test );

The key objects in the ontology are as follows:
rdfs:domain
Test
Activity
rdfs:subclass f
Material object
Specimen
tested in
rdfs:range
specimen for
test
owl:inverseOf
rdfs:subclassOf

The mappings to the ontology are as follows:
ISO 15926
In ISO 15926, Test is a member of class of activity.
In ISO 15926, Specimen is a member of class of arranged individual.
The bi-directional class of participation specimen for test is equivalent to the pair of
properties onSpecimen and testedIn.
JRC:MatDB
Element Test is equivalent to class Test.
Element Specimen is equivalent to class Specimen.
The XML schema does not have a structure equivalent to onSpecimen, because this
relationship is implied by the parent-child relationships

CWA 16200:2010 (E)

83
Test
rdfs:subclassOf
Tensile test
rdfs:subclassOf
rdfs:subclassOf
rdf:type
Tensile test to
ISO 6892-1
Class of activity
hasPowerClass
Activity
Method
rdfs:subclassOf
chosen
method
rdfs:domain rdfs:range


The mappings to the ontology are as follows:
ISO 15926
In ISO 15926, UniaxialTensileTest and TensileTestToISO6892-1 are members of class of
activity.
In ISO 15926, Procedure is a member of class of class.
The property hasProcedure is regarded as a classification.
JRC:MatDB
Element UniaxialTest specifies a classification by its parent child relationship.
Attribute testStandard is equivalent to property hasProcedure.
MatML
Attribute technique is equivalent to the property hasProcedure.
Entity MeasurementTechnique is equivalent to the subclass of the class Procedure that is
concerned with tests.
There does not appear to be a way of classifying a test as a UniaxialTensileTest.
NMC:MatDB
Element TensileTest is equivalent to class UniaxialTensileTest.
Element Procedure is equivalent to class Procedure.
The XML schema does not have a structure equivalent to hasProcedure, because this
relationship is implied by the parent-child relationship.
The class TensileTestToISO6892-1 is specified by attributes and child elements of
Procedure.
ISO 10303-
235
Attribute description of entity product_definition_formation specifies the classification of the
test.
The procedure for the test is specified by a document associated with the product_definition
that is the specimen.

7.5.4 Result of testing activity

statement Test T_101 gives an upper yield strength of 345 MPa.

CWA 16200:2010 (E)

86
ISO 15926
B_101S_101
specimen source
source of
specimen
class of class of
relationship

JRC:MatDB
<Materials>
<Material materialKey="matKey>
<Production batchIdentifier="B_101"/>
</Material>
</Materials>

<Tests>
<Test>
< child element specifying test type >
<Specimen>
<SpecimenGeometryAndProduction materialKey="matKey"
specimenName="S_101"/>
</Specimen>
</child element specifying test type >
</Test>
</Tests>
MatML
MatML does not appear to have a capability to specify the batch from which a specimen is
taken.
NMC:MatDB
NMC:MatDB does not appear to have a capability to specify the batch from which a specimen
is taken.
ISO 10303-
235
#8=PRODUCT_DEFINITION_WITH_ASSOCIATED_DOCUMENTS
('S_101', description , #5,
reference to keywords which specify that the PRODUCT DEFINITION
is a test specimen, and specify the type of test );
#800=PRODUCT_DEFINITION_WITH_ASSOCIATED_DOCUMENTS
('B_101', description , #5,
reference to keywords which specify that the PRODUCT DEFINITION
is a batch );
#801=PRODUCT_DEFINITION_RELATIONSHIP
('B_101 to S_101', 'source of specimen', description, #800, #8);

The key objects in the ontology are as follows:

CWA 16200:2010 (E)

87
rdfs:domain
Material object
Specimen
rdfs:range
source of specimenrdfs:subclassOf


The mappings to the ontology are as follows:
ISO 15926 In ISO 15926, specimenTakenFrom is a member of class of class of relatonship.
JRC:MatDB
The attribute materialKey of element SpecimenGeometryAndProduction which is a child
element of Specimen is equivalent to the property specimenTakenFrom.
MatML
MatML does not appear to have a capability to specify the batch from which a specimen is
taken.
NMC:MatDB
NMC:MatDB does not appear to have a capability to specify the batch from which a
specimen is taken.
ISO 10303-
235
Entity product_definition_relationship with the label attribute ―source of specimen‖ is
equivalent to source of specimen.

CWA 16200:2010 (E)

90
Identify the stages in the value chain of the business process of interest;
Profiling the organisations in the business sector of interest;
Identify deviations from the existing value chain;
Formulate a protocol and interview workshop‘s participants and interested parties;
Formulate a set of modified business models or corrective measures to the existing one;
Select and refine specific business models for intervention and validation;
Propose interventions to the existing business process that offer the best means of a successful
transition to the new model;
Assess the efficacy of the interventions and the additional business venues it modifies and generates.

8.2 Test Certificates for Engineering Materials

8.2.1 General
Mechanical testing certificates provide a guarantee of origin and material specification. While these
certificates are commonly made available in electronic format, these documents are often simply digitized
images embedded in a PDF document, and offer little added-value compared to their conventional paper
counterpart.

A transition to electronic test certificates generated from a content management system offers advantages
over the paper-based alternatives, including that the original data can be traced and the output format can be
tailored to the customer requirements.

8.2.2 Related Cases in Other Domains

In many sectors of economic activity the introduction of computer-readable standards (meaning standardized
data formats) have prompted the production of business models studies and cost benefit analyses. One case
that has been recently reported is the e-clinical trial in the pharmaceutical sector.

See Annex K for reviews of different examples of a transition to more extensive use of computer-readable
standards (meaning standardized data formats) in other business domains.

8.3 Analysis of the Test Certificate Value-Chain

8.3.1 Methodology

The following sections are aimed at presenting a taxonomy of the value chain in the production, issuing and
distribution of test certificates. Paper-based and standardised machine readable tests certificates‘ properties
are also presented and discussed.

Actors involved along the value chain of test certificates, their business interests and sources of business
revenues, constitute the main elements in drawing the taxonomy and are taken as the starting point in
building our business case.

Innovation in the certification process is then seen from the perspective of the stakeholders and discussed
according to their views on the matters.

8.3.2 Test Certificates Value Chain

The value chain for the production, delivery, and consumption of paper-based test certificates involves
standards authorities, materials suppliers, machine manufacturers, software houses, manufacturers, and
their customers.

The typical life cycle of a test certificate is linked to the manufacturing process of the material it relates to. A
manufacturer runs raw materials through a given production line in order to obtain a number of batches of
final product (i.e. sheets of AXi) with certain characteristics.

NOTE AXi indicates the product (engineering material) as follow: A identifies the manufacturer and the
product type; X, the production line or process, and ―i‖ is the batch index i=1, 2, …, n batches obtained through
the given production line.

CWA 16200:2010 (E)

93
distributing and using said certificates and in the disentangling the network of interests presented in figure 24
and 25 above.

How the introduction of standardised machine readable documents will affect the status quo and ride on the
current, ongoing changes in the market for test certificates.

8.5 Business Survey

8.5.1 Structure of the relevant industry (and relative interests of the actors)

The taxonomy presented in section 8.1.4 is hereafter matched against the participants‘ interests in order to
map their roles in the test certificates market and to ascertain the validity of the methodology used.

The stakeholders were shown the classification used comprising a brief description of their stated interests in
the process and the chain of value creation along the process from production to sale of the material
including intermediaries, complementary industries and related uses of the certificates. From the open
discussion with the stakeholders during the plenary session held in Delft on the 8
th
of October 2009, the
hypothesis implicit in the value chain diagrams and the working hypothesis were validated. The strategy to
carry further analysis and the action points were subsequently undertaken.

A survey, aiming at ascertaining the participants‘ view on the business‘ case for the standardisation and the
introduction of electronic material test certificates, was taken into consideration as the most viable ways to
proceed further.

The interview protocol (available in Annex L) has been tailored to be as inclusive and exhaustive as possible
in order to take into consideration the business needs deriving from the variety of businesses‘ objectives
along the value chain and the stakeholders‘ interests in the certification process.

The pro-forma interview was sent to 30 workshop participants between mid-December 2009 and January
2010. Three rounds of reminders were also performed in January. The responses received amount to 14 (or
46.6 % response rate) covering all the actors identified in the value chain presented in Figure 2. The
participants were asked to respond, in writing, to the section(s) of the pro-forma that were applicable to their
role(s) and the current business model adopted by their organisations. A summary of the respondents‘
characteristics is available in Annex M.

8.5.2 Characteristics of the organisations participating in the survey

In the manufacturing sector, materials test certificates are central to both the phase of production and sale.
In the production process the use certification pertains to testing on raw materials while on the sales phase
the certification relates to the products destined to the market. In larger manufacturing companies, testing
and certification of materials has an important place also in the research and development activities of the
organisation. Given this characteristic, manufacturers in the three sectors identified constitute the demand
and the supply side of the value chain.

The material test laboratories’ business is centred on the production of test certificates and material test
reports comprising a battery of tests for a sample material. The respondents have highlighted that their
business objective relates specifically to the needs of the customers regarding the details of reporting: i.e.
some customers require a pass/fail certification, while others require detailed and integrated certificates that
are able to accompany the material throughout its lifecycle.

The production of test certificates is not ―core‖ business for material test machinery manufacturers.
However their involvement in the value chain of said documents is such that their interests on the
standardisation of computer readable certificates are stated as complementary and of strategic support in
the phase of development of their products.

The role of software houses in the test certificate value chain relates specifically to providing the software
necessary to export, handle, archive and update the test results and software products for capturing raw
data and integrate them into the workflow as required by the procedures in place within the customers‘
business. Software houses also provide consulting advice to their costumers in the design development and
implementation of test documents management systems.

CWA 16200:2010 (E)

95
development and production of testing machinery as the content of specific tests influences the machines
output and for marketing and demonstration purposes.

The issues brought out by software houses relate to those aspects of the value chain that act as connector
between the various actors and networks directly with the organisations‘ knowledge management systems
either as major input to the development of the system (i.e. through expert advice or design, development
and implementation of the systems) or as interface between the actors (i.e. software to capture the reading
of the test machine and translate them into readable documents).

NOTE Actual suppliers of test certificate software have not directly contributed to the survey. In order to
address this shortcoming, the comments of software houses and developers made at ELSSI-EMD plenary
sessions and via direct contact have been taken into consideration.

For the reasons adduced, participants involved in activities of research and development have brought to
light various issues relating to the case at hand. Issues pertaining to the conceptualisation of the steps to
integrate electronic formats of certification documents with the information required throughout the lifecycle
of the product they accompany are the main themes of test certification basic R&D carried out in university
laboratories and departments and is better known as material data representation. This research
investigates the problems linked to the dichotomy between paper-based and linked data and their mutual
correspondence. Respondents from industrial R&D background highlight the issues of interoperability
between the certification process and the handling, archiving and updating of the test results between
different R&D sites of the same company, between R&D functions and other operation sites within the
company and between the company and its customers and other interested parties (i.e. auditors). Public
research organisations are instead involved in issues relating to handling, archiving, and managing large
quantity of test data and databases and their application in the respective field of research (i.e. Energy).

The respondents engaged as publishers coincide with software houses; for this reason, the previous
comments reported for the latter, are standing in this case.

The issues brought on the table by third party services providers relate to test certificate management and
their integration with customised services.

The issues highlighted by life cycle analysts are very stringent in terms of certification and conformity to
standards (ASTM, for example) because of regulations and laws concerning quality control and traceability
of the material use in the production process both as raw material or as intermediate products. As a
consequence, also test certificates are required to live up to a certain standard.


8.7 The Business case: analysis and conclusions

While a detailed analysis of the responses for each category of stakeholder is presented in the Annex M,
hereafter we provide the overall view and draw our conclusions.

Collating and analysing the responses given by the participants, the authors report that there is a
generalised need for a computer readable standard test certificate for materials. This need is felt,
albeit for different reasons, by all stakeholders situated along each link of the value chain identified in figure
25. In fact, for all stakeholders involved in the life cycle of test certificates, the business opportunities deriving
from the introduction of a standard computer readable format are self evident and, at times, only identifiable
through the indirect effects they have on the broader organisation‘s value generation activities.

All stakeholders independently highlight (matching our working hypothesis) that introducing standardised
computer readable test certificates will not substantially alter the balance of interests of all parties
within the sector but increases sensibly the efficiency of the day to day operations resulting into
substantial advances at a test certificate management level. In this regard, standardisation will promote
the implementation of state of the art knowledge management and quality control systems as long as the
working parties avoid erecting barriers that can derive from the use of restricted or not shared standard
setting framework.

NOTE A standard setting framework can be defined a set of rules embedded in the network of stakeholders. It
comprises shared issues and technical matters that are necessary when selecting and implementing a standard
setting method, including considerations of all interests of the participants and a shared technological platform that

CWA 16200:2010 (E)

100
9.1.3 Business Model and Economics of Technical Standards

The standards industry is estimated to contribute £2.5 billion to the UK economy alone
[http://www.nccmembership.co.uk/pooled/articles/BF_WEBART/view.asp?Q=BF_WEBART_171204][The
Empirical Economics of Standards, DTI Economics Paper No.12, June 2005].

The introduction of standards-compliant data formats into the standards development process has revenue
generation and licensing implications corresponding to the manner in which the data formats are published.
The conventional publishing mechanism for mechanical testing standards is point-to-point, either as paper-
based media or immutable electronic media, such as read-only PDF format.

Web specifications and standards, such as those managed by W3C and OASIS, are typically published
online as HTTP URI namespaces and are available free-of-charge. Examples include the HTTP and XML
Recommendations, the HTTP URI namespaces for which are published at http://www.w3.org/1999/xhtml/
and http://www.w3.org/XML/1998/namespace, respectively. Since the standards-compliant formats for
mechanical testing described in this document more closely resemble Web specifications than mechanical
testing standards, they too could be published as HTTP URI namespaces. This though introduces revenue
generation and licensing issues, the examination of which is addressed in Clause 9.4.

9.1.4 HTTP URI Namespaces

HTTP URI namespaces are the preferred means for publishing standards in the ICT sector. W3C policy is
described at URIs for W3C Namespaces, and Architecture of the World Wide Web, Volume One. Similarly,
OASIS policy is described at OASIS Naming Guidelines: Filenames, URIs, Namespaces. HTTP URI
namespaces are fundamental to Web architecture, and consequently offer a means for transition to a
Semantic Web of data, where unique occurrences of information (objects) such as a data set or a definition
have an associate HTTP URI namespaces.

NOTE 1 In the context of schemas and ontologies, dereferenceable HTTP URIs are intended primarily for
use by machines, not people, and allow a copy of a resource to be available to be used by a software application.

The utilization of HTTP URI namespaces for standards-compliant data formats applies to indexes and
objects. For example, an index of materials data sets will have a URI, and will be accessed by dereferencing
that URI. Within the index, each materials data set will have URI, and will be associated with URIs
describing relevant administrative metadata, including:
Material type
Material category (containing many materials types)
Property type
Test method
Specimen type
Test laboratory
Operator name

NOTE 2 The list of headings is the start of an administrative metadata set for association with test data. This
is crucial for the long-term curation and reuse of the test data created as it describes the provenance of the data.
The list is not exhaustive, and there are in fact existing metadata standards for creating a proper metadata set for
describing the provenance of the data, such as Dublin Core.

The HTTP URI namespace for an object should be assigned by the organization that defines the object.
Hence the HTTP URI namespace for a materials data set will be assigned by the organization that creates it,
and the URI for an index will be assigned by the organization that creates it. Standardization bodies define
property types, test methods, specimen types, etc. These will have HTTP URI namespaces assigned by the
standardization body.

See Annex P for the ELSSI-EMD contribution to the ICT Forum discussion of namespace URIs.