Proceedings of the Twenty-seventh (2017) International Ocean and Polar Engineering Conference www.isope.

org
San Francisco, CA, USA, June 25-30, 2017
Copyright © 2017 by the International Society of Offshore and Polar Engineers (ISOPE)
ISBN 978-1-880653-97-5; ISSN 1098-6189

A Data Management System for Material Properties and Fracture Assessment at Low Temperature

Downloaded from http://onepetro.org/ISOPEIOPEC/proceedings-pdf/ISOPE17/All-ISOPE17/ISOPE-I-17-584/1257033/isope-i-17-584.pdf/1 by Saudi Aramco, Pierre Martin on 24 April 2025

Vidar Osen 1, Alexandre Kane 1, Odd M. Akselsen 1
1 SINTEF Materials and Chemistry,
Trondheim, Norway

welding qualification.

In this paper, typical classes of materials are revisited to present an

improved data management system and illustrate this framework by
ABSTRACT considering the fracture toughness assessment from testing data of a
structural steel.
Based on the project Arctic Materials (2008-2017), a generic database
platform has been developed for efficient access to material data of DATA SETS
various materials, such as steels/weldments, aluminium alloys,
composites and polymers/coatings. Safe and flexible data processing, Large amount of available results in Arctic Materials (2008-2017)
accessible through a customized web interface, is obtained by unifying projects comprised, from room temperature to -60⁰C as minimum,
test results, test methods and procedures (e.g. tensile testing, fracture structural steels (rolled plates), forgings, castings and pipelines,
mechanics testing). This framework is strengthened by the integration polymers and polymer coatings, composites and aluminum.
of post-processing calculation and visualizing tools to account for the
influence of, e.g., temperature, constraint effects, probabilistic scatter, Polymers & Composites data are typically: stress strain curves in
and scaling on the material response. Its aim is to contribute to an tension. Stress strain curves in compression and DMTA (Dynamic
improved design guideline for materials requirements in Arctic Mechanical Thermal Analysis) data.
conditions (down to -60ºC). This paper will illustrate the framework by
considering the fracture toughness assessment from testing data of a Steel and Aluminum data, for various relevant material grades,
structural steel. include typically: stress-strain curves (of base metal, and weld metal
using notched tensile samples), Charpy V notch tests and Fracture
KEY WORDS: Arctic; low temperature; database; structural steel; toughness results (of weld metal and fusion line). As well as additional
polymers; composites; aluminum; probabilistic scatter, fracture. technical specification related to crack positions (orientation relative to
the weld), number of test specimens (repetition), temperatures,
INTRODUCTION geometries/notch depth, plate thickness, welding procedures,
microstructures description etc.
A large share of the remaining oil and gas reserves is located in the
Arctic. This will move the interest of the industry northwards. Here, the Various examples of use of such existing data using our previous
challenges are quite different from those experienced by the industry so version of the database have been shown by Kane & al (2015). In the
far. For the offshore part of the Arctic, some of these are icing, following section, the improved database framework will be presented
icebergs, low temperature, polar lows (strong winds with drifting and the technologies behind this updated data management system.
snow), long distances and lack of infrastructure and daylight in the
winter months. From a material and structural point of view, two IMPROVED DATA MANAGEMENT SYSTEM
particular challenges will be, low temperatures with a resulting
increased risk of insufficient fracture toughness, and large ice loads, Framework and Technologies
where the expected load must be accounted for in the design stage. The
costs in exploration in the most Northern areas are expected to be The solution, proposed by the Arctic Materials project, includes a web
higher than those elsewhere in the North Sea. Therefore, cost savings application with an underlying database system for fast data access and
are even more important in this region. flexible data processing.

With cost savings in mind, the objective of the proposed database is to The web implementation enable users to export desired tables of data
establish a data management system for material assessment at (e.g. to Microsoft Excel) and data plotting for further analysis while
temperatures relevant for the Arctic region. This may reduce cost in applying filters for different data types.
design and redesign, and not at least, during in materials selection and

310
The system is based on a relational database (RDBMS) and has been The search result is by default presented as tables consisting of columns
tested on both commercial database systems and free and open source that the user can choose, see Figure 3. The available columns is of
database systems. For the web interface available for the Arctic course not restricted to data for one class, but can be chosen from any
Materials project partners, a Microsoft SQL-Server database is related class in the network of relations between classes.
currently being used. The web front-end is realized using C# and
ASP.NET at the server side. The communication with the clients is
partly based on Ajax and Javascript for flexibility and quick response
on user actions.

The database system relies on its own metadata subsystem to describe

the classes of data (materials, specimens, tests, documents, etc.) and

Downloaded from http://onepetro.org/ISOPEIOPEC/proceedings-pdf/ISOPE17/All-ISOPE17/ISOPE-I-17-584/1257033/isope-i-17-584.pdf/1 by Saudi Aramco, Pierre Martin on 24 April 2025

their relations to account for the laboratory working flow (traceability
of data).

As an illustration, a few classes with their relations are presented in

Figure 1. For instance, the class "Specimen" describes the physical
piece of material that is tested. "Specimen" is related to the class Figure 3: Example of data table obtained with use of filtering
"ExpeTest", which describes the actual testing of the specimen. "Test" functionalities.
contains specific information for the physical testing of the particular
specimen (like the test execution date). The raw test results are stored
using the class "TimeSeries". "ExpeTest" has a relation named It is possible at any stage of the filtering process to export the data by
"rawdata" which connects the test with its test results. As we see from downloading a spreadsheet document containing all the data in the
the illustration, the tests are grouped together using the class result table.
"TestSeries".
The user can also easily plot the data by switching from a table view to
a plot view. All results in the result table cells and the points in the plot
is clickable. A click will bring up a "data card" for the clicked object.
This data card will contain information about the object in addition to
links to the related objects. The examples below show "data cards"
related to toughness data (Figure 5) and tensile data (Figure 7) together
with links to related objects like the test series, the specimen and
pictures of the specimen.

Figure 1: Illustration of data classes and their relationships.

The data management system application has been designed according

to predefined set of user requirements and use cases (Kane & al 2015).
A typical user of the system includes scientist and (design) engineers.

Web interface Design and implementation

The web interface is kept as simple as possible. It starts with a page
with functionalities to search the database. The user can search for data
using pre-defined filters in combination with a text search phrase, see
Figure 2. Like a modern search engine, it will search the database for
any relevant information according to the search phrase.

Figure 4: Constraint and temperature effects observed from CTOD tests

data of a 420 MPa grade steel. Clicking on a result will bring up the
relevant material testing card, see Figure 5.

Figure 2: Illustration of searching and filtering functionalities.

311
 POST-PROCESSING PROCEDURE FOR SCATTER AND
FRACTURE TOUGHNESS ASSESSMENT

One of the goals for the database framework proposed here is to be able
to efficiently post-process a large amount of available test results
performed in a large range of testing conditions. The filtering facilities
aforementioned, have thus been combined with statistical post-
processing functionalities (under continuous development) that can
provide a huge time-saver and confidence levels for data analysis and
particularly in presence of scatter inherent to the material properties or
induced by any potential changes of testing conditions.

Downloaded from http://onepetro.org/ISOPEIOPEC/proceedings-pdf/ISOPE17/All-ISOPE17/ISOPE-I-17-584/1257033/isope-i-17-584.pdf/1 by Saudi Aramco, Pierre Martin on 24 April 2025

Post-processing procedures are available via the web interface to
support the scatter assessment. These includes cumulative probability
distributions based on median rank in addition to more advanced 2- and
3-parameter Weibull distributions.

Figure 8 and Figure 9 show examples of the median rank and Weibull
distribution of fracture toughness for weld thermal simulated 420 MPa
Figure 5: Example of material card related to CTOD data of a 420 MPa grade steels.
grade steel, including SEM pictures of post-mortem specimen.
Weibull distributions (with distribution parameters estimated using the
maximum likelihood method) are, for these microstructure and
specimen geometry, in reasonable agreement with the fracture
toughness scatters observed for various temperatures and 10 parallel
tests performed at each temperature.

Figure 6: Plot showing yield strength as function of temperature

obtained from tensile tests data of a 420 MPa grade steel. Clicking on a
result will bring up the following material testing card, see Figure 7.

Figure 8:Cumulative probability of failure (Pf) vs. fracture toughness

for SENB05 specimen of 420 MPa grade weld simulated Intercritically
Reheated Coarse Grained (ICCG) HAZ (microstructure obtained after
two thermal cycles). Markers: median rank, Lines: Two-parameter
Weibull distribution with the maximum likelihood method applied to
find the parameters.

Figure 7: Example of material card related to tensile test data of a 420

MPa grade steel.

312
 The master curve is however known to be very sensitive to the
reference temperature value, an accurate estimation of T0 is thus
crucial. Since T0 is assumed to be constant, for a given specimen
geometry (and constraint condition), another estimation of T0 has been
done based on fracture toughness data at all temperatures. The updated
Master curve shown in Figure 11 exhibits a reasonable predictive
ability against tests temperature for the ICCGHAZ simulated material.

Downloaded from http://onepetro.org/ISOPEIOPEC/proceedings-pdf/ISOPE17/All-ISOPE17/ISOPE-I-17-584/1257033/isope-i-17-584.pdf/1 by Saudi Aramco, Pierre Martin on 24 April 2025

Figure 9: Cumulative probability of failure (Pf) vs. fracture toughness
for SENB02 specimen of 420 MPa grade weld simulated Intercritically
Reheated Coarse Grained (ICCG) HAZ (microstructure obtained after
two thermal cycles). Markers: median rank, Lines: Two-parameter
Weibull distribution with the maximum likelihood method applied to
find the parameters.

Additional post-processing procedure is proposed via the database

interface to evaluate the fracture toughness temperature dependency
using the Master curve approach (Wallin 2002), and the SINTAP
procedure to calculate the reference temperature T0. The master curve Figure 11: Test results and Master curve obtained for 420 MPa grade
approach has shown to describe the temperature dependency and scatter weld simulated ICCCGHAZ with T0=-64oC.
in fracture toughness for many ferritic steels. It assumes that
K0 = 31 + 77 exp(0.019(T-T0))
T0 is the transition temperature and corresponds to the temperature
where the median fracture toughness for a 25 mm thick specimen has In presence of the scatter observed in Figure 4, the extraction of a
the value of 100 MPa√m. characteristic value of fracture toughness requires further investigation
in order to minimize the risk of overestimation and control reliably the
Results based on the median values of the individual distributions are risk of underestimation.
shown in Figure 10, with the reference temperature T0 obtained from
test results performed at -300 C. Various post-processing procedures are proposed (and are under
continue development) in this sense and accessible via the database
interface.

Characteristic fracture toughness values using the minimum-of-three

equivalent recommended in several standard (e.g. MOTE in Bristish
Standard BS 7910) are compared with results obtained using the
SINTAP based procedure (1999) with 20th percentile in Figure 12 and
Figure 13.

Figure 10: Test results and Master curve obtained for 420 MPa grade,
weld simulated ICCCGHAZ with T0=-56oC. T0 obtained from -30oC
test results.

313
 Weibull distribution. This procedure will be made available in the next
step in the database interface.

CONCLUSIONS

A database platform accessible through a customized web interface is

proposed for efficient access to large test data run at low temperatures
for various materials.

Statistical functionalities integrated in this platform are illustrated by

Downloaded from http://onepetro.org/ISOPEIOPEC/proceedings-pdf/ISOPE17/All-ISOPE17/ISOPE-I-17-584/1257033/isope-i-17-584.pdf/1 by Saudi Aramco, Pierre Martin on 24 April 2025

post-processing the fracture toughness results of weld simulated 420
MPa grade structural steel.

In presence of scatter data observed for a weld simulated ICCGHAZ

material, results confirm the reasonable ability to capture the fracture
toughness temperature dependency based on the Master curve approach
as well as to extract reasonably safe characteristic values for fracture
mechanics assessments using the SINTAP procedure.
Figure 12: Dataset of 10 specimens: Fracture toughness for SENB05
specimen of 420 MPa grade weld simulated Intercritically Reheated This database framework is expected to continuously improve by
Coarse Grained (ICCG) HAZ (microstructure obtained after two integrating the research results of Arctic Material project in order to
thermal cycles). Markers: median rank, Lines: Three-parameter strength and extend the assessment capability by better combining
Weibull distribution using the SINTAP procedure to find the material properties data and technical specifications related to crack
parameters. The BS7910 MOTE values are also indicated (second positions (orientation relative to the weld), number of test specimens
lowest of 10). (repetition), temperatures, geometries/notch depth, plate thickness,
welding procedures, microstructures description etc.

ACKNOWLEDGEMENTS
The authors wish to thank the Research Council of Norway for funding
through the Petromaks 2 Programme, Contract No.228513/E30
managed by SINTEF. The financial support from ENI, Statoil, Lundin,
Total, Scana Steel Stavanger, JFE Steel Corporation, Posco, Kobe
Steel, SSAB, Bredero Shaw, Borealis, Trelleborg, Nexans, Aker
Solutions, Kværner Verdal, Marine Aluminium, FMC Kongsberg
Subsea, Hydro and Sapa are also acknowledged.

REFERENCES

Akselsen, O. M., E. Østby, et al. (2011). “Low Temperature Toughness

in SA Welding of 420 MPa Steel”. Proc. 21st Intl Ocean and Polar
Eng Conf, Maui, Hawaii, USA, ISOPE, (4).
British Standard, "Guide to methods for assessing the acceptability of
flaws in metallic structures", BS 7910:2005.
Figure 13: Dataset with more than 10 specimens (19): Fracture Gautier, D. L., K. J. Bird, et al. (2009). “Assessment of undiscovered oil
toughness for SENB05 specimen of 420 MPa grade with crack at and gas in the Arctic”. Science, 324(5931), 1175-1179.
fusion line of weld. Markers: median rank, Lines: Three-parameter Hauge, M. (2012). “Arctic offshore materials and platform
Weibull distribution using the SINTAP procedure to find the winterisation”. International Offshore and Polar Engineering
parameters. The BS7910 MOTE values are also indicated (third lowest Conference, Rhodes, Greece, ISOPE, (4).
of 19). Hauge, M and Østby, E (2013). “Fracture Integrity of Arctic Structures
– How to Determine Characteristic Fracture Toughness for
As described by Hauge and Østby (2013), the SINTAP procedure Engineering Critical Assessment”, Proc. 23st Intl Ocean and Polar
provides a safe but not too high estimation of the characteristic fracture Eng Conf, Alaska, USA, ISOPE, (4).
value in presence of small data set (less than 10) while a small data set Kane, A. Osen, V. Ren, X. Nordhagen H. Grytten F. Akselsen Odd M.
correction is recommended. Zoric J. (2015). "A Data Management System for Material Assessment
at Low Temperature". International Offshore and Polar Engineering
The SINTAP procedure appears however not conservative on data sets Conference, Kona, Hawaii, , ISOPE, (4).
with small shape parameters (large scatter). SINTAP (1999). "SINTAP Procedure, Final version November 1999",
WEM/SINTAP/PROC_7/CONTENTS REGP (05/11/99).
It should also be noted that in presence of large data set, an improved Wallin, K. "Master curve analysis of the ‘‘Euro’’ fracture toughness
procedure is proposed by the authors' aforementioned accounting for dataset". Engineering Fracture Mechanics Vol 69 (2002) 451–481.
both the percentile and the confidence level using a 2-parameter

314