You are on page 1of 9

International Congress on Advances in Welding Science and Technology for

Construction, Energy and Transportation Systems (AWST - 2011)


24-25 October 2011, Antalya, Turkey
AWST-11/112

Advances in Fracture Assessment of Welded Structures: Roles of


Weld Metal Properties
M. Koaka, D. Genkanb
Gedik Welding Inc., Ankara Cad. 34913 eyhli, Pendik, stanbul, Turkey
a

m.kocak@gedik.com , bdgenckan@gedik.com.tr

Abstract
Material tensile and fracture toughness properties are
essential data for assessment of structural integrity of
welded structures. Both base metal (BM) and weld metal
(WM) properties should be available for complete and
non conservative design and structural integrity
assessment of the modern structures. This paper presents,
firstly an overview of the structure of the procedure and
then specifically provides the weld flaw fracture
assessment procedure. Furthermore, this paper also aims
to provide information on the use of the material and
weld metal properties in the fitness for service (FFS)
procedure FITNET.
Keywords: FFS, weld metal, fracture toughness, tensile
properties

The following input data is typically needed to conduct


FFS assessment:
Size, position and orientation of flaw,
Component and weld geometry, fabrication
procedure,
Stresses of all kinds and temperatures including
transients,
Yield or 0.2% proof strength, tensile strength
and elastic modulus, (in some cases a full stressstrain curve is required),
Fatigue/corrosion
propagation data,

fatigue,

S-N

and

crack

1. Introduction
The Fitness for Service (FFS) Procedures (e.g. FITNET,
BS 7910) can be used for the design of a new
component, for support of the fabrication or quality
control process and for assessing the structural
significance of fabrication cracks, or crack like-flaws,
that are detected. Obviously, the FFS procedures also
apply to the clarification of the failure case of a
component. All three application areas require
assessment routes with respective sets of input
parameters which enable FFS analysis to be conducted
with certain accuracy depending on the quality of the
input information.
Therefore, it is advisable to generate relevant material
properties of a component at the fabrication stage, or
retain appropriate materials, especially welds, for later
testing. In particular, the desirability of having accurate
fracture toughness data cannot be emphasized too
strongly and careful tests on weld samples by experts are
advisable. Similarly, fatigue crack growth, creep and
stress corrosion cracking data may be obtained from the
actual materials/welds of the component. The
information required should take account of the material
strain and thermal history and the appropriate
environment.

Fracture toughness ( K Ic or J or CTOD values


or, for some cases, R-curve) data; in some
cases fracture toughness can be estimated from
Charpy V-notch data.
From those listed above, this paper will only address the
material/weld metal properties.
Material properties of the structural component of
interest should generally be generated using international
material testing codes and standards. Alternatively they
should be available based on open literature or
reasonable assumptions. The latter may lead to unconservative predictions or excessive conservatism
depending on the selection of the properties of interest.
The mechanical properties to be used should be
determined at the service temperature of the component.
Principally, material properties of the region where flaw
is detected, is needed for the assessment of the structural
significance of flaws. Efforts should be made to use the
properties of the region of interest. If a component has
exposed to a service or fabrication conditions which may
lead to a change or degradation in material properties,
this should be taken into account and representative
properties should be determined.

The material properties should include all weld metal


tensile properties if the weldment is of interest. Special
testing techniques (e.g. micro-flat tensile) may be needed
for determination of such properties in narrow welds,
such as laser and electron beam welds or for heat
affected zone (HAZ) regions. Hence, all material/weld
metal properties must be adequately defined. If the
temperature varies, the temperature dependence of
material parameters must be specified.
It should be noted that flat tensile specimens containing
weld joint and loaded transverse to the weld seam do not
provide local tensile properties of the weld joint. All
weld metal specimens should be extracted along the weld
seam to determine the tensile properties of weld metal.
The difference between base metal and weld metal
tensile properties creates strength mismatched (M)
situation where plastic deformation of the weld region
and hence flaw or crack should be assessed accordingly.
The pattern of plastic deformation in the neighbourhood
of a section containing a flaw may be influenced if there
is strength mismatch, for example, when the weld metal
and base metal plate of a weldment show different yield
strength and/or strain hardening.

2. FITNET FFS Procedure


The procedure is intended to be used by a broad range of
industries across Europe, and can be used at any stage in
the life of an engineering component, e.g. design,
fabrication, operation, failure analysis or life extension as
shown in Figure 1.
There are four main modules, each covering a particular
failure or damage mechanism: fracture, fatigue, creep
and corrosion (including environmentally-assisted
cracking). These are linked by the use of a common
terminology and a single set of reference compendia
(annexes), e.g. for stress intensity, plastic collapse and
residual stress, so that a particular flaw can be rapidly
analysed for more than one failure mechanism.
Flaws (such as cracks, welding defects and corrosion
damage) can arise during the manufacture and/or use of
metallic components. For safety-critical items such as
pressure vessels and pipelines, the failure of even a
single component due to the presence of a flaw could
threaten human life, as well as having severe economic
and environmental consequences. On the other hand,
some flaws are harmless, as they will not lead to failure
during the lifetime of the component; replacement and/or
repair of such flaws would be economically wasteful.
A fitness-for-service (FFS) procedure allows flaws to be
evaluated consistently and objectively, using fracture
mechanics principles. Although several reputable fitnessfor-service procedures already exist (e.g. BS 7910, API
579-1/ASME FFS-1, SINTAP, R6), they tend to be

aimed at a particular industry sector, or a single failure


mode, or are national documents. There is therefore a
need for a unified European procedure covering a range
of industry sectors.
With this in mind, the European FITNET (FITness-forservice NETwork) consortium was convened in 2002.
The vehicle for the project was a Thematic Network, a
mechanism which promotes international collaboration
on a particular topic, primarily to collate and formalize
existing research work.
The FITNET fracture assessment procedures in
particular represent a significant advance compared with
current published FFS procedures such as API 5791/ASME FFS-1 and BS 7910. There is a hierarchy of
different approaches, designated Options 0 to 5, the
choice between them depending on the quality of
information (in particular, materials property data)
available to the user. This could range from Charpy and
tensile data only (Option 0) through to the constraintdependence of fracture toughness (Option 5). Other
Options allow crack driving force to be calculated
directly from FEA (Option 4), or permit weld metal
strength mismatch to be taken into account (Option 2).

3. Significance of Material Properties


The material properties used in the FFS analysis should,
where possible, be obtained from the specific material
concerned, in the correct product form, and at the
relevant temperature and loading rate. Effects of material
variability, testing and analysis procedures should be
taken into account as well as potential changes in
material properties in service due to thermal aging,
irradiation, and other mechanisms.
If a flaw in the vicinity of weld joint is assessed, weld
metal (and also HAZ) mechanical properties need to be
determined using all weld metal tensile specimens or if
needed using micro-flat-tensile (MFT) specimens for the
HAZ gradient properties. Respectively, fracture
toughness properties of the weld area should be
determined using respective testing procedures.
Weld strength mis-match ratio, M , should be
established for the fracture analysis.
Youngs modulus and Poissons ratio are required for the
temperature at which the component is operated. If no
material specific values are available, use can be made of
generic data. If thermal stress analysis is to be
performed, values of the material conductivity, specific
heat capacity and thermal expansion coefficient will also
be required.
In FITNET FFS procedure the ultimate tensile strength
of a material is designated as Rm and the yield strength
by the general term

. However, with respect to

specific applications it can have a slightly different

International Congress on Advances in Welding Science and Technology for


Construction, Energy and Transportation Systems (AWST - 2011)
24-25 October 2011, Antalya, Turkey
AWST-11/112

Figure1. Structure of the FITNET FFS Procedure

meaning that varies from case to case. For materials with


continuous yielding, i.e. without Lders strain, the yield
strength Y refers to the proof strength of the material

R p 0.2 that is defined for a plastic strain of 0.2%, whereas


for materials with a Lders strain it refers to the lower
yield plateau ReL .
In cases where more than one measurement of

is

available, the lowest or highest value has to be chosen


depending on what is conservative for the special
application. The flow strength, f , can be understood as
an effective yield strength taking into account strain
hardening. In FITNET it is defined as :
(1)

strength is known, the yield strength for ferritic steels


may be estimated from the following equation (2):
Y (T )

Where

Y ( RT )

105
189 MPa
491 1.8 T

(2)

is the chosen yield strength and T is the

temperature of interest in C.
The strength parameters yield strength, tensile strength
and flow strength tend to increase with increasing
loading rate in the absence of dynamic strain aging
effects. In cases of dynamic loading, the use of quasistatic tensile properties will ensure conservatism in the
fracture assessment procedure; although the user should
be aware that fracture toughness may fall with increasing
loading rate.

Temperature and strain rate dependency

Statistical aspects and margins

For situations where the operating temperature is below


room temperature, but only the room temperature yield

The yield and strength parameters are characterised by


variations due to batch effects, different locations in the
plates or bars of material and random errors in testing.

These variations are taken into account in a FITNET FFS


analysis either as lower or upper bound values depending
on what is conservative for the particular application or
as statistical parameters characterizing the respective
scatter bands. Specified minimum properties for the
grade of material can be treated as lower bound values.
Individually determined values need statistical
evaluation. An appropriate measure for this is the
coefficient of variation (COV) which refers to the ratio
of standard deviation to be mean value of the respective
property which has to be determined in a number of tests.
If no specific information is available the default values
in Table 1 can be used.
Table 1. Indicative values of the coefficient of variation for
strength properties
Variable

1 2R
r
Where

r
ln 1
2R

(5)

is formally the axial stress (load divided by

minimum cross section, the non-uniform stress state


being ignored), R is the radius of the curvature of the
neck which is modelled by the arc of a circle and r is
the minimum radius in the neck cross-section. R and r
may be measured directly during the tensile test, e.g.
using photography or a tapered ring gauge.
Strain hardening coefficient
Generally, the strain hardening coefficient N (0 < N <
1) is defined as the slope of the plastic branch of the

Coefficient of variation (default value)

Modulus of elasticity

0.05

Yield strength

0.07

Ultimate tensile strength

0.05

Tensile strength of welds

0.10

stress-strain curve, when plotted in


coordinates. In the FITNET FFS procedure a
conservative estimate for steels is given by equation (6):

Engineering and true stress-strain


For fracture assessment using Options 3, 4 and 5,
individually determined stress-strain curves have to be
available. Special emphasis has to be given to well
defined stress-strain values at strains less than 1%. For
Option 3 true stress, true strain data are mandatory for
large deformation analyses whereas either engineering
and true data may be used for small deformation
analysis. As a rule, J integral analyses can be carried
out assuming small deformation conditions, whereas for
CTOD analyses large deformations conditions need to be
considered. The true stress and strain data, t and t ,
can be determined from the engineering data,
by equations (3) and (4) :

and

0.3 1-

(6)

Rm

Above equation was obtained as a lower bound curve to


a large data set of individually determined N versus
/ Rm pairs for a wide range of steels with yield
Y
strengths between 300 and 1,000 MPa and

/ Rm

ratios between 0.65 and 0.95. Because this equation was


obtained for steels its application to non-ferrous
materials is not recommended. Several definitions of the
strain hardening exponent are used in the literature.
These are commonly also designated as N or n ;
however the corresponding values show considerable
divergence. Therefore, no use must be made of N
values provided in external sources unless it is
demonstrated that the basic definition is compatible to
that given in FITNET FFS procedure.
Lders strain

(3)
(4)
However, since they are based on the assumption of a
homogeneous strain distribution along the gauge length
of the tensile specimen, these equations are applicable
only to the onset of necking. Beyond the maximum load
the true stress should be determined from measurements
of the actual cross section diameter in the necking region.
In addition, since the neck - which by its nature is a mild
notch introduces a complex triaxial stress state further
correction is needed. The so-called Bridgman
correction provides an estimate of the uniaxial stress
that would exist if no necking took place:

For fracture analyses (except Option 0) a distinction must


be made between materials with and without Lders
strain. In many situations the data will be complete
enough to establish the yielding characteristics, but there
are other cases where this may not be the case. For
structural steels, an indication can be obtained from the
yield strength, composition and the process route.
Guidance for this decision is given in Table 2 where
these factors have been grouped according to standard
specifications. It should, however, be recognised that the
presence of Lders strain is affected not only by the
material and its test temperature but also by test method,
loading rate and specimen design. Thus the information
in Table 2 is a generalisation and applies only to the
steels listed.

International Congress on Advances in Welding Science and Technology for


Construction, Energy and Transportation Systems (AWST - 2011)
24-25 October 2011, Antalya, Turkey
AWST-11/112
Table 2. Guidance for determining whether a yield plateau i.e. Lders strain should be assumed
Yield Stress
Range (MPa)

Process Route

As-Rolled

350
Normalised

Controlled Rolled

Composition Aspects

Heat Treatment Aspects

Assume Yield
Plateau

Conventional Steels e.g. EN 10025 grades)


without microalloy additions

NA

Yes

Mo, Cr, V, Nb, Al or Ti present

NA

(No)

EN 10025 type compositions without


microalloy additions

Conventional normalising

Yes

EN 10113 type compositions with microalloy


additions

Conventional normalising

Yes

EN 10113 compositions

Yes

Light TMCR schedules


(
350

Controlled Rolled

EN 10113 compositions

500

Quenched &
Tempered

400)

Yes

Mo or B present with microalloy additions Cr,


V, Nb or Ti

Light -tempering favours no


plateau

(Yes)

Mo or B not present but microalloying


additions Cr, V, Nb or Ti are (V particularly
strong effect)

Heavy tempering

(Yes)

Light tempering

(No)

Tempering to Y < 690

(No)

Quenched &

1050

(Yes)

Heavy tempering favours plateau

Mo or B present with microalloy additions Cr,


V, Nb or Ti
500

Heavy TMCR schedules


(

Yes

400)

Y 690

No

Tempering to Y < 690

Yes

Tempering to

Tempered

Mo or B not present but microalloy additions


Cr, V, Nb or Ti are

Tempering to Y 690

(No)

As-Quenched

All compositions

NA

No

N.B. ( ) =some uncertainty, a sensitivity analysis should be carried out.

For other materials the yielding behaviour should be


individually established. If there is some doubt, it is
conservative to assume the presence of Lders strain. If
no individual measurements of the Lders strain
are
available a conservative estimate can be made as:

(7)
This relation was obtained empirically, but is assumed to
be conservative since the Lders strain is known to be
smaller or even disappear in large-scale tests, in
particular in the presence of bending stress components.
Fracture toughness
In metallurgical terms, materials can fail by one of two
mechanisms: ductile or brittle. The brittle mechanism is

characteristic of ferritic and bainitic (martensitic) steels


at low temperatures, but it can occur in other materials.
Unless crack arrest occurs, the initiation of brittle
fracture can coincide with structural failure. As the
temperature of a ferritic or bainitic steel is increased, the
fracture toughness rises markedly, until a temperature is
reached where brittle fracture is no longer the prevailing
failure mechanism. This is the transition temperature,
above which the fracture mechanism is by ductile
tearing. Ductile fracture is also the failure mechanism of
most other materials and it is characterised by a fracture
resistance curve. In contrast to the brittle fracture
toughness, which appears as a single value of toughness
defining the initiation of brittle fracture, the resistance
curve is a rising function of crack extension. This
resistance curve property means that, in structures made
of ductile materials, structural failure does not generally
coincide with the onset of ductile tearing, but occurs after
some amount of crack growth.

The amount of crack growth leading to structural failure


is dependent upon geometry, loading and the materials
resistance curve. An initiation analysis uses a
characteristic value of toughness equivalent to the onset
of cracking, whether brittle or ductile, while a ductile
tearing analysis needs characteristic values that allow for
ductile tearing.

ductile fracture and showing dimpled region, while


cleavage fracture region is showing the fracture surface
of the CGHAZ of a ferritic steel.

Fracture toughness values characteristic of brittle fracture


tend to be highly scattered and are dependent on the size
and constraint of the specimen. In the ductile-to-brittle
transition region, scatter in fracture toughness is also
significant so that the transition temperature is also likely
to have a degree of uncertainity. In ferritic and bainitic
steels there is also a zone of temperature immediately
above the first appearance of ductile fracture where
brittle fracture can occur after some small amount of
ductile tearing.
In all of these cases, the toughness can be characterised
by the same parameter. This can be written in terms of
K , J or
parameters. Standardized methods are
available for determining all fracture toughness values
needed for this procedure, and the treatments for
determining the characteristic values given in the
following sections are designed around these
standardised methods. In the absence of appropriate
fracture toughness data, Charpy data can be used to give
a conservative estimate of fracture toughness, by using
appropriate correlations. Any of the fracture toughness
parameters can be described as K , J or , depending
on the approach adopted by the user.
Preferably, the test data should be specifically collected
from samples of the actual material of interest. If this is
not possible, it may be obtained from replicate samples.
In either case, care should be taken to ensure that the
data is representative of the material product form of
interest at the appropriate temperature, and any potential
degradation due to manufacture or service is taken into
account. In the case of welded structures, if the tests are
performed on replicate samples, care should be taken to
ensure that they truly represent the original weldments
used in the structure. The samples should be welded with
welding procedures, base materials and consumables as
used for the service application and should take account
of restraint during welding and of PWHT if applicable.
The welding procedures should be appropriately
documented (see EN 288-3 and EN 25817 for instance).
In all cases care should be taken to ensure that the
fracture data have been obtained from specimens whose
fatigue crack tip samples the appropriate weld
microstructures i.e. the data is representative of the
weldment region (weld metal, HAZ) whose toughness is
to be assessed.
It should be mentioned that HAZ (particularly coarse
grained HAZ) region usually lower fracture toughness
and may show brittle fracture as fractography in Figure
2 is clearly showing the weld metal region failed with

Figure 2. Fracture surface showing the ductile (weld metal) and


brittle (CGHAZ) regions of a ferritic weld joint.

Ideally fracture toughness tests for use in modern


assessment methods should not be undertaken with the
aim of measuring maximum load J or CTOD values, as
the resistance curve (R-Curve) is a more correct way of
defining fracture toughness for these cases.
FFS Procedures are usually use fracture toughness data
in terms of K. Hence, the preliminary task is the
conversion of the available fracture toughness data into
units of K .
The relationships for converting J or
given by equations (8) and (9):

J .E /(1

(8)

where E is Youngs modulus and

mRe .E /(1

to K are

is Poissons ratio

(9)

where Re is yield or proof stress and m = 1.5. This


value of the coefficient m was determined from a large
data set on structural steels and represents the 25th
percentile. As such it should give conservative estimates
of toughness. It should be noted, however, that the value
of m = 1.5 is only appropriate for treating data from
deep notch, high constraint specimens and may not be
appropriate for steels with a very low work hardening
exponent i.e. N < 0.05, or for materials other than steel.
For the comparison of data from different size
specimens, adjustment of all results from specimens can
be carried out whose thickness is not 25 mm according
to:

Ki

K25

20 ( K B 20)( B / 25)0.25

(10)

International Congress on Advances in Welding Science and Technology for


Construction, Energy and Transportation Systems (AWST - 2011)
24-25 October 2011, Antalya, Turkey
where K B is the toughness of a specimen of original
thickness B . Note that K B is the original value for the
fracture toughness in a specimen of thickness B (here
B refers to the nominal thickness, regardless of any
side-grooving etc.) and K i is the adjusted value,
equivalent to that for a specimen size of 25 mm.
Use of maximum load fracture toughness data
In some circumstances the only type of fracture
toughness data available may be values relating to the
first attainment of a maximum load plateau (Maximum
load CTOD, J or K J ). Such toughness values
represent ligament size-dependent points on the fracture
toughness resistance curve (R-curve) of the material.
Such values can provisionally be used in this procedure,
particularly for predominantly tension-loaded structures,
but the following aspects should be considered in
connection with their use.
Increasing specimen thickness while maintaining a
constant ligament length leads to an increase in ductilebrittle transition temperature but may also lead to an
increase in the level of upper shelf fracture toughness.
The maximum load fracture toughness is mainly
dependent on the ligament length, and the smaller the
ligament is, the smaller the maximum load fracture
toughness will be. If cleavage occurs after the attainment
of maximum load, the use of maximum load fracture
toughness values will usually be conservative compared
to the cleavage initiation value. Maximum load fracture
toughness values may be used in an analysis where
historic data of this type are the only data available,
provided that the ligament size of the fracture toughness
specimen is equal to or smaller than the corresponding
dimension in the structure.
If the possibility of brittle fracture in the component can
be excluded with confidence, the maximum load fracture
toughness value can be treated as if it was a ductile value
and the size adjustment is not applied. The characteristic
value should be determined from three or more results.
Care should be taken in the case of a high resistance to
tearing since the maximum load fracture toughness in
such cases usually exceeds the toughness corresponding
to the onset of stable crack extension. In such cases, a
full tearing analysis may often be more appropriate.
Where resistance to ductile tearing is low, the maximum
load fracture toughness is closer to the toughness level at
initiation of tearing and can be used with relative
confidence. Where it is possible to establish the amount
of ductile tearing occurring prior to the attainment of the
maximum load, this should be taken into account by
increasing the flaw size in the structure by the same or a
greater amount. In other situations, the maximum load
toughness value should be taken to correspond to 6%
crack growth relative to the ligament.

AWST-11/112
Charpy impact energy - fracture toughness
correlations
If the appropriate fracture toughness data for use in
structural integrity assessments is not available, the use
of correlations between Charpy-V impact energy and
fracture toughness can provide the fracture toughness
value to be used in the assessment.
For the application of the respective correlations, three
different regimes of material behaviour are identified as
follows. Lower shelf is then defined as the temperature
region where the shear fracture appearance (SFA) is
20% and the impact energy is less than 27J. Upper shelf
is the temperature region where SFA = 100% of th
ligament. The range in-between is considered as the
ductile-to-brittle transition region.
A lower bound correlation that can be used in the lower
shelf for a wide range of steels is given by equation (11):

K mat

12 CV

20

25
B

0.25

(11)

20

where K mat is the fracture toughness of the material


(MPa m), CV is the Charpy energy (J) and B is the
crack front length (mm) or the specimen thickness. The
K mat estimate should be based on the CV value
corresponding to the minimum of three tests or its
equivalent.
In case Charpy and fracture toughness data are
unavailable, a very conservative lower bound toughness
for ferritic steels can be taken as :
K mat

20 MPa m

(12)

In the transition region the fracture toughness transition


curve can be described as a function of T27 J as follows:

K mat 20 11 77 exp 0.019 T T27 J 3 C

25
B

0.25

0.25

1
ln
1 Pf
(13)

where Pf is the probability of fracture toughness being


lower than K mat .
Often T27 J may not be known but some other data
describing the transition region might be given for a
material. In such cases it is recommended to use the
procedure involving incomplete transition curves.
When the CVN data does not allow the determination of
T27 J , it can be conservatively estimated from :
where TCv is the temperature at which Charpy data CV
are available and the constant C is a function of the

materials yield strength (

) and upper shelf energy

(14)

USE .

34 C

with

USE
35.1 14.3
Y

fracture toughness properties are to be expected. During


the fracture toughness testing of very narrow weld metal
zones (laser and electron beam welds, or HAZ regions)
crack path deviation occurs towards lower strength
regions as shown in Figure 3 below. Hence, the
toughness value generated from such specimens will not
represent intrinsic fracture toughness properties of the
zone of interest.

(15)

in MPa and USE in J.

If the upper shelf energy is unknown, it should be


estimated as being twice the highest measured impact
energy value.
When Charpy tests exhibit fully ductile behaviour (SFA
= 100%), this does not automatically imply that the
structure itself will also be operating in the upper shelf at
the same temperature. In particular, for thick sections and
for some low carbon and low sulphur steels, the full
thickness material may exhibit transitional behaviour at
temperatures corresponding to upper shelf behaviour in
Charpy specimens. If brittle behaviour can be excluded,
a lower bound estimate of the upper shelf fracture
toughness corresponding to a ductile crack extension of
0.2 mm ( K J 0.2 ), taken as the engineering approximation
of the onset of ductile tearing ( K mat ), can be evaluated
from:

(16)
where CVus is the upper shelf energy (J), E is Youngs
Modulus (MPa) and
is Poissons ratio. Other amounts
of crack extension than 0.2 mm can be substituted into
above equation, if desired. A more user friendly form of
equation;

Kmat

K J 0.2 11.9 CVus 0.545

(17)

The applicability of above written equations is limited to


yield strength values between 170 and 1000 MPa and
upper shelf energies in the range 20 to 300 J. Alternative
upper shelf correlations may be used, provided they can
be proven to have sufficient level of conservatism for the
material to be assessed.
Fracture toughness determination of strength mismatched welds
Strength mis-match affects the constraint conditions near
the crack tip, and hence effects of mis-match on the

Figure 3. Fracture path deviation into the lower strength base


metal regions during the fracture toughness testing of highly
overmatched laser beam welds of ferritic steel.

This situation is a consequence of the remote plasticity


development in the neighbouring base metal, as
illustrated in Figure 4 and hence obtained fracture
toughness values are meaningless. It is obvious that
plastically heterogeneous interfaces (both sides of the
narrow fusion zone with much lower strength level than
fusion zone) near to the crack tips experience high strain
concentrations and this often leads to crack kinking out
of the high strength but lower toughness region.

4. Conclusions
Fitness for Service procedures require special form of
material input data. Correct generation and full
availability of this data is important for the outcome of
the fitness for service analysis of welded structures. This
paper has intended to provide short information on this
matter.

International Congress on Advances in Welding Science and Technology for


Construction, Energy and Transportation Systems (AWST - 2011)
24-25 October 2011, Antalya, Turkey
AWST-11/112

OverOver-Match (OM)

UnderUnder-Match (UM)
B

B
Homogeneous
Base Metal

2H

YSW
YSB < YSW

YSB > YSW

Over-Matched
Plastic Zone

YSW
M=
YSB
Mis-Match Factor

Under-Matched

B
YSW

YSW

YSB < YSW

YSB < YSW

Yield Strength of
Base Metal

LBW and FSW


of AlAl-alloys

Figure 4. Schematic description of crack tip plasticity due to weld strength mis-match.
LBW: Laser Beam Welding, FSW: Friction Stir Welding.

References
This paper has given short summary of the use of
material data in FITNET FFS Procedures and hence
below listed publications are giving wider information on
the subject.
[1]

[2]

[3]

Koak M., Fitness for service analysis of structures using the


FITNET procedure: an overview, In: Offshore Mechanics and
Arctic Engineering (OMAE 2005). Proc., 24th Int. Conf.,
Halkidiki, Greece, 12-17 June 2005. Publ: New York, NY
10016, USA; ASME; 2005.
Seib E. and M. Koak, Fracture analysis of strength
undermatched welds of thin-walled aluminium structures using
FITNET Procedure, IIW-1709-05, Welding in the World, 49 :(
11/12), pp. 58-69, 2005.
Koak M., Seib E., and Motarjemi A. , Treatments of structural
welds using FITNET fitness-for-service procedure: FITNET 06013 2006, Proceedings of the International Conference on
Fitness-for-service (FITNET 2006): 17-19 May, Shell Global
Solutions, Amsterdam, The Netherlands (ISBN 978-3-00021084-6).

[4]

Hadley I., and Moore P., Validation of fracture assessment


procedures through full-scale testing: FITNET 06-018,
Proceedings of the International Conference on Fitness-forservice (FITNET 2006): 17-19 May, Shell Global Solutions,
Amsterdam, The Netherlands (ISBN 978-3-00-021084-6).

[5]

Koak M., FITNET fitness-for-service procedure: an


overview, FITNET 06-04, in Proceedings of the International
Conference on Fitness-for-service (FITNET 2006): 17-19 May,
Shell Global Solutions, Amsterdam, The Netherlands (ISBN
978-3-00-021084-6).

[6]

Seib E., Uz V. M., and Koak M., Fracture analysis of thinwalled laser beam and friction stir welded Al-alloys using the
FITNET procedure, FITNET 06-019, Proceedings of the
International Conference on Fitness-for-service (FITNET 2006):
17-19 May, Shell Global Solutions, Amsterdam, The
Netherlands (ISBN 978-3-00-021084-6).

[7]

Koak M., FITNET fitness-for-service procedure: an


overview, International Institute of Welding, Welding in the
World, vol.51, no.5-6., pp. 94-105May-June 2007.

[8]

Hadley I., Validation of the European FITNET fitness-forservice procedure: incorporation of weld strength mismatch into
fracture assessment (Options 2 and 3). TWI Industrial
Members report 890/2007.

[9]

Koak M., Webster S., Janosch J.J., Ainsworth R.A. and Koers
R., Fitness for Service Analysis of Structures using FITNET
Procedure: An overview, Proc. of the 9 th Int. Conf on
Engineering Structural Integrity: Research, Development and
Application, 15-19 Oct. 2007, Beijing, EMAS Publishing, Vol.
I, pp. 161-179, 2007.

[10] Hadley I., Validation of the European FITNET fitness-forservice procedure: Use of fracture assessment Option 4, TWI
Industrial Members report 893/2008.
[11] Koak M., Webster S., Janosch J.J., Ainsworth R.A., Koers R.,
FITNET Fitness-for-Service (FFS) Procedure (Vol. 1), ISBN:
978-3-940923-00-4., 2008.
[12] Koak M., Hadley I., Szavai S., Tkach Y., Taylor N., FITNET
Fitness-for-Service (FFS) Annex (Vol. 2), ISBN: 978-3940923-01-1., 2008.