Vous êtes sur la page 1sur 14

PLEASE SCROLL DOWN FOR ARTICLE

This article was downloaded by: [ABM Utvikling STM / SSH packages]
On: 21 July 2009
Access details: Access Details: [subscription number 792960683]
Publisher Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Ships and Offshore Structures
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t778188387
Assessment of design criteria for fatigue cracking from weld toes subjected to
proportional loading
I. Lotsberg
a
a
Department for Offshore Structures, Hvik, Norway
Online Publication Date: 01 June 2009
To cite this Article Lotsberg, I.(2009)'Assessment of design criteria for fatigue cracking from weld toes subjected to proportional
loading',Ships and Offshore Structures,4:2,175 187
To link to this Article: DOI: 10.1080/17445300902733998
URL: http://dx.doi.org/10.1080/17445300902733998
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
Ships and Offshore Structures
Vol. 4, No. 2, 2009, 175187
Assessment of design criteria for fatigue cracking from weld toes subjected
to proportional loading
I. Lotsberg

Department for Offshore Structures, DNV, Veritasveien 1, 1322 Hvik, Norway


(Received 20 December 2008; nal version received 9 January 2009)
For fatigue design it is necessary to provide guidelines on how to calculate fatigue damage at weld toes based on S-N data
when the principal stress direction is different from that of the normal direction to the weld toe. Such stress conditions
are found in details in different types of plated structures. Some different fatigue criteria for these stress conditions are
presented in design standards on fatigue design. Criteria used by the International Institute of Welding (IIW), Eurocode,
British Standards and in the DNV (Det Norske Veritas) standards have been assessed against some relevant fatigue test data
presented in the literature. Only proportional loading conditions have been considered here. (By proportional loading it is
understood that the principal stress direction is kept constant during a load cycle.) An alternative equation for calculation
of an equivalent or effective stress range based on stress normal to the weld toe and shear stress at the weld toe has been
proposed. The proposed methodology can be used for nominal S-N curves, which can be used together with a hot spot stress
S-N curve with stresses read out from nite element analysis. The different design criteria are presented in this paper together
with recommendations on analysis procedure.
Keywords: design criteria; fatigue; welded plate structures; weld toe; principal stress direction; proportional loading
1. Introduction
For fatigue design it is necessary to have guidelines on
how to calculate fatigue damage at weld toes based on S-N
data when the principal stress direction is different from
that of the normal direction to the weld toe. Details of
such stress conditions are found in different types of plated
structures, such as at connections with soft brackets and
at tubulars penetrating plates in ship structures, e.g. DNV
CN 30.7 (2005) and Lotsberg (2004). Some fatigue design
standards have advised to use the largest principal stress
range within 45

to the normal to the weld toe together


with an S-N curve derived for stress ranges normal to the
weld toe for fatigue design. Reference is made to the IIW
(1996), British Standards Institution 5400 (1980), BS 7608
(1993) and DNV CN 30.7 (2005).
The International Institute of Welding (IIW) (2007) de-
cided to change the angle for largest principal stress range
direction from45

to 60

, which is now included in the


present version of the IIW fatigue design guidelines. The
same revision was also made in DNV-RP-C203 (2005).
During actual design cases it has been found that the new
criterion can have signicant impact on the design of some
special details and it is observed that designers have difcul-
ties meeting the required fatigue life at these hot spots when
using this procedure. Therefore, it was decided to make a
further assessment of recommended design criteria based

Email: inge.lotsberg@dnv.com
on a review of some relevant fatigue test data from the liter-
ature. This work is presented in more detail in the following.
2. Fatigue test results for inclined welds
from literature
A literature search has been performed in order to nd fa-
tigue test data where the principal stress direction relative
to the weld toe has been a varying parameter. It was found
that such fatigue tests have been performed by Kim and
Yamada (2004) using test specimens as shown in Figure 1.
They presented the test results at an IIWconference in 2004.
Fatigue test data from specimens shown in Figure 1(b) and
test specimens in Figure 1(a), where fatigue cracking oc-
curred at the straight part of the weld toe, are selected for
assessment. The fatigue test data used in the present assess-
ment are listed in Table 1. The following notations are used
on test specimens in this table: Gfor gusset specimen and N
for non-load-carrying cruciform specimen. Specimen no-
tation also include angle as dened in Figure 2 and the
specimen number.
Reference is made to Figure 2 for calculation of stresses
in the test specimen. Equilibrium in the loading direction
of the test specimen gives
(
//
sin +

cos )w/ cos =


1
w (1)
ISSN: 1744-5302 print / 1754-212X online
Copyright C
2009 Taylor & Francis
DOI: 10.1080/17445300902733998
http://www.informaworld.com
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
176 I. Lotsberg
Figure 1. Fatigue test specimens used by Kim and Yamada (2004).
Equilibrium in the transverse direction to the specimen
gives

//
cos =

sin (2)
From these equations the following stresses are derived:

//
=
1
sin cos ,

=
1
cos
2

(3)
The stresses as function of unit stress for varying values of
are shown in Figure 3.
3. Design procedures
3.1. Procedure in Eurocode 3 (2005)
The procedure in Eurocode 3 (2005) is a summation of
calculated fatigue damages from normal and shear stress
ranges at the weld toe. This can be presented in the form of
a design equation as
D

+D

= 1.0 (4)
3.2. Procedure in IIW (2007)
Two alternative procedures are presented in the latest rec-
ommendations on fatigue design by IIW (2007).
1. Principal stress direction.
2. Quadratic interaction of allowable normal stress range
and shear stress range.
The different methodologies are presented in detail in
the following sections:
Principal stress direction in Section 2.2.3.1 of IIW
(2007):
In the case of biaxial stress state at the plate surface,
it is recommended to use the principal stress which is ap-
proximately in line with the perpendicular to the weld toe,
i.e. within 60

. The other principal stress may be anal-


ysed, if necessary, using the fatigue class for parallel welds
in the nominal stress approach. Reference is also made to
Figure 4.
Quadratic interaction of allowable normal and shear
stress ranges of IIW (2007):
The effects of combination of normal and shear stresses
shall be veried by
_

S,d

R,d
_
2
+
_

S,d

R,d
_
2
= CV (5)
where
R,d
or
R,d
is the design resistance stress range
for the specied number of cycles and the appropriate FAT
class for normal and shear stresses at the weld toe.
S,d
or
S,d
are the corresponding design stress ranges. CV is
a comparison value, which is given as 1.0 for proportional
loading in table 4.1 of IIW(2007).
3.3. Procedure by Kim and Yamada (2004)
Kim and Yamada (2004, 2005) proposed to use the follow-
ing expression for effective stress:

Eff
=
1
cos (6)
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
Ships and Offshore Structures 177
Table 1. Fatigue test data (from Kim and Yamada (2004)).
Number of Stress range
Specimen cycles (MPa) Comments
G0-01 216,000 190
G0-02 237,000 190
G0-03 1,564,000 120
G0-04 3,428,000 98
G30-03 603,000 190
G30-04 608,000 190
G45-05 1,447,000 190
G45-06 735,000 204
G45-07 1,278,000 190
G45-08 982,000 190
G45-09 2,270,000 152 Run-out
N0-01 198,000 206
N0-02 170,000 203
N0-03 470,000 160
N0-04 556,000 160
N0-05 1,415,000 136
N0-06 630,000 136
N0-07 990,000 136
N0-08 2,788,000 113
N0-09 6,764,000 113 Runout
N15-01 360,000 206
N15-02 324,000 203
N15-03 479,000 161
N15-04 867,000 160
N15-05 760,000 160
N15-06 1,577,000 136
N15-07 1,739,000 136
N15-08 984,000 136
N15-09 2,366,000 123
N15-10 4,860,000 123 Run-out
K-30-01 502,000 206
K-30-02 389,000 203
K-30-03 1,264,000 174
K-30-04 2,053,000 159
K-30-05 1,620,000 159
K-30-06 6,449,000 138
K-30-07 10,000,000 138 Run-out
K-30-08 10,000,000 123 Run-out
where the stress
1
and the angle are dened in
Figure 2.
3.4. Alternative procedure
A combined stress range (or effective stress range), taking
into account the stress normal and the shear stress along the
weld toe can be expressed in the following form:

Eff
=
_

+
2
//
(7)
where the stress components are explained in Figure 2. The
S-N category will depend on the type of detail in relation
to the normal stress. This will result in different values
as presented in Table 2. The combined stress range should
Figure 2. Denition of symbols and stress components.
be used together with an S-N curve that is selected as if this
stress was acting normal to the weld toe.
The details tested in Figure 1 are classied as E follow-
ing DNV-RP-C203 for small thicknesses of the attachments
and F for larger thicknesses when =0. The test results for
= 0 are presented in Figure 5 together with the E-curve.
(Reference is made to Table 3 for relation between notations
on S-N curves used in DNV-RP-C203 (2005), IIW (2007)
and Eurocode (2005)).
For presentation of mean S-N curves it is assumed that
a standard deviation in logarithmic format is 0.20. (The
design curve is dened as mean minus two standard devia-
tions assuming the test data to follow a normal distribution
in a logarithmic format.)
S-N category C2 may be used for continuous shear
stress in a full penetration weld according to Table A.8 in
DNV-RP-C203. Assuming that the shear stress is classied
as C2, the following equation for combined or effective
stress is derived from Table 2 when the effective stress is
combined with S-N curve E:

Eff
=
_

+0.64
2
//
(8)
The basis for this equation is also illustrated in Figure 6
where
1
=
//
at = 45

. The stress components in


Equation (8) are to be combined with different S-N curves
as shown in Figure 6. (When
1
is acting normal to a weld
toe, it is classied as an E detail or FAT 80. When the detail
is subjected to shear along the weld, S-N curve C2 or FAT
100 should be applied.)
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
178 I. Lotsberg
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Angle (deg)
F
a
c
t
o
r
Tau
Sigma normal
Figure 3. Stress as function of unit stress for varying value of .
4. Comparison of design procedures with fatigue
test data
4.1. Eurocode (2005)
The detail shown in Figure 1 is classied as FAT 80 fol-
lowing Eurocode 3 (2005) and IIW (2007) for = 0. This
is the same as the E-curve in DNV-RP-C203. The test re-
sults for = 0 are presented together with the E-curve in
Figure 5.
The S-N curves for stress range normal to the weld toe
and shear stress can be presented as
N

= a

N
//
= a
//

m
//
(9)
For stress normal to the weld the design S-N curve is FAT
80 with m = 3.0. The design S-N curve for shear stress in
Eurocode 3 and IIW is FAT 100 with m = 5.0.
Figure 4. Figure from IIW (2007) showing stress to be used for fatigue analysis.
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
Ships and Offshore Structures 179
Table 2. Values for based on S-N curves in DNV-RP-C203
(2005).
Stress direction parallel
with the weld
Stress direction normal
to the weld toe C C1 C2
D 0.518 0.646 0.810
E 0.409 0.510 0.640
F 0.322 0.402 0.504
F1 0.254 0.316 0.397
F3 0.201 0.250 0.314
G 0.160 0.199 0.250
A scaling of the stress ranges in the fatigue test data is
performed such that the test data can be presented for n =
10
6
cycles. The following scaling of stress are made for
comparison with Eurocode 3 and IIW (m = 3.0 for normal
stress and 5.0 for shear stress):

10
6 =
t est
_
n
t est
10
6
_
1/3.0

//10
6 =
// t est
_
n
t est
10
6
_
1/5.0
(10)
From the equation for summation of damages in Equation
(4) and (9) the following expression for shear stress resis-
tance for Eurocode 3 is derived:

R,d
=
_
a
//
_
1
n

3.0
S,d
a

__
1/5.0
(11)
Table 3. Relations between notations in DNV-RP-C203
and IIW and Eurocode 3.
DNV-RP-C203 IIW and Eurocode 3
B1 160
B2 140
C 125
C1 112
C2 100
D 90
E 80
F 71
F1 63
F3 56
G 50
W1 45
W2 40
W3 36
This equation together with fatigue test data is shown in
Figure 7. It is observed that this gure shows a rather small
interaction effect between normal and parallel stress ranges
at = 45

.
4.2. IIW(2007)
The IIW quadratic interaction equation on stress reads
_

S,d

R,d
_
2
+
_

S,d

R,d
_
2
= 1.0 (12)
10
100
1000
Number of cycles
S
t
r
e
s
s

r
a
n
g
e

(
M
P
a
)
G0
N0
E design
E mean
100,000 1,000,000 10,000,000
Number of cycles
S
t
r
e
s
s

r
a
n
g
e

(
M
P
a
)
G0
N0
E design
E mean
Figure 5. Test data for principal stress normal to the weld.
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
180 I. Lotsberg
1

//

C2 (FAT 100)
E (FAT 80)
1
//
C2 (FAT 100)
E (FAT 80)
Figure 6. Illustration of stresses
1
=
//
when =45

that are
combined with different S-N curves.
From Equations (9) and (12) the following expression for
shear stress resistance for IIW is derived:

R,d
=
_
a
//
n
_
1/5
_
1
2
S,d
_
n
a

_
2/3
(13)
This equation together with test data is shown in Figure 8.
It is observed that the fatigue test data, also at = 45

,
are in good agreement with the mean line for quadratic
interaction on stress as shown.
4.3. Present proposal
For comparison of Equation (8) with fatigue test data, the
fatigue test data are scaled with respect to stress range to
correspond to 10
6
cycles. The following scaling of stress
range is made for comparison with DNV-RP-C203 (inverse
negative slope of S-N curve m = 3.0):

10
6 =
test
_
n
test
10
6
_
1/3.0

//10
6 =
// test
_
n
test
10
6
_
1/3.0
(14)
The test data for different principal stress range directions
are presented in Figure 9. It is observed that there is a good
correspondence between the test data and the proposed de-
sign equation for effective stress.
4.4. Kim and Yamada (2004) and the present
proposal
A comparison using different equations for effective stress
is presented in Figure 10. The effective stress fromEquation
(6) (Kim and Yamada 2004) is compared with the present
proposal from Equation (8). As the present proposal ts the
test data well, it may be concluded that the procedure by
Kimand Yamada (2004) is slightly on the conservative side.
From Figure 10 it is observed that the effective stress is
reduced by a factor 0.63 at an angle =45

. This means that


using the principal stress within an angle 45

to the normal
to the weld toe becomes conservative for large angles.
0
20
40
60
80
100
120
140
160
0 50 100 150 200 250
Normal stress (MPa)
P
a
r
a
l
l
e
l

s
t
r
e
s
s

(
M
P
a
)
Design
Mean
G0-01
G30-03
G45-05
N0-01
N15-01
N30-01
0
20
40
60
80
100
120
140
160
0 50 100 150 200 250
Normal stress (MPa)
P
a
r
a
l
l
e
l

s
t
r
e
s
s

(
M
P
a
)
Design
Mean
G0-01
G30-03
G45-05
N0-01
N15-01
N30-01
Figure 7. Test data presented in format of interaction equation in Eurocode (2005).
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
Ships and Offshore Structures 181
0
20
40
60
80
100
120
140
160
0 50 100 150 200 250
Normal stress (MPa)
P
a
r
a
l
l
e
l

s
t
r
e
s
s

(
M
P
a
)
Design
Mean
G0-01
G30-03
G45-05
N0-01
N15-01
N30-01
0
20
40
60
80
100
120
140
160
0 50 100 150 200 250
Normal stress (MPa)
P
a
r
a
l
l
e
l

s
t
r
e
s
s

(
M
P
a
)
Design
Mean
G0-01
G30-03
G45-05
N0-01
N15-01
N30-01
Figure 8. Test data presented in format of IIW quadratic interaction equation on stress range components.
5. Comparison of design procedures
and recommended approach
A review of fatigue test data considering principal stress
direction relative to the weld toe geometry has been per-
formed. Based on this assessment one may reconsider the
text related to Figure 4 from IIW (2007).
The IIWmethod with the calculation of allowable stress
ranges for stress normal to the weld and shear stress sep-
arately and using a quadratic interaction equation on these
is considered to t test data very well. A methodology with
adding the damages from these stress components together,
which is used by Eurocode (2005), is not that good.
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
160.00
180.00
0 50 100 150 200 250
Normal stress (MPa)
P
a
r
a
l
l
e
l

s
t
r
e
s
s

(
M
P
a
)
Design
Mean
G0-01
G30-03
G45-05
N0-01
N15-01
N30-01
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
160.00
180.00
0 50 100 150 200 250
Normal stress (MPa)
P
a
r
a
l
l
e
l

s
t
r
e
s
s

(
M
P
a
)
Design
Mean
G0-01
G30-03
G45-05
N0-01
N15-01
N30-01
Figure 9. Test data for principal stress having different angles with the normal to the weld toe compared with proposed effective stress
range.
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
182 I. Lotsberg
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 10 20 30 40 50
Angle between normal to the weld and principal stress direction
E
f
f
e
c
t
i
v
e

s
t
r
e
s
s
Present proposal
Kim and Yamada (2004)
Figure 10. Comparison of effective stress.
An alternative design equation has been proposed that
combines the stress normal to the weld and the shear stress
at the weld toe into an effective stress range that can be
entered into a single S-N curve for calculation of number
of cycles to failure. This design approach is considered to
be efcient for use together with stresses read out from
nite element analyses.
The general expression for effective stress is derived
from Equation (7) where values are derived from Table 2.
If the hot spot stress is derived by extrapolation of stresses to
the weld toe or to the intersection line from read-out-points
t /2 and 3t /2 as explained in DNV-RP-C203, this hot spot
stress should be combined with S-N curve D. This means
that = 0.81 in Equation (7) for calculation of effective
stress range. If the hot spot stress is based on a read-out-
point at t /2, the hot spot stress should be combined with the
E-curve and = 0.64 in Equation (7).
It is realised that the present classication of details
with guidance on S-N curve is not rened enough for some
special details in DNV-RP-C203 (2005). Therefore, a more
detailed classication is proposed as shown in Figure 11
and Table 4.
Figures 11(a and b) are intended to be used for nominal
stress analyses. The selection of the E and F curves depends
on the thickness of attachment as presented in Table A.7 of
DNV-RP-C203.
Figure 11(c) is intended to be used in special cases
when using the hot spot stress methodology as presented in
DNV-RP-C203 (2005). Figure 11(c) can be used together
with the hot spot stress methodology in general.
The stress range in both the two principal directions
should be assessed with respect to fatigue. Here a design
criterion for within an angle 45

to the normal to the


weld has been assessed against fatigue test data. For a prin-
cipal stress direction 45

< 90

, an S-Ncurve for stress


direction parallel with the weld can be used due to the ef-
fective stress reduction factor of 0.63 at = 45

as was
given in Section 4.4.
Table 4. Classication of details and selection of S-N curve.
Angle
in Figure 11
Detail classied as
F for stress direction
normal to the weld
Detail classied as E for
stress direction normal
to the weld
S-N curve when using
the hot spot stress
methodology
030 F E D
3045 E D C2
4560 D C2 C2
6075 C2 C2 C2

7590 C2

C2

C2

A higher S-N curve may be used in special cases. See Table A-3 in DNV-RP-C203 for further information. http://webshop.dnv.com/
global/category.asp?c0=2624&c1=2627.
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
Ships and Offshore Structures 183
C2 C2
F F
E
E
D
D
Principal stress
direction
Weld
toe
Section
C2 C2
F F
E
E
D
D
Principal stress
direction
Weld
toe
Section
(a)
(b)
(c)
C2 C2
E E
D D
Principal stress
direction
Weld
toe
Section
C2 C2
E E
D D
Principal stress
direction
Weld
toe
Section
C2 C2
D D
Principal stress
direction
Weld
toe
Section
C2 C2
D D
Principal stress
direction
Weld
toe
Section

Figure 11. Classication of details and selection of S-N curve: (a) Detail classied as F for stress direction normal to the weld, (b) Detail
classied as E for stress direction normal to the weld, (c) S-N curve when using the hot spot stress methodology.
Different design criteria and interaction equations are
presented in Figure 12 and Figure 13 for comparison of
design criteria at 10
6
and 10
7
cycles respectively.
6. Derivation of hot spot stress using nite element
analysis
Two alternative methods can be used for hot spot stress
derivation in the revised DNV-RP-C203 (2008). These are
described as follows:
Method A:
For modelling with shell elements without any weld
included in the model, a linear extrapolation of the stresses
to the intersection line from the read-out points at 0.5t and
1.5t from the intersection line can be performed to derive
hot spot stress.
For modelling with three-dimensional elements with
the weld included in the model, a linear extrapolation of the
stresses to the weld toe from the read-out points at 0.5t and
1.5t from the weld toe can be performed to derive hot spot
stress.
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
184 I. Lotsberg
0
20
40
60
80
100
120
140
0 20 40 60 80 100 120
Normal stress (MPa)
P
a
r
a
l
l
e
l

s
t
r
e
s
s

(
M
P
a
)
Present proposal
IIW (2007)
Eurocode (2005)
Figure 12. Comparison of design equations at 10
6
cycles.
The notations for stress components are shown in Figure
14 and Figure 15.
The effective hot spot stress to be used together with
the hot spot S-N curve D (FAT 90) is derived as

Eff
= max
_

_
_

+0.81
2
//

1
|
2
|
(15)
where
= 0.90 if the detail is classied as C2 with stress
parallel to the weld at the hot spot (ref. Table A-3 in DNV-
RP-C203 (2008)).
= 0.80 if the detail is classied as C1 with stress
parallel to the weld at the hot spot (ref. Table A-3).
= 0.72 if the detail is classied as C with stress
parallel to the weld at the hot spot (ref. Table A-3).
0
10
20
30
40
50
60
70
80
0 10 20 30 40 50 60
Normal stress (MPa)
P
a
r
a
l
l
e
l

s
t
r
e
s
s

(
M
P
a
)
Present proposal
IIW (2007)
Eurocode (2005)
Figure 13. Comparison of design equations at 10
7
cycles.
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
Ships and Offshore Structures 185
Principal stress
direction
Weld
toe
Section
Fatigue crack
//

//

Principal stress
direction
Weld
toe
Section
Fatigue crack
//

//

//

//


Figure 14. Fatigue cracking along weld toe.
The principal stresses are calculated as

1
=

+
//
2
+
1
2
_
_

//
_
2
+4
2
//

2
=

+
//
2

1
2
_
_

//
_
2
+4
2
//
(16)
The rst equation for effective stress (Equation (15)) is
made to account for the situation with fatigue cracking
along a weld toe as shown in Figure 14 and the second and
third equations are made to account for fatigue cracking
when the principal stress direction is more parallel with
the weld toe as shown in Figure 15.
Principal stress
direction
Weld
toe
Section
Fatigue crack
//

//

Principal stress
direction
Weld
toe
Section
Fatigue crack
//

//

//

//

Figure 15. Fatigue cracking when principal stress direction is more parallel with weld toe.
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
186 I. Lotsberg
Method B: For modelling with shell elements without
any weld included in the model the hot spot stress is taken
as the stress at the read-out point 0.5t away from the inter-
section line.
For modelling with three-dimensional elements with the
weld included in the model the hot spot stress is taken as
the stress at the read-out point 0.5t away from the weld toe.
The effective hot spot stress to be used together with
the hot spot S-N curve D (FAT 90) is derived as

Eff
= max
_

_
1.12
_

+0.81
2
//
1.12
1
1.12 |
2
|
(17)
where ,
1
and
2
are explained under Method A.
The rst equation for effective stress (Equation (17))
is made to account for the situation with fatigue cracking
along a weld toe as shown in Figure 14 and the second and
third equations are made to account for fatigue cracking
when the principal stress direction is more parallel with the
weld toe as shown in Figure 15.
7. Conclusions
The purpose of the present assessment has been to arrive
at guidelines on how to calculate fatigue damage at weld
toes based on S-N data when the principal stress direc-
tion is different from that of the normal to the weld toe.
Some different fatigue criteria have been assessed together
with fatigue test data from the literature. Only proportional
loading has been considered here.
The method used by Eurocode (2005) is to calculate the
fatigue damage due to stress range normal to the weld toe
and the damage due to the shear stress at the weld toe and
then adding the damages together. This sum should be less
than 1.0. It is observed that this method shows somewhat
lowinteraction effect between normal stress and shear stress
when compared with the test data.
IIW(2007) presents two methods for proportional load-
ing. The rst one is to calculate the principal stress at the
weld toe on a nominal basis. If the angle between the prin-
cipal stress and the normal to the weld toe is less than 60

,
this principal stress is used together with S-N curves for a
detail with stress acting normal to the weld toe.
The second method in IIW (2007) is to calculate allow-
able stress ranges for stress normal to the weld and shear
stress separately and use a quadratic interaction equation
on these.
From the present assessment it is found that the rst
method is considered to be conservative and it should be
explained that this approach is conservative for large angles
and that the document includes more accurate alternatives
that can be recommended to be used. The second method is
found to be in good agreement with fatigue test data and is
the preferred methodology based on comparison with test
data.
An alternative equation for calculation of an effective
stress range based on stress normal to the weld toe and shear
stress has been proposed. The equation for effective stress
range reads

Eff
=
_

+
2
//
(18)
where

= stress normal to the weld,

//
= stress parallel with the weld,
= factors from Table 5.
The -factor is derived from the S-N curve constants
in the design standard such that the calculated fatigue life
using this equation for effective stress equals that using
nominal S-N curves for stress parallel with the weld toe.
Equation (18) is considered to be efcient for calcula-
tion of fatigue life when used together with the hot spot
stress concept (or structural stress concept) with stresses
read out from nite element analysis. This methodology
can also be used for nominal stress S-N curves.
The new alternative design approaches are included in
a revision of DNV-RP-C203 that was issued in April 2008.
Table 5. Recommended -factor for design.
-factor
S-N curve for stress
normal to the weld
S-N
classication
C2 (FAT
100) for pure
shear stress
S-N
classication
C1 (FAT
112) for pure
shear stress
D (FAT 90)
When used as nominal S-N curve
or hot spot S-N curve by
extrapolation of stress to the
weld toe or the intersection
line from read-out points 3t /2
and t /2 (t = plate thickness)
0.81 0.64
E (FAT 80)
When used as nominal S-N curve
or hot spot S-N curve by stress
from read-out points t /2 (t =
plate thickness) from the weld
toe or the intersection line
0.64 0.50
F (FAT 71) 0.50 0.40
F1 (FAT 63) 0.40 0.31
F3 (FAT 56) 0.31 0.25
G (FAT 50) 0.25 0.20
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9
Ships and Offshore Structures 187
References
British Standards Institution (1980). BS 5400: steel, concrete and
composite bridges. Code of practice for fatigue. London: BSI.
Part 10.
British Standards Institution. (1993). BS 7608: Code of practice
for fatigue design and assessment of steel structures. London:
BSI.
DNV CN 30.7. 2005. Fatigue assessment of ship structures. Oslo
(Norway): DNV.
DNV-RP-C203. 2005. Fatigue design of offshore steel structures.
Oslo (Norway): DNV.
DNV-RP-C203. 2008. Fatigue design of offshore steel structures.
Oslo (Norway): DNV. Available from: http://webshop.dnv.
com/global/category.asp?c0=2624&c1=2627
Eurocode 3. (2005). EN-1993-1- 9. Design of steel structures, part
19: fatigue. ICS 91.010.30
International Institute of Welding. (1996). Fatigue design of
welded joints and component. Recommendations of IIWjoint
working group XIII-XIV. Edited by A. Hobbacher. Abington
Publishing.
International Institute of Welding. (2007). Recommendations for
fatigue design of welded joints and components. De Gaulle
Cedex (France): IIW. (Document XIII-2151r1/XV-1254r1-
07).
Kim IT, Yamada K. 2004. Fatigue behaviour of llet welded joints
inclined to a uniaxial load. Document No. XIII-2021-04. The
International Institute of Welding.
Kim IT, Yamada K. 2005. Fatigue life evaluation of welded joints
under combined normal and shear stress cycles. Int J Fatigue
27(6):695701. (Pergamon-Elsevier Science Ltd. SCI).
Lotsberg I. 2004. Fatigue design of welded pipe penetrations in
plated structures. Mar Struct. 17(1):2951.
D
o
w
n
l
o
a
d
e
d

B
y
:

[
A
B
M

U
t
v
i
k
l
i
n
g

S
T
M

/

S
S
H

p
a
c
k
a
g
e
s
]

A
t
:

0
9
:
4
8

2
1

J
u
l
y

2
0
0
9

Vous aimerez peut-être aussi