Presentation for DCAMM, 02.

December 2022
—
Fatigue assessment of welded joints with the notch stress
approach – Background and current developments

J. Baumgartner
Agenda

 Fatigue assessment approaches for welded structures

 Effective notch stress approach
 Notch stress approach with reference radii
 Effective stress approach
 Consideration in rules and guidelines
 Summary

Seite 2 05.12.2022 © Fraunhofer LBF - open -

 Fatigue assessment
approaches for welded
structures

Seite 3 05.12.2022 © Fraunhofer LBF - open -

Fatigue assessment approach for welded joints
Properties that influence the fatigue life
Linear-elastic

approaches
stress-based

Seite 4 05/12/2022 © Fraunhofer LBF - open -

Working effort for the assessment with different approaches
Accuracy vs. component complexity

Ref.: J. Samuelsson

Seite 5 05/12/2022 © Fraunhofer LBF - open -

Application of linear-elastic approaches
Determination of S-N curves for welded joints

Seite 6 05/12/2022 © Fraunhofer LBF - open -

FAT class / characteristic value, slope and knee point
Definition within IIW-recommendations and DIN EN 1999-1-3

IIW: FAT class DIN EN 1999-1-3: Δ𝜎𝜎𝐶𝐶

 Defines endurable stresses at N = 2 ⋅ 106 N = 2 ⋅ 106 load cycles
 Is defined for a stress ratio 𝑅𝑅 = 0,5 𝑅𝑅 = 0,0
 Is defined for a probability of survival 𝑃𝑃𝑆𝑆 = 97,7% 𝑃𝑃𝑆𝑆 = 97,7%
 Is based on the stress range Δ𝜎𝜎 = 2 ⋅ 𝜎𝜎𝑎𝑎 Δ𝜎𝜎 = 2 ⋅ 𝜎𝜎𝑎𝑎
 Is based on maximum principal stresses 𝜎𝜎1 𝜎𝜎1

 Stresses normal to weld:

 slope m=3 m=3
 knee point 𝑁𝑁𝑘𝑘 = 1 ⋅ 107 𝑁𝑁𝑘𝑘 = 5 ⋅ 106
 Shear stresses:
 slope m=5 m=5
 knee point 𝑁𝑁𝑘𝑘 = 1 ⋅ 108 𝑁𝑁𝑘𝑘 = 1 ⋅ 108

Seite 7 05/12/2022 © Fraunhofer LBF - open -

 Effective notch stress
approach

Seite 8 05.12.2022 © Fraunhofer LBF - open -

Effective notch stresses
Modeling the radii with r = 1.0 mm

 Rounding of the weld

toe and weld root
notches with a reference
radius of 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 = 1𝑚𝑚𝑚𝑚
 Calculation of maximum
notch stress
 At weld toes, an effective
notch stress of at least
𝜎𝜎
𝐾𝐾𝑊𝑊 = 𝑒𝑒 > 1.6 should be
𝜎𝜎ℎ𝑠𝑠
assumed
(𝜎𝜎𝑒𝑒 : notch stress,
𝜎𝜎ℎ𝑠𝑠 : hot spot stress)

Seite 9 05/12/2022 © Fraunhofer LBF - open -

Background information: Derivation of the FAT 225
Modeling the radii with r = 1.0 mm

 Fatigue tests on welded

transverse stiffeners with
thickness range
8𝑚𝑚𝑚𝑚 ≤ 𝑡𝑡 ≤ 80𝑚𝑚𝑚𝑚
 At different R-ratios
 Evaluation of endurable
notch stresses for survival
probability 𝑃𝑃𝑆𝑆 = 60%

Seite 10 05/12/2022 © Fraunhofer LBF - open -

Background information: Derivation of the FAT 225
Modeling the radii with r = 1.0 mm

Welded joint type Fatigue notch Nominal fatigue Local fatigue strength
(Structural steels) factor Kf strength (R = 0) ∆σ1 = 2 Kf ∆σnA [MPa]

 Average stress range σnA [MPa], Ps =

90% 50% 10%
( R = 0, Ps = 50%):
∆σ1 = 357 MPa

 Average stress range

(R = 0.5, Ps = 50%):
∆σ1 = 325 MPa (fR = 1.10)

 Value for R = 0.5, Ps = 97.7%:

∆σ1 = 236 MPa
(jσ = 1.37, Tσ = 1:1.50)

 FAT – value according to the IIW-numerical system:

∆σ1 = 225 MPa (N = 2·106, R = 0.5, Ps = 97.7%)
Ref.: R. Olivier, V. B. Köttgen

Seite 11 05/12/2022 © Fraunhofer LBF - open -

Background information: Derivation of the FAT 225
Modeling the radii with r = 1.0 mm

 Derivation of FAT225 for IIW-Recommendations in

Joint Working Group 2008

 Document JWG-XIII-XV-197-08: “Database for the

effective notch stress method at steel”

 Evaluation of data with 𝑅𝑅 = 0.4

Seite 12 05/12/2022 © Fraunhofer LBF - open -

Recommendation in the IIW-Recommendation (2016)
FAT-values for steel and aluminium

a) Steel

b) Aluminium

Seite 13 05/12/2022 © Fraunhofer LBF - open -

Mild notches
Additional treatment

 In case the notch is mild, a second check needs to be performed:

 Calculation of the notch stress 𝜎𝜎𝑒𝑒
 Calculation of the structural stress 𝜎𝜎ℎ𝑠𝑠
𝜎𝜎𝑒𝑒
 Calculation of 𝐾𝐾𝑤𝑤 =
𝜎𝜎𝑤𝑤

 𝐾𝐾𝑤𝑤 needs to be at least 1.6. If not, the notch stress needs to be

𝜎𝜎
artificially increased to meet the condition 𝐾𝐾𝑤𝑤 = 𝑒𝑒 = 1.6.
𝜎𝜎𝑤𝑤

Seite 14 05/12/2022 © Fraunhofer LBF - open -

Validation of FAT225
Evaluation by Mikkel Pedersen

 Conclusions:
 For most fillet-welded joints,
the experimental fatigue
data agrees reasonably well
with the current IIW
guidance (FAT 225 S–N curve)
 Butt joints show lower
endurable stress. 𝐾𝐾𝑊𝑊 might
be increased to 𝐾𝐾𝑊𝑊 = 2

Pedersen, M. M. / Mouritsen, O. / Hansen, M. R. / Andersen, J. G. /

Wenderby, J.: Re-analysis of fatigue data of welded joints using
the notch stress approach, 2010. International Journal of
Fatigue , Vol. 32, No. 10, p. 1620-1626

IIW-Rec.: The effect of possible

Seite 15 05/12/2022 © Fraunhofer LBF - open -
misalignment is not included.
Concept according to Neuber / Radaj with rf = 1 mm
Theoretical background

 For welded joints: Worst-case assumption with rreal = 0 mm

 Fictitious (effective) radius: rf = 1.0 mm

Seite 16 05/12/2022 © Fraunhofer LBF - open -

Concept according to Neuber / Radaj with rf = 1 mm
Theoretical background

Seite 17 05/12/2022 © Fraunhofer LBF - open -

Concept according to Neuber / Radaj with rf = 1 mm
Theoretical background

Idea:
Due to the inhomogeneity
of the microstructure not
the max. theoretical value
is accountable for fatigue
but a reduced value.

Ref.: Neuber 1968

Seite 18 05/12/2022 © Fraunhofer LBF - open -

Concept according to Neuber / Radaj with rf = 1 mm
Analytical formulae to determine effective stresses by stress averaging

Ref.: Neuber 1958

Formula 𝜌𝜌1 = 𝜌𝜌 + 2𝜖𝜖 is a simplification.
Only valid for sharp crack-like notches!

Seite 19 05/12/2022 © Fraunhofer LBF - open -

Effective notch stress approach
Summary

 The stress distribution should not be altered if the notches are rounded
 Universal applicability for both weld toe and weld root notches
 Higher application effort is necessary in order to apply the notch stress approach
 Size effects are automatically considered
 Check for mild notches has to be performed. Additional calculation of structural hot spot stress
necessary.

Seite 20 05/12/2022 © Fraunhofer LBF - open -

 Notch stress approach with
reference radii

Seite 21 05.12.2022 © Fraunhofer LBF - open -

Overview: Variants of the notch stress approach
Modeling the radii with r = 1.0 mm

Modelling of the weld toe and weld root radii with a reference radius
Determination of notch stresses as stress concentration factors (max. stresses)
Variations of the concepts:
Reference radius: rref = 1.00 mm
Reference radius: rref = 0.30 mm
Reference radius: rref = 0.05 mm
Real radius: rreal

Stress Hypothesis: Max. Principal or von Mises stresses

Choice of slope k and knee point Nk identical to nominal or structural stress approach

Seite 22 05/12/2022 © Fraunhofer LBF - open -

Notch stress approaches using various reference radii
Available reference radii and their use depending on sheet thickness

Seite 23 05/12/2022 © Fraunhofer LBF - open -

Choice of Radius in Dependence of Sheet
Thickness and Failure Location

- open -
Fatigue assessment of spot-welded welded overlap spec.
Assessment with notch stress approach
Choice of radius: typical gap between sheets
in automotive industry (body-in-white)

Investigations by Zhang in 2000

r = 0.05 mm
Thicknesses:
t = 0.8 mm / t = 0.8 mm
t = 1.5 mm / t = 1.5 mm
Material: FeP04 Approximation of the SIF 𝐾𝐾𝐼𝐼 = 𝑓𝑓𝐼𝐼 lim 𝜎𝜎𝑒𝑒 𝜋𝜋𝜋𝜋
𝜌𝜌→0
with notch stresses
Zhang, G. & Richter, B.: A new approach to the numerical fatigue-life prediction of spot-
welded structures. Fatigue Fract. Eng. Mater. Struct., 2000, 23, 499-508

Seite 25 05/12/2022 © Fraunhofer LBF - open -

Fatigue assessment of spot-welded welded overlap spec.
Assessment with notch stress approach

Investigations on laser beam welded overlap joints

Thicknesses:
t = 0.8 mm / t = 0.8 mm
Eibl, M.
t = 1.9 mm / t = 1.9 mm Berechnung der Schwingfestigkeit
laserstrahlgeschweißter Feinbleche mit …
Material: DC04 PhD-thesis, Technische Universität Darmstadt, 2003

Seite 26 05/12/2022 © Fraunhofer LBF - open -

Fatigue assessment of spot-welded welded overlap spec.
Assessment with notch stress approach

Nominal stress
approach:
High scatter
No applicability
for complex welded
joints

Nominal stress
approach:
Sig. reduction
of scatter
Higher reliability

Eibl, M.
Berechnung der Schwingfestigkeit laserstrahlgeschweißter Feinbleche mit …
PhD-thesis, Technische Universität Darmstadt, 2003

Seite 27 05/12/2022 © Fraunhofer LBF - open -

Fatigue assessment of spot-welded welded overlap spec.
Assessment with notch stress approach

Nominal stress
approach:
High scatter
No applicability
for complex welded
joints

Nominal stress
approach:
Sig. reduction
of scatter
Higher reliability

Eibl, M.
Berechnung der Schwingfestigkeit laserstrahlgeschweißter Feinbleche mit …
PhD-thesis, Technische Universität Darmstadt, 2003

Seite 28 05/12/2022 © Fraunhofer LBF - open -

Fatigue assessment of spot-welded welded overlap spec.
Assessment with notch stress approach

Notch stress approach with

𝑟𝑟ref = 0,05 mm:
Influence of weld geometry
is visible. Possible reasons:
Difference in failure criteria –
Crack initiation and crack
propagation
Influence of material:
High-strength and austenitic
steel show lower endurable
stresses

Störzel, K.; Baumgartner, J.; Bruder, T. & Hanselka, H.

Festigkeitskonzepte für schwingbelastete geschweißte Bauteile
Materials Testing, 2011, 53, 418-426

Seite 29 05/12/2022 © Fraunhofer LBF - open -

Influence of failure location on the endurable notch stresses
Assessment with notch stress approach

Results from a research

project: Differences
between endurable notch Weld toe failure Weld root failure
stresses for weld toe and
weld root failure using
𝑟𝑟ref = 0,05 mm

Bruder, T.; Störzel, K.; Baumgartner, J. & Hanselka, H.

Evaluation of nominal and local stress based
approaches for the fatigue assessment of seam welds
Int. J. Fatigue, 2012, 34, 86-102

Seite 30 05/12/2022 © Fraunhofer LBF - open -

Influence of failure location on the endurable notch stresses
Assessment with notch stress approach

Results from a research

project: Differences
between endurable notch
stresses for weld toe and
weld root failure using
𝑟𝑟ref = 0,05 mm

FAT400?

Baumgartner, J. / Schmidt, H. / Ince, E. / Melz, T. / Dilger, K.

Fatigue assessment of welded joints using stress averaging and
critical distance approaches. 2015. Welding in the World , Vol.
59, No. 5. p. 731-742

Seite 31 05/12/2022 © Fraunhofer LBF - open -

 Effective stress approach

Seite 32 05.12.2022 © Fraunhofer LBF - open -

Notch Stresses from FE-Models of Notched Bars
Varying notch radius r and notch opening angle ω
Normal stresses Shear stresses
Results from a research
𝜔𝜔 = 0°
project: Differences 𝜔𝜔 = 0°
𝜎𝜎𝑒𝑒 = 𝐶𝐶 ⋅ 𝑟𝑟 −0.5
between endurable notch 𝜎𝜎𝑒𝑒 = 𝐶𝐶 ⋅ 𝑟𝑟 −0.5

stresses for weld toe and

weld root failure using
𝑟𝑟ref = 0,05 mm

𝜎𝜎𝑒𝑒 = 𝐶𝐶 ⋅ 𝑟𝑟 −0.33
𝜎𝜎𝑒𝑒 = 𝐶𝐶 ⋅ 𝑟𝑟 −0.20
𝜔𝜔 = 135°
𝜔𝜔 = 135°

an = 10 mm an = 105 mm

Seite 33 05/12/2022 © Fraunhofer LBF - open -

Stress singularity at a sharp V-notch
Influence of loading mode and notch opening angle

Stress singularity depends

on
Notch opening angle
Loading mode:
𝜆𝜆1 : Mode 1 (opening)
𝜆𝜆2 : Mode 2 (shearing)
𝜆𝜆2 : Mode 3 (tearing)

Seite 34 05/12/2022 © Fraunhofer LBF - open -

Stress singularity at a sharp V-notch
Influence of loading mode and notch opening angle

Visualizatio Normal stresses 𝜔𝜔 = 0°

n of stress
singularity
at notch
Relation
between
notch
stress and
opening
angle
depends
on opening
angle

𝜔𝜔 = 150° Ref.: Kranz2008

Seite 35 05/12/2022 © Fraunhofer LBF - open -

Experimental and numerical campaign for advanced assessment approach
Fatigue tests

R-values:
−1 ≤ 𝑅𝑅 ≤ 0.5
Material:
Steel, Rp = 350 MPa

Sheet 1.5 mm / 1.0 mm /

4 mm
thickness: 3.0 mm 2.5 mm

Intended failure
location:
Weld root

Weld toe

Seite 36 05/12/2022 © Fraunhofer LBF - open -

Experimental and numerical campaign for advanced assessment approach
Finite-Element models

Using average, measured local weld geometry

Fixed notch radius of rref = 0,05 mm
Verification of model using experimental strain analysis

t = 1,5 mm

t = 3,0 mm

Seite 37 05/12/2022 © Fraunhofer LBF - open -

Experimental and numerical campaign for advanced assessment approach
Calculation of local stresses and stress gradients

 Maximum notch stress

Maximum Principal
von Mises
 Stress gradient / stress
course in the ligament
At position of
t = 1,5 mm
maximum stress x

Normal to the
surface

t = 3,0 mm

Seite 38 05/12/2022 © Fraunhofer LBF - open -

Effective notch stress approach
Determination of effective stresses

∗
𝜌𝜌
1 Moore /
Neuber: 𝜎𝜎𝑒𝑒𝑒𝑒𝑒𝑒 = ∗ ⋅ � 𝜎𝜎 𝑥𝑥 𝑑𝑑𝑑𝑑 𝜎𝜎𝑒𝑒𝑒𝑒𝑒𝑒 = 𝜎𝜎(𝑥𝑥 = 𝑎𝑎)
𝜌𝜌 0
Peterson:

Seite 39 05/12/2022 © Fraunhofer LBF - open -

Evaluation of endurable effective stresses
Additional test data from literature

Overall 400 fatigue tests (with N < Nk)

Seite 40 05/12/2022 © Fraunhofer LBF - open -

Evaluation of endurable effective stresses
Additional test data from literature

Seite 41 05/12/2022 © Fraunhofer LBF - open -

Evaluation of endurable effective stresses
Results for the Stress Averaging and the TCD Approach

 Minimum scatter for all approaches

 For max. Principal stress:
ρ* = 0.35 mm
(Neuber: ρ* = 0.40 mm)
 Theoretical relationship between a and ρ*
confirmed:
ρ* = 4∙a

Seite 42 05/12/2022 © Fraunhofer LBF - open -

Evaluation of endurable effective stresses
Support factors depending on notch opening angle

Application of this
1 σ(𝑥𝑥 = 0) approach leads to
𝑛𝑛𝑁𝑁𝑁𝑁 = =
𝜂𝜂𝑁𝑁𝑁𝑁 𝜎𝜎𝑒𝑒𝑒𝑒𝑒𝑒 similar scatter
compared to the
numerically derived
support factors

𝑛𝑛𝑁𝑁𝑁𝑁 =
1
=
𝜏𝜏(𝑥𝑥 = 0)
Caution:
𝜂𝜂𝑁𝑁𝑁𝑁 𝜏𝜏𝑒𝑒𝑒𝑒𝑒𝑒
This might not be true
for all notch  load
configurations!

Seite 43 05/12/2022 © Fraunhofer LBF - open -

Evaluation of endurable effective stresses
Support factors depending on notch opening angle

Endurable
effective stress
equals the
endurable
nominal stress
for un-notched
specimens!

Seite 44 05/12/2022 © Fraunhofer LBF - open -

Evaluation of endurable effective stresses
Endurable effective stresses using Neuber’s stress averaging approach

Seite 45 05/12/2022 © Fraunhofer LBF - open -

 Consideration in rules and
guidelines

Seite 46 05.12.2022 © Fraunhofer LBF - open -

Consideration in rules and guidelines
Overview

 Notch stress approach:

IIW-Recommendations 2023 and FKM-Guideline 2020:

 Effective
 Effective stress notch stress
approach: approach:
FKM-Guideline Eurocode 3
2020 FKM-Guidel.
IIW-Recomm.

Seite 47 05/12/2022 © Fraunhofer LBF - open -

 Summary

Seite 48 05.12.2022 © Fraunhofer LBF - open -

Summary
… of the most important results

 The effective notch stress approach with a reference radius of 𝑟𝑟ref = 1 mm is a reliable approach for
the local assessment of welded joints
 For welded thin joints with failure at a sharp weld root notch, the radius of 𝑟𝑟ref = 0.05 mm leads to
reliable results
 In case of weld toe failure at welded thin sheets, two approaches can be applied. The stress-gradient
based support effects can be considered by:
 Using FAT-classes that are depending on the notch opening angle.
 Performing the stress averaging approach based on numerically determined stress gradients and using only one
FAT-class (or using the relation between stress gradients and opening angle)
 The effective stress approach can also be used to evaluate joints that are ground

Seite 49 05/12/2022 © Fraunhofer LBF - open -

Contact
—
Dr. Jörg Baumgartner
Tel. +49 6151 705-474
[email protected]

Fraunhofer LBF
Bartningstr. 47
64289 Darmstadt
www.lbf.fraunhofer.de
Many thanks for your kind
attention
—