Available via license: CC BY 4.0
Content may be subject to copyright.
Welding in the World
https://doi.org/10.1007/s40194-024-01829-y
RESEARCH PAPER
Round robin study on the determination of weld geometry parameters
- Part A: analysis of a reference specimen
Matthias Jung1·Moritz Braun2·Jan Schubnell1·Heikki Remes3
Received: 27 February 2024 / Accepted: 16 August 2024
© The Author(s) 2024
Abstract
The weld toe is known to be a critical point of fatigue failure in many welded constructions. Especially for research purposes
but also for improving fatigue life predictions, the weld toes geometry is often described by a set of parameters, including the
weld toe radius and the flank angle. There is no universal agreement on the definition of the geometry parameters as well as
on measuring routines. To get an overview over used techniques and comparability between research labs, a comprehensive
round robin study was conducted over the past years. Two measuring tasks were given to the participants. Part A: A machined
specimen with well known geometry inspired by a cruciform joint was analyzed and the results were compared with the
actual dimensions of the specimen. Part B: Welded specimens with unknown geometry were measured by the participants
and the results were bench-marked against each other. The present study summarizes the findings of Part A. The study gives
an overview over used measuring techniques, the influence of measuring equipment and the comparability of the results in
the scientific community. Most of the participants achieved good results with their respective measuring methods for radii
larger than 1 mm. Smaller radii tend to be overestimated.
Keywords 3D scans ·Weld classification ·Weld geometry ·Local toe geometry ·Weld toe radius ·Flank angle
IIW Document No. XIII-2993-2023
1 Introduction
As numerous experimental an theoretical studies have shown,
the weld toe is known to be a fatigue critical location in
welded joints [1–7]. This can be explained by the stress con-
centration, caused by the weld bead. The weld geometry and
Recommended for publication by Commission XIII - Fatigue of
Welded Components and Structures.
BMatthias Jung
matthias.jung@iwm.fraunhofer.de
Moritz Braun
moritz.braun@dlr.de
Jan Schubnell
jan.schubnell@iwm.fraunhofer.de
Heikki Remes
heikki.remes@aalto.fi
1Fraunhofer IWM, Freiburg, Germany
2German Aerospace Center (DLR), Institute of Maritime
Energy Systems, Geesthacht, Germany
3Aalto University, Espoo, Finland
the magnitude of the stress concentration is affected by a vari-
ety of influencing factors, like welding technique, shop floor
position, weld type and welding parameters [6–13]. Many
post weld treatment techniques focus specifically on improv-
ing the shape of the weld toe [4,14,15]. Knowledge about
the geometry of the weld is often used in order to assess the
stress concentration. This can be done by simulations or by
utilizing parametric formulae [8,16–22].
A common way for describing a welds shape is based on
a simplified model of the weld, which usually incorporates
at least a weld toe radius and a flank angle (Fig. 1). For deter-
mining the parameters a variety of methods was used in the
past. Manual measurements were done on the weld toe itself
or by obtaining a silicone negative of the welds shape. With
manual testing a dependency on the operator was reported
[23–26]. More recent approaches focus on establishing auto-
mated measuring techniques based on weld profiles obtained
by 3D-scanning [8,22,24,25,27]. By this approach the
dependency on the operator is reduced and the change of
the shape along the weld can be analyzed by repeating the
procedure on different positions.
123
Welding in the World
Fig. 1 Exemplary geometric description of fillet and butt weld [28]
A discussion amongst participants of the IIW Commission
XIII meetings in the past years unveiled large differences
in the way weld toes are usually characterized. To get an
overview over used methods as well as a quantification of
the differences in the obtained results, a Round Robin study
was set up amongst the members of IIW C-XIII. The study
was divided into two parts, the analysis of a small specifically
designed and machined specimen (Part A) and the analysis
of 3 welded specimens from other research projects (Part B).
In the present work, the results for the first part (Part A) of
the study are presented.
2 Study design and specimen description
The goal of the study is to achieve an understanding of dif-
ferences in the weld toe radius measurement results between
different research institutions. In order to compare the results
a Round-Robin study was set up. In total 16 institutions
contributed to the study by analyzing a specimen, that was
specifically designed for this study.
The specimen used in this study was designed with the
overall shape of a cruciform joint in mind. To make sure, that
the geometry of the specimen is well defined, the specimen
was machined from an aluminum sheet using wire-cut electri-
cal discharge machining (EDM). The shape of the specimen
is shown in Fig. 2. Each of the “weld toes” was machined
with a different radius, varying from virtually zero (with a
very small radius caused by the EDM wire with a diameter
of about 0.15 mm) to 5.3 mm. The specimen was anodized to
achieve a dark, almost black surface color and it was labelled
with a coordinate system. Each “weld toe” was denominated
with a letter A-H. The dimensions of the specimen were not
communicated to the participating institutions.
The specimen was sent to each of the participating
institutions. The institutions were instructed to analyze the
Fig. 2 Shape of the round robin specimen
123
Welding in the World
Table 1 Digitizer types used in
the study Digitizer type Measuring principle Number of series
Laser line sensor Optical Light sectional Line projection (2D) 9
Fringe projection sensor Fringe projection (3D) 4
Laser distance sensor Point projection (1D) 1
Line confocal sensor Chromatic confocal surface profiling 1
3D-Microscopy Focal depth variation microscopy 1
Roughness Tactile 1
specimen with the techniques they usually use. To get con-
sistent results a Excel template was sent with the specimen
to be filled out by the participants, which should include
the radius obtained for each of the “weld toes”, a standard
deviation of the measurements if the radii were measured
more than once as well as rudimentary information about
the equipment used, e.g. the digitizer, it’s resolution and if
applicable a description of how the specimen was prepared
for analysis. The results were collected and evaluated by the
studies authors. To increase the number of participants the
studies authors gave an insight into the results during the
Round-Robin period on the IIW Commission XIII annual
and intermediate meetings.
3 Overview over used measurement
principles
Table 1gives an overview over the digitizer types used by
the participants. In general, the digitizer generates a digital
representation of the weld which is then analyzed to acquire
the weld toe geometry parameters. Most digitizers types (5
of 6) can be classified as non-tactile, optical measurement
methods. Among those a further distinction can be made by
measuring principle in light sectional methods, chromatic
confocal surface profiling and focal depth variation [29].
The light sectional methods rely on the triangulation of a
projected light pattern, that is altered by the surface topol-
ogy. The principle is shown in Fig. 3a for the point projection
method. A point is projected from a laser diode onto the sur-
face that should be measured. the apparent shift as seen from
an angle, where a optical line sensor is positioned, relates to
the distance between sensor and surface. Laser line sensors
extend that principle into a second dimension by faning out
the laser point into a straight line and replacing the optical
line sensor by a 2D image sensor. Fringe projection sensors
add a second camera and project a series of patterns onto
the surface. By that, a whole area can be digitized at once. It
should be noted, that for the measurement tasks in this study
a description of a weld profile is needed, which is inherently
a two-dimensional information. With the laser distance sen-
sors this can only be achieved by moving the sensor along
the measuring path. That means, that at least one positional
axis with an applicable sensing apparatus is needed. Laser
line sensors can produce the needed information in a single
measurement as long as the sensor is oriented precisely per-
pendicular to the weld. From fringe projection measurements
the weld can be oriented and the weld profile can be extracted
by applying a cross-sectional plane virtually using common
3D software.
The chromatic confocal surface profiling method pro-
duces a spectrum form white light in such a way, that each
Fig. 3 Measuring principles
123
Welding in the World
Table 2 Number of cross-sections evaluated for each radius
Number of cross-sections
N/A 1 2...10 >10
Number of series 1 1 8 7
color of the spectrum can be assigned a known distance from
the sensor. By this the reflected lights spectrum shows a peak
in one frequency range, which is then related to the distance
from the sensor. The principle is illustrated in Fig. 3b. Focal
depth variation utilizes the limited depth-of-field of a micro-
scope. A sweep over different focal planes and filtering the
defocused areas allows for reconstruction of the 3D surface
information. In the study also one tactile method was used,
which used the information from a roughness and contour
measurement system.
4 Results
The 16 participants reported 17 measurement series in total
(one participant conducted two independent measurement
series with different sensors/approaches). The main content
of the responses is a radius and angle value for each position
on the specimen with given standard deviation if more than
one cross-section was evaluated for each position. That was
the case in 15 of 17 series, as stated in Table 2. In four series,
the specimen was treated with a coating agent to tarnish the
surface, which was usually referred to by the participants
as white powder. Within the responses information about the
resolution and accuracy of the used scanning system was also
requested, but the data that was provided was too inconsistent
to be evaluated.
The measured radii reported by the participants are shown
in Fig. 4a. The median and the interquartile range fit the
machined radii very closely. For the 17 series a variety of dig-
itizer principles was used, with the most common being Laser
line sensors and Fringe projection sensors. An overview is
given in Table 1. Isolating laser line and fringe projection
sensors from the dataset allows for a comparison of the two
most common measuring principles (Fig. 4b).
Figure 5a shows the absolute and relative error of the
radius measurements. While single measurements produced
rather large errors, the majority achieves accurate results for
radii larger than 1mm. The small radii F and G are usually
overestimated.
Figure 5b shows the absolute error of the angle measure-
ments, with the target value being 135° for all radii A-F.
For the majority of participants the angle measurements are
accurate to a few degrees. The larger deviations (larger than
20°) are caused by two of 17 measurement series, both with
unknown evaluation method.
For 6 series a rudimentary description of the evalua-
tion method was included with the returned data. These
descriptions allow for a categorization in ‘automated’, ‘semi-
automated’ and ‘manual’ evaluations, with two series in
each category. The remaining 11 series are categorized as
‘unknown’. In Fig. 6the accuracy of radius and angle for
each individual measurement is shown. While most mea-
Fig. 4 Overview over radius measurements
123
Welding in the World
Fig. 5 Errors of measured values
surements concentrate in the area of low error with respect
to both parameters, a series of measurements shows inac-
curacies in the angle while maintaining low relative radius
error. Other measurements show good angular measurements
while having high relative radius errors. Only two individual
measurements (both semi-automatic) show large deviations
in both radius and angle.
5 Discussion
A possible factor influencing the quality of the weld toe
geometry parameters estimation is the measuring principle
used. In total 5 different non-tactile and one tactile measuring
Fig. 6 Measuring errors for angle and radius
principle were used (see Table 1). On a first glance, the series
with the least accurate results were evaluated based on data
from laser line sensors and laser point sensors. To investigate
if the measuring principle is likely to be the cause for the
deviations, five cross-sections of radii D and E are compared
visually in Fig. 7. The scans seem to be very similar, while
the evaluated radii differ vastly from the machined radii (for
radius D up to 37%, see Table 3. To support this finding the
points in the region of the radius (3.45 mm ≤x≤7.2mm)
where selected form the respective point clouds and the radius
was calculated using a least squares approximation. The least
squares radii are much closer to the machined radius. Another
argument for the independence of the results from the chosen
measuring principle can be derived from Fig. 4b. Although
Fig. 7 Comparison of five 3D scan cross-sections for radius D
123
Welding in the World
Table 3 Radius D obtained from selected scans as measured in the
round robin study and by applying a least-squares method to the area
of interest in the point clouds
Cross-section Radius at position D rDin mm
in Fig. 7Machined Round Robin Least Squares
P1 5.3 4.99 5.35
P2 5.41 5.30
P3 3.31 5.24
P4 5.18 5.19
P5 5.29 5.30
there is a large difference in the results spread, the interquar-
tile range and the median are fitting the machined radii nicely
for laser line sensors and for fringe projection sensors both.
From this finding it can be assumed, that both measuring prin-
ciples are in general capable of producing point clouds that
allow for a adequate evaluate downstream the process. Laser
line sensors and fringe projection sensors seem to be consid-
ered appropriate for the given task by many researchers, as
13 of 17 series were evaluated based on 3D scans obtained
with sensors working with these two principles (Table 1).
Comparing the evaluation methods in Fig. 6shows, that
most series for which the evaluation method was described
achieve at least good results for determining the angle. The
fully automated methods have difficulties determining the
radius accurately, with the all points with larger relative
errors of the radius being related to the very small radius
‘F’ (0.15mm). Semi-automated and manual measurements
show a slightly better performance with the very small radius.
It should be noted, that the large portion of unknown could
potentially alter this finding largely.
To quantify the influence of the measured radius and angle
on a possible fatigue assessment at the weld toe, the notch
stress intensity factor Kfwas calculated from the determined
radii and angles. The notch stress intensity factor Kfcan be
derived from the stress concentration factor (SCF) Ktand
the notch sensitivity η(1).
Kf=Kt
η(1)
For determining the stress concentration factor the for-
mula by Anthes for bending load on a cruciform joint (Eq. 2
with Table 4) was used [16]. The authors are aware, that
there are many different and probably more accurate formu-
Table 4 Parameters used for calculating Ktusing Eq. 2[16]
a0a1a2a3b1b2l1l2l3in °
0.181 1.207 −1.737 0.689 −0.156 0.2070 0.2919 0.3491 3.2830
Fig. 8 Notch stress intensity factors at the weld toe calculated using
the notch sensitivity coefficient approximation by Dietmann [30]and
the stress concentration factor approximation formula by Anthes [16]
for a cruciform joint under bending load
lae available, but, for the purpose of this analysis, the Anthes
formula was deemed appropriate (Table 4).
Ktt
ρ,α
=1+b1·t
ρb2·1+
3
i=0
ai·siniα·t
ρl1+l2·sin (α+l3)
(2)
The notch sensitivity was approximated using the formula by
Dietmann with the yield strength at 0.2 % plastic deformation
Rp,0.2being set to 355MPa (3)[30,31].
η=1+55
Rp,0.22
ρ(3)
The resulting notch stress intensity factors are shown in
Fig. 8and compared with the notch stress intensity factors
calculated using the machined radii and angles. For larger
radii the differences in the measured values only have a small
influence on the stress concentration. However, for smaller
radii the estimations are far off the notch stress intensity fac-
tor calculated by the machined dimensions. Especially for
the smallest radius ‘F’ (0.15mm), the notch stress intensity
factor is underestimated by around 30 % for the median of the
measured values with an interquartile range of 0.41, which
is 15% of the median value. The low outliers on radii ‘B’
(1.2mm), ‘H’ (1.9 mm) and ‘D’ (5.3 mm) can be explained
by a vast underestimation of the flank angle.
6 Conclusion
A round-robin study was set up to compare how different
researchers are characterising welded joints geometrically.
123
Welding in the World
A specimen with eight idealized weld toes was machined
and sent to 16 participants, which analyzed the specimen and
reported their measured angles and radii. The main findings
were:
•Most procedures achieved a decent accuracy with radii
larger than 1mm, while the errors on smaller radii are
usually larger and vary much more amongst the partici-
pants.
•The flank angle was measured accurately by most partic-
ipants.
•Measurements on radii larger than 1 mm have shown to be
adequate for estimating the notch stress intensity factor.
•For smaller radii the notch stress intensity factor is under-
estimated by around 30%.
•Participants with similar measurement systems achieved
similar results.
Originally, it was planned to quantify the influence of the
digitization technique and especially the resolution on the
results. Due to the large differences in the evaluation pro-
cedures used at the participants, this influence can currently
not be asserted. Therefore, a follow-up task was defined to
acquire more detailed information on the digitization and
evaluation methods that were used. The authors are planning
on starting a survey among the participants to get a detailed
description of how radius and angle are derived from the
digital cross-section data. Also, the authors have received
3D point clouds of the specimen from many participants,
from which information about resolution and accuracy can be
derived. The point clouds might also be used to test automated
evaluation procedures with data from different digitizers.
The study was set up with a second part (Part B), that
compares results from three actual welded specimens. The
evaluation of Part B is in process, publication is expected
for 2024. Further work is required in the automation of the
geometric characterisation of welded joints, only four par-
ticipants reported using an automated or a semi-automated
approach. The authors themselves are working on projects
to utilize artificial neural networks for characterising welded
joints.
Acknowledgements The authors want to express their gratefulness to
all participants of the round robin study: Zuheir Barsoum, Gustav Hult-
gren (KTH Stockholm, Sweden); Jonas Hensel (ifs, Braunschweig,
Germany); Jörg Baumgartner (Fraunhofer LBF, Darmstadt, Germany);
Martin Leitner (Montanuniversität Leoben, Austria); Heikki Remes
(Aalto University; Finland); Fabien Lefebvre (CETIM, Senlis, France);
Antti Ahola (LUT University, Lappeenranta, Finland); Daniel Löschner
(University of Applied Sciences Munich, Germany); Christian Dänekas
(Leibniz University Hannover, Germany); Hendrik Bissing (RUB,
Bochum, Germany); Seiichiro Tsutsumi (Osaka University, Japan);
Philippe Thibaux (OCAS, Zwijnaarde, Belgium); Andreas Pittner
(BAM, Berlin, Germany); Andreas Gericke (Fraunhofer IGP, Rostock,
Germany); Matthias Jung, Jan Schubnell (Fraunhofer IWM, Freiburg,
Germany); Finn Renken, Moritz Braun (TUHH, Hamburg, Germany
and German Aerospace Center (DLR), Institute of Maritime Energy
Systems, Geesthacht, Germany). The authors also thank the IIW Com-
mission XIII, led by Kenneth MacDonald, and C-XIII Working Group
4, led by Heikki Remes, for offering a platform for collaboration and
discussion.
Author Contributions All authors contributed to the study conception
and design. The Round-Robin study was mainly administered by Jan
Schubnell and Moritz Braun. The data was analysed by Matthias Jung.
Heikki Remes supervised the Round-Robin study and supported with
thorough discussions of the analysis. The first draft of the manuscript
was written by Matthias Jung and all authors commented on previous
versions of the manuscript. All authors read and approved the final
manuscript.
Funding Open Access funding enabled and organized by Projekt
DEAL. No funding was received to assist with the preparation of
this manuscript. The participation of Matthias Jung at the IIW Annual
Assembly and International Conference 2023 in Singapore was kindly
supported by the German Welding Society with a DVS-IIW Young Pro-
fessionals grant.
Declarations
Conflict of Interest Apart from the aforementioned the authors do not
declare any conflicts of interest.
Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing, adap-
tation, distribution and reproduction in any medium or format, as
long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons licence, and indi-
cate if changes were made. The images or other third party material
in this article are included in the article’s Creative Commons licence,
unless indicated otherwise in a credit line to the material. If material
is not included in the article’s Creative Commons licence and your
intended use is not permitted by statutory regulation or exceeds the
permitted use, you will need to obtain permission directly from the copy-
right holder. To view a copy of this licence, visit http://creativecomm
ons.org/licenses/by/4.0/.
References
1. Hobbacher A (2016) Recommendations for fatigue design of
welded joints and components: IIW Document IIW-2259-15, Ex
XIII-2460-13/XV-1440-13, 2nd edn. IIW document / Interna-
tional Institute of Welding, vol IIW-2259-15. Springer, Cham
and Heidelberg and New York. http://www.springer.com/de/book/
9783319237565
2. Schork B, Zerbst U, Kiyak Y, Kaffenberger M, Madia M, Oech-
sner M (2020) Effect of the parameters of weld toe geometry on
the FAT class as obtained by means of fracture mechanics-based
simulations. Welding in the World 64(6):925–936. https://doi.org/
10.1007/s40194-020-00874-7
3. Nguyen TN, Wahab MA (1995) A theoretical study of the
effect of weld geometry parameters on fatigue crack propagation
life. Eng Fract Mech 51(1):1–18. https://doi.org/10.1016/0013-
7944(94)00241-9
4. Lieurade HP, Huther I, Lefebvre F (2008) Effect of weld quality
and postweld improvement techniques on the fatigue resistance of
extra high strength steels. Welding in the World 52(7–8):106–115.
https://doi.org/10.1007/BF03266658
123
Welding in the World
5. Hultgren G, Barsoum Z (2020) Fatigue assessment in welded
joints based on geometrical variations measured by laser scanning.
Welding in the World 64(11):1825–1831. https://doi.org/10.1007/
s40194-020-00962-8
6. Schork B, Kucharczyk P, Madia M, Zerbst U, Hensel J, Bernhard J,
Tchuindjang D, Kaffenberger M, Oechsner M (2018) The effect of
the local and global weld geometry as well as material defects on
crack initiation and fatigue strength. Eng Fract Mech 198:103–122.
https://doi.org/10.1016/j.engfracmech.2017.07.001
7. Barsoum Z, Jonsson B (2011) Influence of weld quality on the
fatigue strength in seam welds. Eng Fail Anal 18(3):971–979.
https://doi.org/10.1016/j.engfailanal.2010.12.001
8. Renken F, Bock und Polach RUF, Schubnell J, Jung M, Oswald
M, Rother K, Ehlers S, Braun M (2021) An algorithm for statis-
tical evaluation of weld toe geometries using laser triangulation.
Int J Fatigue 149:106293. https://doi.org/10.1016/j.ijfatigue.2021.
106293
9. Holmstrand T, Mrdjanov N, Barsoum Z, Åstrand E (2014) Fatigue
life assessment of improved joints welded with alternative weld-
ing techniques. Eng Fail Anal 42:10–21. https://doi.org/10.1016/j.
engfailanal.2014.03.012
10. Braun M, Kahl A, Willems T, Seidel M, Fischer C, Ehlers S
(2021) Guidance for material selection based on static and dynamic
mechanical properties at sub-zero temperatures. J Offshore Mech
Arct Eng 143(4). https://doi.org/10.1115/1.4049252
11. Harati E, Karlsson L, Svensson L-E, Dalaei K (2015) The relative
effects of residual stresses and weld toe geometry on fatigue life
of weldments. Int J Fatigue 77:160–165. https://doi.org/10.1016/j.
ijfatigue.2015.03.023
12. Jonsson B, Samuelsson J, Marquis GB (2011) Development of
weld quality criteria based on fatigue performance. Welding in the
World 55(11–12):79–88. https://doi.org/10.1007/BF03321545
13. Liinalampi S, Remes H, Lehto P, Lillemäe I, Romanoff J, Porter D
(2016) Fatigue strength analysis of laser-hybrid welds in thin plate
considering weld geometry in microscale. Int J Fatigue 87:143–
152. https://doi.org/10.1016/j.ijfatigue.2016.01.019
14. Shiozaki T, Yamaguchi N, Tamai Y, Hiramoto J, Ogawa K (2018)
Effect of weld toe geometry on fatigue life of lap fillet welded
ultra-high strength steel joints. Int J Fatigue 116:409–420. https://
doi.org/10.1016/j.ijfatigue.2018.06.050
15. Ahola A, Muikku A, Braun M, Björk T (2021) Fatigue strength
assessment of ground fillet-welded joints using 4R method.
Int J Fatigue 142:105916. https://doi.org/10.1016/j.ijfatigue.2020.
105916
16. Anthes RJ, Köttgen VB, Seeger T (1993) Kerbformzahlen von
Stumpfstößen und Doppel-T-Stößen. Schweißen & Schneiden
45:685–688
17. Braun M, Neuhäusler J, Denk M, Renken F, Kellner L, Schubnell
J, Jung M, Rother K, Ehlers S (2022) Statistical characterization
of stress concentrations along butt joint weld seams using deep
neural networks. Appl Sci 12(12):6089. https://doi.org/10.3390/
app12126089
18. Rainer G (1978) Errechnen von Spannungen in Schweißverbindun-
gen mit der Methode der Finiten Elementen. Dissertation, Technis-
che Hochschule Darmstadt, Darmstadt
19. Oswald M, Mayr C, Rother K (2019) Determination of notch fac-
tors for welded cruciform joints based on numerical analysis and
metamodeling. Weldingin the World 63(5):1339–1354. https://doi.
org/10.1007/s40194-019-00751-y
20. Oswald M, Springl S, Rother K (2020) Determination of notch
factors for welded T-joints based on numerical analysis and meta-
modeling: IIW-Doc. XIII-2853-2020
21. Oswald M, Neuhäusler J, Rother K (2020) Determination of notch
factors for welded butt joints based on numerical analysis and meta-
modeling. Welding in the World 64(12):2053–2074. https://doi.
org/10.1007/s40194-020-00982-4
22. Schubnell J, Jung M, Le CH, Farajian M, Braun M, Ehlers S, Fricke
W, Garcia M, Nussbaumer A, Baumgartner J (2020) Influence of
the optical measurement technique and evaluation approach on
the determination of local weld geometry parameters for differ-
ent weld types. Welding in the World 64(2):301–316. https://doi.
org/10.1007/s40194-019-00830-0
23. Lee C-H, Chang K-H, Jang G-C, Lee C-Y (2009) Effect of weld
geometry on the fatigue life of non-load-carrying fillet welded
cruciform joints. Eng Fail Anal 16(3):849–855. https://doi.org/10.
1016/j.engfailanal.2008.07.004
24. Stenberg T, Lindgren E, Barsoum Z (2012) Development of an
algorithm for quality inspection of welded structures. Proc Inst
Mech Eng, Part B: J Eng Manufact 226(6):1033–1041. https://doi.
org/10.1177/0954405412439138
25. Lindgren E, Stenberg T (2011) Quality inspection and fatigue
assessment of welded structures. Master thesis, KTH Stock-
holm, Stockholm, Sweden. https://www.diva-portal.org/smash/
get/diva2:405988/FULLTEXT01.pdf
26. Hammersberg P, Olsson H (2010) Statistical evaluation of welding
quality in production. In: Pedersen MM, Mouritsen O, Hansen M,
Andersen JG (eds) Proceedings of the Swedish conference on light
weight optimized welded structures, Borlänge, Sweden. https://
publications.lib.chalmers.se/records/fulltext/local_146439.pdf
27. Jung M (2018) Entwicklung und Implementierung eines Algo-
rithmus zur Approximation und Bewertung von Kerbfaktoren an
Kehlnähten auf Basis berührungsloser 3D-Vermessung. Masterar-
beit, Karlsruher Institut für Technologie, Karlsruhe
28. Braun M (2021) Assessment of fatigue strength of welded steel
joints at sub-zero temperatures based on the micro-structural sup-
port effect hypothesis. PhD thesis, TUHH Universitätsbibliothek.
https://doi.org/10.15480/882.3782
29. Quinten M (2019) A practical guide to surface metrology. Springer
International Publishing, Cham. https://doi.org/10.1007/978-3-
030-29454-0
30. Dietmann H (1985) Zur Berechnung von Kerbwirkungszahlen.
Konstr 37(2):67–71
31. Radaj D, Vormwald M (2007) Ermüdungsfestigkeit: Grundlagen
Für Ingenieure, 3., neubearbeitete und erweiterte auflage edn.
Springer-Verlag Berlin Heidelberg, Berlin, Heidelberg. https://doi.
org/10.1007/978-3-540-71459-0
Publisher’s Note Springer Nature remains neutral with regard to juris-
dictional claims in published maps and institutional affiliations.
123
