Content uploaded by Johanna Müller
Author content
All content in this area was uploaded by Johanna Müller on Feb 07, 2022
Content may be subject to copyright.
Surface quality parameters for structural components manufactured by
DED-arc processes
Jonas Hensel
a,
⇑
, Anita Przyklenk
b
, Johanna Müller
c
, Markus Köhler
c
, Klaus Dilger
c
a
Chemnitz University of Technology, Chair of Welding Engineering, Reichenhainer Strasse 70, 09126 Chemnitz, Germany
b
Physikalisch-Technische Bundesanstalt (PTB), Bundesallee 100, 38116 Braunschweig, Germany
c
TU Braunschweig, Institute of Joining and Welding, Langer Kamp 8, 38106 Braunschweig, Germany
highlights
Interlayer boundaries and local
minima and maxima of the surface
profile do not generally coincide.
The waviness profile contains
representative features of the surface
topography.
Surface parameters describe the
waviness resulting from different
manufacturing parameters.
graphical abstract
article info
Article history:
Received 9 June 2021
Revised 14 January 2022
Accepted 29 January 2022
Available online 2 February 2022
Keywords:
Additive Manufacturing
Direct Energy Deposition
Metrology
Quality
Surface topography
Construction
abstract
Additive manufacturing of structural metallic components is an emerging technology. The DED-arc (di-
rect energy deposition) process combines efficiency with high degrees of freedom and predictable qual-
ity. Compared to powder-based processes, the surface is relatively coarse as a result of high heat input
and larger molten zones. In industrial applications, the post-treatment of surfaces of DED-arc compo-
nents cannot be guaranteed for economic reasons or due to geometric constraints. Hence, the surface
topography must be evaluated and quality levels need to be defined. Unfortunately, established surface
parameters and their determination cannot be directly transferred to DED-arc components and further
research on this topic is needed. This article demonstrates surface topography features of representative
DED-arc components. Furthermore, waviness profiles were established by applying an alternative cutoff
wavelength to better represent the DED-arc surface characteristics. Based on these waviness profiles, 2-
dimensional height parameters were calculated. The waviness parameters were determined for four
DED-arc specimens made from high strength wire electrodes. The manufacturing parameters were varied
in terms of different energy input and interpass temperature, resulting in varying deposition rates and
surface qualities. The waviness parameters correlated with manufacturing parameters. A mixed deter-
ministic and random nature of the surface topography was verified.
Ó2022 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Advanced architectural and structural design of steel compo-
nents leads to geometric complex and highly individualized struc-
tural members. In construction industry research, additive
manufacturing (AM) has recently been considered as a possible
https://doi.org/10.1016/j.matdes.2022.110438
0264-1275/Ó2022 Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
⇑
Corresponding author.
E-mail address: jonas.hensel@mb.tu-chemnitz.de (J. Hensel).
Materials & Design 215 (2022) 110438
Contents lists available at ScienceDirect
Materials & Design
journal homepage: www.elsevier.com/locate/matdes
alternative approach to conventional manufacturing for the pro-
duction of such steel components. Buchanan and Gardner summa-
rize the current state of the technology for AM of metal in
construction [1]. They discuss possibilities and challenges of AM
in the context of the construction industry, which must adhere
strictly to engineering standards. One example mentioned here is
the resulting wall thickness under consideration of surface wavi-
ness. Fig. 1 contains examples from construction industry showing
nodal connectors for space frames joining standardized tubes of
different shapes and diameters. These components are tailored to
individual global geometries of spaceframes and are designed to
sustain occurring forces and torque.
1.1. Metal AM in construction industry
Feucht et al. investigated interconnect elements and local rein-
forcements of structural members manufactured by Wire Arc Addi-
tive Manufacturing (WAAM) [2]. Laghi et al. experimentally
investigated AM components of stainless steel with a focus on
geometry and mechanical properties [3]. They also addressed the
effects of layer orientation on tensile properties and investigated
buckling. Furthermore, the desire for sustainable production pro-
cesses and structures yields a minimized use of material, promot-
ing topology optimized design according to individual load bearing
requirements. Reimann et al. presented their approach to topology
optimized design of structural nodes to optimize material usage
[4]. DebRoy et al. presented a comprehensive overview of available
technologies and also discussed the resulting microstructure,
material properties, and approaches for quality assessment [5].
Vafadar et al. provided a review of available AM technologies for
metals and related current standards to emphasize the need for
future standard developments [6]. Emerging AM technologies for
the production of structural steel components are direct energy
deposition (DED) methods using energy from laser (DED-LB), elec-
tron beam (DED-EB), or electric arcs (DED-arc). The processes are
abbreviated in accordance with ISO/ASTM 52900 [7]. The feedstock
material for applications in the construction industry is supplied
either as powder or as filament (wire). As summarized by Frazier,
the material available as powder that can be considered for the
cost sensitive construction industry is currently limited to stainless
steel (e.g. 316L, 1.4404) [8]. Wires are available in greater variety of
grades of unalloyed, low alloyed and stainless steels. The DED-arc
process uses conventional gas metal arc welding processes to melt
wire electrodes in the electric arc to generate components layer-
wise. The process can be highly automated to a degree where the
welding torch is positioned and moved automatically. The process
is easily scalable in size and combines process efficiency with a
high degree of geometric freedom at relatively high deposition
rates. Di Angelo et al. studied the effects of several process param-
eters on the component quality, including surface topography [9].
They made use of their analysis of published research articles to
derive optimal build directions during additive manufacturing
and provide important summaries of the effects of different man-
ufacturing parameters.
The surface quality of AM parts depends on the used processes
and their energy input. The deposition rate and hence the melt
pools of typical DED-arc processes are considerably larger com-
pared to DED-laser processes and laser powder bed fusion (LPBF),
which results in thicker layers and a rougher surface topography.
Gonzáles et al. identified the layer structure on the surface topog-
raphy based on height parameters [10]. Lehmann et al. derived the
waviness profile of a nearly net-shape DED-arc component from
micrographs by image analysis [11]. They also related the compo-
nent quality to process parameters. Furthermore, the deviation of
form and shape of DED-arc components are considered to be larger
compared to LPBF due to the high heat input, inhomogeneous ther-
mal shrinkage, and resulting residual stress and distortion. DED-
arc surfaces are hence often subject to post-processing by milling.
Lopes et al. investigated milling of high strength steels manufac-
tured by DED-arc and found a correlation of milling parameters
and surface roughness as well as the layer-dependent microstruc-
ture [12]. In addition to the aforementioned review articles, Prado-
Cerqueira at al. presented an evolutionary design of a robot-based
and cost-effective hybrid manufacturing system combining DED-
arc and milling which could be used for postprocessing of DED-
arc surfaces [13]. To reduce the effort for post-processing, in a large
number of applications it is desirable to limit the post-processing
to functional surfaces, as described by Webster at in the context
of hybrid manufacturing [14]. On the other hand, the surfaces
may be used as manufactured. Mechtcherine et al. investigated
the use of DED-arc rebars for use in concrete applications and
emphasized the importance of the surface topography for shear
transfer [15]. Müller et al. proposed an advanced testing method
that considers surface geometry, microstructure, and internal
defects for component approval [16]. Accordingly, knowledge of
the resulting surface properties is necessary for the design and
operation of the components.
1.2. Surface topography description of metal AM components
Surface topography of technical components is usually charac-
terized by applying the standards ISO 4287 and ISO 4288 [17,18].
The surface of LPBF components was analyzed by Cabanettes
et al. at all levels of the deviation of form [19]. They proved resolu-
tion issues of certain measurement methods such as coordinate
measurement machines and contact stylus approaches. By analyz-
Fig. 1. Examples of nodal connectors for space frames with topology optimized designs. Design and rendering by Christoph Müller, ITE, TU Braunschweig.
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
2
ing roughness profiles, they identified certain process-related
geometry features such as the hatch spacing. On another research,
Cheng at al. compared different measurement methods such as
confocal laser scanning, white light microscopy, and tactile stylus
measurements on LPBF specimens. They found good correlation
between tactile stylus and confocal laser measurements [20].On
the other hand, Leach et al. pointed out that metal AM components
often show high roughness and complex shape, which complicates
the quantitative determination of topography [21]. From their
work, it can be concluded that existing surface measurement tech-
niques and parameters have limited significance due to the mix-
ture of deterministic and random surface geometry of AM parts.
The method of measurement and its spatial resolution affect the
surface topography information included in the measured profile,
as demonstrated by Townsend at al. [22]. The measured profile
contains geometric shape deviations on different geometric levels
as depicted in Table 1. The error of form describes the macroscopic
deviation of the form, for instance the straightness deviation or the
concentricity. The waviness and the roughness describe periodic
shape deviations on smaller scales compared to the error of form.
The waviness describes deviations of shape on longer wavelengths,
while the roughness describes deviations of shape on shorter
wavelengths. The measured surface profile is corrected regarding
the error of form to derive the primary profile (P-profile). The P-
profile contains both waviness and roughness, which are separated
through filtering. A low-pass filter with a cutoff wavelength k
c
is
applied to determine the waviness profile (W-profile). The remain-
ing profile contains shorter wavelengths of the P-profile and is
defined as surface roughness (R-profile). The R-profile is the most
frequently analyzed profile to describe technical surfaces from
milling, turning, blast peening, and other mechanical surface treat-
ments. However, the definition of the cutoff wavelength is very sit-
uational and depends on the measurement method, the surfaces to
be analyzed, and its applications. Liu reported a considerably wide
range of cut-off wavelengths between 0.25 mm and 8 mm applied
for the determination of roughness parameters [23]. The standard
values for the sensing distance l
n
and for k
c
in ISO 4288 are chosen
in dependence of the aperiodic profile height. For coarse surfaces,
these values are l
n
= 40 mm and k
c
= 8 mm. Applying these values
to DED-arc surfaces with layer heights in the typical range of 1 ...
3 mm and waviness in the sub mm-range, it must be noted that the
waviness of the layer structure is not represented accordingly and
the scatter becomes large. Hence, it is conducive to analyze longer
measuring sections and alternatively define the cut-off wave-
length. This explains why LPBF surfaces, with generally small-
scaled surface topography features, may be described by roughness
parameters while the roughness is less meaningful for DED-arc
surfaces.
However, the determination of surface characteristics of AM
components is crucial to fulfil common quality standards. Further-
more, the component properties are affected by the surface shape,
especially in case of cyclic loading where geometric notches pro-
mote the generation of fatigue cracks, often at the surface. While
this is already well known from fusion welding, where weld
notches dominate the fatigue life of welded components (more
details on the importance of weld geometry for fatigue in [24]),
Zerbst et al. approach this issue also for AM components [25].
1.3. General surface characteristics of DED-arc components
Due to the layer by layer material deposition approach, AM
components exhibit characteristic surface properties. Depending
of the deposition strategy, different surface properties can be dis-
tinguished in the spatial directions in DED-arc processes. The sur-
face topography is influenced by the size of the melt pool and
acting process forces, namely the arc force, electromagnetic forces,
gravity, and forces from chemical reactions in the weld bead. Grav-
ity and the arc force are particularly significant for the surface
topography, which act differently on the weld bead depending on
the deposition position and the torch angle. Surfaces of DED-arc
components are provided in this section in order to illustrate typ-
ical surface features. The pictures shown in Figs. 2 to 4 were taken
during own investigations and are presented here for demonstra-
tion purposes only, without going into details of manufacturing
parameters.
As shown in Fig. 2 a), thin walled components exhibit a charac-
teristic surface waviness in building direction evoked by the layer
structure. This waviness is further superimposed by aperiodic
occurring error of form. In the welding direction, the surface
appearance is predominantly characterized by waviness in higher
magnitudes. In contrast, the volumes manufactured without addi-
tional contouring path exhibit significantly differing surfaces, since
the turning points of the deployed deposition pattern are reflected
in the surface structure (Fig. 2 b)). Thereby, a characteristic surface
waviness with the same order of magnitude is formed both in
building and welding direction. Case a) may be described by line
measurements in the building direction, while case b) should be
analyzed by areal measurements and areal surface parameters.
This article focusses on case a) and describes the surface features
on line profile measurements in building direction.
In addition to the deposition strategy, the material (e.g. melt
viscosity, surface tension, solidification behavior) and processing
parameters (e.g. energy input, travel speed, temperature distribu-
tion) can affect the surface formation. Fig. 3 provides an example
for different deposition qualities by means of thin walled alu-
minum samples in cross-sections. In case a) a parallel layer forma-
tion is obtained, while case b) shows a lateral shift of the layers,
resulting in a staggered formation. In order to illustrate the corre-
lation between the layer height and the resulting surface charac-
teristics, the W-profile was graphically derived from the cross-
sections and the profile height was visually magnified to illustrate
local peaks and valleys. Furthermore, the respective layer bound-
aries were marked along the structure. In large parts, minima
and maxima of the profile curve can be assigned to the layer height
Table 1
Deviations of form of technical surfaces (according to [26]).
Level Description Visualization
Error of form Irregularity of form, e.g. straightness deviation.
Describes the coarse shape
Waviness Periodic deviation of shape with intermediate wavelengths. Typically, the waviness
describes irregularities with horizontal spacings larger than the cutoff wavelength k
c
.
Describes the fine shape
Macroscopic roughness
Periodic deviation of shape with short wavelengths. Describes the fine shape
Microscopic roughness
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
3
a) or integer multiples thereof b). Deriving from these qualitative
observations of the DED-arc surfaces, it can be deduced that the
layer height, or integer multiples of the layer height, correlate with
characteristic features of the resulting surface appearance and thus
can be used as an additional input parameter for the determination
of the W-profile. The cross-sections also depict that local minima
may occur at layer boundaries, but also somewhere in the layer
itself. This is a consequence of overlapping weld metal and hence
Fig. 2. Surface appearance of Al-4046 DED-arc samples using different deposition strategies; (a) thin walled sample, (b) zig-zag pattern.
Fig. 3. Graphical determination of surface profiles based on cross-section macrographs of DED-arc aluminum; (a) parallel layer formation, (b) staggered layer formation.
Fig. 4. (a) cross-section macrograph of a steel shell structure, (b) surface with bulge.
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
4
interconnected with melt viscosity, gravity and work position as
well as energy input.
As depicted in Fig. 4 a), the surface profile of DED-arc specimens
made from steel does not necessarily reflect the single layers. One
example is marked where a layer interface is located at a local
maximum of the surface topography. Additionally, the height of
the individual layers may vary along the surface profile. The flow
behavior and the weld penetration vary stochastically. In some lay-
ers, the molten material builds bulges in which more material
remains during solidification, see Fig. 4 b). This creates distinct
waviness, especially at higher deposition rates. Reasons for bulge
building are the melt viscosity, the magnetic forces during welding
and irregularly solidifying weld pools.
The DED-arc process results in larger surface features, which
should be associated to the W-profile by definition (fine shape,
deviation of form with larger wavelength). Accordingly, such sur-
face parameters should be determined on the W-profile.
2. Determination of surface profile data on DED-arc specimens
This article focusses on the determination of surface parameters
based on the analysis of the waviness profiles instead of the rough-
ness profiles, which are most frequently used for casting and
machining processes. In contrast to ISO 4287 and ISO 4288
[17,18], the cut-off wavelength for the separation of roughness
from waviness is chosen in this article as a function of the manu-
facturing process, e.g., the layer height, rather than the expected
roughness parameters. Moreover, measuring lengths are consider-
ably longer to account for scatter on the relatively coarse surface
topography of DED-arc components. In addition, this approach
can be applied to surface data obtained by flexible and low-
priced measuring devices such as laser profile sensors with less
restrictions on the measuring range compared to most industry-
typical stylus instruments.
For demonstration purposes, specimens with varying surface
topography were manufactured. This was achieved by a variation
of the melt pool size resulting from varying energy input. The pen-
etration of the molten material was further influenced by a varia-
tion of the interpass temperature between welding two
subsequent layers.
2.1. DED-arc for specimen manufacturing
The experimental setup consisted of a GMAW (gas metal arc
welding) power source (type Fronius TPS500i), a Kuka 6-axis robot
for handling of the welding torch and a mobile fume exhaust (type
Kemper Filter-Master). The welding parameters were monitored
by means of a weld process data scanner (type HKS Weldscanner)
and the temperature of the specimens was measured using an
infrared pyrometer (type optris CTlaser3MH2). The temperatures
were measured on the top surface of the buildup material in the
middle of the last deposition layer and used to control interpass
temperatures between two subsequent layers. The data was fur-
ther processed to determine cooling times t
8/5
(time range during
cooling from 800 °C to 500 °C). The used shielding gas was 18 %
CO
2
and 82 % Argon.
The feedstock used for manufacturing was a high strength wire
electrode with a diameter of 1.2 mm. Table 2 contents the data
sheet information about chemical composition and mechanical
properties of resulting weld metal provided by the supplier. How-
ever, the mechanical properties are highly dependent on the DED-
arc process parameters and only provided as a general reference.
Furthermore, the composition of the deposited material was mea-
sured by means of spectroscopy and is also provided in Table 2.A
good match was found between nominal and measured chemical
contents. The resulting material properties were also determined,
but beyond the scope of this article and are thus not provided here.
However, it was found that the resulting material parameters are
sensitive to interpass temperature, energy input and buildup
direction.
Four specimens with dimensions of 350 mm 160 mm were
manufactured under variation of process parameters. The parame-
ters applied are typical for the DED-arc process of steel and
resulted in different surface qualities. One exemplary specimen is
shown in Fig. 5. 84 layers were deposited on a 15 mm thick sub-
strate plate made from S355N. The applied process parameters
are provided in Table 3. The parameters varied were the energy
input per layer and the interpass temperature. The chosen energy
input levels were 100 kJ and 170 kJ per layer, representing
nominal deposition rates of 136 cm
3
/hour and 407 cm
3
/hour
respectively. Interpass temperatures were varied at levels of
200 °C and 400 °C. The manufacturing parameters resulted in equal
layer heights of h = 1.9 mm, which were set constant. Fig. 5 con-
tains macroscopic metallographic cross-sections of the four speci-
mens. The increased energy input led to a stronger variation of
layer geometry as well as to stronger variations of the wall
thickness.
After manufacturing and before measurement of profile data, all
surfaces were carefully brushed in order to remove slag and sili-
cates without scratching the surface. Fig. 6 provides photographs
of the four specimens with a close-up magnification of the result-
ing surface topography.
A second set of specimens was manufactured using the same set
of parameters, with additional use of promoted cooling to shorten
the waiting time between subsequent layers. Furthermore, the
variation of cooling time was applied to study its effect on the sur-
face topography. A tubular nozzle with pressurized air was used for
cooling. The air flow was aligned so that the last deposited layers
were always cooled, and the flow velocity was 12 m/s. During
the deposition sequence, the cooling was turned off. The setup
and the resulting t
8/5
-times (time during cooling from 800 °Cto
500 °C) are given in Table 4. The t
8/5
-time as a measure of cooling
was chosen because it is commonly related to the mechanical
properties when welding low-alloyed carbon steels. The cooling
times were measured on the top of the last welded layer at mid
length.
2.2. Laser optical profile determination
For determination of surface profile data, a laser triangulation
sensor was used (type Micro-Epsilon optoNCDT1800) with a work
piece reference distance of 50 mm and a vertical measuring range
of 10 mm (50 mm ± 5 mm). The device uses a laser diode with a
laser wavelength of 670 nm (visible / red). The spot size ranges
between 45 mm at 0 mm (focal position) and 320 mmat±5mm
(begin and end of measuring range). The resolution in vertical z-
direction is 2 mm.
The specimen was held in place while the laser sensor was
moved horizontally over the surface. Except brushing, the surface
topography was determined in the as-manufactured condition.
The nominal distance between the specimen surface and the laser
sensor was 50 mm to ensure measurements near the focal position
of the laser beam. The profile data was measured along five lines
with lengths of 150 mm (Fig. 5). The measured profile was cor-
rected regarding its form and position by a linear fit which repre-
sents the nominal linear shape of the wall-shaped specimens. The
primary surface topography profile (P-profile) was determined by
subtracting the form and position fit from the measured profile.
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
5
Table 2
Nominal and measured properties of wire electrode Böhler Welding 3Dprint AM 80 HD according to the data sheet.
Data sheet Nominal chemical composition (wt.-%) Nominal mechanical properties of weld metal
after heat treatment
CSiMnCrMoNiR
p0.2
R
m
A
5
0.09 0.4 1.7 0.35 0.6 2 820 MPa 920 MPa 20 %
Measured Determined chemical composition (wt.-%)
CSiMnPS CrMoNiAlCuFe
0.11 0.36 1.7 0.01 0.003 0.38 0.6 2.18 0.01 0.055 bal.
Fig. 5. DED-arc specimen geometry with locations of surface profile measurements (left) and metallographic cross sections made from specimens with varying
manufacturing parameters (compare Table 3).
Table 3
Manufacturing parameters of DED-arc process chosen to vary resulting surface topography.
Specimen Welding travel speed (cm/min) Wire feed
(m/min)
Current
(A)
Voltage
(V)
Energy / layer (kJ) Interpass
temperature (°C)
W-25–2-200 25 2 82 ± 1.9 13.8 ± 0.2 100.8 200
W-25–2-400 25 2 81.4 ± 1.0 13.8 ± 0.3 100.1 400
W-45–6-200 45 6 198.3 ± 3.6 18 ± 0.1 176.8 200
W-45–6-400 45 6 188.8 ± 4.2 18 ± 0.2 168.0 400
Fig. 6. Photographs of DED-arc specimens manufactured with different welding parameters and interpass temperatures (a) to (d), compare Table 3 for manufacturing
parameters.
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
6
2.3. Resulting W-profiles of differently manufactured specimens
The DED-arc process, with its relatively large weld pools and
interactions of arc pressure, gravity and surface tension, results
in relatively large magnitudes in the profile height compared to
conventional surface machining or PBF processes. Furthermore,
periodic surface features, as known from milling and turning, are
not common in DED-arc surfaces. The definition of roughness (sur-
face topography with short wavelengths) does not apply to domi-
nant DED-arc surface features which are resulting from layers with
heights in the mm-rage. Additionally, and as described before, the
measurement method affects the resulting P-profile. Typical
roughness parameters in the mm-scale cannot be reliably deter-
mined by the applied laser scanning approach due to resolution
and noise issues. Hence, the surface topography should be ana-
lyzed dominantly regarding its waviness (surface topography with
longer wavelengths).
The P-profile of the surface consists of roughness (R-profile) and
waviness (W-profile), which are separated by Gaussian filtering of
the P-profile. The W-profile will be used to determine surface
parameters (Section 3), while the R-profile will be used to store a
combination of roughness and noise, which will not be analyzed
in the scope of this work. According to ISO 4288, the filter param-
eters for conventional roughness measurements are defined in
dependence of the aperiodic profile deviation, the profile length
and the individual measuring sections. Unfortunately, the parame-
ters used in standards such as ISO 4288 are not directly applicable
to DED-arc components because of the large magnitudes of surface
features. Comparably long measuring sections and single sections
lengths are required to describe the surface topography in a statis-
tically validated manner. Fig. 7 shows the interrelation of P-, R- and
W-profiles for a selected cut-off wavelength of k
c
= h/2 = 0.95 mm.
This cut-off wavelength was chosen after careful investigation of
its smoothening effect while still representing surface topography,
compare Fig. 8. Compared to the approaches defined in ISO 4288,
the W-profile contains local and global waviness, while roughness
and noise are disregarded and not further analyzed.
The cut-off filter length k
c
of the Gaussian filter was varied
between 2 h and h/4 to demonstrate its smoothening effect,
Fig. 8. The profiles are magnified between x = 70 mm and x = 90
mm. It was noted that values of k
c
h (one layer height) led to sig-
nificant trimming of the local minima and maxima. Smaller cut-off
filter lengths of k
c
(h/2) resulted in an accurate description of the
local p-profile peaks on the determined W-profiles. For the follow-
ing analysis, the cut-off filter length was set to k
c
= (h/2)
accordingly.
All surface topography data from differently manufactured
specimens was filtered and W-profiles were derived. Representa-
tive W-profiles of the four specimens are given in Fig. 9.
3. Surface quality parameters on W-profiles
In a first step, the arithmetic mean deviation W
a
and the root
mean square deviation W
q
of the W-profiles of all specimens were
analyzed according to ISO 4287 and Eqs. (1) and (2).
W
a
¼1
l
n
Z
l
n
0
ZðxÞ
jj
dx ð1Þ
Fig. 7. Separation of P-profile into W- and R-profile. Cut off wavelength kc = h/2 with layer height h from DED-arc process.
Table 4
Cooling times from 800 °C to 500 °C and experimental setup used for accelerated cooling.
t
8/5
-time of top layer (s) Experimental setup for accelerated cooling
Specimen Passive cooling Active cooling
W-25–2-200 18 17
W-25–2-400 29 25
W-45–6-200 25 19
W-45–6-400 40 25
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
7
W
q
¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1
l
n
Z
l
n
0
Z
2
xðÞdx
sð2Þ
Fig. 10 contains the data from specimens manufactured under
variation of process parameters and cooling strategy. The W-
profiles were determined by application of a cut-off wavelength
of k
c
= (h/2). The results show a general increase of the surface
parameters with increasing energy input. Mean values of W
a
of
specimens manufactured with low energy input range from
approximately 0.16 mm to 0.24 mm. With higher energy input, this
value increases to 0.22 mm to 0.33 mm. Similar observations were
made for W
q
. Active cooling slightly reduced the surface
parameters.
3.1. Segmentation of W-profiles
The W-profile was divided into individual measuring sections
by two segmentation methods. The first method divides the W-
profile into n single measuring sections with constant length l
s
,
Fig. 11. The second method applies a threshold crossing approach,
where single measuring sections l
c
are divided by locations where
the curve crosses the ordinate at Z = 0 mm. Zero-crossing locations
can be defined in different ways (e.g. from left to right, from right
to left, descending, or ascending). Here, the curve is defined by one
local minimum and one maximum from left to right, Fig. 12.In
both cases, the measuring length l
n
defines the maximum number
of single measuring sections that can be analyzed. The non-
Fig. 8. Effect of k
c
on resulting W-profile. A large cut-off wavelength causes inappropriate smoothing of the P-profile.
Fig. 9. Representative W-profiles of four DED-arc specimens manufactured by different parameters (compare Table 3).
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
8
Fig. 10. Arithmetical mean deviation W
a
and root mean square deviation W
q
of the W-profile.
Fig. 11. Curve segmentation through definition of single measuring sections with constant length.
Fig. 12. Curve segmentation through zero-crossing method.
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
9
periodic signal shape of W-profiles may lead to quite short or long
single measuring sections in the zero-crossing approach, which
automatically enlarges the scatter. Nonetheless, zero-crossing
was applied since it will become part of ISO 16610 and
ISO 21920 for the analysis of surface topography and is recom-
mended by Seewig et al. for analysis of technical surfaces [27].
Parameters can be defined for the entire W-profile (l
n
) as well as
for single measuring sections (l
s
,l
c
). The global parameter W
t
(maximum waviness) is determined as the height difference
between global maximum and global minimum. It can be expected
that this value increases with increasing measuring length l
n
due to
statistical phenomena. This parameter is independent from the
sectioning method. The local waviness W
z
and W
z,c
describe the
height difference between local maximum and minimum and are
determined on individual measuring sections l
s
and l
c
, respectively.
They vary along the W-profile and must be statistically analyzed.
3.2. Factors affecting the determination of waviness parameters
The waviness parameters were determined on the W-profiles of
specimen a) W-25–2-200 (profiles 1 to 5, measuring length
149 mm), compare Fig. 5 and Fig. 6, to demonstrate the effects of
the main evaluation parameters. The W-profiles were derived from
the P-profiles by Gaussian filtering with a cut-off length of
k
c
= (h/2) = 0.95 mm. The height parameters describing the surface
topography of the W-profile depend on three main influencing
parameters:
the length of the single measuring section l
s
,
the sectioning method to determine single measuring sections,
and
the number of single measuring sections to be analyzed.
The length of the single measuring section l
s
was defined in
dependence of the layer height h on three levels 1 h, 5 h and
10 h. Accordingly, the number of single measuring sections varied
between 78 (l
s
= 1 h), 15 (l
s
= 5 h) and 7 (l
s
= 10 h). Fig. 13 depicts
the effect of l
s
on W
z
.W
z
was determined on the individual W-
profiles 1 to 5, and additionally on all five W-profiles at once for
better comparison. Generally, W
z
increased with increasing length
of l
s
. The mean value
l
of W
z
increased from
l
0.20 mm (1 h) to
l
0.55 mm (5 h) and
l
0.75 mm (10 h). Another parameter is
the standard deviation
r
of the distribution of W
z
, which varied
moderately (1 h and 5 h) and significantly (10 h) between the ana-
lyzed profiles, most likely to be explained by the different sample
sizes. Long single measuring sections on W-profiles, and hence
only few W
z
data led to a higher scatter. Furthermore, the individ-
ual profiles 1 to 5 vary regarding their height profiles. Similar
trends regarding
r
were found on profiles 1 and 5 (
r
is large)
respectively profiles 2, 3 and 4 (
r
is small) when medium and long
single measuring sections (5 h and 10 h) were analyzed. This effect
was less significantly captured when analyzing short single mea-
suring sections (1 h).
The median of W
z
in most cases was slightly smaller compared
to the mean values. The maximum and minimum values also
increased with increasing l
s
.
Fig. 14 compares waviness parameters determined by the two
applied sectioning methods. Additionally, the global parameter
W
t
is given. Mean and median values of W
z
and W
z,c
were similar
for all profiles 1 to 5. The parameters determined on zero-crossing
single measuring sections l
c
were slightly smaller than those deter-
mined from l
s
. This can be explained by the large number of short
single measuring sections after the zero-crossing sectioning
approach, which resulted in smaller W
z,c
values (compare Fig. 12
and Fig. 13). The standard deviation
r
of W
z,c
was significantly lar-
ger than
r
of W
z,
which also resulted from the variation in single
measuring section lengths l
c,i
. The maximum waviness ranged
between 1.12 mm and 2.13 mm.
All parameters W
z
,W
z,c
and W
t
depend on the measuring length
l
n
. To improve the significance of these waviness parameters, a def-
inition for the measuring length is required. Furthermore, single
extreme features on individual single measuring sections may sig-
nificantly affect the waviness parameter. This should be avoided,
since the extreme values are represented by the W
t
parameter
and the standard deviation
r
of W
z
. For better comparison of dif-
ferent surface topography profiles, W
z
should be averaged over a
certain number of single measuring sections. The averaged wavi-
ness parameter W
z
is defined as the arithmetic mean value of
W
z,i
over n single measuring sections by
W
z
¼1
nX
n
i¼1
W
z;i
ð3Þ
Fig. 15 contents the averaged waviness parameter W
z
for 15
single measuring sections with a length of 5 h for n = 1, 3, 5 and
10. The influence of the single outlier of W
z
(e.g. section 6,
Fig. 13. Determination of W
z
in dependence of single measuring section length l
s
(h: layer height 1.9 mm; 1 to 5, all: 5 measuring sections on specimen W-25–2-200).
Parameters determined on constant total measuring length l
tot
= 149 mm resulting in different numbers of single measuring sections n.
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
10
W
z
= 1.59 mm) on W
z
(
l
±
r
) decreases as n increases. In addition
to the mean value, the standard deviation remains stable at n = 10.
In conclusion, the number of single measuring sections chosen to
determine averaged parameters should be chosen with regard to
the influence of possible outliers. Consequently, the total amount
of surface topography data should be increased by long measuring
length l
n
or multiple measurements at different locations.
3.3. Approach for comparable determination of waviness parameters
The studies on surface parameter determination (Sections 3.1
and 3.2) have demonstrated the effects of measuring and evalua-
tion parameters. In accordance with the results shown, and for bet-
ter comparison of profile height parameters between different
manufactures, the measured surface profile should preferably
contain
10 single measuring sections,
with a length of 5xlayer height (h),
and an in- and outlet length for the Gaussian filter (k
c
/2).
The W-profile can be reliably determined from the measured
profile through
the application of a form and position fit that accounts for the
nominally intended shape of the component, which results in
the P-profile,
a separation of the P-profile into R- and W-profiles by Gaussian
filtering with a cut-off wavelength of k
c
(h/2).
The W-profile should be divided into single measuring sections
with constant length. This is less sensitive regarding influences
from the form fit of the measured profile. The length of the single
measuring sections affects the individual W
z
parameter. A good
Fig. 14. Influence of sectioning method on waviness parameters and distribution of maximal waviness W
t
.W
z
determined with single measuring section length of
5 h = 9.5 mm. W
z,c
determined by zero crossing approach. W
t
represents the difference of the global minima and maxima on each measuring section. All parameters
determined on constant total measuring length l
tot
= 149 mm.
Fig. 15. Averaging of W
z
over different averaging lengths studying the effect of outliers on W
z
and its standard deviation.
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
11
resolution of W
z
was found when the length l
s
is defined according
to the layer height of the DED-arc process. A single measuring sec-
tion with a length of 5 layers reliably contents local topography
features of the waviness. When determining mean values of W
z
,
the W
z
parameter should be determined on at least 5, preferably
10 single measuring sections to account for outliers.
The determined parameters should be defined in accordance to
the discussed parameters, for instance as
W
t,50h
: Maximal height difference of the entire W-profile over a
length of 50 h (layer height),
W
z
;10;5h: Mean of height differences W
z
on single measuring
sections on the W-profile, determined on 10 subsequent single
measuring sections with lengths of l
s
= 5 h (layer height).
The parameter determination should be conducted on at least
three measured profiles to identify evident trends.
4. Resulting waviness parameters of DED-arc specimens made
from AM80
The waviness parameters were calculated for the eight speci-
mens manufactured with different DED-arc parameters and differ-
ent cooling conditions (Section 2.1). Fig. 16 contains resulting data
from all five laser scan profiles which were used to process the
waviness parameters on single measuring sections W
z
and the
total height of the W-profiles W
t
. A clear effect of the energy input
was proved for both parameters W
z
and W
t
. Low energy input led
to smaller waviness in terms of mean values at smaller standard
deviations. This was likely an effect of the smaller weld bead size
and the lower arc pressure at lower arc energy. An increased inter-
pass temperature also effected the waviness parameters positively.
The accelerated cooling (ac: active cooling) by pressurized air had
no substantial effect on the surface topography. The mean value of
the W
z
parameter was not altered significantly. However, the stan-
dard deviations for both parameters W
t
and W
z
increased for most
applied manufacturing parameters.
The mean values of surface parameters of the individual laser
scan profiles are given in Table 5. Based on the results, homoge-
nous surfaces can be realized at low energy input and high inter-
pass temperatures.
5. Discussion and outlook
The conducted research demonstrated the mixed deterministic
and random nature of surface topography characteristics of DED-
arc components. Layer boundaries do not necessarily correlate
with geometric peaks or valleys on the surface. But the distance
between peaks and valleys can be described to a good approxima-
tion by integers of the layer height. Comparably simple but more
flexible measuring devices such as laser profile sensors may be
used to determine P-profiles of variable length even on curved sur-
faces. The scatter of the Z-data resulting from lower resolution and
reflective surfaces can be larger compared to stylus instruments,
but this may be compensated in the data analysis. The separation
of roughness from waviness with a newly defined cut-off wave-
length as a function of the layer height provides representative
W-profiles for parameter determination and R- profiles containing
a mixture of roughness and noise from measurement. The surface
topography may be described based on height parameters on the
W-profile and the parameters discussed clearly reflect different
surface shapes. The definition of cut-off wavelength and the length
of single measuring sections based on the layer height is hence rec-
ommended. The standard deviation of W
z
values should ideally be
determined to evaluate the inhomogeneity of the surface waviness.
Next to the W-profile, the form fit and its influence on certain sur-
face parameters should be analyzed in greater detail. This informa-
tion could be used, for instance, to identify net sections of near net
shape shell structures. Overall, it was found that the surface
parameters should be determined at different locations of DED-
arc specimens to evaluate the susceptibility of surface features,
such as bulges. However, the total measuring length should not
be chosen based on current technical standards. Short measuring
sections, as recommended for milled surfaces, may be applicable
for investigation on surfaces of LPBF components, but are too short
for DED-arc components.
Furthermore, future work needs to be conducted to estimate
the influence of the measurement method and its resolution on
the quality of surface data. A comparison of optical and tactile
measurement methods should be performed. For this, also 3-
dimensional parameters should be considered. Additionally, the
component behavior under static and cyclic loading conditions
need to be investigated. It is likely that the surface waviness corre-
sponds to static load capacity as it affects the net sections. More-
over, it is widely known that geometric notches are important
Fig. 16. W
z
and W
t
of DED-arc specimens made from steel (AM80) with different manufacturing parameters and cooling conditions.
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
12
for fatigue of welded metals. Here, the interaction of internal irreg-
ularities, such as pores and lack of fusion, and the waviness needs
to be investigated, for instance as demonstrated for LPBF by [28].
With regard to fatigue, additional parameters can be derived
from the W-profile in the future. Liinalampi et al. used gradient
based surface features on thin laser welded components and found
correlations with the fatigue properties [29]. They calculated the
ratio of depth/width for laser weld undercuts. Fig. 17 displays gra-
dient based parameters such as the flank angle band the ratio of
(d
y,i
/d
x,i
) on a DED-arc W-profile. These parameters should be
investigated based on experimental and numerical approaches
and correlated to the fatigue strength.
6. Summary
The surface of DED-arc components contains random and deter-
ministic topography features. The layer height and integers thereof
occur deterministically, while bulges and overlapping weld metal
occur randomly. The surface topography can be described by the
W-profiles and height parameters can be derived. The surface
topography can be imaged to a good approximation by using a
laser scanning sensor. Measured and analyzed profiles from differ-
ently manufactured specimens differ significantly. The cut-off
wavelength, the measuring length, and the length of each measur-
ing section affect the surface parameters and should be provided in
addition to the calculated surface parameters. The waviness of
DED-arc components, expressed by height profiles, increases with
deposition rate. A high interpass temperature decreases the
waviness.
CRediT authorship contribution statement
Jonas Hensel: Conceptualization, Formal analysis, Funding
acquisition, Methodology, Project administration, Software, Super-
vision, Visualization, Writing – original draft. Anita Przyklenk:
Data curation, Formal analysis, Methodology, Software, Validation,
Writing – review & editing. Johanna Müller: Data curation, Inves-
tigation, Validation, Writing – review & editing. Markus Köhler:
Data curation, Investigation, Validation, Writing – review & editing.
Klaus Dilger: Resources, Funding acquisition, Supervision, Writing
– review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
Acknowledgements
The topology optimized nodal connectors presented in Fig. 1 were
designed and rendered by Christoph Müller (ITE, TU
Braunschweig).
Table 5
Surface parameters of DED-arc specimens made from AM80 steel.
Specimen: W-25–2-200 W-25–2-400 W-45–6-200 W-45–6-400
Cooling method: pc ac pc ac pc ac pc ac
W
z,10,5h
(Mean of height differences W
z
on single measuring sections)
0.78
0.58
0.52
0.54
0.62
0.61
0.41
0.47
0.45
0.49
0.60
0.50
0.59
0.49
0.48
0.80
0.82
0.71
0.59
0.76
1.16
0.90
0.91
1.17
1.23
1.27
0.94
1.08
1.10
0.69
0.60
0.70
0.57
0.72
0.74
1.50
1.19
1.51
0.93
0.62
W
t
,
50h
(Maximal height difference of the entire W-profile)
1.58
1.28
1.31
1.12
2.12
1.10
0.78
0.82
0.83
0.85
1.23
1.19
1.06
1.23
1.03
1.46
1.55
1.48
1.26
1.67
2.48
1.54
2.39
2.00
2.58
2.75
2.64
2.09
2.36
1.43
1.69
2.07
1.88
2.4
1.66
3.73
2.68
3.57
2.67
1.66
Fig. 17. W-profile with local minima and maxima to determine wave width d
x,i
, wave height d
y,i
and flank angles.
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
13
Funding acknowledgements
The research presented in this paper is being conducted within
the project ‘‘Wire Arc Additive Manufacturing (WAAM) of Complex
and Refined Steel Components (A07)”. The project is part of the col-
laborative research center ‘‘Additive Manufacturing in Construc-
tion - The Challenge of Large Scale”, funded by the Deutsche
Forschungsgemeinschaft (DFG, German Research Foundation) -
project number 414265976 - TRR 277.
Data availability
The raw data required to reproduce these findings are available
to download as supplementary files.
Appendix A. Supplementary material
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.matdes.2022.110438.
References
[1] C. Buchanan, L. Gardner, Metal 3D printing in construction: A review of
methods, research, applications, opportunities and chalenges, Eng. Struct. 180
(2019) 332–348.
[2] T. Feucht, J. Lange, M. Erven, 3-D-Printing with Steel: Additive Manufacturing
of Connection Elements and Beam Reinforcements, ce/papers 3 (3-4) (2019)
343–348.
[3] V. Laghi, M. Palermo, G. Gasparini, V.A. Girelli, T. Trombetti, Experimental
results for structural design of Wire-and-Arc Additive Manufactured stainless
steel members, J. Constr. Steel Res. 167 (2020) 105858, https://doi.org/
10.1016/j.jcsr.2019.105858.
[4] J. Reimann, P. Henckell, Y. Ali, S. Hammer, A. Rauch, J. Hildebrand, J.P.
Bergmann, Production of Topology-optimised Structural Nodes Using Arc-
based, Additive Manufacturing with GMAW Welding Process, J. Civ. Eng.
Constr. 10 (2021) 101–107. https://doi.org/10.32732/jcec.2021.10.2.101.
[5] T. DebRoy, H.L. Wei, J.S. Zuback, T. Mukherjee, J.W. Elmer, J.O. Milewski, A.M.
Beese, A. Wilson-Heid, A. De, W. Zhang, Additive manufacturing of metallic
components – Process, structure and properties, Prog. Mater Sci. 92 (2018)
112–224, https://doi.org/10.1016/j.pmatsci.2017.10.001.
[6] A. Vafadar, F. Guzzomi, A. Rassau, K. Hayward, Advances in Metal Additive
Manufacturing: A Review of Common Processes, Industrial Applications, and
Current Challenges, Applied Sciences 11 (2021) 1213, https://doi.org/10.3390/
app11031213.
[7] DIN EN ISO/ASTM 52900:2018-06, Additive Fertigung- Grundlagen-
Terminologie (ISO/ASTM DIS52900:2018); Deutsche und Englische Fassung
prENISO/ASTM 52900:2018, Beuth Verlag GmbH, Berlin.
[8] W.E. Frazier, Metal Additive Manufacturing: A Review, J. of Materi Eng and
Perform 23 (6) (2014) 1917–1928, https://doi.org/10.1007/s11665-014-0958-z.
[9] L. Di Angelo, P. Di Stefano, E. Guardiani, Search for the Optimal Build Direction
in Additive Manufacturing Technologies: A Review, JMMP 4 (2020) 71, https://
doi.org/10.3390/jmmp4030071.
[10] J. González, I. Rodríguez, J.-L. Prado-Cerqueira, J.L. Diéguez, A. Pereira, Additive
manufacturing with GMAW welding and CMT technology, Procedia Manuf. 13
(2017) 840–847, https://doi.org/10.1016/j.promfg.2017.09.189.
[11] T. Lehmann, A. Jain, Y. Jain, H. Stainer, T. Wolfe, H. Henein, A.J. Qureshi,
Concurrent geometry- and material-based process identification and
optimization for robotic CMT-based wire arc additive manufacturing, Mater.
Des. 194 (2020) 108841, https://doi.org/10.1016/j.matdes.2020.108841.
[12] J.G. Lopes, C.M. Machado, V.R. Duarte, T.A. Rodrigues, T.G. Santos, J.P. Oliveira,
Effect of milling parameters on HSLA steel parts produced by Wire and Arc
Additive Manufacturing (WAAM), J. Manuf. Processes 59 (2020) 739–749,
https://doi.org/10.1016/j.jmapro.2020.10.007.
[13] J.L. Prado-Cerqueira, J.L. Diéguez, A.M. Camacho, Preliminary development of a
Wire and Arc Additive Manufacturing system (WAAM), Procedia Manuf. 13
(2017) 895–902, https://doi.org/10.1016/j.promfg.2017.09.154.
[14] S. Webster, H. Lin, F.M. Carter III, K. Ehmann, J. Cao, Physical mechanisms in
hybrid additive manufacturing: A process design framework, J. Mater. Process.
Technol. 291 (2021), https://doi.org/10.1016/j.jmatprotec.2021.117048
117048.
[15] V. Mechtcherine, J. Grafe, V.N. Nerella, E. Spaniol, M. Hertel, U. Füssel, 3D-
printed steel reinforcement for digital concrete construction – Manufacture,
mechanical properties and bond behaviour, Constr. Build. Mater. 179 (2018)
125–137, https://doi.org/10.1016/j.conbuildmat.2018.05.202.
[16] J. Müller, M. Grabowski, C. Müller, J. Hensel, J. Unglaub, K. Thiele, H. Kloft, K.
Dilger, Design and Parameter Identification of Wire and Arc Additively
Manufactured (WAAM) Steel Bars for Use in Construction, Metals 9 (2019)
725, https://doi.org/10.3390/met9070725.
[17] Deutsches Institut für Normung, ISO 4287:2009: Geometrical Product
Specifications (GPS) - Surface texture: Profile method - Terms, definitions
and texture parameters, Beuth, Berlin.
[18] Deutsches Institut für Normung, ISO 4288:1996: Geometrical Product
Specifications (GPS) - Surface texture: Profile method - Rules and procedures
for the assesment of surface texture, Beuth, Berlin.
[19] F. Cabanettes, A. Joubert, G. Chardon, V. Dumas, J. Rech, C. Grosjean, Z.
Dimkovski, Topography of as built surfaces generated in metal additive
manufacturing: A multi scale analysis from form to roughness, Precis. Eng. 52
(2018) 249–265, https://doi.org/10.1016/j.precisioneng.2018.01.002.
[20] F. Cheng, S.W. Fu, Y.S. Leong, Research on optical measurement for
additive manufacturing surfaces, in: International Conference on Optical
and Photonics Engineering (icOPEN 2016), China, SPIE, Chengdu, 2017,
p. 102501F.
[21] R.K. Leach, D. Bourell, S. Carmignato, A. Donmez, N. Senin, W. Dewulf,
Geometrical metrology for metal additive manufacturing, CIRP Ann. 68 (2)
(2019) 677–700, https://doi.org/10.1016/j.cirp.2019.05.004.
[22] A. Townsend, N. Senin, L. Blunt, R.K. Leach, J.S. Taylor, Surface texture
metrology for metal additive manufacturing: a review, Precis. Eng. 46 (2016)
34–47, https://doi.org/10.1016/j.precisioneng.2016.06.001.
[23] W. LIU, Optical and XCT Measurement of Additively Manufactured Surfaces,
Fields: journal of Huddersfield student research 7 (2021). https://doi.org/
10.5920/fields.803.
[24] B. Schork, P. Kucharczyk, M. Madia, U. Zerbst, J. Hensel, J. Bernhard, D.
Tchuindjang, M. Kaffenberger, M. Oechsner, The effect of the local and global
weld geometry as well as material defects on crack initiation and fatigue
strength, Eng. Fract. Mech. 198 (2018) 103–122, https://doi.org/10.1016/j.
engfracmech.2017.07.001.
[25] U. Zerbst, M. Madia, G. Bruno, K. Hilgenberg, Towards a Methodology for
Component Design of Metallic AM Parts Subjected to Cyclic Loading, Metals 11
(2021) 709, https://doi.org/10.3390/met11050709.
[26] C.P. Keferstein, M. Marxer, C. Bach (Eds.), Fertigungsmesstechnik, Springer
Fachmedien Wiesbaden, Wiesbaden, 2018.
[27] J. Seewig, P.J. Scott, M. Eifler, B. Barwick, D. Hüser, Crossing-The-Line
Segmentation as a Basis for Rsm and Rc Evaluation, Surf. Topogr.: Metrol.
Prop. 8 (2) (2020) 024010, https://doi.org/10.1088/2051-672X/ab958c.
[28] W. Schneller, M. Leitner, S. Pomberger, F. Grün, S. Leuders, T. Pfeifer, O.
Jantschner, Fatigue strength assessment of additively manufactured metallic
structures considering bulk and surface layer characteristics, Addit. Manuf. 40
(2021) 101930, https://doi.org/10.1016/j.addma.2021.101930.
[29] S. Liinalampi, H. Remes, J. Romanoff, Influence of three-dimensional weld
undercut geometry on fatigue-effective stress, Weld, World 63 (2) (2019) 277–
291, https://doi.org/10.1007/s40194-018-0658-7.
J. Hensel, A. Przyklenk, J. Müller et al. Materials & Design 215 (2022) 110438
14