Poncelet Et Al. (2021) - Critical Assessment of The Impact of Process Parameters On Vertical Roughness of AlSi10Mg

Journal Pre-proof
Critical assessment of the impact of process

parameters on vertical roughness and hardness of
thin walls of AlSi10Mg processed by laser powder
bed fusion
O. Poncelet, M. Marteleur, C. van der Rest, O.

Rigo, J. Adrien, S. Dancette, P.J. Jacques, A.
Simar
PII: S2214-8604(20)31173-8
DOI: https://doi.org/10.1016/j.addma.2020.101801
Reference: ADDMA101801
To appear in: Additive Manufacturing

Received date: 6 July 2020
Revised date: 4 December 2020
Accepted date: 17 December 2020
Please cite this article as: O. Poncelet, M. Marteleur, C. van der Rest, O. Rigo, J.
Adrien, S. Dancette, P.J. Jacques and A. Simar, Critical assessment of the impact
of process parameters on vertical roughness and hardness of thin walls of
AlSi10Mg processed by laser powder bed fusion, Additive Manufacturing,
(2020) doi:https://doi.org/10.1016/j.addma.2020.101801
This is a PDF file of an article that has undergone enhancements after acceptance,
such as the addition of a cover page and metadata, and formatting for readability,
but it is not yet the definitive version of record. This version will undergo
additional copyediting, typesetting and review before it is published in its final
form, but we are providing this version to give early visibility of the article.
Please note that, during the production process, errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier.
Critical assessment of the impact of process
parameters on vertical roughness and hardness of
thin walls of AlSi10Mg processed by laser powder
bed fusion.
O. Poncelet1 , M. Marteleur1 , C. van der Rest1 , O. Rigo2 , J.

Adrien3 , S. Dancette3 , P. J. Jacques1 , A. Simar1
of
1 UCLouvain, Institute of Mechanics, Materials, and Civil Engineering (IMMC),
IMAP, Place Sainte Barbe, 2, 1348 Louvain-la-Neuve, Belgium.
2 Sirris, T-ADD Team, 4102 Seraing, Belgium.
3 University of Lyon, INSA Lyon, CNRS UMR5510, Laboratoire MATEIS,
ro
F-69266 Villeurbanne, France.
E-mail: olivier.poncelet@uclouvain.be
-p
Abstract. This work deals with the impact of prominent process parameters in
laser powder bed fusion (L-PBF) technique, also called selective laser melting
(SLM), on relevant properties of AlSi10Mg such as surface roughness and
re
hardness. More particularly, the influence of scanning strategies, track energy
density (TED), and of the offset between the contours and the bulk hatching
were scrutinized. Basic thin structures such as single wall tracks (SWT) and
ultra thin walls (UTW) were used to isolate the effect of the TED and of the
lP
offset between parallel melt pool tracks on the vertical roughness, respectively. It
is shown that the TED reduces the level of roughness mostly due to a geometrical
effect. The optimum range of TED is shown to be at the beginning of the keyhole
conditions, at the expense of the amount of sub-surface porosities and of hardness.
Increasing the offset also allows reducing the roughness. Thicker structures with
contours and bulk hatching were also used and brought the same conclusions. In
a
this work, low levels of roughness (about 4 µm) ) have been obtained on vertical
walls without any post-treatment but at the expense of the mechanical properties.
rn
Keywords: L-PBF, SLM, AlSi10Mg, roughness, hardness, process parameters,

contours, bulk hatching, sub-surface porosities.
u
Nomenclature: TW: thin walls; SWT: single wall tracks; UTW: utra thin walls;
CF: contour first; BF: bulk first.
Jo
1. Introduction
Laser powder bed fusion (L-PBF, ASTM Standard F2792 (1)), is more and more used
for the processing of metallic components. Owing to their good strength/weight ra-
tio combined with a high thermal conductivity, aluminum alloys, including AlSi10Mg
alloy, are well adapted materials for applications such as heat exchangers processed
by L-PBF. However, defects associated to this specific combination of materials and
process, such as high surface roughness and presence of porosities still limit the ap-
plications where mechanical performances are restrictive. Indeed, surface roughness
2
of parts made of AlSi10Mg alloy by L-PBF is of large concern in aerospace and au-
tomotive industries since it directly impacts the fatigue life (2; 3; 4; 5) and corrosion
properties (6; 7; 8). Despite this large importance of surface roughness, hardly any
studies investigated in details its origin and its evolution with process parameters.
Furthermore, published studies investigated the influence of process parameters on
the top surface roughness (upskin) (9; 10; 11; 12; 13; 14; 15; 16) while side (vertical)
roughness has been studied to a lesser extent (17; 18; 15; 16). While inclined surfaces
lead to a large roughness due to a stair stepping effect related to the intrinsic stratifi-
cation process (19; 20; 21), roughness on vertical surfaces is not yet fully scrutinized.
If it is clear that defects such as spatters, satellites or balling increase the vertical
roughness (17; 18), the relationship between process parameters, scanning strategies
and roughness is not yet clearly established, even though there impact has been ob-
served (22; 23; 15; 16). Moreover, while previous studies showed that increasing the
volumetric energy density (VED) seems to reduce the side roughness (17; 12; 18) below
of
a Ra of 5 µm, the impact of this increase of energy on porosity or on the mechanical
properties is not yet studied. In order to investigate the factors influencing the ver-
tical roughness, thin-wall samples and other structures were processed using a wide
ro
range of parameters. The level of sub-surface porosity as well as the hardness were
simultaneously investigated. This work provides a systematic study to assess the ori-
gin of vertical roughness of L-PBF products and gives guideline to obtain a desirable
net-shaped structure.
-p
In L-PBF, any processed object is built by a succession of parallel and stacked
re
unit cells, called laser tracks, melt pool tracks or single tracks. It is well established
now that a part is also divided into two different areas during the slicing (the strati-
fication process): the bulk hatching (core or in-fill plane) and the contours (the hull,
lP
borders or outer shell) which are the object boundaries (24). Both areas are processed
following different parameters (mainly, laser power and scanning speed). Parameters
for the bulk hatching are logically selected to obtain a porosity-free material while pa-
rameters for the contours are chosen following surface roughness considerations. In the
a
case of L-PBF, roughness can thus be considered as a topographical measurement of

the melt pools constituting the contours. It is also well established now that the level
rn
of bulk porosities is directly linked to the bulk melt pool track stability (25; 26; 27)
and thus related to the process parameters (28; 29; 30; 31). A process window in which
the melt pool tracks are stable can then be established. However, no such conclusion
u
has been drawn yet for the contour process parameters.

Jo
The first part of this work describes a methodology systematically assessing the
processing parameters to reach a low level of vertical roughness. The most basic struc-
ture, i.e. a stack of single tracks, called single wall track (SWT) in this work, is used
as test sample. This structure, mostly used to investigate the L-PBF process behavior
(32; 33), allows highlighting the impact on the roughness of two important process
parameters, i.e. the laser power and the scanning speed. Those parameters can then
be analyzed independently of the building strategies (contours, bulk, ...). Hardness of
SWT was also measured to investigate the evolution of the mechanical properties with
the laser power and the scanning speed, as the microstructure of such alloy quickly
evolves with heat and thus with the process parameters (34). L-PBF AlSi10Mg has a
hierarchical microstructure where large grains present a fine cellular-dendritic solidifi-
cation structure with submicron-sized primary Al-cells and an intercellular Si network.
3
Moreover, this heterogenous microstructure follows the shape of the melt-pools and
is divided in 3 distinct microstructures: the fine zone, the coarse zone and the heat
affected zone (29; 35). This microstructure directly affects the hardness properties of
the alloy and is very sensitive to its thermal environment and to thermal treatments
(36). Indeed, within the melt-pools themselves, hardness gradient have been observed
(from 150 HV in the fine zone to 110 HV in the coarse zone) due to different cooling
rates within the pool (37; 38). Depending on the process parameters and the building
temperature, hardness of as-built L-PBF AlSi10Mg can range from 90 HV (39) to 125
HV (40; 41; 36; 42) or above 130 HV (43; 34).
While only laser power and speed are investigated in the case of SWT, contiguity
of melt pool tracks, e.g. in the case of contours and bulk hatching, can cause a change
of the solidification morphology (44) (due to wetting conditions between tracks), im-
pacting directly the roughness. Accurate modeling of the L-PBF process predicts that
of
gravity, surface tension and subsequent wetting or capillarity forces are responsible for
melt pool stabilities and morphology (45; 46; 47; 48; 49; 50). The second part of this
work deals with the interaction of adjacent melt-pools by using a structure called ultra
ro
thin-walls (UTW), composed of two parallel SWT.
Finally, the third part of this work investigates thicker structures. 600 µm-thick
-p
structures (thin-walls, TW) with a bulk hatching area were processed to confirm the
predictions made from the analysis of the SWT and UTW structures. New process
parameters have to be considered as a result of the bulk hatching area, the scanning
re
strategy and the offset. Indeed, there are two options to process the structures; either
the bulk hatching is processed before the contours (bulk first strategy (BF)) or the
contours are processed first (contours first strategy (CF)). Up to now, reported studies
lP
on roughness have been mostly performed with a BF strategy (13; 18; 51; 52) while
the CF strategy has been less studied (9). In this work, 600 µm-thick structures were
selected because they are representative of the wall thickness in applications such as
heat exchangers, in which the level of roughness is important to keep control of the
a
heat transfer, pressure drop and eventually fatigue resistance.

rn
As a conclusion, the optimized set of parameters was tested on thin-walls and on

bulk cubic structures with a first insight in the mechanical behavior owing to hardness
measurements.
u
Jo
2. Materials and Experimental Procedure
2.1. Manufacturing by L-PBF

®
L-PBF samples of AlSi10Mg alloy were processed with a ProX DMP 200 (3D Sys-
tems Inc.) machine with a laser operating at 1075 nm wavelength, a spot size of 70
µm and 273.6 W maximum power. AlSi10Mg gas atomized spherical powder with an
average diameter of 15-53 µm was provided by TLS Technik. The layer thickness was
fixed at 30 µm for all samples. The build-plate was not heated during the L-PBF
process and is initially at room temperature. The manufacturing was performed in
an argon gas atmosphere, with a residual level of O2 concentration smaller than 500
ppm. Samples were processed without support parts, directly on the build-plate. No
4
post-treatments were applied.
The process window has been identified prior to the present work by studying the
stability of single tracks as proposed in previous works (26; 53). Track energy density
(TED) or linear energy density (LED, defined as the power divided by the scanning
speed) between 120 J/m and 200 J/m have been established as the boundaries of
process window, in good agreement with previous work (26) for AlSi10Mg alloy. A
TED level of 194 J/m (232.5 W at 1200 mm/s) for the bulk hatching was selected in
this work as the standard parameters.
Figure 1 illustrates the three main geometries investigated in the present work.
Single wall track samples (SWT), 10 mm in height and width, were processed by
stacking single tracks (Figure 1.a). The impact of power and scanning speed on the
arithmetic roughness (Ra ) and on the hardness was considered. Roughness was mea-
of
sured on both SWT faces. As no Ra differences between the faces were noticed, results
will be presented as a mean of both faces. It is worth noting that the laser starting
point did not change from one layer to another.
ro
A second set of specimens, called ultra-thin wall samples (UTW) consists in two
parallel tracks or SWT (Figure 1.b), 10 mm in height and width. The impact of the
-p
hatch-space between the two tracks on the roughness was also considered at two differ-
ent TED levels. The face constituted by the first contour is called ”face 1” while the
face constituted by the second contour is called ”face 2”. Ra was measured on each
re
face. The laser starting point did not change between each layer for most samples.
However, one sample was processed with a moving starting point, following the diago-
nal (y-z plane) of the sample (allowing to process the sample presented in Figure 15).
lP
Hardness measurement were not performed on those samples as the microstructure

does not change with the chosen offsets.
Thin-wall samples (TW) are 600 µm-thick walls, 10 mm in height and width with
a
a bulk hatching and contours (Figures 1.c and d). The bulk hatching and the contours
were processed with their own set of parameters (laser power and scanning speed).
rn
Unless mentioned, bulk hatching was processed at 232.5 W (85 % of Pmax ) and at
a scan speed of 1200 mm/s, a hatch space of 100 µm and a scanning rotation of 90
between each layer, as depicted on Figure 1.c and d. Bulk hatching and contours were
u
processed following two distinct strategies. The laser starting point of the contours
did not change from one layer to another. The laser starting point of the bulk changed
Jo
between each layer, as indicates on Figure 1.c and d (between layer ”n” and ”n+1”).
In the first strategy (Figure 1.c), contours are processed first (CF strategy). The
second strategy is the opposite (Figure 1.d), i.e. the bulk hatching is firstly processed
(BF strategy). When one strategy is selected, it is applied to each layer of the same
sample. Within these two strategies, several parameters were varied for the contours:
laser power, scanning speed and the offset. The offset is defined here as the dis-
tance between contours and the bulk hatching (Figures 1.c and 1.d). The contour is
processed asymmetrically as the starting point is always on the same thin-wall face
(depicted on Figure 1.d). The face constituted with the contour processed first is
called ”face 1” and the face constituted with the second contour is called ”face 2”.
Roughness measurements were thus performed on each face individually.
3
3 3 3
3 2.8 3 3
First (CF) First (BF) 5
2
3
2.8 2.8 2.8 3
2.8 First (CF) Thinïwalls

First First
(BF) (CF) (TW)
First Ultraïthin
(BF) wall (UTW) 2.8
3 First
3 (CF) First (BF) 3 2.8
2.6 3 3 First (CF) First (BF) 2.8
First
Thinïwalls (CF) Single
First
(TW) Wall Tracks
(BF)
Thinïwalls
Ultraïthin (TW)wall(SWT)
Ultraïthin
(UTW) wall (UTW) FirstFirst
(CF) Thinïwalls
First
First(BF)(BF) (TW) Ultraïthin wallwall (UTW)
1.8
2.8
(CF) Thinïwalls (TW) Ultraïthin (UTW)
2.6 2.6
3 3 First (CF) First (BF) 2.6 2.8
3 2.6
SingleThinïwalls Contour
Wall Traks (TW)
Single
(SWT)Meltïpool
Ultraïthin
Wall Tracks Bulk
wall
(SWT)x y z
(UTW) 2.8
First
3
2.8(CF) First Thinïwalls
(BF) Thinïwalls
2.8
Single
2.8 (TW)
(TW)
Single Wall Tracks
Ultraïthin
Ultraïthin
Wall Trackswall (SWT)
wall
(SWT)(UTW)(UTW) 3
1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
2.6 Thinïwalls2.6 (TW) Ultraïthin wall (UTW) 2.6
2.4
Contour Bulk Face 1 Bulk hatching
Contour Bulk
(CF) Face 1 Bulk y hatching
3
Powder denudation First (CF)
Single First
Wall (BF) First
Tracks
Contour (SWT)
Meltïpool FirstBulk
(BF) x x zy z 2.6
0
Single Wall Tracks (SWT) Thinïwalls (TW) Ultraïthin
Single wall
Contour (UTW) Meltïpool Bulk
Contour Meltïpool Contour Bulk x y z
Meltïpool Bulk x y z Single Wall Tracks (SWT) 2.6 Wall Tracks (SWT)
2.8
2
2.8 2.6 2.8 2.6 2.8 2.4
1.6
Single Wall Tracks Contour
Thinïwalls Meltïpool
(TW) Powder Bulk xwall
Ultraïthin
Thinïwalls
denudation y z(UTW)
(TW)
2.8
Ultraïthin wall (UTW)
Bulk(SWT)
2.4 2.4 2.4
PowderContour Crossïsection
First (CF)
Meltïpool
denudationPowder Bulk Laser
First
x
denudation (BF)
First
y z path (top(BF)
(CF) First view) FirstContour
(CF) Meltïpool
First (BF) Contour x y 2.6 zMeltïpool
Powder
First (CF) denudation
First
Bulk (BF) x y z
Thinïwalls (TW) Ultraïthin wall (UTW)
2.6 2.4
CF) First (BF)
2.4
2.6
Thinïwalls Indent
(TW) Ultraïthin
Thinïwalls wall (UTW)
(TW) Ultraïthin Contour
wall
Powder
Thinïwalls (UTW) Meltïpool
2.4denudation
(TW) Ultraïthin
Powder
SingleBulk Wall
wall
denudation
x TrackszCrossïsection
y(UTW)
Thinïwalls(SWT)
Single (TW) WallLaser path(SWT)
Tracks
Ultraïthin (top
wall view)
(UTW)
2.4
Face 2 First contour Offset Hatching
2.2
Laser pathpath
(top view)
0.2
2.4
Crossïsection
Powder Laser
denudation
Crossïsection Wall(top view)
Laser pathTracks(top (SWT)
view) Powder 2.4
Crossïsection
denudation pathLaser path (top view)
1.8
2.6 2.4
2.6
Crossïsection LaserWall Laser
pathTracks (top view) 2.6 2.2
x(top view) 2.6
Powder denudation
Crossïsection Laser path (top view)
Single Tracks (SWT)

Single Wall
1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
walls (TW) Ultraïthin 2.2 wallFace (UTW) Single Wall Tracks
Crossïsection Contour
(SWT)
Laser path (top Meltïpool
Single
view) Bulk
Contour yMeltïpool
z (SWT) Bulk x y z
1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
2.4
1
200 µm 300
Layer n Layer
Second
Face 2 First
Laser path
Crossïsection Laser path (top2.4 view) Powder
Powder denudation
Contour Meltïpool
Single Wall
Thinïwalls
First (CF)First
2.2
2 First contour Offset Hatching 2.2
Laser path (top view) 2.2
1.4
z Laser path (top view)
3
Laser path
Crossïsection (top view)
Laser Laser path
(top (top
pathMeltïpool view) view) Crossïsection Face Laser
2 First path
contour (top view)
Offset Hatching
0
Contour Bulk x y z Crossïsection Laser path (top view)
0
Contour Meltïpool 3
BulkContour
x y z path
Laser
2.2 2.4
Meltïpool Bulk
(top view)
Powder x 2.2 2.4
y denudationContour Meltïpool Bulk x y z 2.2
Powder denudation
Wall Tracks 2.2 (SWT)
Second 2contour Single
2.4
Powdertrack FaceLaser
2.4
path (topLaser
view) Face 2 Powder First
2contour
Second contourOffset Hatching
Single track 2.2 2.4
1.6
2.4 2
Face 2 2.2
2 First
Laser path (topPowder
contour
Face view) denudation
Offset
First Hatching
contour denudation
Offset HatchingPowder 22.2
denudation Offsetpath Face (top First
view) contour
denudation Offset Hatching
0.4
First contourCrossïsection
3 Hatching Laser path (top view)
First2 (CF) First (BF) Crossïsection Laser path (top
Crossïsection view) path
2.8 Laser 2 (topFaceview) 2 First contour Second
Offset contour
Layer
Hatching Crossïsection
nSingle
Layer track
n+1 Laser(toppath (top view) 2
Contour Meltïpool Bulk x y z
Layer n Layer n+1 Crossïsection Laser 2 path (top view)Crossïsection Laser path view)
ur Meltïpool Bulk xFace
Second 2 y contour
z2 First Single track Second contour Single track2
Second contour Single track 2
0.2
2 contour
Second Offset
contour Hatching
Single track Face 2 First contour Offset Hatching
0.2
2.2 2.2
Laser path
Layer (top
n view)
Layer Laser
n+1 path (top view)
1.2
2
2.2
Laser path (topLaserview) path (top view) Laser First Second
2.2
(CF) contour
First (BF) Single track 200 µm 300 µm 600 µm
Second contour Single track
2.2 2 2.2 1.8

Thinïwalls (TW) Ultraïthin wall (UTW) path (top view) Laser path (top view)
1.4
3 Layer3 n 2 Layer n+1
contour
200 µm 300 µm 600 µm
Layer
Face1.4 2 First contour Offset Hatching
Laser path (top view)
Crossïsection
Powder denudation
Contour Meltïpool
Single
Thinïwalls (TW)
First2.2(CF) First
er denudation Layer n Layer
Second 200 n+1
µm
contour 300 µm
Single 600
track µm Layer n2FaceLayer n+1
Layer n 2 Layer n+1 Second
2.8
contour Single track
200µm
Second Single
Facecontour
Laser
Crossïsection
Powder denudation
Contour Meltïpool
Single
Thinïwalls
2 1.8
200 µm 300 µm 600 µm
Layer nn+1Layer n+1

Face First contour 2Offset Hatching 200 Layer n(TW)
Layer Face
n+1 21.8First contour Offset Hatching
2contour
First contour Offset Hatching
0.6
Face First (CF)
First contourFirstOffset
Face
1.8
(BF) 2Hatching
Thinïwalls First contour Ultraïthin
Offset wall
HatchingFace(UTW) First Offset Hatching 1.8
First(CF)
µm 300 µm 600 µm 1.8
Single Wall Tracks (SWT)
Single Wall200Tracks1.8
(SWT) 2.6
(top path
(TW) Ultraïthin
µm1.8 300
Layer n3 view)
µm 600
Layer Second
n+1µm300contour Single
µm Second track 3 Single
contour track Layer ncontour
2 Layer
200 n+1
300 600 µm
1.8
2
200 600 (TW)Second 200 contour 300Single 600
track Second
µm contour µm Single track
Traks
0.4
2
1.8
µm µmThinïwalls Single Wall
Ultraïthin µm
2 Tracks
wall µmSecond
(UTW)
1.8 (SWT) µm Single
Second track contour Single track 2 1.8 2 1.6
ïsection Laser path (top
1.2
1.6
1.8
2.4
2.6
2.8
0.4
2.8
1.2
200 µm 300 µm 600 µm
1
µm600
2.6
x Layer zn 300
LayerLayer
n+1 n Single
0
LayerWall n+1Tracks 1.6 1.6

zLayern+1nLayer Layer n+1
2.8
Layer n 1.8Layer n+1 Bulknx Layer
Contour 1.8 Meltïpool Bulk y600
1
1.6
Contour
(SWT) Meltïpool
First (CF)Layer y300 1.6
n Layer n+1
1.6200 µm 300 µm 200 600 n Layer n+1
2 First Offset
1.8
µmµm Firstµm µm µm
3
(BF)
(BF)First (BF)
Wall Tracks
1.6
0.8
path (top1.6 view)
view)(top view)2.2
Wall
200 µm200600µmµm300 2.4 600 µm
1.8 2 µm 200 µm 2.6 300 µm 600 µm 200 µm 300 µm 600 µm

1.8
Contour
.5 Meltïpool Powder Bulk xdenudation
y z Indent 1.8 1.8
1.6
2.8 1.6 1.4
300µm
contourtrack
Powder denudation 3
Thinïwalls 200 (TW)µm 300 µm 600200
Ultraïthin wall
µmµm 300 µm 600 µm
(UTW)
First (CF) First (BF)
0.6
1.8
Layer n Layer n+1
1.8
Powder denudation
1.4 1.4
First (CF) First (BF) Powder denudation
(SWT)
1.4
1
1.6
0.2
1.6 2.4
0.6
3
2 First contour Offset Hatching 3
Crossïsection Laser path (top view) 2.6 1.4
1.6
Bulk x y Bulk
2.8 1.4 2.4
2.
200 µm 300 µm 600 µm
Layer n Layer n+1
Second contour
Face 2 Ultraïthin
Laser
Powder denudation
Contour 1Meltïpool Bulk x y z
Thinïwalls
(TW) Ultraïthin
0.8
Laser path Thinïwalls (TW) Ultraïthin wallFirst (UTW)
8
Crossïsection (top
1.6
view)
Traks (SWT)
1.4
2 Crossïsection Laser
(CF) path 1.6 (top
First (BF) view)
1.6 1.6 1.6
2.2 1.6
µm 600 µm
1.4
1.4
1.2 2.6 1.2 1.21.2
1
contourHatching
2.
d contour Single track 2.8
6
0.8
Single Wall Tracks(BF) (SWT) Laser path Thinïwalls2.2
(top view)
0.4
(BF) Face 2(TW) Ultraïthin wall (UTW) 1.6 1.6 3

0.8
2.2
1.4 First (CF) First 2.8 First contour Offset Hatching 2.4 1.2
Laser path (top 1.4view)
2.
1.4
Powder denudation
0.8 .2 1.2
4
Laser
1.4
2.6 1.2
Contour Meltïpool 1. Bulk
First (CF)x 2ywallz Single
First (BF)
1.4
Face First contour Offset Hatching
path (top view)
1

n Layer n+1
1.4
Wall Tracks (SWT) 1.4 1.4
C
Thinïwalls (TW) 5 Ultraïthin (UTW)
Single
0 0.2 0.4 0.6 0.8 Second
1 contour Single track 1.2 2
1.4 1.21.6 1 1.8 2 1 1
on
1.2
1.4
1.6
1.8
2.2
2.4 u 1.2
2.6
2.8
1.2
1.2 2.4 1.4 Crossïsection Laser path (top view)
First contourwall
1.4 0 2 0.2 3 0.4 0.6 0.80 0.210.2 1.2 0.6 1.4 1.6 1.81 21.2 1.4
20.6
0 3 0.4
0.4 0.6 0.8 0.8 1 1.2 1.4 1.6 1.6 1.8 1.8 2 2
(SWT) Single track
3
Face 2 First contour Offset Powder
2.6
Hatching
denudation
Tracks Thinïwalls Second contour (TW) Single
Ultraïthinntrack wall Bulk
(UTW)
0.6
to
Bulk x y z
0
Contour Meltïpool x y z
0.6
Single Wall (SWT) Layer Layer n+1 (top
2
1 2.8
wall
2.6 2.2 1.4 0.41.4
m 300 µm 1.2 600 µm Laser path view)
rB
1 0 0.2 0.6 0.8 1 1.2 1.4 1.6 1.8 2
z2.6 x y Thinïwalls 1.2

(TW) Ultraïthin wall (UTW)

Crossïsection Laser Layer
path n
(top Layer
view) n+1
1.
1.4
1.6
1.8
2.8
Single Wall Tracks (SWT) 2.4
1
1.2 1.2 1.2 1 1.2

1 0 0.2 0.4 0.6 0.8
8
1.2 1.2 1.4 1.6 1.8 2
1.2 Powder denudation
ul
1
3
Second contour Single track Contour Meltïpool Bulk x y z
path
200 300 600
Offset Face
1.8
µm µm µm 1
0
1.8
k
0
x Face z 2 First contour Offset Hatching
1.
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
3
1
2.2
t (CF) First 1.2(BF) Laser path (top view) 200 µm 300Crossïsection
µm 600
F
µm1.2
6
Contour Meltïpool Bulk y1.8Laser
0.2
2.4 1.8 2.8
0.4
2path (top view)
1 2.8
of
ac
trackSecond
0.4 Powder 0.6 denudation
(UTW)
0.8
2.6
1.4 1.210 1.6 1.40.2 1.8 1.60.4 2

1.
0 0.2 0.8 1 1.2 1.4 1.6
0.4
Second Bu contour Single track First (CF) First (BF) Wall Tracks (SWT)
eThinïwalls (TW) Ultraïthin wall (UTW)2

Layer n Layer n+1 2.4
First (CF) First (BF) 2
4
Face 2 First contour Offset Hatching Single
200 µm 300 µm0.6 600 µm

Layer n Layer n+10.8
Second contour Single1 track
Laser
Crossïsection Laser path (top
Powder denudation
Contour
First
1.2
1
Powder denudation 1.2
1.8
1 1.6
1
1 1
1
1 2.20.6
nïwalls (TW) Ultraïthin Crossïsection Laser path (top Laser
view) path (top view)
0.6wall 0.8 (UTW)
0 0.2 0.4 0.6 0.8 1.2 1.4 0.21 1.6 0.4 1.8 0.6 2
1.2
0 0.2 0.4 00.8 1.2 1.4 1.6 11.8 1.22
1.
0.8 0 0.2 1.4 0.4 1.6 0.6 1.8 0.8 2 1 1.2 1.4 1.6 1.8 2
0.2
B
1
(top
l
2
0.4 1 Layer n k hThinïwalls
Layer n+1
1
0.2 (TW) 0.4 Ultraïthin Thinïwalls
wall
1.6
1 (TW)
(UTW) 1.2 Ultraïthin wall (UTW)
u (CF) First (BF)

0.8
0 0.2
Second 1 1.2
0.6contour
1.4
0.8 0 Single 1 track
1.6 1.8 2 Contour Meltïpool 1.6 Bulk 1.8x y z2
1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
Crossïsection
2.8
Laser path (top1.8view)
lk
2.2 2 1.6
at0 2 0.6 0.8 1.4
0.4
0 0.2 0.4 1.2 Face 1.4 2 First 1.6 contour Offset Hatching
wall (UTW)
200Tracksµm10300(SWT) µm 600
0.2
Laser path (top view) 2.6
200nµm
Layer n Layer

Crossïsection
Contour
Single
First (CF) First (BF)Wall Traks (SWT)

µm
3
2.6
1 0
ha
Hatching
Offset
. 1.4
1
ch
z
gle Wall Layer
0.8 2 nFirst1 Layer n+1
5
Laser 2.2
1.4 pathHatching(top view) 1 200 µm 300 Singlein µm Wall Tracks
600 0.6
µm (SWT) 1 Single1.2Wall Tracks
Powder (SWT)
1.6denudation
1.8
2.4
0.8
ch
path
0.2 0.4 0.6 Face contour
1.2 Offset 1.6 Second 1.8 contour2 0 Single
0.2 track
g 0.4 2 0.8 1.4 1.8 2 1 1
view)denudation
tour First (CF) First Bulk(BF)
0.2
0.4
i
(UTW)
Contour Meltïpool Bulk x Contour y z Meltïpool Bulk x y Laser

Crossïsection z path (top view) 0
0.
0.2 0.4 0.6 0 0.80.2 1 0.4 1.20.6 1.40.8 1.6 1 1.81.2 2 1.4 1.6 1.8 2
1.6
1.4
1.4
g
Face 2 First Layer contour n Offset Hatching
0.6
Meltïpool Bulk x y z
Meltïpool x y z 200 µmcontour
1.8 300 µm Single 600 µmtrack
2.6 1.4
Layer n+1 1.2 2 3
0.6
Second
2 First contour Offset Hatching

2 2.4
Single Wall Tracks (SWT) 2.4
1.2
0
ro
Powder denudation Powder denudation
WallContour
3
2.8
2.6
2.4
2.2
1.8
1.6
1.4
1.2
1
2.2
(top
Thinïwalls (TW) Ultraïthin wall (UTW) Second contour
2
Single track 1.6 0.
Hatching
300
0.6
contour
200 300 600 4
Layer n Layer n+1 µm µm µm
Meltïpool
wder denudation 1
Contour Meltïpool Bulk x y z 1.8
Crossïsection Laser pathcontour
(top view)
0.6 view)
Crossïsection Laser path (top view) Face 2 First Offset Hatching
1
600Layer n Layer n+1

1.2
200
Layer
Second
Laser
Powder denudation
Single
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
0.4
2.4 1.2
(TW) Ultraïthin wall (UTW)

Single Wall LaserTracks (SWT) 1.6
200 µm 300 µm Powder
0.8
2.8
1.6
µm denudation
Tracks (SWT)
1.8 2.2
1.40
Laser path (top view) contour Single track 2.2

1.4
µm µm
ssïsection path (top view) First (CF) First (BF) Second
3
2.8
2.6
2.4
2.2
1.8
1.6
1.4
1.2
2
200 µm 300 µm 600 µm
1.8 1.4
0.4
Crossïsection Laser path (top view) Face 2 First contour 1.6
OffsetFace 2 Ultraïthin
HatchingFirst Layer
contour Offset Hatching
0.8
Laser
n+1
Thinïwalls (TW) n (UTW)
wall Layer n+1
600 300
er path (top view) 1. 1 1
0.2
Single
2.2
1.4
Laser path (top view) 5 0 0.2 0.4
Second 0
contour
0.6 0.2 0.8 0.4 1 1.2
0.6 1.4
0.8 1.6
1 2.6 1.8
1.2 2 1.4 1.6 1.8 2
200 Single
µm 300track
1
1.6
Second contour Single SingleWalltrack
n Layer n+1
path
2
1.6
2
Tracks (SWT)
1.8
µm 600 µm 1.8
Powder denudation
Bulk
1.6 1.2
-p
0.2
e 2 First contour Offset Hatching Face 2 First contour Offset Hatching Layer n Layer n+1
1.80.6
1.2
contour
µm µm 600 µm
1.4
Layer n Bulk Layerx n+1
Contour Meltïpool y z 1.
path
12
0
2
1.2
1.2 2.4
200 µm 300 µm 600 µm
track Single track

3
2.8
2.6
2.4
2.2
1.8
1.6
1.4
1.2
1
ond0.6contour Single track
1.4
200 µm 300 µm Powder 600 µm denudation
(top
1.8
1.8
1.8
x
1.
2 4
1.6
0.8 1 1.2 1.4 1.6 1.8 Layer 2n Layer n+1 1.4 1
2
(topview)
0
y
1.2
Crossïsection Laser path (top view) 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
3
view)
2.8
2.6
2.4
2.2
1.8
1.6
1.4
1.2
1
1.2
er n Laser
Layerpath n+1 (top view)
200 µm 300 µm 600 µm
1.
2 0.8
6
200 µm 300 µm 600 µm
z
1
1.4
1.8
1.2 10 2.2
Laser path (top view) 1.6 1.8 1.6
2
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 2
view)
1.4
Faceµm 2 600Firstµm contour Offset Hatching
1.2 1.
µm 300 Face 2 First contour Offset Hatching 8
re
1
1.4
Layer n Layer n+1

Powder denudation
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

1.6
0.2 Second contour 0.8Single 1 track Second contour Single track

1.6 2
1 1.4 1.4
0.4 0.6 1.2 1.4 1.6 1.8 2
0.8
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

t (CF) First (BF) 1
0 0.2 0.4 Layer n Layer n+1 0.6 0.8 1 1.2 1.4 1.6 1.8 2
1.2
1.6
nïwalls Layer(TW) n Layer Ultraïthin n+1 wall (UTW) 200 µm 300 µm 600 µm
1.8
1.4 1.8
1.2 1.2
1
200Tracksµm 300(SWT) µm 600 µm
1.2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

0.6
gle Wall
1.8
lP
tour First (CF) First Bulk(BF)

1.2
2
1.6 1
1
Meltïpool x y z 0 0.2 0.4 0.6 0.8 0 1 0.2 1.2 0.4 1.4 0.6 1.6 0.8 1.8 1 2 1.2 1.4 1.6 1.8 2
1.4
wder denudation (TW) Ultraïthin

Thinïwalls wall3(UTW)
2
3 3 3
0.4
1 1.4
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Single Wall
ssïsection Laser path (top Tracks (SWT)
2.8 view) 2.8 2.8 2.8
1.6
er path Contour
(top view) Meltïpool Bulk x y z 1.2
0.2
Powder denudation (a) (b) (c) (d) (a) (b) (c) (d) (a) (b) (c) (d) (a) (b) (c) (d)
a
e 2 First contour Offset2.6Hatching 2.6 2.6 2.6

1.8
1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
.4
Crossïsection
0.6
ond contour
0.8
Single track
1
1.2 1.4 1.6 1.8 2
0
2.4
Figure 1. Schematic 2.4 2.4 of the four main2.4
3
2.8
2.6
2.4
2.2
1.8
1.6
1.4
1.2
representation types of samples and their respective

er n Laser
Layerpath n+1 (top view)
rn
Face 2 First contour laser track patterns;

Offset Hatching (a) single wall track (SWT) samples, corresponding to a stack of single
µm 300 µm 600 µm 2.2 2.2 2.2 2.2
0.2 Second 0.4 contour
0.6
tracks;
0.8Single
(b) 1.2
1 track
ultra-thin 1.4
wall 1.6
(UTW) 1.8
samples,
2
corresponding to two parallel SWT; 600 µm-thick
thin-wall
2 (TW) 2samples with (c) contour 2 first (CF) strategy 2 and (d) bulk first (BF) strategy.
Layer n Layer n+1
u
The mentioned thicknesses will be slightly affected by the laser parameters. Location of
200 µm 300 µm 600 1.8µm 1.8 1.8 1.8
hardness indents are shown on the x-z plane.
Jo
1.6 1.6 1.6 1.6
1.4 1.4 1.4 1.4
1.2 Finally, 1.2 optimum parameters 1.2 found with SWT,

1.2 UTW and TW samples were
.4 0.6 0.8 1 tested
1.2 on
1.4 TW1.6samples
1.8 and2 on a 10 mm-side cube. The cube was processed with a CF
1 1 1 1
strategy.
0 The roughness
0.5
0 and 1 hardness
0.5 0 of
1 those
1.5 0.5 samples2
0 1.5 were
1 also
0.5 2 measured.
1.5 1 Hard-
2 1.5 2
ness measurements were performed on the contours and on the bulk hatching in the
0.2 0.4 0.6 0.8 xz-plane.
1 1.2 1.4 1.6 1.8 2
For the present study, 10 batches were processed. The number of samples per
batch was kept similar (60 ± 10) to maintain the job load. Indeed, small roughness
variations can be measured when the job load changes drastically. Each series of
6
the same parameter were done on the same job. A limited number of samples were
produced on different batches to assess reproducibility. Thin walls with two different
contour TED have been produced perpendicularly to each other, in order to check
the impact of the orientation as it can affect the roughness (54). No difference was
observed. This will thus not be further reported. The build-plate is an Al5Mg (wt%
Mg) square of 141 mm by 141 mm and is 15 mm-thick. Samples were widespread on
the plate and equally separated by a distance of 10 mm. The height of the samples
was fixed to 10 mm. Taller samples (100 mm) were produced to check the impact of
the height but no significant difference was found.
2.2. Characterization techniques

Roughness analyses were carried out by white light interferometry, referenced as co-
herent scanning interferometry (CSI) (55) with a MSA-500 Micro System Analyzer
of
(Polytec) built on a vibration-isolation table (SUSS, VIT80x range). CSI is a powerful
method to perform topographical analyses. Measurements were carried out following
the Good Practice Guide No 116 (56) and some other works (57; 58; 59) in order to
ro
obtain reliable data. A x10-objective was used to insure a good trade-off between
field of view (880 x 660 µm map) and adequate lateral resolution (635 nm x 635 nm)
and gave a numerical aperture of 0.3. Raw measurements were analyzed with Poly-
-p
tec’s TMS software 2.0 (Topography Measurement System). The measurements were
carried out in short coherent mode. The cutoff wavelength was set to 800 µm for λc
and 2.5 µm for λs for each profile analyzed in agreement with the EN ISO 13565-1
re
standard. 5 sampling length of 130 µm (limited here by the field of view) were used.
Each measured topographical map was divided into 60 vertical profiles (z-direction)
and a mean arithmetic average roughness (Ra ) was computed out of these profiles. In
lP
this work, Ra was considered because it is one of the most adopted texture parameter
(57) and it is thus largely used. Even though this parameter does not bring much
information about the type of roughness, Ra is mainly used in many applications. On
each sample, 8 maps were acquired (4 per face) for a total area of 3.85 mm2 , insuring a
reliable value of the arithmetic roughness (60). Error bars are the standard deviation
a
between the maps and show how the data are scattered. Indeed, the local nature of
CSI measurements tends to increase the scattering of the data. Rz and Rmax were also
rn
measured on each samples but gave the same trends as Ra (see supplementary mate-
rial). Thus, only Ra will be presented here. It is worth noting that our measurements
cannot directly be compared with Sa measurements as no surface filtering is used here.
u
Hardness was measured with a DuraScan G5 from EMCO-TEST with a Vickers

Jo
tip indenting with 200 g (HV 0.2). The measurements were carried out on x-z cross-
sections (Figure 1) of samples polished down to 1 µm. Ten indents were performed
per measurement to insure reliable data. However, the heterogeneous microstructure
of AlSi10Mg processed by L-PBF (29) may increase the scattering in hardness (38).
As indent sizes were about 50 µm large, they were performed with a 300 µm-space
between them in the building z-direction. Indents were performed in the middle of
the SWT (Figure 1.a). For hardness measurements of TW, indents were performed in
the middle of the contour and in the middle of the bulk (Figure 1.c).
The melt-pool structure of the samples was characterized after grinding and
polishing steps, followed by a chemical etching using a Keller’s etchant. Scanning
7
electron microscopy (SEM) images were performed only on some relevant samples.
Several TW samples were also analyzed by X-ray computed micro-tomography (61; 62)
in order to characterize the internal porosity. A 1 mm-thick section (in y-direction) was
extracted from each sample for the measurements (giving parallelepipedic dimensions
of 600 µm x 1 mm x 10 mm in xyz directions, respectively). The acquisitions were
performed with a vtomex laboratory X-ray computed tomograph (GE Phoenix X-Ray
GmbH) equipped with a 160 kV nano-focus tube, a tungsten transmitting target, and
a 1920 x 1536 pixels Varian detector. Scanning parameters were set at 80 kV tube
voltage and 280 µA current with a voxel size of 2 µm3 . A full analysis consists of 600
2D radiographs and an averaging of 3 images at each step angle. The exposure time
for one radiograph was set to 0.5 s. The volumes were then reconstructed by a filtered
back projection Feldkamp-algorithm. The reconstructed data were finally processed
and visualized with ImageJ/Fiji shareware (63). Data for each sample were binarized
in order to visualize the porosities and to compute a mean density.
of
3. Results and Discussion
ro
3.1. Single wall tracks: effect of laser power and scanning speed on
roughness
-p
Figure 2 summarizes roughness and hardness measurements performed on single wall
track (SWT) samples built at a fixed scanning speed of 200 mm/s (Figure 2.a) or at
a fixed power of 273 W (Figure 2.b), for a large range of TED. Furthermore, several
re
optical microscopy images illustrate how the microstructure changes with the level
of TED. In both cases, the roughness does not change monotonously with the TED
parameter. A minimum Ra can be reached, depending on the laser scanning speed
and power. Topological maps show that surface defects are also different depending
lP
on the TED level.
A minimum roughness of 5.9 µm is reached for a TED level around 1000 J/m
for a scanning speed of 200 mm/s (Figure 2.a). It is worth noting that this TED
a
level corresponds to the transition between self-balling (characterized by a fluctuating

melt pool depth) and keyholing mode in this case. At high energy density, balling
rn
appears due to the Plateau-Rayleigh instability (47; 49; 44; 53), caused by a surface
energy minimization of the melt pool. On the other hand, keyhole instability is ob-
served when the energy injected in the melt pool becomes too large (64). Material
u
vaporization within the melt pool induced by an excessive amount of energy causes
the melt pool to be unstable (65). In keyholing mode, large porosities appear in the
Jo
center of the SWT, as shown on the micrographs in Figure 2.a. The increasing Ra
in keyhole conditions (above a TED of 1000 J/m) is due to an increasing amount of
sintered powder on the side of the SWT samples. Increasing the level of TED also
increases the time spent by the samples at high temperature, thus enhancing the pow-
der sintering. Large clusters of sintered powder then appear at very large TED level,
as shown by the corresponding topographical maps in Figure 2.a. As clusters are not
homogeneously distributed, this increases the roughness scattering. On the contrary,
very few sintered powder is present in the optimum range (around a TED of 1000 J/m).
At fixed power, Figure 2.b shows that Ra evolves similarly with TED as observed
for the previous case. The lowest Ra , 4.5 µm, is obtained near the keyhole conditions
8
as well, around 700 J/m in this case. There is still a roughness decrease at low TED
for 273 W but large porosities appear due to the too low energy density for this struc-
ture. Similarly to Fig. 2.a, powder sintering is also observed for larger TED.
In both cases studied here, a coarsening of the microstructure within the melt-
pools by increasing the level of TED is observed. Melt-pool microstructure of
AlSi10Mg is divided in 3 distinct zones, called fine zone, coarse zone and heat af-
fected zone (HAZ) (29). Fine zones are the core of the melt-pools where the size of
the cells of the microstructure is the smallest. Coarse zones are the border of the
melt-pools, where the size of the cells is larger than in the fine zone. The HAZ are
just outside melt-pool boundaries where the microstructure is globularized. Figure 3
presents SEM micrographs of three SWT for which hardness and roughness are pro-
vided in Figure 2.a. The microstructure within the three zones are affected by the level
of TED (Figure 3.a). The size of the cells within each zone increases by increasing the
of
level of TED. The size of the cells doubled when the power was increased from 109 W
to 274 W (Figure 3.b), explaining the hardness decrease. Indeed, strengthening in ma-
terials is induced by hindering the dislocation glides through solid solution hardening,
ro
precipitation strengthening or by reducing grain size. In the case of the as-built L-PBF
AlSi10Mg, previous studies have shown that strengthening mechanism is governed by
the size of the fine microstructure rather than by the grain size (29; 26; 66), where the
-p
Si network forming the cell boundaries pins dislocations and induces Orowan looping
mechanism (67). Coarsened microstructure leads then to a decrease of the hardness
(37; 38).
re
This microstructure coarsening, due to a longer time at high temperature (29; 68),
induces a decrease of the hardness, from 127 HV to 108 HV at a fixed scanning speed
lP
(Figure 2.a) and from 132 to 108 HV at a fixed power level (Figure 2.b). Indeed,
similar hardness decrease can be observed on AlSi10Mg parts after heat treatment
(54; 69; 70). The heterogenous microstructure observed within the SWT (Figure 3.a)
and the size of the indents ( ± 50 µm) are responsible for the hardness data scatter-
a
ing. Indeed, hardness on the edges of the melt pools (coarse and HAZ zones) is lower
than in the core of the pools (37; 38) (fine zone). Moreover, indents performed on the
rn
thinner samples are near the free-wall surface, which may also induce some scattering
due to border effects on plasticity confinement.
u
However, the observed trends are in line with the microstructure evolution (i.e.
the coarser the microstructure, the softer is the material).
Jo
Figure 2 thus shows that a minimum roughness is obtained for a level of TED
close to the one bringing keyhole instability. A process window minimizing Ra thus
exists, just as it is the case for maximizing the density. However, the required level
of TED is significantly larger to minimize Ra than to maximize the density (120-200
J/m). As a consequence, conditions for contours and bulk hatching must be different.
It is worth mentioning that other conditions of fixed scanning speed and power were
also tested and led to the same conclusions (see supplementary material).
9
2.8
3.2
3.4
3.6
3.8
4.2
4.4
4.6
4.8
2.6
3
5
1

2.4
Powder denudation
3.2
3.4
3.6
3.8
4.2
4.4
4.6
4.8
2.2
3
5
1.2
3.2
3.4
3.6
3.8
4.2
4.4
4.6
4.8
3.2
3.4
3.6
3.8
4.2
4.4
4.6
4.8
of
3
5
1
5
1
2
Layer n Layer n+1
200 µm 300 µm 600 µm
1.4
1.8
3.2
3.4
3.63
3.8
4.2
4.4
4.64
4.8
1.2
1200 mm/s 1500 mm/s
200 mm/s 300 mm/s 400 mm/s 500 mm/s 800

3.2 1.2
3.4
3.6
3.8
4.2
4.4
4.6
4.8 1200 mm/s

3.2
3.4
3.6
3.8
4.2
4.4
4.6
4.8
274 W 246 W 219 W 205 W 192 W 164W 137 W 109 W
3
531
1.2
5 200 mm/s
5
1
1.6
ro
3.2
3.4
3.6
3.8
4.2
4.4
4.6 1.2
4.8
1.6
5
3
3.2
3.4
3.6
3.8
4.2
4.4 1500 mm/s
4.6 3003.6
4.8
1
1.4
3.2
3.4
3.6
3.8
4.2
4.4
4.6
4.8
1.4
1.2
1.4
5
1.2
5 5
3
4200 mm/s 300 mm/s 4.8
5
5
1
12003.2
1
274 W 246 W 219 W 205 W 192 W 164W 137 W 109 W
1.4
5 4.8 5 5
1.2
200 mm/s 300 mm/s 400 mm/s 500 mm/s 800 mm/s 1000mm/s
4.8 4.8
3.2 1.4
3.4
3.6
3.8
4.2
4.4
4.6 1.6
4.8
4.8
3.4
3.8
4.2
4.4
4.6
4.8
mm/s 1500 mm/s
1.8
5
1.2
mm/s
2
5
219W4.8 4.8
1
4.6
2
W 246 W 219 W 205 274
W W 192 246
W 164W 137
W 219 W W205
109
WW274
192 W
W 164W
246 W 137 W109 W W 192
205 200Wmm/s
164W W137 W 109 W0.4mm/s 4.8
-p
1
274300 1200
246
0 mm/sW mm/s
219
400 W 1500
205 mm/s
W
0.6 500 192
0.8mm/s
200W 164W
1 800
mm/s mm/s137
300 W 109
1.41000mm/s
mm/s W 1.8mm/s
1.6 400 2 500 mm/s 800 mm/s 1000mm/s
1.4
1.6
4.6 4.6 4.6

1
0.2 1.2 200 mm/s 300 mm/s 400 mm/s 500 mm/s 800 mm/s 1000mm/s
1.2
1200
200
1.2
1200 mm/s 1500 mm/s
274 W 246 W 219 W 205 W 192 W 164W 137 W 109 W
1.4
274
1.6W 246 1.8
4.6 12004.4
mm/s 1500 mm/s 4.6 4.6 mm/s
1200 mm/s 1500 1200 mm/s 1500 mm/s
4.4 4.4 400
mm/s
4.4 150 150
mm/s
1.2
1.4
2
20 4.420
20
1.2 4.4 1.4
1200 mm/s 1500 mm/s
4.4
1.6
1.8
mm/s
4.2 4.4
274 W 246 W 219
1.6
4.2 4.2
1.8 1
4.2
3.2
3.4
3.6
3.8
4.2
4.43.2
4.63.4
4.83.6
1.4
1.4
18
300
18
18
1.63 274 W
3
53.8
4.2
4.6
1
1500
400 mm/s 500 mm/s 800
W 219 W 3.2
re
44.2 4 140 4.2 4.2
4
4 4 140
1.5
1.5
500mm/s
16 16
2.2
16
mm/s
1.4
1.6
1.8
Hardness [HV]
Hardness [HV]
1.8
2
3.8 3.8 14 3.8 4 3.8 414 4

mm/s 5800 mm/s 1000mm/s
14
2 1.2 2.2 1.4 2.4 1.6 2.6 1.8 2.8 2
1.6
1.2
1.6
1.8 246
130 130
400 mm/s 500 mm/s 800 mm/s 1000mm/s
m]
Ra [[µm]
m]
Ra [[µm]
3.6 12 3.812
12
3.4
3.6
3.8
4.2
4.4
4.6
4.8
3.6 3.63.8
2
3.6 3.8
205 W 192
1200 mm/s 1500 mm/s
200
2.4
1.6
1.8
3 W 205 W 192 W 164W 137 W 109 W
4
1 W 219
mm/s
2.2
10 10
10
lP
2
3.4 3.4
1.6
2
mm/s1000mm/s
3.43.6
a
3.4 3.6 3.6

R
120 120
1.4
8
1.8
88
1.8
274
3.2
2 1.2
3.2
1
3.2
2.2W 164W2.4
1
2.6
3.4 3.2 3.4 6 3.4

1.8
6 6
300 mm/s 400 mm/s 500 mm/s 800 mm/s 1000mm/s
2
1200 mm/s 1500 mm/s
200mm/s
2.2
2.4
W
W 205
2.2
3 3 33 110 110
1.8
1 1.6 1.21.8 1.4 2 1.62.2

1 1.8 4
2.4
1.2 2.6
21.4 2.8
2.2
1.6 2.4
1.83.2
1 2.62
1.2 2.8
2.2
1.4
3 32.4
1.6 1.83.2 4
2.6 4 2.82 3
2.2 2.4
3.2 2.6 2.8 3
1.6
246
mm/s 1000mm/s
1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3

2
2.2
2 22 3
2.8
3
W
2.2
W1.4
3 3
2
3
137 W 109
2.4
2.6
1 1.2 1.4 100 1.6 11.8 1.22 2.2

1.4 2.4 1.2 1.8
1 1.6 2.6 1.4 22.8 100
1.6 2.23 1.8 2.4 2 2.6 2.2 2.8 2.4 3
2.4
2.6
300 mm/s 400 mm/s 500 mm/s 800 mm/s 1000mm/s
a
0 00
219
192 W
0 250 500 750 1000 1250 1500 00 250

250 500 750 1000 1250 1500
1.8
0.5
0.5
2.8 2.8
W2.2
2.2
TED [J/m] TED [J/m]

2.4164W
2.2
2.4
3
2.6
205
2.6
2.8
2.6
rn W
1.6 137
2.6 (a) (b) (c) (d) 2.6 (a) (b) (c) (d)
W
2.2
2
2.4
2.4
192
2.4
2.6
2.6W1.8
2.8
2.8
Figure 2. Evolution
2.4 2.4 as a function of TED for (a) a fixed
of roughness and hardness of SWT
2.8
3
W
2.2
109 W2.8 2
0
scanning speed of 200 mm/s and for (b) a fixed laser power of 273 W. Optical microscopy
164W
2.4
3
2.5
1.5
2.5
1.5
1
u
2.2
3 2.2
2.6
2.8
2.2
2.6
2.6
images of SWT cross-sections illustrate the internal structure of the SWT for specific TED
3
3
3
137 W 2.8
2.4
levels while 2topographical maps corresponding to extreme 2 processing conditions are also
Jo
2.6
2.8
provided.
3
109 W
2.8
3 2.2
2.4
1.8 1.8
2.6
2.8
3
1.6 1.6
3
3.2. Ultra-thin walls: effect of the offset between tracks on roughness

2.6
2.8
2.4
1.4 samples thus allows finding a process 1.4

3
Using SWT window for contours leading to a

minimum 1.2level of roughness. This simple structure 1.2 discriminates in Section 3.1 the
2.8
3
2.6
effect of the scanning speed and laser power from the offset since there is no bulk
hatching in 1this case. The0.5effect of an offset 1 space) between melt pool tracks
(or hatch
0 1 01.5 0.52 1 1.5 2
3
2.8
on the roughness is now investigated by using structures called ultra thin walls (UTW,
Figure 1.b).
3
2
10
1.8
!()$%$ !()$%$ *+,-$
1.6
1.4
109 W HAZ 164 W 274 W
1.2
!"#$"
Power Fine Coarse
1
!"#$%$
0.8
*+,-$
2 µm 3 µm
!"#$"
0.6
of
./012-$ &'#$%$ *+,-$
0.4
ro
!"#$" 0.2
345$
-p
0
&'#$%$
3
2.8
2.6
2.4
2.2
1.8
1.6
1.4
1.2
*+,-$
re
lP
3 ./012-$ 3
!"#$" !"#$"
2.8 345$ 2.8
2.6 (a) (b) (c) (d) 2.6 (a) (b) (c) (d)
a
Figure 3. 2.4SEM micrographs of SWT processed at 2.4

200 mm/s with different power showing
rn
(a) the three different microstructure zones. (b) Zoom on the fine microstructure for the two
2.2 2.2
extreme power. Micrographs were taken on the top of the SWTs. No inhomogeneities were
found in the2 z-direction as SWT are simple structures2 and small in height.
u
1.8 1.8
Figure 4 presents roughness measurements1.6performed on both sides of several
Jo
1.6
UTW samples. Ra is presented as a function of the distance between two parallel
tracks. A1.4
100 µm hatch space was chosen as a1.4starting point and was increased by
steps of 30 µm. The samples were processed with1.2
two levels of TED. Ra of SWT with
1.2
the corresponding TED level are also given as a reference.
1 1
0 0.5 1 01.5 0.52 1 1.5 2
Ra value of ”face 1” (face corresponding to the first processed track ) with a TED
level of 194 J/m is not much impacted by the distance between the tracks. On the
contrary, Ra value of ”face 2” decreases gradually as the offset increases, converging
towards the Ra value of SWT specimen. In the case of a higher TED level, the con-
clusions on roughness behavior are similar. The Ra value of ”face 2” decreases with
the offset.
2
2
40 11
40
35 233 W - 1200 mm/s (TED = 194 J/m)
35 233 W - 1200 mm/s (TED = 194 J/m) 271 W - 800 mm/s (TED = 339 J/m)
2 2
40 271 W - 800 mm/s (TED = 339 J/m) Face 1 30 40 data3 Face 1
30
12
data3 Face 2 data4 Face 2
2
35
Face 1
25 35
Face 2
Face 1
Ra [ m]
1.5
1.5
25 data4 1.5 data5
Ra [ m]
1.5
30 data5 20 30 data6
10
20 data6 data7
25 15 25
[ m]
[ m]
2
Ra [µm]
R [µm]
2
data7
1.5
15 1 1 data8
20 200 µm data8 10 20 200 µm data9
8
1.5
a
10 data9
3
1.5
data10
3
15 data10 5 15
0.5 0.5 data11
1
1 5 data11
2.8
1 2.8
10 First (CF) First (BF) 0 10 data12 First (CF) First (BF)
1
0
3 3 33 3 data12
2.6
Thinïwalls (TW) Ultraïthin wall (UTW) 3 33 3 3 data13
2.6
T
5
4 12 2.82.8 5
m
0.5 3
data13 Single Wall Tracks 0.5(SWT) 0 3 data14 Single Wall Tracks (SWT)
SW
4 0
T
m
m
0 2 4
Contour 6
Meltïpool 8
Bulk x y 10
z 0 2 4
Contour 6
Meltïpool 8 x y 10
Bulk z 12
W
2.82.8
2.82.8 2.8
2.8
2 m 30
0
data14
2.4 2.4
0
10
m 16
0.5 m 19
0 data15
] 4S
0
]0
1
3.8 100
100
100
100 130
130
130
130 160
160
160
160 190
190
190 190 100
100
100
100 130
130
130
130 160
160
160
160 190
190
190
190
10
m1]3
m1]6
m19
0 data15
2.82.22 Crossïsection Laser
0 path (top 3.8
view) 1.5 2.8 data16 Crossïsection Laser path (top view)
0 0.5 2.61
2.6
2.6
2.5 1.5 12.6
2.6
2.5 Offset
T
m
m
m
2.6 2.6
2.6
m
0 0.5 2
T
m
m
0.5
SW
2.2
SW
data16 Laser path (top
Offset tracks view)
DistanceLaser path (top view)
a 0[µ
a [µ
a 0[µ
a 0[µ
Distance between Ra [µm] between Ra [µm]

tracks
R10.5
3.6 3.6 2.42.42.42.4

90
Face 2 First contour Offset Hatching Face 2 First contour Offset Hatching
0
2
2
R10
R13
R16
2.4
2.42.4
10
13
16
19
2.4
(a)contour
(b) Single
(c) track
(d) (a) (b) (c) Single
(d) track
2
2.6 2
Second 2
Second contour
Offset 2.6
3.4 2.2
2.2
2.2 2.2
2
Layer n Layer n+1 3.4 2.2
2.2
2.2 2.2 2 Offset
Layer n Layer n+1
1.8 200 µm 300 µm 600 µm 1.8 200 µm 300 µm 600 µm
3.2 2.4 3.2 2.4
0
2 2 22 2 22 2
0
2
1.5
0.51
0.5
Figure 4. Evolution of the arithmetic roughness of first (in red) and second (in black) contour
2
1.5
of
1.6 1.6
5
5
5
5
3 1.5 samples as a function of the distance3 1.81.81.81.8 between
1.5
3 1.
1.
1.
1.8
1.8
1.
1.8
1.8
of UTW 2.2 1.4
2.2tracks, processed with a TED level 1.4
0
0
3 3
2.5
1.5
2.5
1.5
1
2.8 of (a) 194 J/m and (b) 339 J/m. The 2.8
1.6
1.6
1.6 1.6
1.2
corresponding power and scan speed are indicated
1.6
1.6
1.6 1.6
1.2
3
3
3
on 1.4
1.4
1.4 2
1.4 top 1of the graphs. Corresponding optical microscopy
1.4
1.4 1.4 1 2images of x-z cross-sections are also
ro
1.4
2.6 0 0.5 1 1.5 2.6
1
0 0.2 2 0
0.4 0.6 0.5 0.8 1 1 1.5 1.2 1.4 1.6 1.8 2
1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2
provided. R a level of SWT specimens processed with the same level of TED are indicated as
1
1
1
1.2
1.2
1.2 1.2 1.2
1.2
1.2 1.2
2.4 1.8 2.4 1.8
5
a1 1 reference.
5
5
5
2.
2.
2.
2.
11 11 1 1
0 0 00 0.2
0.20.2 0.4
0.40.4 0.6
0.60.6 0.8
0.80.8 1 1 11 1.2
1.21.2 1.4
1.41.4 1.6
1.61.6 01.8
1.8 01.8
00 20.2
20.220.2 0.4
0.40.4 0.6
0.60.6 0.8
0.80.8 1 11 1 1.2
1.21.2 1.4
1.41.4 1.6
1.61.6 1.8
1.81.8 2 22 2
2.2 0.2 0.4 0.6 0.8 1.2 1.4 1.6
2.2 1.8 20.2 0.4 0.6 0.8 1.2 1.4 1.6 1.8
1.6 1.6
-p
2
2
2
2 2
0 0.5 1 1.5 02 0.5 1 1.5 2
5
5
5
5
0.
0.
0.
0.
1.4 1.4
Micrographs of Figure 4.a show that the two SWT constituting the UTW start
5
5
5
5
1.
1.
1.
1.
to separate 1.2
for a hatch space of 190 µm, leaving pores 1.2 between them. In the case of
re
the higher TED level, the two tracks are not yet separated for an offset of 190 µm due
1
1
1
1
0
0
0
1 1
0
to a better overlapping 0.5
of larger 1
melt-pools. 0 1.5 0.5 2 1 1.5 2
lP
The results of Figure 4 clearly show that the distance between parallel tracks
influence the resulting roughness. Changes of the melt pool morphology between
parallel tracks have already been reported on L-PBF of stainless steel (71). This effect
is even more evidenced when TED increases as a result of the melt pool widening, thus
a
increasing interactions between superimposed melt pool tracks. Optical microscopy

images of Figure 4 show that this interaction induces the melt pool of ”face 2” to flow
rn
outward, especially for the distance of 100 µm. Those protuberant melt pools have a
huge impact on roughness. The distance between the contours and the bulk hatching,
namely the offset, has thus a direct impact on roughness.
u
3.3. Thin walls: effect of the strategy and offset

Jo
Figure 5 presents the evolution of the arithmetic roughness of thin-wall samples (TW)
as a function of the offset, for both building strategies, i.e. ’bulk first’ and ’contour
first’. SEM micrographs of TW surfaces (y-z plane) for the two extreme offsets are
also provided. Figure 5 highlights two first-order effects related to the building strat-
egy and the amplitude of the offset. Increasing the offset has a significant impact on
the BF strategy, decreasing Ra from 35 µm down to 13 µm. However, it is almost
effectless in the case of CF strategy, for which the mean level of Ra is about 9 µm
for all offsets. For the same offset, a CF strategy always leads to a lower Ra than for
the BF strategy. Many other TED levels were also tested and showed similar trends
(see supplementary material). SEM micrographs of Figure 5 also show that the large
level of Ra in BF strategy is mainly due to huge protuberances (of about 150 µm).
12
The amount of protuberances decreases when increasing the offset. Sintered powder
is also observed on the surface but does not impact significantly Ra since the size of
the powder is much smaller than the protuberances. In the case of the CF strategy,
SEM micrographs show no more protuberances. Not much sintered powder can still
be observed.
It is worth noting that the level of Ra of ”face 1” (corresponding to the first

processed contour) is quasi systematically slightly lower than the Ra level of ”face
2” in the case of the BF strategy, for each offset. This difference is less or even not
noticeable anymore in the case of the CF strategy, whatever the value of the offset.
However, the trends of the Ra level observed are similar on each face. As this is a sec-
ond order effect and systematic, faces will no longer be distinguished in what follows.
Ra will then be taken as the mean of the two faces. Some possible explanations for
this effect are provided in supplementary material.
of
Micrographs of Figure 6 illustrate cross-sections in the x-z (Figure 6.a) and x-y
(Figure 6.b) planes of TW samples for three offsets and for both strategies. These
ro
micrographs also show that the excessive roughness with a BF strategy is mainly due
to protuberant or ”flowing-outward” contour melt pools highlighted by black arrows
in Figure 6, as observed on UTW samples. This ”flowing-outward” behavior of the
-p
melt pools of the contours is not observed with a CF strategy, where the contours are
well located. Melt pools of the contours of the last processed layer (top layer) show a
different shape depending on the strategies. With a CF strategy, the top melt pools
re
are pointing vertically (black dashed arrows in Figure 6.a). On the contrary, with a
BF strategy, top melt pools are pointing more and more horizontally when the offset
decreases. This effect indicates that on a given layer, the already solidified bulk melt
lP
pools with a BF strategy disturb the contour tracks. It tends to push out the contour
melt pools towards the unmelted powder.
Figure 5 shows that the value of Ra in the case of a BF strategy converges towards
a
the value of Ra in the case of a CF strategy for large offsets. However, an excessive
increase of the offset, corresponding to a decrease of the overlap between the bulk
rn
hatching and contours, leads to porosities appearing between the bulk and contours
(white arrows on Figure 6), affecting the sample structural integrity.
u
In order to more statistically quantify this effect, some samples were character-
ized by X-ray tomography. For each sample, a mean local density was computed for
Jo
all y-z plane measurements. Figure 7 presents the mean local density of TW samples
following the x-direction for both building strategies and for several offsets. This figure
also shows two reconstructed volumes of TW samples with a 120 µm-offset, showing
porosities. A decrease in the local density near the surface is observed, starting at an
offset of 90 µm for both strategies. This drop becomes extremely significant for an
offset of 120 µm, decreasing the mean local density to 92 % in BF strategy and 97
% in CF strategy, respectively. The reconstructed volumes of the TW samples built
with a 120 µm offset clearly show a large number of sub-surface pores. Interestingly,
an asymmetry in the local density is observed for the BF strategy as it was observed
for roughness (Figure 5). This effect could be attributed to a small thermal effect due
to the asymmetric process itself. Indeed, heat brought by the first processed contour
(Figure 1) slightly increases the temperature of the already solidified top layer so that
2
13
40 2
Face 1
Face 2
40
Bulk first
1.5
1.5
Contour first
30
data3
30 data4
m]
Ra [[µm]
1 data5
!""#$%# !""#$%#
20 data6
Ra [µm]
a
40 data7
20 R
0.5 data8
Bulk first
1
data9
Contour first
10 data10
data3
3 0 data11
data4
30 0 2 4 6 8 10 12
10 data12
data5
data13
data6
0
Ra [µm]
!""#$%# 2.5 0 data720 40 60 80 100 120 !""#$%#
0.5
20 data8
0 Ra [[µm]
Offset m]
1 2 3data9 4 5
of
data10
Figure 5. Evolution of the arithmetic
data11 roughness (Ra ) of thin-walls (TW) samples
as a
10 2
function of the offset for both data12
strategies. The TED value of the contours is the same
than
data13
the TED value of the bulk hatching (232.5 W at 1200 mm/s, corresponding to 194 J/m).
ro
0
SEM micrographs show the surface of some TW samples (identified by black dashed arrows).
3
2.5
1.5
0
Black arrows 1.5
on the2 micrographs indicate examples5 of protuberant contour melt pools. Red
1 3 4
arrows highlight examples of sintered powder particles.
40
1
0 0.5 1
-p 1.5
40
2
re
0 µm 0 µm
3 3 3 3 30 3µm 3 30 µm
40 40
30 60 µm 30 60 µm
5
5
lP
0 µm 90 µm 0 µm 90 µm
40 30 µm 40
data5 30 µm data5
2.5 2.5 32.5 2.5 2.5 2.5
Ra [µm]
Ra [µm]
30 Offset R
= a0[µm]
µm R60
a
[µm]
µm R
120[µm]
µm 30 Offset R
=a0[µm]
µm R60
a
[µm]
µm R
120 [µm]
µm
a a
30
20 µm 90 µm data7 30
20µm 90 µm data7
60 µm data5 data8 60 µm data5 data8
4
30
4
2 2 2 2 302 2
Ra [µm]
Ra [µm]
Contour first
a
Contour first
2.5 90 µm 120 µm data9 90 µm 120 µm data9

20 data5 data7 data10 20 data5 data7 data10
Bulk first
Bulk first
R [µm]
data10
data11
data12
data13
data10
data11
data12
data13
10 µm
Ra [µm]
120 data8 data11 10 µm

120 data8 data11
data3
data4
data5
data6
data7
data8
data9
data3
data4
data5
data6
data7
data8
data9
rn
1.5 20
1.5 1.5 data7 data9
1.5 data12
1.5
20 1.5 data7 data9 data12
5
data10 data13
a
2 data8 data8 data10 data13

3
10 data9 data11 10 data9 data11

0
data10 data12 0 data12
data10
u
1 1 1 1 2
data131
3
1 4 1 51 2 3 4 5
10 data11 1
0 0 0.5 0.5
0 1 0.51.5 0 1 2 100 0.5
1.5 2
1.5 0.50 1 data11
2 1.5data13
10.5 1.51 2 21.5 2
data12 data12
01.5
4
0
Jo
data13
Contour first
data13
2
Contour first
3 1 2 3 3 4 51 2 3 4 5
Bulk first
Bulk first
0 0
data10
data11
data12
data13
data10
data11
data12
data13
2.8 2.8
1 (BF) 2 3 3 4 5 1First (BF) 2 3

data3
data4
data5
data6
data7
data8
data9
First (CF) First

3
First (CF)
data3
data4
data5
data6
data7
data8
data9
3 4 5
3
2.6
Thinïwalls (TW) Ultraïthin wall 1 (UTW) 2.6
12.8 WallFirst
Single 1.5
Tracks (SWT)
(CF) First
0 2 0.5 1Single Wall
2.8 Tracks 1.5
(SWT) 2
x (BF) First (CF) First (BF)
0 3
Contour Meltïpool Bulk y z Contour Meltïpool Bulk x y z

0 3
2.4
Thinïwalls (TW) Ultraïthin wall (UTW) 2.8 Thinïwalls (TW) Ultraïthin wall (UTW) 2.8
1
2.4
1
2.6 2.6
!""#$%# !""#$%#
40
30
20
10
Single Wallpath
Tracks
(top(SWT) Single Laser
Wall Tracks (SWT)
40
30
20
10
Crossïsection Laser view) Crossïsection path (top view)

2.2
Laser Contour
path (top view)Meltïpool Bulk x y z 2.2 Contour
Laser path Meltïpool Bulk x y z
(top view)
2.4
Powder denudation
Face 2 First contour Offset Hatching 2.6 Ra [µm] (a) (b) (c) (d) 2.4
Powder denudation
Face 2 First contour Offset Hatching 2.6 R [µm] (a) (b) (c) (d)
a
2
Second Crossïsection
contour Single trackLaser path (top view) 2 Crossïsection
Second contour Laser path (top view)
Single track
2.2
Laser
Layer n Layer n+1path (top view) 2.2
Layer n Laser
Layer path
n+1 (top view)
200 µm 300Face
µm 2 First
600 µm contour Offset Hatching Face 2 600
Firstµm
200 µm 300
1.8
Figure 6. 2.4
(a) Optical microscopy images of cross-sections in the x-z plane and (b) x-y µm
2
1.8
2.4
2
2
Second contour Single track 2
Layer n Layer n+1 Layer n Layer n+1
1.6
1.8 200 µm 300 µm 600 µm plane corresponding to different building strategies. The dashed black arrows indicate the
1.6
1.8 200 µm 300 µm 600 µm
1.4
1.6
orientation2.2
of contour melt pools. The heat affected
2.2 zone of the respective melt pools are
1.4
1.6
1.2
highlighted in black. The full black arrows indicate examples of disturbed melt pools. The
1.2
1.4
white arrows2indicate porosities. 1.4
1
2
1
40
30
20
10
1
40
30
20
10
1
0 0.2 1.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
1.2
Ra [µm] Ra [µm]
1
0 0.2 0.4 0.6 0.8 1 1.2 1.8
1.4 1.6 1.8 2
1
0 0.2 0.4 0.6 0.8 1.8
1 1.2 1.4 1.6 1.8 2
1.6 1.6
1.4 1.4
1.2 1.2
1 1
0 0.5 1 0 1.5 0.5 2 1 1.5
14
70
offset=120 µm 70
60 5 offset=120 µm
5 3
Bulk 60
first strategy
50 3
Bulk first strategy 3 Contour first strategy
1 Contour first strategy 1 2.8 50
4
2.8
40 2.8
4.5 30 µm
density
density
3 2.6
0.99 30 µm 0.99 40 90 µm
4.5 density 2.6
3 data2
density
2.6
30 2.8 data2
3.8
3.8
2.4
0.98 data2 0.98 30 data3 data3
5 2.4
90 µm
2.4
2.8
data3
local
Face 1ïside
local
2.6 2.2
20 120 µm
meanlocal
4 data2
meanlocal
2.2 0.97 2.2
2.6 0.97 20
3.6
3.6
4
2.4
5 data3 2 data2
10 0.96 2.42
0.96 Face 1ïside
Relative mean
10data3
Relative mean
4.5 2
2.2 1.8
3
3.4
1.8
3.4
2.2
0 1.8
20.95 0.95
100 150 200 250
Relative
300 1.6
Relative
3 4.5 3.5 1.6 0
1.6
2
100 150 200 250 300
(BF)3.5 0.94
2.8 1.8
First4First 0.94 1.4
3.2
First First
(CF) (CF)
3.2
3
(BF) 1.4
2.8 1.4
1.8
Thinïwalls (TW)(TW)
Thinïwalls Ultraïthin wall (UTW)
Ultraïthin wall (UTW)1.6 0.93 32.6 120 µm 1.61.2 0.93 3 1.2
2.6
Single
2.8 Wall Wall
Single Tracks (SWT)
Tracks (SWT)
1.2
1.4 4 data2
3
Contour First
3.5
Meltïpool (CF) First
Bulk Bulk(BF)
x1 y zx y z 0.92 3 0.92
1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Contour Meltïpool 0 1 1001.2
2.4
200 0 1.8 4002 500
300 0.5 600
data3
1
1.4 0 0.2
0 10.4 100
0.6 0.8
1.51 300 1.2 4001.4 2500
1.6 1.8
600 2
Thinïwalls3 (TW) Ultraïthin
0.4 wall (UTW) 2.8 200
12.8
2.4 0.61.2
Powder denudation
2.6 Powder denudation 0
0 0.2
0.5
0.8 1.4 1.6
1.5
Crossïsection [µ m] 2 Crossïsection [µ m]
Single Wallpath
Tracks (SWT) 2.2
2.8
2.8
1.2
Crossïsection Laser (top
Crossïsection Laser path (top view) view) 1
0 0.2 0.4 0.6 3.5 0.8 1 1.2 1.4 1.6 1.8 2
2.2 Laser path Contour
(top3 view)
2.4 Laser path
(top view) 2
4 (a) (b) (c) (d) 0 2.60.4 0.6 0.8 (a) 1 (b)1.2 (c)1.4 (d)1.6 1.8
1
FaceFace
2 First 2
Powder
contour
2 First
denudation
Offset
contour
2.5Hatching
Offset Hatching
3 2.61.8
3.5 0.2 2
2.6
2.6
Second Crossïsection
contour Single track Laser path (top view)
2.2 Second contour Single track
2
Laser path (top view) 3
of
LayerLayer
n Layer n+1 1.6
2 2.5
n Layer n+1 Figure 7. Local mean 2.4 density of some TW3 samples 2.43.5following 4
the x-direction, with (a) a BF
2.4
2.4
200 µm Face 6002 First contour Offset Hatching
1.8 200300
µm µm
300 µm µm 600 µm 1.4
2
Second contour strategy Single track and (b) a CF strategy. Reconstructed volumes for an offset of 120 µm show a huge
Layer n Layer n+1 1.2
2.2 (depicted in grey). Optical 2.2
2.2
2.2
1.6
1.8 200 µm 300 µm 600 amountµm of porosities 1
0
microscopy images show a cross-section in
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
ro
the x-z plane of TW samples built with an offset of 120 µm.
2
1.4
2
2 2
5
4.8
4.6
4.4
4.2
3.8
3.6
3.4
3.2
4.8
4.6
4.4
4.2
3.8
3.6
3.4
3.2
3
1.6
1.2
1.4
1.8 1.8
1 0.2
the initial temperature (T0 ) of the second contour is slightly higher than the initial
0 0 0.20.4
1.2
1
0
0.40.6
0.2
0.60.8
0.4
0.8 1
0.6
1 1.2
0.8
1.21.4
1
1.41.6
temperature of the
1.2
1.61.8
1.4
1.6
1.6 first processed contour, leading
1.8 2 2
to a slightly larger melt pool. This

1.6 1.8
effect is limited in the CF strategy as there is no solidified material between the con-
1.4 1.4
tours, and thus less heat conduction towards the second processed contour.
2 -p
re
1.2 1.2
The offset and the building strategies have a clear impact on the level of rough-
1 the impact of the TED level 1
ness. Figure 8 shows 0 0.5 10 value chosen 0.5for the contours
1.5 21 on Ra 1.5 2
lP
for three offsets and in the case of both strategies. In the case of a BF strategy (Figure
8.a), the level of TED and the offset have a first order effect on the level of roughness.
Ra increases with TED for low TED values then decreases for high TED values, for
the two smaller offsets (0 and 60 µm). Ra slightly increases then decreases with TED
a
for the 120 µm-offset. The highest level of roughness is obtained with an offset of 0 µm.
rn
For the CF strategy (Figure 8.b), only the level of TED has a first order effect.
Indeed, Ra always decreases with TED. The effect of the offset on Ra is rather minor
except at low TED level.
u
The highest level of roughness is observed with an offset of 0 µm, as it is the

case with the BF strategy. This offset is a particular case since it means that there
Jo
is no distance between contours and the bulk, i.e. the bulk (respectively contours)
re-melts contours (respectively the bulk) in CF (respectively BF) strategy. The laser
parameters selected for the bulk could then also have an influence on Ra and have
been investigated.
3.4. Thin walls: effect of bulk parameters

In order to study the effect of the bulk scanning parameters on the level of roughness,
samples were processed with fixed contour parameters with a TED level of 194 J/m
(232.6 W with a scan speed of 1200 mm/s) following both strategies and with vari-
able bulk TED levels for three offset conditions. Figure 9 presents the evolution of
2
2
5 15
5 Bulk first strategy
3
Bulk first strategy Contour first strategy
40 3 40
3 Contour first strategy 3
4040 4.5 40
2.5 40
1.5
1.5
4.5 Ra00[µm]
µm
µm 40 0 µm 0 µm
2.5 30 40 2.5µm
30
30 3
µm 30
2.5 30 µm
R6060
0 [µm]
µm
µm Ra030 µm
[µm]
µm
3 3030 a µm 30 30 60 µm
30 µm R60
a
[µm]
µm
2
Ra [[µm]
Ra [[µm]
90
90
30 µm
µm
m]
m]
µm 90 µm
µm 90 µm
4 data5
data5
60 2.5
µm 30 60
data5 data5
2 20 30 2 20
Ra [µm]
4
Ra [µm]
120 µm 90 µm
R [µm]
120 2
4 [µm]
90 µm µm R
120 [µm]
µm 120 µm
data7 data5
a
2.5 2020 data7
data5 20 20 data7 data7
[µm]
1.5 data8 120 µm
R [µm]
R
120[µm]
data8 µm
4a
data8
a
data8
1
a
data9 data7
R
20
4
4
4
1.5 10 20 data9
data72
1.5 data9 data9
3.5 data10 10
Ra
Wmm/s
mm/s
mm/s
mm/s
Wmm/s
mm/s
mm/s
1400Wmm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
1400 mm/s
mm/s
mm/s
3000 mm/s
1.5 data8
mm/s
mm/s
1025 mm/s
mm/s
1200 mm/s
mm/s
mm/s
mm/s
1700 mm/s
mm/s
mm/s
3000 mm/s
mm/s
1700mm/s
2200mm/s
mm/s
mm/s
3000 mm/s
700 mm/s
233 mm/s
273mm/s
mm/s
800 mm/s
mm/s
mm/s
mm/s
mm/s
27.4 W a
data8
data10 data10
mm/s
273mm/s
2200mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
10 data11 data10
3.5 1032 data9
82.1 W
W
W
W
W
W
WW
data9 10 3
273 W
W
W
W
W data11 10 data11
137 W
178 W
82.1W
W
W
W
W
W
W
W
3000W
1 data11
27.4W
W
1025W
273W
W
W
3.5
3.5
3.5
data12
3.5
3.5
3.5
data10
1200
1400
1025
1700
1200
3000
1700
2200
1025
1200
1025
1200
1400
1700
1400
2200
3000
2200
27.4
27.4
82.1
273
1025
1200
1400
1700
2200
0 0.5 1 data12
data10 1.5 2
27.4
82.1
82.1
233
137
178
800
900
700
900
137
178
233
700
800
900
27.4
82.1
273
data12
137
178
137
233
178
233
700
800
700
900
800
900
data12
137
178
233
700
800
900
data13 10 data11
1 0 10 11.5
data11
data13 0 1 data13 1.5 data13
0 0 0.5
0
1 data12 1.5 10 0.5 2 0.5
data12
1
2
1.5 0 2.8 100 0
2.8 data13100 2001.5 300 2 400
32 200 data13 300 400
3
3
9
0 1 3 40 50
3
3
3
2 0 TED
TED
Contour [J/m]
[J/m] 0.54 1 1 3TED
1.5
Contour [J/m][J/m]
4 TED 2
0.5
0.5
3 1 3 512 23 45 5
0 0.5 1 8.8 0 1.5 [J/m] 10
Bulk TED 2 0 1 0.5 Bulk
1.54TED [J/m] 52
2.6 (a) (b)3 (c) (d) 52 1 3 (a) (b) (c) (d)
2.5
2.5
2.5
2.6
2.5
2.5
1 2 4
2.5
8.6
1
0 0.5 1 1.5 2
of
8.4
Figure 8.2.4Evolution of Ra as a function of contour
2.4 TED level and strategies, for three
2
2
2
2
2
8.2
different offsets at fixed contour scan speed (800 mm/s). (a) Bulk first strategy, (b) contour
first strategy.
31.5
1.5
21.5
8
1.5
1.5
0
1.5
0
2.2 2.2
12
ro
3
2.5
1.5
2.5
1.5
1
7.8
3
40 2
50 1
40 1
1
7.6
1
201
1
2 Bulk first
2
80
70
80
60
70
50
60
40
50
80
30
40
70
20
30
60
10
20
10
30
10
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
80
70
60
50
40
30
20
10
Contour first
40 35 7.4 RZ [µm]
RZ [µm] RZ [µm]
RZ [µm] RZ [µm]
RZ [µm]
2.5
Ra0[µm]µm
7.2
1.5
1.8 1.8
-p
1.5
30 30 µm
3 30 60 µm 7
1.4 21.6 1.8
40 2 2.2 1 2.4 1.2 2.6 1.4 2.8 1.6 3 1.8 2 2.2 2.4 2.6 2.8 3
40 25 90 µm1
[ m]
80
1.6 1.6
Ra [µm]
80 data5
00 µm
µm27.4 W
R [µm]Ra [µm]
re
3 70 2080 120 µm
30 82.1 W 27.4 W
2.5 70 30 µmµm 80W
137 82.1 W 27.4 W
1.5 20 data7
60
a
30
60 70 R60 µm
[µm]
µm 178 W2 13780W 82.1 W 27.4 W
30 0.5
60
1560
data8
a
90
90 µm
data9
µm1.4
233
70W
273 W
178 W
23370W
137 W
178 W
82.1 W
137 W
27.4 W0 µm
82.1 W 1.4
1
50 data5 137 W 60 µm 120 µm

2.5 data5700 2
60mm/s 273 W 233 W 178 W
R [µm]
50 data10
[µm]
12 R [µm] 700 mm/s 273 W 233 W 178 W

RRaa[µm]
120 800 mm/s

µm 0 µm
R [µm]
50 120 60
10
Z
0 40
10 0.5 1 data7 2a µm
data11 1.5 800 mm/s 700
900 mm/s 2 mm/s 273 W 233 W
RZ [µm]
0 1.5
Z
50
1.2
lP
20
20
40
data7 1025
0 mm/s 900
250 mm/s 800
4 mm/s 6 700 mm/s8 0 µm 273 W60 µm 120
10 12 µm 1.2
RZ [µm]
Z
30 40 data12
data8 1025 mm/s 900 mm/s 800 mm/s 700 mm/s
data81200 mm/s
40 mm/s 900 mm/s 60 µm 120 µm
µm
2 30 5 data13
data9
1400
1.5
1200 mm/s 1025 mm/s
40
800 mm/s
1.5 20 30 data91700 mm/s 1400 mm/s 1200 mm/s 1025 mm/s 900 mm/s
20 data10 30 mm/s 1700 mm/s 1400 mm/s 1200 mm/s 1025 mm/s
1.5data10 2200
0
10
10
10 11 2 0
20
data11
3
11
3000 mm/s
4
30 mm/s 1700 mm/s
2200
5
1400 mm/s 1200 mm/s 1
1.5 10 2 2.5 data11
3 20
3.5 0 4 3000 mm/s 2200 mm/s 1700 mm/s
0.51400 mm/s
10 1.5
0.5 21 1.5
0.5
1.5 1 1.510 125

2 data12
data12 2.5 150
3 3.5 175
20 Contour 200
TED [J/m]
4 3000 mm/s 225
2200 mm/s 250
1700 mm/s275
1 1 1.5 10 21
data13 2.5 3 3.5 4 3000 mm/s 2200 mm/s
0 0.5 1 data13 1 1.51.5 10 Bulk 2 TED 2 [j/m]
TED
2.5 [J/m]
3 3.5 4 3000 mm/s
1
a
0.5 1 1.5 2 2.5 3 3.5 4

00
1 11 2 3 4 5
2 1 9. 3 Effect 4 2 5 scanning parameters on R for three offsets as a function of bulk
0 0.5 Figure 0.5
1.5 of the bulk a
0.5
rn
TED level for0 (in black) a bulk first strategy and for (in red) a contour first strategy. Bulk
0 2 4 6 8 10 12
TED level is varied by changing the scanning speed only (indicated on the graph). Contour
0
3
2.5
1.5
0
parameters
0 are0 fixed2 at 854 % of 6Pmax 8and 1200
10 12
mm/s.
0 2 4 6 8 10 12
u
Ra as a function of the bulk TED level for different offsets and strategies. The level
Jo
of TED was increased by decreasing the scanning speed from 1700 mm/s to 800 mm/s.
1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3
With a BF strategy, when there is no offset between bulk tracks and contour
tracks, Ra is large (∼ 32 µm), whatever the bulk TED value (Figure 9.a). In the case
of an offset, Ra slightly increases with the value of the bulk TED level. In the case of
a CF strategy, the offset has a limited effect on Ra .
A small increase of Ra is observed for a 0 µm-offset when the bulk TED value
becomes higher than the contour TED value (Figure 9). Similarly to the analysis of
the influence of the contour parameters, the offset and bulk TED level are first order
parameters for the BF strategy while only the bulk TED level has a significant effect
with a CF strategy.
16
3.5. Trade-off between hardness and roughness

The lowest level of roughness measured on TW samples was 5.9 µm (for a CF strat-
egy, a 90 µm-offset and a TED value of 341 J/m). This level of Ra is higher than the
lowest Ra level obtained with SWT samples. Even tough parameters leading to the
lowest level of Ra for SWT samples are out of the process windows, a set of parameters
corresponding to keyhole conditions (246 W at 300 mm/s, giving a TED level of 820
J/m) was tested on TW samples contours with both strategies and for a cube with CF
strategy. Figure 10 presents the measured roughness and hardness of those samples.
SWT sample corresponding to the same level of TED is also shown as a reference.
Hardness was measured both on contours and on the bulk hatching. The TW-CF and
cube-CF samples present the lowest Ra values measured in the present work, i.e. 4.2
of
µm and 3.9 µm, respectively (Figure 10.b). The TW-BF sample presents a Ra level
of 8.6 µm. This is twice as rough as for the TW-CF sample, but it is still the best
level of roughness in the case of the BF strategy.
ro
As expected from the previous observations on SWT samples, there is a substan-
tial hardness shift between the bulk and the contours for the TW samples (in the case
-p
of both strategies) and for the cube (Figure 10.b) due to different process parameters,
leading to a change in the microstructure (See supplementary materials). Hardness
measured on TW samples and cube bulk hatching are similar (136 HV). The lowest
re
contour hardness is measured on the TW-CF sample (119 HV) but it is still higher
than the hardness of the SWT sample (109 HV), while contour hardness of TW-BF
sample and cube are very similar (∼126 HV). Figure 11 presents representative opti-
lP
cal microscopy images of each sample. They clearly show that the process parameters
lead to a keyhole mode, leaving a lot of large porosities in the middle of the melt pools.
It also shows that the width of the TW samples is larger than the fixed width as the
melt pools are very large in keyhole mode (Figure 10). This very high energy mode
limits the porosities coming from a bad overlap between the bulk and the contours
a
but brings new kind of porosities due to melt pool instability. The different level of
hardness between contours and bulk resulting from the different microstructures might
rn
be a problem for the global mechanical behaviour, especially in thin elements and will
need further investigation.
u
Surprisingly, conditions bringing keyholes are thus the solution to reach a low
level of roughness. However, the resulting sub-surface porosities can be an issue for
Jo
the mechanical performances of the samples. Figure 12 shows optical microscopy im-
ages of TW samples processed with an increasing level of TED. These micrographs
show a drastic increase of the porosities due to the formation of keyholes when TED
level increases. This leads to a trade-off between the level of roughness and the amount
of sub-surface porosities. In this case, TW sample with contour TED of 684 J/m is
a better option than TW sample with contour TED of 773 J/m, even though Ra is
higher.
10 150
10 150
9 145
9 145
8 40 40 140
40 40 Bulk first Bulk first
8 40 140 40
7Bulk first Bulk first 135 Contour
Bulk first firstBulk Contour first first
40 4040 40 40 40
data3
mm/s Ra [ m]Hardness [HV]

7 Contour
Bulk first 40 first Bulk firstContour first data3
135 Contour first Contour first
40 6 40 40 40 130
40
data3 40 data3
30 40first 40 40 30 40 40 data4
data3data4
Hardness [HV]
first0 µm Contour data3
R [ m]
Contour
6 4040 4040 40 4000 µm 40 130
30 5 30 30 40 data330data4 4030
0 µm
data3
30 40 data4
µm
3030 40 0 µm
40 data4 0data5
µm
30µm 0 µm data4 0data5
µm
17
Ra [ m]
30 µm
µm 30
30 µm
µm 30 µm 125 30 30 µm 30 µm
00 µm 0 µm data6
data5 40 data5 0 µm 60 µm 40 0 µm 060µm µm data6 0 µm 060µm
a
µm0data5 µm0data5
µm 60 µm
µmdata4 µm µm 60
5 30 30 40 3030 30data4 303060 0 µm 60
0 30
30 4030 µm 30 30 60 µm µm
R [µm]
R [µm]
30 µm
µm 90 90 µm
30
3030 µmµm 90 90 µm
µm
30 3030125µm 30 µm 30 µm 30 µm
4 30 µm 30 30 µm data730 µm data7
R [ m]
Ra [ m]
Ra [ m]
µm
4 [µm]R [ m]
µm 120 90 90
data5data6 5 data6 0 µm 60 µmdata6
90 µm µm0 µm 60 µmdata6
90 µm µm
60
60 µm
µm data5 data5 60
60 µm
µm data5
30 10 20
3030 3030 0 µm 0 µm
data5
5 30 30data5 30 data5
60 µm 30 60 µm data5 µm150
60 µm 60 µm data5
[µm]
µm
R [µm]
60 µm 3030
Ra [µm]
R [µm]
60 µm 30
20 30 20 20 20 data8
20 30 data5 data8
Ra [µm]
R [µm]
90 90
90 µm
Ra [µm]
90 µm 90
Ra [µm]
µm90
µm data7
4 data6data7
µm
30 µm120 data7 30 µm120 µm90 data7 µm µmBulk
120 µmhatching
[ m]
Ra [ m]
120
mm/s R [ m]
120 µm µm 120
mm/s 4Ra [µm]

90
4 Ra [µm]
data6 µm 120
4 [µm]
µm 90 µm
a a [µm]
µm 90 µm
data5 data5 Bulk hatching
90 µm 90 µm 120 µm µm
60 µm 120 µm
3 30 60
W R [µm] a
a
data5 data5 115
a
data5
a
data5 data7 30 data7 data5
Ra [µm]
60 µmdata7
Ra [µm]
20 6020 µm 20 data5 data9 data9

20 20 20
30 20 209data7 30
data8 data7
µmdata820
data5
data7data8 2020
data5 20 data5
µmdata8
20 20 data7 data5
µmdata8 Contours
data7
[µm]
data7
Ra [µm]
20 data7
20120
[µm]
90 µm145
[µm]
R R[µm]
90 µm
120 120 µmdata8 Contours
R [µm]
R [µm]
120 120 120
a
µm
4 [µm]
µm
WR [µm]
90 µm
R [µm]
90 µm data8
120 µm data8
120 µm 120 µm data8 120 µmdata8 120 µm data8 120 µmdata8
mm/s 4aa
3
a
a
115 data10 data9data10
a
data5 data5
a
data7 data7
20 2020data7 data9 data5 data9 20 data7 data9 data7 data7
R
20 data9
44R R
data9 data7
4R R
20
4 a 44
3.5mm/s 4
2201020
4a 4
mm/s 4
20 data5 20 data7 data9 data7 data9
4 4
data7data8
4a 4
data8 data820data9
4
20data9 data7 20 data7 110 data9
[µm]
[µm]
20 data10 10 120 µmdata8 data9 10 120 µmdata8
a
10
a
data8 [µm]
R [µm]
data10
a
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
1200 mm/s
mm/s
mm/s
2200 mm/s
3.53000 mm/s
mm/s
mm/s
mm/s
Wmm/s
mm/s
mm/s
amm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
amm/s
mm/s
mm/s
data8 data8 a data11 data11
mm/s
mm/s
3.5mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/smm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
8 120 µm 4.5 120 µm R
3.5mm/s
1700mm/s
3.5mm/s
1700 mm/s
3000 mm/s
mm/s
mm/s
mm/s
Wmm/s
mm/s
mm/s
data8 data8140 data8
data10
a
data8 data10 data10
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
a
mm/s
233 mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
data8
27.4 W a
data8
data10 data10 data10 data10 data10
data9data10
3.5mm/s
Rmm/s
mm/s
mm/s
mm/s
900 mm/s
mm/s
Rmm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
data10
mm/s
mm/s
900mm/s
data9 4.52010
data9data9
data9 data7 data7
R
10data11 20
R
10 data7 data11
4
2 10 data9 data9 data9 data9
4
20 10 20data9
data7 data9 10 10 110 data9 data11
WW
WW
W
W
W
W
W
82.1 W
W
W
W
W
W
10 data11 10 10
data9 data9 10 data12
W
W
W
W
W
data1110
W
W
W
W
10 10data10data11 data12
W
W
W
W
W
W
105 data11
W
W
W
W
178 W
82.1W
273W
RaW
W
W
W
W
W
273W
data11
W
W
W
W
W
W
273W
W
data10
W
W
1
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
data8
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
data8
W
3.5
3.5
W
W
W
3.5
3.5
data8 data12 data11 data11
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
data8 data12
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
3.5
mm/s
mm/s
3.5mm/s
mm/s
mm/s
3000 mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
3.5
data10data10
3.5
data11 data10data10
1025
1200
1400
1025
1700
1200
2200
1400
3000
1700
2200
3000
1025
1400
data11
1025
1200
1400
1025
1700
1200
2200
1400
3000
1700
2200
3000
1025
1200
1400
1700
2200
3000
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
mm/s
178 mm/s
mm/s
273mm/s
mm/s
mm/s
273 mm/s
mm/s
mm/s
mm/s
27.4 W a
data10
1025
1200
1025
1400
1200
1700
1400
2200
1700
3000
2200
3000
data10
1025
1200
1025
1400
1200
1700
1400
2200
1700
3000
2200
3000
data10
27.4
82.1
27.4
82.1
273
27.4
82.1
273
data10
322001025
30001200
3.51400
3 2200
27.4
27.4
82.1
data10
273
1025
1200
1400
12001700
2200
3000
data10
3.5mm/s
1200mm/s
mm/s
mm/s
3000mm/s
mm/s
27.4mm/s
273mm/s
2200mm/s
mm/s
273mm/s
mm/s
6
mm/s
31025mm/s
mm/s
6
mm/s
800mm/s
273mm/s
27.4
82.1
27.4
82.1
273
273
27.4
82.1
27.4
82.1
273
273
data10
137
data10
178
233
137
178
700
233
800
900
700
800
137
178
233
700
800
900
27.4
82.1
273
137
233
137
178
700
W800
900
700
800
900
137
178
233
700
800
900
27.4
82.1
273
10 10 data11data12 data11data12 data12 data12
137
178
137
233
178
233
700
800
700
900
800
137
178
137
233
178
233
700
800
700
900
800
900
data12 data12
137
178
233
3.5700
3000800
1700900
137
178
233
700
273 800
900
7 data13 data13 data9 data9 135
10 10 data12 10 10 10 data11 10 data11 data13 data11 data136
W
W
W
W
W
W
W
82.1 W
W
W
W
W
W
W
1040 0 10 10 Bulk
10 40 10 data11 data9 data9 10data13 10 data11
W
W
W
W
W
W
W
W
W
W
data11 data11
6data12
W
W
W
W
W
W
W
data12
W
W
3000W
137 W
27.4W
233 W
W
W
W
W
W
W
W
W
W
6 10data13 40105 data11data13 40
W
900 W
W
W
W
W
800 W
W
137 W
82.1W
137 W
1 6 data11 data11
W
W
W
3.5
3.5
W
W
233W
3.5
3.5
3.5
data13 data13
first 0data12
3.5
first data12 0 data10 0 data10
3.5
data11Bulk
1025
1200
1400
1025
1700
2200
1400
1700
2200
1200
1400
data11
1025
1200
1400
1025
1700
1200
1400
3000
1700
2200
3000
1025
1200
1400
1700
2200
3000
1025
1200
1025
1400
1700
33 1400
2200
1700
3000
32200
1025
1200
1025
1400
1200
31700
1400
2200
1700
3000
32200
3000
27.4
82.1
273
27.4
82.1
273
2.5 3 31025
1200
3 1400
1700
1.5 1.5 2 2 2.5 2.5 3 32200
Ra [ m] 82.1
273
27.4
82.1
273
2 2.5 2.5 3 1025

1200
3 1400
3 1700
1.5 1.5a 2 2 2.5 2.5 3 32200
3000
6
Contour = C 0 data12 data12 Bulk first
6
Bulk first
27.4
82.1
27.4
82.1
Contour = data12
C100
27.4
82.1
27.4
82.1
data100data12 data10
178
233
137
178
700
800
700
900
137
178
233
700
800
900
27.4
82.1
273
178
137
700
233
900
700
800
900
137
178
233
700
800
900
27.4
82.1
273
[HV]
137
178
137
233
178
233
700
800
700
900
800
900
137
178
233
178
233
700
800
700
900
800
3 900
40 data12 data12
137
3178
233
700
800
900
40
137
178
233
1.5 1.5 2 2 2.5 2.5 3 700

800
900
5.5
0 SWT6 9 40 0 100 SW 0TW data13
200 6100 10
300 data13
200 400 10
300 0 Bulkdata13400 100 0 200 100 300 200 400 6 300
5.5
400
40data11 data13 data11 data13
1040 0 10 10 Bulk Contour first 210data13
6data13 1SW TW data13
Cube 0 Contour
2Contour first 0 50Bulk SWT 6data13 40 SW 6 TW Cube Contour Bulk
Bulk = B Bulk first Contour first SW TW Cube Contour first
Contour Bulk
3
9
1 0
0 first 6data12 Cube
data11
9
0 data13 3Contour
data12 Bulkdata11
4data13 3 4 0 = B data130 50 TED data13 data13 Bulk first
3
data13 data13 130
5.5 Contour = C 0
40 0 1 5.5
data3 Contour
246 TED
29 W 5.5
0 TED
(90%[J/m]
data12 P [J/m]
Contour
61001 max 246
3 ) W
ï Bulk
300
first 4
Contour
44SW
(90%first
TED
mm/s
2TW P
data3
4 [J/m]
TED )
TED
data12
1 ï
Cube 300 [J/m]
300
= 820 1
mm/s
J/m
2
3Contour
data12
Contour
TED
Bulk
0 5 Bulkdata13
==C100
1 2
820
40100 3 J/m Contour
40
0SWT12 34
data12 [J/m]
Bulk
TED
5 200data13100 3002 first
3 5.5
[J/m]
45
data3 Contour
5.5
246
200 3TED
4W5 (90%
400 [J/m]
5.5
TED P[J/m]
45246
6 Contour
300
Bulk
) first
ïW300 first
(90% 5data3
SWmm/s
TW
400Pmax TED
)ï=
Cube 820
300 J/m TED
mm/s
Contour = 820 J/m
Bulk
0 SWT 9 Contour
100 0TW200 300 200 400 400
max
0 max
5
0
5
Bulk = B 0 5.5 0 data4data13
first Bulk SW
first TW data13
Cube Contour Contour0
first Bulk Contour SW
first TW Cube SWContour
TW Cube Bulk
Contour first
Contour
SW Cube = B
(90% PBulk
8.8Contour Bulk Bulk first
8.8
1 5.5 2 3 4 5
(a)! 0
(b)! 5.5
8.8
1 2 3 5
(a)! 0 0 mm/s
(b)!
Hardness
0 data13 Bulk 0 TED [J/m]
3
30 Bulk
30
TED 5.5 TED
[J/m] [J/m]0
data3
TED 246 W
[J/m]
data13
(90%
data4 1 Pmax ) ï 300
2 30 TED
1 3 = 820Bulk
TEDJ/m
2 TED
[J/m]
4 5.5[J/m]
3 data4
5 Bulk
TED 30 TED
[J/m]
5.54 [J/m]
data3 5 246 W data4 ) ï 300 mm/s TED = 820 J/m
data3 Contour TED 1P[J/m]
246 WContour
(90% P4 TED [J/m]
2)data5
ï 4300 51
mm/s 2 3 TED 2314J/m Contour
=21820 512 34 53 TED 23[J/m]
4125 Contour
246 34 W TED
5 (90% [J/m]
45246
Pmax )5ïW300 SWmm/s 5data3TED ï= 820 max
J/m TED
2.5
2.5
2 3 4 5
2.5
301 8.6 1data5
246 W (90%
5 52 2Contour )first
3data4ï1300SW
3 2 mm/s
data3
TW 1TED
Cube 1 = 3820
ContourJ/m 5Bulk
40 5
data3 4 Contour first (90% TWPmax )data5
Cube 300 mm/s
Contour = 820 J/m
Bulk
2.5
2.5
2.5
1.5 2 22 2.5 2.5
5
5 max
0 5
data5 5 data4
0
5
8.6 8.8 max
5.5 8.6 0 30
(a)! 5.5
(b)!
8.8
(a)! (b)!
4.5 4.5
30 0 0 data4 30
0 Bulk
data6TED [J/m] 0
data3 Bulk
246 TEDW (90%
data4 [J/m]
Pmax ) ï 300 mm/s 30TED = 820Bulk J/m 4TED [J/m] 5 Bulk
3 data4 30 4TED [J/m] data3 246 W (90%data6
data4 P ) ï 300 mm/s TED = 820 J/m
1 3data55 2 14 1 data6 3 5 2 2 14 1 3 532 2 data6
5 54data5 5
5 max
2.5
2.5
54 3
2.5
2.5
2.5
30 8.62 1data5 2
2.5
2.5
Ra [µm]
Ra [µm]
2 [µm]
Ra [µm]
8.4 4.5
0 4 1 0 158.4 data74 35 8.6data4

data6 4.5
24.5 8.4data6 4data5 3 data7
3.5 data8
5 4 4
4.5
5 30 data5
120 4.5 data7 5 data4
4.55 data6
data5 data7
Ra [µm]
Ra [µm]
data5 5 data7 data6 data6 data5 data6 data8
2
1.5 a 2
20 data8 20 20 20
[µm] 2
2
R [µm]2
data8 data7
4 33.5
Ra [µm]
Ra [µm]
1.5 R [µm]
8.2
4 1 2 18.420 4.5 data7
8.2
3 24.5 8.4data68.2data8
data9
4.5 data7 data95 4 4 5 data7 data9
20 4.5 4.5 data6 data8
4.5 data7 data9
Ra [µm]
2R
2
2
20 data8 20 4 3 20 115 20
2
3.5
1.5 1.5 2
2
3.5
data8 data10
2
data10 data74 data9 data8 4
data10 data74 data9 data8 data10
4.5 4.5
1.5
1.5
8 8.2 8 8.2 8
Ra1.5
1.5
a
1.5
1.5 1.5
3.5
20 data9 data11 data8 data10 4 data9 data11 3.5
20 data9 data11 data8 data10 4 data9 data11
3 47.8
data10 4 data9 data10 3 4
data10 4 data9 4.8data105
3 10 3.2 3.4 3.6 8 3.8
3.5 2
4 10 4.2 4.4 7.8 4.6 4.8 5 3 10 10
1.5
3.2 3.4 3.6 3.8 4 4.2 4.4 4.6
7.8 data12 8
3.5 data12 3.5
1.5
data11 110 data12 3.5 data11 data12
1.5
1.5
1.5
1.5
10 data11data13 data104 3 data13
data11 10 data11data13 data10 4 data11
data12
3.5
3 3 3.5 6 4 4.5 3.5 data11data135 data12data13
data12
3.5
3 3
10 3.2 10 10 10
50 1
1 11
1
601
801
1
3 3.4 3.6 3.8 4 4.2 4.4 7.8 4.67.6 4.8 5 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5
7.6 7.8 data12
3.57.6 3.5 data11data13 data12 3.5
data12
1 1
10 201
601
3
1
6
3
80
70
80
60
70
50
30 40
50
20 30
40
70
30
10 10 20
10
40
30
20
1 10
70 80
70
80
60
70
80 50
60
70 40
50
80
60 30
40
70
50 20
30
60
40 10
20
50
10
40
30
20
10
33.4 3.5 44.4 4.2 44.8 4.65data13 105 4.833 4.5 53.2 data12 5
80
80
70
70
1 60
60
50
50
1 40
40
30
30
20
20
10
10
80
80 70 80
80 70
60
60
50
50
40
40
30
70 2030
30 20
10
20 10
80
70
60
50
201 40
10 30
10 20
10
60 80
50 70
801 60
40 50
601 40
50 1 30
4011 20
30 10
10 data13 33 1 3 10 3
3.2 data12
3.533.4 3.6 data13
3.2
RZ [µm] 3.8
RZZ [µm]
RZ [µm]
RZ [µm] 3.6
Z
RZR[µm]4 3.8
4.2
RZ [µm]
RZ [µm]
RZ [µm]
RZ [µm] 4.6 4.4 3.5
3.4
3 3.6 data13
3.2 3.8
3.4 4
3.6 4.2
3.8 4.4
4 4.6
4.2 4.8
4.4 5
4.6 4.8 5
1
7.4 7.6 RZ [µm]

40 7.6 37.4 40 7.4 33 0 40 40
[µm] [µm]
0 0 3 data13 R 0 3 data13 3
70
60
50
80
80
60
70
50
60
40
50
30
40
70
20
30
10
20
10
40
30
20
10
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
80
70
60
50
40
30
20
80
70
60
50
40
30
20
10
1 2 7.40 =7.2 3.23 3.4 3.6 3.8 5 1 4 4.2 2 4.4 3 4.6 4.8 5 2 53 3.4 3.23 3.4 3.6 3.8 5 4 4.2 4.4 4.6 4.8 5
Contour 3C 3 3.2 1 3 4
3.4 3.62Bulk
3.2
RZ [µm] 53.4 first 3.6
3.8
RZZ [µm]
Z
RZR[µm] 4 4.2 4 4
3.8 RZ [µm]
RZ [µm] 4.4Bulk 4.2 first Contour
4.6 4.4
4.8 4.65 C0 4.83 3 53.2
= 1 3.4
34 Bulk 3.6first
3.2 3.8 4
3.6 4.2 4Bulk
3.8 4 first4.2
4.4 4.6 4.8
4.4 5
4.6 4.8 5
7.2
0
5.5
40
0 0 7.4 RZ [µm]
6 33 3 7.2 RZ [µm]
RZ [µm]
3.236 4 3.4
R [µm]
2.8
[µm] RZ [µm]
RZ [µm]
6 0
5.5
40 100 0 26 5 3 6 3.2 4 3.4
3 6
7
1 2 0Bulk7.2= B
1
3 7 1 4
2
SWT
7.2
2 57 Contour first Bulk
5
43.6 first 3.8 5 1 4
Contour 4.2 Bulk
first2 =4.40B
1
4.6
3 4.8 1SWT 4 2 53
Contour 2.8 first
3 43.6
Bulk first
5 3.8 5 4
Contour first 4.2 4.4 4.6 4.8 5
.4 1 1.6 1.2 1.8 1.4 2 1.6
1 2.2 1.8 1 22.4 2 1.2 2.6 32.2 1.4 2.81 2.4 4 1.6 31.2 2.6 1.85 1.4 2.8 2 1.6 3 2.2 1.8 2.4 2 2.6 2.2
1 2.8 2.4
2 3 2.6
3 2.8
4 3 5
5
7
data3
SW TW Cube
7 SWContour
TW SW
Contour
Cube TWdata3
first Cube
5 Bulk
Contour Contour
Bulk Bulk data3
SW TW Cube SWContour
TW SW
Cube
Contour data3
first
TW Cube
ContourContour
Bulk Bulk Bulk
1.6 1.2 1.8 1.4 2 1.6 2.2 1.8 1 2.4 2
5.5
1.2 2.6 2.2
5.5W
1.4 2.81 2.4
1.6 31.2 2.6 5.5
1.8 1.4 2.8
(a) (b)
2 1.6 3
W 246 ) ï(c)
2.2 1.8
W P(90%(d)) P
2.4 2 2.6 2.2
) mm/s 2.8 2.4
ï= 300
820 mm/s
3 2.6
5.530
= TED = 2.8820 5.5
J/m(a)
3
5.5
(b) (c) ï(d)
246 W (90%) ïPTED )mm/s
ï 300TED
mm/s= 820
TEDJ/m
= 820 J/m
30 2.6 data4
30 246 (90% data3
Pmax
246 (90% 300data4mm/s TED
ï 300
max 30 max
J/m 2.6
TED 820 J/m data4
246 W (90% Pdata3
246 W )(90%
max
300 data4
Pmm/s
max
300
max = 820 J/m
30 data5 data4 data5 30 data5 data4 data5
of
4.5 4.5
5 data65 5 data5 data6 data65 5 data5 data6

Figure 10. 2.4 (a) Arithmetic data6 roughness and (b) hardness 2.4 5 measureddata6 on SWT and TW samples
Ra [µm]
Ra [µm]
Ra [µm]
5a [µm]
4
data7 data7 4
data7 data7
Ra [µm]
Ra [µm]
20 with CF 20 on20a cube with CF 20 data8 Contour

4.5and BF
data8 strategies data7 and
4.5 strategy.
data8 data7 TED data8level is 820 J/m
4.5 4.5 4.5 4.5
20
3.5
while bulk2.2hatching
data9
TED data8
level data9
is 194 J/m. 20
3.5
The2.2 offset was data9
90 µm data9
set todata8 R for the TW sample
ro
6 data10 data9 data10 data10 data9 data10
and the 4cube. data11 4 4 data10 data11 4 data114 4 data10 data11
103.826 data12 3.6 2
3 3
3
10 3.2 3.4 3.6 4 4.2
data11
4.4
6 104.8 3.2 5 3.4
4.6 3
data12 10
3.8 4
data12 4.2 4.4
data11
4.6
data12 4.8 5
5.5 Contour = C 10 data13 10

3.5 3.5 3.5 data12 data13 3.5 data13
3.5 3.5 data12 data13
Bulk = B SWT 1.8 SW TW Cube
data13 Contour SW Bulk TW Cube 1.8 Contour Bulk data13
5
5.5 246 3W (90% P5.5 ) ï 300 mm/s TED = P3820 J/m
-p
4
0 30 3 0 246 W (90% 0 ) 4.2

ï 3003 mm/s 3 TED = 820 J/m
Contour first
5 max
31.6
1 3 3.2 3 23.4433.2 3.6
3.23 53.8
3.4 3.4 43.6
3.6 1 4 3.8
3.8 4.2 4 4.44max
3 134.6
4.2 4.4 34.8
3.2 4.424 3.2
4.6
3.4 3 54.6 3.434.8
4.83.2
3.6 3.6 5 4 43.6
5 3.4
53.8 3.8 4.23.8
4 54.4 4.2
4 4.4 4.84.4
4.64.2 4.6 5 4.6
4.8 4.85 5
1 02 02 51.6
Bulk first
1 52 3 4 5 1 2 3 4 5
data10
data11
data12
data13
5
5
data3
data4
data5
data6
data7
data8
data10 Contour first data9
4.5
1.4 1.4
re
4
4.5
3
4 1.2 !""#$%# 4.5 1.2

data9 Bulk first
data10
data11
data12
data13
4
data11 data3
data12 data4
data13 data5
data6
data7
data8
2 data9
1 1
4
First3.5(CF) First (BF) 0 0.5 4 1 0 1.5 0.5 2 1 1.5 2

lP
Contour first
3
3.5
3
Bulk first

2.83
3.5
data3
data4
data5
data6
data7
data8
Contour 3 Meltïpool
3.2 Bulk
3.4 x y z3.6 3.8 4 4.2 4.4 4.6 4.8 5
Powder denudation 3
Thinïwalls (TW) Ultraïthin wall
3 (UTW)3.2 3.4 3.6 !""#$%#3.8 4 4.2 4.4 4.6 4.8 5
Crossïsection Laser path (top view) !""#$%#
3
2.6
!""#$%#
Single Wall Tracks (SWT) 3
a
Contour Meltïpool Bulk x y z 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5
2
Face2.4 2 First contour Offset Hatching

1
Powder denudation
40
30
20
10

rn
Laser path (topFigure 11. Optical microscopy images of SWT, TW and cube samples. Same samples as in
Layer
2.2 n Layer n+1 Ra [µm]
view)
200 µm 300 µm 600
2 µm
2
Face FigureOffset
First contour 10. Hatching
2
1
Layer n Layer n+1

40
30
20
10
0
u
1.8 200 µm 300 µm 600 µm Ra [µm]

3.6. Assessment of the formation of the vertical roughness
1
1.6
Jo 40
30
20
10
The vertical arithmetic roughness of L-PBF samples is a topographical measurement

1.4
Ra [µm]
of the side of the contour melt pools. Any parameter impacting the morphology of
0.4 0.6 0.8 1 1.2 the contour melt pools should then have an impact on the roughness. The analysis of
1.4 1.6 1.8 2
1.2
SWT samples showed that the lowest level of Ra is attained in or near unstable melt
1
0 0.2 0.4 0.6
pool track conditions, i.e. when TED level reaches the keyhole conditions (Figure 2).
0.8 1 1.2 1.4 1.6 1.8 2
The reason is geometric and is related to the radius of curvature of the melt pools.
This is illustrated in Figure 13, which presents schematically SWT morphologies for
different conditions. In the process window (Figure 13.a), Ra decreases with TED
(Figure 2) since the melt pool width increases and so does the radius of curvature.
When the melt pool becomes unstable, as it is the case in balling conditions (Figure
13.b), the side profile becomes also unstable, increasing Ra . In keyhole conditions,
i.e. at very high TED level, the radius of curvature of the melt pools is so large
3
18
2.8 Power [W]
TED = 684 J/m TED = 773 J/m TED = 912 J/m
R = 5.4 µm
2.6 a
Ra = 4.2 µm Ra = 4.3 µm
2.4
1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
2.2
3
2
3
1
3
0
0
0.2
0.2
1.8
200
Layer n Layer n+1
Powder denudation
Contour Meltïpool Bulk x y z!"
200 µm 300 µm 600 µm
Layer n Layer n+1
Powder denudation
Single
Thinïwalls
First
200 µm 300 µm 600 µm
Layer n Layer
Second
Powder denudation
Thinïwalls
First
1.6
3 3
0.4
0.4
3 3 3 3
µm 300(TW)
1.4
Power [W] 2.8 Power [W]
(CF)
(CF)
2.8
!"#$%&$ TED!"#$%&$
= 684 [W]
J/m TED = 773 !"#$%&$
Power
Power[W]
[W] 2.8 Power [W] Power TEDJ/m TED
= 684 J/m= 912
TEDJ/m
= 773 J/m TED = 912 J/m
Wall
2.82.8 2.8
1.22.6 Ra = 5.4 µm Ra J/m
= 5.4 µm
TED
TED==684684J/m
J/mTED TED
TED==773
773J/m
J/m = TED
684
TED J/m TED
==912
912J/m=TED
J/m 773=J/m
684 TED
2.6
J/m =TED
912= 773 J/m TED = 912 J/m
contour
RR ==5.4
5.4µm
µm2.6 Ra = 5.4 2.6
µm Ra =R4.2= µm
5.4 µmRa = 4.3 µm
Ra = 4.2 µm Ra = 4.3 µm
0.6 Tracks
0.6
2.62.6 aa a
First (BF)
First
12.4 2.4
RRa==4.2 µm RRa==4.3
4.2µm 4.3µmRa = 04.2 µm 0.2 Ra = 4.3
0.4R µm
(TW)n+1
µm = 4.2
0.6 µm 0.8Ra = 4.31 µm 1.2 1.4 1.6 1.8 2
a a a
µm 600
2.42.4 2.4 2.4
Figure 12. Optical microscopy images of TW samples with contours processed near or in
2.2 2.2
(BF)
0.8
0.8
2.22.2 2.2 2.2

keyhole conditions. Increasing the contour TED improves Ra but many sub-surface porosities
of
2 2
Single track
Ultraïthin
2 2
appear. 2 2
(SWT)
1.8 1.8
1.81.8 µm 1.8
1
1.8
1
1.6 1.6
ro
1.61.6 1.6 1.6
1.4 1.4
1.4
1.2
1.2
1.4 1.4 1.4

1.2 1.2
wall (UTW)
1.2
1.2 1.2 1.2
1 1
0 0.2 0.4 0.6 0 0.8 0.2 1 0.4 1.2 0.6 1.4 0.8 1.6 1 1.8 1.2 2 1.4 1.6 1.8 2
-p
1.4
1.4
1 0 0.2 0.4 0.6 1 0.8 1 1.2 1 1.4 1.6 1.8 2

0 0.2 0.4 0.6 0 0.8 0.2 1 0.4 1.2 0.6
0 1.4 0.8
0.2 1.6 1 1.8
0.4 1.2
0.6 2 1.4
0.8 1.6
1 1.8
1.2 2
1.4 1.6 1.8 2
1.6
1.6
re
3 3
3 3 3
1.8
1.8
2.8
2.8 2.8Process window
yholing Unstable parameters 2.8 2.8
Keyholing Unstable parameters Process window
Keyholing Unstable parameters Process window
lP
(a) (b) 2.6

(c) (d) (a) (b) 2.6
(c) (d) (a) (b) (c) (d)
2
2.6
2
2.6 2.6
2.4Figure
2.4
13. Schematic
2.4 representation of cross-sections and profiles of SWT samples for
2.4 2.4
(a) process parameters in the process window, (b) outside the process window (e.g. balling
a
2.2conditions)
2.2 and 2.2 2.2 2.2The dashed red curves is a schematic representation
(c) in keyhole conditions.
of the expected roughness.
2 2 2
rn
2 2
1.8 1.8 1.8

that1.8
the side profile appears flat, inducing
1.8 a very low level of roughness. However,
the
1.6 1.6 instability within
1.6 the melt pools
1.6 1.6causes the contour melt pool tracks to end up
u
with huge porosities, which might reduce mechanical properties, particularly fatigue
1.4 the subsequent
1.4life1.4 (5). Moreover, 1.4 coarsening of the microstructure due to a longer
Jo
1.4
time at high temperatures leads to a drastic hardness drop. Increasing the level of
1.2 1.2 1.2 1.2
TED above the keyhole conditions cancels 1.2 the benefit of large melt pools due to a
1progressive increase
1 of sintered powder
1 particles on the melt pool sides.
0.4 0.6 0.8 1 01.2 10 1.4 0.2 0.5 0
1.6 0.4 1.8 0.6 2 0.8 1 0.51 0 1
1.2
0 1.5
1.4
0.2
1 0.5
1.6
0.4 1.8
0.6 2 1.50.82 1 1 1.2
2 1.51.4 1.6 1.8
2 2
Another factor influencing the melt pool morphology has been highlighted with
UTW samples. The distance between two melt pools was found to be a key factor
inducing roughness. Figure 14 presents a schematic representation of the processing
layer in the case of a UTW sample and two optical microscopy images of real samples
showing the last processed layer. When the second contour is processed (Figure 14.c),
its melt pool is disturbed by the first contour already solidified and seems to point
outward. An asymmetric morphology between both tracks is in fact well observed
1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
1
3
0
0.2
200 µm 300 µm 600 µm

Layer n Layer n+1
Powder denudation
0.4
2
0.6
19
1.8
0.8
3 3 3
1.6
1

2.8 2.8 2.8
1.4
1.2
(a) (b) (c) (d)
(b) (c) (d) (a)2.6(b) (c) (d)
(a) 2.6

2.6
1.2
1.4

2.4 2.4 2.4
200 µm 300 µm 600 µm
1
1.6
Layer n Layer n+1

Powder denudation
2.2 2.2 2.2
0.8
3
1.8
2 2 3 2 3
'($"%&" #$$"%&"
0.6
2.8
1.8 1.8 1.8 2
2.6 (a) (b) (c) (d) (e) (f) (e) (f)
0.4
2.5 2.5
1.6 1.6 1.6
2.4
Figure 14. Schematic representation of the processing steps of a new layer of an UTW
0.2
2.2
1.4 1.4
samples, consisting of1.4
(a) a deposition
2
of a new layer of powder, (b) solidification of the first
2
of
track and (c) solidification of the second track. (d, e) Optical microscopy images of the top
0
3
2.8
2.6
2.4
2.2
1.8
1.6
1.4
1.2
1
2 1.2 1.2 1.2
layer of a UTW sample processed at 194 J/m for a distance between tracks of (d) 160 µm
1.8 1 and (e) 1901 µm.1.5(f) Schematic
1 representation
1.5 of the cross-section of a UTW sample and its
0 0
0.5 0 0.5
1 0.5 1
1.5 1 1.5
2 1.5 2 2
ro
roughness profile on both sides. Red arrows indicate melt pools which increase dramatically
1.6
the roughness.
1
1.4 1 0 0.5 1 1.5 2
0 0.5 1 1.5 2
1.2
1
0
on the samples (Figure 14.d). This observation has been experimentally confirmed
in the case of an UTW sample processed with a laser track starting point shifted for
0.5 1 1.5 2
each layer, following the UTW diagonal, as illustrated in Figure 15. A pattern due
-p
re
to a difference in roughness is clearly visible on each face. In some conditions, melt
pools of the second contour extend outward (red arrows in Figure 14.f) too much
which leads to an excessive roughness. When the distance between tracks increases,
lP
the asymmetry decreases but pores due to a bad overlapping appear (Figure 14.e).
In the case of TW samples, factors influencing the roughness are also depending
on the same two factors. The level of TED tends to decrease the roughness by increas-
a
ing the melt pool size while increasing the offset reduces interactions between contours
and bulk hatching melt pools, thus decreasing the roughness. Moreover, increasing
rn
the melt pool size decreases the offset and should then counteract the beneficial geo-
metrical effect of large melt pools. However, the importance of those effects depends
on the adopted strategy.
u
Figure 16 schematically represents the difference between both strategies. In BF

strategy, the melt pools of the contours are influenced by the solidified melt pools
Jo
of the bulk hatching. Indeed, the solidified bulk tracks impact the wetting of the
melt pools of the contours. In the case of a larger offset, the interactions between
the contour and bulk melt pools are less significant. However, increasing the offset
also increases the amount of sub-surface porosities due to a bad overlapping (30; 72)
between the bulk tracks and the contour tracks, which is expected to be detrimen-
tal to the mechanical properties (particularly fatigue). Cracks can also easily initiate
fracture in tensile tests and thus reduce the mechanical quality of the parts (40).
Micrographs of Figure 16.(a) illustrate cross-sections of CF-TW and BF-TW sam-

ples on which the bulk hatching and the contours of the last layer were not processed.
In the case of the BF strategy, the solidified bulk tracks will disturb the melt pools of
2
2
1.8
1.6
2
3

1.5
1.4
2.8

1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
1.2
1
3
1 Layer 1
0
2.6
Single Wall Tracks (SWT) 20
1.5
200 µm 300 µm 600 µm

Layer 43 Layer 4
1
0.2
2.4
2
Powder denudation

plane
1
(CF) First (BF)
Layer n Layer n+1

Powder denudation
Layer
0.8
2.2
1.8
0.4
planexïz

Layer 32 Layer
0.6
1
2
1.6
0.6
plane
Firstplane
Layer n Layer n+1
0.5
1.8 200 µm 300 µm 600 µm
xïz
0.4
Layer 2 Layer
1.4
0.8
1 Bulk hatching
xïy planexïy
2.5 yïz planeyïz
1.6
0.2
1.2
1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
1
0.5
1
3
0
1.4
0
1.2
2.8
2.6
2 2.4
2.2
1.8
1.6
1.4
1.2
1
0.2
0
3
3
1.5 Face 1.5
1
1.2
200 µm 300 µm 600 µm
Layer n Layer n+1
Powder denudation
3 3 33 3
0.8
1.4
3
3 3 xïy plane xïz plane Layer 1
0.4
1
xïy plane xïz plane Layer
1 1 12.8 Layer
0.6 2 Layer 3 1 Layer1.24
2.8 xïyxïy planexïzxïz plane Layer 0 0.2 0.4 0.8 1.4 1.6 1.8 2
plane plane Layer
0
2.8
2.5
Contour Bulk
3
2.5
xïy plane xïy

xïzplane xïz plane
plane Layer 1 1 Layer
Layer
Layer 2 2 Layer
Layer 3 3 Layer
4 4 4
Layer yïz plane
0.6
Layer 2 3 Layer Contour Bulk Face 1 Bulk hatching
1.6
0.6
Layer
Layer 2 Layer 2.5
3 2Layer 2.5
Layer 2.5
4 3 Layeryïz yïz
4 yïz plane
plane 2.6
plane
2.5 2.5yïz plane yïz plane 2.6 (a) (b) (c) (d) 2.6 (a) (b) (c) (d)
0.4
1.8
2.4
0.8
2
2 2 2.4
2
Figure 2.4
15. UTW sample processed with a starting point track shifted at each layer. (a) 2.2
0.2
2 2
2
Schematic2.2 slicing of the UTW samples,

2.2 showing the laser starting points (red dots) in the x-y 2
of
1.5
plane
1.51.51.5 and the corresponding layers in the x-z plane. The red tracks indicate the first track
0
1.8
1.2
2.8
2.6
2.4
2.2
1.8
1.6
1.4
1.2
1.5 1.5 2
(corresponding to face 1) while 2the black ones indicate the second track (corresponding to 1.6
1.4
face1 2).1.8
1
(b) Picture of the UTW 1.8 sample
1
0
processed
0.5
with1 a level of1.5TED = 194
2
J/m. 1.4
ro
1 1 0 0 0 0.50.50.5 1 1 1 1.51.51.5 2 2 2
1.6
1.2
0 0 0.5 0.5 1 1 1.5 1.5 2 2
1.6 1.6 1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
1.8
the contours.
1.4 A large offset 1.4
limits the impact of those solidified tracks and thus the
roughness. As for the UTW samples, this can be observed by the contour melt pools
-p
2
1.2
orientation (dashed arrows in1.2Figure 16.a) of the last processed layer. In the case of
the CF 1 strategy, contour melt
0 0.5
1 pools act as an ’overflow barrier’. The offset has no
0
1 0.5
1.5 1
2 1.5 2
more effect as long as the bulk TED level is equal or lower than the contour TED
re
level. Indeed, for a small offset, the bulk hatching can re-melt the contours (Figure
16.b). In this case, roughness increases since there are no more contours (Figure 9,
CF with a 0 µm-offset).
lP
Increasing the level of TED with a CF strategy decreases the level of roughness
(Figure 8.b) due to a geometrical effect. The same trend is observed in the case of
the BF strategy when the offset is large enough. For smaller offsets, the trends are
a
different because the increase of the melt pool size with the level of TED decreases
the offset and thus increases the roughness (Figure 8.a). For a large level of TED
rn
(above 220 J/m), the positive geometrical effect overpasses the negative effect of the
decreasing offset, thus reducing the level of roughness.
u
Small variations of the solidified bulk melt pools in the case of a BF strategy, due
to slight variations of some parameters (powder size and compaction, layers rotation,
Jo
laser fluctuation...) (46; 73) can induce an uneven side profile of the bulk hatching
(Figure 16.b). It impacts slightly the roughness since this uneven side profile disturbs
the wetting behavior of the contour melt pools. It is worth noting that powder de-
nudation (27; 71) in BF strategy (Figure 1.b) can also have a small impact on the
contour melt pool geometry due to the induced slight lack of powder near the contour.
4. Conclusions
The present study addressed the origin of the vertical roughness in L-PBF AlSi10Mg
alloy components. Three different structures have been considered to address the key
13
12
11
10
9
8
7
5
µm
1 2 3 4 5
m
m
m
0
5
1.5
2.5
1 2 3 4 5
3
0
Ra [µm]
1.5
2.5
21
3
0
Ra [µm]
10
20
30
40
4
0
10
1
10
20data13
30data9
4090 40
Ra [µm]
120 µm
60 µm
30 µm
0 µm
data12
data11
data10
data8
data7
data52.5
!""#$%#1.5
3 1
10 0.5
201.5
30 2
µm
0.5 0
0
2.5
1
3 3 Ra [µm]
3
3
0
Ra [µm]
35
40 3 3 40 21
10 4
20
30
40
Bulk
40 first Contour tracks Bulk first
10
20
30 2
40
1.5
40 40
0
2
Contour tracks
0
Bulk first Contour tracks
1 2
3 Bulk first Contour first
40 Bulk first Contour first
1
3 2.5Contour2.5 tracks 2.5 Bulk first
3 Contour first data3 Contour Contour
first tracks Contour first data3 layer
Last processed
2.5 data3Last processed layer
Offset30
2.5 Last processed layerContour first
( data5
1 0.5 0 0 1.5 1 0.5
2.8
30 First (BF) data3 Lastdata4
processed layer 30 data3 data4
data3 a [µm]
R
0.5
First (CF) 30 Contour tracksdata4Last processed layer
Ra0[µm]
data4 data5 data4 Bulk tracks data5
1 0 data10
30 tracks 30 data5
1.5 120
2.5
Bulk tracks
2 tracks data5 data4 data6 Ra [µm]
Thinïwalls (TW) 2.5 Ultraïthin wall (UTW)
Contour 30[µm] Bulk
120
90 µm
602 µm
303 µm
data13
data12
data11
data10
data9
data8
data7
1.5 µm
2.5 2
1.5
2.5
10 data8
20 data5
40
Rlayer
1 5data9
2 60
3
&
2.6
23Bulkdata6
R
data5
= 0[µm]
tracks
0 data11
1
2
Single Wall 2.5 Tracks (SWT) 2data6
31data13
µm 3
Last processed a Bulk tracks data5
Ra [µm]
Ra [µm]
a
0
3
90 µm
30
data13
data12
data7
data5
120 µm
90 µm
60 µm
30 µm
data12
data11
data10
data9
data8
data7
Contour Meltïpool2 Bulk data7 <<Overflowdata7>> barrier
10 µm
20
30
40
Ra [µm] layer data7<<Overflow >> barrierdata6
3data6
x y processed
z 2
23
Last <<Overflow >> barrier
µm
data6
10
20
30
40
0
µm
2
Ra [µm]
Ra [µm]
2.4
20denudation <<Overflow
data8 data8 >> barrier 20 data7 data8
Ra [µm]
µm
µm
Powder data7 Bulk >> barrierdata7 Unever
1
20 tracks <<Overflow side profil
1
20 Crossïsection Laser 2 path Bulk
data8 (toptracks
view) data9 20 data9 Unever side
Contour profil side
tracks
Uneven
1.5 data8
profil Fully data9
2.2 2 (top view) 1.5 20 data8 reïmelted contour
Laser path data9 2.5 Unever
<<Overflow Contour
data10 side
1.5
>> profilFully
tracks
barrier reïmelted
Uneven side contour
profil
Fully reïmelted
data9contour data10
1.5
1 0.5
1.5 2.5 <<Overflow >> barrier data10 data9
Face 2 First contour Offset 1.5Fully
Hatching reïmelted Last contour
Fully processed
reïmeltedlayercontour
data11 data10 data10data11
R60[µm]
data10
2
3 data11
120 µm data7 data9
90 µm
30 µm 30 µm data5
0 µm
data13
data12 1.5
data111 data12
data9 data10 data12
data8 data9 data11
data7
data5 data5 data8

R [µm]
0.5
R60
2
Second contour Single track Unever Last
side processed
profil layer
a
10 data12data12 10
120
90 µm
30 µm
043
data13
data12
data11
data10 (b)!
data9
data8
data7
data5
!"#$%&$ 10 data11 data11data12 a
4
Fully reïmelted contour

Unever
<<Overflow >> barrier
Bulk tracks
Last processed layer
Contour
a µm
Layer n Layer1.5 n+1 data11
Unever side Fullyprofil
reïmelted contour
µm
Bulk tracks 1
µm 60 µm 120
[µm]
10 10 10
3
10 data8 data10
20 90 µm data7
30
40
1.5 12 data13 0.5data12data13 1
300 µm 600 µmdata12 Fully reïmelted contour 1data13 0 data12 1.5 2
4
2.8
0
First (CF) 200 Firstµm
2
0 Bulk tracks
µm tracks data9
1.8
(BF) 0 0.5 0.5 1 1 1.5 1.5 2 2
2 side profil 3
1 barrier data132data13
Thinïwalls (TW) Ultraïthin 0wall2data13 (UTW) 0.5 1 1 <<Overflow >> 1.5
120 µm
0 µm
data13
2 1.5 1
2.6 3 0 0.5 1 1.5 2

Single
1.6
Wall Tracks0(SWT)01 <<Overflow >> 3 barrier
0
3
3
2 1
Contour3 Meltïpool 11Bulk x 1 0
y z 2 20.5 3 1 Uneven 41 1.5
side profil
52 2 5 2 3 4 5
1.5 31Uneven 04
5 490 µm
2.8
2.4
1.4 0 First
0 0.5 0 1.5
1
Powder (CF) First (BF)

denudation sideFully
profil
reïmelted2contour
5 1 (BF) 3 4 5
5
1Thinïwalls (TW) 2 Ultraïthin 3 (UTW) 2.8 4 Fully 1 reïmelted 2 3 4 5
90 µm
60 µm
30 µm
0 µm
1.5
R [µm]
Crossïsection Laser path (top wall
view) 3
First (CF) First contour
120
60 µm
30 µm
0 3µm
data13
data12
data11 !""#$%#
data10
data8
data7
data5
Contour tracks
4
2
R
2.6
a
5a [µm]
2.8
Single
path Wall Tracks
(CF)(SWT) Thinïwalls (TW) Ultraïthin wall (UTW)
of
2.2
Laser
1.2 (topFirst
view) First
3 x (BF) 2.6 2.5 3 3
µm 120 µm
Contour Meltïpool
Face 22.6 First contour Thinïwalls Bulk
Offset y z
(TW)Hatching
2.8Single Wall Tracks (SWT)
µm
Last processed layer
4
Contour First (CF) First
Bulk(BF)
2.4
2
Second Powder denudation
contour Single
SingleWalltrack
Tracks (SWT) 1 Meltïpool x y z
!"#$%&$ view)x1.2 y z 1.40 0.5 2
Thinïwalls (TW) 1 Ultraïthin 1.5
wall (UTW) 2
2 1.5
1 2.4
Crossïsection Laser 2.8path (top11Bulk Powder 1.8 denudation 2.8
2 1.5
0 0.2 0.4 0.6 0.8 1.6

LayerLaser
n Layer Contour
n+1 Meltïpool 2.6
0 0.5 Crossïsection Single1WallLaserTracks view)tracks Contour
1.5
(SWT)
Bulk 2 tracks
90 µm
60 µm
30 µm
0 µm
data13
data12
data11
data10
data9
data8
data7
data5
1.5
2.2 2.4 path Powder

(top view)denudation path (top
200 µm 300 µm 600
ro
1.8 µm Offset Hatching 2.2 Contour
2 view)Meltïpool2.5Bulk x y z
3
Face 2 First contour
Crossïsection Laser path (a) (top(b)view) Laser path
(c) 2.4(d) (top (a) (b) (c) (d)
2.6 track Powder denudation 5 2.6
5
2
Second
2.2 contour
Laser Single
path (top view) Face 2 First contour <<Overflow >>
Offset Hatching Last processed layer
barrier
4
Layer n Layer Crossïsection
contour Single Laser
track path (top view)
Facen+1
2 First contour Offset Hatching2.2Second Laser
5
1.6 2
4
1.8 Figure
2002 µm 30016.
µm 600
Second Schematic
2.4
µm
contour Singlerepresentation
track ofn the
Layer Layer path
n+1(top view)
contour laser
2.4 trackside
Unever andprofil
bulk
Bulk laser track as
tracks a
Faceµm2 600
Firstµm
1.4 functionLayer
of the strategies
n Layer n+1 and1.8of the2200 µm 300
offset 1.5the
in (a) x-z 2plane
Fully and in
reïmeltedthe (b)
contour x-y plane.
1.6 1.8 2.2µm 600 µm
200 µm 300
Layer n Layer n+1 2.2 <<Overflow >> barrier
-p 4
2
1.6
2
2
1.2
1.4
1.8 200 µm 300 µm 600 µm
1.6
2 2 Unever side profil
5
factors impacting the level of roughness. 1.4

1 1.5
5
1 1.6
0 0.5 Fully
1 reïmelted1.5contour 2
re
0 1.20.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
1.4
1.8 1.2
1.8
5
1.4
1
0 0.2 Single wall tracks (SWT) specimens have been used to assess the effect of the
1.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
1.6 1.6 1
track energy density (TED) on the roughness and on10the hardness.
1 0.5 It has been
1 shown 1.5
0 0.21.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
2
lP
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

that increasing
1.4 the level of TED has a beneficial impact
1.4 on the reduction of the rough- 1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
ness due to a geometrical effect. A large melt pool leads to a more vertical side-profile
than a small 1.2
melt-pool. However, it has a detrimental 1.2
effect on the hardness due to
the coarsening1 of the microstructure. The optimum level 1 of TED for low roughness
0 0.5 1 1.5
0 0.5
2 1 1.5 2
a
was found to be at the onset of the keyhole conditions. Increasing the TED level above
that value increases the amount of sintered powder particles on the side of SWT sam-
rn
ples and thus cancels the benefit of the geometrical effect. Moreover, it brings a lot of
porosities induced by keyhole laser conditions.
Ultra thin wall (UTW) specimens have been used to study the effect of parallel
u
tracks on the roughness. Distance or offset between melt pools was found to be a
key element to reduce the roughness. A larger offset between melt pools is always
Jo
beneficial to reduce the level of roughness. Indeed, the wetting occurring between a
melt pool and a solidified melt pool next to it leads to a slight overflow of the liquid,
increasing the roughness.
Thin wall (TW) samples with contours and bulk hatching have been used to
analyze the impact of the TED level and of the offset together. The two possible
strategies, contour first (CF) or bulk first (BF) have been investigated. With a CF
strategy, only the level of TED has a first order impact. Increasing the TED level
lowers the level of roughness. With a BF strategy, both the offset and the TED
level have a first order effect. Increasing the offset allows reducing the wetting issue
between the contour melt pool and the solidified bulk hatching melt pools but increases
REFERENCES 22
porosities due to a bad overlapping between the contours and the bulk hatching.
However, increasing the TED level has a mixed effect. On the one hand, it tends to
reduce the roughness due to a geometrical effect. On the other hand, increasing the
size of the melt pools also decreases the offset and thus affects negatively the roughness.
In the different conditions investigated for TW specimens, a CF strategy always gave
the lowest vertical roughness compared to a BF strategy. This strategy should then
be preferably used to process any object. A relatively high level of TED has also to be
used for the contours to obtain a smooth surface. The difference in hardness between
contours and bulk induced by the different parameters might be a problem for the
mechanical properties, especially in thin elements, as damaging mechanisms depend
on the microstructure (74; 75), and requires further investigations.
Acknowledgments
of
The present work is supported by the Walloon Region through the Skywin project
’InHex’. The authors also acknowledge the financial support from the European
Funds for Regional Developments (FEDER) and the Walloon Region in the framework
ro
of the operational program Wallonie-2020.EU (project : IAWATHA/AManUMater,
n101628-722943). This work has been performed with the help of the Lacami
technological platform of UCLouvain. The authors also thank the Wallonia Electronics
-p
and Communications Measurements platform (WELCOME) of UCLouvain for the use
of the white light interferometer.
re
References
[1] Tarasankar DebRoy, HL Wei, JS Zuback, T Mukherjee, JW Elmer, JO Milewski,

lP
Allison Michelle Beese, A Wilson-Heid, A De, and W Zhang. Additive

manufacturing of metallic components–process, structure and properties. Progress
in Materials Science, 92:112–224, 2018.
[2] Nesma T Aboulkhair, Ian Maskery, Chris Tuck, Ian Ashcroft, and Nicola M
a
Everitt. Improving the fatigue behaviour of a selectively laser melted aluminium

alloy: Influence of heat treatment and surface quality. Materials & Design,
rn
104:174–182, 2016.
[3] Ana D Brandão, Johannes Gumpinger, Michael Gschweitl, Christoph Seyfert,
Peter Hofbauer, and Tommaso Ghidini. Fatigue properties of additively
u
manufactured alsi10mg–surface treatment effect. Procedia Structural Integrity,

7:58–66, 2017.
Jo
[4] Sara Bagherifard, Niccolò Beretta, Stefano Monti, Martina Riccio, Michele
Bandini, and Mario Guagliano. On the fatigue strength enhancement of
additive manufactured alsi10mg parts by mechanical and thermal post-processing.
Materials & Design, 145:28–41, 2018.
[5] Emilie Beevers, Ana D Brandão, Johannes Gumpinger, Michael Gschweitl,
Christoph Seyfert, Peter Hofbauer, Thomas Rohr, and Tommaso Ghidini. Fatigue
properties and material characteristics of additively manufactured alsi10mg–effect
of the contour parameter on the microstructure, density, residual stress, roughness
and mechanical properties. International Journal of Fatigue, 117:148–162, 2018.
[6] Marina Cabrini, Sergio Lorenzi, Tommaso Pastore, Simone Pellegrini, Matteo
Pavese, Paolo Fino, Elisa Paola Ambrosio, Flaviana Calignano, and Diego
REFERENCES 23
Manfredi. Corrosion resistance of direct metal laser sintering alsimg alloy. Surface
and Interface Analysis, 48(8):818–826, 2016.
[7] Avi Leon and Eli Aghion. Effect of surface roughness on corrosion fatigue
performance of alsi10mg alloy produced by selective laser melting (slm). Materials
Characterization, 131:188–194, 2017.
[8] P Fathi, M Rafieazad, X Duan, M Mohammadi, and AM Nasiri. On
microstructure and corrosion behaviour of alsi10mg alloy with low surface
roughness fabricated by direct metal laser sintering. Corrosion Science, 2019.
[9] Flaviana Calignano, D Manfredi, EP Ambrosio, Luca Iuliano, and Paolo Fino.
Influence of process parameters on surface roughness of aluminum parts produced
by dmls. The International Journal of Advanced Manufacturing Technology, 67(9-
12):2743–2751, 2013.
[10] Lin-zhi Wang, Sen Wang, and Jiao-jiao Wu. Experimental investigation on
of
densification behavior and surface roughness of alsi10mg powders produced by
selective laser melting. Optics & Laser Technology, 96:88–96, 2017.
ro
[11] Ahmed H Maamoun, Mohamed Elbestawi, Goulnara K Dosbaeva, and Stephen C
Veldhuis. Thermal post-processing of alsi10mg parts produced by selective laser
melting using recycled powder. Additive Manufacturing, 21:234–247, 2018.
-p
[12] Ahmed Maamoun, Yi Xue, Mohamed Elbestawi, and Stephen Veldhuis. Effect of
selective laser melting process parameters on the quality of al alloy parts: Powder
characterization, density, surface roughness, and dimensional accuracy. Materials,
re
11(12):2343, 2018.
[13] Mohsen Mohammadi and Hamed Asgari. Achieving low surface roughness
alsi10mg 200c parts using direct metal laser sintering. Additive Manufacturing,
lP
20:23–32, 2018.
[14] Wenhui Yu, Swee Leong Sing, Chee Kai Chua, and Xuelei Tian. Influence of re-
melting on surface roughness and porosity of alsi10mg parts fabricated by selective
laser melting. Journal of Alloys and Compounds, 792:574–581, 2019.
a
[15] Arfan Majeed, Jingxiang Lv, Yingfeng Zhang, Muhammad Muzamil, Ali Waqas,
Khalid Shamim, Muhammad Ejaz Qureshi, and Fahad Zafar. An investigation
rn
into the influence of processing parameters on the surface quality of alsi10mg

parts by slm process. In 2019 16th international bhurban conference on applied
sciences and technology (IBCAST), pages 143–147. IEEE, 2019.
u
[16] Arfan Majeed, Altaf Ahmed, Abdus Salam, and Muhammad Zakir Sheikh.
Surface quality improvement by parameters analysis, optimization and heat
Jo
treatment of alsi10mg parts manufactured by slm additive manufacturing.

International Journal of Lightweight Materials and Manufacture, 2(4):288–295,
2019.
[17] Bao-Qiang Li, Zhonghua Li, Peikang Bai, Bin Liu, and Zezhou Kuai. Research
on surface roughness of alsi10mg parts fabricated by laser powder bed fusion.
Metals, 8(7):524, 2018.
[18] Tao Yang, Tingting Liu, Wenhe Liao, Eric MacDonald, Huiliang Wei, Xiangyuan
Chen, and Liyi Jiang. The influence of process parameters on vertical surface
roughness of the alsi10mg parts fabricated by selective laser melting. Journal of
Materials Processing Technology, 266:26–36, 2019.
REFERENCES 24
[19] PB Bacchewar, SK Singhal, and PM Pandey. Statistical modelling and

optimization of surface roughness in the selective laser sintering process.
Proceedings of the Institution of Mechanical Engineers, Part B: Journal of
Engineering Manufacture, 221(1):35–52, 2007.
[20] Alberto Boschetto, Luana Bottini, and Francesco Veniali. Roughness modeling
of alsi10mg parts fabricated by selective laser melting. Journal of Materials
Processing Technology, 241:154–163, 2017.
[21] Xuesong Han, Haihong Zhu, Xiaojia Nie, Guoqing Wang, and Xiaoyan Zeng.
Investigation on selective laser melting alsi10mg cellular lattice strut: molten pool
morphology, surface roughness and dimensional accuracy. Materials, 11(3):392,
2018.
[22] A Barari, HA Kishawy, F Kaji, and MA Elbestawi. On the surface
quality of additive manufactured parts. The International Journal of Advanced
of
Manufacturing Technology, 89(5-8):1969–1974, 2017.
[23] F Calignano. Investigation of the accuracy and roughness in the laser powder bed
fusion process. Virtual and Physical Prototyping, 13(2):97–104, 2018.
ro
[24] Charlotte de Formanoir, Umberto Paggi, Thomas Colebrants, Lore Thijs,
Guichuan Li, Kim Vanmeensel, and Brecht Van Hooreweder. Increasing the
productivity of laser powder bed fusion: influence of the hull-bulk strategy on
-p
part quality, microstructure and mechanical performance of ti-6al-4v. Additive
Manufacturing, page 101129, 2020.
re
[25] I Yadroitsev, Pavel Krakhmalev, I Yadroitsava, Sten Johansson, and I Smurov.
Energy input effect on morphology and microstructure of selective laser melting
single track from metallic powder. Journal of Materials Processing Technology,
213(4):606–613, 2013.
lP
[26] Karolien Kempen, Lore Thijs, Jan Van Humbeeck, and J-P Kruth. Processing
alsi10mg by selective laser melting: parameter optimisation and material
characterisation. Materials Science and Technology, 31(8):917–923, 2015.
a
Everitt. On the formation of alsi10mg single tracks and layers in selective laser
rn
melting: Microstructure and nano-mechanical properties. Journal of Materials

[28] Eleftherios Louvis, Peter Fox, and Christopher J Sutcliffe. Selective laser
u
melting of aluminium components. Journal of Materials Processing Technology,

211(2):275–284, 2011.
Jo
[29] Lore Thijs, Karolien Kempen, Jean-Pierre Kruth, and Jan Van Humbeeck. Fine-
structured aluminium products with controllable texture by selective laser melting
of pre-alloyed alsi10mg powder. Acta Materialia, 61(5):1809–1819, 2013.
[30] Nesma T Aboulkhair, Nicola M Everitt, Ian Ashcroft, and Chris Tuck.
Reducing porosity in alsi10mg parts processed by selective laser melting. Additive
Manufacturing, 1:77–86, 2014.
[31] Pei Wei, Zhengying Wei, Zhen Chen, Jun Du, Yuyang He, Junfeng Li, and Yatong
Zhou. The alsi10mg samples produced by selective laser melting: single track,
densification, microstructure and mechanical behavior. Applied surface science,
408:38–50, 2017.
REFERENCES 25
[32] F Calignano, G Cattano, and D Manfredi. Manufacturing of thin wall structures

in alsi10mg alloy by laser powder bed fusion through process parameters. Journal
of Materials Processing Technology, 255:773–783, 2018.
[33] Pierre Lhuissier, Xavier Bataillon, Camille Maestre, Julien Sijobert, Elodie
Cabrol, Philippe Bertrand, Elodie Boller, Alexander Rack, Jean-Jacques Blandin,
Luc Salvo, et al. In situ 3d x-ray microtomography of laser-based powder-bed
fusion (l-pbf)-a feasibility study. Additive Manufacturing, page 101271, 2020.
[34] Luhao Yuan, Dongdong Gu, Kaijie Lin, Qing Ge, Xinyu Shi, Haoran Wang, and
Kaiming Hu. Influence of structural features on processability, microstructures,
chemical compositions, and hardness of selective laser melted complex thin-walled
components. The International Journal of Advanced Manufacturing Technology,
109(5):1643–1654, 2020.
[35] Idan Rosenthal, Adin Stern, and Nachum Frage. Microstructure and mechanical
of
properties of alsi10mg parts produced by the laser beam additive manufacturing
(am) technology. Metallography, Microstructure, and Analysis, 3(6):448–453,
2014.
ro
[36] Michaela Fousová, Drahomı́r Dvorskỳ, Alena Michalcová, and Dalibor Vojtěch.
Changes in the microstructure and mechanical properties of additively
manufactured alsi10mg alloy after exposure to elevated temperatures. Materials
-p
[37] Ming Tang and Petrus Christiaan Pistorius. Anisotropic mechanical behavior of
alsi10mg parts produced by selective laser melting. Jom, 69(3):516–522, 2017.
re
[38] YJ Liu, Z Liu, Y Jiang, GW Wang, Yang Yang, and LC Zhang. Gradient
in microstructure and mechanical property of selective laser melted alsi10mg.
Journal of Alloys and Compounds, 735:1414–1421, 2018.
lP
[39] Le Zhou, Abhishek Mehta, Esin Schulz, Brandon McWilliams, Kyu Cho,
and Yongho Sohn. Microstructure, precipitates and hardness of selectively
laser melted alsi10mg alloy before and after heat treatment. Materials
a
[40] Karolien Kempen, Lore Thijs, Jan Van Humbeeck, and Jean-Pierre Kruth.
Mechanical properties of alsi10mg produced by selective laser melting. Physics
rn
Procedia, 39:439–446, 2012.

Everitt. The microstructure and mechanical properties of selectively laser melted
u
alsi10mg: The effect of a conventional t6-like heat treatment. Materials Science

and Engineering: A, 667:139–146, 2016.
Jo
[42] A Iturrioz, E Gil, MM Petite, F Garciandia, AM Mancisidor, and

M San Sebastian. Selective laser melting of alsi10mg alloy: influence of heat
treatment condition on mechanical properties and microstructure. Welding in
the World, 62(4):885–892, 2018.
[43] Wei Li, Shuai Li, Jie Liu, Ang Zhang, Yan Zhou, Qingsong Wei, Chunze
Yan, and Yusheng Shi. Effect of heat treatment on alsi10mg alloy fabricated
by selective laser melting: Microstructure evolution, mechanical properties and
fracture mechanism. Materials Science and Engineering: A, 663:116–125, 2016.
[44] I Yadroitsev, A Gusarov, I Yadroitsava, and I Smurov. Single track formation
in selective laser melting of metal powders. Journal of Materials Processing
Technology, 210(12):1624–1631, 2010.
REFERENCES 26
[45] Suman Das. Physical aspects of process control in selective laser sintering of
metals. Advanced Engineering Materials, 5(10):701–711, 2003.
[46] Carolin Körner, Elham Attar, and Peter Heinl. Mesoscopic simulation of selective
beam melting processes. Journal of Materials Processing Technology, 211(6):978–
987, 2011.
[47] Guanqun Yu, Dongdong Gu, Donghua Dai, Mujian Xia, Chenglong Ma, and
Qimin Shi. On the role of processing parameters in thermal behavior, surface
morphology and accuracy during laser 3d printing of aluminum alloy. Journal of
Physics D: Applied Physics, 49(13):135501, 2016.
[48] Saad A Khairallah and Andy Anderson. Mesoscopic simulation model of
selective laser melting of stainless steel powder. Journal of Materials Processing
Technology, 214(11):2627–2636, 2014.
[49] Saad A Khairallah, Andrew T Anderson, Alexander Rubenchik, and Wayne E
of
King. Laser powder-bed fusion additive manufacturing: Physics of complex melt
flow and formation mechanisms of pores, spatter, and denudation zones. Acta
Materialia, 108:36–45, 2016.
ro
[50] Min Zheng, Lei Wei, Jing Chen, Qiang Zhang, Chongliang Zhong, Xin Lin, and
Weidong Huang. A novel method for the molten pool and porosity formation
modelling in selective laser melting. International Journal of Heat and Mass
Transfer, 140:1091–1105, 2019. -p
[51] Noriko Read, Wei Wang, Khamis Essa, and Moataz M Attallah. Selective
re
laser melting of alsi10mg alloy: Process optimisation and mechanical properties
development. Materials & Design (1980-2015), 65:417–424, 2015.
[52] U Tradowsky, J White, RM Ward, N Read, W Reimers, and MM Attallah.
lP
Selective laser melting of alsi10mg: Influence of post-processing on the

microstructural and tensile properties development. Materials & Design, 105:212–
222, 2016.
[53] Jean-Pierre Kruth, Ludo Froyen, Jonas Van Vaerenbergh, Peter Mercelis, Marleen
a
Rombouts, and Bert Lauwers. Selective laser melting of iron-based powder.

Journal of materials processing technology, 149(1-3):616–622, 2004.
rn
[54] Leonhard Hitzler, Christoph Janousch, Jochen Schanz, Markus Merkel, Burkhard
Heine, Florian Mack, Wayne Hall, and Andreas Öchsner. Direction and location
dependency of selective laser melted alsi10mg specimens. Journal of Materials
u

[55] Peter de Groot. Coherence scanning interferometry. In Optical measurement of
Jo
surface topography, pages 187–208. Springer, 2011.

[56] Jon Petzing, Jeremy Coupland, and Richard K Leach. Good practice guide No.
116, The measurement of rough surface topography using coherence scanning
interferometry. National Physical Laboratory, 2010.
[57] Andrew Townsend, N Senin, Liam Blunt, RK Leach, and JS Taylor. Surface
texture metrology for metal additive manufacturing: a review. Precision
Engineering, 46:34–47, 2016.
[58] Nicola Senin, Adam Thompson, and Richard K Leach. Characterisation of
the topography of metal additive surface features with different measurement
technologies. Measurement Science and Technology, 28(9):095003, 2017.
REFERENCES 27
[59] Adam Thompson, Nicola Senin, Claudiu Giusca, and Richard Leach. Topography
of selectively laser melted surfaces: a comparison of different measurement
methods. CIRP Annals, 66(1):543–546, 2017.
[60] Carlos Gomez, Rong Su, Adam Thompson, Jack DiSciacca, Simon Lawes,
and Richard K Leach. Optimization of surface measurement for metal additive
manufacturing using coherence scanning interferometry. Optical Engineering,
56(11):111714, 2017.
[61] J-Y Buffiere, E Maire, J Adrien, J-P Masse, and E Boller. In situ experiments with
x ray tomography: an attractive tool for experimental mechanics. Experimental
mechanics, 50(3):289–305, 2010.
[62] Jérôme Adrien, Sylvain Meille, Solène Tadier, Eric Maire, and Layla Sasaki.
In-situ x-ray tomographic monitoring of gypsum plaster setting. Cement and
Concrete Research, 82:107–116, 2016.
of
[63] Johannes Schindelin, Ignacio Arganda-Carreras, Erwin Frise, Verena Kaynig,
Mark Longair, Tobias Pietzsch, Stephan Preibisch, Curtis Rueden, Stephan
Saalfeld, Benjamin Schmid, et al. Fiji: an open-source platform for biological-
ro
image analysis. Nature methods, 9(7):676, 2012.
[64] Wayne E King, Holly D Barth, Victor M Castillo, Gilbert F Gallegos, John W
Gibbs, Douglas E Hahn, Chandrika Kamath, and Alexander M Rubenchik.
-p
Observation of keyhole-mode laser melting in laser powder-bed fusion additive
manufacturing. Journal of Materials Processing Technology, 214(12):2915–2925,
2014.
re
[65] Donghua Dai and Dongdong Gu. Effect of metal vaporization behavior on keyhole-
mode surface morphology of selective laser melted composites using different
protective atmospheres. Applied Surface Science, 355:310–319, 2015.
lP
[66] Jieren Guan, Yehua Jiang, Xiaowei Zhang, and Xiaoyu Chong. Microstructural
evolution and ebsd analysis of alsi10mg alloy fabricated by selective laser
remelting. Materials Characterization, 161:110079, 2020.
a
[67] B Chen, SK Moon, X Yao, G Bi, J Shen, J Umeda, and K Kondoh. Strength and
strain hardening of a selective laser melted alsi10mg alloy. Scripta Materialia,
rn
141:45–49, 2017.
[68] Pauline Delroisse, Pascal J. Jacques, Eric Maire, Olivier Rigo, and Aude Simar.
Effect of strut orientation on the microstructure heterogeneities in alsi10mg
u
lattices processed by selective laser melting. Scripta Materialia, 141:32–35, 2017.

[69] I Rosenthal, R Shneck, and A Stern. Heat treatment effect on the
Jo
mechanical properties and fracture mechanism in alsi10mg fabricated by additive

manufacturing selective laser melting process. Materials Science and Engineering:
A, 729:310–322, 2018.
[70] E Sert, L Hitzler, B Heine, M Merkel, E Werner, and A Öchsner. Influence of the
heat treatment on the microstructure and hardness of additively manufactured
alsi10mg samples. Practical Metallography, 56(2):91–105, 2019.
[71] I Yadroitsev and I Smurov. Surface morphology in selective laser melting of metal
powders. Physics Procedia, 12:264–270, 2011.
[72] Ming Tang, P Chris Pistorius, and Jack L Beuth. Prediction of lack-of-fusion
porosity for powder bed fusion. Additive Manufacturing, 14:39–48, 2017.
REFERENCES 28
[73] Carolin Körner, Andreas Bauereiß, and Elham Attar. Fundamental consolidation
mechanisms during selective beam melting of powders. Modelling and Simulation
in Materials Science and Engineering, 21(8):085011, 2013.
[74] Lv Zhao, Juan Guillermo Santos Macı́as, Lipeng Ding, Hosni Idrissi, and Aude
Simar. Damage mechanisms in selective laser melted alsi10mg under as built
and different post-treatment conditions. Materials Science and Engineering: A,
764:138210, 2019.
[75] Juan Guillermo Santos Macı́as, Chola Elangeswaran, Lv Zhao, Brecht
Van Hooreweder, Jérôme Adrien, Eric Maire, Jean-Yves Buffière, Wolfgang
Ludwig, Pascal J Jacques, and Aude Simar. Ductilisation and fatigue life
enhancement of selective laser melted alsi10mg by friction stir processing. Scripta
Materialia, 170:124–128, 2019.
of
ro
-p
re
a lP
u rn
Jo
Authors statement
We are not interested in recognizing author individual contributions.
f
oo
pr
e-
Pr
n al
ur
Jo
Declaration of interests
☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.
☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests:
f
oo
pr
e-
Pr
n al
ur
Jo

Poncelet Et Al. (2021) - Critical Assessment of The Impact of Process Parameters On Vertical Roughness of AlSi10Mg

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Poncelet Et Al. (2021) - Critical Assessment of The Impact of Process Parameters On Vertical Roughness of AlSi10Mg

Uploaded by

Copyright:

Available Formats

Journal Pre-proof

Critical assessment of the impact of process

O. Poncelet, M. Marteleur, C. van der Rest, O.

To appear in: Additive Manufacturing

O. Poncelet1 , M. Marteleur1 , C. van der Rest1 , O. Rigo2 , J.

Keywords: L-PBF, SLM, AlSi10Mg, roughness, hardness, process parameters,

case of L-PBF, roughness can thus be considered as a topographical measurement of

has been drawn yet for the contour process parameters.

heat transfer, pressure drop and eventually fatigue resistance.

As a conclusion, the optimized set of parameters was tested on thin-walls and on

2. Materials and Experimental Procedure

2.1. Manufacturing by L-PBF

post-treatments were applied.

Hardness measurement were not performed on those samples as the microstructure

First (CF) First (BF) 5

2.8 First (CF) Thinïwalls

Single Tracks (SWT)

Contour Bulk x y z Crossïsection Laser path (top view)

2.2 2 2.2 1.8

Layer nn+1Layer n+1

LayerWall n+1Tracks 1.6 1.6

1.6200 µm 300 µm 200 600 n Layer n+1

1.8 2 µm 200 µm 2.6 300 µm 600 µm 200 µm 300 µm 600 µm

Contour Bulk Face 1 Bulk hatching

(BF) Face 2(TW) Ultraïthin wall (UTW) 1.6 1.6 3

Contour Bulk Face 1 Bulk hatching

Single Wall (SWT) Layer Layer n+1 (top

z2.6 x y Thinïwalls 1.2

(TW) Ultraïthin wall (UTW)

1.2 1.2 1.2 1 1.2

2.4 1.8 2.8

1.4 1.210 1.6 1.40.2 1.8 1.60.4 2

eThinïwalls (TW) Ultraïthin wall (UTW)2

200 µm 300 µm0.6 600 µm

0 0.2 0.4 00.8 1.2 1.4 1.6 11.8 1.22

u (CF) First (BF)

Laser path (top view)

First (CF) First (BF)Wall Traks (SWT)

Contour Meltïpool Bulk x Contour y z Meltïpool Bulk x y Laser

2 First contour Offset Hatching

600Layer n Layer n+1

(TW) Ultraïthin wall (UTW)

Laser path (top view) contour Single track 2.2

Laser path (top view)

track Single track

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 2

Layer n Layer n+1

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

0.2 Second contour 0.8Single 1 track Second contour Single track

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

tour First (CF) First Bulk(BF)

wder denudation (TW) Ultraïthin

e 2 First contour Offset2.6Hatching 2.6 2.6 2.6

representation types of samples and their respective

Face 2 First contour laser track patterns;

1.6 1.6 1.6 1.6

1.4 1.4 1.4 1.4

1.2 Finally, 1.2 optimum parameters 1.2 found with SWT,

2.2. Characterization techniques

Hardness was measured with a DuraScan G5 from EMCO-TEST with a Vickers

on the TED level.