Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 111 (2022) 684–688
www.elsevier.com/locate/procedia

12th CIRP Conference on Photonic Technologies [LANE 2022], 4-8 September 2022, Fürth, Germany

Laser polishing of additive manufactured L-PBF copper parts with visible

laser wavelength of 515 nm – Challenges due to high surface roughness
Patrick Gruba,*, Markus Hofelea, Jochen Schanza, David Kolba, Harald Riegela
a
LaserApplicationCenter, Aalen University, Beethovenstr. 1, D-73430 Aalen, Germany

* Corresponding author. Tel.: +49-7361-576-2662; E-mail address: [email protected]

Abstract

Copper components produced by Laser Powder Bed Fusion (L-PBF) have a high surface roughness of up to Ra = 21.6 µm. The surface has a
structure depth of up to 300 µm, as well as porosity. In order to smooth this surface, laser polishing of copper is being investigated. A challenge
here is the low absorption of the near-infrared laser beams as well as the high thermal conductivity of the copper material. To increase the
absorbed laser radiation, a 1 kW laser source in the visible green wavelength range is used instead of a near-infrared laser source. In the first-time
investigations of laser polishing of copper, the laser beam intensity, the line energy as well as the track overlap are varied. Microscope images,
cross-sections and tactile roughness measurements are used for the analysis. A laser beam intensity of 4.95 kW/cm² is required to completely
remelt the high surface structure depth without porosity. The surface roughness could be reduced by 85.2% from Ra = 21.6 µm to Ra = 3.2 µm.
© 2022 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0)
Peer-review under responsibility of the international review committee of the12th CIRP Conference on Photonic Technologies [LANE 2022]

Keywords: Selective laser melting (SLM); Surface quality; Surface treatment; Surface remelting; Laser Powder bed Fusion (L-PBF)

1. Introduction effects [5, 6]. Therefore, their direct industrial use without
surface post-treatment is limited. A possibility to improve the
Laser Powder Bed Fusion (L-PBF) is one of the most rough surfaces of AM components is offered using non-
common Additive Manufacturing (AM) technique for contact, volume-preserving laser polishing by remelting a thin
industrial manufacturing of complex metal parts. AM of copper layer of material. Thereby, the material peaks flow into the
and its alloys exhibit a wide range of potential applications in valleys due to the surface tensions [7]. In recent years, it has
the fields of optical components, thermal radiators, electrical been shown for various metal alloys like Cobalt-Chromium [8],
assemblies and inductors [1]. However, copper exhibits a less Inconel 718 [9], tool steel [10], maraging steel 1.2709 [11],
laser beam absorption in near-infrared of 2.5% in combination Titanium Ti6Al-4V [12], corrosion resistant steel 316L [13]
with high thermal conductivities greater 393 W/m∙K, which and aluminum [14], that a significant surface roughness
makes it challenging to process [2]. Due to the increasing laser reduction of up to 98% can be achieved using near-infrared
beam intensity and quality, as well as the availability of laser lasers. Based on the authors’ knowledge, laser polishing of pure
sources in the visible wavelength range with higher absorption copper AM components has not yet been investigated due to
in copper, AM with high densities of up to 99.8% has recently the low laser beam absorption in the near-infrared range.
become possible [3, 4]. But it has still one main drawback. For Therefore, in this research work, laser polishing of L-PBF
copper, the L-PBF melting process causes extremely rough produced copper parts using visible laser wavelength of
surface topographies up to 300 µm, consisting of adhering 515 nm is investigated. Special attention is focused on the
powder particles and material accumulations, as well as balling surface topography and the microstructure which are analyzed

2212-8271 © 2022 The Authors. Published by Elsevier B.V.

This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0)
Peer-review under responsibility of the international review committee of the12th CIRP Conference on Photonic Technologies [LANE 2022]
10.1016/j.procir.2022.08.009

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.
 Patrick Grub et al. / Procedia CIRP 111 (2022) 684–688 685

by tactile roughness measurements, optical microscopy and topography of the polished fields are characterized using a
cross-sections. KEYENCE VR-3100 3D optical profilometer with 25x
magnification. The surface structure of the initial part, the
2. Experimental Setup remelting zone geometry and porosity after laser polishing is
analyzed on polished cross-sections in fabrication direction FD
2.1. Material and specimen and polishing direction PD, respectively, which were etched by
use of an iron (III) chloride etchant. Microscopic images with
Laser polishing investigations are executed on additive a 100x magnification were taken using a CARL ZEISS Axio
manufactured pure copper plates. The vertically printed plate Imager.Z2 Vario.
dimensions are 80 mm length, 50 mm width and a material
thickness of 3.3 mm. The fabrication of the plates took place 2.3. Experimental plan and used process parameters
on a TRUMPF TruPrint 1000 L-PBF machine equipped with a
green 515 nm disk laser and a laser spot diameter of 200 µm. Laser polishing is performed on square fields with a size of
The used copper powder has an average grain size of 25 µm. 8 x 8 mm² with an areal bi-directional linear hatching
perpendicular to the fabrication direction. Between the fields is
2.2. Applied laser polishing setup, used measurement devices a distance of 2 mm. The polishing experiments are performed
and evaluation methods with a defocused laser beam diameter dl of 0.5 mm. Process
parameters are investigated by varying the laser power Pl
Laser polishing is carried out using a disk laser TRUMPF between 700 W and 1000 W, the beam velocity vs between
TruDisk 1020 with a visible green wavelength of 515 nm and 50 mm/s and 80 mm/s and the track overlap TO between 60%
a maximum output power of 1000 W. The laser beam is guided and 90%. The resulting average laser beam intensity Iavg can be
through a 50 µm fiber with a Numerical Aperture NA of 0.1 to calculated by Eq. (1). The energy per unit length Edis is given
a 5-axis TRUMPF TruLaser Station 5005 with a maximal axis by the laser power Pl and beam velocity vs, see Eq. (2). The
velocity of 100 mm/s. The laser beam is focused by a TRUMPF areal energy density ED depending on dl, Pl, TO and vs can be
BEO D70 optic with a focal length of 200 mm resulting in a calculated by Eq. (3).
measured focal beam diameter of 67 µm and a Rayleigh length 𝑃𝑃𝑃𝑃𝐿𝐿𝐿𝐿 ∙4
of 0.53 mm. Polishing experiments took place in a polishing 𝐼𝐼𝐼𝐼𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = (1)
𝜋𝜋𝜋𝜋𝜋𝜋𝜋𝜋𝜋𝑙𝑙𝑙𝑙2
chamber under argon atmosphere. The Residual oxygen
𝑃𝑃𝑃𝑃𝑙𝑙𝑙𝑙
content of less than 50 ppm was monitored by an oxygen meter 𝐸𝐸𝐸𝐸𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 = (2)
𝑎𝑎𝑎𝑎𝑠𝑠𝑠𝑠
ZIROX SGM7. During the experiments, the copper plates were
𝑃𝑃𝑃𝑃𝑙𝑙𝑙𝑙
clamped in the polishing chamber with a specimen holder 𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 = (3)
𝑎𝑎𝑎𝑎𝑠𝑠𝑠𝑠 ∗(𝑑𝑑𝑑𝑑𝑙𝑙𝑙𝑙 (1−𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇)
where the vertically printed side was polished. Fig. 1 shows the
laser polishing experimental setup. 3. Results and Discussion

3.1. Initial surface analysis

The vertical printed plates exhibit an average surface

roughness Ra in FD of 21.6 µm and perp. FD of 20.1 µm. The
surface structure consists of closely spaced adhering powder
particles, overlaid by material accumulations in the order of
several hundred micrometers, see Fig. 2 a), b). The surface
exhibits a high surface structure depth of 300 µm with inner
pores, see Fig. 2 c).

Fig. 1. Laser polishing experimental setup.

Quantitative analysis of the surfaces roughness, i.e. Ra, was

performed tactilely using a MarSurf M400 perthometer. The
initial L-PBF surface with Ra >> 2 μm is measured with a cut-
off wavelength of 2500 μm, while for the polished surfaces the
cut-off wavelength was set to 800 μm. Five homogeneously
distributed linear measurements in fabrication direction (in FD)
and perpendicular (perp. FD) are performed on each polished
field to characterize the orientation-dependent surface quality.
Analysis of the surface behaviour is done using an optical
Fig. 2. Initial surface; a) 3D topography, b) microscopic image and c) cross-
microscope CARL ZEISS Axio Zoom V.16 with a 50x section.
magnification. The initial surface topography and the global

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.
 686 Patrick Grub et al. / Procedia CIRP 111 (2022) 684–688

3.2. Influence of energy per unit length

The achievable surface roughness Ra depending on the

energy per unit length Edis by varying the laser power Pl and the
beam velocity vs is given in Fig. 3. Polishing with a laser power
of 700 W leads to a roughness increase in fabrication direction
up to Ra = 25.4 µm.

Fig. 4. Surface topography after laser polishing depending on the laser power
Pl and beam velocity vs with TO = 80 %.

When observing the created surface microstructure at the

center of the polishing fields with a laser powers of 900 W and
less an unclosed remelting zone containing holes with
dimensions partly above a millimeter exists, see Fig. 5. With
Pl = 1000 W independent from the beam velocity vs, a dense
remelting surface containing a wavy microstructure is formed.
At Pl = 1000 W, a completely dense remelted layer is formed
with sporadic pores at the border between the remelting zone
Fig. 3. Roughness Ra as a function of the energy per unit length Edis with a and the base material.
track overlap of TO = 80%.

For a laser power of 700 – 900 W with increasing Edis, the

surface roughness decreases with decreasing beam velocity vs.
At Pl = 1000 W and a laser beam intensity of 4.95 kW/mm², the
resulting roughness Ra is almost independent of Edis. Overall,
the roughness in hatching direction (perpendicular to the
fabrication direction) results in lower roughness values
compared to in fabrication direction. The differences in
roughness between the measurement directions also decrease
with increasing energy per unit length. The highest roughness
reduction of 85.2% to Ra = 5.23 µm perpendicular to FD is
achieved with Edis = 20.0 kJ/m. Over all, the roughness
variation within a polishing field, given by the scattering bars,
is decreasing with higher laser powers and laser beam
intensities.
The surface topography at a laser power of 1000 W shows a Fig. 5. Surface structure depending on the laser power Pl and beam velocity
closed surface with linear elevations up to 100 µm in hatching vs.
direction especially at the first hatching lines, resulting in an
increased roughness Ra perpendicular to the scanning direction, The remelting depth is in the range of 90 - 160 µm, see cross
see Fig. 4 a), b) position 2. At the start and end positions of the section a) in Fig. 6. With insufficient laser beam intensity at a
hatching lines, a surface depression exists, see Fig. 4 a), b) laser power Pl of 700 W, the L-PBF surface structure is
position 1. In the investigations, the beam velocity has no incompletely melted, which can be seen in the still present
significant influence on the created surface topography. porosity in Fig. 6 b). This leads to insufficient wetting of the
Surface remelting with 700 W results in a rough, non-closed melt pool with the solid base material, resulting in balling
surface with periodic spherical or linear material effects, tearing of the weld tracks and individual material
accumulations, regardless the varied beam velocity. accumulations and thus an increase in surface roughness.

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.
 Patrick Grub et al. / Procedia CIRP 111 (2022) 684–688 687

Fig. 6. Cross-section of the remelting zone polished with a beam velocity vs of

50 mm/s; a) Pl = 1000 W; b) Pl = 700 W. Fig. 8. Surface topography after laser polishing depending on the track
overlap TO with a laser power Pl of 1000 W and a beam velocity vs of
3.3. Influence of track overlap 50 mm/s.

The influence of the track overlap TO and the energy density At laser powers of 900 W and less, the track overlap has no
(energy input per area) ED on the surface roughness Ra is given influence on the generated surface microstructure, see Fig. 9.
in Fig. 7. Between a TO of 60 - 80%, the achievable surface With Pl = 1000 W and increasing TO, the residual
roughness is almost unchanged. At TO = 90%, the differences microstructure decreases.
between the measuring orientations increases at energy
densities ED of 280 - 360 J/mm².

Fig. 9. Surface structure depending on the laser power Pl and track overlap
TO with vs = 50 mm/s.

4. Conclusion

Due to high surface roughness and structure depth, laser

polishing of additive manufactured pure copper parts requires
Fig. 7. Roughness Ra depending on the energy density under varying track high laser beam intensity of at least 4.95 kW/mm² to create a
overlap TO with vs = 50 mm/s. stable remelting zone with a closed remelting layer. Insufficient
beam intensities below 4 kW/mm² results in incomplete
At the highest ED of 400 J/mm², the lowest surface roughness remelting of the surface structure and cause unstable wetting of
of Ra = 3.2 µm perpendicular to FD, is achieved. the melt pool on the bulk material below, leading to
With increasing track overlap, the surface topography shows inhomogeneous material accumulation and increased
lower linear surface elevations and material accumulations at a roughness. For laser power below 900 W, increasing energy
sufficient laser power of 1000 W, see Fig. 8. Especially with
density with higher track overlaps has no effect on the created
90% track overlap the flatness is strongly improved. At TO
surface layer and the resulting roughness. At Pl = 1000 W,
greater than 60%, high linear material accumulation occurs
slight roughness improvements with TO = 90% to the lowest
along the first scan lines at the beginning of the polishing
process, position 1 Fig. 8. achievable roughness of Ra = 3.2 µm can be realized. The
created remelting zone is fully dense and exhibits a depth of
90 - 160 µm.

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.
 688 Patrick Grub et al. / Procedia CIRP 111 (2022) 684–688

Acknowledgements [6] Lee, J.-Y., Nagalingam, A.P., Yeo, S.H., A review on the state-of-the-art
of surface finishing processes and related ISO/ASTM standards for metal
additive manufactured components; Virtual and Physical Prototyping,
The authors would like to thank the toolcraft AG for 2021, 16, p. 68.
manufacturing the samples. The work was supported by the [7] Willenborg E., Polieren von Werkzeugstählen mit Laserstrahlung;
Ministry of Science, Research and the Arts Baden- Technische Hochschule Aachen, 2006.
[8] Richter, B., Blanke, N., Werner, C. et al., Effect of Initial Surface Features
Württemberg Geräteprogramm 2019. on Laser Polishing of Co-Cr-Mo Alloy Made by Powder-Bed Fusion; JOM,
2019, 71, p. 912.
References [9] Li, Y., Zhang, Z., Guan, Y., Thermodynamics analysis and rapid
solidification of laser polished Inconel 718 by selective laser melting;
Applied Surface Science, 2020, 511
[1] Tran T.. Chinnappan a.; Lee J. et al. 3D Printing of Highly Pure Copper;
[10]Meylan, B., Calderon, I., Le, Q.T. et al., Investigations of surface defects
Metals, 2019, 9,p. 756
during laser polishing of tool steel; Procedia CIRP, 2020, 94, p. 942.
[2] Hummel, M., Schöler, C., Häusler, A. et al. New approaches on laser micro
[11]Černašėjus, O., Škamat, J., Markovič, V. et al., Surface Laser Processing
welding of copper by using a laser beam source with a wavelength of 450
of Additive Manufactured 1.2709 Steel Parts; Advances in Materials
nm; Journal of Advanced Joining Processes, 2020, p. 100012.
Science and Engeneering, 2019, p. 1.
[3] Roccetti Campagnoli, M., Galati, M., Saboori, A., On the processability of
[12]Ma, C.P., Guan, Y.C., Zhou, W., Laser polishing of additive manufactured
copper components via powder-based additive manufacturing processes:
Ti alloys; Optics and Lasers in Engineering, 2017, 93, p. 171.
Potentials, challenges and feasible solutions; Journal of Manufacturing
[13]Chen, L., Richter, B., Zhang, X. et al., Effect of laser polishing on the
Processes, 2021, 72, p. 320.
microstructure and mechanical properties of stainless steel 316L fabricated
[4] Jiang, Q., Zhang, P., Yu, Z. et al., A Review on Additive Manufacturing of
by laser powder bed fusion; Materials Science and Engineering, 2021, 802
Pure Copper; Coatings, 2021. 11, p. 740.
[14]Hofele, M., Roth, A., Schanz, J. et al., Laser Polishing of Additive
[5] van Hooreweder, B., Lietaert, K., Neirinck, B. et al., CoCr F75 scaffolds
Manufactured Aluminium Parts by Modulated Laser Power;
produced by additive manufacturing: Influence of chemical etching on
Micromachines, 2021, 12, p. 1332.
powder removal and mechanical performance; J Mech Behav Biomed
Mater, 2017, 70, p. 60.

This is a resupply of March 2023 as the template used in the publication of the original article contained errors. The content of the article has remained unaffected.