Surface & Coatings Technology 422 (2021) 127492

Contents lists available at ScienceDirect

Surface & Coatings Technology

journal homepage: www.elsevier.com/locate/surfcoat

Investigation of laser powder bed fusion manufacturing and post-processing

for surface quality of as-built 17-4PH stainless steel
Kyung-Tae Yang a, b, 1, Min-Kyeom Kim a, 1, Dongwon Kim a, Jonghwan Suhr a, *
a
Department of Mechanical Engineering, Sungkyunkwan University, Suwon 440-746, South Korea
b
Mobile Communications R&D Center, LG Electronics, 10, Magokjungang 10-ro, Gangseo-gu, Seoul 07796, South Korea

A R T I C L E I N F O A B S T R A C T

Keywords: Laser powder bed fusion (LPBF) technique enables to manufacture complex parts and difficult-to-cut materials
Laser powder bed fusion with saving the material costs and manufacturing time. However, there still has been technical challenges and
Appearance quality qualities issues for LPBF products to be used in engineering applications. In order to obtain the high-quality
Surface roughness
surface products including minimal defects and voids, this study systematically investigated and determined
Gloss
Touch sensation
LPBF manufacturing process and post-processing using a 17-4PH stainless steel. It was experimentally found that
Mechanical polishing design of LPBF process conditions could only attain the densification of as-built products, while surface qualities
Physical vapor deposition were not as good as commercial products. Therefore, the post-processing was indispensably explored by con­
ducting mechanical polishing, physical vapor deposition coating, and hot isostatic pressing for the high surface
quality and densification of as-built specimens. The surface qualities of the specimens, fabricated with optimum
volumetric energy density of 97.66 J/mm3, were carefully characterized for their surface roughness, gloss,
surface energy and frictional behaviors. After the post-processing, the surface roughness and gloss were
dramatically improved by more than 800% over the as-built parts, and the porosity was reduced by up to 57.8%.
This study demonstrates that optimum LPBF process and post-processing could enhance the qualities of LPBF
products up to the same or beyond level of commercial products, comparing the defects, roughness, gloss, and
touch sensation. The LPBF technique coupled with post-processing, if optimized, can show the great potential for
the use in a wide variety of real-world engineering applications including mobile phones, jewelry, electronic
appliances, artificial teeth and bones.

1. Introduction characteristics, low surface roughness after surface treatment, and high
strength compared with polymers. These materials, particularly stain­
Additive manufacturing (AM), one of enabling technologies of the less steel and aluminum, are expected to be fabricated with the afore­
fourth industrial revolution, has significant potential for the advance­ mentioned properties if they are additively manufactured under suitable
ment of manufacturing technologies owing to its advantages that process conditions [1,2]. Clearly, the cost and time for designing exte­
traditional manufacturing may not offer. This technology can provide rior metal products with good surface quality will be saved owing to the
much more flexibility of the design freedom of products, increase advantages of metal AM.
component performance along with lightweight, and reduce material However, metal AM has limitations in terms of the low surface
costs and time. Recently, metal AM, also known as metal three- quality and defects, caused by the uncontrollable heating and cooling
dimensional (3D) printing, has been intensively studied. Considerable process of metals by a localized heat source [3]. One of the metal AM
efforts have been made in order for this technology to be employed in technologies, laser powder bed fusion (LPBF) process, which is used in
practical applications such as hand-held and wearable devices. this study, melts and fuses dispersed powders using a laser. The process
The metals investigated in this study have been largely used in the parameters of the method are laser power P (W); scanning speed v (mm/
exterior parts of tableware, accessories, watches, cellphones, and auto­ s); line spacing h (mm), which is the distance between the laser lines;
mobiles, because they present luxurious and glossy surface quality and layer thickness d (mm). Volumetric energy density (VED) can be

* Corresponding author.
E-mail address: [email protected] (J. Suhr).
1
These authors contributed equally to this work.

https://doi.org/10.1016/j.surfcoat.2021.127492
Received 6 June 2021; Received in revised form 4 July 2021; Accepted 5 July 2021
Available online 10 July 2021
0257-8972/© 2021 Elsevier B.V. All rights reserved.
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

calculated by combining these parameters as follows: VED = P/(v × h × surfaces.

d) (J/mm3). The LPBF manufacturing process can induce pores, thermal In contrast, the mechanical polishing process used in this study can
deformation, cracks, and delamination in the as-built parts if the process achieve good surface finishing without developing a higher compressive
conditions are unsuitably designed [4–6]. This is because a localized residual stress than milling and grinding; moreover, it completely
heat source at high scan speeds develops a high temperature gradient, eliminates some defects, such as pores and cracks, formed during an
high cooling rate, and large Marangoni effect, which result in unpre­ LPBF process [25]. Mechanical polishing processes are typically con­
dictable particle movements around the melt pools, melt pool flow, ducted in industries manually by skilled workers using rotating abrasive
pores, denudation zones, and residual stress [7,8]. Therefore, because wheels. The surface quality depends on the proficiency of the worker
these defects, such as pore formation, unstable melt pool, and residual because the worker frequently examines the appearance status with the
stress, can induce microcracks, delamination, and chemical corrosion in naked eye, changes the position of the product, and polishes it. There­
the use environment, it is necessary to design process parameters (i.e., fore, currently, owing to the increasing use of metal parts in electronic
VED) [9–12]. products, automation of mechanical polishing is receiving considerable
It is well-reported that process conditions can also affect the focus for overcoming its limitation. However, the issue of designing
morphology of melt pools, which can determine the roughness of the constant pressure and time conditions and appropriate tools for treating
heterogeneous surface of as-built part. For example, an inadequate VED various specimens in the polishing process remains to be addressed.
leads to high surface roughness of the as-built parts, which can not only Plating is one of the methods for providing a characteristic to a
result in poor surface appearance but also degradation of the fatigue surface and changing its physical properties. It can be classified into two
properties [13,14]. Therefore, the surface morphologies and roughness types, namely hydrometallurgy (called wet plating), using liquid as the
of additively manufactured parts have been investigated to determine plating material, and pyrometallurgy (called dry plating), using a gas or
the process effect on their surface quality. plasma. Pyrometallurgy, which is used in this study, is composed of
Tonelli et al. [15] reported that the effect of VED on the surface chemical vapor deposition (CVD) and physical vapor deposition (PVD),
quality of the as-built parts is dominant in the LPBF process. At low VED which has been mainly used in the surface treatment of metals [26].
(≤100 J/mm3), the surface quality was low owing to the lack of heat Deposition of hard coatings on surfaces by the PVD process prevents the
energy; consequently, the powders did not appropriately melt, and the wearing of cutting, forming, and blanking tools, thereby significantly
melt pool could not be fully wet on the baseplate. Hence, high rough­ enhancing their durability. Using the PVD method, Chowdhury et al.
ness, balling effect, and high porosity caused by the lack of fusion were [27] characterized the wear and lubrication mechanisms of metals and
observed after the LPBF process completion. Conversely, Shrestha et al. obtained the wear resistance by coating a CrN monolayer on a Ti6Al4V
[16] noted that a high VED produces a steam column at the laser scan surface. Furthermore, PVD coating has potential as a postprocessing
track and increases the number of pores. It was concluded that the method to achieve high-gloss parts with colors. For example, currently,
appropriate VED should be determined using the laser power, scan the stainless steel external parts of smartphones are easily treated with
speed, line spacing, and layer thickness because a change in the process different colors after the PVD coating process.
parameters might lead to severe defects and deterioration of the part Previous studies reported that VED can be the most important factor
quality. In another case, Dong et al. [17] found that an increase in the determining the surface quality including surface roughness, irregular
line spacing from 75 μm to 200 μm had a detrimental effect on the surface, gloss, and molten pool stability as well as defects such as pores,
surface quality of STS 316L along with a gradual surface unevenness cracks, and delamination in the AM process [12,28,29]. In addition, the
caused by the insufficient melting. A relatively low surface roughness surface quality of the as-built parts was empirically studied by varying
was measured at the line spacing of 75–100 μm, which subsequently process parameters [25,30]. However, there seems to be no detail report
rapidly increased, particularly when the line spacing was ≥150 μm, available on the surface quality of the 17-4PH stainless steel parts
reducing the overlap rate of the molten pools. Qiu et al. [18] noted that investigating the process parameters along with post-processing.
porosity and surface roughness consistently increase when the scanning Here, this study systematically investigates the LPBF manufacturing
speed increases from 2300 mm/s to 3500 mm/s. They also observed that conditions and three different post-processing techniques, including
an increase in the layer thickness (from 20 to 40 μm) exerts deleterious mechanical polishing, PVD coating, and HIP, in order to obtain the high
effects on the surface quality, such as unstable molten pools (necking surface quality as well as minimal defects of as-built 17-4PH stainless
and balling), irregular surfaces, and pore formation, during the process. steel. First, the range of LPBF process conditions was determined by
Surface finishing has been extensively used in the postprocessing of examining the melt pool overlap of one-layer specimens for building as-
exterior metal parts in several industries. Metal AM has a limitation in built products. The LPBF process conditions within the range were
the simultaneous application of additively manufactured as-built parts optimized with as-built products, characterizing the surface morphol­
in different industries (biomedical engineering, aerospace, and electrical ogies and densities. Next, the surface quality of the specimens is care­
appliance) because of their nonuniformity and high surface roughness. fully characterized and evaluated after post-processing. It was
Hence, following metal AM, a surface finishing technique, such as confirmed that the optimized process parameters coupled with the
electrical discharge machining, grinding, milling, mechanical polishing, properly designed post-processing can dramatically improve the surface
electropolishing, or laser polishing, is indispensable for obtaining high quality such as surface roughness, gloss, touch sensation along with high
surface quality of AM products. density of the as-built parts. It is believed that this study can contribute
Electrochemical polishing (ECP) treatment has been studied for to the realization of the metal 3D printing technique in various practical
obtaining high surface quality of additively manufactured parts by the applications including mobile phones, jewelry, electronic appliances,
removal of the rough surfaces. It was observed along with surface artificial teeth and bones.
roughness reduction, surface reflectivity and gloss were improved
[19–21]. Chang et al. [22] and Tyagi et al. [23] improved the surface 2. Materials and methods
quality of an (STS316 lattice) structure by applying ECP, observing the
surface quality level using an optical profilometer. Heintze et al. [24] 2.1. Materials
designed polishing process conditions, such as polishing pressure and
time, to reduce surface roughness and found a negative correlation be­ Stainless steel is relatively less expensive than titanium alloys and
tween gloss and roughness under various polishing process conditions. more suitable for additively manufactured exterior parts than aluminum
However, ECP has a limitation in meeting the surface quality re­ alloys, which are susceptible to hot cracking during the AM process [32].
quirements for high-end products; in addition, there is a risk of Among them, STS630 or 17-4PH (precipitation hardening) stainless
damaging the exterior components of such products while treating their steel contains a high fraction of austenite steel with martensite, showing

2
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

intrinsic characteristics; the precipitation of the highly dispersed nano­ Table 2

scale Cu-rich particles enhances its properties compared to those of LPBF process parameters for fabrication of single-layer specimens.
other austenite stainless steels [31]. Therefore, 17-4PH (Sandvik Osprey No. Laser Scan speed Line spacing Layer thickness VED (J/
Powders), which has excellent properties (corrosion resistance, wear power (W) (mm/s) (mm) (mm) mm3)
resistance, thermal resistance, high strength, and hardness) among most 1 250 400 0.08 0.04 195.3
stainless steel alloys, was selected in this work. Table 1 shows chemical 2 250 500 0.08 0.04 156.3
compositions, supplied by the powder company (Sandvik), and particle 3 250 600 0.08 0.04 130.2
size distribution of 17-4PH stainless steel powders, measured by both 4 250 800 0.08 0.04 97.7
5 250 1100 0.08 0.04 71.0
optical microscope (OM) and scanning electron microscope (SEM). It 6 250 1500 0.08 0.04 52.1
was confirmed that the powders of which size was 20 ± 9.59 μm were 7 250 2200 0.08 0.04 35.5
appropriate for LPBF manufacturing. 8 110 800 0.08 0.04 43.0
9 180 800 0.08 0.04 70.3
10 250 800 0.08 0.04 97.7
2.2. LPBF process and surface characterizations 11 320 800 0.08 0.04 125.0
12 390 800 0.08 0.04 152.3
13 460 800 0.08 0.04 179.7
The LPBF equipment used in this study was an SITI-SLM 250
(Shanghai Industrial Technology Institute-SITI), which can manufacture
as-built parts of sizes smaller than 250 × 250 × 680 mm3. The process manufactured, following the succeeding layer (N + 1th layer) deposited
parameters (laser power and scan speed) were designed to investigate with different building orientations (− 45◦ , +45◦ , − 45◦ , +45◦ , …). The
the printability of the alloy and set the line spacing as 0.08 mm density of as-built parts were measured using an AS3Y balance based on
considering a melt pool width of 49.37 μm–240.04 μm to investigate the Archimedes' principle as defined in ASTM B962-08. The appropriate
heat energy effect on the surface quality, as listed in Table 2. The layer VED range in manufacturing products is derived from the evaluation of
thickness was set as 0.04 mm because the 99% of the 17-4PH powder the printability of single layers. Therefore, a low risk of microcrack
sizes, characterized via OM and SEM images of the powders, were below development around pores was estimated for the optimum parts that
0.04 mm (Table 1). The designed thickness was expected to ensure that were designed and manufactured in this study [35].
the recoater disperses the powders well and increases the powder After quality assurance, the cubic blocks were observed by exam­
packing density on the baseplate or previous layer based on other studies ining the three-dimensional surface morphology, similar to the single-
[33,34]. layer specimens. The morphology of the molten pool overlap and the
Therefore, the laser power range was designed from 90 W to 490 W surface roughness were investigated to determine the process effect on
at a scan speed of 800 mm/s, and the scan speed was ranged from 400 the appearance quality. Heterogeneous peak–valley lengths at the sur­
mm/s to 2200 mm/s at a laser power of 250 W for a line spacing and a faces were quantitatively obtained by examining the three-dimensional
layer thickness of 0.08 mm and 0.04 mm, respectively (Table 2). Single- surface morphology to yield the process conditions for the mechanical
layer specimens were fabricated with dimensions of 10 × 10 mm2 using polishing, which is required because of their poor appearance quality.
different process parameters (VEDs) to evaluate their 3D printability, as
shown in Fig. 1a. Printing quality characteristics, such as molten pool 2.3. Surface finishing processes
characteristics, overlap rate of the molten pools, and carbonization at
the molten pools, were investigated by examining the three-dimensional An as-built part cannot be immediately used as an exterior part
surface morphology, because, as previously reported [12,28,29], eval­ because of its high surface roughness and poor surface quality. There­
uating the printing quality is important for preventing process failure fore, to use the additively manufactured parts as commercial external
during AM. products, surface finishing processes (mechanical polishing and PVD
Appearance quality was determined by obtaining the surface un­ coating) were employed for improving their appearance quality. Herein,
evenness, gloss, and porosity of a specimen using a microscope and a an automatic polishing machine was designed for the test specimens to
wide-area 3D measurement system (VR-3000, Keyence) with contactless exclude the effect of the proficiency of a worker and to ensure constant
optics. To collect 3D data of a surface, structured light is emitted from a pressure, speed, and time and appropriate tools. It was used for surface
transmitter lens and projected onto the surface. Subsequently, the re­ treatment following fine-tuning of the polishing process, and each pro­
flected light is detected by a receiver lens and appears banded and bent cess condition of surface finishing was optimized by characterizing the
based on the changes in the topography of the surface. Finally, the obtained surface morphologies.
triangulation method is used to calculate and measure the height of the First, the volume of material removal during the polishing process
surface. The obtained 3D data (1.9 × 1.4 mm2) were analyzed to design was determined considering the peak–valley lengths on the surfaces of
the process parameters for the multilayered parts to be additively the as-built parts. All cubic blocks fabricated by different parameters
manufactured. (VED ranging from 70 to 100 J/mm3) showed different surface rough­
The printable parameters, which are discussed below, were deter­ ness and peak–valley lengths. Therefore, the polishing process was
mined from the initial parameters (laser power 90–490 W and scanning needed to be designed with appropriate polishing conditions (i.e., con­
speed 400–2200 mm/s) based on the analysis of the 3D printability of tact force, rotating speed, and polishing time).
the single-layer specimens (Table 2). Cubic blocks of 25 mm length were The contact force of the parallel-type spindle head was adjusted in six
additively manufactured under the appropriate VED range of 70–100 J/ steps (50–300 gf) based on the stability of the machine, as listed in
mm3, as listed in Table 3 (Fig. 1a). The as-built blocks were fabricated Table 4. It was observed that the stability of the machine and the pol­
following the scanning strategy (Fig. 1b). Specifically, after scanning the ishing quality were low under low contact forces (50, 100, 150, 200 gf)
infill of the Nth layer using the laser, a contour of the layer was additively compared to those under the high (300 gf) contact force because of the

Table 1
Measured size and chemical compositions of 17-4PH stainless steel powder.
Type Size (μm) Chemical compositions (%)

Fe C Cr Si Mn S P Cu Nb Ni

17-4PH stainless steel 20.01±9.59 Bal. 0.07 15.5–17.5 1.0 1.0 0.04 0.03 3.0–5.0 0.15–0.45 3.0–5.0

3
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

Fig. 1. (a) photographic images of single-layer and multilayer specimens and (b) Schematic of multilayer specimen fabricated by LPBF.

Therefore, rough polishing was conducted using two types of sand pa­
Table 3
pers (120 and 240 grit sandpaper) until surface quality sufficient for
LPBF process parameters for fabrication of multilayer specimens.
high polishing was achieved (Table 4). The as-built specimens were
No. Laser Scan speed Line spacing Layer thickness VED (J/ polished with 120 and 240 grit sand papers for 40 s, respectively, while
power (W) (mm/s) (mm) (mm) mm3)
being repeatedly rotated at 0◦ and 90◦ . Bumps and pores on the surfaces
1 220 800 0.08 0.04 85.9 were eliminated; subsequently, they were treated by the high polishing
2 200 800 0.08 0.04 78.1
method to improve their appearance quality characteristics, such as
3 180 800 0.08 0.04 70.3
4 250 800 0.08 0.04 97.7
gloss and smoothness. In high polishing, the spindle head was equipped
5 250 830 0.08 0.04 94.1 with a wool wheel, using abrasives made by blending blue-colored
6 250 950 0.08 0.04 82.2 oxychromium with beef tallow. Each specimen was treated to three
7 250 1100 0.08 0.04 71.0 stages (grinding times of 40 s, 60 s, and 80 s), and all surface-treated
specimens were compared based on their 3D surface data, as shown in
vibration of the machine caused by the rough surface (Fig. 2a). There­ Fig. 3a–e.
fore, the contact force was optimized to 250 gf, and the rotating speed
was set as 1500 rpm in the range of 1000–3200 rpm, based on the sur­
2.4. PVD coating
face uniformity and stability.
The polishing process employed in this study can be roughly divided
PVD coating achieves the most stable and highest quality coating
into two parts, namely rough and high polishing. Most of the uneven
among all plating methods. PVD coating ensures a dense surface and a
surfaces formed by the melt pools were removed by rough polishing
high adhesion of layers, forming thin film coatings of less than 1 μm
considering the peak–valley lengths. If the spindle head could not fully
thickness having various colors. Various physical properties, such as
cut the surface up to the valley points, pool-like defects were observed.
color, gloss, and abrasion, can be characterized for additively

Table 4
Mechanical polishing processes using sand papers and wool wheel.
Method Level Rough polishing High polishing

120 grit sand paper 240 grit sand paper Wool wheel (abrasive)

Rough polishing Step 1 40 s

Step 2 40 s 40 s
High polishing Step 1 40 s 40 s 40 s
Step 2 40 s 40 s 60 s
Step 3 40 s 40 s 80 s

Fig. 2. Schematics of (a) mechanical polishing and (b) PVD coating.

4
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

Fig. 3. Schematics of surface roughness (a) before and (b–e) after surface finishing processes.

manufactured surfaces depending on the target materials (Ti, Zr, and differences between the target and standard stainless surfaces (95.6 GU).
TiAl) and gases (N2, O2, and C2H2) in the PVD process. PVD has been The gloss values of the surface-treated parts and commercial products
mainly used for the surface treatment of stainless steel exterior parts were quantitatively compared to their surface roughness.
owing to its advantages. Therefore, PVD coating was performed for
improving the surface quality and physical characteristics of the surface- 2.5. Contact angle and friction test
treated 17-4 PH blocks.
PVD coating was performed by breaking down the target material Contact angle and friction coefficient were measured to quantita­
(Ag) into atoms or molecules by first heating or colliding it with Ar+ tively evaluate the surface texture and the stick–slip, which determine
ions. Subsequently the atoms or molecules were condensed and solidi­ the touch sensation of an exterior part. The contact angle was measured
fied on the surface of the product to form a thin film, as shown in Fig. 2b. using drop shape analyzer (DSA25, KRÜSS Instruments) to estimate the
A plasma was developed by injecting inert argon gas into a vacuum adhesion of stainless steel on human skin. Approximately 2 μL of water
chamber and applying power to the target side. The Ar+ ions generated was dropped on a surface more than five times at room temperature, and
in the plasma were irradiated at the target, and the charge density dif­ the average contact angle between the surface and the droplet was used
ference caused the target to eject under the shock energy. The molecules to examine the stick–slip phenomenon between human skin and stain­
of the target material and nitrogen gas were injected into a chamber. A less steel parts.
compound (AgN) was finally deposited on the surface of the product. Friction tests were conducted using a friction coefficient tester
The PVD color of the surface rarely disappears over time, and PVD can (FPT–F1, Labthink) to measure the friction coefficients of the parts, as
achieve a uniform and higher appearance quality than an electro­ shown in Fig. 4a. Fig. 4b presents the friction tester and the method in a
chemical process. schematic form. A silicon sheet in contact with the as-built surface was
Surface gloss is one of the important quality characteristics used to evaluate the surface quality, because the resultant surface
(smoothness, gloss, color, hardness, and strength) of exterior metal characteristic and compression resistance of human skin are almost
parts. In this study, a high-gloss surface was obtained by improving the identical to the silicon sheet. Thus, the touch sensation, in terms of the
surface texture by adopting a surface finishing technology (polishing stick–slip phenomenon on the surface, was quantitatively analyzed
and PVD coating). The gloss of the surface-treated part was measured based on the friction coefficient and the F–d curves. Friction coefficients
using a gloss meter (Micro Gloss Meter, BYK-Gardner) and compared were calculated using the equation, μ = F/P, where normal force P is 0.2
with that of the non-treated one in terms of the measurement scale, gloss kgf and shear force F increases up to the critical value, F. The frictional
units (GU), of the gloss meter. The gloss of the target surface was behaviors of the products before/after postprocessing could be overall
calculated by varying the input while keeping the reflective light toward compared with commercially available parts, considering the surface
the target at an angle of 60◦ . The gloss values were expressed as the gloss roughness, characteristic, and lubricant.

5
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

Fig. 4. (a) Photographic images and (b) schematic of friction measurement.

3. Results and discussion lute deviation from the profile mean line, Ra (μm), which can be
expressed as
3.1. As-built surface morphologies ∫b ∫b ∫b
(Ra )x = |Z(s) |ds, (Ra )y = |Z(s) |ds, (Ra )xy = |Z(s) |ds
3.1.1. Melt pool characteristics with single layer a a a

The single-layer surfaces fabricated under various process parame­ The surface roughness was measured five times in x, y and xy di­
ters, as listed in Table 1, were characterized. The melt pool morphol­ rections (Fig. 1a). It decreased from 3.1 μm to 1.58 μm when the laser
ogies and the surface contours of the single layers were characterized via power increased up to 320 W at 800 mm/s of scanning speed. After the
3D microscopic and SEM images. The melt pool width and height, which surface was carbonated and contaminated with several sputters and
were examined from the data, were important indicators for designing ejected powders, the average surface roughness increased by 2.28 μm.
the process parameters, such as the laser power and the scan speed, This was because the contamination of the surface increased the surface
when the line spacing was set as 0.08 mm. To manufacture defect-free roughness, even if the high energy density induced flat melt pools.
as-built parts, the overlap rates and the melt pool morphologies were Therefore, the deviation of the surface roughness increased with
considered for determining the printable region, which excludes the increasing average roughness at a high VED. In contrast, the surface
regions of non-overlap and carbonate surface. roughness steadily decreased by 1.58 μm when the scanning speed
In designing the process parameters, the overlap rates between the decreased up to 400 mm/s at 250 W of laser power. The high intensity of
melt pool tracks were considered under different laser powers and the laser beam with the low scanning speed leads to a poor surface,
scanning speeds, as shown in Fig. 5a, b. Compared with an appropriate showing black and several sputters and ejected particles.
VED of 97.7 J/mm3 (Fig. 6a), a low VED of less than 70 J/mm3 induced a To build commercial products using LPBF, based on the quality of the
narrow melt pool, which was insufficiently wide to overlap the previous single-layer specimens, a VED of 70–125 J/mm3 was found to present
melt pool at 0.08 mm line spacing. In addition, the metal powders were appropriate ranges of the parameters; Fig. 6a shows the appropriate
insufficiently melted, resulting in unstable melt pool characteristics, surface conditions (Fig. 5a, b). The as-built parts, fabricated with the
such as balling or necking phenomenon (Fig. 6b, c). Contrastingly, a derived VED ranging from 70 to 125 J/mm3, are expected to require
high VED of more than 125 J/mm3 carbonized the metal powders, evaluation and improvement by postprocessing for obtaining high sur­
making the melt pools flat and black. Sputters and ejected particles, face quality characteristics.
which deteriorate the quality of additive manufactured parts, were
observed on the melt pools at a high VED (Fig. 6d, e). 3.1.2. As-built surfaces of blocks
The surface roughness of a single layer was characterized by exam­ As-built cubic parts of sizes 25×25×25 mm3 were fabricated with the
ining the three-dimensional surface morphology using the mean abso­ designed VED ranging from 70.3 to 97.7 J/mm3, as listed in Table 2. The

Fig. 5. Surface roughness of single-layer specimens fabricated using (a) different laser powers at 800 mm/s of scanning speed and (b) different scanning speeds at
250 W of laser power.

6
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

Fig. 6. SEM images of single-layer specimens fabricated with (a) appropriate VED: 97.65 J/mm3, (b, c) low VED: 35.51, 35.16 J/mm3, and (d, e) high VED: 191.41,
195.31 J/mm3.

as-built surfaces of the parts were characterized by examining the three- energy density was sufficient for melting the powders on the previous
dimensional surface morphology. The surfaces were nonuniform and layers, thereby increasing the width/height of the elliptical melt pool
presented a roughness deviation because of the differences in the [38].
accumulated heat energy and the unevenness of the surfaces from the However, the density increased by 2.2% as the laser power increased
repetitive heating and cooling cycles (Fig. 7a–g.-1). The distance be­ from 180 to 250 W, which is inversely proportion to the surface
tween a peak and a valley was in the range of 53.7–135.0 μm, which was roughness (Fig. 8a). The decrease in the scanning speed affected the
longer than the powder layer thickness of 40 μm (Fig. 7a–g.-2). decrease in the surface roughness by 12.94 μm (Fig. 8b). The large melt
High peak–valley points are commonly explained in the literature pool size by the high VED prevented pores, called as lack of fusion, from
based on the above reasons. The low cooling rate and high absolute being formed between the melt pools at line spacing of 0.08 mm and
temperature at an as-built surface, as the layers stack from the substrate, layer thickness of 0.04 mm. The as-built parts were fabricated along
increase the melt pool size and change the melt pool geometry [36]. The with properly overlapped melt pool, without carbonization. The average
denudation caused by particle movement arises under a metal vapor surface roughness was measured to stay within the range of
flow and ambient gas pressure, and it deteriorates the surface quality 12.94–15.92 μm, which is quite favorable for the use in practical
characteristics of the initial layer because of the nonavailability of applications.
powders for depositing subsequent melt pools [37]. Spattering during Therefore, postprocessing to improve the roughness and relative
the process occurs when the powder particles entrained in the flow of densities of the as-built parts was necessary to ensure their use in
the inert gas are ejected away from the melt pool, as shown in Fig. 6b–e. commercially available products. A surface of 135.0 μm or more must be
Spattering degrades the quality of the as-built surfaces because of the cut and polished to obtain the desired appearance quality considering
requirement of a higher powder layer thickness to manufacture the the roughness and peaks of the as-built blocks. The as-built parts fabri­
designed layer thickness [36]. cated with a VED of 97.66 J/mm3 were postprocessed by heat treatment,
Abnormal events gradually increased the surface roughness and surface finishing, and coating for obtaining high-quality products.
peak–valley distance at the top of an as-built part; concurrently, the
roughness at its side was highly dependent on the layer thickness, 3.2. Appearance quality characteristics after surface finishing processes
showing an average value of 15.3 μm. The peak–valley distances at the
top of the as-built surfaces increased from 53.7 μm to 135.0 μm, reaching 3.2.1. Surface roughness and morphologies
a maximum when the laser power decreased from 250 W to 180 W at a Surface treatments, such as mechanical polishing and PVD coating,
scanning speed of 800 mm/s (Fig. 7a–d.-2). The distance also gradually were conducted on the as-built surfaces fabricated using various process
increased from 53.7 μm to 105.4 μm with increasing scanning speed parameters. The as-built surfaces were polished with the selected sand
from 800 mm/s to 1100 mm/s at the laser power of 800 W (Fig. 7a, e–g.- papers and the wool wheel for rough and high polishing, respectively.
2). A low VED as a function of the laser power and the scanning speed The surface roughness of the products based on the laser powers of 180,
exerts a deleterious effect on the surface roughness. This is because the 200, and 220 W were almost the same (3.16, 3.3 and 2.98 μm) after a
insufficient energy to melt the powder layer results in the formation of rough polishing time of 40 s (Fig. 9a). This was because the highest peak
high peaks, balling, and low relative densities, which could deteriorate point of the as-built parts, ranging from 96.4 to 135.0 μm, was sufficient
the surface quality. to be cut with 120 grit sand paper in 40 s. Therefore, the rough polishing
The surface roughness of an as-built part was measured five times in time was designed as 40 s to prevent the surface defects caused by the
the x, y and xy directions (Fig. 1a). The surface roughness of an as-built insufficient polishing process. This time covers an entire as-built part
part steadily decreased by 83.6% and 81.3% when the VED increased fabricated with a VED ranging from 70.3 to 97.7 J/mm. Each surface
from 70.31 (71.02) J/mm3 to 97.66 J/mm3 at the selected laser power roughness and morphology after the rough and high polishing were
and scanning speed, respectively (Fig. 8a, b). This is because the high characterized, as shown in Fig. 9b, for obtaining high appearance

7
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

Fig. 7. (a–g.-1) Surface morphologies and (a–g.-2) profiles of as-built parts fabricated with different VEDs (laser power and scanning speed). (a) 97.66 J/mm3 (250
W, 800 mm/s), (b) 85.94 J/mm3 (220 W, 800 mm/s), (c) 78.13 J/mm3 (200 W, 800 mm/s), (d) 70.31 J/mm3 (180 W, 800 mm/s), (e) 94.13 J/mm3 (250 W, 830
mm/s), (f) 82.24 J/mm3 (250 W, 950 mm/s), and (g) 71.02 J/mm3 (250 W, 1100 mm/s) at 0.04 mm of layer thickness and 0.08 mm of line spacing.

quality. The roughness was improved by 1.8 μm after the high polishing, surface of an as-built part was improved by 856% in terms of the surface
during which the roughly polished surface was polished for 80 s with the roughness; therefore, it could be used for products, such as tableware,
wool wheel. accessories, watches, and cellphones, with luxurious and glossy surface
The surface roughness obtained by polishing with sand papers of 120 quality characteristics.
and 240 grit were 3.16 and 3.02 μm, respectively, showing a peak and However, defects were occasionally found on the polished surfaces
valley point (approximately 27 μm) (Fig. 10a). The surface roughness because of the pore distribution in the as-built parts after the LPBF
could be reduced by 1.8 μm after the high polishing with the wool wheel, process (Fig. 11a, b). The defects characterized with a surface contour
as shown in Fig. 10b. The surface was formed with a peak and valley had a width of 0.2 mm and a depth of 0.016 mm. The edges of the de­
point ranging from 0 to 9.6 μm, aligning to the polishing direction. The fects, which show bright colors in Fig. 11b, were overpolished during the

8
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

Fig. 8. Surface roughness and densities of as-built parts fabricated with (a) different laser powers (180, 200, 220, and 250 W) and (b) scanning speeds (800, 830,
950, and 1100 mm/s), as listed in Table 2.

Fig. 9. Surface roughness of (a) as-built and roughly polished parts fabricated with laser powers of 180, 200, and 220 W, and (b) surface roughness after rough/high
polishing and PVD coating.

Fig. 10. Surface morphologies and profiles of parts after (a) rough and (b) high polishing.

9
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

Fig. 11. Surface morphologies of parts after high polishing and PVD coating observed from (a) 3D microscopic images and (b) 2D microscopic images and (c) surface
profiles of the parts.

high gloss polishing process because the high pressure, induced by the the appearance quality characteristics but also the properties of the as-
difference in the contact areas, was applied to the boundary (Fig. 11c). built parts for exterior products.
The densities were 7.67, 7.61, 7.54, and 7.50 g/cc when they are
fabricated with laser powers of 250, 220, 200, and 180 W (VED of 97.7, 3.2.2. Surface gloss
85.9, 78.1, and 70.3 J/mm3) (Fig. 12a). The densities were also changed The surface gloss, which is affected by surface roughness and
as 7.67, 7.55, 7.50, and 7.51 g/cc when the scanning speed changes as coating, is one of the main appearance quality characteristics for
800, 830, 950, and 1100 mm/s (VED of 97.7, 94.1, 82.2, and 71.0 J/ designing exterior parts. Therefore, each surface was polished and
mm3), respectively. The low densities originated from the presence of treated with PVD coating to obtain high surface gloss, similar to
several pores of sizes approximately 42.35±54.57 μm in the as-built commercially available exterior parts, such as stainless steel watches
parts. and smartphones. The surface gloss of the as-built and surface-treated
The irregular shaped pores, as shown in Fig. 12b, were induced by parts were compared to commercially available parts, as shown in
the lack of fusion because of the inability of a low VED to fully melt a Fig. 13a. The surface gloss increased from almost 0 to 433 GU following
previous layer or an adjacent weld bead [39]. Spherical pores belong to rough polishing with 120 and 240 grit papers and high gloss polishing
another type of pores, which could be attributed to the entrapment of for 80 s. The surface gloss was inversely proportional to the roughness,
gas at a keyhole collapse and/or to the metallurgical pores in the representing a linear correlation with an R-squared of 0.890. The gloss
spherical particles (Fig. 12b). Such pores have been reported to cause quality of the polished parts was higher than that of the plastic exterior
the degradation of not only the appearance quality characteristics but parts of a mobile phone coated by ultraviolet (UV) curing or black/silver
also the tensile and fatigue properties. Therefore, the pores should be anodizing techniques. After high gloss polishing, the surface gloss was
eliminated or reduced by heat treatment to obtain high properties and enhanced by 800 GU, when the PVD coating was conducted on 17-4PH
appearance quality. stainless steel (Fig. 13b). A higher surface gloss could be achieved
In this study, the densities were increased via hot isostatic pressing compared to the gloss of a stainless steel watch (gloss of 540 GU).
(HIP), heating and keeping the temperature up to 1100◦ C and the The improvement in the surface gloss was also observed via the
pressure up to 1100 bar for 1 h, and subsequently cooling in air. Because reflective images displayed in Fig. 14a–c. Although the printed letters on
most of the irregular and spherical pore sizes were reduced to less than the floor could not be seen on the as-built surface, the letters began
24.47 μm after the HIP (Fig. 12b), defects on the surfaces could not be faintly appearing on the surface of 1.8 μm roughness after the high gloss
observed with the naked eye. The densities of the parts fabricated with polishing. Fig. 14c exhibits that the gloss of a surface is outstandingly
the different laser powers and printing speed were distributed from 7.50 enhanced by the PVD coating. The mechanical postprocessing and the
g/cc to 7.67 g/cc, which were improved subsequently by 7.70 g/cc after chemical coating process could achieve appearance quality parameters
the HIP (Fig. 12a). It could be estimated that the HIP enhanced not only better than those of commercial products in terms of the surface gloss,

Fig. 12. (a) Densities of HIP-treated and as-built parts fabricated with different powers and speeds, as listed in Table 2, and (b) microscopic images of pores
distributed on vertically cut parts.

10
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

Fig. 13. (a) Surface roughness and gloss of commercially available parts and additively manufactured parts after surface finishing process and (b) comparison of
surface gloss of parts.

Fig. 14. Surface gloss of (a) as-built part and parts after (b) high polishing and (c) PVD coating.

while improving the smoothness and texture of the parts.

3.2.3. Surface energy and friction

The touch sensation of products is determined by the physiological
and psychological perceptions when human skin contacts their surfaces
[40]. The sensation could be quantitatively investigated in terms of
tribological aspects to improve human perception of a smooth/rough,
slippery/grippy, or soft/hard surface. Smooth/rough, slippery/grippy,
and soft/hard sensations are highly related to the surface roughness,
friction coefficient, and hardness, respectively. In this study, the
smooth/rough and slippery/grippy sensations on the as-built surfaces
were examined before/after postprocessing. The surface characteristics
were investigated using the contact angle of water and diiodomethane
on the surface, which could determine the surface free, dispersive, and
polar energies. The surface energies were highly related to the stick–slip
friction on the surface.
The surface roughness of an as-built part is inversely proportional to
the contact angle of water because 17-4PH stainless steel is hydrophilic,
which causes more dispersion of the liquid on the surface under a
favorable contact with the surface (Fig. 15) [41]. The average contact
Fig. 15. Surface roughness and water contact angle on surface before/after
angle of water decreases from 98.57◦ to 83.4◦ after the PVD coating was
post processing.
conducted on the highly polished surface. The coating changed not only
the surface gloss but also the surface condition, decreasing the contact
angle (Fig. 15). The change in the roughness and the surface condition
influenced the frictional behaviors, which determined the touch
sensation.
γSG = γSL + γLG ∙cosθ (1)

11
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

Fig. 16. Contact angles and surface energies of (a) silicon, (b) roughly polished, (c) highly polished, and (d) PVD-coated parts.

(
γSL = γSG + γLG − 2 γd SG ∙γd LG
)1/2
− 2(γp SG ∙γp LG )1/2 (2) of the surface tension decreased by 12.94 m∙N/m under low roughness
because of the hydrophilicity of the surface. The PVD coating changed
( )1/2 the surface conditions, inducing an increase in the surface free energy to
γLG (1 + cosθ) = 2 γd SG ∙γd LG + 2(γp SG ∙γp LG )1/2 (3)
142.96 m∙N/m.
The increase in the dispersion term increases the total surface free
γSG = γdSG + γpSG (4)
energy because it contributes more to the total surface free energy that
(√̅̅̅̅̅̅̅̅̅̅̅̅̅̅ √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ )2 the polar term. Therefore, the surface free energy of the products
∆γ = (γSG )Si + (γ SG )Metal − (γSG )Si − (γSG )Metal (5) increased following the post processing, such as rough/high polishing
and PVD coating, increasing the work of adhesion, ∆γ, between the
The surface free energy can be expressed in terms of the dispersion silicon and product surfaces (Fig. 16a–d). The works of adhesion at the
and polar surface tension based on Young's and Owen− Wendt equa­ roughly and highly polished surfaces were 65.23 m∙N/m and 71.27
tions. Eq. (4) was derived when the Owen− Wendt equations, Eqs. (2), m∙N/m, respectively. The adhesion of the PVD-coated surface slightly
(3), were used in Young's equation (Eq. (1)). ∆γ is the change in the increased by 75.70 m∙N/m owing to the change in the surface condition,
interfacial free energy per unit area (work of adhesion) when the silicon and it is expected to promote the adhesion between silicon and the
and metallic surfaces contact in the friction test [42]. It was derived products while maintain the human sensation of a grippy surface.
from the calculated surface free energy of the silicon, imitating a human The kinetic coefficient and the maximum force were determined
finger, and from the products before/after postprocessing. from the F–d curves obtained from the friction tests. The friction be­
Comparing the surface conditions of the mechanically polished and tween the additively manufactured and silicon parts, which can imitate
coated products showed that the contact angle of diiodomethane a finger touch, was investigated to compare slippery/grippy sensation.
decreased to 47.83◦ , which implies an increase in the dispersed As postprocessing, such as mechanical polishing and PVD coating, is
component for the surface tension (Fig. 16b–d). However, the polar term conducted on the surface, the low roughness and the coating facilitate

Fig. 17. (a) Kinetic friction coefficients and maximum force of additively manufactured parts before/after surface finishing processes and (b) F–d curves of parts in
friction tests.

12
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

the stick phenomenon, as shown in Fig. 17b. The kinetic coefficients of which showed 16.05% increase in the work of adhesion when compared
the PVD-coated surface increased by 32.4%, comparing to those of the to the one of the rough polishing. The difference in adhesion energies
as-built surface. between as-built and PVD coating specimens increased the kinetic fric­
The maximum force of the PVD-coated surface, which is highly tion coefficients by 32.4%. It should be noted here that it is quite similar
related to the stick phenomenon, also increased by 38.7% compared to to those of commercially available stainless steel products.
that of the as-built surface (Fig. 17a). This is because the low roughness This study can demonstrate that the high surface quality of 17-4PH
and the PVD coating materials (AgN) on the surface increase the work of stainless steel as-built parts were achieved up to the commercially
adhesion, which is highly related to the increase in the friction coeffi­ available product-level of the surface roughness, gloss, touch sensation,
cient. The results indicated that low roughness causes a user to sense and density by coupling the optimization of additive manufacturing
grip and smoothness more than slipping and roughness. Comparing to a parameters along with properly designed post-processing techniques. It
commercially available STS sheet, the post-treated parts showed the can also show the feasibility of the use of LPBF technique in a wide
desired appearance quality characteristics—touch sensation and surface variety of real-world engineering applications including mobile phones,
gloss—only showing 10.8% difference in the kinetic friction coefficient. jewelry, electronic appliances, artificial teeth and bones.
Therefore, the 17-4PH stainless steel parts fabricated with 97.66 J/mm3
of VED and treated with heat treatment, mechanical polishing, and PVD CRediT authorship contribution statement
coating achieved high surface quality characteristics for being used in
commercial products. Kyung-Tae Yang and Min-Kyeom Kim contributed equally to this
work, preparing the specimens, doing the experiment, analyzing the
4. Conclusions data and writing the manuscript; Dongwon Kim assisted to do the
experiment; Jonghwan Suhr guided this work and give a financial
This study systematically investigated the LPBF manufacturing support.
conditions to determine the optimum process parameters for as-built 17-
4PH stainless steel with minimal defects/voids, and also to obtain the
high surface quality by conducting three post-processing techniques Declaration of competing interest
including HIP treatment, mechanical polishing, and PVD coating. The
surface quality of the specimens was carefully characterized and Jonghwan Suhr reports financial support was provided by Korea
evaluated. Ministry of Trade Industry and Energy. Jonghwan Suhr reports financial
The surface morphologies and conditions of single layer specimens support was provided by Korea Hydro and Nuclear Power Co Ltd.
under different laser powers and scanning speeds were characterized by Kyung-Tae Yang reports financial support was provided by LG
examining the three-dimensional surface morphology. Carbonization, Electronics.
one of the defects generated by the high energy intensity, was observed
when the laser power reached 320 W at a scanning speed of 800 mm/s or Acknowledgements
the speed reached 600 mm/s at the power of 250 W. The process pa­
rameters, at which this defect and also undesirable non-overlapping This work was supported by LG Electronics Inc., the Technology
melt pool were generated, were ruled out for the determination of the Innovation Program (20013794, Center for Composite Materials and
optimal process conditions. Accordingly, the laser power of 180–250 W Concurrent Design) funded by the Ministry of Trade, Industry & Energy
with a scanning speed of 800 mm/s and the scanning speed of 800–1100 (MOTIE, KOREA), and Korea Hydro & Nuclear Power Co., Ltd.
mm/s with a power of 250 W were down-selected to fabricate cubic
shaped blocks with high relative densities. The as-built parts were suc­ References
cessfully fabricated along with properly overlapped melt pool, impor­
tantly without carbonization. The average surface roughness was [1] N. Haghdadi, M. Laleh, M. Moyle, S. Primig, Additive manufacturing of steels: a
review of achievements and challenges, J. Mater. Sci. 56 (2021) 64–107, https://
measured to stay within the range of 12.94–15.92 μm, which is quite doi.org/10.1007/s10853-020-05109-0.
favorable for the use in practical applications. [2] N.T. Aboulkhair, M. Simonelli, L. Parry, I. Ashcroft, C. Tuck, R. Hague, 3D printing
Some defects were observed on the surfaces is owing to the low of aluminium alloys: additive manufacturing of aluminium alloys using selective
laser melting, Prog, Mater. Sci. 106 (2019), 100578, https://doi.org/10.1016/j.
densities of the as-built parts. Therefore, the process parameters were
pmatsci.2019.100578.
adjusted and then the HIP treatment was applied in order to minimize [3] J.V. Gordon, S.P. Narra, R.W. Cunningham, H. Liu, H. Chen, R.M. Suter, J.L. Beuth,
defects/voids thereby securing the high densities of the parts. The A.D. Rollett, Defect structure process maps for laser powder bed fusion additive
densities of the as-built parts fabricated with the VED ranging from manufacturing, Addit. Manuf. 36 (2020), 101552, https://doi.org/10.1016/j.
addma.2020.101552.
70.31 to 97.66 J/mm3 were enhanced up to 7.70 g/cc. Comparing the [4] D. Herzog, K. Bartsch, B. Bossen, Productivity optimization of laser powder bed
densities before/after the HIP, the optimum process parameter was fusion by hot isostatic pressing, Addit. Manuf. 36 (2020), 101494, https://doi.org/
determined as the VED of 97.66 J/mm3; the laser power, scanning 10.1016/j.addma.2020.101494.
[5] W. Stopyra, K. Gruber, I. Smolina, T. Kurzynowski, B. Kuźnicka, Laser powder bed
speed, layer thickness, and line spacing were found to be 250 W, 800 fusion of AA7075 alloy: influence of process parameters on porosity and hot
mm/s, 40 μm, and 80 μm, respectively. cracking, Addit. Manuf. 35 (2020), https://doi.org/10.1016/j.
The average surface roughness of the as-built parts is not suitable for addma.2020.101270.
[6] F. Chu, K. Zhang, H. Shen, M. Liu, W. Huang, X. Zhang, E. Liang, Z. Zhou, L. Lei,
exterior products, and gloss value, the most important metal property is J. Hou, A. Huang, Influence of satellite and agglomeration of powder on the
almost 0 Gu. Hence, automatic mechanical polishing was conducted to processability of AlSi10Mg powder in laser powder bed fusion, J. Mater. Res.
improve the surface quality of specimens. The fine surface finishing Technol. 11 (2021) 2059–2073, https://doi.org/10.1016/j.jmrt.2021.02.015.
[7] S.A. Khairallah, A.T. Anderson, A. Rubenchik, W.E. King, Laser powder-bed fusion
could achieve 856% reduction in the surface roughness and produced an additive manufacturing: physics of complex melt flow and formation mechanisms
outstanding improvement in the gloss (up to 433 Gu) with irregular of pores, spatter, and denudation zones, Acta Mater. 108 (2016) 36–45, https://
surface and bumps removal. PVD coating of the surface also resulted to doi.org/10.1016/j.actamat.2016.02.014.
[8] J.D. Roehling, W.L. Smith, T.T. Roehling, B. Vrancken, G.M. Guss, J.T. McKeown,
188.9% improvement in the gloss compared to the high polishing.
M.R. Hill, M.J. Matthews, Reducing residual stress by selective large-area diode
Touch sensations of the as-built parts were analyzed by character­ surface heating during laser powder bed fusion additive manufacturing, Addit.
izing the surface roughness, contact angle, and friction behaviors. The Manuf. 28 (2019) 228–235, https://doi.org/10.1016/j.addma.2019.05.009.
work of adhesions, ∆γ, between the silicon and products was found to be [9] C.J. Huang, X.C. Yan, C.Y. Chen, Y.C. Xie, M. Liu, M. Kuang, H.L. Liao, Additive
manufacturing hybrid Ni/Ti-6Al-4V structural component via selective laser
dependent on the surface roughness and the surface conditions. The high melting and cold spraying, Vacuum. 151 (2018) 275–282, https://doi.org/
mechanical polishing and the PVD coating generated a grippy surface, 10.1016/j.vacuum.2018.02.040.

13
K.-T. Yang et al. Surface & Coatings Technology 422 (2021) 127492

[10] B. Zhao, H. Wang, N. Qiao, C. Wang, M. Hu, Corrosion resistance characteristics of beam irradiation, Addit. Manuf. 33 (2020), 101125, https://doi.org/10.1016/j.
a Ti-6Al-4V alloy scaffold that is fabricated by electron beam melting and selective addma.2020.101125.
laser melting for implantation in vivo, Mater. Sci. Eng. C. 70 (2017) 832–841, [26] H. Li, Z.W. Chen, M. Ramezani, Wear behaviours of PVD-TiN coating onTi-6Al-4V
https://doi.org/10.1016/j.msec.2016.07.045. alloy processed by laser powder bed fusion or conventionally processed, Int. J.
[11] J.J. de Damborenea, M.A. Arenas, M.A. Larosa, A.L. Jardini, C.A. de Carvalho Adv. Manuf. Technol. 113 (2021) 1389–1399, https://doi.org/10.1007/s00170-
Zavaglia, A. Conde, Corrosion of Ti6Al4V pins produced by direct metal laser 021-06714-2.
sintering, Appl. Surf. Sci. 393 (2017) 340–347, https://doi.org/10.1016/j. [27] M.S.I. Chowdhury, B. Bose, K. Yamamoto, L.S. Shuster, J. Paiva, G.S. Fox-
apsusc.2016.10.031. Rabinovich, S.C. Veldhuis, Wear performance investigation of PVD coated and
[12] S. Pal, G. Lojen, R. Hudak, V. Rajtukova, T. Brajlih, V. Kokol, I. Drstvenšek, As- uncoated carbide tools during high-speed machining of TiAl6V4 aerospace alloy,
fabricated surface morphologies of Ti-6Al-4V samples fabricated by different laser Wear. 446–447 (2020), 203168, https://doi.org/10.1016/j.wear.2019.203168.
processing parameters in selective laser melting, Addit. Manuf. 33 (2020), 101147, [28] M. Mohammadi, H. Asgari, Achieving low surface roughness AlSi10Mg_200C parts
https://doi.org/10.1016/j.addma.2020.101147. using direct metal laser sintering, Addit. Manuf. 20 (2018) 23–32, https://doi.org/
[13] M. Bezuidenhout, G. Ter Haar, T. Becker, S. Rudolph, O. Damm, N. Sacks, The 10.1016/j.addma.2017.12.012.
effect of HF-HNO3 chemical polishing on the surface roughness and fatigue life of [29] Y. Tian, D. Tomus, P. Rometsch, X. Wu, Influences of processing parameters on
laser powder bed fusion produced Ti6Al4V, Mater. Today Commun. 25 (2020), surface roughness of Hastelloy X produced by selective laser melting, Addit. Manuf.
101396, https://doi.org/10.1016/j.mtcomm.2020.101396. 13 (2017) 103–112, https://doi.org/10.1016/j.addma.2016.10.010.
[14] Y. Madhavi, N. Narasaiah, A. Jyothirmayi, R.K. L, Influence of surface-roughness [30] Y. Kaynak, O. Kitay, The effect of post-processing operations on surface
on the corrosion-fatigue behavior of MAO coated 6061-T6 Al alloy assessed in NaCl characteristics of 316L stainless steel produced by selective laser melting, Addit.
medium, Surf. Coat. Technol. 414 (2021), 127102, https://doi.org/10.1016/j. Manuf. 26 (2019) 84–93, https://doi.org/10.1016/j.addma.2018.12.021.
surfcoat.2021.127102. [31] Y. Sun, R.J. Hebert, M. Aindow, Effect of heat treatments on microstructural
[15] L. Tonelli, A. Fortunato, L. Ceschini, CoCr alloy processed by selective laser melting evolution of additively manufactured and wrought 17-4PH stainless steel, Mater.
(SLM): effect of laser energy density on microstructure, surface morphology, and Des. 156 (2018) 429–440, https://doi.org/10.1016/j.matdes.2018.07.015.
hardness, J. Manuf. Process. 52 (2020) 106–119, https://doi.org/10.1016/j. [32] J.H. Martin, B.D. Yahata, J.M. Hundley, J.A. Mayer, T.A. Schaedler, T.M. Pollock,
jmapro.2020.01.052. 3D printing of high-strength aluminium alloys, Nature. 549 (2017) 365–369,
[16] S. Shrestha, Y. Kevin Chou, A numerical study on the keyhole formation during https://doi.org/10.1038/nature23894.
laser powder bed fusion process, J. Manuf. Sci. Eng. 141 (2019) 1–9, https://doi. [33] H. Irrinki, M. Dexter, B. Barmore, R. Enneti, S. Pasebani, S. Badwe, J. Stitzel,
org/10.1115/1.4044100. R. Malhotra, S.V. Atre, Effects of powder attributes and laser powder bed fusion (L-
[17] Z. Dong, Y. Liu, W. Wen, J. Ge, J. Liang, Effect of hatch spacing on melt pool and PBF) process conditions on the densification and mechanical properties of 17-4 PH
as-built quality during selective laser melting of stainless steel: modeling and stainless steel, Jom. 68 (2016) 860–868, https://doi.org/10.1007/s11837-015-
experimental approaches, Materials (Basel). 12 (2018), https://doi.org/10.3390/ 1770-4.
ma12010050. [34] H. Chen, Q. Wei, Y. Zhang, F. Chen, Y. Shi, W. Yan, Powder-spreading mechanisms
[18] C. Qiu, C. Panwisawas, M. Ward, H.C. Basoalto, J.W. Brooks, M.M. Attallah, On the in powder-bed-based additive manufacturing: experiments and computational
role of melt flow into the surface structure and porosity development during modeling, Acta Mater. 179 (2019) 158–171, https://doi.org/10.1016/j.
selective laser melting, Acta Mater. 96 (2015) 72–79, https://doi.org/10.1016/j. actamat.2019.08.030.
actamat.2015.06.004. [35] T. Fischer, B. Kuhn, D. Rieck, A. Schulz, R. Trieglaff, M.B. Wilms, Fatigue cracking
[19] U.S. Kim, J.W. Park, High-quality surface finishing of industrial three-dimensional of additively manufactured materials-process and material perspectives, Appl. Sci.
metal additive manufacturing using electrochemical polishing, Int. J. Precis. Eng. 10 (2020), https://doi.org/10.3390/app10165556.
Manuf. - Green Technol. 6 (2019) 11–21, https://doi.org/10.1007/s40684-019- [36] T.G. Fleming, S.G.L. Nestor, T.R. Allen, M.A. Boukhaled, N.J. Smith, J.M. Fraser,
00019-2. Tracking and controlling the morphology evolution of 3D powder-bed fusion in situ
[20] A. Bernhardt, J. Schneider, A. Schroeder, K. Papadopoulous, E. Lopez, F. Brückner, using inline coherent imaging, Addit. Manuf. 32 (2020), 100978, https://doi.org/
U. Botzenhart, Surface conditioning of additively manufactured titanium implants 10.1016/j.addma.2019.100978.
and its influence on materials properties and in vitro biocompatibility, Mater. Sci. [37] M.J. Matthews, G. Guss, S.A. Khairallah, A.M. Rubenchik, P.J. Depond, W.E. King,
Eng. C. 119 (2021), 111631, https://doi.org/10.1016/j.msec.2020.111631. Denudation of metal powder layers in laser powder bed fusion processes, Acta
[21] B. Seo, H.K. Park, H.G. Kim, W.R. Kim, K. Park, Corrosion behavior of additive Mater. 114 (2016) 33–42, https://doi.org/10.1016/j.actamat.2016.05.017.
manufactured CoCr parts polished with plasma electrolytic polishing, Surf. Coat. [38] L. Scime, J. Beuth, Melt pool geometry and morphology variability for the Inconel
Technol. 406 (2021), 126640, https://doi.org/10.1016/j.surfcoat.2020.126640. 718 alloy in a laser powder bed fusion additive manufacturing process, Addit.
[22] S. Chang, A. Liu, C.Y.A. Ong, L. Zhang, X. Huang, Y.H. Tan, L. Zhao, L. Li, J. Ding, Manuf. 29 (2019), 100830, https://doi.org/10.1016/j.addma.2019.100830.
Highly effective smoothening of 3D-printed metal structures via overpotential [39] J.P. Oliveira, A.D. LaLonde, J. Ma, Processing parameters in laser powder bed
electrochemical polishing, Mater. Res. Lett. 7 (2019) 282–289, https://doi.org/ fusion metal additive manufacturing, Mater. Des. 193 (2020) 1–12, https://doi.
10.1080/21663831.2019.1601645. org/10.1016/j.matdes.2020.108762.
[23] P. Tyagi, T. Goulet, C. Riso, F. Garcia-Moreno, Reducing surface roughness by [40] X. Liu, Z. Yue, Z. Cai, D.G. Chetwynd, S.T. Smith, Quantifying touch-feel
chemical polishing of additively manufactured 3D printed 316 stainless steel perception: Tribological aspects, Meas. Sci. Technol. 19 (2008), https://doi.org/
components, Int. J. Adv. Manuf. Technol. 100 (2019) 2895–2900, https://doi.org/ 10.1088/0957-0233/19/8/084007.
10.1007/s00170-018-2890-0. [41] R. Ploeger, S. Musso, O. Chiantore, Contact angle measurements to determine the
[24] S.D. Heintze, M. Forjanic, V. Rousson, Surface roughness and gloss of dental rate of surface oxidation of artists’ alkyd paints during accelerated photo-ageing,
materials as a function of force and polishing time in vitro, Dent. Mater. 22 (2006) Prog. Org. Coat. 65 (2009) 77–83, https://doi.org/10.1016/j.
146–165, https://doi.org/10.1016/j.dental.2005.04.013. porgcoat.2008.09.018.
[25] T.S.N. Sankara Narayanan, J. Kim, H.E. Jeong, H.W. Park, Enhancement of the [42] A. Le Gal, X. Yang, M. Klüppel, Evaluation of sliding friction and contact mechanics
surface properties of selective laser melted maraging steel by large pulsed electron- of elastomers based on dynamic-mechanical analysis, J. Chem. Phys. 123 (2005),
https://doi.org/10.1063/1.1943410.

14