ScienceDirect
Available online at www.sciencedirect.com
Procedia CIRP 124 (2024) 157–162
2212-8271 © 2024 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 the 13th CIRP Conference on Photonic Technologies [LANE 2024]
10.1016/j.procir.2024.08.090
Keywords: PBF-LB; AlSi10Mg; DoE; density; surface roughness; volumetric energy density; parameter Optimization; upskin; downskin; pore size
1. Introduction
Additive Manufacturing (AM) is a rapidly developing field
with an expected compound annual growth rate of 13.9 % until
2028, where the global market revenue would increase from
EUR 10.5 billion in 2023 to a projected EUR 20 billion in 2028,
with PBF-LB being the predominant technology overall [1]. It
is an additive powder bed-based manufacturing process in
which metal powder is melted layer by layer selectively with a
laser beam to produce a component [2, 3]. This technology is
widely used in the aerospace,defense,automotive and medical
industries to produce highly specialized parts, often
topologically optimized to reduce weight, for specific
applications[4, 5].
Nomenclature
ET exposure time, in µs
HD hatch distance, in µm
LT layer thickness, in µm
Plaser power,in W
13th CIRP Conference on Photonic Technologies [LANE 2024], 15-19 September 2024, Fürth, Germany
Parameter analysis to correlate density with surface roughness and
productivity in Powder Bed Fusion –Laser Beam (PBF-LB) of AlSi10Mg
-Invited Paper-
Hong Tiat Tana,*, Elena Lopeza, Alex Selbmanna, David Karlb, Lukas Stepiena, Frank Bruecknera
aFraunhofer Institute for Material and Beam Technology IWS, Winterbergstr. 28, 01277 Dresden, Germany
bChair of Advanced Ceramic Materials, Institute of Material Science and Technology, Technische Universität Berlin, Straße des 17. Juni 135, 10623 Berlin,
Germany
* Corresponding author. Tel.: +49-351 83391 3876.E-mail address:hong.tiat.tan@iws.fraunhofer.de
Abstract
This work investigates the development of manufacturing parameters and correlations of parts manufactured via PBF-LB with AlSi10M g using
Design of Experiments (DoE). The main goal of this research is to gain a comprehensive understanding of the parameter space and find
correlations of relative density with surface roughness and productivity by considering laser powers up to 400 W. The influence of different
process parameters, combined and represented as the Volumetric Energy Density (VED) is investigated. Here, a density of above 99.5 % could
be achieved within a wide VED range of 40 –188 J/mm³. The measured mean downskin surface roughness ranges from 12 –68 µm depending
on their overhang angle and measurement method, whereas the mean top upskin surface roughness ranges between 4 –48 µm with different
parameter combinations of laser power (P) and layer thickness (LT). The approach developed here enablesthe efficient development of suitable
process parameters and reduces the optimization effort of the PBF-LB process through the rapid characterization of manufactured part quality in
terms of density, which directly influences its mechanical strength.
© 2024 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 the 13th CIRP Conference on Photonic Technologies [LANE 2024]
158 Hong Tiat Tan et al. / Procedia CIRP 124 (2024) 157–162
PD point distance, in µm
PR productivity, in mm3/s
Ra,Saarithmetic mean roughness, in µm
Rdq, Sdq root mean square slope of a roughness profile
Rz,Szsum of peak height and largest valley depth, in µm
VED volumetric energy density, in J/mm3
Among the different materials used in PBF-LB, AlSi10M g
is one of the most intensively researched Al-alloys in AM [6, 7].
This is due to its favorable properties such as high specific
strength-to-mass ratio, ductility, and corrosion resistance along
with good thermal and electrical conductivity [8]. However, the
optimization of PBF-LB parameters for Al-alloys is
challenging, especially for low-power lasersas Al-alloys
dissipate heat rapidly due to their high thermal conductivity,
which can result in shallow melt pools[9]. Al-alloys tend to
have high reflectivity, making it difficult to achieve high energy
absorption rates[9 –12]. What is more, Al-alloys form o xid e
layers instantaneously starting at room temperature, which can
be advantageous for the corrosion resistance of the final product
but requires high-powered lasersin PBF-LB processing to
disrupt the oxidation layer during manufacturing in order for
the powder particles to melt and fuse [13]. On top of that, the
vast number of parameters in PBF-LB makes it a challenging
task to optimize the manufacturing process [5, 14].
The focus of process parameter optimization is usually on
achieving the highest possible relative component density and
minimal pore size in order to achieve the best mechanical
strength [13]. However, reduced surface roughness and
production time are further competing objectives in the
development of suitable process parameters [15]. For example,
the production time could be reduced by increasing the LT but
the surface roughness thereby increases significantly [16].
Besides, the ideal operating window of PBF-LB narrows as the
layer thickness is increased [17].
Various authors found that the ideal VED for PBF-LB of
AlSi10Mg is around 60 J/mm3[13, 18]. In fact, the density of a
part increases significantly with increasing energy density but
decreases after reaching a maximum, most likely due to the
formation of keyholes [19]. Giovagnoli et al. [20] s ummarize d
that a wide range of VED 30 –110 J/mm3resulted mostly in a
density of above 97 %. Despite using ahigh LT of 90 μm,
Mercurio et al. [14] were able to achieve an as-built density of
more than 99 % at VED between 22.59 and 35.13 J/mm3, with
a majority of the pores caused by lack of fusion and spattering.
Wang et al. [19] investigated the effect of VED on the
roughness of the surface perpendicular to the building direction,
in other words, the top surface. They found that the surface
roughness decreases with increasing VED as partially melted
powder is eliminated with an increase of liquid melt due to
higher energy absorbed by the powder particles, whereby the
lowest mean surface roughness of Ra4.127 μm is obtained at
131 J/mm3when using a laser power of 400 W.
Calignano et al. [21] established that the laser scanning speed
is highly significant on the top surface roughness of a part,
followed by laser power and hatching distance.
Praneeth et al. [22] investigated front-facing roughness or
surface perpendicular to the build platform and concluded that
laser power and scan speed have the most effect on the surface
roughness.
This study aims to derive the relationship between the
manufacturing parameters and the target variables, namely
relative density, surface roughness and productivity,in order to
find suitable parameter combinations for different purposes.
For instance, the desire for high relative density, which
sacrifices productivity,dominates in lightweight or aerospace
applications [23]. Surface roughness is of primary importance
for heat exchangers and other flow-through geometries [24]. It
is,therefore,of high scientific and economic relevance to gain
a comprehensive overview of the parameter space and their
relationship between different objectives to allow for a fast yet
effective parameter optimization.
2. Material and Methods
AlSi10Mg powder feedstock material produced via gas
atomization was provided by Heraeus Additive Manufacturing
Gmb H,with the powder composition conforming to the
EN AC 43000 standard. The morphology and particle size
distribution (PSD)of the powder were characterized using
scanning electron microscopy (SEM) and via high-resolution
digital camera using a CAMSIZER X2 from Microtrac Retsch
GmbH. As seen in (a) of Fig. 1, the majority of the powder
particles have a spherical shape, typical for gas-atomized
metallic feedstocks. In (b), the d10, d50 and d90 of the powder
are 21.5 µm, 32.7 µm and 46.5 µm respectively.
(a)
(b)
Fig. 1. (a) Powder morphology observed under SEM; (b) PSD of powder used
The AM 400 PBF-LB machine equipped with a 400 W
ytterbium fiber pulsed wave laser and a 75 µm laser spot
diameter from Renishaw plc. is used to produce specimens for
this work. 15 x10 x10 mm3cuboids are manufactured for the
density experiments. The metallography image analysis
method is used to characterize the density.The specimens are
sliced in two planes, namely parallel (xz) and perpendicular
(xy) to the build direction. The former is analyzed to spot
keyhole porosity, whereas the latter is done to check for
porosity due to uneven powder deposition. They are embedded
in epoxy, polished and viewed under the Leica MEF 4 light
microscope to determine their densities, whereby an average of
both planes is taken. The experiments are planned based on the
DoE approach and evaluated using the statistical software
Minitab 19 from Minitab Gmb H, where the most significant
parameters are determined and evaluated further. The
parameters are eventually converted into the VED using
Hong Tiat Tan et al. / Procedia CIRP 124 (2024) 157–162 159
equation (1), where P, ET, PD, HD and LT are laser power,
exposure time, point distance, hatch distance and layer
thickness. The VED of 26 –188 J/mm3is investigated in this
work.
𝑉𝑉𝑉𝑉𝑉𝑉 =
𝑃𝑃 ∙ 𝑉𝑉𝐸𝐸
𝑃𝑃𝑉𝑉 ∙𝐻𝐻𝑉𝑉 ∙𝐿𝐿𝐸𝐸
(1)
The productivity, PR is calculated using equation (2).
𝑃𝑃𝑃𝑃 =
𝑃𝑃𝑉𝑉
𝑉𝑉𝐸𝐸
∙𝐻𝐻𝑉𝑉 ∙𝐿𝐿𝐸𝐸
(2)
In this work, the Surfcom Touch 50 from Accretech
(Europe) GmbH along with a VK-X100 laser scanning
microscope (LSM) from Keyence Corporation,is used for the
tactile and surface roughness measurements,respectively. The
initial R-value roughness measured in this study are Ra,Rzand
Rdq, whilst the S-value roughness are Sa,Szand Sdq, which are
defined and described in DIN EN ISO 4287 and
DIN EN ISO 25178 respectively [25, 26]. Separate specimens
with different overhang angles of 30 –75° are manufactured
with the same parameter.This prevents the influence of other
neighboring samples in case of individual build failures and the
downskin surface is measured using both apparatuses. This is
done to evaluate the suitability and repeatability of different
measurement methods for PBF-LB parts as well as the
minimu m manufacture angle. The type of surface roughness
with comparable values across both measurement techniques is
used for further evaluation in this work. The upskin surface
roughness is measured using the top surface of the cuboids
manufactured during the density characterization. The surface
roughness specimens are measured three times in the same
direction at different regions,and an average is taken as the
roughness value. This work is a subpart of a large project,
where a substantial amount of specimens are manufactured for
parameter optimization of the PBF-LB process.
3. Results and discussion
3.1. Density evaluation
Through the DoE evaluation excluding LT, the laser P is
found to be the most significant parameter affecting the
density. Fig. 2illustrates the relationship between the density,
laser P, average pore size and VED. The size of the circle
represents the average pore size relative to one another, with
the largest being 76.76 µm,whereas the smallest is 8.83 µm.
Generally, the average pore size decreases with increasing
density. Lack of fusion is the main contributor to porosities
below a VED of 100 J/mm3, while the pores are mostly round
above that. Also, the density range narrows towards being fully
dense as laser P increases and lack of fusion decreases. Once
the parameters are summarized as VED, a trend of higher
density at higher VED could be seen, albeit being scattered
along the line of best fit.
(a)
(b)
Fig. 2. (a) Graph of density against P with average pore size; (b) Graph of
density against VED with average pore size
A VED of above 100 J/mm3is required to achieve densities
above 98 % consistently,but it is noted that higher densities
could be achieved at lower VEDs when the full 400 W laser
power is used. Besides, the smallest average pore size is
achieved at VED 110 –135 J/mm3. When taking LT of
25 –75 µm into account, as shown in Fig. 3, where the
relationship between density, LT, VED as well as average pore
size is displayed, the average pore size decreases similarly with
increasing density. Additionally, the density decreases while
the average pore diameter increases with increasing LT. The
density increases along with VED. However, there are a
number of outliers between 50 –100 J/mm3where the density
fluctuates between a big range of around 87 –99 % for both
LT 50 μm and LT 75 μm.
(a)
(b)
Fig. 3. (a) Graph of density against LT with average pore size; (b) Graph of
density against VED at different LT with average pore size
At a fixed LT,a similar VED could lead to different
densities,but when comparing different LTs,a lower VED
leads to lower density. When parts of similar achieved densities
at different LTs but with almost identical VED of ca.
115 J/mm3are compared, the average pore size is larger at the
higher LT. Fig. 4shows an example of a part in (a) with a
density above 99.9 %, whereby little to no key hole porosity is
160 Hong Tiat Tan et al. / Procedia CIRP 124 (2024) 157–162
present, along with a part in (b) with a density of only ca. 95 %,
where there is a massive lack of fusion. In short, VED that is
widely used in literature could only give an estimation of the
achievable density range and might be an over simplification
of the manufacturing parameters as combinations of different
significance could appear to have similar VED but contrasting
densities.
(a)
(b)
Fig. 4. (a) Metallographic micrograph of a part with a density of 99.92 %;
(b) Metallographic micrograph of a part with a density of 95.09%
3.2. Surface roughness evaluation
The initial surface roughness specimens manufactured are
shown in Fig. 5. The minimum manufacture angle without
defect is 35° to the build platform. A lower angle is not possible
as it would lead to build failure and damage to the silicone
recoating wiper. Fig. 6shows the graph of downskin Ra, Sa, Rz,
Sz, Rdq and Sdq against the build angle. Generally, it shows the
trend of a steady increase in surface roughness with a
decreasing angle to the build platform and a steep increase
beyond 40°. Although the 30° specimen could be
manufactured, the downskin surface is too rough and out of the
measurable range of the tactile measuring device.
Comparing (a), (b),and (c) graphs against each other in
Fig. 6, both tactile and optical Rashows the most consistent
reading between different measuring techniques. The Ravalues
are comparable to the Savalues that take the whole surface into
account. Although both Rzvalues are comparable, the Szvalue
is much higher when an entire surface is taken into account.
However, both tactile and optical Rdq, along with the optical
Sdq, are not comparable with each other, although the same part
and region are measured. Although the Sameasured using an
LSM provides a comprehensive three-dimensional topology of
the surface, it takes a substantially longer time to measure
compared to the tactile method. For example, 3 min is required
for each tactile measurement compared to 15 –30 minfor a
single surface measurement on the LSM,depending on the
resolution and measured surface area. Hence, the tactile Rais
taken to quantify the surface roughness in this work.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
Fig. 5. Downskin surface wit h t he angle to build a platform of (a)30°;
(b) 35°; (c) 40°; (d) 45°(e) 53°; (f)60°; (g) 75°
(a)
(b)
(c)
Fig. 6. (a) Graph of downskin Ra, Saagainst angle to build platform;
(b) Graph of downskin Rz, Szagainst angle to build plat form; (c) Graph of
downskin Rdq, Sdq against angle to build platform
3.3. Correlation of density against productivity and roughness
Fig. 7shows the plot of density against the productivity of
the cubes manufactured with a fixed LT of 25 μm. Generally,
the density decreases with increasing productivity,apart from
a few outliers near 8 mm3/s. At the same productivity of about
4mm3/s, it can be seen that a higher power favors higher
density at the same productivity since productivity is not
directly influenced by laser power,as seen in equation (2).
Although laser power is not a function of productivity, a
high laser power should be chosen as it favors a high density
without sacrificing productivity. Based on the trend lines
plotted in Fig. 7, the slope or degree of decrease in density as
productivity increases is higher when a lower laser power is
used. Since VED is proportional to the laser power, it decreases
as a lower power is used,prompting a lack of fusion to occur
at lower productivity compared to when a higher laser power is
used. Here, it is expected that the density decreases with
increasing manufacturing speed or productivity up to a point
where the parts could not be manufactured due to a massive
lack of fusion.
However, the density values scatter massively around the
curve of best fit in Fig. 7, even at a constant calculated
productivity. Also, there is an anomaly towards the end of the
curve where densities higher than expected are recorded. This
Hong Tiat Tan et al. / Procedia CIRP 124 (2024) 157–162 161
could be due to the pure mathematical simplification of
equation (2) that does not take the melt-pool properties into
account. There might be a tight process window that allows a
high density to be achieved despite the high productivity that
should generally result in a low density. This phenomenon is
not investigated further in this work and could be part of a
future work.
Fig. 7. Graph of density against productivity at a fixed LT of 25 µm
Next, the top upskin surface roughness of a part could be
related to its density,as shown in Fig. 8. Generally, the density
decreases as the top surface roughness increases. For parts with
a high density, the top surface appears shiny and smooth. The
tactile measuring device could pick up pores formed on the top
surface. Also, the top surface roughness tends to be higher
when a higher LT of 50 μm or 75 μm is used. This could be due
to more powder being melted which might affect the stability
of the melt pool as more molten swirl is formed. Also, the drop
in density along with the increase in top surface roughness
escalates when a lower power is applied. The density drops due
to lack of fusion and the top roughness increases as the result
of low density.
Fig. 8. Graph of density against top upskin Raincluding different laser power
and layer thickness
The findings and correlations in Fig. 7and Fig. 8further
confirm the state of the art that density, surface roughness and
productivity are competing objectives in the PBF-LB process,
as shown in Fig. 9, that affect each other negatively when one
or the other are favored over the other. Hence, compromises
and boundaries have to be defined before an optimization is
conducted to find the suitable parameter for a specific use case.
Fig. 9. Competing objectives in the PBF-LB process
4. Conclusion
This work presents novel findings on the correlation
between density and surface roughness for PBF-LB parameters
of Al-alloys , which should be further investigated using a wider
range of data. It has been proven that density, surface roughness
and productivity are competing objectives. With that,
conventional parameter optimization workflow could be
shortened by implementing a faster yet non-destructive build
quality characterization method and this increases the
productivity on the process chain level.
This work also suggests that VED could only give an
estimation of the achievable density range. However, it might
be an over simplification of the manufacturing parameters as
combinations of different significance could appear to have
similar VED but contrasting densities and vice versa. Adensity
of above 99.5 % with an average pore size as small as 8.83 µm
is achieved within a wide VED range of 40 –188 J/mm3,
although state of the art (Refer Chapter 1) recommends a
specific VED of 60 J/mm3or a range of 30 –110 J/mm3to gain
high relative densities.The arithmetic mean roughness, Raand
Sashowed the most consistent and comparable values between
different measuring techniques, where the optical LSM
roughness is more representative of an entire surface,but the
tactile roughness measurements are faster. The values converge
to the measured optical values when an average of more
measurements are taken.
Future research will focus on concurrently optimizing these
competing objectives based on pre-defined goals. Also, other
non-destructive characterization methods,such as surface eddy
current measurements,should be conducted on suitable
materials to verify any potential correlations.
Acknowledgements
The authors would like to thank the Bundesministerium für
Wirtschaft und Klimaschutz (BMWK) for funding the FAST
project through the Luftfahrtforschungsprogramm (LuFo) with
the Grant No. 20L2105B1.
References
[1] Munsch M, Schmidt-Lehr M, Wycisk E, Führer T , AMPOWER Report
2024 -Management Summary, 2024.
[2] Verein Deutscher Ingenieure e.V., Additive Fertigungsverfahren
Grundlagen, Begriffe, Verfahrensbeschreibungen, Beuth Verlag GmbH,
Berlin 25.030, 2014.
162 Hong Tiat Tan et al. / Procedia CIRP 124 (2024) 157–162
[3] Deutsches Institut für Normung e. V., DIN EN ISO/ASTM 52900:2022-03,
Addit ive Fertigung –Grundlagen –T erminologie,Beut h Verlag GmbH,
Berlin 01.040.25; 25.030, 2022.
[4] Czink S, Holoch J, Renz R, Schulze V, Albers A, Dietrich S, Process-
Specific Topology Optimization Method Based on Laser-Based Additive
Manufact uring of AlSi10Mg Components: Material Characterization and
Evaluation, Processes 11, 2023; 648.
[5] Calignano F, Galati M, Iuliano L, A Metal Powder Bed Fusion Process in
Indust ry: Qualification Considerations, Machines 7, 2019; 72.
[6] Limbasiya N, Jain A, Soni H, Wankhede V, Krolczyk G, Sahlot P, A
comprehensive review on the effect of process parameters and post-process
treatments on microstructure and mechanical properties of selective laser
melt ing of AlSi10Mg, Journal of Materials Research and Technology 21;
2022. p. 1141–1176.
[7] Wong K K, Ho J Y, Leong K C, Wong T N, Fabrication of heat sinks by
Selective Laser Melting for convective heat transfer applications, Virtual
and Physical Prototyping 11; 2016. p. 159–165.
[8] Fella E, Selective Laser Beam Melting of AlSi10Mg Alloy. Master's Thesis,
Milan, Italy, 2016.
[9] Yu W, Sing S L, Tian X, Chua C K, Effects Of Re-Melting Strategies On
Densification Behaviour And Mechanical Properties Of Selective Laser
Melted AlSi10Mg Parts, Nanyang Technological University, 2018.
[10] Aboulkhair N T, Maskery I, Tuck C, Ashcroft I, Everitt N M, The
microstructure and mechanical properties of selectively laser melted
AlSi10Mg: The effect of a conventional T6-like heat treatment, Materials
Science and Engineering: A 667; 2016. p. 139–146.
[11] Aboulkhair N T, Maskery I, Tuck C, Ashcroft I, Everitt N M, On the
formation of AlSi10Mg single tracks and layers in selective laser melting:
Microstructure and nano-mechanical properties, Journal of Materials
Processing Technology 230; 2016. p. 88–98.
[12] Kiani P, Dupuy A D, Ma K, Schoenung J M, Directed energy deposition
of AlSi10Mg: Single track nonscalability and bulk properties, Materials &
Design 194, 2020; 108847.
[13] Ferro P, Meneghello R, Razavi S M J, Berto F, Savio G, Porosit y Inducing
Process Parameters in Selective Laser Melted AlSi10Mg Aluminium
Alloy, Phys Mesomech 23; 2020. p. 256–262.
[14] Mercurio V, Calignano F, Iuliano L, Sust ainable production of AlSi10Mg
parts by laser powder bed fusion process, Int J Adv Manuf Technol 125;
2023. p. 3117–3133.
[15] Strano G, Multi-objective Optimization in Addit ive manufacturing. PhD
T hesis, Exeter, England, 2012.
[16] Paradise P, Patil D, van Handel N, Temes S, Saxena A, Bruce D, Suder A,
Clonts S, Shinde M, Noe C, Godfrey D, Hota R, Bhate D, Improving
Productivity in the Laser Powder Bed Fusion of Inconel 718 by Increasing
Layer Thickness: Effects on Mechanical Behavior, J. of Mat eri Eng and
Perform 31; 2022. p. 6205–6220.
[17] Saunders M, X marks the spot -find t he ideal process parameters for your
metal AM parts, 2019.
[18] Read N, Wang W, Essa K, Attallah M M, Selective laser melting of
AlSi10Mg alloy: Process optimisation and mechanical properties
development, Materials & Design (1980-2015) 65; 2015. p. 417–424.
[19] Wang L, Wang S, Wu J, Experimental investigation on densification
behavior and surface roughness of AlSi10Mg powders produced by
selective laser melting, Optics & Laser Technology 96; 2017. p. 88–96.
[20] Giovagnoli M, Silvi G, Merlin M, Di Giovanni M T , Optimisation of
process parameters for an additively manufactured AlSi10Mg alloy:
Limitations of the energy density-based approach on porosity and
mechanical properties estimation, Materials Science and Engineering: A
802, 2021; 140613.
[21] Calignano F, Manfredi D, Ambrosio E P, Iuliano L, Fino P, Influence of
process parameters on surface roughness of aluminum parts produced by
DMLS, Int J Adv Manuf Technol 67; 2013. p. 2743–2751.
[22] Praneeth J, Venkatesh S, Sivarama Krishna L, Process parameters
influence on mechanical properties of AlSi10Mg by SLM, Materials
Today: Proceedings 2023.
[23] Lu C, Shi J, Simultaneous consideration of relative density, energy
consumption, and build time for selective laser melting of Inconel 718: A
mult i-objective optimization study on process parameter selection, Journal
of Cleaner Product ion 369, 2022; 133284.
[24] Burkhardt T, Abrasives Strömungsschleifen von selektiv
laserstrahlgeschmolzenen Bauteilen für Luft-und Raumfahrtapplikationen.
Diplomarbeit, Dresden, Germany, 2022.
[25] Deutsches Institut für Normung e. V., DIN EN ISO 4287:2010-07,
Geometrische Produktspezifikation (GPS) –Oberflächenbeschaffenheit:
Tastschnittverfahren –Benennungen, Definitionen und Kenngrößen der
Oberflächenbeschaffenheit, Beut h Verlag GmbH, Berlin 01.040.17;
17.040.30, 2020.
[26] Deutsches Institut für Normung e. V., DIN EN ISO 25178-2:2012-09,
Geometrische Produktspezifikation (GPS) –Oberflächenbeschaffenheit:
Flächenhaft –Teil 2: Begriffe und Oberflächen-Kenngrößen, Beuth Verlag
GmbH, Berlin 17.040.30, 2012.
