Process Parameters Development of Selective Laser Melting of Lunar
Regolith for On-Site Manufacturing Applications
Miranda Fateri* and Andreas Gebhardt
FH Aachen University, Goethestr. 1, Aachen 52064, Germany
Selective Laser Melting (SLM) is a candidate for on-site manufacturing as its characteristics of energy source and powder-
based fabrication process are suitable for use with in situ material. The feasibility of using lunar regolith simulant to create
objects with SLM process is investigated in this study. The process parameters are optimized and multiple objects are fabricated.
A qualitative chemical analysis is carried out with scanning electron microscopy using energy-dispersive X-ray emission. Lastly,
properties such as particle size distribution, particle shape, and crystal structure of the lunar simulant powder as well as the
crystallinity and hardness of the fabricated objects are investigated.
The moon, the only other celestial body besides the
Earth man has ever stepped on, is the closest and the
easiest for human space travel; namely, the moon is on
average 348,000 km away from the Earth, while Venus
and Mars, the two closest planets to Earth, are on aver-
age 42 and 78 million kilometers away, respectively. The
moon is also the most investigated and well known due
to the extensive number of previous missions by NASA
and other space agencies. Recently, space agencies world-
wide are considering manned space missions to celestial
bodies albeit with the objective to establish a base instead
of a short visit. A lunar base is not only extremely valu-
able for long-term scientiﬁc projects, but it will also serve
as a test bed for bases in nearby planets as well as poten-
tially serving as a refueling stopover for missions to Mars
and other planets. To achieve that the objective of estab-
lishing a long-term base on a celestial body, large-scale
on-site manufacturing is required; initially, it may be
possible to take parts of the base modularly from Earth
and assemble on-site, but eventually as the base grows,
long-term sustainability will indeed depend on on-site
manufacturing using in situ resources. According to
NASA, “In-situ resource utilization will enable the
affordable establishment of extraterrestrial exploration
and operations by minimizing the materials carried from
Such establishment will mean that manufactur-
ing in a very broad scale will be required, from precision
manufacturing of gears, bearings, fasteners, and electrical
components to large-scale construction.
Numerous papers have been published on the appli-
cability and feasibility of conventional and additive man-
ufacturing (AM) processes using lunar regolith as raw
material for both large-scale and small part fabrication.
Regarding large-scale construction, numerous ideas both
underground and aboveground have been investigated;
one pervasive fact in all studies is the extensive use of
lunar regolith for bulk construction as well as thermal
and radiation shielding.
Additionally, the scarcity of
manpower will mean that processes, especially mundane
tasks, must be thoroughly automated. With regard to
this, Khoshnevis et al.
studied the applicability of Con-
tour Crafting (CC) system which was initially developed
to fabricate 3D objects out of various materials such as
polymer, ceramic slurry, and cement. Following the CC
research, Khoshnevis et al.
investigated the feasibility of
sulfur concrete extrusion with different sulfur concentra-
tions. Sulfur concrete was chosen as a candidate as the
material has a low melting temperature (120°C), it is
recyclable and it cures fast (after 2 min of extrusion);
however, Grugel et al.
studied the behavior of sulfur
concrete cubes subjected to lunar temperature cycles and
reported unfavorable results. The evaluation results of
cycled and noncycled samples have shown that the ten-
sile strength of the cubes exposed to the lunar thermal
cyclic variation has dropped to one-ﬁfth of that of the
noncycled ones. After 80 cycles, material disintegration,
due to the differences in thermal expansion coefﬁcients
of the constituting materials of the concrete, was macro-
scopically observed. Based on this, sulfur concrete extru-
sion by CC could be suitable for certain lunar structures
such as underground or near the poles that will not
experience cyclic temperature variations.
Another technology that has been investigated for
lunar base construction is D-shape process which was
developed by Monolite UK. Regarding this, Alta SpA
et al. studied the feasibility of D-shape technology that
©2014 The American Ceramic Society
Int. J. Appl. Ceram. Technol., 1–7 (2014)
has the capability to agglomerate inert materials using
a special binding ink for in situ application. In the
D-shape process, the lunar regolith must ﬁrst be mixed
with additional magnesium oxide and the process also
requires a binding salt. For lunar applications, a new
injection method was developed in which the ink is
injected directly inside the regolith powder bed being
surrounded by powder all around which shields the ink
from evaporation and freezing due to the vacuum.
part of the conceptual feasibility study, a 1.5-ton, hollow
closed-cell structure, building block was fabricated.
The production of small objects, secondary to the
construction of the building structure itself, has not been
as thoroughly investigated. The aim must be to once
again produce as much as possible with in situ materials
using highly automated processes with the least amount
of machinery. In contrast, here on Earth, manufacturing
is normally carried out using unique material with large
and heavy equipment in specialized factories. On the
other hand, on-site manufacturing using AM processes
enables fabrication of parts with minimal infrastructure.
With respect to this, Balla et al.
studied the feasibility
of direct fabrication of objects using lunar regolith simu-
lant with Laser Engineering Net shaping. Cylindrical
shape objects were fabricated, and the parts were ana-
lyzed using microscopy and spectroscopy to evaluate
their microstructure. The investigation has demonstrated
the applicability of LENS process for in situ lunar con-
struction although further research is required to deter-
mine the suitability of the mechanical property of the
parts for engineering applications.
In a previous investigation, the present authors
investigated the direct sintering/melting of lunar regolith
using SLM technology,
which was based on previous
studies on SLM of common beach sand.
the capability of SLM of lunar regolith was
established and 3D object were fabricated and presented.
This paper presents further research on this topic focused
on process parameter optimization, morphology, and
mechanical tests of the fabricated objects.
For this investigation, a SLM machine embedded
with an Yb:YAG ﬁber laser with a spot size of approxi-
mately 15 lm and a maximum power of 100 Watts is
used. The device has a printable build space with a
diameter of 60 mm and a maximum z-direction height
of 27 mm.
For raw material, JSC-1A lunar regolith
simulant procured from orbital technologies has been
used. This simulant is designed to have a similar compo-
sition as the lunar mare areas and has a grain size smaller
than 1 mm with a melting temperature of 1100–
1125°C. In Fig. 1a, the particle size distribution of the
JSC-1A as well as the lunar regolith is shown; the rego-
lith has a very wide range of particle size distribution
with the largest particles in the millimeter range, while
the smallest particles are a few micrometers in diameter.
Figure 1b shows the obtained simulant particle size dis-
tribution as tested using static laser light-scattering
method, which was further conﬁrmed using SEM.
The results of the tests show that 90 percentile point
of our sample, Fig. 1b, has a diameter less than 200 lm
which is slightly smaller than the 300 lm for the 90
mass percentile point shown in Fig. 1a.
The powder particle size and its distribution play
an important role in SLS/SLM processes similar to
conventional sintering processes. In SLS/SLM, the laser
is irradiated on a spot for a predetermined amount of
time, and then, it is moved to a new spot; thus, a
Fig. 1. (a) Particle size distribution of the JSC-1A and lunar
(b) Particle size distribution of JSC 1A obtained from
2International Journal of Applied Ceramic Technology—Fateri and Gebhardt 2014
variation in particle size directly alters the irradiated
energy density per unit mass.
In previous research,
using the raw simulant, the
fabricated parts exhibited a heterogeneous structure and
rough surface when high dimensional accuracy was a pri-
ority. Based on this, to reduce the variation in particle
size and equalize the energy absorption within the pow-
der bed, the simulant was ﬁltered through a 63 lm grid
mesh. As shown in Fig. 1, the ﬁltered powder represents
approximately 55% of the obtained JSC-1A simulant;
this value corresponds to approximately 27–62% of the
The surface proﬁle of the powder particles is studied
using SEM process. Figure 2 shows the SEM image of
the lunar regolith simulant in different magniﬁcations.
SEM is done using the backscattering electron mode that
superimposes the topographical information of the more
usual secondary electron mode with a contrast controlled
by differences in chemical composition, thus making vis-
ible inhomogeneities or foreign precipitations. However,
the composition of the powder is homogeneous.
As it can be seen in Fig. 2a, particles are granular
and irregular in shape. The powder particles feature frag-
ments with sharp edges as well as holes due to fracturing.
Additionally, small particles in nanometer range which
act as dust in the process can be seen in Fig. 2b.
Granular particles cannot be spread across the pow-
der bed as smoothly as spherical powder as their shapes
are random; additionally, the volume gaps that are
inconsistent between the different particles lead to a het-
erogeneous mass distribution and thus energy intensity,
which may lead to an inconsistent melt pool or voids
within the part due to lack of material at a spot.
The material composition of the lunar regolith as
reported by NASA and the lunar simulant including tol-
erances are shown in Fig. 3.
Additionally, EDX spectrum of the obtained simu-
lant powder is shown in Fig. 4. According to the chemi-
cal analysis, lunar regolith simulant consists mainly of
silicon, aluminum, calcium, and oxygen.
Fig. 2. (a) SEM image of the JSC-1A particles, (b) Magniﬁed
image of particles.
Fig. 3. Chemical composition (wt%) of lunar regolith and simu-
Fig. 4. EDX spectrum of lunar regolith simulant.
www.ceramics.org/ACT SLM of Lunar Regolith 3
Based on the weight percent measurements of the
EDX analysis and the elements mass calculation, the
results are similar to the data available for the simulant.
The results indicate that the material properties closely
match those of anorthite as expected to be present on
Additionally, the absorption of the lunar regolith
simulant in various wave lengths has been tested. As it
can be seen in Fig. 5, at a frequency of 1070 nm (reci-
procal value of wave number) corresponding to the Yb:
YAG ﬁber laser frequency, the regolith exhibits an
absorption unit of 1.1 which according to the Lambert–
Beer law (A=lg(I
/I)) represents 92.06% absorption
SLM/SLS of the ceramic and glass material could fol-
low different approaches regarding different applications.
A high geometrical accuracy could be achieved by sintering
the powder as melting the whole particles could lead to
shrinkage, baling effect, or formation of heat affected zone
(HAZ) around the melt pool; however, improved mechan-
ical properties could be achieved by SLM due to a stable
melt pool which results in parts with a density close to the
original material. Other properties such as crystallization
and color of the material could be realized by thermal
postprocessing or process modiﬁcation.
To keep the geometry reasonably accurate and to
provide acceptable mechanical properties, the following
strategy has been chosen: melting of lines which sur-
round the sintered powder particles. Based on previous
studies, it was shown that the different particles might
have also different sintering/softening/melting point.
new set of process parameters are ana-
lyzed. Regarding this, process parameters are set to be
50 W, 50 mm/s which showed an acceptable micro-
scopic view with a relatively homogenous surface. One-
layer object using these parameters is built, and the layer
penetration depth is measured. Figure 6 shows the pol-
ished cut surface of one layer of lunar regolith simulant
processed with SLM. The measurements are based on
the average of the minimum melt pool penetration
depths obtained from multiple single-layer objects.
Based on this and considering overlapping between
layers, although it is better to keep the re-exposure area
as small as possible to avoid thermal stress, the layer
thickness is chosen to be 100–300 lm. The utilization
of the optimum process parameter set enabled the
increase of the layer thickness beyond one single particle
layer which leads to a reduced processing time.
Furthermore, a 10 mm 910 mm 93 mm cube
has been fabricated using the newly obtained process
parameter set and scan strategy. Figure 7 shows the top
and 3D view of the fabricated sample as well as micro-
scopic view of cross-sectional view of the part. The
checker board pattern visible in the microscopic view of
the cut surface in Fig. 7c exhibits the molten tracks (dar-
ker lines) crossing each other surrounding sintered parti-
cles in square areas with a distance of 200 lm.
The surface morphology of the raw powder, a sin-
tered sample, and a molten part using SLM are shown
in Fig. 8a–c, respectively. In Fig. 8a, the individual
grains of the raw powder are clearly visible with explicit
edges. In the sintered part shown in Fig. 8b, the powder
grains are still visible albeit some sections of the grains
are now completely consolidated and others are adhered
together. In Fig. 8c, an unequivocal and complete melt
pool can be seen; the individual powder grains have
Fig. 5. Absorption spectrum of lunar regolith simulant versus
Fig. 6. Cross-sectional view of a single layer of lunar regolith
simulant fabricated using SLM.
(a) (b) (c)
Fig. 7. (a) Top view of cube geometry fabricated using SLM of
lunar regolith, (b) 3D view of the cube geometry fabricated using
SLM of lunar regolith, (c) Process layer pattern.
4International Journal of Applied Ceramic Technology—Fateri and Gebhardt 2014
completely molten within the melt pool which has then
solidiﬁed in a solid layer.
The fabricated part was analyzed using an X-ray dif-
fraction (XRD) machine to check whether the sample
had developed any crystallinity. The X-ray diffractograms
of the surfaces of the SLM-processed fabricated sample
were measured in the range of 40°≤2h≤140°using
CrKaradiation, while the JSC-1AC powder was tested
using CuKaradiation. As it is shown in Fig. 9, the
as-received JSC-1A powder is crystalline; however,
Fig. 10 illustrates that the SLM-processed samples are
amorphous. The crystal structure present in the powder
was eliminated during the melting and resolidiﬁcation.
As the XRD results show, the SLM process transformed
the crystalline structure of lunar regolith simulant into
The hardness of a part is also an important property
for engineering applications as such a surface hardness
test was conducted using the Berkovich method. The
load is held for 10 s at the maximum load. The test that
was conducted using a CSM nanoindenter machine was
repeated 12 times on each sample. The process used a
diamond tip by applying a maximum force of 160 mN
at a load rate of 320 mN per minute. Figure 11 shows
the loading procedure as well as the penetration depth of
diamond tip in the part.
The results show that the part surface had a hard-
ness of HVIT =1245 vickers. In comparison, a fused
silica reference part had a hardness value of HVIT =865
Furthermore, surface roughness as one of the repre-
sentative properties of part resolution is measured to be
an average value of Ra of 1.5 lm and Rz of 7.5 lm for
the last irradiated layer using sintering–melting method.
In the next step, various 3D geometries have been
built with a focus on different applications. Figure 12
shows multiple net shape matrix objects
30 mm 930 mm and 15 mm 915 mm which have
been built using a laser power of 50 W and scan speed
of 500 mm/s for the outer contours, while the inner pro-
ﬁle was irradiated using a laser power of 50 W and a
scan speed of 50 mm/s. The geometry has been selected
as it demonstrates the capability of SLM of lunar rego-
(a) (b) (c)
Fig. 8. (a) SEM image of lunar simulant powder, (b) Sintered
part embedded within a molten zone, (c) Fully melted and resolid-
Fig. 9. X-Ray diffractogram of lunar simulant (CuKaradia-
Fig. 10. X-Ray Diffractogram of fabricated SLM process parts
using lunar regolith simulant (CrKaradiation).
Fig. 11. (a) Penetration depth to normal force considering the
time, (b) Indentation graph.
Fig. 12. Multiple net shape objects fabricated using SLM process
with lunar simulant.
www.ceramics.org/ACT SLM of Lunar Regolith 5
lith for fabrication of miniature thin wall objects. Addi-
tionally, the matrix structure is one of the fundamental
objects needed in any kind of powder-based fabrication
technique as it can be used as a ﬁlter for the process
itself. The ﬁlter could be used for the raw powder prepa-
ration for the SLM process as well as recycling excess
unused powder from the machine. Furthermore, the
structure can be used for matrix-reinforcement composite
structures and possibly as a ﬁlter for liquid and gaseous
Based upon the successful fabrication of the various
matrix structures, a microscopic net shape matrix object
was also considered. Regarding this, a net shape matrix
object with cubic walls of 200 lm9200 lm and cubic
gaps of 250 lm9250 lm has been built. The micro-
scopic images of the part in two different magniﬁcations
are shown in Fig. 13. While some cracks are visible
throughout the parts under microscopy, the part exhib-
ited satisfactory properties which can potentially be
improved using a heated powder bed.
To further demonstrate the engineering applicability
of SLM process for lunar applications, a pair of M3 and
M5 nuts was fabricated. Ceramic fasteners can be used
for low-stress, high-temperature or corrosive applications.
The nuts were fabricated with an internal thread and
could be threaded on a screw as shown in Fig. 14.
Lastly, as shown in Fig. 15, multiple gears in vari-
ous sizes have been built to further validate the adequacy
of SLM of lunar regolith simulant in producing func-
tional parts. Similarly to fastener applications, the gears
can be used for low-stress, high-temperature or corrosive
environments. Using the obtained processe parameters,
the gears could be fabricated with reasonably high
dimensional accuracy as such, the gear pair teeth could
engage one another.
The fabricated objects had a high surface hardness,
but exhibited brittle behavior. Postprocessing as well as
an inﬁltration process could improve the parts
mechanical properties. Additionally, the compressive
strength of the material could be improved by adding
elements such as copper and iron to the base regolith.
To further investigate the feasibility of SLM process
for lunar application, the process should be tested
under vacuum/lunar atmospheric condition as well as
under lunar gravity to ensure that the process func-
tions adequately. Additionally, the feasibility of replac-
ing the laser beam heat source with other efﬁcient
alternatives such as solar lens should be studied. The
simulant could be used to study the applicability of
the SLM process for various applications such as sub-
strates for solar cells. Furthermore, in addition to the
JSC-1A simulant, the Minnesota Lunar Simulant
(MLS) which contains high titanium basalt could be
also tested using the SLM process.
The feasibility of SLM process using Yb:YAG ﬁber
laser with JSC-1A lunar regolith simulant has been inves-
tigated in this study. The raw powder properties were
examined, and the particle size distribution of the pow-
der was reduced to equalize the energy distribution
within the powder bed. One-layer objects were fabricated
to determine the melt pool penetration depth, and 3D
Fig. 13. Microscopic net shape geometry fabricated from lunar
simulant using SLM process.
Fig. 14. Nut geometry fabricated using SLM of lunar regolith.
Fig. 15. Multiple gear geometry built using SLM of lunar rego-
6International Journal of Applied Ceramic Technology—Fateri and Gebhardt 2014
test geometries were built using a developed scanning
strategy which melted speciﬁc tracks with a sintered area
between the tracks. Using this strategy, high geometrical
accuracy as well as favorable part quality has been
achieved. The RDX investigation of the raw powder and
SLM-processed parts has revealed that the raw lunar reg-
olith simulant has a crystalline structure, while the SLM
processed parts exhibit a completely amorphous struc-
ture. It is shown that relative dense parts could be
fabricated if the process parameters such as powder parti-
cle size and distribution, fabricating strategy, laser pow-
der, and scan speed are optimized. The surface hardness
tests of the fabricated objects showed a high surface
hardness value of HVIT =1245 vickers, which is 44%
higher than the value of a reference fused silica part
using the same equipment. Lastly, various practical and
geometrically complicated test parts in the millimeter
and micrometer range were fabricated and presented.
The investigation established the feasibility of using lunar
regolith with SLM or other similar AM processes to fab-
ricate practical engineering objects for on-site manufac-
turing on the moon.
1. NASA, “Ames Technology Capabilities and Facilities”, Available at http://
Utiliza14.html (accessed September 2, 2012).
2. J. Kaplicky and D. Nixon, Lunar Bases and Space Activities of the 21st Cen-
tury. ed., W. W. Mendell. Lunar and Planetary Institute, 1  275–279,
3. F. Ruess, J. Sch€anzlin, and H. Benaroya, J. Aerosp. Eng., 19  133–157
4. B. Khoshnevis, Int. J. Ind. Syst. Eng., 1  301–320 (2006).
5. B. Khoshnevis, “Extraterrestrial Construction Using Contour Crafting”, Solid
Freeform Fabrication symposium, Austin, TX, 2012.
6. R. N. Grugel, Adv. Space Res. NASA, 50  1294–1299 (2012).
7. L. C. Simonsen, M. J. DeBarro, and J. T. Famer, Second Conference on Lunar
Bases and Space Activities of the 21st Century. eds., W. W. Mendell. NASA
Conferences Publication, Hampton, VA, 3166, Vol. 1, 579–592, 1992.
8. G. Cesarettia, E. Dinib, X. DeKestelierc, V. Collad, and L. Pambaguiane,
Acta Astronaut., 93 430–450 (2014).
9. V. K. Balla and L. B. Roberson, Rapid Prototyp. J., 18  451–457 (2012).
10. M. Fateri, A. Gebhardt, and M. Khosravi, Int. Mech. Eng. Congress. Expo.,
2A V02AT02A008–V02AT02A008 (2013).
11. M. Fateri and M. Khosravi, LPI Contrib., 1679 4368 (2012).
12. J. L. Carter, D. S. McKay, L. A. Taylor, and D. W. Carrier, Proc. Space Re-
sour. Roundtable, VI 886–887 (2004).
13. R. V. Morries, R. Score, C. Dardano, and G. Heiken, Handbook of lunar
soils, Johnson Space Center Planetary Materials Branch, Houston, TX, 67,
14. Anonymous, Material Safety Data Sheet, Lunar Mare Regolith Simulant
JSC-1, Orbitec Technologies Corporation, Retrieved 2014.
www.ceramics.org/ACT SLM of Lunar Regolith 7