Materials Science and Engineering A 383 (2004) 191–200

Direct laser sintering of iron–graphite powder mixture

A. Simchi a,∗ , H. Pohl b
a Department of Materials Science and Engineering, Sharif University of Technology, P.O. Box 11365-9466, Azadi Avenue, Tehran, Iran
b Fraunhofer Institute for Manufacturing and Advanced Materials (IFAM), Wiener Strasse, 28359 Bremen, Germany

Received 17 January 2004

Abstract

In the present work, the role of graphite addition on the laser sintering of iron powder was studied. Powder mixtures containing iron
and 0.4, 0.8, 1.2, and 1.6 wt.% graphite were prepared by blending elemental powders. These powders were sintered layer-by-layer under
nitrogen atmosphere using a continuous wave CO2 laser beam. A laser power of 70–225 W, scan rate of 50–600 mm s−1 , scan line spacing
of 0.1–0.3 mm, and layer thickness of 0.1 mm was used. It was found that the processing parameters play a key role on the densification of
the iron–graphite powder mixtures. The addition of graphite enhances the densification of the iron powder and improves the surface quality
of the laser sintered parts when optimized manufacturing conditions are applied. The graphite content has a significant influence on the
internal pore structure of the sintered parts. They are gradually changed from interconnected networks to closed and spherical shaped pores
with increasing graphite content. The metal matrix structure consists of different phases such as ferrite, austenite, and tempered martensite,
which highlights the heterogeneous distribution of dissolved carbon in the iron matrix. This article presents the experimental details of the
microstructural evolution in laser sintered iron–graphite powder mixtures. The role and key importance of graphite addition to iron powder
in the laser sintering process is addressed.
© 2004 Elsevier B.V. All rights reserved.

Keywords: Direct laser sintering; Iron powder; Graphite; Densification; Microstructure; Rapid prototyping

1. Introduction Among different RP techniques, the direct metal laser

Rapid prototyping (RP) technology is a relatively new
material additive manufacturing process which enables the
quick fabrication of a three-dimensional part of arbitrary
shape directly from CAD data [1]. Unlike the conventional
subtractive manufacturing methods such as machining, this
process is unconstrained by the limitations attributed to
specially design tooling and fixturing [2]. Therefore, almost
any shape of geometry with variations in size and complex-
ity can be produced to a high degree of accuracy. A wide
variety of materials such as polymers, waxes, papers, met-
als, ceramics, and composites are currently used for rapid
prototyping, and for the fabrication of functional prototypes
by means of various methods [3]. These methods build
parts layer-by-layer using photochemical, laser sintering,
extrusion layering, gluing, materials deposition, etc [4].
∗ Corresponding author. Tel.: +98 21 616 5226; fax: +98 21 600 5717.
E-mail addresses: simchi@sharif.edu (A. Simchi),
pohl@ifam.fraunhofer.de (H. Pohl).
sintering (DMLS) process exhibits a high potential for the
net-shape fabrication of prototypes and short series tooling
for plastic injection molding and die casting [5,6]. In this
method, a localized region of a thin powder layer is con-
solidated by the energy of a focused laser beam according
to a sliced 3D CAD model. This enables the fabrication
of complex shape parts without using conventional cost in-
tensive shaping methods. Nevertheless, the potential of this
process has not been explored in terms of material versatil-
ity, quality, and precision [7]. Parts of full density are rarely
achieved and thus a post-treatment, e.g. epoxy or metal
infiltration, is often required. In this particular context, the
process may be useful in synthesizing porous materials [8].
Recently, considerable research effort has been reported
in the area of laser sintering. The main trends of these
investigations were directed towards material develop-
ment, microstructural characterization, and consolidation
mechanism. Direct laser sintering of metal powders such
as iron [9], nickel [10], titanium [7], bronze [11], steels
(plain carbon [12], low alloy [13–15], stainless [16,17], and

0921-5093/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.msea.2004.05.070
192 A. Simchi, H. Pohl / Materials Science and Engineering A 383 (2004) 191–200
high speed steels [18–20]), and nickel-base alloys [1,21]
have been reported. The fabrication of composite parts,
e.g. Cu-TiB2 [22] and TiC-Al2 O3 [23], ceramic artifacts
[24,25], and hard metals [26,27] by the laser sintering pro-
cess were also studied. Although recent advances in DMLS
have improved the technology considerably, the method es-
sentially relies on empirical, experimental knowledge and
still lacks a strong theoretical basis [7]. In fact, not much
work has previously been reported on the basic principles
of this process and on the powder bonding mechanisms.
This may be attributed to the complex nature of the process,
which exhibits multiple modes of heat, mass and momen-
tum transfer, and chemical reactions. Although some efforts
was put in the enhancement of the basic knowledge of this
process [24,28–31], significant research and development
are still required for the fabrication of high performance
engineering parts with controlled microstructure.
Among different material systems, DMLS of steel pow-
ders is of particular importance regarding to the process ben-
efits in the fabrication of prototype tools directly from the
CAD models. This rapid tooling method leads to compressed
time to market solution, having a huge impact on chang-
ing product design and test procedure [2,32]. Nevertheless,
there is still a lack in understanding how the microstructure
and properties of the processed material are influenced by
the processing parameters. DMLS of iron powder and the
effects of powder characteristics and manufacturing pa-
rameters have been reported previously [9]. In the present
work, the laser sintering of iron–graphite powder mixtures
was studied in order to evaluate the role of graphite on the
densification and the attendant microstructural features of
iron powder in DMLS process. Plain carbon steels were
examined since the presence of other alloying elements,
e.g. Cu and Mo, may influence the bonding mechanism
and the resulting microstructure. Not much work has been
previously reported on the laser sintering of iron–graphite
powder mixture. To date, it has been known that laser sin-
tered iron–graphite parts are porous and their microstructure
and mechanical properties are substantially different from
the same produced by conventional sintering [12]. This
article presents the sintering behavior, surface morphology,
and microstructure of iron–graphite powder mixtures (up
to 1.6 wt.% graphite) processed by a continuous wave CO2
laser beam. The mechanism of particle bonding and the
role of graphite on the laser sintering of iron powder are
addressed.
Table 1
Characteristics of iron powders used in this study
Iron powder D10 (␮m) D50 (␮m)
Water atomized 34 68
Carbonyl 4 10
a Particle size distribution slope: 2.56/log(D90 /D10 ).
2. Experimental
The starting powders used for the laser sintering of
iron–graphite alloys were as follows: water atomized iron
powder (ASC 100.30 Hoeganaes, 0.0612 wt.% O); carbonyl
iron powder (BASF, 0.12 wt.% O); fine graphite powder
(2 ␮m, Aldrich). The characteristics of the iron powders,
accessed according to the MPIF standard test methods [33],
are given in Table 1. The particle size distribution was deter-
mined using Coulter LS130 laser particle size analyzer. The
surface area was measured through BET method using nitro-
gen gas by a NOVA220 (Quantachrome Instruments, USA).
The as received atomized iron powder was screened using
a 0.1 mm sieve and the subseive powder was mixed with
20 wt.% carbonyl iron in a tumbling mixer for 20 min. This
way decreases the mean particle size of the powder bed
and increases the bulk density of the mixture which in fact
enhances the sintering rate. Homogeneous powder blends of
0.4, 0.8, 1.2, and 1.6 wt.% graphite and iron were prepared
by mixing the elemental powders in a tumbling mixer for
40 min. For comparison, one batch of the plain iron powder
mixture without graphite addition was also prepared.
Rectangular test specimens with dimensions of 10 mm ×
10 mm × 7 mm were produced using EOSINT M250Xtended
laser sintering machine (Electro Optical Systems GmbH,
Germany). Fig. 1 shows a schematic picture of this instru-
ment. The machine consists of a powder handling system, a
continuous wave carbon dioxide laser with maximum power
of 250 W and 0.4 mm spot size, laser optics and a process
computer. The 3D CAD model of the test specimens was
prepared using the Pro/Engineer software. The model was
then converted to triangulated surface model in the standard
STL format. The STL format was sliced to thin horizontal
layers of 0.1 mm thickness. The data preparation step was
followed by the laser sintering process. First, a steel base
plate was placed on the building platform (XY table) and
leveled. Then, a powder layer (about 0.1 mm in thickness)
was spread on the base plate using a moving wiper (mechan-
ical re-coater). Thereafter, the laser sintering started by scan-
ning the powder according to the STL file under a nitrogen
atmosphere. The working output power of P = 70–225 W,
scan rate of v = 50–600 mm s−1 , and scan line spacing
of h = 0.1–0.3 mm was applied. The powder bed temper-
ature was kept constant at 80 ◦ C during processing. This
process was repeated and the test specimens were produced
layer-by-layer according to the sliced STL file.
D90 (␮m) SW a Mean Surface Apparent Tap
value area density density
(␮m) (m2 g−1 ) (g cm−3 ) (g cm−3 )
108 5.1 69.4 0.219 3.1 3.8
29 3 13.4 0.420 3.4 4.0
 A. Simchi, H. Pohl / Materials Science and Engineering A 383 (2004) 191–200 193

Fig. 1. Schematic representation of the direct laser sintering apparatus.

After removing the samples from the build plate, the

density of the specimens was measured by using the volu-
metric method. Each processing condition was repeated at
least two times and the result of the density measurement
was expressed using the mean value. The standard devi-
ation is <0.02 g cm−3 . Microstructural evaluations were
performed by optical and scanning electron microscopy
(SEM). Samples for metallographic examination were pre-
pared using standard techniques and etched in 2% nital
(2 ml HNO3 per 100 ml CH3 OH). A LEO 438 VP scanning
electron microscopy was used to take SEM micrographs.
The total carbon content of laser sintered parts was deter-
mined using LEO 444 instrument. The local concentration
of carbon in the iron matrix was also evaluated by X-ray
photoelectron spectroscopy (XPS). A HMV-2000 Shimadzu
micro-hardness tester was used to determine the hardness
of the microstructural phases using a 0.025 kg load.
3. Results
3.1. Density
Fig. 2 show the sintered density of the iron–graphite
powder mixtures as a function of laser power at scan rate of
75 mm s−1 and scan line spacing of 0.1 and 0.3 mm. The re-
sults demonstrate the importance of the scan line spacing (h)
on the densification of the blends. While at h = 0.1 mm, the
density of iron powder was not influenced by the graphite ad-
ditions, significant densification was obtained at h = 0.3 mm
(Fig. 3). It was previously shown [9], that in the case of plain
iron, overlapping of scan lines increases the sintered density,
even though at very intense laser energy input the density
Fig. 2. Sintered density of iron–graphite powder mixtures (Fe–xG: x shows
wt.% of graphite in the mixture) as a function of laser power at scan rate
of 75 mm s−1 , layer thickness of 0.1 mm, and scan line spacing of: (a)
0.1 mm and (b) 0.3 mm.
would reach to a plateau, i.e. higher densification cannot be
obtained due to delamination of the sintered layers. In oppo-
site to this observation, the laser sintering of iron–graphite
powder mixtures seems to be more sensitive to the scan line
spacing. Similar results were obtained when other manu-
facturing parameters were taken into account. For instance,
Fig. 4 shows the iso-density curves of examined materials in
a laser power-graphite content diagram. One can notice that
the effect of laser power on the density is more pronounced
with increasing graphite content. Therefore, the densifica-
tion kinetics of iron powder is enhanced by the addition of
graphite powder and it is more sensitive to the laser energy
input. Fig. 5 shows the effect of the laser scan rate on the
sintered density of iron and iron–0.8 wt.% graphite pow-
ders. A significant improvement (∼15% theoretical density)
of the densification was obtained by adding graphite to iron
powder at different scan rates. In addition, the critical scan
194 A. Simchi, H. Pohl / Materials Science and Engineering A 383 (2004) 191–200

Fig. 3. Effect of graphite content on the sintered density of iron–graphite
powder mixtures dependent on laser power (P) and scan line spacing (h);
scan rate is 75 mm s−1 and layer thickness is 0.1 mm.
rate (the highest scan rate for which sound parts can be
obtained) increased from 150 mm s−1 for iron powder to
600 mm s−1 for the iron–0.8 wt.% graphite powder mixture.
3.2. Microstructure
Fig. 6 shows the pore structure of laser sintered parts at
laser power of 215 W, scan rate of 75 mm s−1 , scan line
spacing of 0.3 mm, and layer thickness of 0.1 mm. Gener-
ally, the microstructure consists of large pores. However,
the connectivity, the orientation, and the shape of the pores
depend on the graphite content of the processed powder
Fig. 5. Effect of laser scan rate on the sintered density of iron powder
and iron–0.8 wt.% graphite powder mixture processed at laser power of
215 W, scan line spacing of 0.3 mm, and layer thickness of 0.1 mm.
blends. The pore structure of laser sintered iron is shown
in Fig. 6a. A continuous network of interconnected pores
around fully dense iron agglomerates is visible. The ori-
entation of pore channels is aligned towards the 45◦ angle
of the building direction. With the addition of graphite to
the iron powder, no continuous network of pore channels
was observed. Instead, irregular shaped pores surrounded by
a fully dense iron matrix are formed Fig. 6b and c). The
microstructure consists of some small and spherical pores
while no pore orientation can be realized. With increas-
ing graphite content, the total porosity decreases gradually,
and more spherical pores were found in the microstructure
(Fig. 6d).

Fig. 4. Iso-density curves (relative to the theoretical density, %) of sintered iron–graphite powder mixtures as a function of laser power and graphite
content; scan rate is 75 mm s−1 , scan line spacing is 0.3 mm, and layer thickness is 0.1 mm.
 A. Simchi, H. Pohl / Materials Science and Engineering A 383 (2004) 191–200 195

Fig. 6. Polished section of laser sintered parts shows the effect of graphite addition on the pore structures on a section cut parallel to the building
direction; laser power is 215 W, scan rate is 75 mm s−1 , scan line spacing is 0.3 mm, and layer thickness is 0.1 mm.

Fig. 7 shows the characteristic microstructure of laser sin-
tered iron–1.2 wt.% graphite powder mixture processed at
laser power of 215 W, scan rate of 75 mm s−1 , scan line spac-
ing of 0.3 mm, and layer thickness of 0.1 mm. The local car-
bon concentration analyzed by XPS and the microhardness
values at different regions of the microstructure are listed
in Table 2. This information clearly indicates the heteroge-
neous distribution of carbon in the iron matrix, which led to
the formation of different phases. Similar results were ob-
tained for the other sintered parts, although the heterogeneity
is more pronounced in the case of higher graphite content.
For instance, Fig. 8 shows the characteristic microstructure
of laser sintered iron–1.6 wt.% graphite powder mixture. A
ferrite core surrounded by a carbide phase, white areas of re-
tained austenite, lattes of martensite, and the eutectic mixture
of austenite and cementite known as ledeburite are visible.

Fig. 7. Characteristic microstructure of laser sintered iron–1.2 wt.% graphite powder mixture on a section cut parallel to the building direction shows
heterogeneous carbon dissolution in the iron matrix resulting in the varying local hardness values (see Table 2). Laser power is 215 W, scan rate is
75 mm s−1 , scan line spacing is 0.3 mm, and layer thickness is 0.1 mm.
196 A. Simchi, H. Pohl / Materials Science and Engineering A 383 (2004) 191–200

Table 2
The local carbon concentration measured by XPS method and microhard-
ness values of marked areas in Fig. 7
Area Possible phases Carbon Hardness
content (wt.%) (HV0.025)
I Ferrite 0.1 137
II Tempered martensite 1.6 463
III High carbon austenite 1.9 476
3.3. Surface morphology
Fig. 9 illustrates the effect of graphite addition on the
surface morphology of laser sintered iron processed at laser
power of 215 W, scan rate of 75 mm s−1 , scan line spac-
ing of 0.3 mm, and layer thickness of 0.1 mm. In the case
of plain iron powder, the solid particles were bonded to-
gether to form a network of large agglomerates (Fig. 9a).
Between the iron agglomerates, large pore channels are
visible. Here, it is important to point out that the scan
pattern used in this study for direct laser sintering was
alternating from layer-to-layer while the same line spac-
ing in X- and Y-directions was equal. This suppresses the
formation of columnar agglomerates parallel to the scan
direction. When 0.8 wt.% graphite was mixed with iron
powder, smaller solid agglomerates were formed (Fig. 9b).
Consequently, a smoother surface was obtained. At graphite
content of 1.6 wt.%, a relatively smooth and almost fully
dense sintered surface was obtained (Fig. 9c). In fact, a pro-
gressive transition from highly rippled surface consisting
of large solid agglomerates to a smooth surface with fully
dense sintered layer occurred with increasing the graphite
content.
4. Discussion
It is known that the laser sintering of metals is a very
complex process. When iron powder is irradiated by a laser
beam, multiple modes of mass, heat and momentum transfer,
and chemical reactions occur. The local temperature gradi-
ents in the molten pool can give rise to surface tension gra-
dient and associated Marangoni effect [34,35]. Due to the
capillary instability effect, the liquid iron breaks up to a row
of spheres to reduce the surface area [20]. The formation
of iron balls is likely to occur when the molten metal does
not wet the underlying substrate due to an oxide layer being
present on the substrate and on the surface of the melt [31].
The melting is followed by rapid solidification, in which a
network of metallic agglomerates is formed (Fig. 9a). Con-
sequently, the part surface is very rough and the microstruc-
ture consists of large pores and fully dense iron agglomerates
(Fig. 6a). Here, it is important to point out that if a sorted
scanning pattern is used, e.g. sorted in X-direction, rows
of columnar iron agglomerates and oriented pores would
be formed parallel to the scan- and building-direction, re-
spectively [9]. In this study, an alternating pattern, chang-
ing from one layer to the next layer was used which dimin-
ishes the formation of oriented microstructure as explained
above.
The addition of graphite to iron powder enhances the den-
sification and influences the microstructural development
during laser sintering. The significant effects of graphite ad-
dition can be highlighted as follows:
• The sintered density increases with increasing graphite
content of the powder mixture when scan line spacing of
0.3 mm is applied (Figs. 2 and 3).

Fig. 8. Characteristic microstructure of laser sintered iron–1.6 wt.% graphite powder mixture on a section cut parallel to the building direction; laser
power is 215 W, scan rate is 75 mm s−1 , scan line spacing is 0.3 mm, and layer thickness is 0.1 mm.
 A. Simchi, H. Pohl / Materials Science and Engineering A 383 (2004) 191–200
Fig. 9. SEM images of laser sintered surfaces using alternating scan-
ning strategy in X- and Y-directions; laser power is 215 W, scan rate is
75 mm s−1 , scan line spacing is 0.3 mm, and layer thickness is 0.1 mm.
• The structure of pores is changed from a continuous net-
work of interconnected pores to closed and spherical like
shape with increasing graphite content (Fig. 6).
• The metal matrix structure is heterogeneous and with in-
creasing the graphite content this heterogeneity increases
(Figs. 7 and 8).
• The surface of laser sintered iron–graphite powder mix-
ture is smoother than that of sintered iron, and the forma-
tion of iron balls and agglomerates is less likely to occur
(Fig. 9).
197
These observations may be attributed to several micro-
and macro-events (e.g. heat and mass transfer, and chemi-
cal reactions) occurring during laser sintering. However, the
phenomena, which would be much more effective than the
others, are highlighted and explained below.
4.1. Increasing the absorptance of powder bed
It is evident that the absorption rate of a powder bed signif-
icantly influences the working temperature and the sintering
kinetics. Basically the absorption rate of metals depends on
their electrical conductivity, surface condition, and the laser
wavelength [36]. In this context, the formation of oxides on
the surface of powder particles as well as the multiple re-
flection of the laser beam between the particles enhances
the absorption rate significantly [7,37]. These effects result
in a lower laser power input requirement for melting of iron
powder. Nevertheless, when powder is fusing during laser
irradiation and forms a cluster-like shape, the laser beam is
reflected mainly by this cluster, resulting in a decrease in
the absorptance [38].
In the case of iron–graphite powder mixture, the iron
particles are covered with graphite lamella during the mix-
ing procedure [39,40]. This graphite layer appears to the
incident radiation as grey body increasing the absorptance
considerably. On the hand, during the laser sintering of
iron–graphite powder mixtures, the formation of iron ag-
glomerates is less likely to occur (Fig. 9). In other words,
the addition of graphite to iron powder suppresses the for-
mation of iron clusters, therefore the reflectivity decreases.
This phenomenon increases the working temperature and
enhances the sintering kinetics which results in higher at-
tainable sintered density.
4.2. Changing the surface tension and viscosity of iron
melt pool
Particle melting by laser irradiation may result in the for-
mation of convective streams within the molten pool. Such
streams play an important role in material transport dur-
ing laser processing. Simultaneous heat and mass transport
are governed by thermokinetic and thermocapillary effects,
but primarily by surface tension gradient [35]. This gradi-
ent results in shear stress and convective movement of the
melt pool (Marangoni effect). The sign and magnitude of
the surface tension gradient to temperature (dγ/dT ) deter-
mines the flow pattern of the material [13,34]. The changes
in surface tension lead to different molten tracks formed by
line scanning of the powder bed. On the other hand, it is
known that the surface tension of the iron melt depends on
the amounts of dissolved elements like carbon, oxygen, ni-
trogen, boron, and phosphorous [41]. Generally, these ele-
ments decrease the surface tension of molten iron. However,
the effect of carbon is more pronounced. During the laser
sintering of iron–graphite powder mixture, it is most likely
that the melting of the iron powder occurs due to the lower
198 A. Simchi, H. Pohl / Materials Science and Engineering A 383 (2004) 191–200
Fig. 10. Carbon concentration of laser sintered parts versus the graphite
content of the powder mixture at laser powers (P) of 100 and 215 W.
Scan rate of 75 mm s−1 , scan line spacing of 0.3 mm, and layer thickness
of 0.1 mm was applied.
melting point of iron in contrast to graphite (1537 ◦ C for iron
and >3500 ◦ C for graphite). However, due to the high tem-
perature and associated high diffusion rate, graphite tends
to dissolve in the iron melt. The dissolution of carbon into
the iron decreases the surface tension and viscosity of the
molten pool considerably. Fig. 10 shows the combined car-
bon of the laser sintered parts as a function of the graphite
content at two laser powers of 100 and 215 W. The results
show that not much carbon loss occurred during laser sinter-
ing in nitrogen, and it was almost independent on the laser
power. Nevertheless, the carbon loss was increased when the
concentration of graphite in the powder mixture was higher.
It is possible that some amounts of graphite got oxidized
due to the presents of trace amounts of oxygen in the sinter-
ing chamber (present in the inert gas and dissolved in iron
powder) or got vaporized off [12]. Anyway, with both de-
creasing the surface tension and the viscosity of the molten
pool, the flow of the melt is improved and the formation of
iron agglomerates is less likely to occur. These effects sup-
press the formation of large pores and increase the attainable
density. The sintered surface layer is also smoother (Fig. 9).
However, metallographic observations demonstrated that
carbon is not homogenously dissolved in the iron matrix
(Figs. 7 and 8) even though the diffusion rate of carbon into
the iron is very high at the operating temperature. This can be
attributed to heterogeneous distribution of graphite particles
in the powder bed, e.g. due to powder segregation during the
course of layer coating. In addition, localized mass transfer
and material convection within the molten pool caused by the
temperature and chemical gradients, in accordance with the
subsequent rapid solidification process result in the forma-
tion of different phases and in a heterogeneous microstruc-
ture. For instance, in the case of iron powder containing high
amounts of graphite, the formation of iron–graphite eutec-
tic was detected (Fig. 8). This observation demonstrates that
during the course of laser sintering, carbon dissolved into
the iron melt, thereby the low melting point eutectic phase
was formed. This melt was subsequently quenched due to
the high cooling rate, and transformed to ledburitic struc-
ture. Murali et al. [12] have shown that during sintering of
iron–0.78 wt.% graphite powder mixture using an Nd:YAG
laser (pulsed, average power 100 W) carbon is not com-
pletely dissolved into the iron, and the iron globules are hav-
ing graphite particles at their periphery. In the present work,
a continuous CO2 laser beam with higher average power of
215 W was used. Under this condition, almost complete dis-
solution of carbon into the iron was obtained (free graphite
was not observed or analyzed). However, rapid quenching
due to the high cooling rate suppressed complete homoge-
nizing of carbon in the iron matrix, thereby different phases
were formed.
4.3. Decreasing the oxygen content
The amount of oxygen present during the heating, melt-
ing and fusion of the metal powder during laser sintering
strongly influences the densification and the resulting mi-
crostructural features. It is known that the presence of oxy-
gen allows surface oxides and slags to form as the powder
is heated and melted by the scanning laser beam [16]. The
formation of oxide layer on the surface of powder particles
significantly increases the absorption rate of CO2 laser radi-
ation [37]. This changes the temperature–time history dur-
ing sintering and increases the melt volume allowing surface
tension to become more dominant. With reference to the
Marangoni effect, this leads to the formation of different scan
tracks: wide and shallow for a low oxygen content or deep
and narrow for a high oxygen concentration [18]. Another
concern is the liquid metal surface tension which influences
the wetting angle between the solid and the liquid phases
that can disrupt bonding between rastered lines and individ-
ual layers [16,24]. The formation of a larger melt volume,
dominating the surface tension force, and decreasing the
wetting angle causes the melt pool to solidify into agglomer-
ates. Fig. 11 depicts the influence of dissolved carbon on the
oxygen concentration of laser sintered iron–graphite powder
mixtures. The oxygen concentration of laser sintered iron
part (C ∼ 0.07 wt.%) is about 0.13 wt.% which is consider-
ably higher than that of the original powder (0.06 wt.%). This
observation demonstrates that oxygen is incorporated during
the laser processing, although the powder was sintered un-
der nitrogen atmosphere. Additions of 0.4 wt.% graphite to
the iron powder decreased the oxygen content considerably
to about 0.04 wt.%. Lower oxygen concentration was mea-
sured when powder mixtures comprising of higher graphite
contents were processed. Nevertheless, the effect of graphite
addition on the oxygen level is less pronounced at higher
concentrations. These results indicate that some amounts of
graphite acted as a reducing agent during the course of laser
sintering. This can be one source of carbon loss as it was
 A. Simchi, H. Pohl / Materials Science and Engineering A 383 (2004) 191–200
Fig. 11. Effect of combined carbon on the oxygen concentration of sintered
iron–graphite parts under nitrogen at laser power of 215 W, scan rate of
75 mm s−1 , scan line spacing of 0.3 mm, and layer thickness of 0.1 mm.
shown in Fig. 10. Anyway, decreasing the oxygen concen-
tration leads to lower melt volume and reduces the dominat-
ing influence of the surface tension. In addition, the wetting
angle between the solid and liquid phases is changed and
better bonding between rastered lines and individual layers
are obtained. Under this condition, the formation of iron ag-
glomerates is less likely to occur. As cited above, the scan
track is also influenced, resulting in the formation smoother
surfaces (Fig. 9).
The experiments results showed that when a scan line
spacing of 0.1 mm was used, the sintered density is almost
independent of the graphite content (Figs. 2 and 3). It is
known that intensifying the energy input results in the devel-
opment of residual thermal stresses. These stresses are re-
sponsible for reduced part performance as well as warpage,
loss of edge tolerance and even delamination of layered de-
posited parts [42,43]. Additionally, it has been shown above
that during laser sintering of iron–graphite powder mix-
tures the formation of hard and brittle phases, e.g. marten-
site and/or tempered martensite, is likely to occur. These
phases decrease the toughness of the sintered part, thereby
delamination of the layers due to the development of thermal
stresses is more susceptible. Under this condition, although
addition of graphite to iron powder enhances the sintering
kinetics, the formation of large cracks decreases the sintered
density quite similar to sintered iron as presented elsewhere
[9].
5. Conclusions
The influence of graphite addition on the laser sintering
of iron powder was investigated. It was found that if very
intense laser energy input is applied, due to the formation of
199
large cracks the sintered density would be almost indepen-
dent of the graphite content. In other words, although almost
fully dense iron agglomerates are formed, the delamination
of the sintered layers decreases the density. On the other
hand, the results showed that graphite addition can increase
the sintering kinetics of iron powder considerably if proper
DMLS parameters are applied. This effect is mainly due to
the impact of dissolved carbon, i.e. it changes the surface
tension and viscosity of the iron melt formed during laser
irradiation. Moreover, a decrease in oxygen content of the
iron melt through in-situ reduction by graphite and/or dis-
solved carbon plays a significant role on the material con-
vection in the molten pool. This results in the formation of
different scan tracks, i.e. wide and shallow for a low oxygen
content and deep and narrow for a high oxygen concentra-
tion, which in fact changes the pore structure and the surface
roughness of the sintered layers. The structure of internal
pores is changed from large and interconnected network to
spherical and closed pores. Important is the fact that the dis-
solution of carbon into the iron matrix was almost complete
when a continuous laser source was used. Since the sinter-
ing cycle in DMLS process is very short, the time for the
homogenizing of carbon in the iron matrix is too short, re-
sulting in the formation of very heterogeneous microstruc-
ture. For instance, iron ferrite with low carbon concentration
was observed nearby high carbon retained austenite when
evaluating the microstructure of laser sintered iron–graphite
powder blends. Low and high carbon martensite/tempered
martensite were also detected, owning to the high cooling
rate of the melt pool caused by underlying layers.
Acknowledgements
A. Simchi would like to thank the Office of Vice President
for Research and Technology, Sharif University of Technol-
ogy for partial financial support. The authors also acknowl-
edge Dr. Frank Petzoldt, Fraunhofer Institute IFAM, for his
input during valuable discussions.
References
[1] F. Abe, K. Osakada, M. Shiomi, K. Uematsu, M. Matsumoto, The
manufacturing of hard tools from metallic powders by selective laser
melting, J. Mater. Process. Technol. 111 (2001) 210.
[2] D. King, T. Tansey, Alternative materials for rapid tooling, J. Mater.
Process. Technol. 121 (2002) 313.
[3] T. Wohlers, Rapid Prototyping, Tooling, and Manufacturing: State
of the Industry, Annual Worldwide Progress Report 2003, Wohlers
Associates, Inc., Fort Collins, CO, 2003.
[4] P.T. Pham, S. Dimov, Rapid Manufacturing: The Technolo-
gies and Applications of Rapid Prototyping and Rapid Tooling,
Springer–Verlag, London, 2001.
[5] J. Haenninen, DMLS moves from rapid tooling to rapid manufac-
turing, Metal Powder Report, September 2001, pp. 25–29.
[6] D. Atkinson, Rapid Prototyping and Tooling: A Practical Guide,
Strategy Publication, Herts, 1997.
200 A. Simchi, H. Pohl / Materials Science and Engineering A 383 (2004) 191–200
[7] P. Fischer, V. Romano, H.P. Weber, N.P. Karapatis, E. Boillat, R.
Glardon, Acta Mater. 51 (2003) 1651.
[8] A.V. Gusarov, T. Laoui, L. Froyen, V.I. Titov, Int. J. Heat Mass
Transfer 46 (2003) 1103.
[9] A. Simchi, H. Pohl, Effects of laser sintering processing parameters
on the microstructure and densification of iron powder, Mater. Sci.
Eng. A359 (2003) 119.
[10] N.P. Karapatis, G. Egger, P.-E. Gygax, R. Glardon, in: Proceedings
of the Solid Freeform Fabrication Symposium, University of Texas,
Austin, 1999, pp. 255–263.
[11] G.B. Prabhu, D.L. Bourell, in: Proceedings of the Solid Freeform Fab-
rication Symposium, University of Texas, Austin, 1993, pp. 317–324.
[12] K. Murali, A.N. Chatterjee, P. Saha, R. Palai, S. Kumar, S.K. Roy,
P.K. Mishra, A. Roy Choudhury, J. Mater. Process. Technol. 136
(2003) 179.
[13] A. Simchi, F. Petzoldt, H. Pohl, Int. J. Powder Metall. 37 (2) (2001)
49.
[14] A. Simchi, F. Petzoldt, H. Pohl, H. Löffler, P/M Sci. Technol. Briefs
3 (1) (2001) 5.
[15] A. Simchi, F. Petzoldt, H. Pohl, J. Mater. Process. Technol. 141
(2003) 319.
[16] C. Hauser, T.H.C. Childs, K.W. Dalgarno, R.B. Eane, in: Proceedings
of the Solid Freeform Fabrication Symposium, University of Texas,
Austin, 1999, pp. 265–272.
[17] C. Hauser, T.H.C. Childs, K.W. Dalgarno, in: Proceedings of the
Solid Freeform Fabrication Symposium, University of Texas, Austin,
1999, pp. 273–280.
[18] H.J. Niu, I.T.H. Chang, Selective laser sintering of gas and water
atomized of high speed steel powders, Scripta Mater. 41 (1) (1999)
25.
[19] H.J. Niu, I.T.H. Chang, Scripta Mater. 39 (1) (1998) 67.
[20] H.J. Niu, I.T.H. Chang, Scripta Mater. 41 (1) (1999) 1229.
[21] K.W. Dalgarno, C.S. Wright, Powder Met. Prog. 1 (1) (2001) 70.
[22] C.C. Leong, L. Lu, J.Y.H. Fuh, Y.S. Wong, Mater. Sci. Eng. A338
(2002) 81.
[23] A. Slocombe, L. Li, J. Mater. Process. Technol. 118 (2001) 173.
[24] D.L. Bourell, H.L. Marcus, J.W. Barlow, J.J. Beaman, Int. J. Powder
Metall. 28 (4) (1992) 369.
[25] R. Lenk, Adv. Eng. Mater. 2 (1–2) (2002) 40.
[26] T. Laoui, L. Froyen, J.P. Kruth, Proc. PM World Congress, Granada
5 (1998) 394.
[27] T. Laoui, J. Bonse, J.P. Kruth, L. Froyen, K.U. Leuven, Int. J. Adv.
Manufact. Technol. 14 (3) (1998) 271.
[28] D.E. Bunnell, S. Das, D.L. Bourell, J.B. Beaman, H.L. Marcus, in:
Proceedings of the Solid Freeform Fabrication Symposium, Univer-
sity of Texas, Austin, 1995, pp. 440–447.
[29] Y.P. Kathuria, Surf. Coat. Technol. 116–119 (1999) 643.
[30] G. Lewis, E. Schlienger, Mater. Des. 21 (2000) 417.
[31] S. Das, in: Proceedings of the Solid Freeform Fabrication Sympo-
sium, University of Texas, Austin, 2001, pp. 85–109.
[32] D. King, T. Tansey, J. Mater. Process. Technol. 132 (2003) 42.
[33] Standard Test Methods for Metal Powder and Powder Metallurgy
Products, Metal Powder Industries Federation, Princeton, NJ, 1986.
[34] L. Pawlowski, J. Thermal Spray Technol. 8 (1999) 279.
[35] D.I. Pantelis, G. Pantazopoulos, in: N.B. Dahorte (Ed.), Lasers in
Surface Engineering, ASM International, Materials Park, OH, 1998,
pp. 357–394.
[36] Y. Arata, Plasma, Electron, and Laser Beam Technology: Develop-
ment and Use in Materials Science, ASM, Metals Park, OH, 1986.
[37] N. Tolochko, S.E. Mozzharrov, N.V. Sobolenko, Yu.V. Khlopkov,
I.A. Yadroitsev, V.B. Mikhailov, J. Adv. Mater. 2 (2) (1995) 151.
[38] Y. Kizaki, H. Azuma, S. Yamazaki, H. Sugimoto, S. Takagi, Jpn. J.
Appl. Phys. 32 (1A) (1993) 205.
[39] C. Gierl, Vorsintern von PM-Eisenbasiswerkstoffen, M.Sc. Thesis,
1996, Vienna University of Technology, Vienna, Austria.
[40] A. Simchi, H. Danninger, C. Gierl, Powder Metall. 44 (2) (2001)
148.
[41] B. Bergquist, Powder Metall. 42 (4) (1999) 331.
[42] X.C. Li, J. Stampfl, F.B. Prinz, Mater. Sci. Eng. A282 (2000) 86.
[43] A.H. Nickel, D.M. Barnett, G. Link, F.B. Prinz, in: Proc. Solid
Freeform Fabrication Symposium, University of Texas, Austin,
1999.