International Materials Reviews
ISSN: 0950-6608 (Print) 1743-2804 (Online) Journal homepage: https://www.tandfonline.com/loi/yimr20
Laser additive manufacturing of metallic
components: materials, processes and
mechanisms
D D Gu, W Meiners, K Wissenbach & R Poprawe
To cite this article: D D Gu, W Meiners, K Wissenbach & R Poprawe (2012) Laser additive
manufacturing of metallic components: materials, processes and mechanisms, International
Materials Reviews, 57:3, 133-164, DOI: 10.1179/1743280411Y.0000000014
To link to this article: https://doi.org/10.1179/1743280411Y.0000000014
Published online: 12 Nov 2013.
Submit your article to this journal
Article views: 22141
Citing articles: 782 View citing articles
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=yimr20
Laser additive manufacturing of metallic
components: materials, processes and
mechanisms
D. D. Gu*1,2, W. Meiners2, K. Wissenbach2 and R. Poprawe2
Unlike conventional materials removal methods, additive manufacturing (AM) is based on a novel
materials incremental manufacturing philosophy. Additive manufacturing implies layer by layer
shaping and consolidation of powder feedstock to arbitrary configurations, normally using a
computer controlled laser. The current development focus of AM is to produce complex shaped
functional metallic components, including metals, alloys and metal matrix composites (MMCs), to
meet demanding requirements from aerospace, defence, automotive and biomedical industries.
Laser sintering (LS), laser melting (LM) and laser metal deposition (LMD) are presently regarded
as the three most versatile AM processes. Laser based AM processes generally have a complex
non-equilibrium physical and chemical metallurgical nature, which is material and process
dependent. The influence of material characteristics and processing conditions on metallurgical
mechanisms and resultant microstructural and mechanical properties of AM processed
components needs to be clarified. The present review initially defines LS/LM/LMD processes
and operative consolidation mechanisms for metallic components. Powder materials used for AM,
in the categories of pure metal powder, prealloyed powder and multicomponent metals/alloys/
MMCs powder, and associated densification mechanisms during AM are addressed. An in depth
review is then presented of material and process aspects of AM, including physical aspects of
materials for AM and microstructural and mechanical properties of AM processed components.
The overall objective is to establish a relationship between material, process, and metallurgical
mechanism for laser based AM of metallic components.
Keywords: Additive manufacturing, Rapid prototyping, Rapid manufacturing, Direct metal laser sintering, Selective laser melting, Direct metal deposition,
Laser engineered net shaping, Metals, Alloys, Metal matrix composites, Microstructure, Mechanical property, Review
Introduction
Since the first technique for additive manufacturing
(AM) became available in the late 1980s and was used to
fabricate models and prototypes,1–3 AM technology has
experienced more than 20 years of development and
is presently one of the rapidly developing advanced
manufacturing techniques in the world.4 Different to the
material removal method in conventional machining processes, AM is based on a completely contrary
discipline, i.e. material incremental manufacturing
(MIM).5 Additive manufacturing implies layer by layer
shaping and consolidation of feedstock (typically
powder materials) to arbitrary configurations, normally
using a computer controlled laser as the energy resource.
1
College of Materials Science and Technology, Nanjing University of
Aeronautics and Astronautics, Yudao Street 29, 210016 Nanjing, China
Fraunhofer Institute for Laser Technology ILT/Chair for Laser Technology
LLT, RWTH Aachen, Steinbachstraße 15, Aachen D-52074, Germany
2
*Corresponding author, email dongdonggu@nuaa.edu.cn
ß 2012 Institute of Materials, Minerals and Mining and ASM International
Published by Maney for the Institute and ASM International
DOI 10.1179/1743280411Y.0000000014
First, the computer aided design (CAD) model of the
object to be produced is mathematically sliced into thin
layers. The object is then created by selective consolidation of the deposited material layers with a scanning
laser beam. Each shaped layer represents a cross-section
of the sliced CAD model. Therefore, AM is, also called
solid freeform fabrication, digital manufacturing, or
e-manufacturing.6 Additive manufacturing technology,
which involves a comprehensive integration of materials
science, mechanical engineering, and laser technology, is
regarded as an important revolution in manufacturing
industry.7
Rapid prototyping (RP) and rapid manufacturing
(RM) are two widely recognised synonyms of AM
technology.4 In the historical subsequence, a series of
processes for RP were primarily established. Then,
considerable research efforts proved that some of these
processes could also be used for manufacturing,
especially for small runs. Thus, ‘rapid prototyping’ was
combined with ‘manufacturing’ to give ‘rapid manufacturing’. As compared to the phrases RP and RM, AM is
International Materials Reviews
2012
VOL
57
NO
3
133
Gu et al.
Laser additive manufacturing of metallic components
regarded as a more general designation that directly
reflects the processing strategy of this advanced manufacturing technology.
Based on the similar processing philosophy, the
established AM techniques are versatile. The initially
developed AM techniques include stereolithography
apparatus,8 laminated object manufacturing,9 fused
deposition modelling,10 three-dimensional printing11
and selective laser sintering.12–14 These AM processes
are typically applied for the fabrication of prototypes
made from low melting point polymers as communication or inspection tools. The capability of producing
physical objects in a short period directly from CAD
models helps to shorten the production development
steps. Nevertheless, the production of conceptual prototypes made from polymers has no longer been the
current research focus of AM, because it enters a mature
development stage. The next natural development of
AM techniques is to produce complex shaped functional
metallic components, including metals, alloys and metal
matrix composites (MMCs) that cannot be easily
produced by the conventional methods, in order to
meet the demanding requirements from aerospace,15,16
automotive,17,18 rapid tooling19–21 and biomedical22,23
industrial sectors. Actually, components produced by
AM are no longer used merely as visualisation tools, but
to be used as real production parts (i.e., end-use
products) which have basic mechanical properties meeting the industrial requirements. To satisfy the demands
for AM fabrication of cost effective and end-use metallic
components, three typical processes in terms of laser
sintering (LS), laser melting (LM) and laser metal
deposition (LMD) have been developed. Different
institutions and companies use different phrases to
denominate these three most prevailing variants of
AM technology, as revealed in Table 1.
Being capable of processing a wide range of metals,
alloys, ceramics and MMCs, LS/LM/LMD are presently regarded as the most versatile AM processes.
Nevertheless, laser based AM techniques generally
involve a complex non-equilibrium physical and chemical metallurgical process, which exhibits multiple modes
of heat and mass transfer,37–40 and in some instances,
chemical reactions.41,42 The microstructural features
(grain size, texture, etc.) and resultant mechanical
properties (strength, hardness, residual stress, etc.) are
normally difficult to be tailored for a specific material
processed with AM technology. A large amount of
existent literature reveals that the complex metallurgical
phenomena during AM processing are strongly material
and process dependent and governed by both powder
characteristics (e.g. chemical constituents, particle
shape, particle size and its distribution, loose packing
density, and powder flowability) and processing parameters (e.g. laser type, spot size, laser power, scan speed,
scan line spacing and powder layer thickness).41–44 In
this respect, significant emphasis should be paid on both
design strategy of powder materials and control
methods of laser process, in order to achieve the feasible
metallurgical mechanism for powder consolidation in
LS/LM/LMD processes and resultant favourable microstructural and mechanical properties. Therefore, a
comprehensive review on the materials design, process
control, property characterisation and metallurgical
Table 1 Different categories and phrases of additive manufacturing processes
General phrase
Two widely recognised
synonymous phrases
Three typical
processes*
Synonyms from different
institutions/companies
Additive
manufacturing
Rapid prototyping
and rapid manufacturing
Laser sintering
Selective laser sintering; The
University of Texas at Austin, USA
Direct metal laser sintering; EOS company;
for EOSINT M 250 machine
equipped with CO2 laser
The same direct metal laser
sintering phrase but different processing
mechanism; EOS company; for EOSINT M
270/280 machine equipped with fibre laser
Selective laser melting; widely used in Europe
Direct metal laser remelting; The University of
Liverpool, UK; presently merged
into selective laser melting
Lasercusing; Sauer product GmbH, Germany
Direct metal deposition; The University
of Michigan, USA
Laser engineered net shaping (LENS);
widely used in USA; LENS is a trademark
of Sandia National Laboratory and the
United States Department of Energy, USA
Directed light fabrication; Los Alamos
National Laboratory, USA
Direct laser deposition; The
University of Manchester, UK
Direct laser fabrication; The
University of Birmingham, UK
Laser rapid forming; Northwestern
Polytechnical University and The Hong
Kong Polytechnic University, China
Laser melting deposition; Beihang University, China
Laser melting
Laser metal
deposition
Ref.
16
24
25
26
27
28
17, 29
30, 31
32
33
34
35
36
*In this review, we use the basic phrases (i.e. laser sintering, laser melting and laser metal deposition) to denominate the three most
prevailing variants of additive manufacturing technology for fabrication of metallic components.
134
International Materials Reviews
2012
VOL
57
NO
3
Gu et al.
Laser additive manufacturing of metallic components
1 Classification of AM processes based on different mechanisms of laser–material interaction: * 5 partial melting
mechanism is occasionally applied for LMD to create porous components with the residual porosity required.45,46
theories for LS/LM/LMD of a wide variety of metallic
powders is particularly necessary.
The basic intent of this article is to review the current
status of research and development in AM of enduse metallic components, including metals, alloys
and MMCs, with particular emphasis on strategies of
powder materials design and laser process control. The
classification of currently prevailing AM processes for
metallic components and the operative consolidation
mechanisms are given in the section on ‘Classification of
AM processes and metallurgical mechanisms’. The
section on ‘Classes of materials for AM and processing
mechanisms’ classifies the ever reported metallic materials used for AM, both commercially available and
experimentally developed powders, and the associated
bonding and densification mechanisms during laser
processing. The section on ‘Material/process considerations and control methods’ presents an in depth review
of the materials aspects of AM processes, including
physical aspects of materials for AM, microstructural/
mechanical properties of AM processed parts and
structure/property stability of AM fabricated parts.
The dependence of these microstructural/mechanical
properties on material/process parameters will be
elucidated. This review, therefore, seeks to establish
the relationship between material, process and metallurgical mechanism of various AM processes.
Classification of AM processes and
metallurgical mechanisms
Although AM processes share the same MIM philosophy, each AM process has its specific characteristics in
terms of usable materials, processing procedures and
applicable situations. The capability of obtaining high
performance metallic components with controllable
microstructural and mechanical properties also shows
a distinct difference for various AM processes. As
revealed in Fig. 1, according to the different mechanisms
of laser–powder interaction (i.e. prespreading of powder
in powder bed before laser scanning versus coaxial
feeding of powder by nozzle with synchronous laser
scanning) and the various metallurgical mechanisms (i.e.
partial melting versus complete melting), the prevailing
AM technology for the fabrication of metallic components typically has three basic processes: LS, LM and
LMD. Their deposition mode, deposition rate, processing conditions and attendant microstructural/mechanical properties are summarised in Table 2 and will be
addressed in detail as follows.
Laser sintering
Laser sintering is a typical AM process based on the
layer by layer powder spreading and subsequent laser
sintering. As schematically shown in Fig. 2, the LS
system normally consists of a laser, an automatic
powder layering apparatus, a computer system for
process control and some accessorial mechanisms (e.g.
inert gas protection system and powder bed preheating
system). Different types of lasers are used, including
CO2,54 Nd : YAG,55 fibre lasers,56 disc lasers,57 etc. The
choice of laser has a significant influence on the
consolidation of powders, mainly because:
(i) the laser absorptivity of materials greatly depends
on the laser wavelength
(ii) the operative metallurgical mechanism for powder densification is determined by the input laser
energy density.
The general processing procedures of LS include:
(i) a substrate for part fabrication is fixed on the
building platform and levelled
(ii) the protective inert gas is fed into the sealed
building chamber to reduce the interior oxygen
content below a required standard
International Materials Reviews
2012
VOL
57
NO
3
135
Gu et al.
Laser additive manufacturing of metallic components
(iii) a thin layer of the loose powder with a thickness
normally below 100 mm is deposited on the
substrate by the layering mechanism
(iv) the laser beam scans the powder bed surface to
form layer wise profiles according to CAD data
of the components to be built
(v) the above procedures including powder spreading and laser treatment are repeated and the
components are built in a layer by layer manner
until completion.
During LS, the duration of laser beam on any powder
particle depends on beam size and scan speed and is
typically between 0?5 and 25 ms.58 Under this extremely
short thermal cycle, the processing mechanism must be
rapid and, thus, solid state sintering mechanism is not
feasible. Melting/solidification approach is the only
mechanism suitable for the rapid consolidation of powder
during LS.59,60 As is implied in its name, LS is processed
based on a liquid phase sintering (LPS) mechanism
involving a partial melting of the powder (i.e. semisolid
consolidation mechanism). So far, LS has demonstrated
the feasibility in processing multicomponent metal
powder and prealloyed powder.61,62 Powder characteristics and laser processing conditions are required to be
carefully determined in order to realise the favourable
metallurgical mechanism for powder consolidation.
The multicomponent powder mixture is generally
composed of the high melting point metallic component,
acting as the structural metal, the low melting point
metallic component, taking as the binder, and a
small amount of additives such as fluxing agent or
deoxidiser.63,64 The operative LS temperature is carefully
determined between these two different melting temperatures by adjusting laser processing parameters. The
binder, thus, melts completely to form liquid phase, while
the structural metal remains its solid cores in the liquid.
Densification of the solid/liquid system occurs as a result
of the rearrangement of solid particles under the influence
of capillary forces exerted on them by the wetting liquid.
The liquid/solid wetting characteristics and the capillary
force exerted on particles determine the particle rearrangement rate and resultant success of LS. Laser melting of
a multicomponent Cu based powder consisting of pure
Cu powder and prealloyed SCuP powder has been
performed by Zhu et al.47,65 The SCuP with lower
melting point (645uC) acts as the binder, while the Cu
2 Schematic of LS apparatus53
with higher melting point (1083uC) acts as the structural
metal (Fig. 3a). Gu et al.54 have applied LS to process Ni–
CuSn–CuP system consisting of high melting point Ni as
the structural metal. The LS processed material is
composed of unmelted Ni solids (Fig. 3b), revealing a
semisolid LPS mechanism involved in LS process.
In contrast to pure metals with congruent melting
point, prealloyed powder exhibits a mushy zone between
solidus and liquidus temperatures, within which liquid
and solid phases coexist during melting/solidification
process (Fig. 4a). As laser processing parameters are
optimised, the preferable LS temperature is in the mushy
zone to produce a semisolid system. This process,
termed supersolidus liquid phase sintering (SLPS), acts
as the feasible metallurgical mechanism for LS of
prealloyed powders.66 As illustrated in Fig. 4b, prealloyed particles melt incongruently and become mushy
once a sufficient amount of liquid is formed along grain
boundaries. The liquid flows and wets solid particles and
grain boundaries, leading to a rapid densification of
semisolid system by means of rearrangement of solid
particles and solution reprecipitation process. Niu
et al.68 have demonstrated that SLPS mechanism is
operative during LS of high speed steel powder. The
Table 2 Comparisons of some representative AM processes*
Process
Deposition
mode
Layer thickness/
mm
DMLS
Laser sintering
20–100
SLM
Laser melting
20–100
DMD
LENS
Laser cladding
Laser cladding
254
130–380
Depend on laser
spot size, scan speed
and size, number, and
complexity of parts
Depend on laser spot
size, scan speed and
size, number, and
complexity of parts
0.1–4.1 cm3 min21
…
DLF
Laser cladding
200
10 g min21 (1 cm3 min21)
Deposition rate
Dimensional
accuracy/mm
Surface roughness/mm
Ref.
High, ¡0.05
14–16
24, 47
High, ¡0.04
9–10
48, 49
…
x–y plane ¡0.05;
z axis ¡0.38
¡0.13
y40
61–91
17, 29
30, 50
y20
51, 52
*AM, additive manufacturing; DMLS, direct metal laser sintering; SLM, selective laser melting; DMD, direct metal deposition; LENS,
laser engineered net shaping; DLF, directed light fabrication.
136
International Materials Reviews
2012
VOL
57
NO
3
Gu et al.
Laser additive manufacturing of metallic components
3 Microstructures of LS processed a Cu–SCuP (Ref. 47) and b Ni–CuSn–CuP (Ref. 54) multicomponent powder
thick ring microstructure reprecipitated around the
austenitic grain boundaries indicates the formation of
liquid phase along grain boundaries within particles
during SLPS (Fig. 4c).
It should be noted that LS of prealloyed powders
through SLPS mechanism requires a strict control of
laser processing parameters to realise the incongruent
melting of particles within the mushy zone. However,
due to the localised, rapid nature of thermal cycle during
LS, there exists a significant difficulty in controlling the
sintering temperature between solidus and liquidus,
which in turn handicaps the successful operation of
SLPS mechanism. Processing problems (e.g. insufficient
densification, heterogeneous microstructures and properties, etc.) tend to occur in LS processed prealloyed
powders. Therefore, post-processing treatment such as
4 a portion of idealised temperature–composition equilibrium phase diagram for prealloyed binary metal system, b schematic of SLPS densification of prealloyed
particles67 and c microstructural development during
LS of high speed steel powder68
furnace post-sintering,69 hot isostatic pressing (HIP),70
or secondary infiltration with a low melting point
material71 is normally necessary to obtain sufficient
mechanical properties.
Laser melting
Driven by the demand to produce fully dense components with mechanical properties comparable to those
of bulk materials and by the desire to avoid time
consuming post-processing cycles, LM has been developed. Laser melting shares the same processing apparatus and procedures with LS. The only difference is that
LM of metallic powders is based on a complete melting/
solidification mechanism. The idea of full melting is
supported by the continuously improved laser processing conditions in recent years (e.g. higher laser power,
smaller focused spot size, smaller layer thickness, etc.),
leading to significantly improved microstructural and
mechanical properties as relative to those of early time
LS processed components.72 Accordingly, LM shows
better suitability to produce full dense parts approaching
99?9% density in a direct way, without post-infiltration,
sintering or HIP.73 Simchi74 and Niu et al.75 have
processed M2 high speed steel powder using LM and LS
methods, respectively. The densification rate, surface
smoothness and microstructural homogeneity of LM
processed material under optimal processing conditions
show a significant improvement upon those of LS
processed material (Fig. 5).
Another major advance of LM lies in its high
feasibility in processing nonferrous pure metals, e.g.
Ti,76 Al,77 Cu,78 etc., which to date cannot be well
processed using LS partial melting mechanism. Early
attempts to process pure metals using LS are proved to
be unsuccessful, due to the considerably high viscosity
and resultant balling phenomenon caused by the limited
liquid formation.79,80 In contrast, the density of LM
processed pure metals is highly controllable and can be
improved significantly up to 99?5% through the full
melting mechanism of LM.77,78
Nevertheless, LM requires a higher energy level,
which is normally realised by applying good beam
quality, high laser power and thin powder layer
thickness (i.e. long building time). Consequently, LM
suffers from or is at a significant risk for the instability
of molten pool due to the full melting mechanism used.
A large degree of shrinkage tends to occur during
liquid–solid transformation, accumulating considerable
stresses in LM processed parts.81,82 The residual stresses
arising during cooling are regarded as key factors
International Materials Reviews
2012
VOL
57
NO
3
137
Gu et al.
Laser additive manufacturing of metallic components
5 Surface morphologies of M2 high speed steel components processed by a LM74 and b LS75
responsible for the distortion and even delamination of
the final products. Pogson et al.’s work83 on LM of Cu–
75%H13 reveals that the incorporation of Cu into tool
steel during LM produces the over heating Cu rich
region around the austenite grain boundaries, which
increases the risk of cracking by hot tearing (Fig. 6).
Also, the melt instabilities may result in spheroidisation
of the liquid melt pool (known as balling effect) and
attendant interior porosity. Therefore, proper care
should be paid in the reasonable selection of both laser
processing and powder depositing parameters to determine a suitable process window, in order to yield a
moderate temperature field to avoid the overheating of
LM system.
It is noted that the period for rapid development of
LM technology is from the year 2000. In contrast, the
intensive research attempts on LMD technology has
started from the year 1993 – the production of metallic
parts with favourable mechanical properties by LMD
has been reported in the nineties. For instance,
Mazumder et al. have reported direct metal deposition
(DMD) fabrication of fully dense aluminium 1100 parts
as early as 1993, demonstrating to provide metal
properties equivalent to a wrought process.17,29,84 In
contrast, LM production of complex shaped aluminium
components meeting industrial standards has been
successfully performed at the Fraunhofer ILT in 2008.77
Laser metal deposition
Process overview
Although the processing strategy of LMD follows the
general MIM principle, the manner of powder supply
changes from the prespreading in LS/LM processes to
the coaxial feeding in LMD process (Fig. 1). The LMD
powder delivery system consists of the specially designed
powder feeder that delivers powder into a gas delivery
system via the nozzles. The high energy laser beam is
delivered along the z axis in the centre of the nozzle
array and focused by a lens in a close proximity to the
workpiece. Moving the lens and powder nozzles in the z
direction controls the height of the focuses of both laser
and powder. The workpiece is moved in the x–y
direction by a computer controlled drive system under
the beam/powder interaction zone to form the desired
cross-sectional geometry. Consecutive layers are additively deposited, producing a three-dimensional component. With the integration of multi-axis deposition
system, multiple material delivery capability, and, in
some instances, the patented closed loop control
system,3,85,86 Laser metal deposition can coat, build,
138
International Materials Reviews
2012
VOL
57
NO
3
and rebuild components having complex geometries,
sound material integrity and dimensional accuracy.
Accordingly, LMD has a highly versatile process
capability and can be applied to manufacture new
components, to repair and rebuild worn or damaged
components and to prepare wear and corrosion resistant
coatings.87
The DMD, LENS and Directed light fabrication
(Table 1) are regarded as three representative processes
of LMD technology. It is worth noting that the DMD
technology developed by Mazumder’s group at the
University of Michigan is equipped with a feedback
system that provides a closed loop control of dimensional accuracy during deposition process. The feedback
loop is, thus, regarded as a unique feature of DMD that
differentiates from LENS and Directed light fabrication
processes.88
Constitutes of DMD system
A typical DMD system is schematically depicted in
Fig. 7 and some of the main features are as follows.88,89
Patented closed loop feedback control for DMD process
This unique system serves as the key tool for producing
a near-net shape product.85,86 High speed sensors collect
melt pool information, which is directly fed into a
dedicated controller that adjusts the input processing
parameters to maintain dimensional accuracy and
material integrity.
Coaxial nozzle with local shielding of melt pool
The coaxial nozzle design is based on a patent90 and
offers equal deposition rates in any direction. Inert gas
blown through the nozzle helps both in powder delivery
and shielding the deposit from oxidation. Shielding
strategy is a delicate balance between the adequate
pressure to drive away the ambient air and the powder
delivery without causing excessive disturbance within
the molten pool.
Six-axis computer aided manufacturing (CAM) software
for AM
Six-axis DMD CAM software for AM, which includes
an integrated DMD database with process recipes as a
part of the software, builds a CAM tool path directly
from CAD data. Contour, surface and volume deposition paths are provided in three dimensions, and,
accordingly, multilayer deposition paths can be prepared in a single operation. Simulation and collision
detection modules are included and, thus, enable the
user to detect any possible collision of the processing
Gu et al.
Laser additive manufacturing of metallic components
6 Distortion and crack formation in LM processed Cu–
H13 powder83
head and the part while creating the deposition tool
path.
Directed light fabrication vision system
The DMD vision system has been developed for
deposition on small objects with fine features. The
system locates the coordinate position of a part in the
machine and allows easy tool path generation for
accurate deposition. This eliminates manual part pickup, which is practically impossible for very small
components with fine structures. Faster operation and
better repeatability improves productivity considerably.
Unique applications of LMD/DMD technology
Similar as powder bed based LS/LM processes, LMD/
DMD technology has been applied successfully in direct
building near-net shape three-dimensional components,
covering a broad range of industries.87 Besides the nearnet shape part manufacturing capability, LMD/DMD,
as an enabling technology that allows the right material
to be added to the right place,88 has some unique
capabilities/features that are absent in LS/LM processes.
Repair and remanufacturing
Repairing of worn components is typically cost saving
versus purchasing new parts. Also, when a worn part is
rebuilt, the potential exists to repair that component in
such a manner that it will have a longer wear life than a
new part. The use of LMD/DMD technology opens new
technical opportunities for repairing components previously considered non-repairable by conventional
methods.91 The application areas best suited for LMD/
DMD are turbine blades/vanes repairs.87 The concentrated heat of the laser, typically for Nd : YAG and fibre
laser beams, allows blade tip build-up with minimum
distortion. The vision system and closed loop feedback
system offer precision part pick-up and restoration,
leading to a quality product that requires minimal post
grinding. Another feasible application of LMD/DMD is
the repair of drive shafts.91 Bearing, seal, and coupler
surfaces on shafts, which are typically considered nonrepairable by conventional welding techniques, act as
the great candidates for build-up and repair utilising
LMD/DMD. Furthermore, the LMD/DMD deposits
are metallurgically bonded to the substrate, not
mechanically bonded like spray or chroming processes.91
Cladding and hardfacing
Cladding and hardfacing are actually a form of repair
build-up applied to deposit new layer(s) of material on a
substrate. Multiple layers can be deposited to form
7 Schematic of closed loop DMD system88
shapes with complex geometry. These two variants of
LMD/DMD have been used for material surface
property modification and for the repair and manufacturing of multilayer coatings.92 Cladding and hardfacing
using CO2 lasers have proved to be highly successful.91
Combining the flexible LMD/DMD system with the new
fibre lasers improves on this success. POM Group Inc.
has developed large DMD workstations (DMD 105D)
for hardfacing and repair/cladding of large dies, moulds
and components.93 The fibre laser having the shorter
wavelength can achieve equivalent deposition rates with
y50% of the wattage required by a CO2 laser.91 The
favourable result is similar production rates with less
stress conveyed into the part being cladded. The surface
finish of the cladding may be left as deposited or ground
to finish dimension.
Designed material
One of the unique characteristics of closed loop DMD
technology is that multiple materials can be deposited at
different parts of a single component with high
precision. This capability can be utilised to develop a
new class of optimally designed materials, i.e. a class of
artificial materials with properties and functions that do
not exist in natural environments. In other words, a
material system can be designed and fabricated for a
chosen performance.
Mazumder’s group has developed a new methodology
for design, representation and fabrication of the
performance based ‘designed material’ using multiple
material deposition by DMD. The methodology
involves the computer integration of three key technologies, i.e. homogenisation design method (HDM),
heterogeneous solid modelling and DMD.94 The HDM
is applied to determine the optimal shape and topology
of a macroscale structural component and, subsequently, the HDM output is converted into a CAD
model using geometric modelling techniques. This
enhanced HDM can be used for material design to
control Young’s moduli, shear moduli, Poisson’s ratios
and even thermal expansion coefficients.29 An object
with material attributes as heterogeneous object and the
corresponding solid model are referred to as heterogeneous solid modelling. Heterogeneous objects are
International Materials Reviews
2012
VOL
57
NO
3
139
Gu et al.
Laser additive manufacturing of metallic components
8 Microstructures of LMD processed Ti–xMo graded alloy with progressively increasing Mo contents102
mainly classified into multimaterial objects, which have
distinct material domains, and functionally graded
materials (FGMs), which are a new class of composites
that possess continuous material variation along with
the geometry.89
The development of FGMs by LMD/DMD is
regarded as a basic strategy for ‘designed material’ by
tailoring the compositions and microstructures during
deposition.95,96 Since LMD uses the coaxially supplied
powder feedstock, it has the ability to produce FGMs by
selectively depositing different elemental powders into
the molten pool at specific locations in the structure
during part buildup.97–101 The adaptation of multiple
powder feeders in a LMD/DMD system makes it
possible. Dissimilar powder materials can be placed
into separate powder hoppers. Computer control
system, which is integrated into the powder feed system,
enables the user to vary the deposit composition as a
function of position. Shin et al.89 have introduced an
integrated design and fabrication system for heterogeneous objects, especially FGMs. A variant design
paradigm and a constructive representation scheme for
FGMs are primarily described. A discretisation based
process planning method, which converts continuous
material variation into stepwise variation, is then
proposed. The DMD process, which can take advantage
of the proposed process planning method, is applied to
prepare rectangular and circular graded parts of Cu–
xNi, in order to reveal how the material compositions
change during deposition and, accordingly, to verify the
proposed design–fabrication cycle of FGMs. Collins
et al.102 have deposited the compositionally graded
binary Ti–xMo alloys, from elemental Ti to Ti–25
at-%Mo, within a 25 mm length part using LMD. The
microstructures across the graded alloy correspond to
those typically observed in a/b-Ti alloys, but the
microstructural scale is significantly refined. Interesting
microstructure gradients are tailored across the alloy
140
International Materials Reviews
2012
VOL
57
NO
3
(Fig. 8). The ability to achieve such substantial changes
in composition/microstructure across rather limited
length makes LMD a highly attractive candidate for
developing novel structured FGM components with
unique properties. It is widely accepted that the ability
to produce near-net shape components with graded
compositions from elemental powders using LMD may
potentially be a feasible route for manufacturing
unitised structures for high demanding aerospace
applications.102
More important, the methodology for ‘designed
material’ has been extended from the design of
compositions/microstructures of materials to the creation of microscopic structures with particular behaviours. These microscopic structures are effectively
artificially designed materials and their behaviours are
essentially artificial properties. Many of these properties
are technologically interesting (e.g. extraordinary piezoelectricity), physically unusual (e.g. negative Poisson’s
ratio) or unavailable in nature (e.g. ductile metals with
negative thermal expansion).94 The designed materials
are regarded as a revolutionary departure from the
present material selection methods. One creative demonstration is firstly disclosed in Mazumder et al.’s research
work on the homogenisation DMD process using a
combination of Ni and Cr. Figure 9 shows a structure
designed by HDM and fabricated by DMD, which
exhibits negative thermal expansion dL/L<–0?00065 at
150uC and maintains such a unique property up to
300uC.29,94,103
Metallurgical mechanisms of LMD/DMD process
Molten pool behaviour
During LMD/DMD, the laser beam creates a mobile
molten pool on the substrate into which powder is
injected. A continuous, stable and precise feeding of
powders into the molten pool is, thus, of primary
importance. Then, the molten pool size has been
Gu et al.
Laser additive manufacturing of metallic components
9 a design and b realisation of negative coefficient of thermal expansion using DMD (green, light colour, Ni; blue, dark colour, Cr)103
identified as a critical parameter for maintaining optimal
building conditions.104–107 A photograph of a single line
LMD of 316 stainless steel by Hofmeister et al.108 shows
the presence of molten pool with a clear contour
(Fig. 10a). The formation of dimensionally steady
molten pool with a small heat affected zone and an
uninterrupted solidification front is preferable. Real
time thermal imaging of molten pool size and its
morphology (Fig. 10b) is used as a feedback mechanism
to determine temperature gradient and cooling rate and to
control LMD process. The effects of laser processing
parameters (e.g. laser power and scan speed) on the molten
pool features have been investigated both by modelling109–111
and experiments.112–114 For a constant scan speed, the
geometry of the molten pool depends on the input heat
distribution. The laser power is adjusted to make sure that
the pool size is in the predefined range. Cooling of the pool
is accomplished primarily by conduction of heat through
the part and substrate.113 Depending on the substrate
temperature and laser energy input, cooling rates at solid–
liquid interface are varied from 103 to 104 K s–1.109 This
flexibility allows the control of the final microstructures
and properties of LMD processed parts.
Thermal and kinetic history
Different to LS/LM, LMD involves the computer
controlled three-dimensional shaping of molten materials through a deposition head, using the powder injected
into a molten pool created by a focused high power laser
beam. Accordingly, LMD accommodates a wide range
of materials and deposition styles. The applicable
materials are primarily from the prealloyed powders of
the determined compositions. In particular, high melting
point alloys have demonstrated a unique applicability
for LMD,115 due to a precision, point by point complete
melting mechanism of LMD. Various parts have been
fabricated from nickel based alloys, titanium alloys,
steels, and other specialty materials (see the section on
‘For LM and LMD: alloys powder’).
Nevertheless, due to the layer by layer additive nature
of LMD, the complex thermal histories are experienced
repeatedly in different regions of the deposited material.
The thermal histories of LMD normally involve melting
and numerous reheating cycles at a relatively lower
temperature.116 Such complicated thermal behaviour
during LMD results in the complex phase transformations and microstructural developments.34,117 There,
consequently, exist significant difficulties in tailoring
compositions/microstructures required. On the other
hand, the use of a finely focused laser to form a rapidly
traversing molten pool may result in considerably high
solidification rate and melt instability. Complicated
residual stresses tend to be locked into the parts during
the building process, due to the thermal transients
encountered during solidification.118–120 The presence of
residual stresses causes deformation or, in the worst
instance, cracks formation in LMD processed components. The uncontrollability of compositions/microstructures and the formation of residual stresses are regarded
as two major difficulties associated with LMD.
The understanding of the origin of these defects aids
in improving controllability of either LMD process or
final microstructural/mechanical properties. Actually, a
series of complex physical phenomena including heat
transfer, phase changes, mass addition and fluid flow are
involved in the molten pool during LMD. Interactions
between the laser beam and the coaxial powder flow are
of a primary consideration, including the attenuation of
beam intensity and temperature rise of powder particles
before reaching the pool.39 The temperature and velocity
fields, liquid/gas interface, and energy distribution at
liquid/gas interface in the pool should be monitored, in
order to further control the melt pool width and length,
and the resultant height and width of solidified cladding
tracks.40 Therefore, the knowledge of temperature,
velocity and composition distribution history is essential
for an in depth understanding of the process and
subsequent microstructure evolution and properties.121
10 a photograph of single line LMD build and b side view of molten pool showing temperature in kelvin108
International Materials Reviews
2012
VOL
57
NO
3
141
Gu et al.
Laser additive manufacturing of metallic components
11 a heterogeneous microstructures of LS processed Ti122 and b partially melted particle surface of LMD processed porous Ti (Ref. 130)
Classes of materials for AM and
processing mechanisms
For LM and LMD: pure metals powder
Pure metals that have been applied for various AM
processes are listed in Table 3. As relative to alloys, pure
metals are not the focus of AM technology, mainly due
to the following two reasons. First, the relatively weak
nature of pure metals, e.g. limited mechanical properties
and poor anti-oxidisation/anticorrosion capabilities,
makes them less attractive as candidate materials for
AM. Second, the unsuccessful early attempts to process
pure metals through partial melting mechanism by LS
have lasted a long period without any significant
progress before a successful application of LM.62 For
instance, the LS processed Ti, due to a partial melting
mechanism applied, typically has a heterogeneous
microstructure and consists of three different regions:
(i) the cores of unmelted grains
(ii) the melted surface of grains
(iii) the residual pores (Fig. 11a).122
Currently, the move from LS to LM represents a major
advance in AM of nonferrous pure metal components in
industrial practice.128
It is worth noting that though LMD is normally
processed based on a complete melting mechanism to
yield a fully dense component (Fig. 1), recent research
efforts by Bandyopadhyay et al.45,46,129,130 on LMD of
pure Ti and Ta through a partial melting mechanism
(Table 3) have demonstrated a high potential to produce
Table 3 Pure metals components produced by various AM processes*
Metal Powder characteristics
Process Laser type
Bonding mechanism Mechanical properties
Ref.
122
Ti
Spherical shape;
LS
Gaussian particle size
distribution, mean size
8 mm, maximum size 30 mm
Pulsed Nd : YAG laser
Partial melting in
a narrow surface
layer of particles
Ti
Spherical shape;
average size 45 mm
LM
Pulsed Nd : YAG laser
Complete melting
of powder
Ti
Commercially pure;
particle size 50–150 mm
LMD
Nd : YAG laser, 500 W
Ta
99.5% purity;
particles size
45–75 mm
LMD
Nd : YAG laser, 500 W
Cu
…
LM
Q switched krypton flash
lamp pumped Nd :
YAG laser, 90 W
Partial melting of
powder surface
(avoid complete
melting of powder
to form desired
porous structure)
Partial melting of
powder surface
(avoid complete
melting of powder
to form desired
porous structure)
Complete melting
of powder
Au
24 carat gold;
mean particle
size 24 mm; tap
density 10.3 g cm–3
LM
Continuous wave
Complete melting
ytterbium fibre laser, 50 W of powder
72% theoretical
density; microhardness
250–340 HV;
compressive yield
strength 260 MPa
Tensile strength 300
MPa; torsional fatigue
strength 100 MPa;
microhardness 600–
1000 HV (after
laser gas nitriding)
Porosity 35–42 vol.-%;
Young’s modulus 2–45
GPa; 0.2% proof
strength 21–463 MPa
(similar to human
cortical bone)
Porosity 27–55 vol.-%;
Young’s modulus 1.5–20
GPa; 0.2% proof
strength 100–746 MPa
International Materials Reviews
2012
VOL
57
NO
3
125
45
126
Tentative experiments
on LM of Cu powder
layers to produce
simple three-dimensional
structures
Minimum internal
127
porosity 12.5%;
maximum microhardness
29 HV
*AM, additive manufacturing; LS, laser sintering; LM, laser melting; LMD, laser metal deposition.
142
123, 124
Gu et al.
Laser additive manufacturing of metallic components
12 a, c oriented martensite plates containing acicular hcp phase in LM processed Ti–6Al–4V and b, d a–b biphasic
microstructure developed in heat treated material161
complex shaped porous implants with functionally
graded porosity used for load bearing biomedical
applications. According to their design philosophy,
complete melting of the powder is avoided by using
low laser powers to partially melt the metal powder
surface (Fig. 11b). The surface melted powders join
together due to the presence of liquid metal at the
particle interfaces, leaving some interparticle residual
porosity. As against solid state sintering in the conventional powder metallurgy (PM) route of porous metals,
the inherent brittleness can be eliminated. Furthermore,
by changing scan speeds, the interaction time between
powder particles and laser beam can be varied, creating
different porous structures with various final porosities.
For LM and LMD: alloys powder
So far, a large amount of prealloyed powders have been
applied for various AM processes, as reviewed in
Table 4. A majority of research efforts have been
focused on Ti based, Ni based and Fe based alloys
powder, among which some material and process
combinations have entered a mature phase of practical
applications. Additive manufacturing of Al based alloys
might be the next research focus to face the big challenge
in laser processing of nonferrous alloys with high
reflectivity to laser energy. Almost all the existent work
on AM of prealloyed powders is based on a complete
melting mechanism using LM or LMD, due to a
relatively easy process controllability as compared to
SLPS mechanism associated with LS (Fig. 4). Therefore,
laser resource with high energy densities, e.g. high
powered CO2 laser, Nd : YAG laser and fibre laser, is
generally required to yield a favourable bonding
mechanism (Table 4). Once the processing parameters
are optimised to obtain fully dense parts (except for
porous materials if needed), attention is focused on
residual stresses and microstructures. The control of as
built microstructures is strongly influenced by the large
undercooling degree during rapid solidification of laser
generated molten pool.159 The following sections give an
overview of four representative alloys used for AM,
especially focusing on microstructural development and
its mechanism.
Ti based alloys
Ti based alloys processed by AM, typically Ti–6Al–
4V, are mainly used in the aeronautical34,131,160 and
medical128,133 fields, because of their unique chemical
and mechanical features along with well documented
biocompatibility. Recent study by Facchini et al.161 has
disclosed the change in mechanical properties with
microstructures of Ti–6Al–4V produced by LM.
Owing to the formation of unique hcp martensitic
microstructure (Fig. 12a and c), the tensile strength of
LM manufactured parts is higher than that of hot
worked parts, whereas the ductility is lower. A postprocessing heat treatment causes the transformation of
the metastable martensite into a biphasic a–b matrix
(Fig. 12b and d), resulting in an increase in ductility and
a reduction in strength. The stabilisation of microstructures contributes to the improvement of the ductility.
This study has evidenced how it is possible to obtain a
fully dense material and control the martensite transform in Ti–6Al–4V alloy through the variation of LM
conditions.
Ni based alloys
Ni based superalloys, e.g. Inconel 625, 718 and Rene 41,
88DT (Table 4), due to an improved balance of creep,
damage tolerance, tensile properties and corrosion/
International Materials Reviews
2012
VOL
57
NO
3
143
144
International Materials Reviews
Ti–6Al–4V
Ti based
2012
VOL
57
NO
3
Inconel 625
(Ni–22Cr–5Fe–3.5Nb–
9Mo–0.4Al–0.4Ti–0.1C)
Waspaloy (Ni–13.5Co–19.5Cr–
4.2Mo–2.0Fe–0.7Si–1.0Mn–
1.4Al–3.0Ti–0.5Cu)
Inconel 625 (64.61Ni–
21.25Cr–8.45Mo–
4.65Nb–1.06Fe)
Inconel 718 (Ni–19Cr–18Fe–
0.5Al–1Ti–3Mo–5Nb–0.042C)
Rene 88DT (Ni–
17Cr–14Co–4.2W–
4Mo–3.3Ti–2.2Al–
0.7Nb–0.04C–
0.03O–0.02N)
Ni based
Ni based
Ni based
Ni based
Ni based
Ti–4Al–1.5Mn
Ti–25V–
15Cr–2Al–0.2C
Ti based
Ti based
Ti–6Al–4V
Ti–6Al–4V
Ti based
Ti based
Compositions{
Alloy
LMD
LMD
Particle size 44–150 mm
DMD
LM
LM
LMD
LMD
LMD
LM
DMD
Process{
Gas atomised; spherical
shape; particle
size 44–150 mm
Gas atomised; powder
diameter 45–135 mm
Average particle
size 63 mm
Ar atomised; spherical
shape; particle
size 45–420 mm
Spherical shape; 95%
particle size ,20 mm
Gas atomised; oxygen
content 0.19 wt-%
Spherical shape;
particle size 25–45 mm
Spherical shape;
particle size 25–45 mm
Gas atomised; spherical
shape; particle
size 2100z325 mesh
Powder characteristics
Table 4 Alloys components produced by various AM processes*
Continuous wave
CO2 laser, 5 kW
Continuous wave
CO2 laser, 5 kW
CO2 laser, 6 kW
Nd : YAG pulsed
laser, 550 W
Diffusion cooled
slab CO2 laser,
5 kW
Continuous
wave fibre laser
CO2 laser, 1.75 kW
Free from defects like crack, bonding error
or porosity; as deposited microstructure
mostly consists of columnar dendrites;
very high hardness 254¡6 HV
Tensile strength 845 MPa (as deposited)
and 1240 MPa (heat treated); 0.2%
yield strength 590 MPa (as deposited)
and 1133 MPa (heat treated); elongation
11% and reduction in area 26% (as deposited)
Tensile strength 1400–1440 MPa; 0.2%
yield strength 1010–1030 MPa; elongation
16.5–17.5% and reduction in area 17.5–18%
(HIPzheat treated)
Ultimate tensile strength 1030¡50
MPa (horizontal) and 1070¡60 MPa
(vertical); 0.2% yield strength 800¡20
MPa (horizontal) and 720¡30 MPa
(vertical); Young’s modulus 204.24¡4.12
MPa (horizontal) and 140.66¡8.67 MPa
(vertical); elongation about 8–10% (both directions)
Maximum 99.7% density
Tensile strength 1163¡22 MPa,
yield strength 1105¡19 MPa,
ductility y4% (as deposited); tensile
strength 1045¡16 MPa, yield strength
959¡12 MPa, ductility y10.5¡1%
(950uC annealed)
Approximately 100% density; tensile
strength .1000 MPa; breaking elongation 12%
Tensile strength 1211¡31 MPa;
yield strength 1100¡12 MPa; breaking
elongation 13.0¡0.6% (annealed);
Young’s modulus 118.000¡2.300 MPa
Tensile strength y1100 MPa/20uC;
ductility 2–4%; fatigue properties 650
MPa/450uC, 300 MPa/550uC, 200 MPa/650uC
Impact toughness 599¡57 kJ m–2
(as deposited), 888¡33 kJ m–2 (955uC annealed)
CO2 laser, 6 kW
Ytterbium fibre
laser, 200 W
Nd : YAG laser
Mechanical properties
Laser type
141
140
139
138
136, 137
135
134
133
132
131
Ref.
Gu et al.
Laser additive manufacturing of metallic components
International Materials Reviews
2012
VOL
Stainless steel 316L (Fe,
0.08C, 2.00Mn, 0.045P,
0.03S, 0.75Si, 16–18Cr,
10–14Ni, 2–3Mo, 0.12Cu, 0.10N)
Fe–15Cr–2Mn–16B–
4C–2Mo–1Si–1W–1Zr (at-%)
Fe based
Al based
Fe based
Al–40Ti–10Si (at-%)
AISI 4340 high strength low alloy steel
(Fe–0.42C–2.63Ni–0.90Cr–
0.74Mn–0.45Mo–0.29Si)
Fe based
Fe based
High speed steel M2
(Fe–0.86C–0.33Si–
0.37Mn–1.25Cr–1.97V–
5.23Mo–6.32W)
Tool steel H13 (Fe–0.40C–
0.93Si–0.35Mn–
5.31Cr–0.30Mo–1.07V–0.016P–
0.005S–0.006O–0.048N)
Fe based
Fe based
Fe based
LS
LMD
LMD
Spherical shape;
particle size 53–173 mm
Gas atomised; spherical
shape; particle
size 10–110 mm
Mechanically alloyed partially
amorphous and
nanocrystalline powder
DMD
DMD
LS
Continuous wave CO2 laser, 1.5 kW
Continuous wave Nd : YAG laser
…
Fibre coupled diode laser, 1 kW
Microhardness 745.2 HV; specific
wear rate 4.0461027 mm3 N21m21
Maximum hardness 690 HK;
yield strength 1505 MPa;
ultimate strength
1820 MPa; failure strain 6%;
reduction in area 10%
Maximum porosity 4.13%;
microhardness 681–480 HV;
Microhardness decreases
and amount of tempered
martensite increases from
the upper to the lower layers.
Porosity 5.07 vol.-%; tension
modulus 193.47 GPa; yield stress
419.0 MPa; ultimate tensile strength
826.9 MPa; failure strain 28.95%
Microhardness y900 HV (9.52 GPa)
151
150
149
148
15, 17
147
Maximum density 88.2%;
microhardness 560–1020 HV0.05
Continuous wave CO2 laser
CO2 laser, 4.5 kW
146
Maximum density y84%
Nd : YAG laser, 90 W
LM
143, 144
Density .99.5%
145
142
Ref.
Tensile strength 855 MPa;
yield strength 682 MPa; elongation
30.3% and reduction in area 45.8%
(high temperature tensile tests at 800uC)
Mechanical properties
Successful fabrication
of 2062065 mm object with
140 mm thick inner compartment walls
Continuous
wave fibre laser
Q switched Nd :
YAG laser, 90 W
Continuous wave
CO2 laser, 8 kW
Laser type
LM
LM
LMD
Process{
Gas atomised; mostly
spherical shape; particle
size 2140/z325 mesh
Particle size –70 mesh
Gas atomised;
near spherical
shape; 80% particle
size ,22 mm
Gas atomised;
particle size ,45 mm
Spherical shape; 95%
particle size ,20 mm
Gas atomised; spherical
shape; particle size
1–56 mm, 80%,22 mm
Ar atomised
Rene 41 (Ni, 18.0–
20.0Cr, 10.0–12.0Co,
9.00–10.5Mo, 1.40–
1.80Al, 3.00–3.50Ti, 0.06–
0.12C, 0.003–0.010B,
Fe(5.00, Zr(0.07,
Si(0.50, Mn(0.50, P(
0.015, S(0.015)
Stainless steel 316L
(Fe–16.73Cr–13.19Ni–
0.017C–0.71Si–
2.69Mo–1.69Mn)
Stainless steel Inox 904L
(Fe, 23–28Ni, 19–23Cr,
4–5Mo, 1–2Cu, Mn(2, Si(1,
C(0.02, P(0.045, S(0.035)
Tool steel H13 (Fe–0.4C–1.0Si–
0.4Mn–0.03S–5.2Cr–
1.5Mo–1.0V–0.3Ni)
Ni based
Fe based
Powder characteristics
Compositions{
Alloy
Table 4 Continued
Gu et al.
Laser additive manufacturing of metallic components
57
NO
3
145
146
International Materials Reviews
2012
VOL
57
NO
3
61.78Co–
29.37Cr–6.52Mo–0.23C–
0.69Mn–0.68Si, Ni, Ti,
Fe, S, P, N, O trace
Co–10.10Ni–26.41Cr–
7.31W–0.81C–0.44Si
Hovadur K220 (Cu–
2.4Ni–0.4Cr–0.7Si)
Cu–30Ni alloy (Cu,
29.0–33.0 Ni, 0.4–1.0 Fe,
1.0 Mn, Zn(0.5, 0.45 C,
Pb(0.02, P(0.02, S(0.02)
Co based
Gas atomised; mostly
spherical shape; particle
size 2100/z325 mesh
…
Particle size 40–100 mm
Near spherical shape;
mean particle size 50 mm
Gas atomised; particle
size –100/z325 mesh
…
Powder characteristics
DMD
LM
LMD
LMD
LM
LM
Process{
CO2 laser, 5 kW
Continuous wave fibre laser, 1 kW
Continuous wave CO2 laser, 5 kW
Nd : YAG laser, 500 W
Ytterbium fibre laser
Continuous wave fibre laser
Laser type
Maximum porosity 1.47%; microhardness
115–130 HV; ultimate tensile strength
240.49 MPa; yield strength 317.16 MPa;
elongation 13.9%
156
Tensile strength 946.5 MPa; elongation
27%; microhardness 540 HV
y99.9% density
158
157
155
154
152, 153
Ref.
Fully dense; hardness 40 HRC,
equivalent to CoCrMo wrought material
y100% density; microhardness
150 HV0.025; tensile strength 355
MPa (horizontal) and 280 MPa (vertical);
0.2% yield strength 250 MPa
Maximum density 89.5%
Mechanical properties
*AM, additive manufacturing; DMD, direct metal deposition; LM, laser melting; LMD, laser metal deposition; LS, laser sintering; HIP, hot isostatic pressing.
{Unless indicated, the chemical compositions are in wt-%.
{Besides LS process, AM of materials in Table 4 is based on a complete melting mechanism.
Cu based
Cu based
Co based
6061 Al alloy
Al–10Si–Mg
(EOS GmbH, Germany)
Compositions{
Al based
Al based
Alloy
Table 4 Continued
Gu et al.
Laser additive manufacturing of metallic components
Gu et al.
Laser additive manufacturing of metallic components
13 a longitudinal microstructure of LMD processed Rene 41; b size difference of c9 precipitate in c cellular dendritic and
d interdendritic regions142
oxidation resistance, are normally developed for high
performance components in jet engines and gas
turbines.162,163 As precipitate hardened PM superalloys,
Rene alloys are strengthened by the precipitation of
ordered L12 intermetallic Ni3(Al,Ti) c9 phase. The total
amount of Al and Ti elements in Rene alloys is y6
wt-%.141 Inconel alloys are Nb modified Ni based
superalloys and their high temperature strength is
developed by solid solution strengthening or precipitation strengthening. In precipitation strengthening varieties, a fine dispersion of D022 ordered c0 or L12 ordered
c9 precipitates is expected.140 Wang et al.142 have
produced Rene 41 components using LMD and found
that ultra fine directionally solidified columnar grains
with a primary arm spacing of y35 mm are formed
along the deposited direction, due to the high thermal
gradient and solidification cooling rate (Fig. 13a). The c9
precipitate in interdendritic zones has a smaller size and
a more uniform morphology than that in dendritic cores
(Fig. 13b–d), due to larger supersaturation of elements
and longer growth time of c9 in dendrites than that
located in interdendritic spaces.142
However, there is a high cracking susceptivity during
LM/LMD of Ni based superalloys, because of a high
amount of alloying elements and c9/c0 forming elements.
Crack characterisations in LMD fabricated Rene 88DT
(Fig. 14a) and LM processed Waspaloy (Fig. 14b)
have been investigated by Huang et al.141 and Mumtaz
et al.138 respectively. For LMD, cracks mainly nucleate
and propagate in the overlap zone between two adjacent
deposited passes. The overlapping degree has a significant effect on the size and amount of cracks. Two
typical kinds of cracks, i.e. long cracks (3–10 mm) and
short cracks (100–300 mm), are formed with different
overlapping (Fig. 14a). The formation of short cracks is
mainly attributed to the boundary liquation cracking.164
It is difficult to eliminate all the short cracks merely by
adjusting LMD processing parameters.141 Post-processing steps, e.g. HIP, are required to realise a substantial
improvement of mechanical properties. Comparatively,
the formation of Waspaloy parts by means of LM can
be controlled by manipulating processing conditions. A
definition of a feasible process window allows for the
fabrication of near fully dense (99?7%) components by
LM.138
Fe based alloys
Though research reports on AM of Fe based alloys
(typically steels) are abundant (Table 4), it seems that
the progress is not very significant. Simply in the review
of densification, the obtained density of AM processed
steels generally cannot reach a full density. Therefore,
AM of steels is still in the stage of pursuing the
14 Cracks formation in a LMD processed Rene 88DT141
and b LM processed Waspaloy138
International Materials Reviews
2012
VOL
57
NO
3
147
Gu et al.
Laser additive manufacturing of metallic components
15 Laser melting processed Al–10Si–Mg a thin wall component and b valves152
fully dense components. Nevertheless, some reports on
DMD/LMD of steels have started to focus on further mechanical properties besides the densification
rate.15,148,149 The difficulty in AM of steels is primarily
ascribed to the special chemical properties of the main
elements in steels. Both the matrix element Fe and the
primary alloying element Cr are very active to oxygen. A
certain degree of oxidation, thus, cannot be avoided
under normal powder handling and AM conditions.165
Consequently, balling phenomena are more likely to
occur during laser processing, due to a contamination
layer of oxide being present on the surfaces of steel melt,
severely degrading AM densification and attendant
mechanical properties. On the other hand, the carbon
content of steels is a critical factor in determining AM
processability. Normally, AM processed tool steels and
high speed steels demonstrate a limited densification
response (Table 4), since the high carbon content has a
detrimental effect. Investigations by Wright et al.166
reveal that as the carbon content increases, so does the
thickness of the carbon layer segregated on the melt
surface. Such carbon layer has the same detrimental
influence as oxide layer, reducing wettability and causing
the melt to spheroidise rather than flow across the
underlying surface. Furthermore, the formation of
complex interfacial carbides at grain boundaries
increases the brittleness of AM processed high carbon
content steels.166 Childs et al.’s results167,168 indicate that
elevating the heat flow in the powder being treated
favours the dissolution of carbides and, accordingly,
homogenises the distribution of alloying elements.
Therefore, besides the optimisation of laser type and
parameters, a thin powder layer thickness less than
100 mm is recommended for LM, in order to realise a
sufficiently high volumetric energy density for both
powder consolidation and elemental homogeneity.169–171
Al based alloys
Except for the research work by Mazumder et al.,29
Louvis et al.154 and Buchbinder et al.,152 very little
research work has been reported on AM of Al based
alloys by LM or LMD. There are a number of
difficulties in a successful LM/LMD of Al based
powders. First, the high reflectivity (.91%)154 and high
thermal conductivity of Al significantly increase laser
power required for melting. Second, the high susceptivity of Al based alloys to oxidation acts as a main
obstacle to the effective melting. The adherent thin oxide
films on molten Al reduce wettability. Oxide also causes
problems when stirred into the molten pool, since the
entrapped oxide generates regions of weakness within
the part. Third, as to LM, it critically depends on being
able to spread a thin powder layer, which is difficult
148
International Materials Reviews
2012
VOL
57
NO
3
because Al powders are light with poor flowability.
Consequently, Al based powders are unsuitable for
many existing powder deposition mechanisms, even
though they are effective for other metallic powders of
the same particle shape and size distribution.154
Louvis et al.154 have studied the oxidation mechanisms in different positions of the molten pool during LM
of 6061 and Al–12Si alloys. The oxide film on the upper
surface of the pool evaporates under laser beam.
Marangoni forces that stir the pool are the most likely
mechanism by which these oxide films are disrupted,
allowing fusion to the underlying layer. However, the
oxides at the sides of the pool remain intact and, thus,
create regions of weakness and porosity, as the pool fails
to wet the surrounding material. Further research on
LM of Al based alloys should be primarily orientated
towards new methods of controlling oxidation process
and disrupting the formed oxide films.
Recently, the Fraunhofer ILT has successfully qualified LM for Al–10Si–Mg functional prototypes
(Fig. 15). The static and dynamic tests demonstrate that
the mechanical properties of LM processed Al–10Si–Mg
specimens obtain at least the mechanical properties of
serial produced die cast Al–10Si–Mg components
according to EN 1706 specifications. Furthermore, it is
found that preheating significantly increases dimensional and shape accuracy of LM processed Al–10Si–
Mg thin wall parts.152,153 These inspiring results are of
major importance to future industrial applications of
AM technology for Al based alloys.
For LS and LMD: multicomponent metals/alloys
powder mixture
Multicomponent metallic powders are initially designed
for LS, using different binder and structural particles.
As an early developed AM process for metallic
materials, LS is performed based on a partial melting
mechanism. The application of such a semisolid
mechanism lowers the requirements for high powered
lasers. Also, the formation of thermal stresses and
resultant deformation/cracks is expected to be alleviated,
due to the limited thermodynamics and shrinkage rate
of a semisolid LS system.172 As revealed in Table 5,
multicomponent metallic powder systems can be classified as three categories:
For LS: distinct binder and structural metal with significant
difference in melting points
In this category, the structural metals have a distinctly
higher melting point than the metallic binder, e.g.
Cu versus SCuP (645uC),47 and Cu versus CuSn
(840uC).181,187 Normally, the particle size of the binder
is smaller than that of the structural metal, in order to
International Materials Reviews
Fe–20Ni–15Cu–15Fe3P
Fe–20Ni–15Cu–15Fe3P
Fe–0.8C
(–2.5Cu, 1.0Si, 1.0Ti)
Fe–4B(–9Ti)
Cu–40SCuP
Cu–30CuSn–10CuP
Fe based
Fe based
Fe based
Fe based
Cu based
Cu based
Cu based
Cu–W
Fe–15Cu–15W
Fe based
Fe based
Electrolytic Cu, dendritic shape, mean
particle size 40 mm; prealloyed SCuP,
spherical shape, particle size 5–20 mm;
P acts as flux to protect Cu oxidisation
Irregular Cu, particle size
28–75 mm; ellipsoidal
CuSn 11–46 mm; spherical
CuP 5–24 mm; homogeneous
powder mixture by ball mixing
coarse and fine
powders with a broad
particle size distribution
Cu mean size 15 mm; W–20Cu
mean size 0.24 mm;
submicron/micron system increases
flowability of powder mixture
Fe 80%,22 mm, Fe–B 100%,
45 mm, Ti 100%,40 mm
Fe irregular shape, particle size
5–10 mm; Cu dendrite shape, mean
size 40 mm; W prismatic
shape, mean size 4.25 mm
Spherical Fe,50 mm, spherical
Ni 5 mm, spherical Cu,50 mm, spherical
Fe3P,50 mm; dissolution of P lowers
surface tension and
oxidation rate of melts
Spherical Fe,50 mm, spherical Ni 5 mm,
spherical Cu,50 mm,
spherical Fe3P,50 mm;
dissolution of P lowers surface
tension and oxidation rate of melts
Water atomised Fe powder (0.5% oxygen)
d50558 mm, Cu d50530 mm, Ti
d50,25 mm, Si d50,8 mm
Cu binder; Ni, Mo alloying elements
(y5 wt-%); C decreases surface
tension and viscosity of Fe base;
particle size 30–45 mm
Fe–(0.4, 0.8,
Water atomised/carbonyl Fe powder;
1.2, 1.6)C (graphite)
mean particle size 69.4 mm/
13.4 mm; fine graphite powder 2 mm
Fe–29Ni–8.3Cu–1.35P
Spherical Ni and Fe, irregular
(EOS GmbH, Germany) Cu particles; particle size Cu
32¡22 mm, Fe 3.6¡5.0 mm, Ni 6¡2 mm
Fe based
Fe–C–Cu–Mo–Ni
Fe based
Powder characteristics/considerations
Materials system
Category
LS
LS
LS
LM
Partial melting of powder
Partial melting of powder
Partial melting of powder
Complete melting of powder
Complete melting of powder
Complete melting of powder
LM
LM
Partial melting of powder
LS
Partial melting of powder
LS
Partial melting of powder
Partial melting of powder
LS
LS
Partial melting of powder
LS
Process Bonding mechanism
Table 5 Multicomponent metals/alloys powder systems processed by different AM processes
Relative density 94.8%
Relative density 94.6%; fracture strength
169.2 MPa; hardness 101.7 HB
179
Fe–0.8C maximum relative density 94%,
minimum roughness Ra 38 mm; Cu, Ti and
Si have negative effect on surface
quality and densification
Fe–4B minimum roughness Ra 49 mm;
mean microhardness 838.2HV;
Ti increases porosity
Relative density 65%; roughness Ra
14–16 mm; hardness 40¡7HR 15T
182
181
47
180
178
177
Relative density 91%; bending strength
630 MPa; roughness Ra 10–30 mm
Density 6.29 g cm–3; Brinell hardness
84.72 kg mm22; roughness Ra 7.41 mm;
bending strength 316 MPa
175
Porosity 2.6%; microhardness 381¡30
HV (dendritic regions), 260¡15 HV
(non-dendritic regions); roughness Ra 18.2 mm
(top surface), 12.6 mm (side surface)
High residual porosity; minimum surface
roughness Ra 23 mm; W particles
reduces part distortion
176
174
173
Ref.
Porosity 22–34%; microhardness
137–476 HV0.025
Porosity ,5 vol.-%; bending strength
900 MPa; microhardness 450–1000 HV0.025
Mechanical properties
Gu et al.
Laser additive manufacturing of metallic components
2012
VOL
57
NO
3
149
Solidification and subsolidus cracking
susceptibility and porosity formation
…
183
184, 185
Microhardness Ti2Ni phase
y600 HV, TiNi phase y244 HV,
Ti2Ni/TiNi alloy y310 HV; high wear resistance
Full density; tensile strength 600–
186
650 MPa (longitudinal) and 550–600
MPa (transverse); ductility y0.6% (both directions)
facilitate its complete melting. Also, a mixture of small
sized binder particles and relatively larger structural
particles favours an improvement in the loose packing
density of the whole powder system.61,173 This favours a
fast spreading of the molten binder by capillary forces
and a rapid rearrangement of solid particles, providing a
direct condition for a better densification of LS
processed components. The sufficient wetting of the
structural solids by the surrounding liquid plays a
crucial role in forming a sound interfacial bonding
between the remaining solids and the solidified binder.58
However, due to the considerably different melting
points and/or other mismatch in chemical/physical
properties, the remaining solids have a high tendency
of debonding along particle boundaries, resulting in an
inherent intercrystalline weakness. Gu et al.54 have
characterised the fracture surface of LS processed Ni–
CuSn–CuP powder and observed large sized brittle
dimples (Fig. 16a) and corresponding debonded Ni
particles (Fig. 16b). The weakness caused by debonding
in a fraction of areas significantly lowers the mechanical
properties of LS processed components, especially the
tensile strength.
Complete melting of powder;
in situ reactive alloying
Complete melting of powder;
phase evolutions aRazbRazbz
Ti2NiRb/B2zTi2Ni
Complete melting of powder; formation
of binary Ti2Ni/TiNi B2 intermetallic
alloy
Complete melting of prealloyed powder
International Materials Reviews
LMD
LMD
Ar atomised; spherical shape;
particle size 70–75 mm; oxygen
content ,0.06–0.1 wt-%
For LS: multiple constituents without significant difference
in melting points
Elemental powder blends in
nominal composition of 52.04Ti–47.96Ni
2012
As indicated in Table 5, Fe based powders consisting of
multiple kinds of constituents, which have the nominal
chemical compositions corresponding to a certain type
of steel, can be classified as the second category. Wang
et al.’s work175,188 on LS of Fe–29Ni–8?3Cu–1?35P
powder has disclosed the presence of Fe rich ferrite a-Fe
(Fig. 17a) and Ni rich phase (Fig. 17b) in LS processed
material, revealing that the Fe and Ni particles are only
partially melted during LS. Work of Simchi et al.173 on
LS of Fe–C–Cu–Mo–Ni powder has also revealed the
formation of a heterogeneous microstructure consisting
of unmelted constituents (Fig. 17c), due to the incomplete melting and diffusion of alloying elements.
Nevertheless, a general comparison reveals that almost
full density is achievable for this category of materials by
LS (Table 5), even though the constituents have not
melted completely. It is noticed that LM has also been
applied to process multicomponent powders. Although
Kruth et al.’s work177,178 on LM of Fe–20Ni–15Cu–
15Fe3P has proved a certain degree of enhancement of
densification and bending strength as relative to LS
processed parts (Table 5), their work179 on LM of Fe–
0?8C(–2?5Cu, 1?0Si, 1?0Ti) and Chen et al.’s work180 on
LM of Fe–4B(–9Ti) reveal that the multiple Si, Ti and
Cu constituents have a negative effect on densification of
Fe based parts. The detrimental effect is ascribed to their
high tendency to form oxides and carbides during LM
process with a significantly elevated energy input and a
complete liquid formation.
For LMD: intermetallics from elemental constituents
Intermetallic c-TiAl, Ti–47Al–
2.5V–1Cr (at-%)
Intermetallic Ti–Ni
Intermetallic Compositionally
graded Ni–Al
Intermetallic Compositionally
graded Ti–Ni
LMD
LMD
Gas atomised Al and water atomised
Ni; both particle sizes 45–75 mm
From elemental Ti to Ti–23.2 at-%Ni
Mechanical properties
Process Bonding mechanism
Powder characteristics/considerations
Materials system
Category
Table 5 Continued
150
101
Laser additive manufacturing of metallic components
Ref.
Gu et al.
VOL
57
There are growing research attempts to produce intermetallics components, including compositionally graded
intermetallics, via reactive in situ alloying from a blend
of elemental powders using LMD (Table 5). In situ
reactive alloying by LMD can be successfully achieved
by delivering elemental powders from two (or more)
powder feeders101 or using blown powder cladding
technique with mixed powder of pure elements.92 The
rapid exothermic reactions, which are normally involved
NO
3
Gu et al.
Laser additive manufacturing of metallic components
16 Fracture surface of LS processed Ni-CuSn-CuP multicomponent powder: a brittle dimples; b debonded solid particles54
during liquid formation of intermetallics, ensure the
homogeneity of in situ alloying of intermetallic
compounds.189,190 For multiple powder feeders, the
phase formation and microstructure evolution of in situ
alloyed intermetallics can be controlled along the
deposition direction by regulating the ratio of feed rates
of different powders.101 For blended elemental powders,
the chemical composition of as deposited parts can be
controlled the same as the premixed elemental powders
by keeping the identity of the divergence angles of the
elemental powder streams.191 The in situ reactive
formation of intermetallics by LMD has the following
potential advantages:
(i) raw material cost savings by eliminating the
production steps required for prealloyed
powders
(ii) suitability for fabricating a compositionally
graded structures and materials
(iii) decrease in laser energy requirements by using
reaction generated heat.101
The earliest research on in situ formation of novel
Ni70Al20Cr7Hf3 intermetallic alloys using laser cladding
was reported by Mazumder et al. in the last eighties.192 A
10 kW CO2 laser with mixed powder feed has been used
to produce Ni–A1–Cr–Hf alloys with an extended solid
solution of Hf in a near stoichiometric Ni3A1 matrix. The
laser cladding parameters, microstructure evolution and
oxidation resistance behaviour have been investigated.193
Wang’s group has performed systematic researches on
LMD fabrication of intermetallic alloys (e.g. c-TiAl,186
Ti–Ni,184,185 and CoTi194) and transition metal silicides
(e.g. Ti–Ni–Si,195 Ti–Co–Si,196 Mo–Ni–Si,197 Cr–Ni–
Si,198 and Co–Mo–Si199). In particular, the microstructural development, dry sliding wear resistance, and high
temperature wear resistance of LMD processed intermetallic components have been comprehensively studied.
Metal matrix composites
Ex situ MMCs
Ceramics reinforced MMCs exhibit an optimum combination of metallic matrix and stiffer and stronger
ceramic reinforcements. As to ex situ MMCs powders,
the ceramic reinforcing particles are added exteriorly
into the metal matrix, having each individual
particles.200 The MMCs powders are normally obtained
by mechanically alloying a mixture of different powder
components.201 The powder particles are repeatedly
fractured, cold welded, and refractured during
milling,202 producing MMCs powders with required
characteristics for AM. In a broad sense, ex situ MMCs
can be classified as multicomponent systems, with the
matrix metal and ceramic reinforcement acting as the
binder and structural material, respectively.
Additive manufacturing of MMCs, as a unique
method to obtain a designed composite material with
comprehensive properties normally not available with a
single metal or alloy,92 has already attracted growing
interest. WC–Co is the most intensively studied MMCs
for AM, including LS work by Wang et al.,203
Kumar,204 and Gläser,205 and LMD work by Xiong
et al.114,206 and Picas et al.207 Gläser has disclosed that a
high LS density is obtainable when applying the
spherical WC–Co particles, yielding a structure comparable with conventionally sintered hard metal. Xiong
et al.114,206 have fabricated bulk WC–Co MMCs using
LMD, starting from the high energy ball milled powder
consisting of nanostructured WC crystallites in Co
matrix. Microstructures with alternating layers are
observed, which is relevant to the thermal behaviour
of LMD. Variations in hardness result from the change
in cooling rate along specimen height. Other preliminary
researches have been performed on LS of ex situ MMCs
in terms of TiC/(Fe,Ni),208 SiC/Fe,209,210 SiC/Al–4?5Cu–
17 Microstructures of LS processed a, b Fe–29Ni–8?3Cu–1?35P175,188 and c Fe–C–Cu–Mo–Ni173 powders
International Materials Reviews
2012
VOL
57
NO
3
151
Gu et al.
Laser additive manufacturing of metallic components
microstructural refinement and improves the particulate
dispersion homogeneity (Fig. 19), due to the unique
metallurgical functions of RE:
(i) decreasing surface tension of the melt
(ii) resisting grain growth coarsening
(iii) increasing heterogeneous nucleation rate.
In situ MMCs
18 Fracture surface of LS processed TiC/(Fe,Ni) MMCs208
3Mg,211 SiC/Al–7Si–0?3Mg,212 WC–Co/Cu,213 ZrB2/Cu,
TiB2/Cu,214 and ZrB2/Zr.215 Laser metal deposition
of MMCs, e.g. (Ti,W)C/Ni,216 Ni coated TiC/
Inconel625,217 Ni coated TiC/Ti–6Al–4V,218 TiC/Ti–
48Al–2Cr–2Nb,219 TiC/Ti–6Al–2Zr–1Mo–1V,220 TiO2/
Ti,221 and Y2O3/Fe–Cr–Al,222 has also been reported.
The problems in terms of gas entrapment, particulate
aggregation and interfacial microcracks are regarded as
the main obstacles to obtain full density MMCs
components with favourable microstructural homogeneity. In particular, the strength and stability of the
interfacial region between ceramic reinforcement and
metal matrix govern the mechanical response of MMCs.
Failure that initiates by interfacial debonding is likely to
occur when MMCs have weak interfaces. For example,
LS processed TiC/(Fe,Ni) MMCs subjected to bending test show ductile fracture of metal matrix, but
brittle fracture and debonding around TiC particles
(Fig. 18).208 The key factor accounting for this problem
is the poor wettability between ceramics and metals. One
effective strategy is to encapsulate the ceramic particles
with a metal coating, in order to modify interfacial
structure and promote wettability. Zheng et al.217,218
have applied the Ni coated TiC to reinforce Inconel 625
and Ti–6Al–4V. This approach effectively alleviates the
formation of voids or cracks at metal/ceramic interface
and prevents clustering of ceramic particles in LMD
processed MMCs. On the other hand, Gu et al.’s
work223,224 on LS of WC reinforced Cu MMCs has
revealed that the addition of a trace amount of rare
earth (RE) compounds, e.g. La2O3 and RE–Si–Fe, can
improve laser processability of MMCs. A comparative study illustrates that RE elements favour the
The development of novel in situ MMCs via an AM
route, in which the constitutions are synthesised by
chemical reactions between elements, exhibits more
significant advantages. In situ formed ceramic reinforcement is thermodynamically stable, leading to less
degradation in elevated temperature applications.
Furthermore, the ceramic/metal interfaces within in situ
MMCs are generally cleaner and more compatible,
yielding stronger interfacial bonding and elevated
mechanical properties of the final products.225 Additive
manufacturingof in situ MMCs components represents
an important direction in AM research fields to fulfil
the future demand of novel materials with unique
properties.
The production of in situ MMCs requires a complete
melting of the starting materials to form an in situ
reaction system. Therefore, both LM and LMD have a
potential applicability. The formation of in situ reinforcement, in a broad sense, can be regarded as a bottomup method starting with atoms in the liquid to form the
required phases. Combined with the highly non-equilibrium nature of laser processing, it provides a high
possibility to create unique microstructures of in situ
phases.
The earliest report on non-equilibrium DMD synthesis of in situ Fe–Cr–C–W composites is provided by
Choi and Mazumder,226 offering an opportunity to
produce a novel wear resistant material. The composition and volume fraction of carbides can be controlled
by controlling the preheating temperature, power
density, and traverse speed. Mostly M6C or M23C6 type
carbides precipitate in the matrix. The diamond shaped
M6C carbides show good tribological characteristics.
Zhong et al.227 have reported on NiAl intermetallic
matrix composites reinforced with TiC particles
obtained by in situ LMD with coaxial feeding of Ni/
AlzTiC powder mixture. The microstructure of LMD
processed material consists of partially melted TiC,
dispersively precipitated fine TiC particles, and refined
b-NiAl phase matrix. Banerjee et al.228,229 have applied
LMD to deposit in situ TiB/Ti–6Al–4V and TiB/Ti
19 Microstructures of LS processed WC–Co/Cu MMCs a without and b with La2O3 addition223
152
International Materials Reviews
2012
VOL
57
NO
3
Gu et al.
Laser additive manufacturing of metallic components
a 700 W; b 800 w; c 875 W ; d 900 W ( Ref. 231)
20 Morphologies of in situ TiC reinforcement in LM processed Ti–Al–C powder at different laser powers
MMCs from a powder blend of Ti–6Al–4V (or Ti) and
elemental B. A unique microstructural feature of LMD
processed MMCs is the formation of highly refined
nanometre scale TiB precipitates within the grains of aTi. The ability to produce such an ultrafine dispersion of
TiB precipitates in near-net shape MMCs is highly
beneficial from the viewpoint of applicability of these
novel materials. Wang et al.230 have also prepared TiB/
Ti–6Al–4V MMCs by LMD of premixed powders of
TiB2 and Ti–6Al–4V. The modulus, yield and ultimate
strength, and wear resistance of Ti–6Al–4V are generally
increased by incorporation of TiB, but that the ductility
is decreased. Gu et al. have paid considerable research
efforts on LM fabrication of in situ MMCs such as TiC/
Ti–Al (from Ti–Al–graphite powder),231 WC/Ni (from
W–Ni–graphite powder),232 TiN/Ti5Si3 (from Ti–Si3N4
powder),233 and TiC/Ti5Si3 (from Ti–SiC powder).234
Although it has experienced long term development,
LMD/LM preparation of in situ MMCs still encounters
some few challenges. The most significant one is the
unpredictability and/or uncontrollability of the formation of in situ microstructures during processing. The
non-equilibrium metallurgical process of LMD/LM
makes it rather difficult to control the crystallisation
and growth morphology of in situ phases. For instance,
in Gu et al.’s work231 on LM of Ti–Al–C blended
powder, the morphologies of in situ TiC experience a
successive change: a laminated shapeRan octahedron
shapeRa truncated near-octahedron shapeRa nearspherical shape, on increasing the applied laser powers
(Fig. 20). As phase constitution and crystal structure
may significantly influence the final mechanical properties of MMCs, it is highly necessary to be able to
understand and control them during LMD/LM process.
Material/process considerations and
control methods
General physical aspects and design strategies
of materials for AM
In spite of two different AM approaches, LS/LM
process and LMD/DMD process share some common
physical mechanisms. This section focuses on general
physical aspects and corresponding materials considerations of AM processes.
Absorptance
Processes of AM generally involve a direct interaction of
powders with laser beam. The determination of absorptance of powders is particularly important to thermal
development, because it allows one to determine a
suitable processing window free of a non-response of
powder due to an insufficient laser energy input or a
pronounced material evaporation due to an excessive
energy input.235 The absorptance is defined as the ratio
of the absorbed radiation to the incident radiation.
Dissimilar as dense materials, only a fraction of the
incident radiation is absorbed by the outer surface of
particles. Another part of the radiation penetrates
through the interparticle voids into the depth of
the loose powder layer. The absorptance of pores
approaches that of a grey body.235 The absorptance of
powders has a direct influence on the optical penetration
depth d of the radiation, which is defined as the depth at
which the intensity of the radiation inside the material
falls to 1/e (y37%) of the original value. Owing to the
multiple reflection effect, the d measured in powders is
larger than in bulk materials.236
To understand the absorption mechanism of powders
to laser radiation, Fischer et al.237 have considered two
different energy coupling mechanisms, i.e. bulk coupling
and powder coupling. In a first step, the energy is
absorbed in a narrow layer of individual particles
determined by the bulk properties of the material,
leading to a high temperature of particle surfaces during
interaction. After thermalisation of the energy, heat
flows mainly towards the centre of particles until a local
steady state of the temperature within the powder is
obtained. Afterwards, the surrounding powder properties are responsible for the further thermal development.
Tolochko et al.235 have experimentally determined the
absorptance of a number of powders, with two different
wavelengths of 1?06 and 10?6 mm obtained by Nd : YAG
and CO2 lasers. For metals and carbides, the absorptance of powders decreases with increasing wavelength;
whereas for oxides, the absorptance increases with
increasing wavelength. The change in powder thermophysical properties, particle rearrangement, phase transitions, and melt oxidation during laser processing affect
the absorptance. Also, the absorptance of powders is
time and process dependent. Generally, the greater the
absorptance of powder, the less the laser energy output
required. That is why the laser radiation absorbing
additives are of interest for AM applicable powders.
Simchi’s work212 has proved that the addition of 5
vol.-%SiC increases the densification of Al–7Si–0?3Mg
powder during LS, mainly due to a higher effective
absorptance in the presence of SiC (SiC of 0?68 versus Al
of 0?06 under CO2 laser).238 Nevertheless, these additives
should be carefully selected to yield appropriate microstructural and mechanical properties of AM processed
powder.
International Materials Reviews
2012
VOL
57
NO
3
153
Gu et al.
Laser additive manufacturing of metallic components
Surface tension and wettability
The liquid–solid wetting characteristics are crucial for a
successful AM process. The wetting behaviour of a
partially melted LS system involves the wetting between
structural metal and liquid binder as well as the wetting
between the molten system and the solidified preprocessed layer. For the completely melted LM/LMD
systems, the second kind of wetting behaviour prevails.
The wetting of a solid by a liquid is related to the surface
tension of solid–liquid csl, solid–vapour csv and liquid–
vapour clv interfaces. Wettability can be defined by the
contact angle h (Ref. 58)
cos h~
csv {csl
clv
(1)
The liquid wets the solid as coshR1. Das165 has defined
a spreading coefficient
S~csv {csl {clv
(2)
to describe the wetting behaviour and, normally, a large
positive S favours spreading of the liquid. Conversely, if
csl.csv, h.90u and, accordingly, the liquid spheroidises
rather than wetting the solid substrate, so as to have
minimum surface energy. Das165 has disclosed that the
contamination layer of oxide being present on the
surface of melts and on the previously processed layer is
a severe impediment to a sound wettability and causes
defects such as balling. Essentially, the poor wettability
of a molten metal with oxidation inside is due to its
wetting nature similar as a metal/ceramic system.239
In order to mitigate oxidation, AM process must be
conducted in a protective atmosphere using high purity
inert gases. However, these environments alone cannot
warrant a complete wetting. Owing to the high reactivity
at melting temperatures, most metals will easily form
oxides even under very low partial pressure of oxygen.165
A certain degree of oxidation cannot be avoided under
normal AM conditions. To achieve a good wetting,
reduction of surface oxides is necessary to form clean
metal/metal interfaces. When choosing materials, fluxing
agents or in situ deoxidisers can be considered. These
additives are added in small quantities to the powders,
either mixed or prealloyed with the matrix constituent,
to aid wetting activity. In Kruth et al.,177 Zhu et al.47
and Gu et al.’s239 work, P element is added in the form
of prealloyed Fe3P, SCuP and Cu3P to Fe based and Cu
based powder systems, which are effective in enhancing
wetting behaviour and LS densification. Rare earth
elements La and Ce also contribute to the improvement
of wettability during LS of WC–Co/Cu MMCs.223,234
Viscosity
Besides the favourable wettability, it is required that the
viscosity of the melt is low enough such that it
successfully spreads on the previously processed layer
and, in the case of LS, surrounds the solid structural
particles. For a LS system consisting of a solid–liquid
mixture, the viscosity of the molten material m is
expressed as58
1{wl {2
m~m0 1{
(3)
wm
where m0 is the base viscosity that includes temperature
terms, Ql is the volume fraction of liquid phase and Qm is
a critical volume fraction of solids above which the
154
International Materials Reviews
2012
VOL
57
NO
3
mixture has essentially infinite viscosity. As to an LM or
LMD system with a complete liquid formation, the
dynamic viscosity of the liquid is defined by240
16 m 1=2
m~
c
(4)
15 kT
where m is the atomic mass, k the Boltzmann constant,
T the temperature and c the surface tension of the liquid.
Agarwala et al.’s results58 reveal that particle bonding
during LS is controlled by m0. This viscosity decreases
with increasing the working temperature, which in turn
leads to better rheological properties of the liquid in
conjunction with solid particles and, accordingly, an
improved densification. In respect of viscosity, the
metallic systems with a strong formation tendency of
intermetallic compounds are difficult to process, because
the intermetallics are generally brittle and may increase
the viscosity of the melt.101 On the other hand, the
dynamic viscosity m should be high enough to prevent
balling phenomena.58 This can be best obtained by
controlling a right solid/liquid ratio during LS, or by
varying the processing conditions to yield a feasible
operative temperature during LM/LMD.
Microstructural properties of AM processed
parts
Surface morphology and roughness
Laser sintering/LM: laser powder bed approach
Generally, the microstructural properties of AM processed parts include the exterior surface microstructure
and the interior grain microstructure. Balling phenomena are regarded as the typical microstructure occurred
on surfaces of laser processed parts using LS/LM from a
bed of loose powder. The broadly recognised definition
of balling effect is concluded as follows, combined the
previous studies by Niu et al.,241 Tolochko et al.,80
Das165 and Simchi et al.61 During LS/LM, laser
scanning is performed line by line and the laser energy
causes melting along a row of powder particles, forming
a continuous liquid scan track in a cylindrical shape.
The diminishing in the surface energy of the liquid track
is going on until the final equilibrium state through the
breaking up of the cylinder into several metallic
agglomerates in spherical shape (so called balling effect).
Balling phenomena may result in the formation of
discontinuous scan tracks and poor interline bonding
property as a current layer is processed. Furthermore,
during layer by layer LS/LM process, balling effect is a
severe impediment to a uniform deposition of the fresh
powder on the previously processed layer and tends
to cause porosity and even delamination induced by
poor interlayer bonding in combination with thermal
stresses.165
Balling effect is a complex metallurgical process that is
controlled by both powder material properties and laser
processing conditions. Comprehensive studies of balling
effect during LS/LM of multicomponent Cu based
powder and 316L stainless steel powder, including its
physical nature and control methods, are presented in
Gu et al.’s work.242,243 Three kinds of balling mechanisms during LS of Cu–30CuSn–10CuP powder are
disclosed. Scanning the initial tracks onto a cold powder
bed gives rise to the ‘first line scan balling’, due to the
high thermal gradients imposed on the melt. Using a
higher scan speed leads to the ‘shrinkage induced
Gu et al.
Laser additive manufacturing of metallic components
21 Microstructures of a LS processed 316L,243 b LMD processed 316L246 and c LM processed Fe–Ni–Cu–Fe3P (Ref. 178)
balling’, due to a significant capillary instability. The
‘splash induced balling’ with the formation of a large
amount of micrometre scale balls prevails at a high laser
power combined with a low scan speed, because of the
considerably low viscosity and long lifetime of liquid.
The following control methods have proved feasible in
decreasing balling tendency during LS/LM of 316L
powder:
(i) increasing the volumetric energy density
(ii) adding a trace amount of H3BO3 and KBF4
deoxidant.243
Recent work by Mumtaz and Hopkinson48,49 has
investigated LM of Inconel 625 using pulse shape
control to vary the energy distribution within a single
laser pulse, which is effective in attaining parts with
minimum balling effect and surface roughness. High
peak power tends to reduce top and side surface
roughness as recoil pressures flatten out the melt pool
and to reduce balling formation by increasing wettability
of the melt. Ramping up energy distribution can reduce
the maximum peak power required to melt material and
reduce material spatter generation due to a localised
preheating effect. Ramping down energy distribution
prolongs melt pool solidification, allowing more time for
molten material to redistribute and, accordingly, reducing the top surface roughness of parts.
Laser metal deposition/DMD: coaxial powder feeding
approach
Laser powder bed approach is currently the preferred
technology for manufacturing small components which
normally require a good surface finish.34 In contrast, the
surface roughness of components produced by LMD/
DMD approach is typically higher, due to the presence
of relatively larger molten pool induced by larger sized
laser spot and melt deposition mechanism applied.
Control of surface and wall roughness is, therefore, an
important issue for LMD/DMD components to reduce
post-processing costs. Normally, four directions with
respect to the cladding should be considered for the
measurements of surface roughness, i.e. the length and
width directions on the top surface, and the horizontal
and vertical directions on the walls.17 As indicated in
Mazumder et al.’s work on DMD of aluminium 1100
and H13 tool steel components, the roughness perpendicular to the cladding direction on the top surface is
y5% rougher than that parallel to the cladding. In
contrast, the roughness in the vertical direction on the
side wall was y3% larger than that in the horizontal
direction.29 The directions perpendicular to the cladding
direction on the top surface and in the vertical direction on the walls, therefore, are of primary importance
for determining the maximum roughness of DMD
components.
Laser power, traverse speed and powder flow rate are
found to be three important parameters influencing the
roughness of DMD components. The wall roughness is
directly related to layer thickness and may be increased
by depositing thicker layers, due to the variation of
beam diameter caused by defocusing. On the other
hand, using higher deposition velocities normally makes
the wall surface rougher. Mazumder et al. have
proposed a sound explanation of this phenomenon.29
At higher velocities, the cladding at the part edges
normally is unable to catch as much powder as the
internal cladding. Consequently, there is not sufficient
time for the cladding to build to the required height,
producing gaps in the cladding passes at the sample
edges. In this regard, reducing the traverse speed of the
deposition around the outline of the component favours
a decrease in wall roughness. Furthermore, the application of three sensor system proves to be effective in
improving the height control of DMD process and,
accordingly, reduces the surface roughness average of
the fabricated parts by y14%.29
Grain size and structure
The key to the mechanical properties of AM processed
components is the solidification microstructure. The
high energy laser interaction gives rise to superfast
heating and melting of materials, which is inevitably
followed by a rapid solidification on cooling. Laser
based AM processes normally offer high heating/cooling
rates (103–108 K s–1)244 at the solid/liquid interface in a
small sized molten pool (y1 mm).245 Furthermore, the
rates of quenching that occurs by conduction of heat
through the substrate are sufficiently fast to produce a
rapid solidification microstructure. Therefore, as a
characteristic of AM processed materials, grain refinement is generally expected, due to an insufficient time for
grain development/growth. For instance, the conventional dendritic solidification features of Fe based
materials are not well developed after AM, but showing
a directional cellular microstructure, due to the insufficient growth of secondary dendrite arms, e.g. LS and
LMD processed 316L powder243,246 (Fig. 21a and b) and
LM processed Fe–Ni–Cu–Fe3P powder178 (Fig. 21c).
On the other hand, either chemical concentration or
temperature gradients in molten pool may generate
surface tension gradient and resultant Marangoni
convection,53,61 making the solidification as a nonsteady state process. Meanwhile, rapid solidification
has the kinetic limitation of crystal growth that normally
follows the direction of maximum heat flow. The
International Materials Reviews
2012
VOL
57
NO
3
155
Gu et al.
Laser additive manufacturing of metallic components
22 a LS processed Al50Ti40Si10 with partially amorphous
and nanocrystalline microstructures,151 b LM processed TiCx/Ti nanocomposites and c its formation
mechanism247
simultaneous but competitive action of the above two
mechanisms, i.e. a non-equilibrium solidification nature
versus a localised directional growth tendency, may
result in a variety of crystal orientations with a localised regularity.234 Therefore, AM processed metallic
materials may have the inherent, more or less, anisotropic characteristics.
Recent research attempts have demonstrated that
laser based AM may be a useful strategy to consolidate a
number of unconventional powders with novel microstructures (e.g. amorphous and nanostructured powders). Singh et al.151 have applied LS to process
mechanically alloyed Al50Ti40Si10 powder with partially amorphous and nanocrystalline microstructures.
Following laser irradiation, the coexistence of these two
156
International Materials Reviews
2012
VOL
57
NO
3
novel microstructures is well attained (Fig. 22a). Our
recent work247 has used LM to consolidate high energy
ball milled nanostructured TiCp/Ti powder to prepare
bulk form TiCx/Ti nanocomposites. The substoichiometric TiC0?625 with a hexagonal crystal structure acts as
the reinforcement, having a lamellar nanostructure with
a mean thickness ,100 nm (Fig. 22b). The successful
formation of nanoscale TiCx is due to the action of
microscopic pressure, which is induced by evaporative
recoil of laser irradiation and surface tension of liquid,
on (111) plane of hexagonal TiCx crystals (Fig. 22c). In
essential, the successful AM of these novel structured
amorphous and nanocrystalline materials is attributed
to the unique non-equilibrium metallurgical nature of
laser irradiation.
Another important feature that is intrinsic to AM
processed components is the microstructural difference,
both in grain size and its structure, between the bottom
and top of a part along laser deposition direction.
Hofmeister et al.’s research245 has focused on grain size
variations in LMD processed 316 stainless steel and H13
tool steel powder. The microstructural scale at the
bottom of 316 parts, where conductive cooling is
highest, is 4?2–4?8 mm. Above the base (z.4 mm) the
average increases to 5?4 mm. At the bottom of H13 parts
the mean microstructural scale is 4?8–6?4 mm, and near
the top (z520 mm) the average is 7?4 mm. Wu et al.’s
work34 on LMD of b-Ti alloy also reveals that there is a
tendency for coarsening of b grains in the reheated
region near the top of previously processed layer.
Towards the top of the part, the b grains coarsen
throughout the whole of each layer, as the whole region
remains hot. Therefore, the occurrence of grain coarsening is due to:
(i) considerable remelting of the top of previous
layer
(ii) long term thermal accumulation.
Basically, the different thermal histories of different
layers of the part lead to the variation of microstructures
along the height direction, as the conduction, convection, and radiation conditions change.
Microstructural features of AM processed components are significantly influenced by the processing
parameters applied. Mazumder et al. have performed a
comparative study on microstructures of DMD processed H13 tool steel using two extreme processing
conditions.17 At high specific energy combined with a
high material deposition rate, the solidifying material is
held at a higher temperature for a longer time and,
therefore, the local temperature gradients are smaller. In
this case, the grains are coarsened and mostly equiaxed,
approximately 10–16 mm across (Fig. 23a). In contrast,
a considerably fine microstructure is formed in DMD
part, as a lower specific energy and a smaller material
addition rate are settled (Fig. 23b). A lower specific
energy is realised by using a faster traverse speed in this
case and, therefore, there is no sufficient time for the
laser to have any annealing effect on the material.
Furthermore, the profile of molten pool becomes narrow
at a higher speed and, accordingly, the local temperature
gradients are enhanced throughout the whole cladding
pass, producing the columnar grains within the majority
of DMD part. Layer thickness is another major factor in
determining the microstructures of DMD components.
Its influence is dependent on other parameters, e.g.
Gu et al.
Laser additive manufacturing of metallic components
a power 1200 W, velocity 8?5 mm s21, powder 8?0 g min21, layer thickness 1?37 mm, pass overlap 27%; b power
1200 W, velocity 50?8 mm s21, powder 4?8 g min21, layer thickness 0?254 mm, pass width overlap 66% (Ref. 17)
23 Microstructures of DMD processed H13 tool steel using different parameters
power, velocity, specific energy and powder mass flow
rate. As the specific energy is lowered, the thinner layer
thickness is required, because there is less energy per unit
area to melt powder. The coarsening of microstructures
normally occurs as the applied layer thickness increases,
due to a decrease in cooling rate.17 Furthermore,
Hofmeister et al.245 have confirmed that the microstructural scale of LMD components is more sensitive to
variations in z height (i.e. layer thickness) than to
changes in laser power and scan speed, due to the
predominance of heat conduction condition of the
substrate on cooling rate and resultant microstructures.
Mechanical properties and performance aspects
of AM processed parts
Densification level
The densification level is a fundamental property that
determines other mechanical behaviours of AM processed components. As revealed in Tables 3–5, near full
density components made from metals, alloys and
blended/composite powders can presently be fabricated
under the optimised processing conditions, especially by
LM/LMD based on a full melting mechanism. As a
general rule, a proper increase in the applied laser energy
density leads to higher part density, as confirmed in
Kruth et al.’s work128 on LM of Ti–6Al–4V (Fig. 24).
Nevertheless, for an excessive energy input, the presence
of overheated liquid with a too low viscosity may
aggravate balling effect and thermal stress, hence
inducing porosity/cracks formation.62 The suitable
processing window for a material and process combination is normally very narrow, making it difficult to
optimise the processing conditions. Additive manufacturing is a complicated shaping process, which follows a
processing routine from a ‘line’ to a ‘layer’ and then to a
‘bulk’. Additive manufacturing starts with a single line
scanning, introducing two main parameters, i.e. laser
power P and scan speed v. The completion of multiple
scan lines produces a layer. Here, another parameter, i.e.
hatch spacing h, is involved. The layer by layer
consolidation yields a bulk component, which requires
a suitable layer thickness d to be determined. The
individual P, v, h and d all have great influence on
densification of powder and, meanwhile, these parameters are inter-affected. In order to evaluate the
combined effect of these parameters and, thus, improve
the controllability of AM process, an integrated factor
termed ‘volumetric energy density’ (VED, kJ mm23) is
defined
VED~
P
vhd
(5)
24 Parameter study for part density and microstructure of LM processed Ti–6Al–4V (Ref. 128)
International Materials Reviews
2012
VOL
57
NO
3
157
Gu et al.
Laser additive manufacturing of metallic components
Gu et al.’s work on LS of W–Cu (Ref. 182) and Cu–
CuSn–CuP (Ref. 187) powder reveals that setting
VED of about 0?6–0?8 kJ mm23 and 0?16–0?23 kJ mm23
respectively favours a better yield of high density parts.
Simchi,74 and Hao et al.248 have also applied the VED to
integrally control energy input and melting mechanism
during LS/LM of Fe based powders, which have
demonstrated efficient in achieving a high densification
response.
Residual stress and strength
In general, residual stresses are considerably large in
layer by layer fabricated AM parts. Theoretical and
experimental studies by Kruth et al.249 have disclosed
that the residual stress profile consists of two zones of
large tensile stresses at the top and bottom of a LS/LM
processed part, and a large zone of intermediate
compressive stress in between. The magnitude and shape
of the residual stress profile depend on:
(i) the geometric height of the part
(ii) the material properties; and
(iii) laser scanning strategy and processing conditions.
The elastic modulus and coefficient of thermal expansion (CTE) are two most important material properties
that determine the level of residual stresses. The stresses
can be controlled by using material with a low CTE.208
Also, for MMCs parts, a reasonable selection of the
ceramic reinforcement which has a similar CTE as the
matrix metal is preferred.234 Furthermore, phase transformation may be detrimental or beneficial with respect
to residual stresses. Normally, the formation of brittle
phases during AM may promote stress cracking.
Whereas, some controlled phase transformations may
have the potential to reduce or eliminate stresses and
deformation.250 For instance, in carbon steels the
martensitic transformation leads to a volume increase
that can reach a large value of 4%,61 so that the natural
shrinkage that takes place during liquid phase processing is compensated by the material expansion after
phase transformation. Nevertheless, further systematic
studies are still required to quantify the role of phase
transformations in stress control for AM processed
metallic components.
On the other hand, care should be taken to optimise
laser processing conditions to control residual stresses.
For LS/LM process, laser scanning strategy that is being
used to melt the powder has a significant influence on
the residual stresses being developed. Normally, the
stresses are larger perpendicular to the scan direction
than along the scan direction.249 A subdivision of the
surface in smaller sectors leads to a lower stress value. A
scanning geometry with short raster lines is recommended. Also, the preheating of the substrate favours a
reduction of the residual stress level, due to a decreased
temperature gradient.251 For DMD/LMD processes,
Mazumder et al. have obtained some important understanding of stress generation and accumulation.15 It is
found that the tool path location is a critical factor for
the management of residual stress and resultant distortion. Normally, locations deposited during the last path
show residual compressive stress, since they are not
stress relieved. The other locations are deposited in
earlier paths and are subsequently stress relieved,
showing negligible residual stress.
Residual stress accumulation induced by rapid cooling and uncontrolled phase transformations may result
158
International Materials Reviews
2012
VOL
57
NO
3
in stress cracking and interlayer/interface debonding.
Normally, the cracks in AM produced components can
be divided into microscopic and macroscopic cracks.
The microscopic cracks are typically formed during
rapid solidification, which accordingly belong to the hot
cracking. Their formation is ascribed to the interruption
of liquid film at grain boundaries in the solidification
temperature range, due to the action of the tensile
stress.164,252 The macroscopic cracks are normally
regarded as the cold cracking.253 The combined influence of the low ductility of material itself and the stress
induced part deformation accounts for their propagation. The formation of microscopic and macroscopic
cracks, especially the latter, significantly lowers the
dimensional accuracy, ductility and strength of AM
fabricated components. As revealed in Tables 4, LM/
LMD processed Ti based parts have mechanical properties that are equivalent or superior to the wrought
counterparts. However, for Ni based and Fe based
alloys, post-processing such as HIP and furnace annealing/strengthening is required to favour stress relief and/
or microcrack healing, in order to realise a substantial
improvement in the final properties. Nevertheless, Zhao
et al.’s work141 reveals that the large macroscopic cracks
cannot be completely healed and eliminated through the
diffusion bonding during heat treatment.
Hardness and wear performance
Hardness is a commonly investigated mechanical
property for almost all AM processed components
(Tables 3–5). In most cases, the hardness of laser
processed materials is superior to conventionally PM
or casting materials. On the premise of a sufficiently
high densification without the formation of cracks, the
remaining of a reasonable level of residual stresses in
laser processed components favours the enhancement of
hardness.232 Associated with hardness property, recent
researches start to study the wear and tribology
performance of AM processed components. Kruth
et al.254 have investigated the wear behaviour of
prealloyed tool steel produced by LS/LM, showing that
AM technique is capable to offer excellent surface wear
properties. The densification level of AM processed
parts has a fundamental influence on wear performance.
Better wear resistance is obtained for fully dense
components. In order to further enhance the hardness
and wear property of unreinforced metals and alloys,
ceramic reinforcement is introduced to prepare MMCs
components using AM. In Ramesh et al.’s work,210
the microhardness and wear rate of LS processed
SiC/Fe MMCs respectively show y1?7-fold increase
and y66?7% decrease upon the unreinforced Fe.
Mazumder et al. have reported the in situ synthesis of
Fe–Cr–C–W MMCs using DMD process which leads to
the development of a suitable alternate for cobalt
bearing wear resistant alloys.255 Setting specific energy
input of 9?447 kJ cm22 and preheating temperature at
y500uC produces best possible combination of wear
and hardness properties and the microstructure is
comprised of MC, M7C3 and M6C types of carbides
with ferrite matrix. Our recent work234 has applied
LM to prepare in situ TiC/Ti5Si3 MMCs with novel
reinforcement architecture. The uniformly dispersed TiC
reinforcement has a unique network distribution and a
near nanoscale dendritic morphology (Fig. 25a). The in
situ TiC/Ti5Si3 MMCs have a considerably low friction
Gu et al.
Laser additive manufacturing of metallic components
25 a microstructures of LM processed in situ TiC/Ti5Si3 MMCs and b worn surface234
coefficient of 0?2 and a reduced wear rate of
1?4261024 mm3 N21m21. The high wear resistance is
attributed to the formation of adherent and strain
hardened tribolayer on the worn surface during sliding
(Fig. 25b).
Structure/property stability of AM processed
parts
Since AM production involves a long term line by line
and layer by layer localised material deposition, the
main laser processing parameters, especially the focused
beam size and output laser power, will inevitably exhibit
a certain fluctuation. Under the combined influence
of the periodic change in laser scanning pattern, a
significant thermodynamic instability may be generated
in the molten pool and the melt inside. Furthermore, the
protective atmosphere in the sealed processing chamber,
due to the continuous release of metal vapour and/or gas
impurity from the melted powder, changes significantly,
particularly during the long time AM process for large
sized components. Consequently, AM processed metals,
alloys, and MMCs parts may have the structural
differences and properties instability, hence influencing
their practical application reliance.256 Nevertheless, a
comprehensive understanding of material design, process control and metallurgical mechanisms for various
AM processes, as systemically presented in this review,
hopefully helps to overcome the structure/property
instability of AM fabricated metallic components.
Summary and prospective view
Essential of AM
Additive manufacturing technology, also widely known
as RP or RM, has a more than 20-year history of
development and, in one sense, has started to enter
mature growth stage. At present, AM has become
competitive with traditional manufacturing techniques
in terms of cost, speed, reliability and accuracy.
Therefore, AM is believed by many experts that it is a
‘next generation’ technology. The word ‘rapid’ in RP/
RM phrases is relative; it can typically produce
components in a few hours, although it varies significantly depending on the type of machine being used and
the size, number and complexity of parts being produced
simultaneously. The concept ‘rapid’ is largely reflected
by its processing philosophy: a direct shaping from loose
powder to bulk form parts, without having to invest the
time or resource to develop tooling for support.
Unique application areas
Applications of AM technology have been realised in a
variety of industries including aerospace, military, automotive, dental, medical, etc. The primary application is to
fabricate intricate aero- or land-based engine components in complex geometries out of hard to machine
materials.87,91 AM produces shapes close enough to the
final product to eliminate the need for rough machining.
Second, the tooling industry applies AM to produce
functional tool components, in particular the small batch
or one-off parts. One of the most promising applications
is to manufacture plastic injection tools and die cast
tooling.15,29,257 Rapid tooling is, therefore, considered as
an important subcategory of AM.62,258 Third, AM has
found its place in medical devices manufacturing,
including the specialty surgical instruments and prosthetic implants.257 Medical implants have to be extremely
flexible to fit in a specific patient. Also, the weight of these
implants is required to be as light as possible while still
ensuring proper structural and mechanical characteristics. This is the reason that porous metallic parts with
particular configurations are normally desired. Additive
manufacturing technology has demonstrated to be a
favourable solution.259
Future research interests
Researches on laser based AM of metallic components,
as reviewed in the present article, are interdisciplinary,
integrating materials science, metallurgical engineering,
mechanical engineering and laser technology. Significant
research and understanding are still required in the
aspects of materials preparation and characterisation,
process control and optimisation, and theories of
physical and chemical metallurgy for each AM process.
Combining the opinions proposed by other experts in
AM research fields,62,92,94 the authors have summarised
the following issues that are of particular significance for
future development of AM technology.
Extension of AM applicable powders
The basic role of powder material properties in a
successful AM production has long been recognised.
Certain powder materials which have sound processability in conventional PM routes are found inapplicable
for AM processes. The unique processing manner of
AM brings forward some special requirements of
applicable powders. For instance, high flowability is a
primary consideration for powders, since AM processing is based on powder spreading (LS/LM) or powder
feeding (LMD/DMD). Further studies in terms of
International Materials Reviews
2012
VOL
57
NO
3
159
Gu et al.
Laser additive manufacturing of metallic components
chemical and physical properties, preparation technique
and characterisation method of powder materials are
required to fit a successful AM application. Material
researches should be extended to multiple systems and
forms, including prealloyed/blended/composite Fe, Ni,
Ti, Al, Cu and Mg based powders, in order to realise a
diversity of AM applicable materials.
Development of novel materials and ‘designed materials’
The application of AM technology to prepare novel
structured high performance functional components is
of unique interest. The special MIM processing procedure and highly non-equilibrium nature of AM favour
the formation of bulk form materials with unique
microstructures and properties. It provides a beneficial
method to develop new materials, such as nanophase,
amorphous, functionally gradient and porous materials.
On the other hand, the unique integration of homogenisation design, heterogeneous modelling and LMD/
DMD process (as reviewed in the section on ‘Unique
applications of LMD/DMD technology’) offers a
revolutionary approach for manufacturing ‘designed
materials’ with properties and functions which do not
currently exist.
Establishment of AM process database
Comprehensive knowledge is involved in AM processes,
including laser technology, material science, PM and
rapid solidification.92 The suitable AM processing data
for various metallic materials should be accumulated.
Combined with the optimisation in powder material
design and preparation, the corresponding optimal AM
processing parameters should be experimentally determined. After a sufficient accumulation, the material
process database can be established, realising a simplified, precise and stable control of AM treatment of
versatile powder materials for industrial applications.
Microstructure development and metallurgical mechanism
Additive manufacturing processes offer a promising
potential for development of novel bulk form materials
of designed compositions, microstructures and properties. However, due to the significant non-equilibrium
nature of laser processing and the complicated mutual
influence of material and process parameters, the
unpredictability and/or uncontrollability of the formation of phases and microstructures in an AM route still
remain as a major challenge. The underlying physical
and chemical metallurgical mechanisms responsible for
the variation of microstructural and mechanical properties should be determined, in order to give a strong
theoretical basis for AM processes.
Theoretical modelling and simulation
The existent reports on the theoretical modelling and
simulation of AM processes are mostly focused on the
relatively macroscopic thermal field,38 stress field,119 and
volume shrinkage,260 based on a heat transfer model or
heat–stress coupled model with a necessary consideration of melting/solidification phase transformations, but
few have incorporated the microscopic fluid flow
calculations due to the involved complexity.261 The
theoretical study of the metallurgical thermodynamics
and kinetics behaviours of the melt within nonequilibrium molten pool is of particular importance,
including the mass transfer and fluid flow, crystal
nucleation and growth, and melting and mixing
160
International Materials Reviews
2012
VOL
57
NO
3
behaviour of key alloying/additive elements, thereby enabling the microstructure to be tailored
according to the local performance requirements of the
component.262
Acknowledgements
One of the authors (D. D. Gu) gratefully appreciates the
financial supports from the Alexander von Humboldt
Foundation Germany, the National Natural Science
Foundation of China (grant nos. 51054001 and
51104090), the Aeronautical Science Foundation of
China (grant no. 2010ZE52053), the Natural Science
Foundation of Jiangsu Province (grant no. BK2009374),
and the NUAA Research Funding (grant no. NS2010156).
References
1. J. J. Beaman and C. R. Deckard: ‘Selective laser sintering with
assisted powder handling’, US patent no. 4938816, 1990.
2. D. L. Bourell, H. L. Marcus, J. W. Barlow, J. J. Beaman and
C. R. Deckard: ‘Multiple material systems for selective beam
sintering’, US patent no. 5076869, 1991.
3. J. Koch and J. Mazumder: ‘Apparatus and methods for
monitoring and controlling multi-layer laser cladding’, US patent
no. 6122564, 2000.
4. I. Gibson, D. W. Rosen and B. Stucker: ‘Additive manufacturing
technologies: rapid prototyping to direct digital manufacturing’,
1st edn, 17; 2010, New York, Springer Science/Business Media.
5. C. K. Chua, K. F. Leong and C. S. Lim: ‘Rapid prototyping:
principles and applications’, 2nd edn, 11; 2003, Singapore, World
Scientific Publishing Company.
6. P. K. Venuvinod and W. Ma: ‘Rapid prototyping: laser-based and
other technologies’, 1st edn, 6; 2001, Norwell, Kluwer Academic
Publishers.
7. L. Lü, J. Y. H. Fuh and Y. S. Wong: ‘Laser-induced materials and
processes for rapid prototyping’, 1st edn, 6; 2001, Norwell,
Kluwer Academic Publishers.
8. A. P. West, S. P. Sambu and D. W. Rosen: Comput. Aid. Des.,
2001, 33, 65–79.
9. J. Park, M. J. Tari and H. T. Hahn: Rapid Prototyp. J., 2000, 6,
36–49.
10. R. W. Gray, D. G. Baird and J. H. Bohn: Rapid Prototyp. J.,
1998, 4, 14–25.
11. F. Ribeiro: Comput. Control Eng. J., 1998, 9, 31–38.
12. D. L. Bourell, H. L. Marcus, J. W. Barlow and J. J. Beaman: Int.
J. Powder Metall., 1992, 28, 369–381.
13. T. H. C. Childs, M. Berzins, G. R. Ryder and A. Tontowi: Proc.
Inst. Mech. Eng. B, 1999, 213B, 333–349.
14. D. T. Pham, S. Dimov and F. Lacan: Proc. Inst. Mech. Eng. B,
1999, 213B, 435–449.
15. J. Mazumder, J. Choi, K. Nagarathnam, J. Koch and D. Hetzner:
JOM, 1997, 49, 55–60.
16. S. Das, M. Wohlert, J. J. Beaman and D. L. Bourell: JOM, 1998,
50, 17–20.
17. J. Mazumder, A. Schifferer and J. Choi: Mater. Res. Innovat.,
1999, 3,118–131.
18. D. H. Mueller and H. Mueller: ‘Experiences using rapid
prototyping techniques to manufacture sheet metal forming
tools’, Proc. Int. Symp. on ‘Automotive technology and automation’, Dublin, Ireland, September 2000. No. 00SE008.
19. N. Levy, R. Schindel and J. P. Kruth: CIRP Ann., 2003, 52, 589–
609.
20. J. Choi and J. Mazumder: ‘Rapid manufacturing by laser aided
direct metal deposition process: issues and examples’, Proc.
ASME Des. Eng. Techn. Conf. and Comput. Inform. Eng.
Conf., Pittsburgh, PA, USA, September 2001, ASME, 333–
338.
21. D. King and T. Tansey: J. Mater. Process. Technol., 2003, 132,
42–48.
22. P. A. Webb: J. Med. Eng. Technol., 2000, 24, 149–153.
23. K. F. Leong, C. M. Cheah and C. K. Chua: Biomaterials, 2003,
24, 2363–2378.
24. J. Hänninen: Adv. Mater. Process., 2002, 160, 33–36.
25. EOS Electro Optical Systems, www.eos.info.
Gu et al.
26. N. K. Tolochko, V. V. Savich, T. Laoui, L. Froyen, G. Onofrio,
E. Signorelli and V. I. Titov: Proc. Inst. Mech. Eng. L, 2002, 216,
267–270.
27. S. Pogson, P. Fox and W. O’Neill: ‘Experimental investigation of
the direct metal laser re-melting of dissimilar metals’, Proc. 3rd
Natl Conf. on ‘Rapid prototyping, tooling and manufacturing’,
High Wycombe, UK, June 2002, Buckinghamshire Chilterns
University College, 139–147.
28. Sauer product – Prototyping, Lasercusing, www.sauerproduct.
com.
29. J. Mazumder, D. Dutta, N. Kikuchi and A. Ghosh: Opt. Laser
Eng., 2000, 34, 397–414.
30. C. Atwood, M. Griffith, L. Harwell, E. Schlienger, M. Ensz,
J. Smugeresky, T. Romero, D. Greene and D. Reckaway: ‘Laser
engineered net shaping (LENSTM): a tool for direct fabrication of
metal parts’, Proc. Int. Cong. on ‘Applications of lasers and
electro-optics’, Orlando, FL, USA, November 1998, Laser Institute of America, E1-7 vol.2.
31. J. Brooks, C. Robino, T. Headley, S. Goods and M. Griffith:
‘Microstructure and property optimization of LENS deposited
H13 tool steel’, Proc. 10th Symp. on ‘Solid freeform fabrication’,
Austin, TX, USA, August 1999, The University of Texas at
Austin, 375–382.
32. G. K. Lewis and E. Schlienger: Mater. Des., 2000, 21, 417–423.
33. A. J. Pinkerton and L. Li: Thin Solid Films, 2004, 453–454, 600–
605.
34. X. Wu: Mater. Sci. Technol., 2007, 23, 631–640.
35. X. Lin, T. M. Yue, H. O. Yang and W. D. Huang: Metall. Mater.
Trans. A, 2007, 38A, 127–137.
36. Y. D. Wang, H. B. Tang, Y. L. Fang and H. M. Wang: Mater.
Sci. Eng. A, 2010, A527, 4804–4809.
37. B. Xiao and Y. W. Zhang: J. Manuf. Sci. Eng., 2008, 130, 041002.
38. K. Dai and L. Shaw: Acta Mater., 2004, 52, 69–80.
39. X. He and J. Mazumder: J. Appl. Phys., 2007, 101, 053113.
40. X. He, G. Yu and J Mazumder: J. Phys. D: Appl. Phys., 2010, 43,
015502.
41. A. Manthiram, D. L. Bourell and H. L. Marcus: JOM, 1993, 45,
66–70.
42. H. Asgharzadeh and A. Simchi: Mater. Sci. Eng. A, 2005, A403,
290–298.
43. E. C. Santos, M. Shiomi, K. Osakada and T. Laoui: Int. J. Mach.
Tool Manuf., 2006, 46, 1459–1468.
44. J. P. Kruth, X. Wang, T. Laoui and L. Froyen: Assem. Autom.,
2003, 23, 357–371.
45. V. K. Balla, S. Bodhak, S. Bose and A. Bandyopadhyay: Acta
Biomater., 2010, 6, 3349–3359.
46. A. Bandyopadhyay, F. Espana, V. K. Balla, S. Bose, Y. Ohgami
and N. M. Davies: Acta Biomater., 2010, 6, 1640–1648.
47. H. H. Zhu, L. Lu and J. Y. H. Fuh: J. Mater. Process Technol.,
2003, 140, 314–317.
48. K. A. Mumtaz and N. Hopkinson: J. Mater. Process Technol.,
2010, 210, 279–287.
49. K. Mumtaz and N. Hopkinson: Rapid Prototyping J., 2009, 15,
96–103.
50. M. L. Griffith, D. M. Keicher, C. L. Atwood, J. A. Romero, J. E.
Smugeresky, L. D. Harwell and D. L. Greene: ‘Free form
fabrication of metallic components using laser engineered net
shaping (LENSTM)’, Proc. 7th Symp. on ‘Solid freeform fabrication’, Austin, TX, USA, August 1996, The University of Texas at
Austin, 125–132.
51. D. J. Thoma, G. K. Lewis, J. O. Milewski, K. C. Chen and R. B.
Nemec: ‘Rapid fabrication of materials using directed light
fabrication’, Proc. THERMEC ’97, Wollongong, Australia,
July 1997, University of Wollongong. Available at: http://
www.osti.gov/energycitations/product.biblio.jsp?osti_id=542062
52. J. O. Milewski, G. K. Lewis, D. J. Thoma, G. I. Keel, R. B.
Nemec and R. A. Reinert: J. Mater. Process. Technol., 1998, 75,
165–172.
53. A. Simchi and H. Pohl: Mater. Sci. Eng. A, 2003, A359, 119–128.
54. D. D. Gu, Y. F. Shen and Z. J. Lu: Mater. Des., 2009, 30, 2099–
2107.
55. P. Fischer, V. Romano, A. Blatter and H. P. Weber: Laser Phys.
Lett., 2005, 2, 48–55.
56. H. Schleifenbaum, W. Meiners, K. Wissenbach and C. Hinke:
CIRP J. Manuf. Sci. Technol., 2010, 2, 161–169.
57. T. Kaiser and G. J. Albrecht: Laser Techn. J., 2007, 4, 54–57.
58. M. Agarwala, D. Bourell, J. Beaman, H. Marcus and J. Barlow:
Rapid Prototyping J., 1995, 1, 26–36.
59. S. Das, J. J. Beaman, M. Wohlert and D. L. Bourell: Rapid
Prototyping J., 1998, 4, 112–117.
Laser additive manufacturing of metallic components
60. A. N. Chatterjee, S. Kumar, P. Saha, P. K. Mishra and A. Roy
Choudhury: J. Mater. Process Technol., 2003, 136, 151–157.
61. A. Simchi, F. Petzoldt and H. Pohl: Int. J. Powder Metall., 2001,
37, 49–61.
62. J. P. Kruth, G. Levy, F. Klocke and T. H. C. Childs: CIRP Ann.,
2007, 56, 730–759.
63. M. W. Khaing, J. Y. H. Fuh and L. Lu: J. Mater. Process.
Technol., 2001, 113, 269–272.
64. J. P. Kruth, L. Froyen, M. Rombouts, J. van Vaerenbergh and
P. Mercells: CIRP Ann., 2003, 52, 139–142.
65. H. H. Zhu, L. Lu and J. Y. H. Fuh: Mater. Sci. Eng. A, 2004,
A371, 170–177.
66. G. B. Prabhu and D. L. Bourell: ‘Supersolidus liquid phase
selective laser sintering of prealloyed bronze powder’, Proc. 4th
Symp. on ‘Solid freeform fabrication’, Austin, TX, USA, August
1993, The University of Texas at Austin, 317–324.
67. J. X. Liu and M. L. Ryneson: ‘Blended powder solid supersolidus
liquid phase sintering’, US patent no. 7070734 B2, 2006.
68. H. J. Niu and I. T. H. Chang: Scr. Mater., 1998, 39, 67–72.
69. M. Agarwala, D. Bourell, J. Beaman, H. Marcus and J. Barlow:
Rapid Prototyping J., 1995, 1, 36–44.
70. S. Das, M. Wohlert, J. J. Beaman and D. L. Bourell: Mater. Des.,
1999, 20, 115–121.
71. S. Kumar and J. P. Kruth: Mater. Des., 2007, 28, 400–407.
72. J. P. Kruth, P. Mercelis, J. Van Vaerenbergh, L. Froyen and
M. Rombouts: Rapid Prototyping J., 2005, 11, 26–36.
73. R. Poprawe, P. Loosen, H. D. Hoffmann: ‘The future of high
power laser techniques’, Proc. 16th Int. Symp. on ‘Gas flow,
chemical lasers, and high-power lasers’, Gmunden, Austria (ed. D.
Schuöcker), 634602; 2007, SPIE.
74. A. Simchi: Mater. Sci. Eng. A, 2006, A428, 148–158.
75. H. J. Niu and I. T. H. Chang: J. Mater. Sci., 2000, 35, 31–38.
76. A. Fukuda, M. Takemoto, T. Saito, S. Fujibayashi, M. Neo,
Deepak K. Pattanayak, T. Matsushita, K. Sasaki, N. Nishida, T.
Kokubo and T. Nakamura: Acta Biomater., 2011, 7, 2327–2336.
77. D. Buchbinder and K. Wissenbach: ‘Additive manufacturing of
aluminum components with minimal distortion’, Fraunhofer ILT
annual report, Fraunhofer ILT, Aachen, Germany, 2008.
78. D. Becker and K. Wissenbach: ‘Additive Manufacturing of
copper components’, Fraunhofer ILT annual report, Fraunhofer
ILT, Aachen, Germany, 2009.
79. N. K. Tolochko, M. K. Arshinov, A. V. Gusarov, V. I. Titov,
T. Laoui and L. Froyen: Rapid Prototyping J., 2003, 9, 314–326.
80. N. K. Tolochko, S. E. Mozzharov, I. A. Yadroitsev, T. Laoui,
L. Froyen, V. I. Titov and M. B. Ignatiev: Rapid Prototyping J.,
2004, 10, 78–87.
81. F. Verhaeghe, T. Craeghs, J. Heulens and L. Pandelaers: Acta
Mater., 2009, 57, 6006–6012.
82. M. Shiomi, K. Osakada, K. Nakamura, T. Yamashita and
F. Abe: CIRP Ann, 2004, 53, 195–198.
83. S. Pogson, P. Fox, W. O’Neill and C. J. Sutcliffe: Mater. Sci. Eng.
A, 2004, A386, 453–459.
84. J. L. Koch and J. Mazumder: ‘Rapid prototyping by laser
cladding’, Proc. Int. Cong. on ‘Applications of lasers and electrooptics’, Vol. 77, (ed. P. Denny et al.), 556–565; 1993, Orlando, FL,
Laser Institute of America.
85. J. Mazumder and D. D. Voelkel: ‘Method and apparatus for
noncontact surface contour measurement’, US patent
no. 5446549, 1995.
86. J. Mazumder, D. Morgan, T. W. Skszek and M. Lowney: ‘Direct
metal deposition apparatus utilizing rapid-response diode laser
source’, US patent no. 7765022, 2010.
87. B. Dutta, V. Singh, H. Natu, J. Choi and J. Mazumder: Adv.
Mater. Process., 2009, 167, 29–31.
88. B. Dutta, S. Palaniswamy, J. Choi, L. J. Song and J. Mazumder:
Adv. Mater. Process., 2011, 169, 33–36.
89. K. H. Shin, H. Natu, D. Dutta and J. Mazumder: Mater. Des.,
2003, 24, 339–353.
90. A. W. Hammeke: ‘Laser spray nozzle and method’, US patent
no. 4724299, 1988.
91. R. P. Mudge and N. R. Wald: Weld J., 2007, 86, 44–48.
92. M. Zhong and W. Liu: Proc. Inst. Mech. Eng. C, 2010, 224C,
1041–1060.
93. The POM Group Inc., www.pomgroup.com.
94. J. Mazumder: JOM, 2000, 52, 28–29.
95. R. K. Oruganti, A. K. Ghosh and J. Mazumder: ‘Fabrication and
creep properties of compositionally graded nickel base alloys’,
Proc. Symp. on ‘Power metallurgy alloys and particulate materials
for industrial applications’ St Louis, MO, USA, October 2000,
TMS, 123–133.
International Materials Reviews
2012
VOL
57
NO
3
161
Gu et al.
Laser additive manufacturing of metallic components
96. R. K. Oruganti and A. K. Ghosh: Metall. Mater. Trans. A, 2003,
34A, 2643–2653.
97. K. I. Schwendner, R. Banerjee, P. C. Collins, C. A. Brice and
H. L. Fraser: Scr. Mater., 2001, 45, 1123–1129.
98. R. Banerjee, P. C. Collins and H. L. Fraser: Metall. Mater. Trans.
A, 2002, 33A, 2129–2138.
99. R. Banerjee, P. C. Collins, D. Bhattacharyya, S. Banerjee and
H. L. Fraser: Acta Mater., 2003, 51, 3277–3292.
100. R. R. Unocic and J. N. DuPont: Metall. Mater. Trans. A, 2003,
34A, 439–445.
101. W. P. Liu and J. N. DuPont: Metall. Mater. Trans. A, 2003, 34,
2633–2641.
102. P. C. Collins, R. Banerjee, S. Banerjee and H. L. Fraser: Mater.
Sci. Eng. A, 2003, A352, 118–128.
103. J. Mazumder: ‘Direct metal deposition: process control, properties
and applications’, Proc. Int. Conf. on ‘Processing and manufacturing of advanced materials – processing, fabrication, properties,
applications’, Berlin, Germany, August 2009. Available at: http://
claim.engin.umich.edu/papers/mazumderthermec09/index.html
104. L. Wang, S. Felicelli, Y. Gooroochurn, P. T. Wang and M. F.
Horstemeyer: Mater. Sci. Eng. A, 2008, A474, 148–156.
105. L. Wang, S. D. Felicelli and J. E. Craig: J. Manuf. Sci. Eng., 2009,
131, 041019.
106. J. Yu, X. Lin, J. J. Wang, J. Chen and W. D. Huang: Appl. Surf.
Sci., 2010, 256, 4612–4620.
107. H. Tan, J. Chen, F. Y. Zhang, X. Lin and W. D. Huang: Int. J.
Mach. Tool Manuf., 2010, 50, 1–8.
108. W. Hofmeister, M. Wert, J. Smugeresky, J. A. Philliber, M.
Griffith and M. Ensz: JOM, 1999, 51, http://www.tms.org/pubs/
journals/JOM/9907/Hofmeister/Hofmeister-9907.html.
109. B. Zheng, Y. Zhou, J. E. Smugeresky, J. M. Schoenung and E. J.
Lavernia: Metall. Mater. Trans. A, 2008, 39A, 2228–2236.
110. A. Vasinonta, J. L. Beuth and M. Griffith: J. Manuf. Sci. Eng.,
2007, 129, 101–109.
111. L. Wang and S. Felicelli: Mater. Sci. Eng. A, 2006, A435–436,
625–631.
112. G. K. Lewis and E. Schlienger: Mater. Des, 2000, 21, 417–423.
113. B. Zheng, Y. Zhou, J. E. Smugeresky, J. M. Schoenung and E. J.
Lavernia: Metall. Mater. Trans. A, 2008, 39A, 2237–2245.
114. Y. H. Xiong, J. E. Smugeresky and J. M. Schoenung: J. Mater.
Process. Technol., 2009, 209, 4935–4941.
115. D. Gill: ‘Laser engineered net shaping’, www.sandia.gov/mst/pdf/
LENS.pdf
116. M. L. Griffith, M. E. Schlienger, L. D. Harwell, M. S. Oliver,
M. D. Baldwin, M. T. Ensz, M. Essien, J. Brooks, C. V. Robino,
J. E. Smugeresky, W. H. Hofmeister, M. J. Wert and D. V.
Nelson: Mater. Des, 1999, 20, 107–113.
117. L. Qian, J. Mei, J. Liang and X. Wu: Mater. Sci. Technol., 2005,
21, 597–605.
118. P. Pratt, S. D. Felicelli, L. Wang and C. R. Hubbard: Metall.
Mater. Trans. A, 2008, 39A, 3155–3163.
119. L. Wang, S. D. Felicelli and P. Pratt: Mater. Sci. Eng. A, 2008,
A496, 234–241.
120. P. Rangaswamy, M. L. Griffith, M. B. Prime, T. M. Holden, R. B.
Rogge, J. M. Edwards and R. J. Sebring: Mater. Sci. Eng. A,
2005, A399, 72–83.
121. P. S. Mohanty and J. Mazumder: Metall. Mater. Trans. B, 1998,
29B, 1269–1279.
122. P. Fischer, H. Leber, V. Romano, H. P. Weber, N. P. Karapatis,
C. André and R. Glardon: Appl. Phys. A, 2004, 78A, 1219–1227.
123. E. C. Santos, K. Osakada, M. Shiomi, Y. Kitamura and F. Abe:
Proc. Inst. Mech. Eng. C, 2004, 218C, 711–719.
124. T. Laoui, E. Santos, K. Osakada, M. Shiomi, M. Morita, S. K.
Shaik, N. K. Tolochko, F. Abe and M. Takahashi: Proc. Inst.
Mech. Eng. C, 2006, 220C, 857–863.
125. B. V. Krishna, S. Bose and A. Bandyopadhyay: Acta Biomater.,
2007, 3, 997–1006.
126. S. R. Pogson, P. Fox, C. J. Sutcliffe and W. O’Neill: Rapid
Prototyping J., 2003, 9, 334–343.
127. M. Khan and P. Dickens: Gold Bull., 2010, 43, 114–121.
128. B. Vandenbroucke and J. P. Kruth: Rapid Prototyping J., 2007,
13, 196–203.
129. B. V. Krishna, S. Bose and A. Bandyopadhyay: Acta Biomater.,
2007, 3, 997–1006.
130. W. C. Xue, B. V. Krishna, A. Bandyopadhyay and S. Bose: Acta
Biomater., 2007, 3, 1007–1018.
131. G. P. Dinda, L. Song and J. Mazumder: Metall. Mater. Trans. A,
2008, 39A, 2914–2922.
162
International Materials Reviews
2012
VOL
57
NO
3
132. W. Meiners and K. Wissenbach: ‘Manufactoring of individual
titanium implants using selective laser melting (SLM)’, www.ilt.
fraunhofer.de/eng/101937.html
133. D. A. Hollander, M. von Walter, T. Wirtz, R. Sellei, B. SchmidtRohlfing, O. Paar and H. J. Erli: Biomaterials, 2006, 27, 955–963.
134. X. Wu, R. Sharman, J. Mei and W. Voice: Mater. Des., 2004, 25,
103–109.
135. X. J. Tian, S. Q. Zhang, A. Li and H. M. Wang: Mater. Sci. Eng.
A, 2010, A527, 1821–1827.
136. I. Yadroitsev, L. Thivillon, Ph. Bertrand and I. Smurov: Appl.
Surf. Sci., 2007, 254, 980–983.
137. I. Yadroitsev, M. Pavlov, Ph. Bertrand and I. Smurov:
‘Mechanical properties of samples fabricated by selective laser
melting’, Proc. 14th Eur. Conf. on ‘Rapid prototyping and manufacturing’, Paris, France, June 2009. Available at: http://code80.
net/afpr/content/assises/2009/actes_aepr2009/papiers/s4_5.pdf.
138. K. A. Mumtaz, P. Erasenthiran and N. Hopkinson: J. Mater.
Process. Technol., 2008, 195, 77–87.
139. G. P. Dinda, A. K. Dasgupta and J. Mazumder: Mater. Sci. Eng.
A, 2009, A509, 98–104.
140. X. M. Zhao, J. Chen, X. Lin and W. D. Huang: Mater. Sci. Eng.
A, 2008, A478, 119–124.
141. X. M. Zhao, X. Lin, J. Chen, L. Xue and W. D. Huang: Mater.
Sci. Eng. A, 2009, A504, 129–134.
142. J. Li and H. M. Wang: Mater. Sci. Eng. A, 2010, A527, 4823–
4829.
143. R. Morgan, C. J. Sutcliffe and W. O’Neill: J. Mater. Sci., 2004,
39, 1195–1205.
144. R. H. Morgan, A. J. Papworth, C. Sutcliffe, P. Fox and W.
O’Neill: J. Mater. Sci., 2002, 37, 3093–3100.
145. I. Yadroitsev, Ph. Bertrand and I. Smurov: Appl. Surf. Sci., 2007,
253, 8064–8069.
146. J. W. Xie, P. Fox, W. O’Neill and C. J. Sutcliffe: J. Mater.
Process. Technol., 2005, 170, 516–523.
147. A. Simchi and H. Asgharzadeh: Mater. Sci. Technol., 2004, 20,
1462–1468.
148. S. Bhattacharya, G. P. Dinda, A. K. Dasgupta and J. Mazumder:
Mater. Sci. Eng. A, 2011, A528, 2309–2318.
149. Y. Xue, A. Pascu, M. F. Horstemeyer, L. Wang and P. T. Wang:
Acta Mater., 2010, 58, 4029–4038.
150. B. Zheng, Y. Zhou, J. E. Smugeresky and E. J. Lavernia: Metall.
Mater. Trans. A, 2009, 40A, 1235–1245.
151. S. S. Singh, D. Roy, R. Mitra, R. V. S. Rao, R. K. Dayal, B. Raj
and I. Manna: Mater. Sci. Eng. A, 2009, A501, 242–247.
152. D. Buchbinder, W. Meiners, K. Wissenbach, K. MüllerLohmeier, E. Brandl and N. Skrynecki: ‘Rapid manufacturing
of aluminium parts for serial production via selective laser melting
(SLM)’, Proc. Int. Conf. on ‘Rapid manufacturing’,
Loughborough, UK, July 2009, University of Loughborough,
1–12.
153. E. Brandl, U. Heckenberger, V. Holzinger and D. Buchbinder:
Mater. Des., 2012, 34, 159–169.
154. E. Louvis, P. Fox and C. J. Sutcliffe: J. Mater. Process. Technol.,
2011, 211, 275–284.
155. G. D. J. Ram, C. K. Esplin and B. E. Stucker: J. Mater. Sci.,
2008, 19, 2105–2111.
156. C. F. Xue, Y. Dai and X. L. Tian: Trans. Nonferr. Met. Soc.
Chin., 2006, 16, S1982–S1985.
157. D. Becker, W. Meiners and K. Wissenbach: ‘Additive manufacturing of copper alloys by selective laser melting’, Proc. Conf. on
‘Lasers in Manufacturing’, Munich, Germany, June 2009, WLT,
195–199.
158. S. Bhattacharya, G. P. Dinda, A. K. Dasgupta, H. Natu, B. Dutta
and J. Mazumder: J. Alloy Compd, 2011, 509, 6364–6373.
159. M. Boccalini and H. Goldenstein: Int. Mater. Rev., 2001, 46, 92–
115.
160. X. Wu, J. Liang, J. Mei, C. Mitchell, P. S. Goodwin and
W. Voice: Mater. Des., 2004, 25, 137–144.
161. L. Facchini, E. Magalini, P. Robotti, A. Molinari, S. Höges and
K. Wissenbach: Rapid Prototyping J., 2010, 16, 450–459.
162. Y. M. Li, H. O. Yang, X. Lin, W. D. Huang, J. G. Li and Y. H.
Zhou: Mater. Sci. Eng. A, 2003, A360, 18–25.
163. C. P. Paul, P. Ganesh, S. K. Mishra, P. Bhargava, J. Negi and
A. K. Nath: Opt. Laser Technol., 2007, 39, 800–805.
164. M. L. Zhong, H. Q. Sun, W. J. Liu, X. F. Zhu and J. J. He: Scr.
Mater., 2005, 53, 159–164.
165. S. Das: Adv. Eng. Mater., 2003, 5, 701–711.
166. C. S. Wright, M. Youseffi, S. P. Akhtar, T. H. C. Childs,
C. Hauser, P. Fox and J. Xie: Mater. Sci. Forum, 2006, 514–516,
516–523.
Gu et al.
167. T. H. C. Childs, C. Hauser and M. Badrossamay: Proc. Inst.
Mech. Eng. B, 2005, 219B, 339–357.
168. T. H. C. Childs, C. Hauser and M. Badrossamay: CIRP Ann.,
2004, 53, 191–194.
169. S. Tsopanos, R. A. W. Mines, S. McKown, Y. Shen, W. J.
Cantwell, W. Brooks and C. J. Sutcliffe: J. Manuf. Sci. Eng., 2010,
132, 041011.
170. I. Yadroitsev, I. Shishkovsky, P. Bertrand and I. Smurov: Appl.
Surf. Sci., 2009, 255, 5523–5527.
171. A. V. Gusarov, I. Yadroitsev, Ph. Bertrand and I. Smurov: Appl.
Surf. Sci., 2007, 254, 975–979.
172. H. H. Zhu, L. Lu and J. Y. H. Fuh: Proc. Inst. Mech. Eng. B,
2006, 220B, 183–190.
173. A. Simchi, F. Petzoldt and H. Pohl: J. Mater. Process Technol.,
2003, 141, 319–328.
174. A. Simchi and H. Pohl: Mater. Sci. Eng. A, 2004, 383A, 191–200.
175. Y. Wang, J. Bergström and C. Burman: J. Mater. Process
Technol., 2006, 172, 77–87.
176. H. H. Zhu, J. Y. H. Fuh and L. Lu: Proc. Inst. Mech. Eng. C,
2003, 217C, 139–147.
177. J. P. Kruth, S. Kumar and J. van Vaerenbergh: Rapid Prototyping
J., 2005, 11, 287–292.
178. J. P. Kruth, L. Froyen, J. van Vaerenbergh, P. Mercelis, M.
Rombouts and B. Lauwers: J. Mater. Process. Technol., 2004,
149, 616–622.
179. M. Rombouts, J. P. Kruth, L. Froyen and P. Mercelis: CIRP
Ann., 2006, 55, 187–192.
180. X. C. Chen, J. W. Xie and P. Fox: Mater. Sci. Technol., 2004, 20,
715–719.
181. D. D. Gu, Y. F. Shen, S. Q. Fang and J. Xiao: J. Alloys Compd,
2007, 438, 184–189.
182. D. D. Gu and Y. F. Shen: J. Alloy Compd, 2009, 473, 107–115.
183. X. Lin, T. M. Yue, H. O. Yang and W. D. Huang: J. Eng. Mater.
Technol., 2009, 131, 041002.
184. F. Gao and H. M. Wang: Mater. Charact., 2008, 59, 1349–1354.
185. F. Gao and H. M. Wang: Intermetallics, 2008, 16, 202–208.
186. H. P. Qu and H. M. Wang: Mater. Sci. Eng. A, 2007, A466, 187–
194.
187. D. D. Gu, Y. F. Shen, J. L. Yang and Y. Wang: Mater. Sci.
Technol., 2006, 22, 1449–1455.
188. Y. Wang, J. Bergström and C. Burman: Int. J. Fatigue, 2006, 28,
1705–1715.
189. X. J. Xu, X. Lin, M. C. Yang, J. Chen and W. D. Huang: J. Alloys
Compd, 2009, 480, 782–787.
190. X. Lin, T. M. Yue, H. O. Yang and W. D. Huang: Acta Mater.,
2006, 54, 1901–1915.
191. F. Y. Zhang, J. Chen, H. Tan, X. Lin and W. D. Huang: Opt.
Laser Technol., 2009, 41, 601–607.
192. S. Sircar, C. Ribaudo and J. Mazumder: Metall. Trans. A, 1989,
20A, 2267–2277.
193. C. Ribaudo. S. Sircar and J. Mazumder: Metall. Trans. A, 1989,
20A, 2489–2497.
194. Y. Xue and H. M. Wang: Intermetallics, 2009, 17, 89–97.
195. L. X. Dong and H. M. Wang: J. Alloy Compd, 2008, 465, 83–89.
196. Y. Xue and H. M. Wang: J. Alloy Compd, 2008, 464, 138–145.
197. Y. L. Gui and H. M. Wang: Int. J. Refract. Met. Hard Mater.,
2007, 25, 433–439.
198. Y. L. Fang, H. B. Tang and H. M. Wang: Intermetallics, 2006, 14,
750–758.
199. Y. Liu and H. M. Wang: Mater. Sci. Eng. A, 2005, A396, 240–250.
200. J. LLorca: Prog. Mater. Sci., 2002, 47, 283–353.
201. S. C. Tjong: Adv. Eng. Mater., 2007, 9, 639–652.
202. C. Suryanarayana: Prog. Mater. Sci., 2001, 46, 1–184.
203. X. C. Wang, T. Laoui, J. Bonse, J. P. Kruth, B. Lauwers and
L. Froyen: Int. J. Adv. Manuf. Technol., 2002, 19, 351–357.
204. S. Kumar: J. Mater. Process. Technol., 2009, 209, 3840–3848.
205. T. Gläser: ‘Laser sintering of tungsten carbide with cobalt binder
phases’, Fraunhofer IPT annual report, Fraunhofer IPT, Aachen,
Germany, 2005.
206. Y. H. Xiong, J. E. Smugeresky, L. Ajdelsztajn and J. M.
Schoenung: Mater. Sci. Eng. A, 2008, A493, 261–266.
207. J. A. Picas, Y. Xiong, M. Punset, L. Ajdelsztajn, A. Forn and
J. M. Schoenung: Int. J. Refract. Met. Hard Mater., 2009, 27,
344–349.
208. A. Gåård, P. Krakhmalev and J. Bergström: J. Alloys Compd,
2006, 421, 166–171.
209. C. K. Srinivasa, C. S. Ramesh and S. K. Prabhakar: Rapid
Prototyping J., 2010, 16, 258–267.
210. C. S. Ramesh, C. K. Srinivas and B. H. Channabasappa: Wear,
2009, 267, 1777–1783.
Laser additive manufacturing of metallic components
211. S. K. Ghosh, P. Saha and S. Kishore: Mater. Sci. Eng. A, 2010,
A527, 4694–4701.
212. A. Simchi and D. Godlinski: Scr. Mater., 2008, 59, 199–202.
213. D. D. Gu and Y. F. Shen: Appl. Surf. Sci., 2008, 254, 3971–3978.
214. B. E. Stucker, W. L. Bradley, P. T. Eubank, B. Bozkurt,
S. Norasetthekul: ‘Manufacture and use of ZrB2/Cu or TiB2/Cu
composite electrodes’, US patent no. 5933701, 1999.
215. C. N. Sun, T. Baldridge and M. C. Gupta: Mater. Lett., 2009, 63,
2529–2531.
216. Y. Xiong, M. Kim, O. Seo, J. M. Schoenung and S. Kang: Powder
Metall., 2010, 53, 41–46.
217. B. Zheng, T. Topping, J. E. Smugeresky, Y. Zhou, A. Biswas,
D. Baker and E. J. Lavernia: Metall. Mater. Trans. A, 2010, 41A,
568–573.
218. B. Zheng, J. E. Smugeresky, Y. Zhou, D. Baker and E. J.
Lavernia: Metall. Mater. Trans. A, 2008, 39A, 1196–1205.
219. W. P. Liu and J. N. DuPont: Metall. Mater. Trans. A, 2004, 35A,
1133–1140.
220. D. Liu, S. Q. Zhang, A. Li and H. M. Wang: J. Alloys Compd,
2010, 496, 189–195.
221. V. K. Balla, P. D. DeVasConCellos, W. C. Xue, S. Bose and
A. Bandyopadhyay: Acta Biomater., 2009, 5, 1831–1837.
222. J. C. Walker, K. M. Berggreen, A. R. Jones and C. J. Sutcliffe:
Adv. Eng. Mater., 2009, 11, 541–546.
223. D. D. Gu and Y. F. Shen: J. Phys. D: Appl. Phys., 2008, 41,
095308.
224. D. D. Gu and Y. F. Shen: J. Mater. Res., 2009, 24, 3397–3406.
225. S. C. Tjong and Z. Y. Ma: Mater. Sci. Eng. R, 2000, R29, 49–113.
226. J. Choi and J. Mazumder: J. Mater. Sci., 1994, 29, 4460–4476.
227. M. L. Zhong, X. Y. Xu, W. J. Liu and H. Q. Sun: J. Laser Appl.,
2004, 16, 160–166.
228. R. Banerjee, A. Genç, P. C. Collins and H. L. Fraser: Metall.
Mater. Trans. A, 2004, 35A, 2143–2152.
229. R. Banerjee, A. Genç, D. Hill, P. C. Collins and H. L. Fraser: Scr.
Mater., 2005, 53, 1433–1437.
230. F. Wang, J. Mei and X. Wu: J. Mater. Process. Technol., 2008,
195, 321–326.
231. D. D. Gu, Y. F. Shen and G. B. Meng: Mater. Lett., 2009, 63,
2536–2538.
232. D. D. Gu and W. Meiners: Mater. Sci. Eng. A, 2010, A527, 7585–7592.
233. D. D. Gu, Y. F. Shen and Z. J. Lu: Mater. Lett., 2009, 63, 1577–
1579.
234. D. D. Gu, Y. C. Hagedorn, W. Meiners, K. Wissenbach and
R. Poprawe: Surf. Coat. Technol., 2011, 205, 3285–3292.
235. N. K. Tolochko, T. Laoui. Y. V. Khlopkov, S. E. Mozzharov,
V. I. Titov and M. B. Ignatiev: Rapid Prototyping J., 2000, 6, 155–
160.
236. A. V. Gusarov and J. P. Kruth: Int. J. Heat Mass Trans., 2005, 48,
3423–3434.
237. P. Fischer, V. Romano, H. P. Weber, N. P. Karapatis, E. Boillat
and R. Glardon: Acta Mater., 2003, 51, 1651–1662.
238. A. Grabowski, M. Nowak and J. Śleziona: Opt. Laser Technol.,
2005, 43, 233–246.
239. D. D. Gu and Y. F. Shen: Metall. Mater. Trans. B, 2006, 37B,
967–977.
240. I. Takamichi and I. L. G. Roderick: ‘The physical properties of
liquid metals’, 1st edn; 1993, Oxford, Clarendon Press.
241. H. J. Niu and I. T. H. Chang: Scr. Mater., 1999, 41, 1229–1234.
242. D. D. Gu and Y. F. Shen: J. Alloy Compd, 2007, 432, 163–166.
243. D. D. Gu and Y. F. Shen: Mater. Des., 2009, 30, 2903–2910.
244. M. Das, V. K. Balla, D. Basu, S. Bose and A. Bandyopadhyay:
Scr. Mater., 2010, 63, 438–441.
245. W. Hofmeister, M. Griffith, M. Ensz and J. Smugeresky: JOM,
2001, 53, 30–34.
246. A. J. Pinkerton and L. Li: Appl. Surf. Sci., 2003, 208–209, 411–
416.
247. D. D. Gu, W. Meiners, Y. C. Hagedorn, K. Wissenbach and R.
Poprawe: J. Phys. D: Appl. Phys., 2010, 43, 295–402.
248. L. Hao, S. Dadbakhsh, O. Seaman and M. Felstead: J. Mater.
Process. Technol., 2009, 209, 5793–5801.
249. P. Mercelis and J. P. Kruth: Rapid Prototyping J., 2006, 12, 254–265.
250. G. Link, T. Huntley, A. Nickel, R. Leitgeb, T. Nguyen and
F. Prinz: ‘Reducing part deformation by inducing phase
transformation’, Proc. 10th Symp. on ‘Solid freeform fabrication’,
Austin, TX, USA, August 1999, The University of Texas at
Austin, 727–734.
251. Y. C. Hagedorn, J. Wilkes, W. Meiners, K. Wissenbach and
R. Poprawe: Phys. Proced., 2010, 5, 587–594.
252. J. Chen, X. Lin, T. Wang, H. Yang and W. D. Huang: Rare Met.
Mater. Eng., 2003, 32, 183–186.
International Materials Reviews
2012
VOL
57
NO
3
163
Gu et al.
Laser additive manufacturing of metallic components
253. J. Chen, H. O. Yang, Y. M. Li and W. D. Huang: Appl. Laser,
2002, 22, 300–304.
254. S. Kumar and J. P. Kruth: Adv. Eng. Mater., 2008, 10, 750–753.
255. J. Choi, S. K. Choudhuri and J. Mazumder: J. Mater. Sci., 2000,
35, 3213–3219.
256. H. M. Wang, S. Q. Zhang and X. M. Wang: Chin. J. Lasers, 2009,
36, 3204–3209.
257. I. Palčič, M. Balažic, M. Milfelner and B. Buchmeister: Mater.
Manuf. Process, 2009, 24, 750–753.
164
International Materials Reviews
2012
VOL
57
NO
3
258. S. S. Dimov, D. T. Pham, F. Lacan and K. D. Dotchev: Assem.
Autom., 2001, 21, 296–302.
259. B. V. Krishna, W. C. Xue, S. Bose and A. Bandyopadhyay: JOM,
2008, 60, 45–48.
260. K. Dai and L. Shaw: Acta Mater., 2005, 53, 4743–4754.
261. H. Qi, J. Mazumder and H. Ki: J. Appl. Phys., 2006, 100,
024903.
262. R. J. Moat, A. J. Pinkerton, L. Li, P. J. Withers and M. Preuss:
Acta Mater., 2009, 57, 1220–1229.