Optics and Laser Technology 106 (2018) 427–441

Contents lists available at ScienceDirect

Optics and Laser Technology

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

Review

Review on thermal analysis in laser-based additive manufacturing

Zhaorui Yan a, Weiwei Liu a,⇑, Zijue Tang a, Xuyang Liu a, Nan Zhang a, Mingzheng Li a,b,
Hongchao Zhang a,c
a
School of Mechanical Engineering, Dalian University of Technology, Dalian 116024, China
b
School of Engineering, University of Liverpool, Merseyside L69 3GH, UK
c
Department of Industrial Engineering, Texas Tech University, Lubbock, TX 79409, USA

a r t i c l e i n f o a b s t r a c t

Article history: Laser-based Additive Manufacturing (LAM) has been widely used in different industries. However, the
Received 14 November 2017 quality and repeatability of the components and parts produced by LAM have hindered the spread of this
Received in revised form 1 February 2018 technique. The better understanding of the LAM underlying mechanism can provide insight into acquir-
Accepted 25 April 2018
ing high-quality products. Among researches on underlying physics, there are efforts to study the thermal
Available online 10 May 2018
behavior, as one important part of the complex mechanisms, and its influence on the product quality. This
review is presented to comprehensively analyze different approaches to study the thermal behavior. The
Keywords:
relationship between thermal behavior and product quality is identified and some recommendations for
Additive manufacturing
In-situ monitoring
future research are discussed.
Thermal model Ó 2018 Elsevier Ltd. All rights reserved.
Thermal behavior
Product quality
Numerical simulation

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
2. In-situ temperature monitoring sensors available in LAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
2.1. Contact measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
2.2. Contactless measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
3. Thermal models in LAM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
3.1. Melt pool behaviors in LAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
3.2. Empirical models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
3.3. Analytical models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
3.4. Numerical models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
3.4.1. Governing equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
3.4.2. Heat source models and boundary conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
3.4.3. Convection and solidification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
3.4.4. Temperature-dependent material properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
3.4.5. Computing cost. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
3.4.6. Geometry and material additive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
4. Parameter-thermal behavior-quality relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
4.1. Parameters-thermal behavior relationship in LAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
4.2. Residual stress and deformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
4.3. Solidification and microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
4.4. Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
4.5. Process temperature control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
5. Future research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

⇑ Corresponding author.
E-mail address: liuww@dlut.edu.cn (W. Liu).

https://doi.org/10.1016/j.optlastec.2018.04.034
0030-3992/Ó 2018 Elsevier Ltd. All rights reserved.
428
6. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
1. Introduction
Laser-based Additive Manufacturing (LAM) is gaining rapid pro-
gress in recent years and widely adopted in different industries
such as tool-making [1], aviation and aerospace [2] and biomedical
[3]. As the laser beam irradiates the surface of the substrate, the
metal substrate fuses and liquid melt pool forms. A product is pro-
duced by joining mental material layer by layer from a 3-D model.
Compared to the conventional Metallurgy and Subtractive Manu-
facturing, LAM has a number of advantages:
(1) The high energy density and small heat affected zone lead to
a high cooling rate in the melt pool. This contributes to finer
grain size and better mechanical properties of products [4].
(2) Due to high temperature in the melt pool, it is feasible to
process difficult-to-machine or refractory material [5].
(3) By changing the chemical composition of the input metal
material, LAM can be used to process functionally graded
components [6].
(4) Since the material is deposited layer by layer, the compo-
nents with complex structure, which is usually difficult to
produce by traditional processing methods, can be produced
by LAM [5].
Laser-based Powder Bed Fusion (PBF) and laser-based Direct
Energy Deposition (DED) are two categories in LAM. Laser-based
PBF uses the laser as a heat source to selectively fuse powder
which is preplaced layer by layer. Among various PBF systems,
Selective Laser Sintering (SLS), first created at the University of
Texas [7], is a technology that fuses the substrate and binds pow-
der together to make a solid product. Compared to SLS, Selective
Laser Melting (SLM) uses a comparatively high-power laser to fully
fuse the powders and a fully dense product can be produced [8].
Laser-based DED process uses a laser to fuse metal powders, which
are injected into the melt pool by a feeder. In different literature
and systems, laser-based DED is also called Laser Metal Deposition
(LMD) [9], Laser Engineered Net Shaping (LENS) [4] and Laser Clad-
ding (LC). Laser Cladding is successfully used in repairing machine
part and improving corrosion, erosion and wear resistance [10].
However, the quality of the product cannot fully meet the
requirement of the industrial application and hamper further
development of LAM. The quality issues such as dimensional and
form errors, unexpected mechanical property and internal defects
are still challenging due to a serial of reasons: There is need to opti-
mize the process parameters for the product quality to meet the
requirements, but the underlying physics process is not completely
clear at present. This physics process involves heat and mass trans-
fer, convection of liquid metal and non-equilibrium solidification
due to high heating and cooling rate and temperature gradient.
Insufficient understanding of the formation mechanisms of defects
results in higher defect ratios with respect to conventional produc-
tion systems [11]. The material experience a serial of the phase
transition processes, that is solid-liquid-solid and the solid phase
transformation process [12]. The microstructure evolution influ-
ences the final mechanical properties of the product [13]. Thermal
expansion and cooling of material cause thermal stress, which,
leads to distortion and cracking of product. The issue of prevention
and control of thermal stress is especially necessary for producing
large parts [14]. In the underlying physics, thermal behavior, as a
Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
key factor governing the transformation of microstructure, evolu-
tion of stress and pattern of fluid convection, has a significant influ-
ence on acquiring high-quality products.
To solve quality issues, many high-level literature reviews
about online monitoring and control have been provided in recent
years [15,16]. Schoinochoritis et al. [17] summarized the numerical
models in powder bed additive manufacturing and concluded that
the temperature field was what most studies target at. These pub-
lished works give an overview of the monitoring and control meth-
ods or numerical models, but there is few published review
summarizing the investigation of thermal behavior and relation-
ship between temperature distribution and the product quality.
In this paper, we focus on the thermal behavior of LAM and con-
sider it as the process signature that links the process parameters
and the product quality. This review is organized as follows. In Sec-
tions 2 and 3, the temperature monitoring methods and mathe-
matical models of thermal analysis are discussed respectively.
The parameters-thermal behavior-quality relationships are illus-
trated in Section 4. At last, the conclusion of this review is reached
and some future research opportunities are proposed.
2. In-situ temperature monitoring sensors available in LAM
The accuracy and reliability of the temperature process signa-
ture should meet the requirement of providing insight into the
quality of the product. However, the severe monitoring condition
imposes restrictions on the possible temperature sensors.
Although the LAM product demands dozens of hours to be com-
pleted, the process is extremely fast at the scale of the laser process
zone [18]. The size of melt pool is rather small compared to that of
the product. Due to these restrictions, high speed of data acquisi-
tion and high resolution are needed for sensors to capture signa-
tures of the melt pool. Researchers have explored new forms of
temperature measurement technique and appropriate methods
applied in LAM process, including the contact measurement and
contactless measurement.
2.1. Contact measurement
Thermocouples are sensors that carry out contact measurement
of temperature by fixing the one end of the sensor on the metal.
Thermocouples consist of two metal wires of various materials
with one end connected to form a closed circuit, in which potential
difference is detected and converted into temperature. Although
there are a number of advantages such as low cost, high accuracy
and convenient calibration, thermocouples have limited use in
LAM because of some weaknesses. First, the temperature distribu-
tion is disturbed by the initial temperature of the thermocouples
for the direct contact between the sensors and the part. Second,
the thermocouples are fixed on the part by welding. As a result,
they are liable to interfere with the laser spot and fail to trace
the movement of the melt pool with the moving laser head.
The major usage of thermocouples is in-situ monitoring of sub-
strate temperature. Griffith et al. [19] first monitored the temper-
ature by implanting a type C thermocouples into one-line width
wide shell boxes during fabrication. Ya et al. [20] inserted four
thermocouples into the substrate at different depths below the
top surface of the substrate along the cladding direction and the
other four perpendiculars to the cladding direction. Zhang et al.
 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
[21] arranged the thermocouples in diagonal to in-situ monitor the
thermal signature. Segerstark et al. [22] set up protective sheets to
shield the thermocouples from the laser radiation not absorbed by
the substrate. Results show that the protective sheets have limited
impact on the accuracy of thermocouples and effectively protect
the thermocouples from damaging. Among various kinds of ther-
mocouples, type K thermocouples are preferred for the lower cost and
type S is another choice for excellent comprehensive properties.
One exploration of new contact measurement method was per-
formed by integrating fiber Fabry–Perot cavity sensor into the
component. The sensor consists of an optical fiber end and a reflec-
tive splice to form the optical cavity. The feasibility of temperature
measurement with the range up to 1000 °C was conducted exper-
imentally while two embedded technology was discussed [23].
2.2. Contactless measurement
The IR temperature sensors are regarded as the contactless
measurement based on the principle that the infrared radiation
of an object varies with its temperature. A blackbody’s radiation,
at certain wavelength k and absolute temperature TA, is calculated
by the Plank’s law:
C1
LðT A ; kÞ ¼
k5 ðeC 2 =ðkT A Þ  1Þ
where L is the radiation of blackbody, C1 and C2 are the first and the
second Plank constant, respectively. The radiation of an object can
be expressed as the product of ideal blackbody’s radiation and the
spectral-directional emissivity:
Lr ðT A ; k; h; /Þ ¼ eðT A ; k; h; /ÞLðT A ; kÞ
C1
¼ eðT A ; k; h; /Þ
k5 ðeC2 =ðkT A Þ  1Þ
Temperature can be restored based on the IR data acquired by
sensors, and different methods have been used including single-
wavelength method, dual-wavelength method and multi-
wavelength method. The principle of the dual-wavelength method
is gray body approximation, that is to say, the two emissivities in
two close wavelengths are equal. The multi-wavelength method
is a combination of spectrum radiation and algorithms to restore
the true temperature. Typical IR-thermometry sensors in published
literature mainly included a photodiode, pyrometer, IR-camera and
spectrophotometer.
Compared with thermocouples, the IR sensors do not need to
contact the object directly, and hence, not interfere with the tem-
perature distribution. Short response time and the ability to mon-
itor the moving melt pool make IR thermometry widely adopted.
One known challenge of IR thermometry is the unknown emissiv-
ity during the process. The emissivity, which was introduced in for-
mula (1.2), varies with different temperature, wavelengths and the
surface characteristic. The majority of models of emissivity are
empirical models on the basis of the experimental data, which lack
universality and can only be applied in particular processing situ-
ation. Moreover, the emissivity data of liquid metal is rare due to
the restriction of sensors. Several methods have been used to
determine the emissivity and restore the absolute temperature.
(1) Emissivity can be set at a fixed value. Liu et al. [24] used the
value of 0.3 as the emissivity value for the infrared camera. Rodri-
guez et al. [25] fabricated an artificial blackbody as a reference and
calculated the average emissivity. Doubenskaia et al. [26] defined
the emissivity value under the assumption that the brightness
temperature remaining stable at liquid-solid transition corre-
sponds to the known melting point of TiAl6V4. These studies using
fixed emissivity can neither acquire real-time emissivity nor mod-
ify the temperature value in situ. (2) Multi-wavelength method has
429
Fig. 1. Experimental setup for optical diagnostics of DMD process [41].
been used to store the emissivity, and it has been developed in
models [27,28] and emissivity retrieve algorithm [29,30]. Hagqvist
et al. [31,32] designed an emissivity compensated spectral pyrom-
eter to retrieve true temperature and emissivity simultaneously for
ð1:1Þ the metallic thermometry with varying emissivity. The key step of
this algorithm was to minimize the weighted sum of absolute dif-
ferences between the theoretical and the reconstructed spectra. (3)
True temperature can be in-situ restored with different sensors in
addition to the spectrophotometer. Köhler et al. [33] designed an IR
temperature monitoring system based on spatial emissivity com-
pensation using a CMOS camera. Zhirnov et al. [34] retrieved the true
temperature with an IR camera in SLM process under an assumption
that the temperature dependence of emissivity was not considered.
ð1:2Þ In some studies, the single sensor was used in in-situ tempera-
ture monitoring. Carcel et al. [35] measured the surface tempera-
ture with a dual-wavelength pyrometer during processing thin
wall. Results showed that, when depositing LMD layers, the
microstructure and geometry were influenced by the temperature
and the cooling rate. Similarly, a dual-wavelength pyrometer was
put into use in [36]. Lei et al. [37] monitored the size and the shape
variation of melt pool based on a CCD-based detector, and obtained
the temperature field of the melt pool by the dual-wavelength
method. Doubenskaia et al. [26] monitored the brightness temper-
ature when depositing TiAl6V4 with IR thermography and the true
temperature was restored. The center temperature evolution illus-
trated different stages of melt pool: cooling of the liquid phase,
liquid-solid phase transition and cooling of the solid phase. IR ther-
mography was also widely used in PBF process [38,39]. And Köhler
et al. [33] monitored the temperature distribution with a CMOS
camera and a dual-wavelength method was applied. Some sensors
based on multi-wavelength method were used in recent studies. De
Baere et al. [9] monitored the spectral radiance of seven locations
along the deposition direction in the melt pool. The least square
method and double-wavelengths methods were applied to estimate
the temperature of different positions. Combining the 2D tempera-
ture distribution and the multi-wavelength method, a hyperspec-
tral line camera was applied to estimate the true temperature
distribution, the emissivity of the melt pool and their error [40].
Multi-sensor system, applying diverse and complementary
monitoring tools, is valuable to a comprehensive analysis of tem-
perature during processing. Bi et al. [41] set up a monitoring sys-
tem including thermocouples, a photodiode and a CCD camera.
First, the temperature dependence on the process parameters
was investigated [10] and then, it was extended to the process
parameters – thermal signature – geometry and microstructure
relationship [42]. Different sensors were integrated into the clad-
ding head for comprehensive monitoring [41]. Smurov et al. [43]
430 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
Fig. 2. Thermographic images captured by pyrometer (a) and IR camera (b) [46,47].
Table 1
In-situ temperature monitoring techniques and sensors in different publication.
Sensor Dimension
Thermocouples Spot
Monochromatic pyrometer Spot
Dual-wavelength pyrometer
Spectrometer or Multi-wavelength pyrometer Spot
2D pyrometer Line
Hyperspectral line camera Line
IR-camera Plane
CCD Plane
CMOS Plane
Photodiode Spot
comprehensively analyzed the deposition process by a monochro-
matic pyrometer, a multi-wavelength pyrometer and an IR-
camera as shown in Fig. 1. Diverse temperature data fusion were
presented to synthetically analyze the parameters dependence
[44] and the additive material influence [45]. In the literature
[39], a similar system was applied in PBF, which focused on the
mechanism of the heat and mass transfer of pre-deposited inter-
metallic TiAl powder with optical diagnostics. Liu et al. [24] per-
formed optical diagnostics on high power direct diode laser
cladding with a system comprised of a high-speed CCD camera,
a pyrometer, and an infrared camera. The variation of the bright-
ness temperature of the molten pool was measured by the
pyrometer and the temperature distribution, molten pool size,
and cooling rate at different processing parameters was studied
by using the infrared camera. Another monitoring system was
present in the literature [46,47], and Marshall et al. built a thin
wall and a rod respectively. A dual-wavelength pyrometer was
mounted coaxially to monitor the melt pool temperature and an
IR-camera was tilted and viewed the whole product’s temperature
distribution as illustrated in Fig. 2. Lane et al. [18] fabricated a
nickel alloy 625 part conducted on a commercial LPBF machine,
which was incorporated with a thermal camera, a high-speed vis-
ible camera and a photodiode. The photodetector signal and cam-
era images were merged simultaneously to have a better
understanding of the melt pool behaviors. For a generalization
of the studies surveyed above, monitoring techniques and sensors
are summarized in Table 1.
3. Thermal models in LAM
There has been a continuous effort in developing various ther-
mal models, including empirical, analytical and numerical models.
Principle Publications
Thermoelectric effect [19,20,41,21,22,48,49,50,24]
Single wavelengths/Dual wavelengths [35,43,24,46,47,36]
Multi-wavelength [45,44,40,39]
Single wavelength [43,44]
Multi wavelengths [40,51]
Single wavelength [52,26,45,44,53,24,54,38,55,47,47,18,39]
Dual wavelengths [19,41,10,37,43]
Dual wavelengths [33]
Photoelectric effect [10,42,41,18]
Physics-based analytical and numerical models are crucial to pre-
dicting the outcome of LAM process that accounts for the changes
in material properties [16], and the empirical models are mostly
used in control systems. In this section, the melt pool behaviors
are first introduced, and then, the empirical, analytical and numer-
ical models are discussed in details.
3.1. Melt pool behaviors in LAM
The substrate is irradiated and heated by laser and conduction
dominate the heat transfer process. When the temperature of the
substrate is locally higher than the melting point, the liquid melt
pool generates. High temperature gradient occurs at the surface
of the melt pool, and the resulting ‘‘Marangoni force” drives the liq-
uid metal flow. As the melt pool moves with the laser spot, melting
and solidification processes evolve continuously, so the latent heat
of phase change, heat transfer and flow behaviors in the binary
solid-liquid phase region have to be taken into account. Heat trans-
fers from the product to the surrounding by convection and radia-
tion. The materials properties (including thermo-physical
properties and mechanical properties) depend on temperature.
The interaction of laser, powder and melt pool including the atten-
uation effect is not completely clear today as shown in Fig. 3. Based
on the heat transfer, mass transfer and momentum transfer of melt
pool, the thermal models can be developed to reveal key informa-
tion inside the melt pool that in-situ monitoring is unable to do it.
3.2. Empirical models
Empirical models are developed experimentally to estimate the
response between process parameter and melt pool temperature.
Empirical models with transient response relating the laser power
to the temperature were presented in [57]. Tang et al. [58] mod-
 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441 431

post-processing. Details of numerical models are summarized in

Table 2.

3.4.1. Governing equation

The numerical model is developed on energy conservation,
momentum conservation and mass conservation. In heat conduc-
tion model, only energy conservation is considered and the heat
conduction equation is expressed as follow:
     
@T @ @T @ @T @ @T
qC p ¼ k þ k þ k þ q_ ð3:1Þ
@ s @x @x @y @y @z @z

where q, k and Cp represent the density, thermal conductivity and

Fig. 3. Schematic overview of the interaction zone between laser radiation and
powder bed [56].
eled the melt pool temperature using powder flow rate, laser
power, and traverse speed. The time constant of the models was
identified by experiments data and the least-squares method. Song
et al. [59,60] developed a dynamic empirical model experimentally
employing the subspace method. The pulse durations and laser
powers were regulated randomly and a fourth-order state space
model was applied to identify this experiment results to acquire the
relationship between the laser power and the melt pool temperature.
3.3. Analytical models
The well-known ‘‘Rosenthal solution” showed the temperature
distribution for a moving point heat source [61]. The dimensionless
solution [62] of a traveling Gaussian heat distribution was given by
Eagar and Tsi in 1983. Vasinonta et al. [63] developed a process
map based on the 2D ‘‘Rosenthal solution” [61]. Other process
maps are given in the literature [63,64]. Gockel et al. [65] proposed
a closed-form solution model to modify the ‘‘Rosenthal solution”
considering the effect of free edges. Lumped-parameter models
were put forward to predict the melt pool temperature and geom-
etry for the control system. Doumanidis and Kwak [66] proposed a
lumped-parameter model for a moving and concentrated source,
which compensated measurement delays in a closed-loop control
system. Following this line, some Multi-Input Multi-Output
(MIMO) models were developed. Sammons et al. [67] introduced
a solidification term and developed a layer dependent model of
the LMD process in order to design layer-to-layer control system.
Wang et al. [68] proposed to approximate and parameterize the
entire material transfer rate lf altogether based on Kwak’s model
[66]. Cao and Ayalew [69] proposed a MIMO coupled nonlinear
process model which was expressed in Hammerstein form. Models
in this line are semi-empirical, which means some parameters in
the models have to be designed by experiments.
In above models, modeling process involves certain assumption
and simplification, so some of the details are neglected for the fea-
sibility of the solution. Numerical models are preferred for more
detailed understanding of complicated physics.
3.4. Numerical models
specific heat of the material, respectively. q_ is the heat source term
and T is temperature. Because the heat conduction equation is easy
to converge, it has been widely adopted [14,56,70]. If taking convec-
tion of the liquid metal into account, the governing equation [71] is
expressed as:
 !
@q
þ r q V ¼ Ms
@t
 
@V ! ! ! !
q þ V r V ¼ lr2 V rp þ M s  V þF ð3:2Þ
@t
 
@T !
q þ V rT ¼ rðkrT Þ þ SH
@t
!
where l is the viscosity of the materials. Ms is a mass source, V is
the motion velocity of the melt, and F is the body force (e.g., gravity
and buoyancy forces). SH is the source item of energy equation and
can be defined by:
 ! 
@
SH ¼ q DH þ r  V H ð3:3Þ
@t
Eq. (3.2) was used in the study of transient heat transfer of fluid
flow, solidification and interface geometry [71–73]. In 2D models,
governing equation can be simplified [20,74].
3.4.2. Heat source models and boundary conditions
LAM and fusion welding share similar heat source models,
mainly including Gaussian surface heat source [72,73,75], top-hat
distribution [49,70,76] and Goldak’s double ellipsoid heat source
[77–79]. The heat source was implemented in either the governing
equations as the source term [80] or in the model as boundary con-
dition [73].
In DED process, the attenuation effect of the laser was given by
Leer-Lambert law [72]. Manvatkar et al. [80] used volumetric
Gaussian heat source model to simulate the energy attenuation
by particles in flight. Tabernero et al. [81] modeled the energy
attenuation of the laser and powder interaction process consider-
ing the particle size and the beam divergence. In LBF process, the
laser energy decreased as the beam penetrates deep into the pow-
der bed, which depended on the material absorbance, porosity and
particle diameter [71,76]. Khairallah et al. [82] put forward that
powder particles were inhomogeneously heated by the laser beam
and melting was non-uniform. Otherwise, other boundary condi-
tions include convective and radiative heat transfer [56]. Gouge
et al. [79] proposed that melt pool was exposed to the shielding
gas and the forced convection instead of natural convection should
be considered.

Numerical models are more economical and available to pre-
sent complex analysis inside the melt pool, compared with the
experimental investigation. The main procedures of numerically
modeling include pre-processing (geometry modeling, boundary
conditions, initial conditions, heat loads and material properties),
processing (choosing proper governing equation and solving) and
3.4.3. Convection and solidification
The convection pattern has a strong impact on the heat transfer
and temperature distribution in the liquid metal. Manvatkar et al.
[80] showed that by ignoring the effect of convection, the cooling
rates in additive manufacturing were over-estimated by about
twice of the correct values. This result revealed that convection
432 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441

Table 2
Summary of numerical models.

Category Material Dimension Geometry Year Software Content Publications

DMD TA6V 3 Thin wall 2008 COMSOL Temperature [90]
LM TiAl6V4 3 3D part 2010 ANSYS Temperature and [105]
stress
DED Chromium powder 3 Single track 2011 COMSOL Temperature and [83]
AISI/SAE 1020 steel plate fluid flow
LPD IN625 powder and superalloy plate 3 Multi-tracks 2011 ANSYS Temperature and [48]
stress
DED TiAl6V4 2 Single track 2012 COMSOL Temperature and [74]
fluid flow
DED NiCoCrAlY powder and RT300A plate 3 Single track 2013 ANSYS Temperature and [97]
geometry
DED TiAl6V4 3 Large product 2014 CUBIC Temperature and [14]
stress
LC ASTM A36 plate and H13 powder 3 Single track 2014 ANSYS Temperature and [49]
stress
SLM 316 stainless steel 3 Mesoscopic 2014 ALE3D Fluid flow and [8]
model multiphysics temperature
code
SLM AlSi10Mg powder 3 Multi tracks 2014 ANSYS Temperature [102]
and multi
layers
DED SS316 3 Thin wall 2014 N Temperature and [80]
fluid flow
AM Ti64 1 or 3 Thin wall 2014 CUBIC Temperature [88]
DED InconelÓ 625 3 Thin wall 2015 N Temperature [79]
SLM TiC reinforced AlSi10Mg 3 Single track 2015 FLUENT Temperature and [71]
fluid flow
LC 304 powder and 45# steel substrate 3 Single track 2015 ANSYS Temperature [75]
PBF TiAl6V4 3 3D complex 2015 COMSOL Temperature [94]
parts
PBF InconelÓ 718 3 Cylinder 2016 N Temperature and [78]
deformation
DED InconelÓ 718 3 Thin wall 2016 FORTRAN Temperature and [98]
fluid flow
DED TiAl6V4 3 Single track 2016 N Temperature [77]
LHW H13 3 3D part 2016 ABQUS Temperature and [106]
stress
PBF InconelÓ 718 3 Single track 2016 ANSYS Temperature [56]
DED DIN 2393 steel 3 Multi tracks 2016 COMSOL Temperature and [20]
geometry
DED Ti48Al2Cr2Nb 3 Thin wall 2016 ANSYS Temperature and [117]
microstructure
PBF chromium molybdenum base and maraging-based powder 2 Thin wall 2017 ANSYS Temperature and [107]
and carbon steel and stainless steel substrate stress
DED 38MnVS carbon steel substrate and Co–Cr–W powder 3 Single track 2017 COMSOL Temperature and [73]
mass transfer
DED InconelÓ 718 3 Single track 2017 FORTRAN Temperature and [72]
fluid flow
PBF SS316, TiAl6V4 and IN718 3 Single track 2017 N Temperature, fluid [91]
flow, and stress
PBF TiAl6V4 and IN718 3 Single track 2017 ABAQUS Temperature and [84]
stress
DMLS TiAl6V4 3 Single track 2017 COMSOL Temperature and [89]
stress
LC AISI D2, InconelÓ 718 3 2012 Matlab Temperature [81]
attenuation
PBF 316L 3 Single track 2016 ALE3D Fluid flow and [82]
temperature
SLM AlSi10Mg 3 Single track 2016 N Temperature, flow [86]
and morphology
SLM InconelÓ 718 3 Single track 2017 N Porosity [87]
mechanism
SLM Iron-based powder 3 Multi tracks 2016 ABAQUS Temperature and [95]
DirectSteelÒ Powder stress
DED Martensitic stainless steel 3 Multi tracks 2015 MARC Stress evolution [108]

should not be neglected in numerical models. The liquid metal is
driven by ‘‘Marangoni effect” and under the influence of gravity,
buoyancy and impact force from the shielding gas and vaporized
alloy, which makes the flow pattern complex. The N-S equation
has poor convergence, so the simplified method of increasing the
heat conductivity in heat conduction equation has been widely
adopted to estimate the effect of fluid flow [20,69,70]. Vásquez
et al. [83] developed a 3D quasi-stationary multi-physics model
considering the ‘Marangoni effect’ and the quality of vapor mass
escaping from the pool, which was calculated by the analytical
form of Stefan’s condition under a quasi-steady state. For transient
heat transfer and fluid flow model, the fluid flow is acquired by
generally solving the N-S equation. Morville et al. [74] simulated
the shape of melt pool surface by moving meshing in a transient
 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441 433

Fig. 4. Computed velocity magnitude at different times: (a) 10 ms, (b) 300 ms, (c) 500 ms, and (d) 700 ms [73].

Fig. 5. Calculated porosity of mesoscopic scale model [87].

434 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
2D longitudinal thermal model and the main physics phenomena
such as surface tension effect and addition of powder were taken
into account. A 3D moving meshing model was developed by
Gan et al. [73], and thermal behavior, fluid flow, melting–solidifica-
tion and multicomponent mass transfer were investigated. Gu and
Yuan [71] developed a transient three-dimensional CFD model to
investigate the influence of the TiC reinforcement weight fraction
on heat transfer and fluid flow during SLM. The computed velocity
at different time was shown in Fig. 4. The influence of convection
was also considered [72,84]. Another factor that affects tempera-
ture distribution is latent heat released during solidification. There
were mainly two ways to estimate the latent heat. One method
was to calculate enthalpy change according to the definition of
latent heat of phase change [72,73] and the other one was to
increase the specific heat [20,49,83]. The latter one method
required very precise control of time stepping to improve the sim-
ulation of phase change [85].
Models at the mesoscopic scale were widely used in under-
standing the underlying physics and the formation mechanism of
defects in SLM process. Khairallah and Anderson [8,82] proposed
a 3D mesoscopic model by using ALE3D simulation software. The
melt flow of liquid 316L was analyzed and the formation pores,
spatter, and denudation zones were further explored. Yu et al.
[86] developed an FVM model, which considered the powder-to-
solid transition to analyze the thermal behavior, the effect of laser
power on the surface morphology and balling phenomenon. Xia
et al. [87] studied the evolution and mechanism of porosity using
a transient mesoscale model with randomly-packed powder bed.
Results showed that at high scanning speed, a short lifespan was
produced and velocity and depth of the melt pool were evidently
reduced. The porosity transformed from metallurgical porosity to
open porosity as the scanning speed increased as shown in Fig. 5.
3.4.4. Temperature-dependent material properties
The materials properties undergo nonlinear dependence on the
temperature change and phase change. A suitable definition of
materials properties is necessary to an accurate model. The
temperature-dependence heat physics properties include density,
conductivity and heat capacity. Because temperature has little
influence on the density of solid metal and liquid metal is always
assumed to be incompressible, density was frequently defined to
be constant [49,80,88,89], while sometimes it was defined as tem-
perature dependence [74,85,90]. Heat conductivity and specific
heat are two physics properties that affect the temperature distri-
bution. The thermal conductivity of a material determines the rate
Fig. 6. Multi-scale methodology for fast prediction of part distortion and residual
stress [95].
at which heat transfers through the material and the amount of the
energy needed for a certain temperature rise can be found using
the specific heat capacity [17]. Most literature indicated that heat
conductivity and specific heat were defined to be temperature
dependence [14,56,90,91]. Gu et al. [71] considered in PBF, the
effective thermal conductivity of loose metallic powder was con-
trolled by gas in pores and was about 100 times lower than that
of the dense material, and the effective thermal conductivity was
restored to actual value when the temperature was higher than
the melting point. The physical properties in the mushy zone dur-
ing solidification were defined to be isotropy and the weighted
average of that of solid and liquid phase [69,83]. In thermomechan-
ical models, physics properties such as elastic coefficient, yield
strength and coefficient of thermal expansion [48,49,84] were tem-
perature dependence.
3.4.5. Computing cost
The high computing cost of thermal model prevents it from fur-
ther adoption in industrial application, especially the simulation of
large products. Bringing down the computing time without sacri-
ficing accuracy would be necessary for numerical models of large
products. Denlinger et al. [14] used the adaptive meshing, which
meant merging the elements layer by layer and coarsening the
mesh of layers to reduce the number of nodes. The mesh coarsen
algorithm was also applied in [92,93]. Another approach was pro-
posed by Kundakcioglu et al. [94] Scanning route was broken into
laser position points according to the laser beam diameter, and
then a file containing 4D data (time and three position coordinates)
was formed. This file was transferred to Gaussian heat model and a
discrete heat source was simulated instead of a traditionally con-
tinuous model. Li et al. [95] developed a multiscale model to sim-
ulate the residual stress and distortion during SLM, as well as to
reduce the computing time. The model consisted of microscale
laser scan model, mesoscale layer hatch model and macroscale part
model as shown in Fig. 6.
3.4.6. Geometry and material additive
Geometry model can be defined based on the fitting of the
experiment results [96] or assumed to be known geometry
[14,56] in heat conduction models. To simulate the addition of
material, a number of studies used the method of ‘‘element birth
and death”. This method consists of the quiet element method
[49,75,79] and the inactive element method [14,88]. In quiet ele-
ment method, the elements are considered during the analysis
with low heat conductivity and specific heat value. In the inactive
element method, the elements are not contained until they are
activated by the certain condition. Michaleris [88] compared these
two methods and proposed a hybrid quiet-inactive metal deposi-
tion method to slightly accelerate computer run times. Zhang
et al. [70] used the elements stacking method to simulate the pro-
cess of adding powder when the geometry of the track was
predetermined.
The geometry can also be predicted in some models. Liu et al.
[97] calculated the melt pool size and simulated the geometry by
activating the elements, the number of which was calculated by
mass conservation. Lee and Farson [98] used the volume of fluid
method to predict the shape in multilayer single track laser addi-
tive manufacturing deposition. Another method adopted by Li
et al. [72] was the level set method to capture the liquid/gas free
interface. This method transformed an interface tracking problem
into a partial differential equation, making its implementation rel-
atively easier than the volume of fluid method [99]. In addition, the
dynamic shape of the liquid/gas interface can be also simulated by
moving mesh based on arbitrary Lagrangian-Eulerian method. Two
kinds of velocities were considered at liquid/gas interface – the
fluid flow velocity and boundary moving velocity due to mass
 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
addition. 2D model and 3D model using this method were pre-
sented in [74] and [73] respectively.
4. Parameter-thermal behavior-quality relationships
In this section, first we compared results between the in-situ
monitoring and the thermal modeling, and discussed how process
parameters influence dynamic melt pool, temperature distribution
and temperature-dependent variables. Then, the thermal behavior
was further extended to analyze the product quality. The product
quality is not directly related to temperature, but has dependence
on the residual stress, deformation, microstructure and defects of
the product, so the effect of temperature on these issues was
studied.
4.1. Parameters-thermal behavior relationship in LAM
It has been extensively reported that process parameters, such
as laser power, scan speed and powder feeding rate, have different
influence on the product quality. The temperature distribution
comparison between monitoring [44] and modeling [56] is pre-
sented in Fig. 7. Temperature showed a similarly increasing trend
with higher laser power and slower scan speed, though different
parameter values were chosen for laser cladding and selective laser
melting. Similar trends were reported in [75,80]. Muvvala et al.
[36] monitored the thermo-cycle of LBF and found that tempera-
ture dropped when the scan speed was lower than 800 mm/min,
because evaporation from the molten pool surface become signifi-
cant. The temperature distribution of the melt pool in the thermal
Fig. 7. Temperature distribution of model (a), (b) [56] and monitoring results (c), (d) [44].
435
image [26] and thermal model [71] are presented in Fig. 8. The high
temperature area was comet-like shape with the sharp gradient at
the head of the melt pool and low gradient at the tail of the melt
pool where the metal cooled down and solidified. It was also found
that monotonous increase of melt pool size (length, width and
depth) with laser power and scan speed was obvious [44,56]. Melt
pool underwent contraction and expansion during the period of
pulsed-wave laser and the temperature fluctuated with the same
frequency of the pulse laser compared with the melt pool in con-
tinuous wave [72].
Temperature gradient and cooling rate, deduced from tempera-
ture distribution or thermo-cycle, are crucial to the end quality of
the product. The cooling rate decreased with higher laser power
and increased with larger scan speed [100], which was reported
in monitoring data [36] and model results [80]. However, different
trend was illustrated in [56] that the cooling rate increased with
both the laser power and the scan speed. The reason for this differ-
ence may be that an overall enhanced conductivity method was
adopted to consider the convection, which was unable to simulate
the local fluid flow and its impact on the local temperature of the
tailing zoon and bottom of the melt pool [56]. Li et al. [72] reported
that cooling rates along the solidification front using pulse laser
were almost twice those using continue laser. Beuth and Klingbeil
[101] first drew a process map for temperature gradient analyti-
cally. Li et al. [102] developed a numerical model and found that
the maximum temperature gradient was in the front melt pool in
which the maximum laser intensity located. In the depth direction,
the temperature gradients decreased with the larger distance from
the surface of the melt pool [70,102], and the temperature gradients
436 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
Fig. 8. Temperature distribution of melt pool of model (a) [71] and monitoring results (b) [26].
perpendicular to the scan direction decreased as the powder feed-
ing rate slowed down because of the interaction between laser
beam and particles [43].
Since parameters influence the temperature synthetically, it is
significant to combine the parameters and studies the comprehen-
sive effect. The effective energy and the powder deposition density
was utilized by Farshidianfar et al. [100] to study the parameter
dependence of cooling rate. Muvvala et al. [36] used the line energy
to investigate the energy distribution. Mukherjee et al. [103] stud-
ied the role of non-dimensional numbers, including non-
dimensional heat input, Peclet numbers, Marangoni numbers and
Fourier number in processing. Islam et al. [104] chose volumetric
energy input to describe the amount of energy input of SLM
process.
4.2. Residual stress and deformation
The rapid thermal cycling causes high temperature gradient
near the melt pool and which, in turn, leads to undesirable defor-
mation and dimensional distortion in the product. Residual stress
resulting from inhomogeneous thermal strain is generated in the
product. The presence of high tensile stress in the product has a
crucial effect on crack sensitivity, decreasing the component life
expectancy and leaving the fabricated parts prone to premature
failure [49].
Deformation can be measured either in situ or after processing.
Roberts et al. [105] measured the deformation using a Confocal
Laser Scanning Microscope after processing. The cases of in-situ
monitoring the deformation were limited. Denlinger et al. [50]
employed a laser displacement sensor to monitor the distortion
of the free end of the substrate when the other end was clamped.
The temperature data was recorded with thermocouples located
on the substrate bottom. A thin wall was processed and a new
parameter called the dwell time was defined to study its influence
on the distortion and the residual stress. A similar method with
thermocouples and displacement sensors were also adopted by
Nie et al. [106] in laser hot-wire additive manufacturing. Furumoto
et al. [107] processed a thin wall of different material on the stain-
less steel or carbon steel substrate and a strain gauge was used to
measure the stress history of the substrate. The resulting stress and
stress history were compared when different material of powder
and substrates were used.
Numerical models can be used to predict the residual stress and
deformation. The temperature distribution in the thermal models
was applied as load on the nodes and stress and displacement
fields are simulated. Mukherjee et al. [84] developed a thermal-
fluid coupled numerical model, and then a thermo-mechanical
model was presented with ABQUS. Results showed the influence
of layer thickness and heat input on the residual stress and defor-
mation. Furumoto et al. [107] reported that the deformation of the
component was related to the coefficient of thermal expansion,
mechanical strength and temperature gradient, and among which,
the coefficient of thermal expansion was the decisive factor. Farah-
mand et al. [49] developed a thermo-elastic-plastic model and
found that there was high concentrated stress in the last tracks
because of the large cooling rate and the stress release of the pre-
vious tracks. This characteristic of stress distribution was also
reported in the literature [107]. Fang et al. [108] investigated the
effect of martensitic transformation on the stress evolution. The
model revealed the influence of material properties, temperature
induced plasticity and the phase transformation temperature on
the stress field was considered. Zhang et al. [48] developed a ther-
momechanical model of multi-bead pulsed laser metal deposition.
The maximum temperature gradient located in the interface
between depositions and substrate and which, the thermal stresses
concentrated on.
4.3. Solidification and microstructure
The cooling rate, temperature gradient and fluid flow are crucial
to the metal solidification, which, in turn, determined the grain
morphology, size and orientation [71,72]. Herzog et al. [109] pre-
sented typical microstructures for additive manufacturing, espe-
cially specific grain structures that resulted from the complex
thermal cycle and high cooling rates. Microstructure can be studied
experimentally by analyzing the thermal signature. Doubenskaia
et al. [45] studied the effect of different TiC content on the melt
pool temperature and ultimate microstructure in DED process.
Muvvala et al. [36] compared cooling rate and microstructure at
CW and PW laser. The stacks of columnar dendrites had different
orientation because of the repeated re-melting and solidification
during one period of PW laser as shown in Fig. 9. Higher cooling
rate could cause severe elemental segregation and formation of
Laves, which led to decrease in the layer hardness. Farshidianfar
et al. [100] studied the effect of melt pool temperature and cooling
rate on the microstructure using a thermography. It was proved
that the size of the solidification structure and mode of solidification
could be defined by the cooling rate calculated from thermography
 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
Fig. 9. Effect of cooling rate on stray grains layer thickness [36].
Fig. 10. Metallographic cross sections of voids and keyhole-mode [115].
data. Marshall et al. [47] fabricated a Ti-6Al-4V cylindrical stub and
microstructure of different location was predicted using the cool-
ing rate and temperature gradient. The prediction was correspond-
ing to the experiment results except for those of first three layers.
Many current works focused on experiment design method to
choose proper process window and acquired desirable microstruc-
ture, which was time and cost consuming [110]. A relatively new
method to control microstructure compared to this was proposed
by Farshidianfar et al. [111]. In this method, a feedback control of
cooling rate was developed and a controlled microstructure can
be achieved. The difficulties lied in this method was the complexity
of in-situ monitoring the cooling rate, which was resolved and
elaborated.
Microstructure can be predicted by using numerical model. Gu
and Yuan [71] developed a thermal-fluid coupled model and sim-
ulated the rearrangement mechanism of TiC particles. By increas-
ing the TiC content, Marangoni convection was intensified and
then, a novel ring-structure appeared in the microstructure, which
was verified by experimental observation. Zhang et al. [70] calcu-
lated the cooling rate R, temperature gradient G and G/R of differ-
ent position and which, confirmed to the experimental results. The
microstructure morphology of the cladding layer changed from
cellular to columnar and finer dendrites going from the bottom
to the top surface of the clad. Li et al. [72] studied the microstruc-
ture of CW laser and PW laser numerically. High cooling rate and
solidification rate contributed to finer grain size and the tilt angle
of the normal direction of the solidification front matched a den-
drites tilt angle.
4.4. Defects
Typical defects that may lead to later product failure include
pores, balling effect, and cracking [112]. Ali et al. [113] studied
the effect of different combination of parameters on the porosity.
Results showed that low power and low exposure generated irreg-
ularly shaped pores because of insufficient energy for melting
while spherical pores generated because of excess heat input. Shi
437
et al. [56] reported the role of melt pool temperature and liquid
flow played in the formation of pores. Keyhole-mode was widely
used in fusion welding [114] to study the formation of porosity.
Three steps were proposed for keyhole-mode formation. King
et al. [115] observed the keyhole-mode and void of different
parameters experimentally in SLM process as shown in Fig. 10.
Normalized enthalpy was used to express the threshold for the
transition from the conduction mode to keyhole mode. Khairallah
et al. [82] illustrated the complex physics such as recoil pressure
and Marangoni effect and the formation mechanism of pores. The
physics process concerned had intense dependence on the temper-
ature. Islam et al. [104] monitored the process with a pyrometer
and visual inspection and explored how input energy affected
the emergence of balling. Hu et al. [116] studied the effect of heat
input on the cracks and formation mechanism of the cracks. The
liquation film in the grain boundary and sufficient tensile stresses
should be the necessary conditions for the formation of the cracks.
Then, a method of reducing the residual stress to control the cracks
was proposed. Yan et al. [117] simulated the cooling rate effect on
the cracks and an insulating brick was used to control the heat and
prevent cracks. Hassler et al. [55] proposed a defect detection
method using thermography, but the marks can only be seen
before laser scan and there was no evident inclination visible of
the flaws in temperature profile after the deposition. Krauss et al.
[53] proposed another method that used effective heat diffusivity
as a key indicator and layer-wise monitored the process. The devi-
ation of effective heat diffusivity was related to the delamination.
The parameters-thermal behavior-quality relationship is illus-
trated in Fig. 11, some important factors that link to thermal trans-
fer such as temperature gradient, cooling rate and heat input are
given along the arrows.
4.5. Process temperature control
Combining sensing and models, many process temperature con-
trol models have been designed to acquire better product quality
such as PID controller [118], a state-space model [59], a combined
438 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
Fig. 11. Parameters-thermal behavior-quality relationship.
linear state feedback and PI controller [119]. The PID controllers
are developed based on process signature without a model or rel-
atively simplistic single-input single-output (SISO) empirical mod-
els. Tang and Lander [58] concluded that pure temperature control
cannot guarantee a uniform track morphology because consistent
melt pool size cannot be achieved by only regulating temperature
online. Physics-based multi-input multi-output (MIMO) model
[68,120], which has been discussed in Section 3.3, was developed
to control the temperature and geometry simultaneous.
5. Future research
Thermal signature plays an important role in LAM. In-situ mon-
itoring and modeling are two complementary methods that have
been widely used to study temperature distribution and thermal
behavior. The unknown emissivity of liquid metal, lacking a gener-
ous model or adequate experimental data, is the main obstacle to
the IR temperature monitoring. In the future, the emissivity data
can be accumulated and temperature-emissivity empirical models
can be developed experimentally. The true temperature field of
melt pool can be restored based on emissivity model. In addition,
new algorithms can be proposed to extract true temperature and
emissivity separately from the spectrum of melt pool.
Currently, the spatial resolution of thermometer is not sufficient
to provide thermal signature in the scale of metal powder particles
and most numerical models simulated the temperature distribu-
tion on the macroscopic scale. The powder particles behaviors
and the temperature distribution in such scale influence the liquid
flow and the product quality [121]. The process of being partial
irradiated, heated, melting and consolidation of particles is not
completely clear. In further work, new monitoring sensors can be
employed to provide monitoring data in the scale of powder parti-
cles. Measurement in the powder scale can provide detailed infor-
mation about the interaction between laser and material and the
formation mechanism of inner defects.
The material is deposited layer by layer to make a product. The
residual heat that accumulated by the last layers has an influence
on the temperature distribution of the depositing material, which,
in turn, change the convection state of the melt pool. However, the
geometry of models that study the thermal behavior and fluid flow
was mostly restricted to single track processing. Therefore, the
influence of residual heat on the convection state has not been con-
sidered. In the future, numerical models simulating the heat trans-
fer and convection pattern in the actual product can be developed.
The difficulties lie in proposing a method to save computing time
without sacrificing the accuracy when solving the N-S equation.
The SISO temperature control system is unable to guarantee the
geometry precision, which can be proved in [58]. Layer-to-layer
temperature control model [122] and multi-input multi-output
model [120] were developed to control the thermal signature
and the geometry of the track simultaneous. These control models
only consider the width of the melt pool or the height of the thin
wall. However, the deformation due to the high temperature gradi-
ent and inhomogeneous materials properties will result in the
dimensional deviation, which is not included in the current geom-
etry control system. Further study can focus on reducing the defor-
mation by the feedback control of thermal signature and reheating
process. Currently, little investigation reported successfully devel-
oped control scheme to migrate the thermal stress or acquire
expected microstructure because it is difficult to in situ monitor
either of them directly. As the temperature is a process signature
that can correlate the parameters with the thermal stress and the
microstructure, thermal stress and microstructure can be con-
trolled based on the thermal signature. Controlling of the
microstructure based on the cooling rate of the melt pool was pro-
posed by Farshidianfar et al [111]. In addition to the thermal signa-
ture, other temperature-dependent process signatures can also be
used in in-situ monitoring and feedback control such as thermal
diffusivity, spectrum signature and emissivity.
6. Conclusions
In this paper, an overview of the research about the tempera-
ture and thermal phenomenon in Laser-based Additive Manufac-
turing is provided. Typical monitor equipments and thermal
models have been summarized and the relationship between
process parameters, thermal behavior and product quality was
presented. The conclusions are listed as follows:
(1) Great efforts have been made to in-situ monitor the process
temperature since the work of Griffith et al. [19]. Low
accuracy, resulted from unknown emissivity and spatial
 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
resolution still hampered the further development of this
field. Methods to reduce the measurement errors and new
monitoring techniques were summarized, and we proposed
some temperature-dependent variables which can be moni-
tored and controlled.
(2) Physics-based thermal models include analytical model,
semi-empirical lumped-parameter models and numerical
models. Numerical models are crucial to simulate the melt
pool behaviors and temperature distribution of the product.
Then, microstructure, residual stress and deformation can be
predicted.
(3) Thermal behavior is a factor that influences the eventual
quality of the product. Parameters-thermal behavior-
quality relationship is presented and some mechanisms that
influence the residual stress, deformation, microstructure
and defects are discussed. This review can serve to resolve
quality issues by understanding the thermal behavior.
Acknowledgement
This work is jointly supported by the 973 Basic Research and
Development Plan of China (No. 2011CB013402), the Fundamental
Research Funds for the Central Universities (No. DUT16QY21), the
Natural Science Foundation of China (No. 51305063), Specialized
Research Fund for the Doctoral Program of Higher Education (No.
2013004112002), and Collaborative Innovation Center of Major
Machine Manufacturing in Liaoning.
References
[1] J. Song, Research progress of laser cladding forming technology, J. Mech. Eng.
46 (14) (2010) 29.
[2] A. Uriondo, M. Esperon-Miguez, S. Perinpanayagam, The present and future of
additive manufacturing in the aerospace sector: a review of important
aspects, Proc. Inst. Mech. Eng., Part G: J. Aerosp. Eng. 229 (11) (2015) 2132–
2147.
[3] I. Yadroitsev, P. Krakhmalev, I. Yadroitsava, Selective laser melting of Ti6Al4V
alloy for biomedical applications: temperature monitoring and
microstructural evolution, J. Alloy. Compd. 583 (2014) 404–409.
[4] M.L. Griffith et al., Understanding the microstructure and properties of
components fabricated by Laser Engineered Net Shaping (LENS), in: Solid
Freeform and Additive Fabrication-2000, 2000. 625, pp. 9–20.
[5] W.E. Frazier, Metal additive manufacturing: a review, J. Mater. Eng. Perform.
23 (6) (2014) 1917–1928.
[6] W. Liu, J.N. DuPont, Fabrication of functionally graded TiC/Ti composites by
laser engineered net shaping, Scripta Mater. 48 (9) (2003) 1337–1342.
[7] J.J. Beaman et al., Laser sintering of layers of powder to build up a part – with
computer controlled treatment and handling to produce high bulk density,
Univ Texas System (Texa-C), Univ Texas System (Texa-C).
[8] S.A. Khairallah, A. Anderson, Mesoscopic simulation model of selective laser
melting of stainless steel powder, J. Mater. Process. Technol. 214 (11) (2014)
2627–2636.
[9] D. De Baere et al., Spectroscopic monitoring and melt pool temperature
estimation during the laser metal deposition process, J. Laser Appl. 28 (2)
(2016) 022303.
[10] G. Bi et al., Identification and qualification of temperature signal for
monitoring and control in laser cladding, Opt. Lasers Eng. 44 (12) (2006)
1348–1359.
[11] M. Grasso, B.M. Colosimo, Process defects and in situ monitoring methods in
metal powder bed fusion: a review, Meas. Sci. Technol. 28 (4) (2017) 044005.
[12] H. Wang, Materials’ fundamental issues of laser additive manufacturing for
high-performance large metallic components, Acta Aeron. Astron. Sin. 35 (10)
(2014) 2690–2698.
[13] Z. Zhao et al., Evolution of plastic deformation and its effect on mechanical
properties of laser additive repaired Ti64ELI titanium alloy, Opt. Laser
Technol. 92 (2017) 36–43.
[14] E.R. Denlinger, J. Irwin, P. Michaleris, Thermomechanical modeling of additive
manufacturing large parts, J. Manuf. Sci. Eng.-Trans. ASME 136 (6) (2014)
01007.
[15] G. Tapia, A. Elwany, A review on process monitoring and control in metal-
based additive manufacturing, J. Manuf. Sci. Eng. 136 (6) (2014) 060801.
[16] M. Mani et al., A review on measurement science needs for real-time control
of additive manufacturing metal powder bed fusion processes, Int. J. Prod.
Res. 55 (5) (2016) 1400–1418.
[17] B. Schoinochoritis, D. Chantzis, K. Salonitis, Simulation of metallic powder
bed additive manufacturing processes with the finite element method: a
439
critical review, Proc. Inst. Mech. Eng., Part B: J. Eng. Manuf. 231 (1) (2017) 96–
117.
[18] J.N. Zalameda et al., Multiple sensor detection of process phenomena in laser
powder bed fusion, Spie Commerc. + Sci. Sens. Imaging 9861 (2016) 986104.
[19] M.L. Griffith et al., Understanding thermal behavior in the LENS process,
Mater. Des. 20 (2–3) (1999) 107–113.
[20] W. Ya, B. Pathiraj, S. Liu, 2D modelling of clad geometry and resulting thermal
cycles during laser cladding, J. Mater. Process. Technol. 230 (2016) 217–232.
[21] K. Zhang, X.F. Shang, W.J. Liu, Realtime measurement of temperature field
during direct laser deposition shaping, Adv. Mater. Res. 143–144 (2010) 521–
526.
[22] A. Segerstark, J. Andersson, L.-E. Svensson, Evaluation of a temperature
measurement method developed for laser metal deposition, Sci. Technol.
Weld. Joining 22 (1) (2016) 1–6.
[23] J. Mathew et al., Integrating fiber fabry-perot cavity sensor into 3-D printed
metal components for extreme high-temperature monitoring applications,
IEEE Sens. J. 17 (13) (2017) 4107–4114.
[24] S. Liu, P. Farahmand, R. Kovacevic, Optical monitoring of high power direct
diode laser cladding, Opt. Laser Technol. 64 (2014) 363–376.
[25] E. Rodriguez et al., Approximation of absolute surface temperature
measurements of powder bed fusion additive manufacturing technology
using in situ infrared thermography, Addit. Manuf. 5 (2015) 31–39.
[26] M. Doubenskaia et al., Definition of brightness temperature and restoration of
true temperature in laser cladding using infrared camera, Surf. Coat. Technol.
220 (2013) 244–247.
[27] T. Duvaut, Comparison between multiwavelength infrared and visible
pyrometry: application to metals, Infrared Phys. Technol. 51 (4) (2008)
292–299.
[28] C.-D. Wen, C.-T. Lu, Suitability of multispectral radiation thermometry
emissivity models for predicting steel surface temperature, J. Thermophys.
Heat Transfer 24 (3) (2010) 662–665.
[29] J. Xing, F. Wu, An independent retrieval algorithm of multi-wavelength
pyrometer, in: 2015 Fifth International Conference on Instrumentation and
Measurement, Computer, Communication and Control (Imccc), 2015, pp.
943–946.
[30] J. Xing, R.S. Rana, W. Gu, Emissivity range constraints algorithm for multi-
wavelength pyrometer (MWP), Opt. Express 24 (17) (2016) 19185–19194.
[31] P. Hagqvist et al., Emissivity compensated spectral pyrometry for varying
emissivity metallic measurands, Meas. Sci. Technol. 25 (2) (2014) 025010.
[32] P. Hagqvist et al., Emissivity compensated spectral pyrometry—algorithm and
sensitivity analysis, Meas. Sci. Technol. 25 (2) (2014) 025011.
[33] H. Köhler, C. Thomy, F. Vollertsen, Contact-less temperature measurement
and control with applications to laser cladding, Weld. World 60 (1) (2015) 1–
9.
[34] I. Zhirnov et al., New approach of true temperature restoration in optical
diagnostics using IR-camera, J. Therm. Spray Technol. 26 (4) (2017) 648–660.
[35] T. Dreischuh et al., Improved laser metal deposition (LMD) of nickel base
superalloys by pyrometry process control, in: Proc Spie, 2010, 7751, pp.
775123-775123-9.
[36] G. Muvvala, D. Patra Karmakar, A.K. Nath, Online monitoring of thermo-cycles
and its correlation with microstructure in laser cladding of nickel based super
alloy, Opt. Lasers Eng. 88 (2017) 139–152.
[37] J.B. Lei, Z. Wang, L.F. Liu, Design of forming shape measurement system for
laser molten pool in laser fabricating, Appl. Mech. Mater. 37–38 (2010) 327–
330.
[38] H. Krauss, T. Zeugner, M.F. Zaeh, Thermographic process monitoring in
powderbed based additive manufacturing, in: 41st Annual Review of Progress
in Quantitative Nondestructive Evaluation, vol. 34, 2015, 1650, pp. 177–183.
[39] M. Doubenskaia, A. Domashenkov, I. Smurov, Study of the laser melting of
pre-deposited intermetallic TiAl powder by comprehensive optical
diagnostics, Surf. Coat. Technol. 321 (2017) 118–127.
[40] W. Devesse et al., High resolution temperature estimation during laser
cladding of stainless steel, in: Laser Assisted Net Shape Engineering 9
International Conference on Photonic Technologies Proceedings of the Lane
2016, 2016, 83, pp. 1253–1260.
[41] G. Bi et al., Development and qualification of a novel laser-cladding head with
integrated sensors, Int. J. Mach. Tools Manuf. 47 (3–4) (2007) 555–561.
[42] G. Bi et al., Investigation on the direct laser metallic powder deposition
process via temperature measurement, Appl. Surf. Sci. 253 (3) (2006) 1411–
1416.
[43] I. Smurov et al., Optical monitoring in laser cladding of Ti6Al4V, J. Therm.
Spray Technol. 21 (6) (2012) 1357–1362.
[44] I. Smurov, M. Doubenskaia, A. Zaitsev, Comprehensive analysis of laser
cladding by means of optical diagnostics and numerical simulation, Surf.
Coat. Technol. 220 (2013) 112–121.
[45] M. Doubenskaia et al., Complex analysis of elaboration of steel–TiC
composites by direct metal deposition, J. Laser Appl. 25 (4) (2013) 042009.
[46] G.J. Marshall, S.M. Thompson, N. Shamsaei, Data indicating temperature
response of Ti-6Al-4V thin-walled structure during its additive manufacture
via Laser Engineered Net Shaping, Data Brief 7 (2016) 697–703.
[47] G.J. Marshall et al., Understanding the microstructure formation of Ti-6Al-4V
during direct laser deposition via in-situ thermal monitoring, Jom 68 (3)
(2016) 778–790.
[48] C. Zhang, L. Li, A. Deceuster, Thermomechanical analysis of multi-bead pulsed
laser powder deposition of a nickel-based superalloy, J. Mater. Process.
Technol. 211 (9) (2011) 1478–1487.
440 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
[49] P. Farahmand, R. Kovacevic, An experimental–numerical investigation of heat
distribution and stress field in single- and multi-track laser cladding by a
high-power direct diode laser, Opt. Laser Technol. 63 (2014) 154–168.
[50] E.R. Denlinger et al., Effect of inter-layer dwell time on distortion and residual
stress in additive manufacturing of titanium and nickel alloys, J. Mater.
Process. Technol. 215 (2015) 123–131.
[51] W. Devesse, D. De Baere, P. Guillaume, High resolution temperature
measurement of liquid stainless steel using hyperspectral imaging, Sensors
(Basel) 17(1) (2017).
[52] M. Doubenskaia, P. Bertrand, I. Smurov, Optical monitoring of Nd:YAG laser
cladding, Thin Solid Films 453–454 (2004) 477–485.
[53] H. Krauss, T. Zeugner, M.F. Zaeh, Layerwise monitoring of the selective laser
melting process by thermography, Phys. Proc. 56 (2014) 64–71.
[54] S. Moylan et al., Infrared thermography for laser-based powder bed fusion
additive manufacturing processes, in: 40th Annual Review of Progress in
Quantitative Nondestructive Evaluation: Incorporating the 10th International
Conference on Barkhausen Noise and Micromagnetic Testing, vols. 33a & 33b,
2014, 1581, pp. 1191–1196.
[55] U. Hassler et al., In-situ monitoring and defect detection for laser metal
deposition by using infrared thermography, Phys. Proc. 83 (2016) 1244–1252.
[56] Q. Shi et al., Effects of laser processing parameters on thermal behavior and
melting/solidification mechanism during selective laser melting of TiC/
Inconel 718 composites, Opt. Laser Technol. 84 (2016) 9–22.
[57] J.P. Kruth et al., Feedback control of selective laser melting, Virtual and Rapid
Manufacturing (2008) 521–527.
[58] L. Tang, R.G. Landers, Melt pool temperature control for laser metal
deposition processes-part I: online temperature control, J. Manuf. Sci. Eng.-
Trans. ASME 132 (1) (2010) 011010.
[59] L. Song, J. Mazumder, Feedback control of melt pool temperature during laser
cladding process, IEEE Trans. Control Syst. Technol. 19 (6) (2011) 1349–1356.
[60] L. Song et al., Control of melt pool temperature and deposition height during
direct metal deposition process, Int. J. Adv. Manuf. Technol. 58 (1–4) (2011)
247–256.
[61] D. Rosenthal, The Theory of Moving Source of Heat and Its Application to
Metal Treatments, 1946, 68.
[62] T.W. Eagar, N.S. Tsai, Temperature-fields produced by traveling distributed
heat-sources, Weld. J. 62 (12) (1983) S346–S355.
[63] A. Vasinonta, J.L. Beuth, M.L. Griffith, A process map for consistent build
conditions in the solid freeform fabrication of thin-walled structures, J.
Manuf. Sci. Eng.-Trans. ASME 123 (4) (2001) 615–622.
[64] A. Vasinonta, J. Beuth, M. Griffith, Process maps for laser deposition of thin-
walled structures, in: Solid Freeform Fabrication Proceedings, August 1999,
1999, pp. 383–391.
[65] J. Gockel, N. Klingbeil, S. Bontha, A closed-form solution for the effect of free
edges on melt pool geometry and solidification microstructure in additive
manufacturing of thin-wall geometries, Metall. Mater. Trans. B 47 (2) (2015)
1400–1408.
[66] C. Doumanidis, Y.M. Kwak, Geometry modeling and control by infrared and
laser sensing in thermal manufacturing with material deposition, J. Manuf.
Sci. Eng.-Trans. ASME 123 (1) (2001) 45–52.
[67] P.M. Sammons, D.A. Bristow, R.G. Landers, Height dependent laser metal
deposition process modeling, J. Manuf. Sci. Eng. 135 (5) (2013) 054501.
[68] Q. Wang et al., Physics-based multivariable modeling and feedback
linearization control of MeltPool geometry and temperature in directed
energy deposition, J. Manuf. Eng.-Trans. ASME 139 (2) (2017) 021013.
[69] X.Q. Cao, B. Ayalew, Control-oriented MIMO modeling of laser-aided powder
deposition processes, Am. Control Conf. (Acc) 2015 (2015) 3637–3642.
[70] Z. Zhang, P. Farahmand, R. Kovacevic, Laser cladding of 420 stainless steel
with molybdenum on mild steel A36 by a high power direct diode laser,
Mater. Des. 109 (2016) 686–699.
[71] D. Gu, P. Yuan, Thermal evolution behavior and fluid dynamics during laser
additive manufacturing of Al-based nanocomposites: underlying role of
reinforcement weight fraction, J. Appl. Phys. 118 (23) (2015) 233109.
[72] S. Li et al., Melt-pool motion, temperature variation and dendritic
morphology of Inconel 718 during pulsed- and continuous-wave laser
additive manufacturing: a comparative study, Mater. Des. 119 (2017) 351–
360.
[73] Z. Gan et al., Numerical simulation of thermal behavior and multicomponent
mass transfer in direct laser deposition of Co-base alloy on steel, Int. J. Heat
Mass Transf. 104 (2017) 28–38.
[74] S. Morville et al., 2D longitudinal modeling of heat transfer and fluid flow
during multilayered direct laser metal deposition process, J. Laser Appl. 24(3)
(2012) 032008-1-9.
[75] H. Ju et al., Test and temperature field of finite element simulation about the
effect of scanning speed on 304 stainless layer’s properties by laser cladding.
Mater. Res. Innov. 19(sup8) (2015) S8-9-S8-13.
[76] R. Martukanitz et al., Toward an integrated computational system for
describing the additive manufacturing process for metallic materials, Addit.
Manuf. 1–4 (2014) 52–63.
[77] L.-E. Lindgren et al., Simulation of additive manufacturing using coupled
constitutive and microstructure models, Addit. Manuf. 12 (2016) 144–158.
[78] A.J. Dunbar et al., Experimental validation of finite element modeling for laser
powder bed fusion deformation, Addit. Manuf. 12 (2016) 108–120.
[79] M.F. Gouge et al., Modeling forced convection in the thermal simulation of
laser cladding processes, Int. J. Adv. Manuf. Technol. 79 (1–4) (2015) 307–
320.
[80] V. Manvatkar, A. De, T. DebRoy, Spatial variation of melt pool geometry,
peak temperature and solidification parameters during laser assisted
additive manufacturing process, Mater. Sci. Technol. 31 (8) (2014)
924–930.
[81] I. Tabernero et al., Modelling of energy attenuation due to powder flow-laser
beam interaction during laser cladding process, J. Mater. Process. Technol.
212 (2) (2012) 516–522.
[82] S.A. Khairallah et al., Laser powder-bed fusion additive manufacturing:
physics of complex melt flow and formation mechanisms of pores, spatter,
and denudation zones, Acta Mater. 108 (2016) 36–45.
[83] F. Vásquez, J.A. Ramos-Grez, M. Walczak, Multiphysics simulation of laser–
material interaction during laser powder deposition, Int. J. Adv. Manuf.
Technol. 59 (9–12) (2011) 1037–1045.
[84] T. Mukherjee, W. Zhang, T. DebRoy, An improved prediction of residual
stresses and distortion in additive manufacturing, Comput. Mater. Sci. 126
(2017) 360–372.
[85] S. Safdar et al., An anisotropic enhanced thermal conductivity approach for
modelling laser melt pools for Ni-base super alloys, Appl. Math. Model. 37 (3)
(2013) 1187–1195.
[86] G. Yu et al., On the role of processing parameters in thermal behavior, surface
morphology and accuracy during laser 3D printing of aluminum alloy, J. Phys.
D: Appl. Phys. 49(13) (2016).
[87] M. Xia et al., Porosity evolution and its thermodynamic mechanism of
randomly packed powder-bed during selective laser melting of Inconel 718
alloy, Int. J. Mach. Tools Manuf 116 (2017) 96–106.
[88] P. Michaleris, Modeling metal deposition in heat transfer analyses of additive
manufacturing processes, Finite Elem. Anal. Des. 86 (2014) 51–60.
[89] X. Zhao et al., Numerical modeling of the thermal behavior and residual stress
in the direct metal laser sintering process of titanium alloy products, Addit.
Manuf. 14 (2017) 126–136.
[90] P. Peyre et al., Analytical and numerical modelling of the direct metal
deposition laser process, J. Phys. D Appl. Phys. 41 (2) (2008) 025403.
[91] T. Mukherjee et al., Mitigation of thermal distortion during additive
manufacturing, Scripta Mater. 127 (2017) 79–83.
[92] H. Runnemalm, S. Hyun, Three-dimensional welding analysis using an
adaptive mesh scheme, Comput. Methods Appl. Mech. Eng. 189 (2) (2000)
515–523.
[93] E.R. Denlinger, J.C. Heigel, P. Michaleris, Residual stress and distortion
modeling of electron beam direct manufacturing Ti-6Al-4V, Proc. Inst.
Mech. Eng. Part B-J. Eng. Manuf. 229 (10) (2015) 1803–1813.
[94] E. Kundakcioglu, I. Lazoglu, S. Rawal, Transient thermal modeling of laser-
based additive manufacturing for 3D freeform structures, Int. J. Adv. Manuf.
Technol. 85 (1–4) (2015) 493–501.
[95] C. Li et al., A multiscale modeling approach for fast prediction of part
distortion in selective laser melting, J. Mater. Process. Technol. 229 (2016)
703–712.
[96] Y.W. Zhang, A. Faghri, Thermal modeling of selective area laser deposition of
titanium nitride on a finite slab with stationary and moving laser beams, Int.
J. Heat Mass Transf. 43 (20) (2000) 3835–3846.
[97] L. Hao et al., Three-dimensional numerical simulation of transient
temperature field and coating geometry in powder feeding laser cladding,
Chin. J. Lasers 40 (12) (2013) 1203007.
[98] Y. Lee, D.F. Farson, Simulation of transport phenomena and melt pool shape
for multiple layer additive manufacturing, J. Laser Appl. 28 (1) (2016) 012006.
[99] H. Ki, P.S. Mohanty, J. Mazumder, Modeling of laser keyhole welding: part I.
Mathematical modeling, numerical methodology, role of recoil pressure,
multiple reflections, and free surface evolution, Metall. Mater. Trans. a-Phys.
Metall. Mater. Sci. 33(6) (2002) 1817–1830.
[100] M.H. Farshidianfar, A. Khajepour, A.P. Gerlich, Effect of real-time cooling rate
on microstructure in Laser Additive Manufacturing, J. Mater. Process.
Technol. 231 (2016) 468–478.
[101] J. Beuth, N. Klingbeil, The role of process variables in laser-based direct metal
solid freeform fabrication, Jom 53 (9) (2001) 36–39.
[102] Y. Li, D. Gu, Parametric analysis of thermal behavior during selective laser
melting additive manufacturing of aluminum alloy powder, Mater. Des. 63
(2014) 856–867.
[103] T. Mukherjee et al., Dimensionless numbers in additive manufacturing, J.
Appl. Phys. 121 (6) (2017) 064904.
[104] M. Islam et al., Temperature profile and imaging analysis of laser additive
manufacturing of stainless steel, Phys. Proc. 41 (2013) 835–842.
[105] I.A. Roberts et al., Experimental and numerical analysis of residual stresses in
additive layer manufacturing by laser melting of metal powders, Key Eng.
Mater. 450 (2010) 461–465.
[106] Z. Nie et al., Experimental study and modeling of H13 steel deposition using
laser hot-wire additive manufacturing, J. Mater. Process. Technol. 235 (2016)
171–186.
[107] T. Furumoto et al., Study on deformation restraining of metal structure
fabricated by selective laser melting, J. Mater. Process. Technol. 245 (2017)
207–214.
[108] J.X. Fang et al., The effects of solid-state phase transformation upon stress
evolution in laser metal powder deposition, Mater. Des. 87 (2015) 807–814.
[109] D. Herzog et al., Additive manufacturing of metals, Acta Mater. 117 (2016)
371–392.
[110] Y.-J. Liang et al., Experimental optimization of laser additive manufacturing
process of single-crystal nickel-base superalloys by a statistical experiment
design method, J. Alloy. Compd. 697 (2017) 174–181.
 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
[111] M.H. Farshidianfar, A. Khajepour, A. Gerlich, Real-time control of
microstructure in laser additive manufacturing, Int. J. Adv. Manuf. Technol.
82 (5–8) (2015) 1173–1186.
[112] S.K. Everton et al., Review of in-situ process monitoring and in-situ
metrology for metal additive manufacturing, Mater. Des. 95 (2016) 431–
445.
[113] H. Ali et al., In-situ residual stress reduction, martensitic decomposition and
mechanical properties enhancement through high temperature powder bed
pre-heating of Selective Laser Melted Ti6Al4V, Mater. Sci. Eng., A 695 (2017)
211–220.
[114] R. Lin et al., Numerical study of keyhole dynamics and keyhole-induced
porosity formation in remote laser welding of Al alloys, Int. J. Heat Mass
Transf. 108 (2017) 244–256.
[115] W.E. King et al., Observation of keyhole-mode laser melting in laser powder-
bed fusion additive manufacturing, J. Mater. Process. Technol. 214 (12)
(2014) 2915–2925.
441
[116] Y.L. Hu et al., Effect of heat input on cracking in laser solid formed DZ4125
superalloy, Opt. Laser Technol. 86 (2016) 1–7.
[117] L. Yan et al., Simulation of cooling rate effects on Ti–48Al–2Cr–2Nb crack
formation in direct laser deposition, Jom 69 (3) (2016) 586–591.
[118] D. Salehi, M. Brandt, Melt pool temperature control using LabVIEW in Nd:
YAG laser blown powder cladding process, Int. J. Adv. Manuf. Technol. 29 (3–
4) (2005) 273–278.
[119] W. Devesse et al., Hardware-in-the-loop control of additive manufacturing
processes using temperature feedback, J. Laser Appl. 28 (2) (2016) 022302.
[120] X.Q. Cao, B. Ayalew, Multivariable predictive control of laser-aided powder
deposition processes, Am. Control Conf. (Acc) 2015 (2015) 3625–3630.
[121] M.J. Matthews et al., Denudation of metal powder layers in laser powder bed
fusion processes, Acta Mater. 114 (2016) 33–42.
[122] L. Tang, R.G. Landers, Melt pool temperature control for laser metal
deposition processes-part II: layer-to-layer temperature control, J. Manuf.
Sci. Eng.-Trans. ASME 132(1) (2010).