Professional Documents
Culture Documents
2018 Review On Thermal Analysis in Laser-Based Additive Manufacturing
2018 Review On Thermal Analysis in Laser-Based Additive Manufacturing
Review
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 Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441
6. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
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.
comprehensively analyzed the deposition process by a monochro- Physics-based analytical and numerical models are crucial to pre-
matic pyrometer, a multi-wavelength pyrometer and an IR- dicting the outcome of LAM process that accounts for the changes
camera as shown in Fig. 1. Diverse temperature data fusion were in material properties [16], and the empirical models are mostly
presented to synthetically analyze the parameters dependence used in control systems. In this section, the melt pool behaviors
[44] and the additive material influence [45]. In the literature are first introduced, and then, the empirical, analytical and numer-
[39], a similar system was applied in PBF, which focused on the ical models are discussed in details.
mechanism of the heat and mass transfer of pre-deposited inter-
metallic TiAl powder with optical diagnostics. Liu et al. [24] per- 3.1. Melt pool behaviors in LAM
formed optical diagnostics on high power direct diode laser
cladding with a system comprised of a high-speed CCD camera, The substrate is irradiated and heated by laser and conduction
a pyrometer, and an infrared camera. The variation of the bright- dominate the heat transfer process. When the temperature of the
ness temperature of the molten pool was measured by the substrate is locally higher than the melting point, the liquid melt
pyrometer and the temperature distribution, molten pool size, pool generates. High temperature gradient occurs at the surface
and cooling rate at different processing parameters was studied of the melt pool, and the resulting ‘‘Marangoni force” drives the liq-
by using the infrared camera. Another monitoring system was uid metal flow. As the melt pool moves with the laser spot, melting
present in the literature [46,47], and Marshall et al. built a thin and solidification processes evolve continuously, so the latent heat
wall and a rod respectively. A dual-wavelength pyrometer was of phase change, heat transfer and flow behaviors in the binary
mounted coaxially to monitor the melt pool temperature and an solid-liquid phase region have to be taken into account. Heat trans-
IR-camera was tilted and viewed the whole product’s temperature fers from the product to the surrounding by convection and radia-
distribution as illustrated in Fig. 2. Lane et al. [18] fabricated a tion. The materials properties (including thermo-physical
nickel alloy 625 part conducted on a commercial LPBF machine, properties and mechanical properties) depend on temperature.
which was incorporated with a thermal camera, a high-speed vis- The interaction of laser, powder and melt pool including the atten-
ible camera and a photodiode. The photodetector signal and cam- uation effect is not completely clear today as shown in Fig. 3. Based
era images were merged simultaneously to have a better on the heat transfer, mass transfer and momentum transfer of melt
understanding of the melt pool behaviors. For a generalization pool, the thermal models can be developed to reveal key informa-
of the studies surveyed above, monitoring techniques and sensors tion inside the melt pool that in-situ monitoring is unable to do it.
are summarized in Table 1.
3.2. Empirical models
3. Thermal models in LAM Empirical models are developed experimentally to estimate the
response between process parameter and melt pool temperature.
There has been a continuous effort in developing various ther- Empirical models with transient response relating the laser power
mal models, including empirical, analytical and numerical models. 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
Numerical models are more economical and available to pre- 3.4.3. Convection and solidification
sent complex analysis inside the melt pool, compared with the The convection pattern has a strong impact on the heat transfer
experimental investigation. The main procedures of numerically and temperature distribution in the liquid metal. Manvatkar et al.
modeling include pre-processing (geometry modeling, boundary [80] showed that by ignoring the effect of convection, the cooling
conditions, initial conditions, heat loads and material properties), rates in additive manufacturing were over-estimated by about
processing (choosing proper governing equation and solving) and 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.
should not be neglected in numerical models. The liquid metal is et al. [83] developed a 3D quasi-stationary multi-physics model
driven by ‘‘Marangoni effect” and under the influence of gravity, considering the ‘Marangoni effect’ and the quality of vapor mass
buoyancy and impact force from the shielding gas and vaporized escaping from the pool, which was calculated by the analytical
alloy, which makes the flow pattern complex. The N-S equation form of Stefan’s condition under a quasi-steady state. For transient
has poor convergence, so the simplified method of increasing the heat transfer and fluid flow model, the fluid flow is acquired by
heat conductivity in heat conduction equation has been widely generally solving the N-S equation. Morville et al. [74] simulated
adopted to estimate the effect of fluid flow [20,69,70]. Vásquez 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].
2D longitudinal thermal model and the main physics phenomena at which heat transfers through the material and the amount of the
such as surface tension effect and addition of powder were taken energy needed for a certain temperature rise can be found using
into account. A 3D moving meshing model was developed by the specific heat capacity [17]. Most literature indicated that heat
Gan et al. [73], and thermal behavior, fluid flow, melting–solidifica- conductivity and specific heat were defined to be temperature
tion and multicomponent mass transfer were investigated. Gu and dependence [14,56,90,91]. Gu et al. [71] considered in PBF, the
Yuan [71] developed a transient three-dimensional CFD model to effective thermal conductivity of loose metallic powder was con-
investigate the influence of the TiC reinforcement weight fraction trolled by gas in pores and was about 100 times lower than that
on heat transfer and fluid flow during SLM. The computed velocity of the dense material, and the effective thermal conductivity was
at different time was shown in Fig. 4. The influence of convection restored to actual value when the temperature was higher than
was also considered [72,84]. Another factor that affects tempera- the melting point. The physical properties in the mushy zone dur-
ture distribution is latent heat released during solidification. There ing solidification were defined to be isotropy and the weighted
were mainly two ways to estimate the latent heat. One method average of that of solid and liquid phase [69,83]. In thermomechan-
was to calculate enthalpy change according to the definition of ical models, physics properties such as elastic coefficient, yield
latent heat of phase change [72,73] and the other one was to strength and coefficient of thermal expansion [48,49,84] were tem-
increase the specific heat [20,49,83]. The latter one method perature dependence.
required very precise control of time stepping to improve the sim-
ulation of phase change [85]. 3.4.5. Computing cost
Models at the mesoscopic scale were widely used in under- The high computing cost of thermal model prevents it from fur-
standing the underlying physics and the formation mechanism of ther adoption in industrial application, especially the simulation of
defects in SLM process. Khairallah and Anderson [8,82] proposed large products. Bringing down the computing time without sacri-
a 3D mesoscopic model by using ALE3D simulation software. The ficing accuracy would be necessary for numerical models of large
melt flow of liquid 316L was analyzed and the formation pores, products. Denlinger et al. [14] used the adaptive meshing, which
spatter, and denudation zones were further explored. Yu et al. meant merging the elements layer by layer and coarsening the
[86] developed an FVM model, which considered the powder-to- mesh of layers to reduce the number of nodes. The mesh coarsen
solid transition to analyze the thermal behavior, the effect of laser algorithm was also applied in [92,93]. Another approach was pro-
power on the surface morphology and balling phenomenon. Xia posed by Kundakcioglu et al. [94] Scanning route was broken into
et al. [87] studied the evolution and mechanism of porosity using laser position points according to the laser beam diameter, and
a transient mesoscale model with randomly-packed powder bed. then a file containing 4D data (time and three position coordinates)
Results showed that at high scanning speed, a short lifespan was was formed. This file was transferred to Gaussian heat model and a
produced and velocity and depth of the melt pool were evidently discrete heat source was simulated instead of a traditionally con-
reduced. The porosity transformed from metallurgical porosity to tinuous model. Li et al. [95] developed a multiscale model to sim-
open porosity as the scanning speed increased as shown in Fig. 5. ulate the residual stress and distortion during SLM, as well as to
reduce the computing time. The model consisted of microscale
3.4.4. Temperature-dependent material properties laser scan model, mesoscale layer hatch model and macroscale part
model as shown in Fig. 6.
The materials properties undergo nonlinear dependence on the
temperature change and phase change. A suitable definition of
3.4.6. Geometry and material additive
materials properties is necessary to an accurate model. The
temperature-dependence heat physics properties include density, Geometry model can be defined based on the fitting of the
experiment results [96] or assumed to be known geometry
conductivity and heat capacity. Because temperature has little
influence on the density of solid metal and liquid metal is always [14,56] in heat conduction models. To simulate the addition of
material, a number of studies used the method of ‘‘element birth
assumed to be incompressible, density was frequently defined to
be constant [49,80,88,89], while sometimes it was defined as tem- and death”. This method consists of the quiet element method
[49,75,79] and the inactive element method [14,88]. In quiet ele-
perature dependence [74,85,90]. Heat conductivity and specific
heat are two physics properties that affect the temperature distri- ment method, the elements are considered during the analysis
with low heat conductivity and specific heat value. In the inactive
bution. The thermal conductivity of a material determines the rate
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
Fig. 6. Multi-scale methodology for fast prediction of part distortion and residual kinds of velocities were considered at liquid/gas interface – the
stress [95]. fluid flow velocity and boundary moving velocity due to mass
Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441 435
addition. 2D model and 3D model using this method were pre- image [26] and thermal model [71] are presented in Fig. 8. The high
sented in [74] and [73] respectively. 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
4. Parameter-thermal behavior-quality relationships pool where the metal cooled down and solidified. It was also found
that monotonous increase of melt pool size (length, width and
In this section, first we compared results between the in-situ depth) with laser power and scan speed was obvious [44,56]. Melt
monitoring and the thermal modeling, and discussed how process pool underwent contraction and expansion during the period of
parameters influence dynamic melt pool, temperature distribution pulsed-wave laser and the temperature fluctuated with the same
and temperature-dependent variables. Then, the thermal behavior frequency of the pulse laser compared with the melt pool in con-
was further extended to analyze the product quality. The product tinuous wave [72].
quality is not directly related to temperature, but has dependence Temperature gradient and cooling rate, deduced from tempera-
on the residual stress, deformation, microstructure and defects of ture distribution or thermo-cycle, are crucial to the end quality of
the product, so the effect of temperature on these issues was the product. The cooling rate decreased with higher laser power
studied. and increased with larger scan speed [100], which was reported
in monitoring data [36] and model results [80]. However, different
4.1. Parameters-thermal behavior relationship in LAM trend was illustrated in [56] that the cooling rate increased with
both the laser power and the scan speed. The reason for this differ-
It has been extensively reported that process parameters, such ence may be that an overall enhanced conductivity method was
as laser power, scan speed and powder feeding rate, have different adopted to consider the convection, which was unable to simulate
influence on the product quality. The temperature distribution the local fluid flow and its impact on the local temperature of the
comparison between monitoring [44] and modeling [56] is pre- tailing zoon and bottom of the melt pool [56]. Li et al. [72] reported
sented in Fig. 7. Temperature showed a similarly increasing trend that cooling rates along the solidification front using pulse laser
with higher laser power and slower scan speed, though different were almost twice those using continue laser. Beuth and Klingbeil
parameter values were chosen for laser cladding and selective laser [101] first drew a process map for temperature gradient analyti-
melting. Similar trends were reported in [75,80]. Muvvala et al. cally. Li et al. [102] developed a numerical model and found that
[36] monitored the thermo-cycle of LBF and found that tempera- the maximum temperature gradient was in the front melt pool in
ture dropped when the scan speed was lower than 800 mm/min, which the maximum laser intensity located. In the depth direction,
because evaporation from the molten pool surface become signifi- the temperature gradients decreased with the larger distance from
cant. The temperature distribution of the melt pool in the thermal the surface of the melt pool [70,102], and the temperature gradients
Fig. 7. Temperature distribution of model (a), (b) [56] and monitoring results (c), (d) [44].
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- was applied as load on the nodes and stress and displacement
ing rate slowed down because of the interaction between laser fields are simulated. Mukherjee et al. [84] developed a thermal-
beam and particles [43]. fluid coupled numerical model, and then a thermo-mechanical
Since parameters influence the temperature synthetically, it is model was presented with ABQUS. Results showed the influence
significant to combine the parameters and studies the comprehen- of layer thickness and heat input on the residual stress and defor-
sive effect. The effective energy and the powder deposition density mation. Furumoto et al. [107] reported that the deformation of the
was utilized by Farshidianfar et al. [100] to study the parameter component was related to the coefficient of thermal expansion,
dependence of cooling rate. Muvvala et al. [36] used the line energy mechanical strength and temperature gradient, and among which,
to investigate the energy distribution. Mukherjee et al. [103] stud- the coefficient of thermal expansion was the decisive factor. Farah-
ied the role of non-dimensional numbers, including non- mand et al. [49] developed a thermo-elastic-plastic model and
dimensional heat input, Peclet numbers, Marangoni numbers and found that there was high concentrated stress in the last tracks
Fourier number in processing. Islam et al. [104] chose volumetric because of the large cooling rate and the stress release of the pre-
energy input to describe the amount of energy input of SLM vious tracks. This characteristic of stress distribution was also
process. 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
4.2. Residual stress and deformation
induced plasticity and the phase transformation temperature on
the stress field was considered. Zhang et al. [48] developed a ther-
The rapid thermal cycling causes high temperature gradient
momechanical model of multi-bead pulsed laser metal deposition.
near the melt pool and which, in turn, leads to undesirable defor-
The maximum temperature gradient located in the interface
mation and dimensional distortion in the product. Residual stress
between depositions and substrate and which, the thermal stresses
resulting from inhomogeneous thermal strain is generated in the
concentrated on.
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 4.3. Solidification and microstructure
failure [49].
Deformation can be measured either in situ or after processing. The cooling rate, temperature gradient and fluid flow are crucial
Roberts et al. [105] measured the deformation using a Confocal to the metal solidification, which, in turn, determined the grain
Laser Scanning Microscope after processing. The cases of in-situ morphology, size and orientation [71,72]. Herzog et al. [109] pre-
monitoring the deformation were limited. Denlinger et al. [50] sented typical microstructures for additive manufacturing, espe-
employed a laser displacement sensor to monitor the distortion cially specific grain structures that resulted from the complex
of the free end of the substrate when the other end was clamped. thermal cycle and high cooling rates. Microstructure can be studied
The temperature data was recorded with thermocouples located experimentally by analyzing the thermal signature. Doubenskaia
on the substrate bottom. A thin wall was processed and a new et al. [45] studied the effect of different TiC content on the melt
parameter called the dwell time was defined to study its influence pool temperature and ultimate microstructure in DED process.
on the distortion and the residual stress. A similar method with Muvvala et al. [36] compared cooling rate and microstructure at
thermocouples and displacement sensors were also adopted by CW and PW laser. The stacks of columnar dendrites had different
Nie et al. [106] in laser hot-wire additive manufacturing. Furumoto orientation because of the repeated re-melting and solidification
et al. [107] processed a thin wall of different material on the stain- during one period of PW laser as shown in Fig. 9. Higher cooling
less steel or carbon steel substrate and a strain gauge was used to rate could cause severe elemental segregation and formation of
measure the stress history of the substrate. The resulting stress and Laves, which led to decrease in the layer hardness. Farshidianfar
stress history were compared when different material of powder et al. [100] studied the effect of melt pool temperature and cooling
and substrates were used. rate on the microstructure using a thermography. It was proved
Numerical models can be used to predict the residual stress and that the size of the solidification structure and mode of solidification
deformation. The temperature distribution in the thermal models could be defined by the cooling rate calculated from thermography
Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441 437
data. Marshall et al. [47] fabricated a Ti-6Al-4V cylindrical stub and et al. [56] reported the role of melt pool temperature and liquid
microstructure of different location was predicted using the cool- flow played in the formation of pores. Keyhole-mode was widely
ing rate and temperature gradient. The prediction was correspond- used in fusion welding [114] to study the formation of porosity.
ing to the experiment results except for those of first three layers. Three steps were proposed for keyhole-mode formation. King
Many current works focused on experiment design method to et al. [115] observed the keyhole-mode and void of different
choose proper process window and acquired desirable microstruc- parameters experimentally in SLM process as shown in Fig. 10.
ture, which was time and cost consuming [110]. A relatively new Normalized enthalpy was used to express the threshold for the
method to control microstructure compared to this was proposed transition from the conduction mode to keyhole mode. Khairallah
by Farshidianfar et al. [111]. In this method, a feedback control of et al. [82] illustrated the complex physics such as recoil pressure
cooling rate was developed and a controlled microstructure can and Marangoni effect and the formation mechanism of pores. The
be achieved. The difficulties lied in this method was the complexity physics process concerned had intense dependence on the temper-
of in-situ monitoring the cooling rate, which was resolved and ature. Islam et al. [104] monitored the process with a pyrometer
elaborated. and visual inspection and explored how input energy affected
Microstructure can be predicted by using numerical model. Gu the emergence of balling. Hu et al. [116] studied the effect of heat
and Yuan [71] developed a thermal-fluid coupled model and sim- input on the cracks and formation mechanism of the cracks. The
ulated the rearrangement mechanism of TiC particles. By increas- liquation film in the grain boundary and sufficient tensile stresses
ing the TiC content, Marangoni convection was intensified and should be the necessary conditions for the formation of the cracks.
then, a novel ring-structure appeared in the microstructure, which Then, a method of reducing the residual stress to control the cracks
was verified by experimental observation. Zhang et al. [70] calcu- was proposed. Yan et al. [117] simulated the cooling rate effect on
lated the cooling rate R, temperature gradient G and G/R of differ- the cracks and an insulating brick was used to control the heat and
ent position and which, confirmed to the experimental results. The prevent cracks. Hassler et al. [55] proposed a defect detection
microstructure morphology of the cladding layer changed from method using thermography, but the marks can only be seen
cellular to columnar and finer dendrites going from the bottom before laser scan and there was no evident inclination visible of
to the top surface of the clad. Li et al. [72] studied the microstruc- the flaws in temperature profile after the deposition. Krauss et al.
ture of CW laser and PW laser numerically. High cooling rate and [53] proposed another method that used effective heat diffusivity
solidification rate contributed to finer grain size and the tilt angle as a key indicator and layer-wise monitored the process. The devi-
of the normal direction of the solidification front matched a den- ation of effective heat diffusivity was related to the delamination.
drites tilt angle. The parameters-thermal behavior-quality relationship is illus-
trated in Fig. 11, some important factors that link to thermal trans-
4.4. Defects fer such as temperature gradient, cooling rate and heat input are
given along the arrows.
Typical defects that may lead to later product failure include
pores, balling effect, and cracking [112]. Ali et al. [113] studied 4.5. Process temperature control
the effect of different combination of parameters on the porosity.
Results showed that low power and low exposure generated irreg- Combining sensing and models, many process temperature con-
ularly shaped pores because of insufficient energy for melting trol models have been designed to acquire better product quality
while spherical pores generated because of excess heat input. Shi 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
linear state feedback and PI controller [119]. The PID controllers influence of residual heat on the convection state has not been con-
are developed based on process signature without a model or rel- sidered. In the future, numerical models simulating the heat trans-
atively simplistic single-input single-output (SISO) empirical mod- fer and convection pattern in the actual product can be developed.
els. Tang and Lander [58] concluded that pure temperature control The difficulties lie in proposing a method to save computing time
cannot guarantee a uniform track morphology because consistent without sacrificing the accuracy when solving the N-S equation.
melt pool size cannot be achieved by only regulating temperature The SISO temperature control system is unable to guarantee the
online. Physics-based multi-input multi-output (MIMO) model geometry precision, which can be proved in [58]. Layer-to-layer
[68,120], which has been discussed in Section 3.3, was developed temperature control model [122] and multi-input multi-output
to control the temperature and geometry simultaneous. model [120] were developed to control the thermal signature
and the geometry of the track simultaneous. These control models
5. Future research only consider the width of the melt pool or the height of the thin
wall. However, the deformation due to the high temperature gradi-
Thermal signature plays an important role in LAM. In-situ mon- ent and inhomogeneous materials properties will result in the
itoring and modeling are two complementary methods that have dimensional deviation, which is not included in the current geom-
been widely used to study temperature distribution and thermal etry control system. Further study can focus on reducing the defor-
behavior. The unknown emissivity of liquid metal, lacking a gener- mation by the feedback control of thermal signature and reheating
ous model or adequate experimental data, is the main obstacle to process. Currently, little investigation reported successfully devel-
the IR temperature monitoring. In the future, the emissivity data oped control scheme to migrate the thermal stress or acquire
can be accumulated and temperature-emissivity empirical models expected microstructure because it is difficult to in situ monitor
can be developed experimentally. The true temperature field of either of them directly. As the temperature is a process signature
melt pool can be restored based on emissivity model. In addition, that can correlate the parameters with the thermal stress and the
new algorithms can be proposed to extract true temperature and microstructure, thermal stress and microstructure can be con-
emissivity separately from the spectrum of melt pool. trolled based on the thermal signature. Controlling of the
Currently, the spatial resolution of thermometer is not sufficient microstructure based on the cooling rate of the melt pool was pro-
to provide thermal signature in the scale of metal powder particles posed by Farshidianfar et al [111]. In addition to the thermal signa-
and most numerical models simulated the temperature distribu- ture, other temperature-dependent process signatures can also be
tion on the macroscopic scale. The powder particles behaviors used in in-situ monitoring and feedback control such as thermal
and the temperature distribution in such scale influence the liquid diffusivity, spectrum signature and emissivity.
flow and the product quality [121]. The process of being partial
irradiated, heated, melting and consolidation of particles is not 6. Conclusions
completely clear. In further work, new monitoring sensors can be
employed to provide monitoring data in the scale of powder parti- In this paper, an overview of the research about the tempera-
cles. Measurement in the powder scale can provide detailed infor- ture and thermal phenomenon in Laser-based Additive Manufac-
mation about the interaction between laser and material and the turing is provided. Typical monitor equipments and thermal
formation mechanism of inner defects. models have been summarized and the relationship between
The material is deposited layer by layer to make a product. The process parameters, thermal behavior and product quality was
residual heat that accumulated by the last layers has an influence presented. The conclusions are listed as follows:
on the temperature distribution of the depositing material, which,
in turn, change the convection state of the melt pool. However, the (1) Great efforts have been made to in-situ monitor the process
geometry of models that study the thermal behavior and fluid flow temperature since the work of Griffith et al. [19]. Low
was mostly restricted to single track processing. Therefore, the accuracy, resulted from unknown emissivity and spatial
Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441 439
resolution still hampered the further development of this critical review, Proc. Inst. Mech. Eng., Part B: J. Eng. Manuf. 231 (1) (2017) 96–
117.
field. Methods to reduce the measurement errors and new
[18] J.N. Zalameda et al., Multiple sensor detection of process phenomena in laser
monitoring techniques were summarized, and we proposed powder bed fusion, Spie Commerc. + Sci. Sens. Imaging 9861 (2016) 986104.
some temperature-dependent variables which can be moni- [19] M.L. Griffith et al., Understanding thermal behavior in the LENS process,
tored and controlled. Mater. Des. 20 (2–3) (1999) 107–113.
[20] W. Ya, B. Pathiraj, S. Liu, 2D modelling of clad geometry and resulting thermal
(2) Physics-based thermal models include analytical model, cycles during laser cladding, J. Mater. Process. Technol. 230 (2016) 217–232.
semi-empirical lumped-parameter models and numerical [21] K. Zhang, X.F. Shang, W.J. Liu, Realtime measurement of temperature field
models. Numerical models are crucial to simulate the melt during direct laser deposition shaping, Adv. Mater. Res. 143–144 (2010) 521–
526.
pool behaviors and temperature distribution of the product. [22] A. Segerstark, J. Andersson, L.-E. Svensson, Evaluation of a temperature
Then, microstructure, residual stress and deformation can be measurement method developed for laser metal deposition, Sci. Technol.
predicted. Weld. Joining 22 (1) (2016) 1–6.
[23] J. Mathew et al., Integrating fiber fabry-perot cavity sensor into 3-D printed
(3) Thermal behavior is a factor that influences the eventual metal components for extreme high-temperature monitoring applications,
quality of the product. Parameters-thermal behavior- IEEE Sens. J. 17 (13) (2017) 4107–4114.
quality relationship is presented and some mechanisms that [24] S. Liu, P. Farahmand, R. Kovacevic, Optical monitoring of high power direct
diode laser cladding, Opt. Laser Technol. 64 (2014) 363–376.
influence the residual stress, deformation, microstructure [25] E. Rodriguez et al., Approximation of absolute surface temperature
and defects are discussed. This review can serve to resolve measurements of powder bed fusion additive manufacturing technology
quality issues by understanding the thermal behavior. 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.
Acknowledgement 220 (2013) 244–247.
[27] T. Duvaut, Comparison between multiwavelength infrared and visible
pyrometry: application to metals, Infrared Phys. Technol. 51 (4) (2008)
This work is jointly supported by the 973 Basic Research and 292–299.
Development Plan of China (No. 2011CB013402), the Fundamental [28] C.-D. Wen, C.-T. Lu, Suitability of multispectral radiation thermometry
emissivity models for predicting steel surface temperature, J. Thermophys.
Research Funds for the Central Universities (No. DUT16QY21), the
Heat Transfer 24 (3) (2010) 662–665.
Natural Science Foundation of China (No. 51305063), Specialized [29] J. Xing, F. Wu, An independent retrieval algorithm of multi-wavelength
Research Fund for the Doctoral Program of Higher Education (No. pyrometer, in: 2015 Fifth International Conference on Instrumentation and
2013004112002), and Collaborative Innovation Center of Major Measurement, Computer, Communication and Control (Imccc), 2015, pp.
943–946.
Machine Manufacturing in Liaoning. [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
References emissivity metallic measurands, Meas. Sci. Technol. 25 (2) (2014) 025010.
[32] P. Hagqvist et al., Emissivity compensated spectral pyrometry—algorithm and
[1] J. Song, Research progress of laser cladding forming technology, J. Mech. Eng. sensitivity analysis, Meas. Sci. Technol. 25 (2) (2014) 025011.
46 (14) (2010) 29. [33] H. Köhler, C. Thomy, F. Vollertsen, Contact-less temperature measurement
[2] A. Uriondo, M. Esperon-Miguez, S. Perinpanayagam, The present and future of and control with applications to laser cladding, Weld. World 60 (1) (2015) 1–
additive manufacturing in the aerospace sector: a review of important 9.
aspects, Proc. Inst. Mech. Eng., Part G: J. Aerosp. Eng. 229 (11) (2015) 2132– [34] I. Zhirnov et al., New approach of true temperature restoration in optical
2147. diagnostics using IR-camera, J. Therm. Spray Technol. 26 (4) (2017) 648–660.
[3] I. Yadroitsev, P. Krakhmalev, I. Yadroitsava, Selective laser melting of Ti6Al4V [35] T. Dreischuh et al., Improved laser metal deposition (LMD) of nickel base
alloy for biomedical applications: temperature monitoring and superalloys by pyrometry process control, in: Proc Spie, 2010, 7751, pp.
microstructural evolution, J. Alloy. Compd. 583 (2014) 404–409. 775123-775123-9.
[4] M.L. Griffith et al., Understanding the microstructure and properties of [36] G. Muvvala, D. Patra Karmakar, A.K. Nath, Online monitoring of thermo-cycles
components fabricated by Laser Engineered Net Shaping (LENS), in: Solid and its correlation with microstructure in laser cladding of nickel based super
Freeform and Additive Fabrication-2000, 2000. 625, pp. 9–20. alloy, Opt. Lasers Eng. 88 (2017) 139–152.
[5] W.E. Frazier, Metal additive manufacturing: a review, J. Mater. Eng. Perform. [37] J.B. Lei, Z. Wang, L.F. Liu, Design of forming shape measurement system for
23 (6) (2014) 1917–1928. laser molten pool in laser fabricating, Appl. Mech. Mater. 37–38 (2010) 327–
[6] W. Liu, J.N. DuPont, Fabrication of functionally graded TiC/Ti composites by 330.
laser engineered net shaping, Scripta Mater. 48 (9) (2003) 1337–1342. [38] H. Krauss, T. Zeugner, M.F. Zaeh, Thermographic process monitoring in
[7] J.J. Beaman et al., Laser sintering of layers of powder to build up a part – with powderbed based additive manufacturing, in: 41st Annual Review of Progress
computer controlled treatment and handling to produce high bulk density, in Quantitative Nondestructive Evaluation, vol. 34, 2015, 1650, pp. 177–183.
Univ Texas System (Texa-C), Univ Texas System (Texa-C). [39] M. Doubenskaia, A. Domashenkov, I. Smurov, Study of the laser melting of
[8] S.A. Khairallah, A. Anderson, Mesoscopic simulation model of selective laser pre-deposited intermetallic TiAl powder by comprehensive optical
melting of stainless steel powder, J. Mater. Process. Technol. 214 (11) (2014) diagnostics, Surf. Coat. Technol. 321 (2017) 118–127.
2627–2636. [40] W. Devesse et al., High resolution temperature estimation during laser
[9] D. De Baere et al., Spectroscopic monitoring and melt pool temperature cladding of stainless steel, in: Laser Assisted Net Shape Engineering 9
estimation during the laser metal deposition process, J. Laser Appl. 28 (2) International Conference on Photonic Technologies Proceedings of the Lane
(2016) 022303. 2016, 2016, 83, pp. 1253–1260.
[10] G. Bi et al., Identification and qualification of temperature signal for [41] G. Bi et al., Development and qualification of a novel laser-cladding head with
monitoring and control in laser cladding, Opt. Lasers Eng. 44 (12) (2006) integrated sensors, Int. J. Mach. Tools Manuf. 47 (3–4) (2007) 555–561.
1348–1359. [42] G. Bi et al., Investigation on the direct laser metallic powder deposition
[11] M. Grasso, B.M. Colosimo, Process defects and in situ monitoring methods in process via temperature measurement, Appl. Surf. Sci. 253 (3) (2006) 1411–
metal powder bed fusion: a review, Meas. Sci. Technol. 28 (4) (2017) 044005. 1416.
[12] H. Wang, Materials’ fundamental issues of laser additive manufacturing for [43] I. Smurov et al., Optical monitoring in laser cladding of Ti6Al4V, J. Therm.
high-performance large metallic components, Acta Aeron. Astron. Sin. 35 (10) Spray Technol. 21 (6) (2012) 1357–1362.
(2014) 2690–2698. [44] I. Smurov, M. Doubenskaia, A. Zaitsev, Comprehensive analysis of laser
[13] Z. Zhao et al., Evolution of plastic deformation and its effect on mechanical cladding by means of optical diagnostics and numerical simulation, Surf.
properties of laser additive repaired Ti64ELI titanium alloy, Opt. Laser Coat. Technol. 220 (2013) 112–121.
Technol. 92 (2017) 36–43. [45] M. Doubenskaia et al., Complex analysis of elaboration of steel–TiC
[14] E.R. Denlinger, J. Irwin, P. Michaleris, Thermomechanical modeling of additive composites by direct metal deposition, J. Laser Appl. 25 (4) (2013) 042009.
manufacturing large parts, J. Manuf. Sci. Eng.-Trans. ASME 136 (6) (2014) [46] G.J. Marshall, S.M. Thompson, N. Shamsaei, Data indicating temperature
01007. response of Ti-6Al-4V thin-walled structure during its additive manufacture
[15] G. Tapia, A. Elwany, A review on process monitoring and control in metal- via Laser Engineered Net Shaping, Data Brief 7 (2016) 697–703.
based additive manufacturing, J. Manuf. Sci. Eng. 136 (6) (2014) 060801. [47] G.J. Marshall et al., Understanding the microstructure formation of Ti-6Al-4V
[16] M. Mani et al., A review on measurement science needs for real-time control during direct laser deposition via in-situ thermal monitoring, Jom 68 (3)
of additive manufacturing metal powder bed fusion processes, Int. J. Prod. (2016) 778–790.
Res. 55 (5) (2016) 1400–1418. [48] C. Zhang, L. Li, A. Deceuster, Thermomechanical analysis of multi-bead pulsed
[17] B. Schoinochoritis, D. Chantzis, K. Salonitis, Simulation of metallic powder laser powder deposition of a nickel-based superalloy, J. Mater. Process.
bed additive manufacturing processes with the finite element method: a 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 [80] V. Manvatkar, A. De, T. DebRoy, Spatial variation of melt pool geometry,
distribution and stress field in single- and multi-track laser cladding by a peak temperature and solidification parameters during laser assisted
high-power direct diode laser, Opt. Laser Technol. 63 (2014) 154–168. additive manufacturing process, Mater. Sci. Technol. 31 (8) (2014)
[50] E.R. Denlinger et al., Effect of inter-layer dwell time on distortion and residual 924–930.
stress in additive manufacturing of titanium and nickel alloys, J. Mater. [81] I. Tabernero et al., Modelling of energy attenuation due to powder flow-laser
Process. Technol. 215 (2015) 123–131. beam interaction during laser cladding process, J. Mater. Process. Technol.
[51] W. Devesse, D. De Baere, P. Guillaume, High resolution temperature 212 (2) (2012) 516–522.
measurement of liquid stainless steel using hyperspectral imaging, Sensors [82] S.A. Khairallah et al., Laser powder-bed fusion additive manufacturing:
(Basel) 17(1) (2017). physics of complex melt flow and formation mechanisms of pores, spatter,
[52] M. Doubenskaia, P. Bertrand, I. Smurov, Optical monitoring of Nd:YAG laser and denudation zones, Acta Mater. 108 (2016) 36–45.
cladding, Thin Solid Films 453–454 (2004) 477–485. [83] F. Vásquez, J.A. Ramos-Grez, M. Walczak, Multiphysics simulation of laser–
[53] H. Krauss, T. Zeugner, M.F. Zaeh, Layerwise monitoring of the selective laser material interaction during laser powder deposition, Int. J. Adv. Manuf.
melting process by thermography, Phys. Proc. 56 (2014) 64–71. Technol. 59 (9–12) (2011) 1037–1045.
[54] S. Moylan et al., Infrared thermography for laser-based powder bed fusion [84] T. Mukherjee, W. Zhang, T. DebRoy, An improved prediction of residual
additive manufacturing processes, in: 40th Annual Review of Progress in stresses and distortion in additive manufacturing, Comput. Mater. Sci. 126
Quantitative Nondestructive Evaluation: Incorporating the 10th International (2017) 360–372.
Conference on Barkhausen Noise and Micromagnetic Testing, vols. 33a & 33b, [85] S. Safdar et al., An anisotropic enhanced thermal conductivity approach for
2014, 1581, pp. 1191–1196. modelling laser melt pools for Ni-base super alloys, Appl. Math. Model. 37 (3)
[55] U. Hassler et al., In-situ monitoring and defect detection for laser metal (2013) 1187–1195.
deposition by using infrared thermography, Phys. Proc. 83 (2016) 1244–1252. [86] G. Yu et al., On the role of processing parameters in thermal behavior, surface
[56] Q. Shi et al., Effects of laser processing parameters on thermal behavior and morphology and accuracy during laser 3D printing of aluminum alloy, J. Phys.
melting/solidification mechanism during selective laser melting of TiC/ D: Appl. Phys. 49(13) (2016).
Inconel 718 composites, Opt. Laser Technol. 84 (2016) 9–22. [87] M. Xia et al., Porosity evolution and its thermodynamic mechanism of
[57] J.P. Kruth et al., Feedback control of selective laser melting, Virtual and Rapid randomly packed powder-bed during selective laser melting of Inconel 718
Manufacturing (2008) 521–527. alloy, Int. J. Mach. Tools Manuf 116 (2017) 96–106.
[58] L. Tang, R.G. Landers, Melt pool temperature control for laser metal [88] P. Michaleris, Modeling metal deposition in heat transfer analyses of additive
deposition processes-part I: online temperature control, J. Manuf. Sci. Eng.- manufacturing processes, Finite Elem. Anal. Des. 86 (2014) 51–60.
Trans. ASME 132 (1) (2010) 011010. [89] X. Zhao et al., Numerical modeling of the thermal behavior and residual stress
[59] L. Song, J. Mazumder, Feedback control of melt pool temperature during laser in the direct metal laser sintering process of titanium alloy products, Addit.
cladding process, IEEE Trans. Control Syst. Technol. 19 (6) (2011) 1349–1356. Manuf. 14 (2017) 126–136.
[60] L. Song et al., Control of melt pool temperature and deposition height during [90] P. Peyre et al., Analytical and numerical modelling of the direct metal
direct metal deposition process, Int. J. Adv. Manuf. Technol. 58 (1–4) (2011) deposition laser process, J. Phys. D Appl. Phys. 41 (2) (2008) 025403.
247–256. [91] T. Mukherjee et al., Mitigation of thermal distortion during additive
[61] D. Rosenthal, The Theory of Moving Source of Heat and Its Application to manufacturing, Scripta Mater. 127 (2017) 79–83.
Metal Treatments, 1946, 68. [92] H. Runnemalm, S. Hyun, Three-dimensional welding analysis using an
[62] T.W. Eagar, N.S. Tsai, Temperature-fields produced by traveling distributed adaptive mesh scheme, Comput. Methods Appl. Mech. Eng. 189 (2) (2000)
heat-sources, Weld. J. 62 (12) (1983) S346–S355. 515–523.
[63] A. Vasinonta, J.L. Beuth, M.L. Griffith, A process map for consistent build [93] E.R. Denlinger, J.C. Heigel, P. Michaleris, Residual stress and distortion
conditions in the solid freeform fabrication of thin-walled structures, J. modeling of electron beam direct manufacturing Ti-6Al-4V, Proc. Inst.
Manuf. Sci. Eng.-Trans. ASME 123 (4) (2001) 615–622. Mech. Eng. Part B-J. Eng. Manuf. 229 (10) (2015) 1803–1813.
[64] A. Vasinonta, J. Beuth, M. Griffith, Process maps for laser deposition of thin- [94] E. Kundakcioglu, I. Lazoglu, S. Rawal, Transient thermal modeling of laser-
walled structures, in: Solid Freeform Fabrication Proceedings, August 1999, based additive manufacturing for 3D freeform structures, Int. J. Adv. Manuf.
1999, pp. 383–391. Technol. 85 (1–4) (2015) 493–501.
[65] J. Gockel, N. Klingbeil, S. Bontha, A closed-form solution for the effect of free [95] C. Li et al., A multiscale modeling approach for fast prediction of part
edges on melt pool geometry and solidification microstructure in additive distortion in selective laser melting, J. Mater. Process. Technol. 229 (2016)
manufacturing of thin-wall geometries, Metall. Mater. Trans. B 47 (2) (2015) 703–712.
1400–1408. [96] Y.W. Zhang, A. Faghri, Thermal modeling of selective area laser deposition of
[66] C. Doumanidis, Y.M. Kwak, Geometry modeling and control by infrared and titanium nitride on a finite slab with stationary and moving laser beams, Int.
laser sensing in thermal manufacturing with material deposition, J. Manuf. J. Heat Mass Transf. 43 (20) (2000) 3835–3846.
Sci. Eng.-Trans. ASME 123 (1) (2001) 45–52. [97] L. Hao et al., Three-dimensional numerical simulation of transient
[67] P.M. Sammons, D.A. Bristow, R.G. Landers, Height dependent laser metal temperature field and coating geometry in powder feeding laser cladding,
deposition process modeling, J. Manuf. Sci. Eng. 135 (5) (2013) 054501. Chin. J. Lasers 40 (12) (2013) 1203007.
[68] Q. Wang et al., Physics-based multivariable modeling and feedback [98] Y. Lee, D.F. Farson, Simulation of transport phenomena and melt pool shape
linearization control of MeltPool geometry and temperature in directed for multiple layer additive manufacturing, J. Laser Appl. 28 (1) (2016) 012006.
energy deposition, J. Manuf. Eng.-Trans. ASME 139 (2) (2017) 021013. [99] H. Ki, P.S. Mohanty, J. Mazumder, Modeling of laser keyhole welding: part I.
[69] X.Q. Cao, B. Ayalew, Control-oriented MIMO modeling of laser-aided powder Mathematical modeling, numerical methodology, role of recoil pressure,
deposition processes, Am. Control Conf. (Acc) 2015 (2015) 3637–3642. multiple reflections, and free surface evolution, Metall. Mater. Trans. a-Phys.
[70] Z. Zhang, P. Farahmand, R. Kovacevic, Laser cladding of 420 stainless steel Metall. Mater. Sci. 33(6) (2002) 1817–1830.
with molybdenum on mild steel A36 by a high power direct diode laser, [100] M.H. Farshidianfar, A. Khajepour, A.P. Gerlich, Effect of real-time cooling rate
Mater. Des. 109 (2016) 686–699. on microstructure in Laser Additive Manufacturing, J. Mater. Process.
[71] D. Gu, P. Yuan, Thermal evolution behavior and fluid dynamics during laser Technol. 231 (2016) 468–478.
additive manufacturing of Al-based nanocomposites: underlying role of [101] J. Beuth, N. Klingbeil, The role of process variables in laser-based direct metal
reinforcement weight fraction, J. Appl. Phys. 118 (23) (2015) 233109. solid freeform fabrication, Jom 53 (9) (2001) 36–39.
[72] S. Li et al., Melt-pool motion, temperature variation and dendritic [102] Y. Li, D. Gu, Parametric analysis of thermal behavior during selective laser
morphology of Inconel 718 during pulsed- and continuous-wave laser melting additive manufacturing of aluminum alloy powder, Mater. Des. 63
additive manufacturing: a comparative study, Mater. Des. 119 (2017) 351– (2014) 856–867.
360. [103] T. Mukherjee et al., Dimensionless numbers in additive manufacturing, J.
[73] Z. Gan et al., Numerical simulation of thermal behavior and multicomponent Appl. Phys. 121 (6) (2017) 064904.
mass transfer in direct laser deposition of Co-base alloy on steel, Int. J. Heat [104] M. Islam et al., Temperature profile and imaging analysis of laser additive
Mass Transf. 104 (2017) 28–38. manufacturing of stainless steel, Phys. Proc. 41 (2013) 835–842.
[74] S. Morville et al., 2D longitudinal modeling of heat transfer and fluid flow [105] I.A. Roberts et al., Experimental and numerical analysis of residual stresses in
during multilayered direct laser metal deposition process, J. Laser Appl. 24(3) additive layer manufacturing by laser melting of metal powders, Key Eng.
(2012) 032008-1-9. Mater. 450 (2010) 461–465.
[75] H. Ju et al., Test and temperature field of finite element simulation about the [106] Z. Nie et al., Experimental study and modeling of H13 steel deposition using
effect of scanning speed on 304 stainless layer’s properties by laser cladding. laser hot-wire additive manufacturing, J. Mater. Process. Technol. 235 (2016)
Mater. Res. Innov. 19(sup8) (2015) S8-9-S8-13. 171–186.
[76] R. Martukanitz et al., Toward an integrated computational system for [107] T. Furumoto et al., Study on deformation restraining of metal structure
describing the additive manufacturing process for metallic materials, Addit. fabricated by selective laser melting, J. Mater. Process. Technol. 245 (2017)
Manuf. 1–4 (2014) 52–63. 207–214.
[77] L.-E. Lindgren et al., Simulation of additive manufacturing using coupled [108] J.X. Fang et al., The effects of solid-state phase transformation upon stress
constitutive and microstructure models, Addit. Manuf. 12 (2016) 144–158. evolution in laser metal powder deposition, Mater. Des. 87 (2015) 807–814.
[78] A.J. Dunbar et al., Experimental validation of finite element modeling for laser [109] D. Herzog et al., Additive manufacturing of metals, Acta Mater. 117 (2016)
powder bed fusion deformation, Addit. Manuf. 12 (2016) 108–120. 371–392.
[79] M.F. Gouge et al., Modeling forced convection in the thermal simulation of [110] Y.-J. Liang et al., Experimental optimization of laser additive manufacturing
laser cladding processes, Int. J. Adv. Manuf. Technol. 79 (1–4) (2015) 307– process of single-crystal nickel-base superalloys by a statistical experiment
320. design method, J. Alloy. Compd. 697 (2017) 174–181.
Z. Yan et al. / Optics and Laser Technology 106 (2018) 427–441 441
[111] M.H. Farshidianfar, A. Khajepour, A. Gerlich, Real-time control of [116] Y.L. Hu et al., Effect of heat input on cracking in laser solid formed DZ4125
microstructure in laser additive manufacturing, Int. J. Adv. Manuf. Technol. superalloy, Opt. Laser Technol. 86 (2016) 1–7.
82 (5–8) (2015) 1173–1186. [117] L. Yan et al., Simulation of cooling rate effects on Ti–48Al–2Cr–2Nb crack
[112] S.K. Everton et al., Review of in-situ process monitoring and in-situ formation in direct laser deposition, Jom 69 (3) (2016) 586–591.
metrology for metal additive manufacturing, Mater. Des. 95 (2016) 431– [118] D. Salehi, M. Brandt, Melt pool temperature control using LabVIEW in Nd:
445. YAG laser blown powder cladding process, Int. J. Adv. Manuf. Technol. 29 (3–
[113] H. Ali et al., In-situ residual stress reduction, martensitic decomposition and 4) (2005) 273–278.
mechanical properties enhancement through high temperature powder bed [119] W. Devesse et al., Hardware-in-the-loop control of additive manufacturing
pre-heating of Selective Laser Melted Ti6Al4V, Mater. Sci. Eng., A 695 (2017) processes using temperature feedback, J. Laser Appl. 28 (2) (2016) 022302.
211–220. [120] X.Q. Cao, B. Ayalew, Multivariable predictive control of laser-aided powder
[114] R. Lin et al., Numerical study of keyhole dynamics and keyhole-induced deposition processes, Am. Control Conf. (Acc) 2015 (2015) 3625–3630.
porosity formation in remote laser welding of Al alloys, Int. J. Heat Mass [121] M.J. Matthews et al., Denudation of metal powder layers in laser powder bed
Transf. 108 (2017) 244–256. fusion processes, Acta Mater. 114 (2016) 33–42.
[115] W.E. King et al., Observation of keyhole-mode laser melting in laser powder- [122] L. Tang, R.G. Landers, Melt pool temperature control for laser metal
bed fusion additive manufacturing, J. Mater. Process. Technol. 214 (12) deposition processes-part II: layer-to-layer temperature control, J. Manuf.
(2014) 2915–2925. Sci. Eng.-Trans. ASME 132(1) (2010).