Additive Manufacturing: Balaji Soundararajan, Daniele Sofia, Diego Barletta, Massimo Poletto

Additive Manufacturing 47 (2021) 102336
Contents lists available at ScienceDirect
Additive Manufacturing
journal homepage: www.elsevier.com/locate/addma
Review
Review on modeling techniques for powder bed fusion processes based on

physical principles
Balaji Soundararajan a, *, Daniele Sofia a, b, Diego Barletta a, Massimo Poletto a
a
Department of Industrial Engineering, University of Salerno, Via Giovanni Paolo II, 132 – 84084 Fisciano, Salerno, Italy
b
Sense Square Srl, Piazza Vittorio Emanuele 10, Penta 84084, Salerno, Italy
A R T I C L E I N F O A B S T R A C T
Keywords: Powder bed fusion is one of the most common types of additive manufacturing processes using granular mate
Additive manufacturing rials. In this field, improvements in the processing of materials and process repeatability with desired part quality
Powder bed Fusion are needed to allow the implementation of powder bed fusion process at full scale across various industrial
Numerical simulations
sectors. The microstructure and the thermo-mechanical properties of the components vary according to the
Heat transfer
process parameters adopted in the manufacturing process. Physically accurate modeling of the powder bed
fusion process can play a vital role in the qualification of components and materials and also can help in pre
dicting the experimental observations. This paper aims to provide a comprehensive review on the state of the art
referred to numerical simulation of powder bed fusion, by covering both mesh based approaches such as finite
element modeling, finite volume modeling and mesh free methods such as the discrete element modeling, the
lattice Boltzmann method, the optimal transportation method and the smoothed particle hydrodynamics
approach. Combined approaches of different methods are discussed as well. A case study comparing two different
types of models for a similar type of problem has been provided. Moreover, means of experimental techniques
used to validate the numerical models are briefly discussed.
1. Introduction ideal for low volume, high value and user specific industries such as
aerospace, automotive and medical [2]. Although there are lots of ad
The term Additive Manufacturing (AM) refers to manufacturing of vantages in using this process, Quality Assurance (QA) of the compo
parts by fusing materials layer by layer sometimes also referred as rapid nents produced by AM processes seems to be a major challenge [3].
prototyping, 3D printing or freeform fabrication. Although there are Despite continuous technological advancements in the field of AM, the
different types of AM processes, Powder Bed Fusion (PBF) type is one of lack of process repeatability acts as a major barrier for the industry-wide
the widely used in industries as a wide spectrum of materials including implementation.
polymers, metals and ceramics can be used [1]. In PBF, the feedstock PBF includes complex phenomena involving mechanisms such as
material is in the form of powder and parts are built by sintering or heat absorption, high thermal gradients, local melting and solidification
melting the powder bed layer by layer using Laser or electron beam as of particles, phase change and Marangoni convection which are not yet
energy source. Fig. 1 shows the basic working principle of a Laser based fully understood [4]. As a result, the parts produced by PBF may have
PBF process. A rolling pin prepares the powder bed in the building poor thermal and mechanical characteristics compared to industrial
chamber and the laser system heats the cross section of the specimen to standards or may involve difficulty in estimating distortion behavior
be printed. The input to the laser scanner is a Stereolithography file [5]. Therefore, understanding these physical phenomena included in
(STL) which contains the Computer Aided Design (CAD) model to be PBF is vital to overcome the related problem and attain a successful
printed. Excess powder is collected in the reservoir and reused. Once this manufacturing process. The formation of solidified droplet of material
layer is fused, the rolling pin prepares the next layer. Selective Laser (ball), also known as balling, is one of the common defects during PBF
sintering (SLS), Selective Laser Melting (SLM) and Electron Beam process which occurs when the melt pool solidifies into such shapes
Melting (EBM) are some of the widely used PBF processes. The design instead of layers [6]. There are three detrimental effects of the balling
freedom and part optimization strategies offered by this PBF process is phenomenon in PBF process: (1) increased surface roughness; (2)
* Corresponding author.
E-mail address: bsoundararajan@unisa.it (B. Soundararajan).
https://doi.org/10.1016/j.addma.2021.102336
Received 7 April 2021; Received in revised form 2 August 2021; Accepted 15 September 2021
Available online 21 September 2021
2214-8604/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
B. Soundararajan et al. Additive Manufacturing 47 (2021) 102336
of the part. As a result, roughness and poor surface characteristics

represent a major defect in such processes [17,18]. ‘Stair-stepping’ ef
fect due to layer-by-layer production and roughness of the finished part
are two different issues faced in PBF parts. Part orientation during
construction and balling effects also influences the surface quality and
the kind of defects [7,19].
Another important issue to be considered in the PBF process is the
powder reusability. In fact, in the process, only a small portion of the
powder used goes in into the manufactured part, in spite of the high cost
of the powder, requiring a proper engineering to combine the proper
particle size distribution ad flowability. Therefore, the possibility to
reuse powder directly affects the affordability of the manufacturing
process. Several authors have studied the effect of powder reusability on
final specimen mechanical and microstructural properties [20–22].
Popov et al. [20] concluded that recycling Ti-6Al-4 V powder causes
various defects in the final specimen manufactured by EBM, as tensile
Fig. 1. Sketch of a Laser based Powder bed fusion process: (a) roller for powder tests had a negative effect on elongation and area reduction. Also, fa
bed preparation; (b) moving mirror for the laser scanning over the powder bed tigue results shows that the specimen demonstrated a shorter lifetime
surface; (c) laser source; (d) powder storage used to feed the powder to the (cycles to breakage) than those of new powder. This may be mainly due
roller by the movement of the bottom piston; (e) building chamber hosting the to surface oxidation which results in lack of fusion in EBM. Chan
powder bed for the fusion action of the laser beam, the piston at the bottom
drasekar et al. [23] studied the powder spreadability using process log
lowers the bed to host new powder layers; (f) manufacturing specimen; (g)
data and concluded that the Inconel 718 powder agglomerates over
reservoir to collect the e powder in excess from the layer formation.
recycling whereas Ti-6Al-4 V powder tends to sinter quickly irrespective
of recycling. In the case of reuse of Ti-6Al-4 V powder in the EBM pro
cess, the oxygen content increased from 0.08 wt% to 0.19 wt% with
increase in powder reuse times from 0 to 21 [24]. The powder becomes
less spherical, and the PSD gets narrower as reuse times increases. Due to
long exposure of vacuum and high temperatures, moisture and satellite
particles are removed thereby improving the flowability.
From the above presented overview on the many issues affecting the
PBF processes, it appears that the process needs to be optimized to
reduce the defects of the produced parts. In order to minimize the cost
involved in trial and error experimental methods, a great help may
derive from the possibility of using process design tools and from the
proper characterization of materials, that may be associated to stan
dardized certifications [25]. Therefore, it is crucial to develop accurate
numerical models of PBF processes. A possible approach is to is to apply
to the description of the process data driven models based on statistical
principles or artificial intelligence algorithm. These models can be very
Fig. 2. Number of publications with keywords "heat transfer" and "powder bed
fusion" in the literature of the last 10 years (source: Google scholar, useful in managing the process in condition of limited knowledge, but do
December 2020). not allow to fully understand how the different phenomena occurring in
the process interact with each other. For this reason, in this review,
increased pores formation between the balls; (3) interference of formed models based on first principles are considered. In physically based
spheres with the powder deposition blade [7]. Therefore, balling results models, the governing equations of the problem, more or less simplified
in irregular layer deposition, poor interlayer connection after fusion and version of mass, momentum and energy balances, which bear appro
finally low and irregular part density and quality. Internal porosity in priate constitutive equation of the material used and are usually stated
the manufactured parts is another critical defect of PBF process since it in the form of Partial Differential Equations (PDEs), are written first with
affects the fatigue performances and promotes the crack growth and the appropriate boundary conditions, which are used for the solution.
propagation in parts [8]. A lot of work investigating the porosity in PBF There are numerous papers published in recent years on thermal
process has been done [9–11]. Residual stresses in PBF processes are due modeling of PBF process, especially in the last ten years. In particular,
to strong thermal gradients as a result of point heating and the fast there has been a 3600% increase in number of papers on this topic from
cooling down of molten top layers. Furthermore, repeated heating and the year 2010–2020 with an increasingly growing number of the year
cooling cycles occur during the process due to the laser movement over throughout, as shown in Fig. 2. A detailed review on existing types of
the surface, which also gives rise to residual stresses due to different numerical simulations should compare mesh based continuum ap
local thermal deformation. Cracking occurs where, in the thermal his proaches, such as Finite Element Method (FEM), Finite Volume Method
tory of the material, the tensile stresses are higher than the ultimate (FVM) [26], with discrete mesh free approaches, such as Discrete
strength of solid material [12,13]. When residual stresses exceed the Element Method (DEM), Lattice Boltzmann Method (LBM), Optimal
adhesion ability between two layers, delamination occurs [14]. The PBF Transportation Method (OTM) and Smoothed Particle Hydrodynamics
produced parts may also exhibit dimensional, geometric and surface (SPH). The FEM was first introduced by Turner et al. [27] and is a widely
inaccuracies and deviations. Both part shrinkage and volume enlarge used simulation technique in PBF process [28] for its simplicity and
ment are common phenomena [15]. A combination of warping and resource efficient characteristics, although it includes uncertainties as it
shrinkage, called the curling phenomena is also seen in some parts [16]. treats the powder bed as a continuum. Although the model can include
Curling occurs due to uneven shrinkage between top and bottom sur temperature-dependent material properties and phase changes to
faces of the part. Usually, surface and thermal treatments are required improve the model accuracy, it still has uncertainties as it treats the
for parts made using PBF methods as it affects the fatigue performances powder bed as a continuum and often neglects effects of hydrodynamics
such as surface tension which causes significant inaccuracies in
2
basis of the surface hit by the beam mostly determined by the illumi
nated spot size. The powder bed allows a deep penetration of the energy
source due to multiple scattering and reflection from the surface [37,
38]. After sintering/melting of a given layer, the energy source beam is
inactive while travelling allowing some time for cooling before starting
to work on new layer [33]. Heat escapes the system primarily through
radiation and convection from top surface. For SLS and SLM, the
building chamber is usually inert gas environment, whereas convection
is negligible in case of EBM as the build chamber is maintained in vac
Fig. 3. Main energy fluxes and contributions appearing in the heat balance uum pressure [39]. Due to heat conduction in powder bed, the absorbed
applied to the melt pool in a PBF processes [47]. heat energy is distributed according to the density, porosity and
reflectivity of the packed powder bed. The general energy balance
predicting melt pool behavior [29]. The DEM was first discussed by equation of the above system is described as follows (Fig. 3):
Cundall and Strack [30]. It is one of the least used yet promising tech
QL = QM + Qcond + Qconv + Qrad
nique for simulating PBF process, because this non continuum approach
well simulates the discrete nature of powder beds [31]. In FEM, a con where QL is the power supplied by the Laser, QM is the negligible
tinuum domain is broken down into a finite number of elements thereby contribution of sensible heat required to reach melting temperature of
forming a Two-Dimensional (2D) or a Three-Dimensional (3D) mesh. the material, Qcond is the power dissipated through conduction in
With this approach the research for a solution for a differential equation powder bed, Qconv is the power dissipated through convection and
in the space is reduced to the solution of a finite number of algebraic Qrad is the power dissipated through radiation from the powder bed
equations [32]. The FEM has become a widely used technique for pre surface.
dicting layer surface temperature, residual stresses, porosity, and During the process, as the heat source moves, the powder melts and
geometrical distortion of parts produced by PBF process [33]. Other forms a melt pool. The melt pool is usually in micrometer range and has
numerical simulations such as FVM and LBM methods are primarily used a short lifetime. The convective flux in the melt pool occurs due to (1)
to study the hydrodynamics of melt pool. Some of the commercial temperature dependent surface tension in the flow (Marangoni effect)
software used to solve are ANSYS, ABAQUS, COMSOL, FLUENT/CFX or (2) back pressure of the evaporating surface (3) buoyancy due to vari
custom-built codes. Monte Carlo (MC) method is also used to simulate ation of temperature dependent density [40]. Other factors such as fluid
heat absorption in the powder bed and ray tracing of energy source. The viscosity, when it is low, and gravity play a minor role. Thermal
DEM is a numerical approximation method for studying the behavior of expansion in the heated material causes thermal stresses. On the one
particle interactions from a Newtonian mechanics perspective [34]. It is hand, high surface tension along with the material’s wetting ability
a widely used numerical methods to simulate granular media [35,36]. promotes the formation of stable melt pools and consequent smooth
Some of the commonly used DEM software are EDEM, MFix, LIGGGHTS surfaces. On the contrary, low wetting ability of the material produce
and LAMMPS. instable melt pools in which the surface tension induces the formation of
A common feature of numerical methods is to make calculations at single molten balls [41]. Marangoni effects force the fluid to move away
certain specified points known as nodes and then interpolating the re from the maximum temperature in the center of the melt pool, and
sults over entire domain such as volume or surface. Generally, an surface tension is lower, and increase the convective heat transport [42].
assumption is made on how the unknown is going to vary over space and In case of SLM where the melt pool temperatures are high, some of the
time, often based on conservation equations. After acquiring the results, material evaporates and results in high local gas pressures which drives
there are ways to check numerical correlation accuracy and minimiza the fluid motion away from the laser resulting in the “keyhole” forma
tion of errors. The ultimate aim of these studies applied to PBF is to tion, which allows Laser to penetrate into the material, by forming a
adjust the process parameters, such as the input power of the energy vapor capillary [43]. Apart from convection, volatilization of elements,
source, the scan strategy and the powder characteristics to achieve the evaporation and condensation of materials are also noted which impacts
desired qualities of products and to minimize manufacturing costs. the local and global material composition [44]. In case of laser melting
In this work the existing numerical simulation techniques used in at atmospheric pressure, ejected particles are trapped towards the melt
PBF processes are critically reviewed. The Section ’Numerical modeling’ pool due to the formation of flow from laser plume. In case of, laser
describes the heat transfer mechanism in typical PBF process, the nu heating at reduced chamber pressure, although the particles entrain
merical approach towards solving equation, the heat source models and ment is increased, but due to the expansion of laser plume, the particles
the material properties required. The Section ’Mesh based approaches’ are pushed away thereby giving rise to a stable melt pool, and reduced
critically reviews studies using FEM and FVM methods, two of the most porosity [45]. On the contrary in a high-pressure inert gas environment,
used methods. The Section ’Mesh free methods’ critically reviews using Argon, the effect of pressure increases the temperature in the melt
studies using DEM, LBM, SPH and OTM models as well as other ’com pool and in the vapor plume generated by the laser, producing higher
bined’ models. Section 5 includes a case study on comparing two spatter and ionization of metal vapors, which result in low surface
different modeling approaches for a given material. The following sec smoothness and continuity. However, using Helium, for its high thermal
tion is dedicated to the understanding of the effects of process param conductivity and diffusivity, it was possible to obtain a good surface
eters. Furthermore, the means for the experimental validation of finish even at a higher scan speed, with a positive effect in the part
numerical models is provided in a separate section. The last section building rate [46].
contains summary of the literature. When a new powder layer is spread on top of existing layers, the
previous layer characteristics play an important role while processing
2. Numerical modeling the new layer. In general, a high relative density with high powder
flowability is desired with properties related to particle size, surface
2.1. Heat transfer in PBF process topology and shape distribution.
During the PBF process, the powder bed is heated by a Laser or an

electron beam, where the photon or electron energy is transformed into 2.2. Numerical approach
thermal energy by absorption. The heating is done selectively on the
In PBF simulations, the process can be simulated for 3D [48–53] or
3
Table 1 cubes in the upper layer [69]. In DEM approach, each particle is treated
Various numerical simulations techniques applied to PBF processes. individually in an array of a number N of particles with predefined size
Model Applications References distribution [34,70]. Table 1 shows the different numerical approaches
used in PBF modelling:
Finite Element Methods Heat transfer and mechanical [71–73]
(FEM) characterization As described in Fig. 4, the modeling is divided into micro, meso and
Finite Volume Methods Heat transfer and melt pool dynamics [74–76] macroscopic models based on the variations in spatial scales of the
(FVM) problems. The particle sizes used in Powder Bed Fusion is usually in tens
Lattice Boltzmann Method Heat transfer and melt pool dynamics [77,78] of micrometers. The microscopic analysis includes beam penetration
(LBM)
Ray tracing Beam absorption and ray tracing [79]
and microstructure and grain structure evolution. The mesoscopic
Discrete Element Method Powder bed preparation and heat [70,80, analysis is the distribution of particles, analysis of formation and prop
(DEM) transfer 81] agation of pores by studying melt pool and solidification under heat. The
macroscopic thermo mechanical models resolve for the part level re
sidual stresses and deformations. Computational efforts required to
2D [54–56]. Although most authors have used 3D for a more realistic
implement depends upon the processing power of the computer, as
simulation of PBF process, some authors have used 2D to simulate single
sumptions such as 2D or 3D, linear or nonlinear material properties and
particle layers as it involves lesser computational effects. Conduction is
level of accuracy required. In case of mesh-based approach, each
the major form of heat transfer whereas radiation from the top surface
element size is in µm range given the specimen dimensions in few mm or
has been neglected in some studies [48,57–59] for the sake of simplicity,
cm and each time-step is in µs range given the entire simulation time of
especially for materials involving lower temperatures. Some authors
few minutes to hours. A lot of authors use single or few track or single or
have included radiation as well [55,56,60–62]. Gravity effects are usu
few layer simulation to study most of the process parameters.
ally considered by applying force loads to elements [63]. Gravity models
can track nodal displacements especially in over hanging structures.
Effect of phase change can be modeled by including latent heat ac 2.3. Heat source modeling
counting for material melting [64–66]. With phase changes, the material
properties have to be modified in the model. The powder particles can be The thermal and optical properties of the powder bed, along with the
modeled as a continuum property [67,68] or can be modeled explicitly laser/heat source characteristics influences the heat transfer in powder
as particles, e.g. in the form of cubes stacked so that they are separated bed and corresponding melt pool characteristics, especially in the case of
within each layer but are connected to 4 cubes in the lower layer and 4 SLS and SLM. The incident heat input load can be modeled as heat flux or
temperature. Most authors have used heat flux as input load [50,82–84],
Fig. 4. Different numerical approaches to simulations used in PBF methods classified according to various scales and modelling methods.
4
Fig. 5. Distribution of heat flux sources used to simulate thermal energy inputs due to the laser beam in the modeling the physics of the melt pool: (a) Surface
Gaussian (b) Volumetric Gaussian and (c) Egg-shaped Volumetric [85].
only (Fig. 5).

When the laser hits the powder bed surface, internal reflections of the
rays occur inside the bed and melt pool and hence, the depth of pene
tration of rays is comparable to the layer thickness. Multiple scattering
causes the absorptivity of the powder bed to be more than the evaluated
on a single layer at the surface [86]. Therefore, assuming the absorbed
energy to be limited to the surface instead of considering the absorption
in the volume of powder bed significantly reduces the accuracy of the
heat source models [85].
The Gaussian heat flux load with intensity I(r) at a radial distance r
from the center on the powder bed surface is given by [64]:
2AP 2r2
I(r) = exp( − )
πω2 ω2
where A is the surface absorptivity of the powder bed, P is the power of

the beam, ω is the radius where the local energy intensity decreases
from the intensity of the center by a factor of 1/e. Because of the simi
Fig. 6. Example of constant power density distribution pattern used to model larity of SLS/SLM processes with laser welding, some authors [87,88]
the laser energy input at the bed surface [54].
have adopted in their models surface heat sources and as the laser beam
can only penetrate up to several microns (20 µm for Nickel [89] and
which follows Gaussian volumetric intensity distribution, which is 65 µm for Titanium [90]). The egg shaped volumetric heat source in
nothing but the laser energy is distributed into volumetric form. Mishra troduces non-symmetrical heat distribution caused by the moving laser
et al. [85] analyzed and compared three different types of heat sources [91]. Contuzzi et al. [48] and Matsumoto et al. [53] have used constant
such as Gaussian Volumetric, surface Gaussian and Egg-shaped volumetric. circular power density as energy input applied to five elements in a cross
The volumetric heat sources give rise to melt pool temperatures, ge layout, as shown in Fig. 6. Few authors have used fixed temperature load
ometry and flow field that were significantly different from surface as input [89,92] mainly due to reduced computational costs. In the case
Gaussian. The volumetric heat sources also provided values closer to of DEM modeling of PBF, Gaussian Volumetric heat source [31,70] and
experimental results. The surface Gaussian results in very high surface fixed circular heat source [34] and even ray tracing methods [93] have
temperature due to the fact that total energy is absorbed in the top layer been used. Mirkoohi et al. [94] defined a new 3D semi elliptical which
5
Fig. 7. Modeling feature of the Laser PBF heat source: the most common inputs required and the different modeling approach of heat absorption.
Fig. 9. Gaussian volumetric heat source distributes laser power into the grey
shaded volume (Left) and ray tracing model where the heat is distributed into
packets of energy portions (rays) that are absorbed into powder bed (right).
bed porosity and surface areas [38].

Fig. 8. 2D illustration of optical penetration of a laser beam in a powder bed
Beer-Lambert model assumes the laser beam is mainly absorbed at the
and the coexistence of laser reflection and radiative exchange in a layer. “S”
denotes powder layer thickness [85].
surface and its intensity drops exponentially with depth of penetration
[95].
predicts melt pool geometry more realistically. I(z) = Io ∗ e− A(z)
In order to estimate the absorbed heat energy due to laser and ab
sorptivity of powder bed, there are three major types of numerical Where I(z) is variation of intensity in z direction (depth), Io is intensity
models as described in Fig. 7. of incident energy, A is the absorptivity of powder bed.
Radiation transfer model is a derived analytical solution from a ho Ray tracing RT: A more precise approach to depict the laser-particle
mogeneous continuum radiation transfer equation (RTE) using powder interaction is ray tracing methods, where the laser beam is divided
6
depends on the level of penetration among the particles. According to

experimental results by Fischer et al. [90], the OPD in spherical Nickel
alloy powder with size of 20 µm is measured to be 20 µm whereas for a
powder size of 50–75 µm, OPD is found to be 200 µm. For a 25 µm size
Titanium alloy powder, OPD is around 65 µm. For AlSl 316 L spherical
powder of size 45 µm, Foroozmehr et al. [58] interpolated the results for
OPD to be 170 µm. Yang [98] deduced that in case of Aluminum alloys,
as particles sizes increases, the penetration depth decreases. For polymer
powders, the penetration depth is 200–250 µm for an incident CO2 laser
with 1.8 W power [99].
2.4. Material properties
Accurate information on material properties is important in the

simulation of PBF processes. Due to elevated temperatures being
involved, temperature dependent material properties should be stated
along with the phase change. The major material properties affected are
density (Kg/m3), thermal conductivity (W/m K) and specific heat ca
pacity (J/Kg K). Masubuchi and Agapakis [100] state that the temper
ature dependent material property is a vital factor in the analysis of
thermal stresses.
Density is one of the most important thermo-physical properties as it
influences thermal conductivity [101] and laser absorptivity of the
powder bed [102]. Density is a temperature dependent property and
most of the times, the density of the final specimen is higher than the
density of the powder bed [48]. The relation between temperature and
density is considered to be linear by Yin et al. [62]. Studies with
changing densities are time consuming yet accurate [14,57,103]. Other
studies have used two different densities, one for the powder and the
other for solid state [49,54,55,62,63]. In these studies, bulk powder
density, including the interparticle voids, varies between 40% and 60%
of solid density at ambient conditions, which is also the case if the
Fig. 10. Temperature dependence of the most relevant material properties of powder bed is not mechanically compressed. The bulk powder density is
In718: (a) Thermal conductivity, k, and density, ρ, vs temperature (b) Specific measured by weighing the powder and dividing the mass by the volume
heat capacity, Cp , and mean thermal expansion coefficient, α, vs tempera
′
inside the sample holder [104]. Dong et al. [105] used the following
ture [116]. equation to describe the rate of variation of the powder bed density:
E
ρt+∆t = ρt + ∆t∙ρsolid ∙A∙exp(− )
into moving discrete energy portions that are traced in the domain. Each
R∙T
E
1 + ∆t∙A∙exp(− )
ray has some energy and when it hits the particle, some part of the en
R∙T
ergy is absorbed and some reflected (Fig. 8) [85]. This procedure con where ρsolid is the density of solid, E is the activation energy, R is the
tinues until the ray is considered to disappear either by reflecting outside universal gas constant and ρt is the density of powder bed obtained from
the powder bed or its energy becomes negligible after several internal previous time step. Some authors have calculated the powder bed den
reflections. The total energy in-coupling is the ratio between absorbed sities as follows [65,68]:
and total input energy and is influenced by the material, mixture ratio,
πρsolid
mean particle shape and size distribution and wavelength of the laser (in ρbed =
6
case of SLS/SLM). The energy absorbed by each particle is gathered
during the simulation and is used in the end to calculate total energy in where ρbed is the density of the powder bed and ρsolid is the density of the
coupling as well as energy absorbed. From this, we can deduce the solid.
absorbed energy as a function of depth in the powder bed. Thermal conductivity is also a temperature-dependent property and is
Lasers emit power in pulses and in some cases in continuous manner. defined as the rate at which heat is conducted through the body and
Wessels et al. [96] compared ray tracing method with Gaussian volu usually measured as Watt per meter Kelvin. Thermal conductivity of
metric methods (Fig. 9) and concluded that the former remarkably im powder bed is thermal conductivity increases with increase and lower
proves the accuracy of heat absorption and vaporization. The author than that of a continuum of solid made of the particle material. In fact,
integrated ray tracing method of irradiation into the Optimal Trans particles touches only on very small contact areas though which
formation Mesh-free (OTM) method to solve the surface tension effects. conductive heat transfer is effective. Other mechanisms of heat transfer
The author also concluded that in ray tracing method, the input heat by radiation or interstitial gas conduction and convection prove to be
energy is confined close to the irradiated surface whereas volumetric less effective than solid conduction. In order to account for these effects,
heat sources penetrate within the part which determine a lower esti usually in powder beds, an effective thermal conductivity is employed
mation of the vaporization and, therefore, recommends ray tracing for which is dependent on various factors such as the size and the shape of
mesh-free methods. particles, the solid volume fraction and the thermal conductivity of the
Optical Penetration Depth (OPD) is defined as the depth inside the interstitial gas [106]. Some authors have also proved that the powder
powder bed at which the intensity of the radiation falls to 1/e (or 37%) bed thermal conductivity is as function of the relative density of powder
of the original value [97]. The OPD is highly dependent on material or its porosity [102]. Higher thermal conductivity leads to lower tem
absorptivity. For materials with similar absorption, the value of OPD is peratures in the powder bed, which could be compensated by increasing
highly influenced by powder shape and size [90]. Absorbed energy the power of the energy source [107]. The effective thermal
7
material property which indicates the ability of the material to expand

under thermal gradient and influences the thermal strain. The CTE in
creases with increase in temperature and phase transition from solid to
liquid has slight effect on change of this value [119]. Jiang studied the
effect of CTE on carbon nano tube (CNT) reinforced Aluminum alloy
AlSi10Mg in LPBF process. It was found that the CTE value increased
with the addition of CNTs [120].
3. Mesh based approaches
3.1. Meshing
The quality of the simulation results in mesh based approaches very

much depends upon the characteristics of the mesh. A poor mesh quality
impacts the computational efficiency, increases the computational time
and may lead to unstable solutions [121,122]. Meshing can be either
Fig. 11. An example of FEM adaptive mesh. A Fine mesh used in the area of uniform or non-uniform. Adaptive meshing is a type of non-uniform
interest where the heat source is located and a coarse mesh is used in rest of the mesh scheme and one of the most used techniques used in FEM based
domain [128]. approaches. It is characterized by a mesh density that varies across
different regions [123]. Furthermore, non-uniform meshing can be static
conductivity of powder bed is also dependent on the neck formed be or dynamic. Static meshing does not change with time and may include
tween the particles during sintering and the diameter of the neck con both finer and coarser meshes. Dynamic meshes change with time. This
trols the heat transfer significantly. Larger the diameter, higher the technique is often used to provide finer meshes only where and when it
thermal conductivity and vice versa [108]. Gusarov et al. [109] calcu is needed. In the case of a moving heat source, the region where the fine
lated the contact thermal conductivity in powder bed for particles which mesh is deployed moves with the heat source, therefore the region left
has contact radius to particle radius ratio below 0.3. Usually, the solid by the heat source, with a fine mesh in previous time step, is restored to a
thermal conductivity for different temperatures is determined by ex coarse mesh. This approach combines high accuracy with low compu
periments. The powder bed effective thermal conductivity is then tational efforts, therefore it has been widely used [25,48,49,63,64,124,
calculated according to porosity and thermal conductivity of gas in pores 95,125–127]. An example of sample adaptive mesh is reported in Fig. 11
The following equation describes effective thermal conductivity of [128].
packed powder bed [101]: Studies with static non uniform meshing uses finer mesh for powder
bed and coarser mesh for metal substrate and the mesh does not change
Keff =
ρr ∙Ks (T)
/ as the heat source move [49,64,82]. A uniform mesh is easy to imple
1 + Ф∙Ks (T) Kg ment but requires a larger amount of calculations, hence longer
computational times [54,65,129].
where ρr is relative density of powder bed, Ks is the solid thermal con Gouge et al. [130] discussed meshing in multi scale modeling which
ductivity as a function of temperature T, Kg is the gas thermal conduc involves splitting the analysis into two different scales: small-scale
tivity and Ф is the empirical co-efficient which is equal to (local) and part-scale meshing. Small scale mesh involves construction
0.02∙102(0.7− ρr ) . Yuan et al. [110] proposed experiments to calculate of small 5 mm x 5 mm x 1 mm blocks. The mesh here has an element
thermal conductivities of polyamide 12 (PA12) powder for use in laser length and width equal to melt pool radius while depth is equal to single
sintering. It was noted that in loose form, thermal conductivity increased layer thickness and used adaptive meshing whereas part-scale meshing
with higher temperatures and in bulk form, thermal conductivity involves meshing for whole component where individual layers are
decreased with increasing temperatures. Few authors proposed accurate assorted together to form smallest element. Grouping more than a single
calculation of effective thermal conductivity calculation using DEM layer results in larger elements, fewer equations to solve and quicker
methods [111–114]. Moser et al. [115] proposed computation of results. Williams et al. [73] also suggested a part scale meshing by
effective thermal conductivity of finite powder bed depths specifically grouping layers for solving predicting residual stresses and distortions
for SLS applications using DEM simulations and used MFiX open source (Fig. 12). Other authors have also used similar grouping techniques
software. It is important to note that the analysis was done when a thin where they group layers of 50 µm each into reduced number of thicker
layer of powder is present on top of a solid already sintered surface with layers of 1 mm [49,131,132].
a very high temperature where radiation effects are important. An Element birth and death is another technique used to simulate the
example of the effect of temperature on the most significant physical deposition of powder layer on substrates and consequently on solidified
properties in PBF are reported in Fig. 10 [116]. particles during PBF process. The geometry of the whole powder bed at
Specific heat capacity is yet another temperature dependent property the end of the process including the model to be printed are meshed.
and is defined as the amount of heat required to increase the tempera During the simulation of the process, all the elements other than the first
ture of the one kilogram of material by one Kelvin. The specific heat active layer are deactivated or ’killed’ and the deactivated elements
capacity has a nonlinear relationship with temperature and phase contribute to zero heat transfer. Once the current layer is completed, the
change [107]. For the sake of simplicity, many authors have used a next layer is activated, similar to the process. Numerous studies have
linear interpolation or extrapolation [48,60–65]. In studies where the implemented this approach [73,133–135].
melt pool dynamics is analyzed, the specific heat capacity for solid and
liquid phase is defined [68,117]. 3.2. Finite element method
Latent heat is defined as the heat energy absorbed or released during
a phase change. The temperature of the material is unchanged during Thermo-mechanical modeling: During the PBF, the powder bed un
the process. Thermal conductivity and specific heat changes signifi dergoes transient temperature changes which is important for assessing
cantly during the melting of powders and while solidifying back to form the thermal stresses and associated residual stresses in the final
tracks [118]. component. There are two major types of thermo-mechanical analysis
Co-efficient of thermal expansion (CTE) is a temperature dependent used which are 1) coupled and 2) uncoupled analysis. Fig. 13 shows the
8
Fig. 12. Grouping layers into large simulated sections to reduce computational efforts: (a) actual geometry and mesh (b) bridge component showing division of
blocks [73].
Fig. 13. Flow chart of major steps in the application of a sequential Thermo-mechanical FEA model.
flow chart of sequential thermo-mechanical model: analysis, the displacements and temperatures fields are solved simulta
In the numerical simulation of PBF, the pre-processing procedure is neously. In this way the effect of the temperature on the displacements
the first step in which the user provides the input parameters such as and the displacements on the temperature can are mutually accounted
temperature-dependent material properties (density, specific heat, for. While thermal expansion of the material has effects on material
thermal conductivity, emissivity of powder bed, latent heat), process displacements, the effect of displacement on temperature is limited to
parameters, input Computer Aided Design (CAD) model, heat source radiation effects.
model (volumetric or surface), thermal boundary conditions, process Considering the heat conduction in a 3D volume, expressed by the
parameters and mesh scheme for FEM studies (mesh type, mesh strategy, Fourier’s law:
number of elements). Once the pre-processing is ready, the simulation ( ) ( ) ( )
δ kδT δ kδT δ kδT δT
for thermal analysis is done. In this step the partial differential equations + + + q = ρCp
δx δx δy δy δz δz δt
of the governing system are discretized for each element (mesh) and all
the equations are grouped to form a system of equations. The solution of
where k is the thermal conductivity (in an isotropic material is assumed
this equation is found in the post processing step where temperature
and it is uniform in all directions), T is the temperature, q is the rate of
results are found. In order to find the residual stresses and distortions, a
heat generation given to the system, ρ is the density of the material (a
mechanical analysis is needed. In case uncoupled analysis is adopted, for
further assumption is made on its constant value), Cp is the specific heat
each time increment a thermal analysis is performed first after which the
(also assumed to be constant) and t is the time. In order to consider the
output temperature field is used as an input for the following mechanical
phase change, changes in enthalpy (dH) can be introduced in the above
analysis. With uncoupled analysis [33], a widely used procedure, the
equation in place of the product Cp∙dT:
output of the mechanical analysis is not used to solve the temperature
field in the same step, therefore the number of equations is reduced and dH = Cp∙dT
the solution is faster. Instead, in case of coupled thermo-mechanical
The energy balance on the external surfaces can be written as fol
9
Fig. 14. Temperature history of SLM at specific points on the surface of Iron based powder when exposed to scanning pattern described in the small pane on the
right [136].
where α is the co-efficient of thermal expansion of material and Tref is the

reference temperature. The plastic strain is implemented by von-mises
yield criterion and the Prandtl-Ruess flow rule:
( )
f = σm − σy εq , T ≤ 0
where f is the yield function, σm is the von Mises’ stress, σy is the yield
stress, εq equivalent plastic strain.
Authors implementing FEA are discussed in summary.
3.3. FEA: simulation results
3.3.1. Temperature fields, residual stresses and distortions

Temperature field history of the layer can provide input for deter
mining thermal stresses, residual stresses and distortions. The temper
Fig. 15. Experimental measurement of residual stress as a function of the base
plate temperature in a metallic model made by selective laser melting pro ature profiles of the scanned areas in the layer undergo similar thermal
cess [142]. cycles as shown in Fig. 14 [136]:
Roberts [135] simulated temperature fields using FEA and found that
lows: laser scanning of additional layers accounts for peak temperatures in
( ) underlying layers. He also witnessed that the temperature of the bottom
δT ( ) most layer experienced slight yet steady raise in temperature as the
k + h(T − T0 ) + σb ε T4 − T0 4 = q
δn number of layers increased, which could account for the increase in
residual stresses during cooling once the building process is complete. As
where n is the normal vector to the surface, h is the heat transfer coef the laser input power increases with decreasing laser speed, the peak
ficient of the gas medium, T0 is the ambient temperature, σ is Stefan- temperature of the layer increases [60,62,137]. Zhang et al. [60] found
Boltzmann constant and ε is the surface emissivity. that preheating the layer can significantly increase the maximum layer
In the case of quasi-static mechanical analysis, the governing stress temperature allowing for less energy input.
equilibrium equation is written as follows [130]: Residual stresses measured in experiments are classified as Type 1
∇∙σ = 0 macroscopic stresses that acts on the overall geometry and causes global
distortions and type 2 microscopic stresses acting at the individual
where σ is the stress. The mechanical constitutive law for elastic granular level [138,139] and type 3 stresses are at atomic scale,
behavior is although type 1 stresses are often discussed in the scope of AM. FEA is
one of the most effective means to predict the residual stresses and
σ = C∙εe distortions [140]. As discussed before in the meshing section, authors
Coherently, the total strain εtotal, assuming small deformations and have used grouping techniques to merge multiple layers in single ele
thermo-elasto-plasticity of the material, is given as ments for the reduced computational efforts. Williams et al. [73] esti
mated that with this grouping method the calculated distortion were
εtotal = εe + εp + εT kept within 10% of the experimental results. Some researchers have
focused their studies on single layer and single beam line in order to
Where C is the fourth order isotropic material stiffness tensor, εe , εp , εT
understand relation between residual stresses and the distortions related
are the elastic, plastic and thermal strain vectors respectively.
to process parameters related to the energy source, such as, speed,
Thermal strain is computed as:
power, spot size [54,61,141]. This method provides insights on scan
εT = α (T − Tref ) strategy, and speed as well although multilayer effects are not included.
An important factor to be considered is the effect of the temperature of
10
Fig. 16. FEA model showing effects of pre-heating of base plate on residual stresses (y-axial Cauchy stress, MPa). (a) non-heated baseplate (b) heated baseplate. A
heated baseplate exhibits lower residual stress [131].
history, residual stress and warping and shrinkage at part scale. An in
crease of power gives rise to increase in shrinkage and warping which is
due to the increase of thermal strain at higher energy densities.
3.3.2. Melt pool

In order to model the powder melting and the subsequent solidifi
cation of the powder bed, conservation of mass, momentum and energy
needs to be considered. Melt pool ensures bonding of the particles within
the layer, as it also affects the porosity formations and volume defects.
The point which achieves the maximum temperature in the melt pool is
called ’center of melt pool’ and often it is present not exactly in the
center of the laser illuminated spot, but rather displaced backwards
according to the scan direction [117]. This offset between the two
centers increases with increasing beam power [67]. Generally, the size
of the melt pool is about 10% larger than that of the illuminated area
[146]. Gusarov et al. [117] concluded that the melt pool shape and its
cross section gets narrower and longer with increasing scan speed. They
also observed that the region behind the melt pool, which is made of
solidified material, exhibits a larger thermal conductivity than the ma
Fig. 17. Melt pool dimensions according to scan speed: 120 mm/s (Left) and
terial present in the other position around the melt pool that is in the
240 mm/s (right). The broken circles in the figure indicate the projections of
powder form. The increase in length of melt pool with the increasing
laser beam. λ denotes optical thickness, which is a function of layer thickness
and particle diameter [147]. scan speed is attributed to the fact that the cooling time of the molten
regions is only limitedly reduced by the speed increase [64]. However,
the depth of the melt pool decreases with the speed, since the interaction
powder bed, which will significantly reduce the residual stresses. Shiomi
time becomes shorter [62,64,137]. Loh et al. [51] in their model
et al. [142] found that residual stresses reduced by 40% when the base
considered the direct powder to solid transition, the evaporation of the
plate supporting the bed was heated up to 160 ◦ C (Fig. 15). Buchbinder
material and the overall top layer volume shrinkage. They found that,
[143] reduced distortions to almost zero mm by heating the base plate to
under the effect of laser heating, the molten material sinks thereby
250 ◦ C. These results are also validated through FEA by other authors
removing pores that are present initially between the powder particles.
[131,141,144] (Fig. 16). Also, uniform distribution of residual stresses
This reduces the powder layer volume in correspondence of the molten
can be achieved by using thicker base plates [145].
areas as shown by Gusarov et al. [147] in Fig. 17.
Shrinkage and warping in the parts produced by PBF process are
predicted using macro-scale based FEM based thermo-mechanical
models. The shrinkage due to powder fusion is due to material state 3.4. Finite volume methods
transformations from power to solid and subsequent thermal contraction
cooling down. Li et al. [119] studied the shrinkage and warping using Finite Element Methods are often used to simulate PBF processes.
modelling and experimental techniques. Polymer PA12 powder was The main disadvantage of FEM if compared with FVM or FDM is the
analyzed as it represents semi-crystalline polymer which exhibits implementation of volume shrinkage and the simulation of the deposi
shrinkage and warping when subjected to high thermal gradients. A tion of new layers [74]. In several studies using FEM, the effect of vol
model combining the heat transfer, thermo-mechanical and material ume shrinkage and vaporization effects are not considered [108,129,
crystallization kinetics is implemented to predict the temperature 148]. Khairallah et al. [149] simulated the melt flow fluid dynamics by
11
FVM method [75,76,151]. Also a study considering shrinkage of volume

due to evaporation based on conservation of mass and successive layer
deposition using FVM is proposed [74]. In this study, the new volume
occupied by the powder bed and the new density of the molten material
are inversely proportional to each other to simulate mass conservation.
The powder bed volume is updated after each time step. Powder bed
layer deformation is followed at each instant and empty cells are deac
tivated (Fig. 18).
3.5. Phase field method

Fig. 18. Powder bed volume updated after densification and volume shrinkage
in FVM approach. At each time step, bed volume is updated. Heat convection
Phase Field Method (PFM) is primarily used for studying micro
and radiation are included in the computational domain [74].
structural phenomena such as neck growth, solidification, phase trans
formation and densification. PFM is based on the theory of irreversible
FVM by employing the Volume of Fluid (VOF) method which resolves thermodynamics and is useful in studying time dependent evolving
the surface flow and is able to track the fluid-gas interface. Qiu et al. morphologies which is difficult to implement in traditional sharp-
[150] implemented the Navier Stokes momentum equations in which interface model [152]. Wang [153] utilized PFM to study solid state
the fluid dynamics is driven by interfacial forces including temperature sintering where the model includes various physical processes such as
dependent surface tension, the recoil pressure due to liquid evaporation rigid body translations, rotation of particles, grain growth and various
and the drag force of the liquid phase flowing in the surrounding porous diffusion mechanisms such as grain boundary diffusion, volume diffu
bed. In the energy equation, heat losses due to evaporation, convection sion and surface diffusion. A multi-scale model analyzing temperature
and radiation are also considered. Similarly, other authors have used distribution due to laser heating and subsequent phase field model was
Fig. 19. Solid state SLS of two 316 L stainless steel particles of 30 µm diameter. Input process parameters are Laser power of 21 W, spot size of 0.8 mm and scan
speed of 1 mm/s. (a) temperature profile predicted by the thermal model (b), (c) and (d) neck growth predicted by the PFM at the beginning of sintering at (4.5 s),
peak temperature (5 s) and final status (5.5 s) respectively. Scale bar on the three paned on the right denotes 10 µm [154].
Fig. 20. Laser power vs Neck size in laser sintering process. (a) thermal model predicted temperature histories in fixed positions for changing laser power (b)
predicted neck size after SLS for various input laser powers by a PF model [154].
12
Fig. 21. DEM overview (left) and variables associated with collision mechanics for spherical particles (right). Pi and Pj are two different colliding particles, vi and vj
are velocity vectors of colliding particles, f n and f s are normal and shear force, ̂ ̂ s are unit normal vectors. [34].
X n and X
achieved in the thermal model is from 1200 to 1500 K with a neck size of
7.9–19.6 µm, as shown in Fig. 20. The Arrhenius equation relates the
diffusion coefficient and temperature and this increase in neck size is
due to greater atomic diffusion rate resulting from a higher temperatures
[154].
A powder based three-dimensional heat transfer model followed by a
two-dimensional non-isothermal PF model for simulation of micro
structure in SLS was performed by Wang et al. [152]. The temperature
Fig. 22. Metaball: two particles of different radii move towards each other, information from the thermal model is used as an input to the 2D PF
advancing in time from left to right [155]. model to simulate the effect of layer wise sintering of 316 L stainless
steel powder on densification and porosity. A thin layer thickness
resulted in reduced porosity. Also, a narrow particle size distribution
means reduced powder volume and higher thermal conductivity. This
gives rise to a larger heating zone, full grain coalescence and reduced
part porosity.
4. Mesh free methods
4.1. Discrete element modeling
The DEM is a numerical approximation method capturing the way

particles interact and accounts for the discrete nature of the powder bed.
Steuben et al. [155] extended the DEM beyond elementary Newtonian
mechanics and included other physical interactions between particles,
demonstrating the possibility to simulate granular systems capturing
granular dynamics as well as energy inputs and conservation and
transfer involved in PBF. Hence, the term Multiphysics DEM or MDEM
for poly-dispersed particle system was proposed. In the description of
fine granulates or powders, a DEM simulation contains a large number of
Fig. 23. Average temperature of powder bed with time when a zig-zag pattern
of uniform laser scan is implemented [81].
particles, a subset of which may be “bonded,” or attached to each other
and some are unbonded. Both bonded and unbonded particles collide
and produce reaction forces. These forces contribute to changes in
studied by Zhang and Liao [154]. Here, a three-dimensional thermal
particle velocity and position. The steps performed in DEM imple
model was developed to study the temperature evolution and the tem
mentation are: collision detection, contact behavior of particles,
perature history is given as an input for the two-dimensional phase field
boundary conditions, bonding and time step integration as summarized
model to simulate microstructure evolution. The simulations are con
in Fig. 21. Due to the extensive computational efforts, DEM for PBF
ducted on two equal sized 316 L stainless steel particles subjected to
processes have not been used widely. Some authors have implemented
solid state sintering. Fig. 19a, shows the thermal profile from the ther
DEM in PBF but they do not include important parameters such as
mal model which is given as an input to the PFM. Fig. 19b, c and d shows
powder compaction by re-coating blade and thermally induced distor
the PF model predicted neck formation. It is understood that the sin
tions [93,156,157]. Steuben et al. [155] introduced the concept of
tering occurs at temperatures higher than half of the melting point.
’metaballs’, which refer to the representation of
The neck size increases as the input laser power increases. By
melting-bonding-solidification process (Fig. 22).
increasing the input energy from 17 to 25 W, the maximum temperature
Ganieriwala and Zohdi [81] implemented DEM for SLS considering
13
in a similar model accounted also for new particle contacts and particle
coarsening while sintering by implementing a coarsening model. Dayal
and Gambaryan-Roisman [31] studied heat transfer with moving heat
source in SLM using DEM and ray tracing approach to model laser
irradiation and heat absorption in powder bed (Fig. 24). The model was
developed considering conduction between particles and neglecting the
change of thermal conductivity of inert gas with pressure. Moser et al.
[115,162] used DEM to compute effective thermal conductivity and
effective radiative properties of powder bed for SLS applications. Xin
et al. [163] developed 3D numerical heat transfer in SLS using DEM and
implemented it using EDEM software. They also used Monte-Carlo ray
tracing method for laser irradiation of powder bed. They compared the
results with FEM and found that DEM simulations depicts the granular
Fig. 24. Temperature distribution of powder bed during scanning of laser nature with lesser efforts than FEM, in particular for its ability to
predicted by a DEM simulation [31]. correctly predict the effects of the thermal history experienced by the
particles, useful in the determination of the residual stresses (Fig. 25).
particle to particle heat conduction component only since, with the Authors have extensively used DEM for powder bed preparation as
metal particles they modelled, it prevails on convection and radiation by well, in order to simulate the powder deposition and recoating mecha
more than one order of magnitude. Fig. 23 shows that the average nisms. Parteli and Pöschel [164] used LIGGGHTS for DEM simulation of
temperature of a metal powder bed increases linearly with time as laser powder bed deposition through roller. Haeri et al. [165] explored the
passes over it. A zig-zag pattern of scan was implemented and assumed a effects of particle shapes on deposition using LAMMPS code.
uniform laser intensity.
Some other studies have considered sintering, densification of 4.2. other mesh-free methods
powder packing as a result of viscous flow [158] due to surface tension
[159]. Parhami and McMeeking [160] investigated the problem of free 4.2.1. Lattice Boltzmann method (LBM)
and pressure assisted sintering in three dimensions. Martin et al. [161] This method is a mesoscopic modelling method which models mostly
Fig. 25. Comparison between the temperature at three different points along the depth by using FEM (Left) and by using DEM (right) for SLS of polymer powders.
Z = 0 correspond to the top surface [163].
Fig. 26. 3D Lattice Boltzmann simulation of 316 L Stainless steel powders melted by 200 W laser. The two figures highlight the effects of inclusion of temperature
dependent surface tension inducing Marangoni convection (left) with respect to a model assuming constant temperature (right). Inclusion of temperature dependent
surface tension gives rise to increased size of melt pool. Temperature color scale is in K [170].
14
Fig. 27. Spatial discretization with the OTM method [178].
Fig. 28. Coupled DEM-FDM working flowchart [180].
fluid dynamic problems treated as pseudo-particles represented by ve conditions, surface tension effects, phase transition as well as wetting
locity distribution functions. The fluid particles stay and interact on the and the effect of wetting on balling. Klassen et al. [171,172] imple
nodes of a lattice grid. One of the first applications of LBM to PBF pro mented evaporation model in AM using LBM.
cesses is due to Attar and Koerner [166] and by Koerner et al. [167,168].
More recently other authors have used LBM as well to study the hy 4.2.2. Smoothed particle hydrodynamics (SPH)
drodynamic phenomena and material-electron beam interaction [71, This method is a particle like, Lagrangian and mesh free method
72], and powder bed preparation [163]. Markl et al. [78] proposed widely used to simulate fluid dynamic problems [173]. In SPH, particles
three–dimensional free surface LBM for EBM where powder particles are are augmented with spatial ’smoothing length’ over which their prop
taken into account. The model includes fluid dynamic effects, like the erties are interpolated by a kernel function. Weirather et al. [174]
dynamics of the melt pool, capillarity and wetting, as well as thermal implemented 3D multi-phase, weakly compressible SPH approach for
effects, like heat conduction and transport, energy absorption and phase SLM. Viscous forces, surface tension, thermo-capillary forces, heat
transitions. Ammer et al. [77] implemented 3D LBM for EDM process as conduction and phase changes are considered. The method, however,
well and additionally describes in detail about the modelling of Electron neglects the effects of internal reflections of energy source and plasma
beam and energy absorption. Lavery et al. [170] proposed 3D LBM code formation, which are important to be characterized. More recently,
in Fortran code for 316 L Stainless steel powder with a particle size of Fürstenau et al. [175] proposed 3D SPH model based on Graphic Pro
50 µm and a layer thickness of 100 µm. It was found that by imple cessing Unit (GPU) implementation for SLM. They have neglected the
menting temperature dependent surface tension (inducing Marangoni) formation of residual stresses and the mechanical equilibrium of solid
the resulting melt pool size was larger than that calculated using con phase is not calculated. This approach resulted in significantly reduced
stant surface tension (Fig. 26). Körner et al. [167] developed a 2D nu computational times as compared to current CFD method. However, in
merical method based on LBM as well, including free surface boundary visco-elastic free flow, SPH methods suffer from tensile instabilities
15
Fig. 29. Temperature evolution of 316 L Stainless steel powder bed. Top row: Top view of DEM model of powder bed with laser heating, second row: cross sectional
view including substrate. [180].
which is due to formation of group of particles when the visco-elastic volumetric heat source model are considered. A multi-scale approach
fluid is in the state of stretching which results in fracture behavior of combining molecular dynamics (MD), DEM and FEM was developed to
the fluid [176]. understand sintering, particle flow and residual stresses [182]. The
coupling mechanism is described in Fig. 30.
4.2.3. Optimal transportation meshfree (OTM)
This method is a form of meshfree Galerkin approach and was 5. Comparison between FEM and LBM: a case study
initially developed by Li et al. [177] for simulating fluid flows and
fluid-structure interactions. This method combines optimal trans This section describes the comparison of Powder bed fusion of high
portation theory with material point sampling and mesh free interpo temperature material, Ti6-Al-4 V from two different groups of authors
lation. Primary variables and their time derivatives are nodal both employing two different numerical approach such as FEM (mesh
parameters whereas mass, volume, heat conductivity and stress are based) and LBM (mesh free), as well as two different AM technique such
material points (Fig. 27). While computing, a search algorithm estab as SLM and EBM respectively. Both groups have implemented phase
lishes connectivity between material points and nodes. Meshfree changes, mass and energy losses due to evaporation as it significantly
methods make use of explicit time integration along with lumping influences the temperatures achieved during the process. Both the
techniques as it also aids parallelization concepts. Wessels et al. [178] groups have validated their models with experimental results. The
proposed OTM to better simulate the moving boundaries between the evaporation during high temperatures generates recoil pressure which
melt pool and the solid material. The numerical model includes straining either helps in deeper penetration of melt or creates melt pool
induced by melting and solidification and the related instabilities.
thermo-mechanical model to address the consolidation behavior. Effects Karayagiz et al. [50] developed a three-dimensional FEM model
such as Marangoni convection and recoil pressure have been neglected. implemented in COMSOL to study the Ti6-Al-4 V alloy with constant
Other methods includes combined multi-scale models where authors laser power of 50 W as Gaussian heat input. Conduction, convection and
have combined DEM and FEM for obtaining faster simulation results radiation were included in the study. A model to study the thermal
with good accuracy [70,179]. Haddad et al. [179] proposed a novel history of the surface along with material evaporation based on a new
approach in combining DEM and FEM for heat conduction calculations physics based approach was developed. Generally during evaporation,
although not in the scope of AM. A suitable coupling matrix was used to both heat and mass are lost, but here the mass loss has been neglected
integrate both models. Ganeriwala and Zohdi [180] developed a considering only heat lost through evaporation. The authors have
reduced model that combines DEM and FDM for SLS, in which the base employed a heat sink on the surface of the powder bed which accounts
substrate is modeled using FDM and powder bed heating is simulated for heat loss due to evaporation. The heat sink is described as
using DEM (Fig. 28). By simulating a small domain, using high level
Qsink = − mv X Lv
programming language (Fortran in this case) and using a binning algo
rithm, results can be obtained in few hours (Fig. 29). Although, for larger where mv is the evaporative mass rate and Lv is the latent heat of
layer continuum models, FEM approaches can be effectively used in evaporation. The parameter mv is calculated based upon the fraction of
coupling with DEM as the part level modeling of residual stresses and vapor phase, which is an internal variable in COMSOL. The powder layer
distortions are easily calculated using FEM. was assumed to be a continuum medium with a constant absorptivity.
Multi scale approaches includes combining analyses of atomic, An initial powder bed porosity of 0.35 was assumed with a layer
powder and component level models at the level of melt pool including thickness of 30 µm. The volume shrinkage due to melting of powder and
Marangoni recirculation and free surface to overall part residual stresses fluid dynamic effects such as Marangoni convection, wetting and
calculation and distortions. The Multi-scale FEM (MsFEM) is used in capillary forces and surface tension were neglected. The authors
such studies where the computational efforts were reduced significantly employed a variable mesh with fine mesh size of 25 µm near laser
as compared to conventional DEM [181]. Ammer et al. [77] developed a interaction zone and coarser mesh away from laser. The experiments
three-dimensional LB method for simulating Titanium alloy using EBM, were performed in a custom built LPBF system with a ThermaViz sensor
in which powder bed generation, surface tension and wetting and a
16
Fig. 30. Multi-scale approach (a) modeling framework for multi-physics multi-scale LPBF process (b) coupling between different scales and experimental valida
tion [182].
which is a dual-wavelength pyrometer. In the simulation, the authors

have achieved a peak temperature history of 3315 K which is close to the
boiling point of the Ti6-Al-4 V alloy, as against an unrealistic peak
temperature of 6000 K when evaporation model was not included.
Experiential thermographic data gives a peak of 3100 K as well. The
predicted size of the heat affected zone are 211 µm width and 72 µm
depth. The experimental results for the heat affected zone are
218 ± 21 µm width and 64.4 ± 1.5 µm depth, which is in line with
simulation, with an error percentage of 3.2% for the width and 10.8%
for the depth. The predicted depth of melt pool was found to be 22 µm
for the given power which is just enough to melt the powder. However,
experimental melt pool depth was not provided. The width of melt pool
in the experiments was found to be 170 ± 13 µm and the simulated
result was 144 µm. This 15% error is mainly due to the neglected Mar
angoni convection, which is not easy to implement in FEM.
Fig. 31. Micrograph cross sections of experimental single track: solid line in Klassen et al. [171] developed a two-dimensional free surface LBM
dicates molten area and discontinuous line indicates predicted molten technique to study the same type of Ti6-Al-4 V using electron beam as
depth [171]. heat input. The author here has implemented evaporation based on a
17
Fig. 32. Scan strategies. (a) "S" pattern and (b) Fractal pattern [89].
Fig. 33. (a) Out-in scanning (b) 45◦ line scanning [184].
Fig. 34. (a) Temperature and melt-pool lifetime vs scan speed at fixed power of 200 W (b) Temperature and melt-pool lifetime vs scan speed vs Laser power at fixed
scan speed of 200 mm/s [67].
18
B. Soundararajan et al.
Table 2
Summary of models.
Model & Description Inputs Outputs Limitations references
references
Finite Element 1. Assumes continuum powder bed, solves heat Temperature dependent material Layer Temperature, residual Not an accurate depiction of the PBF process, as [25,48–53,55,56,
Methods conduction equation sometimes considering radiation properties stresses and distortions, Melt FEM fails to replicate the particle nature of the 58–69,71–73,82,95,
and convection Laser power, Beam size, Scan speed, pool and heat affected zone system. Accurate analysis of melt pool formation is 102,103,123–137,
2. Melt pool hydrodynamics and associated effects are Powder bed model, preheat temperature, dimensions not possible and power spreading cannot be 140–146,148,197]
often neglected optical properties of bed simulated
3. Frequently used software: COMSOL, ABAQUS, ANSYS Solidus and liquidus temperature
Computational 1. Mass, Momentum and energy equations are Temperature dependent material Layer Temperature, Melt pool Difficulty in implementation, longer [74–76,88,149–151,
Fluid Dynamics numerically solved using FDM/FVM coupled with properties including latent heat of fusion dimensions, surface roughness, computational times compared to FEM, poor 198,199]
Volume of Fluids methods and vaporization, Laser input, Powder bed keyhole defects, porosity, representation of particle nature of the system,
2. Considers capillary, wetting, Marangoni effects, recoil model, preheat temperature, optical balling and non-uniformity prone to numerical interpolation errors
pressure and vaporization properties of bed
3. Frequently used software: FLUENT, OpenFoam and Solidus and liquidus temperature, Surface
Flow3D tension coefficients and Gravity input
Lattice Boltzmann 1. considers individual particles Solving discretized Heat input, powder bed model, Layer Temperature, Melt pool Choice of ideal lattice mesh is the basis to avoid [4,77,78,167–172,
19
Method Boltzmann equation on a lattice, Temperature dependent material dimensions, surface roughness, numerical instabilities and in case of free-surface 200]
2. Considers capillary, wetting, Marangoni effects, recoil properties, viscosity keyhole defects, porosity, flows upgradation from 2D to 3D is not easy to
pressure and vaporization balling and non-uniformity implement
Smooth Particle 1. Lagrangian mesh-free model in which governing Heat input, powder bed model, Layer Temperature, Melt pool Prone to tensile instabilities, computationally [174,175]
Hydrodynamics equations are discretized by a moving set of particles Temperature dependent material dimensions, surface tension, expensive than FVM
2. Highly adaptive in nature and suited in treating large properties, viscosity, surface tension co- thermo-capillary forces and
deformation, moving interfaces and fluid-structure efficient phase changes
interactions and can provide the time dependent of the
field variables.
Discrete Element 1. Powder deposition and heat transfer studies possible System of particles with geometry, heat powder bed preparation, Layer Computationally expensive, Lacks physical [31,34,70,80,81,115,
modeling 2. Well suited for realistic sintering/melting simulation input, temperature dependent material temperature, phase change/ representation of Melt flow and convection is often 155–157,159,
of random packing of particles based on Newtonian properties solidification ignored 161–165]
interactions of a system of particles which includes
contact models, friction and heat transfer, ability to
establish/break inter-particle forces and influence of
external forces
3. Some of the software used are LIGGGHTS, LAMMPS,
Additive Manufacturing 47 (2021) 102336

EDEM
Table 3
Mesh based models.
Author Year Process Material Analysis type, Heat transfer Validation Major contribution
software and mesh modes
strategy used considered
Denlinger et al. 2017 LPBF Inconel 718 FEA, Dynamic non Conduction, FE results were compared with A novel FE model to predict the in-
[128] uniform meshing convection and experimental data from Dunbar situ response of LPBF workpieces to
using Patran & radiation et al. [201]. get insights into residual stress, effect
solved using Pan of layer interaction and model for
solver multi-layer building
Chen et al. [95] 2017 SLM Ceramics Level set 3D FEA, Conduction, Nil An FE model for SLM to study
adaptive mesh convection and thermo-mechanical behavior at the
radiation scale of bead. Heat transfer and fluid
flow study to predict shape of the
bead
Montgomery 2020 LPBF Inconel 625 Novel Powder FEM Conduction Results were compared with Power absorption and melt pool
et al. [71] experiments performed on geometry study
EOSINT 270 LBPF machine
Foteinopoulos 2018 LPBF Aluminium Finite Difference Conduction, Comparison of results between A 2D adaptive mesh strategy to
et al. [202] alloy with adaptive & non convection and adaptive and non-adaptive mesh decrease computational time and
adaptive mesh radiation memory.
Williams et al. 2018 LPBF Inconel 718 FEA using ABAQUS, Conduction FE results were compared with A pragmatic approach to model PBF
[73] and 316 L Element birth & experimental data from Dunbar to predict residual stresses across a
Stainless steel death meshing et al. [201] realistic workpiece
Han et al. [203] 2016 SLM Al-Al2O3 FEA using ANSYS, Conduction, Results compared with Using different combinations of
non-uniform static convection experiments conducted on process parameters to predict
meshing Renishaw AM250 thermal behavior and exploring
volumetric energy density for
building quality parts
Cheng et al. 2020 SLM Inconel 718 FEA using ABAQUS, Conduction, Results were compared with EBM A raster scan strategy was developed
[52] adaptive mesh convection and machine data from Cheng et al. to study melt pool in a single scan
radiation [82] track
Mukherjee et al. 2017 LPBF Inconel 718 & FEA using ABAQUS Conduction, Results were compared with Direct Energy Deposition based
[197] Ti-6Al-4 V convection and individual experimental data numerical model with heat transfer
radiation and fluid flow to predict temperature
fields and residual stresses
Masoomi et al. 2018 LPBF Ti-6Al- 4 V CFD using COMSOL Conduction, Results were compared with data Thermal response of single track
[204] convection and from Criales et al. [205] LPBF considering constant and
radiation locally variable heat transfer
coefficient
Dai et al. [206] 2015 SLM TiC & FVM using Fluent Conduction, Results were compared with YLR- Effect of Laser energy per unit length
AlSi10Mg software convection and 200-SM ytterbium fiber laser, on melt pool and resultant top surface
radiation PMG3 optical microscope and S- finish were predicted
4800 field emission Scanning
Electron Microscope
Mokrane et al. 2018 SLS Polymers FVM, uniform mesh Conduction, Results were compared with Global view of simulating SLS for
[74] convection and Riedlbauer et al. [207] polymers
radiation
Knudsen layer that forms at the direct interface between the condensed strategy, Particle Size Distribution (PSD) and mechanical characteristics
matter and vapor phase. A simplified evaporation model based on this such as final component residual stresses, dimensional accuracy, surface
concept, considering local loss of mass and heat is represented. The roughness and strength are very important in obtaining an optimized
authors have included hydrodynamic properties such as wetting and product. The material properties are a functions of process parameters.
capillary action, convective mass and heat transfer along with phase Design of Experiments (DoE) is a most common approach where a
changes. However, Marangoni convection was not considered. A spatial certain specimen is built and the best parameter set-up is interpolated.
resolution of 1 µm is considered in the simulation. Experiments were The aforementioned simulation techniques are applied to this optimi
conducted at different input power and scan speed and the melt pool zation procedure in order to replace the majority of experiments by
depth is close to simulated results as shown in Fig. 31. At 120 W power modeling. The model is validated and or calibrated based on experi
and 0.5 m/s scan speed, both the experimental and numerical depth of mental results. Some of the major process parameters are discussed
penetration is 60 µm. Mass losses were also calculated in experiments below.
and simulation and they are in good agreement with each other. The Scanning Strategy is associated to the residual stresses and distor
mass loss was found to be a function of beam width whereas the melt tions. Lot of authors have investigated their relationship and Ma and Bin
pool geometry is weakly dependent on the beam width. The peak tem [89] used two strategies such as "S" pattern and fractal pattern in SLS as
perature reached in simulation for a given power of 300 W is 3250 K. shown in Fig. 32. The Fractal pattern offers smaller distortions due to
Both the above two process shows good agreement with experiments symmetric temperature field than "S" pattern.
but LBM proves to be more accurate in determining the melt pool More recently, Ali et al. [183] investigated the effect of scan strategy
characteristics and agrees better with experiments in spite of the fact on SLM parts by varying the scan vector length, direction, order of
that Marangoni convection is not accounted for in both models. scanning and rotation of each layers. Six scan strategies were studied
and the lowest residual stress build-up were witnessed in 90◦ , alter
6. Process parameters optimization nating scanning strategy for SLM Ti6Al4V parts. Cheng et al. [184]
studied eight different scan strategies on In718 and concluded that
In PBF, process parameters such as Laser speed, Laser power, Scan out-in scanning pattern gave rise to maximum stresses along the X and Y
20
Table 4
Mesh free models and coupled types.
Author Year Process Material Analysis type, Heat transfer Validation Major contribution
software and modes considered
mesh strategy
used
Ganeriwala 2014 SLS Stainless DEM Conduction Nil A 3D DEM is proposed to prepare powder bed and
et al. [81] steel laser sintering of powders. Change in material
properties due to phase changes are considered
Gobal et al. 2017 SLS Stainless DEM Conduction, Results compared with An adaptive DEM for thermo-mechanical behavior
[208] steel convection and various authors of SLS. This method increases the size of the
radiation elements in DEM based on particle temperature of
relative position to laser. A faster computational
time is achieved.
Gobal et al. 2017 SLS FEM/DEM Conduction, Nil Combined FEM-DEM models to account for dynamic
[70] convection and changes in particle contact areas in the heat affected
radiation zone. Faster implementation than DEM
Moser et al. 2016 SLS NA DEM/Monte Carlo Conduction, Results compared with Calculation of thermal conductivity in finite powder
[115] using MFiX radiation various authors bed depth and in case of multi-layer SLS
softwae
Moser et al. 2015 SLS NA DEM/Monte Carlo Conduction, Results compared with Optical properties of various types of powder beds
[162] using MFiX radiation [209] were calculated
softwae
Dayal et al. 2017 SLM Stainless Discrete particle Conduction, Results are validated by Laser energy absorption in metal powders
[31] steel method/ Monte radiation solving test cases and
Carlo comparison with known
solutions
Ganeriwala 2016 SLS Coupled DEM/ Conduction, Results were compared A DEM model for powder spreading and sintering
et al. [180] Finite difference convection and with [198] and FD model to predict the underlying substrate.
approach radiation Material properties changes due to phase change is
included. Faster computational times achieved
Xin et al. 2017 SLS polymer DEM/Monte Carlo Conduction, Results compared with A radiative sub model to predict Scattering,
[163] using EDEM convection and [99] reflection
software radiation
Xin et al. 2017 SLS polymer DEM/Monte Carlo Conduction, Results compared with Coupled DEM/Monte Carlo model to implement
[210] radiation [99] laser scattering
directions while the residual stresses are reduced while scanning was
Table 5
done at 45◦ amongst all tested cases (Fig. 33).
Summary of influence of process parameters on the PBF processes.
The Laser power is another important process parameter which in
Parameters Influence on PBF process References fluences the maximum temperature of the layer and has opposite effects
Laser Input power and scan Distinguishes energy [61] to the increase of the scan speed. Fig. 34 by Li and Gu [67] shows the
speed requirement for melting and relation between temperature, scan speed, Laser power and the time at
sintering, life time of melt pool which the material remains liquid. As we can see, the molten pool life
and subsequent formation of
pores
time increased thee times when the laser power was doubled from 150 to
Scan strategy Propagation of temperature [83,171] 300 watts. Similarly, when the scan speed is decreased from 400 to
during building and resulting 100 mm/s, there was a significant increase of 6 times of the cooling
effect on residual stress on final time. A combination of increased scan speed and low Laser power yields
specimen
a low temperature and very short melt-pool lifetime resulting in poor
Particle size distribution and Meltability/sinterability, packing [172,173]
shape efficiency, pores, effective wettability and increased number of pores in SLM parts.
thermal conductivity and surface The energy density (ED) which is defined as the ratio of Laser power
roughness. Wider range of PSD to the product of scan speed, layer thickness and beam diameter. This
generates higher powder bed influences microstructure of the specimen. A lower ED decreases the
density. This directly affects the
thermo-mechanical behavior of
melt pool size and temperature gradient, thereby increasing the cooling
the specimen rate resulting in fine-grained microstructure. This restricts dislocation
Layer thickness Improper depth of penetration of [195] movement thereby increasing microhardness [185]. Bang et al. [186]
input power results in lack of described about the effect of ED on cellular structures formation in
fusion and poor mechanical
which the cell size increases with ED.
properties
Powder Flowability Uniform spreading of the powder [173] Particle Size Distribution (PSD) is a vital factor in determining the
and powder with narrow range of quality of the PBF process. The packing of powder on the PBF system is
PSD provides better flowability an important factor in determining the properties such as porosity in the
and generates parts with higher bed, effective thermal conductivity and thermal expansion. Brika et al.
tensile strength and increased
hardness
[187] investigated three Ti-6Al-4 V powder lots and their effects on
Thickness of base plate Better build support and uniform [138] LPBF process in terms of powder bed surface roughness, packing density,
distribution of residual stresses in ability to print and final thermo-mechanical properties of printed parts.
the specimen It was shown that a powder mead of more spherical particles demon
Material properties (thermal Energy absorption into powder [96]
strates a better rheological behavior and efficient packing thereby
conductivity, specific heat bed, Heat transfer in powder bed
capacity, surface roughness and environment providing a dense powder bed. This in turn yields final parts with
and emissivity) excellent thermo-mechanical properties, desired surface finish and
dimensional accuracy. Also, the presence of a large quantities of fine
21
Fig. 35. Existing PBF modeling approaches and their suitability on applications.
particles adversely affects the properties of final parts due to higher discussed and using mesh free model and coupled types is reported in
inter-particle forces. Liu et al. [188] concluded that powders with Table 4. Finally a summary of the paper dealing on the effect of the main
diverse range of particle sizes can provide higher packaging density, and parameters is reported in Table 5. The following points are understood
yield higher density parts with smooth surface finish even at low input from the review:
energy intensities. Powders with narrower range of particle size distri
bution has better flowability, yields parts with higher Ultimate Tensile • FEM based approaches are widely used methodologies for PBF pro
Strength and increased hardness. cesses mostly adopting SLM and SLS, whereas EBM is given least
attention.
7. Experimental validation • Steel and Titanium alloys are widely used materials less negatively
affected by thermal effects. Instead, larger thermal effect are found
There has been a lot of in-situ experimental set-ups used to capture for polymer based and ceramics, which still require a better
the PBF process in real time. The numerical models are usually validated understanding.
and calibrated with the help of experimental results. The temperature • Most of the thermo-mechanical models are based on FE models.
history of layer, the melt pool characteristics and the laser track ge Outputs such as temperature field, residual stresses and distortion
ometry and temperatures are usually measured for comparison. A lot of are widely discussed using FEM. On the other hand, equally impor
authors have used Infra-red (IR) thermography for this purpose tant fluid dynamic problems such as melt pool characteristics are
[1189–191]. Two types of IR thermography is generally used such as more complex to be described by FEM and a less treated with this
co-axial and off-axial systems [192]. Co-axial systems make use of the approach.
optical path of the power source (Laser) while off-axial system is placing • Approaches able of resolution at the particle scale, such as FV, LBM,
the IR sensor at an angle overlooking the powder bed. In the PBF pro SPH and OTM methods, can describe melt pool features and the
cesses, the heat exchange, dynamic characteristics of the powder heat melting and solidification process quite accurately. For an accurate
ing, melting and solidification process is called the Process signatures prediction of the melt pool temperature, it is necessary to consider
[193]. The observable process signatures are (1) melt pool (2) scan path radiation and evaporation contributions to the heat balance. For the
(3) the slice or layer temperature and (4) powder bed homogeneity. The higher input energies the formation of a “keyhole welding”, requires
thermal images are in turn used to compare and validate thermal the inclusion of recoil pressure terms related to the high evaporation
models. Peyre et al. [99] used a laser power-meter placed below the rates. The interconnection between different phenomena accounted
powder bed. This latter is spread on a NaCl window, 95% transparent to by these methods results in the need of high computational power,
the laser beam, to measure the power of the transmitted laser beam as a significantly larger than FEM.
function of the powder thickness. This information, in turn, is used to • DEM proves to be more accurate in simulating the powder bed
measure the fusion depth. Residual stresses can be measured in the post preparation, heat absorption and conduction in powder bed. How
processing, that is once the sintered/melted specimen is ready, using ever also in this case the computational efforts needed are much
Non Destructive Testing (NDT) methods such as Neutron diffraction higher than in FEM. Combined models also prove to be quicker and
[194,195]. Scanning Electron Microscopy (SEM) is used to visualize the gives better results.
melt pool, pores, surface roughness and microstructure of the specimen
[196]. Fig. 35 showcases the most suitable modeling tools according to the
modeling objective. Several issues still need to be addressed especially in
the optimization of the tradeoff between accuracy and computational
8. Conclusion
efforts. Parallel computing, cloud-based computing and reduced order
models are also being developed to reduce the computational efforts.
A comprehensive review on modeling of PBF processes including
Secondly, the uncertainties involved while implementing boundary
conventional mesh based and mesh free approaches has been carried
conditions and material properties in solving a complex phenomenon
out. The method considered are centered on heat transfer but always
such as PBF, the results cannot be fully validated by experiments. The
include other related physical aspects. A summary of the models dis
material properties for example, mostly includes assumptions and non-
cussed is provided in Table 2. A summary of the papers discussed and
linear approaches due to simplicity and similarly, inaccurate
using mesh based model is reported in Table 3. A summary of the papers
22
implementation of boundary conditions also leads to inaccurate results. [16] A.A. Mousa, Experimental investigations of curling phenomenon in selective laser
sintering process, Rapid Prototyp. J. (2016), https://doi.org/10.1108/RPJ-12-
This is the reason of the existence of errors in model results while
2013-0132.
matching the experimental results. Data driven modeling approaches [17] J.C. Fox, S.P. Moylan, B.M. Lane, Effect of process parameters on the surface
are being developed to deal with this issue. Also, most of the meshes roughness of overhanging structures in laser powder bed fusion additive
used in case of Mesh based approaches are quadrilateral (2D) or hex manufacturing, Procedia CIRP 45 (2016) 131–134, https://doi.org/10.1016/j.
procir.2016.02.347.
ahedral (3D) or even cubical which is far from the spherical or elongated [18] W.J. Sames, F.A. List, S. Pannala, R.R. Dehoff, S.S. Babu, The metallurgy and
nature of the particles. Therefore, a hybrid/combined modeling seems processing science of metal additive manufacturing, Int. Mater. Rev. (2016),
interesting in such scenarios. A closed loop control system on the whole https://doi.org/10.1080/09506608.2015.1116649.
[19] A. Townsend, N. Senin, L. Blunt, R.K. Leach, J.S. Taylor, Surface texture
SLS process is also being researched to develop optimized and repro metrology for metal additive manufacturing: a review, Precis. Eng. (2016),
ducible parts. A new method of Particle Finite Element Method (PFEM) https://doi.org/10.1016/j.precisioneng.2016.06.001.
able to use updated Lagrangian description to model the motion of nodes [20] V.V. Popov, A. Katz-Demyanetz, A. Garkun, M. Bamberger, The effect of powder
recycling on the mechanical properties and microstructure of electron beam
(particles) in fluid-solid interaction areas is a potential tool which could melted Ti-6Al-4 V specimens, Addit. Manuf. (2018), https://doi.org/10.1016/j.
accurately simulate PBF as well. addma.2018.06.003.
[21] L. Cordova, T. Bor, M. de Smit, S. Carmignato, M. Campos, T. Tinga, Effects of
powder reuse on the microstructure and mechanical behaviour of Al–Mg–Sc–Zr
Funding alloy processed by laser powder bed fusion (LPBF), Addit. Manuf. (2020), https://
doi.org/10.1016/j.addma.2020.101625.
This work has received funding from the European Union’s Horizon [22] L. Denti, A. Sola, S. Defanti, C. Sciancalepore, F. Bondioli, Effect of powder
recycling in laser-based powder bed fusion of Ti-6Al-4V, Manuf. Technol. (2019),
2020 research and innovation programme under the Marie Skłodowska- https://doi.org/10.21062/ujep/268.2019/a/1213-2489/mt/19/2/190.
Curie grant agreement MATHEGRAM No 813202. [23] S. Chandrasekar, et al., Investigating the effect of metal powder recycling in
electron beam powder bed fusion using process log data, Addit. Manuf. (2020),
https://doi.org/10.1016/j.addma.2019.100994.
Declaration of Competing Interest [24] H.P. Tang, M. Qian, N. Liu, X.Z. Zhang, G.Y. Yang, J. Wang, Effect of powder
reuse times on additive manufacturing of Ti-6Al-4V by selective electron beam
melting, JOM (2015), https://doi.org/10.1007/s11837-015-1300-4.
The authors declare the following financial interests/personal re [25] E. Boillat, S. Kolossov, R. Glardon, M. Loher, D. Saladin, G. Levy, Finite element
lationships which may be considered as potential competing interests: and neural network models for process optimization in selective laser sintering,
Proc. Inst. Mech. Eng., Part B: J. Eng. Manuf. 218 (2004) 607–614, https://doi.
Balaji Soundararajan reports financial support was provided by Euro org/10.1243/0954405041167121.
pean Union’s Horizon 2020 research and innovation programme under [26] C. Meier, R.W. Penny, Y. Zou, J.S. Gibbs, A.J. Hart, Thermophysical phenomena
the Marie Sklodowska-Curie grant agreement MATHEGRAM No in metal additive manufacturing by selective laser melting: fundamentals,
modeling, simulation, and experimentation, Annu. Rev. Heat. Transf. (2018),
813202.
https://doi.org/10.1615/annualrevheattransfer.2018019042.
[27] M.J. TURNER, R.W. CLOUGH, H.C. MARTIN, L.J. TOPP, Stiffness and deflection
References analysis of complex structures, J. Aeronaut. Sci. (1956), https://doi.org/10.2514/
8.3664.
[28] T.A. Krol, C. Seidel, J. Schilp, M. Hofmann, W. Gan, M.F. Zaeh, Verification of
[1] R.J. Williams, et al., In situ thermography for laser powder bed fusion: effects of
structural simulation results of metal-based additive manufacturing by means of
layer temperature on porosity, microstructure and mechanical properties, Addit.
neutron diffraction, Phys. Procedia 41 (2013) 849–857, https://doi.org/10.1016/
Manuf. (2019), https://doi.org/10.1016/j.addma.2019.100880.
j.phpro.2013.03.158.
[2] T. Pereira, J.V. Kennedy, J. Potgieter, A comparison of traditional manufacturing
[29] T. Moges, G. Ameta, P. Witherell, A review of model inaccuracy and parameter
vs additive manufacturing, the best method for the job, Procedia Manuf. 30
uncertainty in laser powder bed fusion models and simulations, J. Manuf. Sci.
(2019) 11–18, https://doi.org/10.1016/j.promfg.2019.02.003.
Eng., Trans. ASME (2019), https://doi.org/10.1115/1.4042789.
[3] M.N. Esfahani L. Bian W. Tian “-SITU LAYER-WISE Qual. Monit. LASER-BASED
[30] P.A. Cundall, O.D.L. Strack, A discrete numerical model for granular assemblies,
Addit. Manuf. USING IMAGE Ser. Anal.,” 2019 202 212.
Géotechnique 29 (1) (1979) 47–65, https://doi.org/10.1680/geot.1979.29.1.47.
[4] A. Bauereiß, T. Scharowsky, C. Körner, Defect generation and propagation
[31] R. Dayal, T. Gambaryan-Roisman, Heat transfer in granular medium for
mechanism during additive manufacturing by selective beam melting, J. Mater.
application to selective laser melting: a numerical study, Int. J. Therm. Sci.
Process. Technol. (2014), https://doi.org/10.1016/j.jmatprotec.2014.05.002.
(2017), https://doi.org/10.1016/j.ijthermalsci.2016.11.014.
[5] D. Chantzis, S. Van-Der-Veen, J. Zettler, W.M. Sim, “An industrial workflow to
[32] E. Madenci I. Guven Finite Elem. Method Appl. Eng. Using Ansys® 2006.
minimise part distortion for machining of large monolithic components in
[33] B. Schoinochoritis, D. Chantzis, K. Salonitis, Simulation of metallic powder bed
aerospace industry”, J. Chem. Theory Comput. 9 (2013) 4517–4525, https://doi.
additive manufacturing processes with the finite element method: a critical
org/10.1016/j.procir.2013.06.103.
review, Proc. Inst. Mech. Eng., Part B: J. Eng. Manuf. (2017), https://doi.org/
[6] J.P. Kruth, L. Froyen, J. Van Vaerenbergh, P. Mercelis, M. Rombouts, B. Lauwers,
10.1177/0954405414567522.
Selective laser melting of iron-based powder, J. Mater. Process. Technol. 149
[34] J.C. Steuben, A.P. Iliopoulos, J.G. Michopoulos, Discrete element modeling of
(2004) 616–622, https://doi.org/10.1016/j.jmatprotec.2003.11.051.
particle-based additive manufacturing processes, Comput. Methods Appl. Mech.
[7] R. Li, J. Liu, Y. Shi, L. Wang, W. Jiang, Balling behavior of stainless steel and
Eng. (2016), https://doi.org/10.1016/j.cma.2016.02.023.
nickel powder during selective laser melting process, Int. J. Adv. Manuf. Technol.
[35] B.K. Mishra, R.K. Rajamani, The discrete element method for the simulation of
(2012), https://doi.org/10.1007/s00170-011-3566-1.
ball mills, Appl. Math. Model. (1992), https://doi.org/10.1016/0307-904X(92)
[8] P. Edwards, A. O’Conner, M. Ramulu, Electron beam additive manufacturing of
90035-2.
titanium components: properties and performance, J. Manuf. Sci. Eng. Trans.
[36] S. Luding, Cohesive, frictional powders: contact models for tension, Granul.
ASME (2013), https://doi.org/10.1115/1.4025773.
Matter (2008), https://doi.org/10.1007/s10035-008-0099-x.
[9] H. Gu, H. Gong, D. Pal, K. Rafi, T. Starr, B. Stucker, “Influences energy density
[37] N.K. Tolochko, T. Laoui, Y.V. Khlopkov, S.E. Mozzharov, V.I. Titov, M.B. Ignatiev,
porosity microstruct. sel. laser melted 17-4PH Stainl. Steel,” (2013).
Absorptance of powder materials suitable for laser sintering, Rapid Prototyp. J.
[10] J.A. Slotwinski, E.J. Garboczi, K.M. Hebenstreit, Porosity measurements and
(2000), https://doi.org/10.1108/13552540010337029.
analysis for metal additive manufacturing process control, J. Res. Natl. Inst.
[38] A.V. Gusarov, J.P. Kruth, Modelling of radiation transfer in metallic powders at
Stand. Technol. (2014), https://doi.org/10.6028/jres.119.019.
laser treatment, Int. J. Heat. Mass Transf. (2005), https://doi.org/10.1016/j.
[11] C. Weingarten, D. Buchbinder, N. Pirch, W. Meiners, K. Wissenbach, R. Poprawe,
ijheatmasstransfer.2005.01.044.
Formation and reduction of hydrogen porosity during selective laser melting of
[39] M. Markl, C. Körner, Multiscale modeling of powder bed–based additive
AlSi10Mg, J. Mater. Process. Technol. (2015), https://doi.org/10.1016/j.
manufacturing, Annu. Rev. Mater. Res. (2016), https://doi.org/10.1146/
jmatprotec.2015.02.013.
annurev-matsci-070115-032158.
[12] L.N. Carter, C. Martin, P.J. Withers, M.M. Attallah, The influence of the laser scan
[40] B. Dikshit, G.R. Zende, M.S. Bhatia, B.M. Suri, Convection in molten pool created
strategy on grain structure and cracking behaviour in SLM powder-bed fabricated
by a concentrated energy flux on a solid metal target, Phys. Fluids (2009),
nickel superalloy, J. Alloy. Compd. (2014), https://doi.org/10.1016/j.
https://doi.org/10.1063/1.3210763.
jallcom.2014.06.172.
[41] I. Yadroitsev, A. Gusarov, I. Yadroitsava, I. Smurov, Single track formation in
[13] N.J. Harrison, I. Todd, K. Mumtaz, Reduction of micro-cracking in nickel
selective laser melting of metal powders, J. Mater. Process. Technol. (2010),
superalloys processed by Selective Laser Melting: a fundamental alloy design
https://doi.org/10.1016/j.jmatprotec.2010.05.010.
approach, Acta Mater. (2015), https://doi.org/10.1016/j.actamat.2015.04.035.
[42] E. Louvis, P. Fox, C.J. Sutcliffe, Selective laser melting of aluminium components,
[14] M.F. Zäh, S. Lutzmann, Modelling and simulation of electron beam melting, Prod.
J. Mater. Process. Technol. (2011), https://doi.org/10.1016/j.
Eng. (2010), https://doi.org/10.1007/s11740-009-0197-6.
jmatprotec.2010.09.019.
[15] B.M. Sharratt, Non-destructive techniques and technologies for qualification of
additive manufactured parts and processes: a literature review, Dep. Natl. Def.
Can. (2015).
23
[43] M. Geiger, K.H. Leitz, H. Koch, A. Otto, A 3D transient model of keyhole and melt [72] E.R. Denlinger, Thermomechanical model development and in situ experimental
pool dynamics in laser beam welding applied to the joining of zinc coated sheets, validation of the laser powder-bed fusion process ✶ ✶should note that some
Prod. Eng. (2009), https://doi.org/10.1007/s11740-008-0148-7. content from this chapter was recently published: Denlinger, Erik R., et al.
[44] S. Das, Physical aspects of process control in selective laser sintering of metals, “Thermomechanical model development and in, Thermo-Mech. Model. Addit.
Adv. Eng. Mater. (2003), https://doi.org/10.1002/adem.200310099. Manuf. (2018).
[45] P. Bidare, I. Bitharas, R.M. Ward, M.M. Attallah, A.J. Moore, Laser Powder Bed [73] R.J. Williams, C.M. Davies, P.A. Hooper, A pragmatic part scale model for
Fusion At Sub-atmospheric Pressures, in: Int. J. Mach. Tools Manuf, 130–131, residual stress and distortion prediction in powder bed fusion, Addit. Manuf.
2018, pp. 65–72, https://doi.org/10.1016/j.ijmachtools.2018.03.007. (2018), https://doi.org/10.1016/j.addma.2018.05.038.
[46] P. Bidare, I. Bitharas, R.M. Ward, M.M. Attallah, A.J. Moore, Laser powder bed [74] A. Mokrane, M. Boutaous, S. Xin, Process of selective laser sintering of polymer
fusion in high-pressure atmospheres, Int. J. Adv. Manuf. Technol. (2018), https:// powders: Modeling, simulation, and validation, Comptes Rendus - Mec. (2018),
doi.org/10.1007/s00170-018-2495-7. https://doi.org/10.1016/j.crme.2018.08.002.
[47] D. Sofia M. Granese D. Barletta M. Poletto “Laser Sinter. unimodal Distrib. Glass [75] F.J. Gürtler, M. Karg, K.H. Leitz, M. Schmidt, Simulation of laser beam melting of
Powders Differ. size,” 2015 doi: 10.1016/j.proeng.2015.01.180. steel powders using the three-dimensional volume of fluid method, Phys.
[48] S. Kolossov, E. Boillat, R. Glardon, P. Fischer, M. Locher, 3D FE simulation for Procedia 41 (2013) 881–886, https://doi.org/10.1016/j.phpro.2013.03.162.
temperature evolution in the selective laser sintering process, Int. J. Mach. Tools [76] P. Yuan, D. Gu, Molten pool behaviour and its physical mechanism during
Manuf. (2004), https://doi.org/10.1016/j.ijmachtools.2003.10.019. selective laser melting of TiC/AlSi10Mg nanocomposites: simulation and
[49] N. Contuzzi, S.L. Campanelli, A.D. Ludovico, 3D finite element analysis in the experiments, J. Phys. D. Appl. Phys. (2015), https://doi.org/10.1088/0022-
selective laser melting process, Int. J. Simul. Model. (2011), https://doi.org/ 3727/48/3/035303.
10.2507/IJSIMM10(3)1.169. [77] R. Ammer, M. Markl, U. Ljungblad, C. Körner, U. Rüde, Simulating fast electron
[50] K. Karayagiz, et al., Numerical and experimental analysis of heat distribution in beam melting with a parallel thermal free surface lattice Boltzmann method,
the laser powder bed fusion of Ti-6Al-4V, IISE Trans. (2019), https://doi.org/ Comput. Math. Appl. (2014), https://doi.org/10.1016/j.camwa.2013.10.001.
10.1080/24725854.2018.1461964. [78] M. Markl, R. Ammer, U. Ljungblad, U. Rüde, C. Körner, Electron beam absorption
[51] L.-E. Loh, et al., Numerical investigation and an effective modelling on the algorithms for electron beam melting processes simulated by a three-dimensional
selective laser melting (SLM) process with aluminium alloy 6061, Int. J. Heat. thermal free surface lattice boltzmann method in a distributed and parallel
Mass Transf. vol. 80 (. 2015) 288–300, https://doi.org/10.1016/j. environment, Procedia Comput. Sci. 18 (2013) 2127–2136, https://doi.org/
ijheatmasstransfer.2014.09.014. 10.1016/j.procs.2013.05.383.
[52] B. Cheng K. Chou “Melt Pool. Evol. Study Sel. Laser melting,” 2020. [79] O.B. Kovalev, I.O. Kovaleva, V.V. Belyaev, Ray tracing method for simulation of
[53] Y.M. Arısoy, L.E. Criales, T. Özel, Modeling and simulation of thermal field and laser beam interaction with random packings of powders, AIP Conference
solidification in laser powder bed fusion of nickel alloy IN625, Opt. Laser Proceedings (2018), https://doi.org/10.1063/1.5027340.
Technol. (2019), https://doi.org/10.1016/j.optlastec.2018.08.016. [80] E.J.R. Parteli, DEM simulation of particles of complex shapes using the
[54] M. Matsumoto, M. Shiomi, K. Osakada, F. Abe, Finite element analysis of single multisphere method: Application for additive manufacturing, AIP Conference
layer forming on metallic powder bed in rapid prototyping by selective laser Proceedings 1542 (2013), https://doi.org/10.1063/1.4811898.
processing, Int. J. Mach. Tools Manuf. (2002), https://doi.org/10.1016/S0890- [81] R. Ganeriwala, T.I. Zohdi, Multiphysics modeling and simulation of selective laser
6955(01)00093-1. sintering manufacturing processes, Procedia CIRP 14 (2014) 299–304, https://
[55] L. Van Belle G. Vansteenkiste J.C. Boyer “Comparisons Numer. Model. Sel. Laser doi.org/10.1016/j.procir.2014.03.015.
Melting,” 2012 doi: 10.4028/www.scientific.net/KEM.504-506.1067. [82] B. Cheng, S. Price, J. Lydon, K. Cooper, K. Chou, On process temperature in
[56] P.K. Bai, J. Cheng, B. Liu, W.F. Wang, Numerical simulation of temperature field powder-bed electron beam additive manufacturing: model development and
during selective laser sintering of polymer-coated molybdenum powder, Trans. validation, J. Manuf. Sci. Eng. Trans. ASME (2014), https://doi.org/10.1115/
Nonferrous Met. Soc. China (2006), https://doi.org/10.1016/S1003-6326(06) 1.4028484.
60264-1. [83] S.I. Shahabad, et al., Heat source model calibration for thermal analysis of laser
[57] I.A. Roberts, C.J. Wang, R. Esterlein, M. Stanford, D.J. Mynors, A three- powder-bed fusion, Int. J. Adv. Manuf. Technol. (2020), https://doi.org/
dimensional finite element analysis of the temperature field during laser melting 10.1007/s00170-019-04908-3.
of metal powders in additive layer manufacturing, Int. J. Mach. Tools Manuf. [84] Z. Zhang, et al., 3-Dimensional heat transfer modeling for laser powder-bed fusion
(2009), https://doi.org/10.1016/j.ijmachtools.2009.07.004. additive manufacturing with volumetric heat sources based on varied thermal
[58] A. Foroozmehr, M. Badrossamay, E. Foroozmehr, S. Golabi, Finite element conductivity and absorptivity, Opt. Laser Technol. (2019), https://doi.org/
simulation of selective laser melting process considering optical penetration 10.1016/j.optlastec.2018.08.012.
depth of laser in powder bed, Mater. Des. (2016), https://doi.org/10.1016/j. [85] A.K. Mishra, A. Aggarwal, A. Kumar, N. Sinha, Identification of a suitable
matdes.2015.10.002. volumetric heat source for modelling of selective laser melting of Ti6Al4V powder
[59] J. Bai, R.D. Goodridge, S. Yuan, K. Zhou, C.K. Chua, J. Wei, Thermal influence of using numerical and experimental validation approach, Int. J. Adv. Manuf.
CNT on the polyamide 12 nanocomposite for selective laser sintering, Molecules Technol. (2018), https://doi.org/10.1007/s00170-018-2631-4.
(2015), https://doi.org/10.3390/molecules201019041. [86] C.D. Boley, S.A. Khairallah, A.M. Rubenchik, Calculation of laser absorption by
[60] D.Q. Zhang, Q.Z. Cai, J.H. Liu, L. Zhang, R.D. Li, Select laser melting of W-Ni-Fe metal powders in additive manufacturing, Addit. Manuf. Handb.: Prod. Dev. Def.
powders: simulation and experimental study, Int. J. Adv. Manuf. Technol. (2010), Ind. (2017).
https://doi.org/10.1007/s00170-010-2641-3. [87] Y. Li, D. Gu, Thermal behavior during selective laser melting of commercially
[61] B. Song S. Dong H. Liao C. Coddet “Process Parameter Sel. Sel. Laser melting pure titanium powder: numerical simulation and experimental study, Addit.
Ti6Al4V Based Temp. Distrib. Simul. Exp. Sinter.,” 2012 doi: 10.1007/s00170- Manuf. (2014), https://doi.org/10.1016/j.addma.2014.09.001.
011-3776-6. [88] Y.C. Wu, et al., Numerical modeling of melt-pool behavior in selective laser
[62] J. Yin, H. Zhu, L. Ke, W. Lei, C. Dai, D. Zuo, Simulation of temperature melting with random powder distribution and experimental validation, J. Mater.
distribution in single metallic powder layer for laser micro-sintering, Comput. Process. Technol. (2018), https://doi.org/10.1016/j.jmatprotec.2017.11.032.
Mater. Sci. (2012), https://doi.org/10.1016/j.commatsci.2011.09.012. [89] L. Ma, H. Bin, Temperature and stress analysis and simulation in fractal scanning-
[63] W. Jiang, K.W. Dalgarno, T.H.C. Childs, Finite element analysis of residual based laser sintering, Int. J. Adv. Manuf. Technol. (2007), https://doi.org/
stresses and deformations in direct metal SLS process, SFF Symp. (2002). 10.1007/s00170-006-0665-5.
[64] A. Hussein, L. Hao, C. Yan, R. Everson, Finite element simulation of the [90] P. Fischer, V. Romano, H.P. Weber, N.P. Karapatis, E. Boillat, R. Glardon,
temperature and stress fields in single layers built without-support in selective Sintering of commercially pure titanium powder with a Nd:YAG laser source, Acta
laser melting, Mater. Des. (2013), https://doi.org/10.1016/j. Mater. (2003), https://doi.org/10.1016/S1359-6454(02)00567-0.
matdes.2013.05.070. [91] N. Yadaiah, S. Bag, Development of egg-configuration heat source model in
[65] R.B. Patil, V. Yadava, Finite element analysis of temperature distribution in single numerical simulation of autogenous fusion welding process, Int. J. Therm. Sci.
metallic powder layer during metal laser sintering, Int. J. Mach. Tools Manuf. (2014), https://doi.org/10.1016/j.ijthermalsci.2014.06.032.
(2007), https://doi.org/10.1016/j.ijmachtools.2006.09.025. [92] C. Li, C.H. Fu, Y.B. Guo, F.Z. Fang, A multiscale modeling approach for fast
[66] N. Shen K. Chou “Thermal Model. Electron beam Addit. Manuf. Process - Powder prediction of part distortion in selective laser melting, J. Mater. Process. Technol.
Sinter. Eff.,” 2012 doi: 10.1115/MSEC2012-7253. (2015), https://doi.org/10.1016/j.jmatprotec.2015.10.022.
[67] Y. Li, D. Gu, Parametric analysis of thermal behavior during selective laser [93] T.I. Zohdi, Rapid simulation of laser processing of discrete particulate materials,
melting additive manufacturing of aluminum alloy powder, Mater. Des. (2014), Arch. Comput. Methods Eng. (2013), https://doi.org/10.1007/s11831-013-9092-
https://doi.org/10.1016/j.matdes.2014.07.006. 6.
[68] K. Antony, N. Arivazhagan, K. Senthilkumaran, Numerical and experimental [94] E. Mirkoohi, D.E. Sievers, H. Garmestani, K. Chiang, S.Y. Liang, Three-
investigations on laser melting of stainless steel 316L metal powders, J. Manuf. dimensional semi-elliptical modeling of melt pool geometry considering hatch
Process. (2014), https://doi.org/10.1016/j.jmapro.2014.04.001. spacing and time spacing in metal additive manufacturing, J. Manuf. Process.
[69] F.R. Liu, Q. Zhang, W.P. Zhou, J.J. Zhao, J.M. Chen, Micro scale 3D FEM (2019), https://doi.org/10.1016/j.jmapro.2019.07.028.
simulation on thermal evolution within the porous structure in selective laser [95] Q. Chen, G. Guillemot, C.A. Gandin, M. Bellet, Three-dimensional finite element
sintering, J. Mater. Process. Technol. (2012), https://doi.org/10.1016/j. thermomechanical modeling of additive manufacturing by selective laser melting
jmatprotec.2012.05.010. for ceramic materials, Addit. Manuf. (2017), https://doi.org/10.1016/j.
[70] A. Gobal, B. Ravani, Physical modeling for selective laser sintering process, addma.2017.02.005.
J. Comput. Inf. Sci. Eng. (2017), https://doi.org/10.1115/1.4034473. [96] H. Wessels, T. Bode, C. Weißenfels, P. Wriggers, T.I. Zohdi, Investigation of heat
[71] C. Montgomery, J. Beuth, L. Sheridan, N. Klingbeil, “Process Mapp. Inconel 625 source modeling for selective laser melting, Comput. Mech. (2019), https://doi.
Laser Powder bed Addit. Manuf.,” (2020). org/10.1007/s00466-018-1631-4.
24
[97] L. Dowling, J. Kennedy, S. O’Shaughnessy, D. Trimble, A review of critical ANSYS, Mater. Sci. Technol. (2015), https://doi.org/10.1179/
repeatability and reproducibility issues in powder bed fusion, Mater. Des. vol. 1743284714Y.0000000703.
186 (2020), 108346, https://doi.org/10.1016/j.matdes.2019.108346. [126] A.J. Dunbar, E.R. Denlinger, M.F. Gouge, P. Michaleris, Experimental validation
[98] Y. Yang, D. Gu, D. Dai, C. Ma, Laser energy absorption behavior of powder of finite element modeling for laser powder bed fusion deformation, Addit.
particles using ray tracing method during selective laser melting additive Manuf. (2016), https://doi.org/10.1016/j.addma.2016.08.003.
manufacturing of aluminum alloy, Mater. Des. 143 (2018) 12–19, https://doi. [127] D. Pal et al. “Enhancing Simul. Addit. Manuf. Process. Using spatiotemporal
org/10.1016/j.matdes.2018.01.043. multiscaling,” 2014.
[99] P. Peyre, Y. Rouchausse, D. Defauchy, G. Régnier, Experimental and numerical [128] E.R. Denlinger, M. Gouge, J. Irwin, P. Michaleris, Thermomechanical model
analysis of the selective laser sintering (SLS) of PA12 and PEKK semi-crystalline development and in situ experimental validation of the Laser Powder-Bed Fusion
polymers, J. Mater. Process. Technol. (2015), https://doi.org/10.1016/j. process, Addit. Manuf. (2017), https://doi.org/10.1016/j.addma.2017.05.001.
jmatprotec.2015.04.030. [129] K. Dai, L. Shaw, Distortion minimization of laser-processed components through
[100] K. Masubuchi, J. Agapakis, Analysis and control of residual stresses, distortion control of laser scanning patterns, Rapid Prototyp. J. (2002), https://doi.org/
and their consequences in welded structures, Springer (1982). 10.1108/13552540210451732.
[101] G. Bugeda, M. Cervera, G. Lombera, Numerical prediction of temperature and [130] M. Gouge, E. Denlinger, J. Irwin, C. Li, P. Michaleris, Experimental validation of
density distributions in selective laser sintering processes, Rapid Prototyp. J. thermo-mechanical part-scale modeling for laser powder bed fusion processes,
(1999), https://doi.org/10.1108/13552549910251846. Addit. Manuf. (2019), https://doi.org/10.1016/j.addma.2019.06.022.
[102] M. Shiomi, A. Yoshidome, F. Abe, K. Osakada, Finite element analysis of melting [131] N.E. Hodge, R.M. Ferencz, R.M. Vignes, Experimental comparison of residual
and solidifying processes in laser rapid prototyping of metallic powders, Int. J. stresses for a thermomechanical model for the simulation of selective laser
Mach. Tools Manuf. (1999), https://doi.org/10.1016/S0890-6955(98)00036-4. melting, Addit. Manuf. (2016), https://doi.org/10.1016/j.addma.2016.05.011.
[103] T.H.G. Childs G. Hauser M. Badrossamay “Selective Laser Sinter. (melting) Stainl. [132] M.F. Zaeh, G. Branner, Investigations on residual stresses and deformations in
Tool. Steel Powder.: Exp. Model.,” 2005 doi: 10.1243/095440505X8109. selective laser melting, Prod. Eng. (2010), https://doi.org/10.1007/s11740-009-
[104] M. Rombouts, L. Froyen, A.V. Gusarov, E.H. Bentefour, C. Glorieux, 0192-y.
Photopyroelectric measurement of thermal conductivity of metallic powders, [133] E.R. Denlinger, J.C. Heigel, P. Michaleris, Residual stress and distortion modeling
J. Appl. Phys. (2005), https://doi.org/10.1063/1.1832740. of electron beam direct manufacturing Ti-6Al-4V, Proc. Inst. Mech. Eng. Part B J.
[105] L. Dong, A. Makradi, S. Ahzi, Y. Remond, Three-dimensional transient finite Eng. Manuf. (2015), https://doi.org/10.1177/0954405414539494.
element analysis of the selective laser sintering process, J. Mater. Process. [134] L.E. Lindgren, A. Lundbäck, M. Fisk, R. Pederson, J. Andersson, Simulation of
Technol. (2009), https://doi.org/10.1016/j.jmatprotec.2008.02.040. additive manufacturing using coupled constitutive and microstructure models,
[106] M. Shapiro et al. “Characterization Powder Beds. Therm. Conduct.: Eff. Gas. Press. Addit. Manuf. (2016), https://doi.org/10.1016/j.addma.2016.05.005.
Therm. Resist. Part. Contact Points,” 2004 doi: 10.1002/ppsc.200400943. [135] I.A. Roberts, Investigation of residual stresses in the laser melting of metal
[107] T. Chen, Y. Zhang, Thermal modeling of laser sintering of two-component metal powders in additive layer manufacturing, J. Mater. Process. Technol. (2012),
powder on top of sintered layers via multi-line scanning, Appl. Phys. A Mater. Sci. https://doi.org/10.1007/s00170-007-1181-y.
Process. (2007), https://doi.org/10.1007/s00339-006-3739-1. [136] C. Li J.F. Liu Y.B. Guo Z.Y. Li “A Temp. -Thread multiscale Model. Approach Effic.
[108] N.K. Tolochko, M.K. Arshinov, A.V. Gusarov, V.I. Titov, T. Laoui, L. Froyen, Predict. Part distortion Sel. Laser melting,” 2020.
Mechanisms of selective laser sintering and heat transfer in Ti powder, Rapid [137] R. Li, Y. Shi, J. Liu, H. Yao, W. Zhang, Effects of processing parameters on the
Prototyp. J. (2003), https://doi.org/10.1108/13552540310502211. temperature field of selective laser melting metal powder, Powder Metall. Met.
[109] A.V. Gusarov, T. Laoui, L. Froyen, V.I. Titov, Contact thermal conductivity of a Ceram. (2009), https://doi.org/10.1007/s11106-009-9113-z.
powder bed in selective laser sintering, Int. J. Heat. Mass Transf. (2003), https:// [138] P.J. Withers, Residual stress and its role in failure, Rep. Prog. Phys. (2007),
doi.org/10.1016/S0017-9310(02)00370-8. https://doi.org/10.1088/0034-4885/70/12/R04.
[110] M. Yuan, T.T. Diller, D. Bourell, J. Beaman, Thermal conductivity of polyamide [139] P.J. Withers, H.K.D.H. Bhadeshia, Residual stress part 1 - measurement
12 powder for use in laser sintering, Rapid Prototyp. J. (2013), https://doi.org/ techniques, Mater. Sci. Technol. (2001), https://doi.org/10.1179/
10.1108/RPJ-11-2011-0123. 026708301101509980.
[111] H.W. Zhang, Q. Zhou, H.L. Xing, H. Muhlhaus, A DEM study on the effective [140] J.L. Bartlett, X. Li, An overview of residual stresses in metal powder bed fusion,
thermal conductivity of granular assemblies, Powder Technol. (2011), https:// Addit. Manuf. (2019), https://doi.org/10.1016/j.addma.2019.02.020.
doi.org/10.1016/j.powtec.2010.09.008. [141] G. Vastola, G. Zhang, Q.X. Pei, Y.W. Zhang, Controlling of residual stress in
[112] T. Tsory, N. Ben-Jacob, T. Brosh, A. Levy, Thermal DEM-CFD modeling and additive manufacturing of Ti6Al4V by finite element modeling, Addit. Manuf.
simulation of heat transfer through packed bed, Powder Technol. (2013), https:// (2016), https://doi.org/10.1016/j.addma.2016.05.010.
doi.org/10.1016/j.powtec.2013.04.013. [142] M. Shiomi, K. Osakada, K. Nakamura, T. Yamashita, F. Abe, Residual stress within
[113] Y.T. Feng, K. Han, C.F. Li, D.R.J. Owen, Discrete thermal element modelling of metallic model made by selective laser melting process, CIRP Ann. - Manuf.
heat conduction in particle systems: basic formulations, J. Comput. Phys. (2008), Technol. (2004), https://doi.org/10.1016/S0007-8506(07)60677-5.
https://doi.org/10.1016/j.jcp.2008.01.031. [143] D. Buchbinder, W. Meiners, N. Pirch, K. Wissenbach, J. Schrage, Investigation on
[114] G. Weidenfeld, Y. Weiss, H. Kalman, A theoretical model for effective thermal reducing distortion by preheating during manufacture of aluminum components
conductivity (ETC) of particulate beds under compression, Granul. Matter 6 using selective laser melting, J. Laser Appl. (2014), https://doi.org/10.2351/
(2004) 121–129, https://doi.org/10.1007/s10035-004-0170-1. 1.4828755.
[115] D. Moser, S. Pannala, J. Murthy, Computation of effective thermal conductivity of [144] X. Zhao, A. Iyer, P. Promoppatum, S.C. Yao, Numerical modeling of the thermal
powders for selective laser sintering simulations, J. Heat. Transf. (2016), https:// behavior and residual stress in the direct metal laser sintering process of titanium
doi.org/10.1115/1.4033351. alloy products, Addit. Manuf. (2017), https://doi.org/10.1016/j.
[116] D. Dye, O. Hunziker, S.M. Roberts, R.C. Reed, Modeling of the mechanical effects addma.2016.10.005.
induced by the tungsten inert-gas welding of the IN718 superalloy, Metall. Mater. [145] T. Furumoto T. Ueda M.S. Abdul Aziz A. Hosokawa R. Tanaka “Study on
Trans. A Phys. Metall. Mater. Sci. (2001), https://doi.org/10.1007/s11661-001- reduction of residual stress induced during rapid tooling process: Influence of
0149-z. heating conditions on residual stress,” 2010 doi: 10.4028/www.scientific.
[117] A.V. Gusarov, I. Yadroitsev, P. Bertrand, I. Smurov, Heat transfer modelling and net/KEM.447-448.785.
stability analysis of selective laser melting, Appl. Surf. Sci. (2007), https://doi. [146] K. Dai, L. Shaw, Thermal and stress modeling of multi-material laser processing,
org/10.1016/j.apsusc.2007.08.074. Acta Mater. (2001), https://doi.org/10.1016/S1359-6454(01)00312-3.
[118] M. Boivineau, et al., Thermophysical properties of solid and liquid Ti-6Al-4V [147] A.V. Gusarov, I. Yadroitsev, P. Bertrand, I. Smurov, Model of radiation and heat
(TA6V) alloy, Int. J. Thermophys. (2006), https://doi.org/10.1007/PL00021868. transfer in laser-powder interaction zone at selective laser melting, J. Heat.
[119] J. Li, S. Yuan, J. Zhu, S. Li, W. Zhang, Numerical model and experimental Transf. (2009), https://doi.org/10.1115/1.3109245.
validation for laser sinterable semi-crystalline polymer: Shrinkage and warping, [148] D. Dai, D. Gu, Thermal behavior and densification mechanism during selective
Polymer (2020), https://doi.org/10.3390/POLYM12061373. laser melting of copper matrix composites: simulation and experiments, Mater.
[120] L.Y. Jiang, et al., Thermal expansion behavior of CNT reinforced AlSi10Mg Des. (2014), https://doi.org/10.1016/j.matdes.2013.10.006.
composite fabricated via laser powder bed fusion, Mater. Res. Express (2019), [149] S.A. Khairallah, A.T. Anderson, A. Rubenchik, W.E. King, Laser powder-bed fusion
https://doi.org/10.1088/2053-1591/ab5b5b. additive manufacturing: physics of complex melt flow and formation mechanisms
[121] V.N. Parthasarathy, S. Kodiyalam, A constrained optimization approach to finite of pores, spatter, and denudation zones, Acta Mater. (2016), https://doi.org/
element mesh smoothing, Finite Elem. Anal. Des. (1991), https://doi.org/ 10.1016/j.actamat.2016.02.014.
10.1016/0168-874X(91)90004-I. [150] C. Qiu, C. Panwisawas, M. Ward, H.C. Basoalto, J.W. Brooks, M.M. Attallah, On
[122] D. Vartziotis, J. Wipper, M. Papadrakakis, Improving mesh quality and finite the role of melt flow into the surface structure and porosity development during
element solution accuracy by GETMe smoothing in solving the poisson equation, selective laser melting, Acta Mater. (2015), https://doi.org/10.1016/j.
Finite Elem. Anal. Des. (2013), https://doi.org/10.1016/j.finel.2012.11.004. actamat.2015.06.004.
[123] Z. Luo, Y. Zhao, A survey of finite element analysis of temperature and thermal [151] G. Yu, D. Gu, D. Dai, M. Xia, C. Ma, Q. Shi, On the role of processing parameters in
stress fields in powder bed fusion Additive Manufacturing, Addit. Manuf. (2018), thermal behavior, surface morphology and accuracy during laser 3D printing of
https://doi.org/10.1016/j.addma.2018.03.022. aluminum alloy, J. Phys. D. Appl. Phys. (2016), https://doi.org/10.1088/0022-
[124] N. Patil, et al., A Generalized Feed Forward Dynamic Adaptive Mesh Refinement 3727/49/13/135501.
and Derefinement Finite Element Framework for Metal Laser Sintering - Part I: [152] X. Wang, Y. Liu, L. Li, C.O. Yenusah, Y. Xiao, L. Chen, Multi-scale phase-field
Formulation and Algorithm Development. J. Manuf. Sci. Eng. Trans. ASME, 2015, modeling of layer-by-layer powder compact densification during solid-state direct
https://doi.org/10.1115/1.4030059. metal laser sintering, Mater. Des. (2021), https://doi.org/10.1016/j.
[125] K. Zeng, D. Pal, H.J. Gong, N. Patil, B. Stucker, Comparison of 3DSIM thermal matdes.2021.109615.
modelling of selective laser melting using new dynamic meshing method to
25
[153] Y.U. Wang, Computer modeling and simulation of solid-state sintering: a phase [182] J. Zhang, Y. Zhang, W.H. Lee, L. Wu, H.H. Choi, Y.G. Jung, A multi-scale multi-
field approach, Acta Mater. (2006), https://doi.org/10.1016/j. physics modeling framework of laser powder bed fusion additive manufacturing
actamat.2005.10.032. process, Met. Powder Rep. (2018), https://doi.org/10.1016/j.mprp.2018.01.003.
[154] X. Zhang, Y. Liao, A phase-field model for solid-state selective laser sintering of [183] H. Ali, H. Ghadbeigi, K. Mumtaz, Effect of scanning strategies on residual stress
metallic materials, Powder Technol. (2018), https://doi.org/10.1016/j. and mechanical properties of Selective Laser Melted Ti6Al4V, Mater. Sci. Eng. A
powtec.2018.08.025. (2018), https://doi.org/10.1016/j.msea.2017.11.103.
[155] J.C. Steuben A.P. Iliopoulos J.G. Michopoulos “Recent Dev. multiphysics Discret. [184] B. Cheng, S. Shrestha, K. Chou, Stress and deformation evaluations of scanning
Elem. Method Addit. Manuf. Model. Simul.,” 2017 doi: 10.1115/DETC2017- strategy effect in selective laser melting, Addit. Manuf. (2016), https://doi.org/
67597. 10.1016/j.addma.2016.05.007.
[156] T.I. Zohdi, A direct particle-based computational framework for electrically [185] M.A. Pekok, R. Setchi, M. Ryan, Q. Han, D. Gu, Effect of process parameters on the
enhanced thermo-mechanical sintering of powdered materials, Math. Mech. microstructure and mechanical properties of AA2024 fabricated using selective
Solids (2014), https://doi.org/10.1177/1081286513505472. laser melting, Int. J. Adv. Manuf. Technol. (2021), https://doi.org/10.1007/
[157] T.I. Zohdi, Additive particle deposition and selective laser processing-a s00170-020-06346-y.
computational manufacturing framework, Comput. Mech. (2014), https://doi. [186] G.B. Bang, et al., Effect of process parameters for selective laser melting with
org/10.1007/s00466-014-1012-6. SUS316L on mechanical and microstructural properties with variation in
[158] A. Jagota, P.R. Dawson, Simulation of the viscous sintering of two particles, chemical composition, Mater. Des. (2021), https://doi.org/10.1016/j.
J. Am. Ceram. Soc. (1990), https://doi.org/10.1111/j.1151-2916.1990.tb05117. matdes.2020.109221.
x. [187] S.E. Brika, M. Letenneur, C.A. Dion, V. Brailovski, Influence of particle
[159] K. Mori, M. Ohashi, K. Osakada, Simulation of microscopic shrinkage behaviour morphology and size distribution on the powder flowability and laser powder bed
in sintering of powder compact, Int. J. Mech. Sci. (1998), https://doi.org/ fusion manufacturability of Ti-6Al-4V alloy, Addit. Manuf. (2020), https://doi.
10.1016/S0020-7403(97)00144-6. org/10.1016/j.addma.2019.100929.
[160] F. Parhami, R.M. McMeeking, A network model for initial stage sintering, Mech. [188] B. Liu R. Wildman C. Tuck I. Ashcroft R. Hague “Investigaztion the effect of
Mater. (1998), https://doi.org/10.1016/S0167-6636(97)00034-3. particle size distribution on processing parameters optimisation in selective laser
[161] C.L. Martin, L.C.R. Schneider, L. Olmos, D. Bouvard, Discrete element modeling of melting process.,” 2011.
metallic powder sintering, Scr. Mater. (2006), https://doi.org/10.1016/j. [189] D. Hagen, A. Chen, J.J. Beaman, D. Kovar, Moisture Eff. Sel. Laser Flash Sinter.
scriptamat.2006.05.017. Yttria-Stab. Zirconia (2019) 873–891.
[162] D. Moser, S. Pannala, J. Murthy, Computation of effective radiative properties of [190] R.B. Dinwiddie M.M. Kirka P.D. Lloyd R.R. Dehoff L.E. Lowe G.S. Marlow.
powders for selective laser sintering simulations, JOM (2015), https://doi.org/ “Calibrating IR cameras for in-situ temperature measurement during the electron
10.1007/s11837-015-1386-8. beam melt processing of Inconel 718 and Ti-Al6-V4,” 2016 doi: 10.1117/1
[163] L. Xin, M. Boutaous, S. Xin, D.A. Siginer, Numerical modeling of the heating phase 2.2229070.
of the selective laser sintering process, Int. J. Therm. Sci. (2017), https://doi.org/ [191] J. Alldredge J. Slotwinski S. Storck S. Kim A. Goldberg T. Montalbano “In-Situ
10.1016/j.ijthermalsci.2017.05.017. monitoring and modeling of metal additive manufacturing powder bed fusion,”
[164] E.J.R. Parteli, T. Pöschel, Particle-based simulation of powder application in 2018 doi: 10.1063/1.5031504.
additive manufacturing, Powder Technol. (2016), https://doi.org/10.1016/j. [192] M. Grasso, V. Laguzza, Q. Semeraro, B.M. Colosimo, In-Process Monitoring of
powtec.2015.10.035. Selective Laser Melting: Spatial Detection of Defects Via Image Data Analysis.
[165] S. Haeri, Y. Wang, O. Ghita, J. Sun, Discrete element simulation and experimental J. Manuf. Sci. Eng. Trans. ASME, 2017, https://doi.org/10.1115/1.4034715.
study of powder spreading process in additive manufacturing, Powder Technol. [193] M. Mani, S. Feng, B. Lane, A. Donmez, S. Moylan, R. Fesperman, Measurement
(2017), https://doi.org/10.1016/j.powtec.2016.11.002. science needs for real-time control of additive manufacturing powder-bed fusion
[166] E. Attar, C. Körner, Lattice Boltzmann method for dynamic wetting problems, processes. Additive Manufacturing Handbook: Product Development for the
J. Colloid Interface Sci. (2009), https://doi.org/10.1016/j.jcis.2009.02.055. Defense Industry, 2017.
[167] C. Körner, E. Attar, P. Heinl, Mesoscopic simulation of selective beam melting [194] Z. Wang, E. Denlinger, P. Michaleris, A.D. Stoica, D. Ma, A.M. Beese, Residual
processes, J. Mater. Process. Technol. (2011), https://doi.org/10.1016/j. stress mapping in Inconel 625 fabricated through additive manufacturing:
jmatprotec.2010.12.016. method for neutron diffraction measurements to validate thermomechanical
[168] C. Körner, A. Bauereiß, E. Attar, Fundamental consolidation mechanisms during model predictions, Mater. Des. (2017), https://doi.org/10.1016/j.
selective beam melting of powders, Model. Simul. Mater. Sci. Eng. (2013), matdes.2016.10.003.
https://doi.org/10.1088/0965-0393/21/8/085011. [195] L. Papadakis, G. Branner, A. Schober, K.H. Richter, T. Uihlein, “Numerical
[169] M. Markl, C. Körner, Powder layer deposition algorithm for additive modeling of heat effects during thermal manufacturing of aero engine
manufacturing simulations, Powder Technol. (2018), https://doi.org/10.1016/j. components,”, 2012.
powtec.2018.02.026. [196] S. Tammas-Williams, H. Zhao, F. Léonard, F. Derguti, I. Todd, P.B. Prangnell, XCT
[170] N.P. Lavery S.G.R. Brown J. Sienz J. Cherry “A review of Computational analysis of the influence of melt strategies on defect population in Ti-6Al-4V
Modelling of Additive Layer Manufacturing – multi-scale and multi-physics,” components manufactured by selective electron beam melting, Mater. Charact.
2014 doi: 10.13140/RG.2.1.3103.0884. (2015), https://doi.org/10.1016/j.matchar.2015.02.008.
[171] A. Klassen, T. Scharowsky, C. Körner, Evaporation model for beam based additive [197] T. Mukherjee, W. Zhang, T. DebRoy, An improved prediction of residual stresses
manufacturing using free surface lattice Boltzmann methods, J. Phys. D. Appl. and distortion in additive manufacturing, Comput. Mater. Sci. (2017), https://
Phys. (2014), https://doi.org/10.1088/0022-3727/47/27/275303. doi.org/10.1016/j.commatsci.2016.10.003.
[172] A. Klassen, V.E. Forster, C. Körner, A multi-component evaporation model for [198] S.A. Khairallah, A. Anderson, Mesoscopic simulation model of selective laser
beam melting processes, Model. Simul. Mater. Sci. Eng. (2017), https://doi.org/ melting of stainless steel powder, J. Mater. Process. Technol. (2014), https://doi.
10.1088/1361-651X/aa5289. org/10.1016/j.jmatprotec.2014.06.001.
[173] M.B. Liu, G.R. Liu, S. Li, ?“Smoothed particle hydrodynamics ? a meshfree [199] C. Panwisawas, et al., Mesoscale modelling of selective laser melting: thermal
method?,”, Comput. Mech. (2004) https://doi.org/10.1007/s00466-004-0573-1. fluid dynamics and microstructural evolution, Comput. Mater. Sci. (2017),
[174] J. Weirather, et al., A smoothed particle hydrodynamics model for laser beam https://doi.org/10.1016/j.commatsci.2016.10.011.
melting of Ni-based alloy 718, Comput. Math. Appl. (2019), https://doi.org/ [200] A. Cattenone, S. Morganti, F. Auricchio, Basis of the lattice Boltzmann method for
10.1016/j.camwa.2018.10.020. additive manufacturing, Arch. Comput. Methods Eng. (2020), https://doi.org/
[175] J.P. Fürstenau, H. Wessels, C. Weißenfels, P. Wriggers, Generating virtual process 10.1007/s11831-019-09347-7.
maps of SLM using powder-scale SPH simulations, Comput. Part. Mech. (2020), [201] A.J. Dunbar, et al., Development of experimental method for in situ distortion and
https://doi.org/10.1007/s40571-019-00296-3. temperature measurements during the laser powder bed fusion additive
[176] X. Xu, P. Yu, A technique to remove the tensile instability in weakly compressible manufacturing process, Addit. Manuf. (2016), https://doi.org/10.1016/j.
SPH, Comput. Mech. (2018), https://doi.org/10.1007/s00466-018-1542-4. addma.2016.04.007.
[177] B. Li, F. Habbal, M. Ortiz, Optimal transportation meshfree approximation [202] P. Foteinopoulos, A. Papacharalampopoulos, P. Stavropoulos, On thermal
schemes for fluid and plastic flows, Int. J. Numer. Methods Eng. (2010), https:// modeling of additive manufacturing processes, CIRP J. Manuf. Sci. Technol.
doi.org/10.1002/nme.2869. (2018), https://doi.org/10.1016/j.cirpj.2017.09.007.
[178] H. Wessels, C. Weißenfels, P. Wriggers, Metal particle fusion analysis for additive [203] Q. Han R. Setchi S.L. Evans C. Qiu “Three-Dimens. Finite Elem. Therm. Anal. Sel.
manufacturing using the stabilized optimal transportation meshfree method, Laser melting Al-Al2O3 Powder,” 2016.
Comput. Methods Appl. Mech. Eng. (2018), https://doi.org/10.1016/j. [204] M. Masoomi, J.W. Pegues, S.M. Thompson, N. Shamsaei, A numerical and
cma.2018.04.042. experimental investigation of convective heat transfer during laser-powder bed
[179] H. Haddad, M. Guessasma, J. Fortin, Heat transfer by conduction using DEM-FEM fusion, Addit. Manuf. (2018), https://doi.org/10.1016/j.addma.2018.06.021.
coupling method, Comput. Mater. Sci. (2014), https://doi.org/10.1016/j. [205] L.E. Criales, Y.M. Arısoy, B. Lane, S. Moylan, A. Donmez, T. Özel, Laser powder
commatsci.2013.08.033. bed fusion of nickel alloy 625: experimental investigations of effects of process
[180] R. Ganeriwala, T.I. Zohdi, A coupled discrete element-finite difference model of parameters on melt pool size and shape with spatter analysis, Int. J. Mach. Tools
selective laser sintering, Granul. Matter (2016), https://doi.org/10.1007/s10035- Manuf. (2017), https://doi.org/10.1016/j.ijmachtools.2017.03.004.
016-0626-0. [206] D. Dai, D. Gu, Tailoring surface quality through mass and momentum transfer
[181] H.W. Zhang, Q. Zhou, Y.G. Zheng, A multi-scale method for thermal conduction modeling using a volume of fluid method in selective laser melting of TiC/
simulation in granular materials, Comput. Mater. Sci. (2011), https://doi.org/ AlSi10Mg powder, Int. J. Mach. Tools Manuf. (2015), https://doi.org/10.1016/j.
10.1016/j.commatsci.2011.04.019. ijmachtools.2014.09.010.
[207] D. Riedlbauer, M. Drexler, D. Drummer, P. Steinmann, J. Mergheim, Modelling,
simulation and experimental validation of heat transfer in selective laser melting
26
of the polymeric material PA12, Comput. Mater. Sci. (2014), https://doi.org/ [209] J. Zhou, Y. Zhang, J.K. Chen, Numerical simulation of laser irradiation to a
10.1016/j.commatsci.2014.06.046. randomly packed bimodal powder bed, Int. J. Heat. Mass Transf. (2009), https://
[208] A. Gobal, B. Ravani, An adaptive discrete element method for physical modeling doi.org/10.1016/j.ijheatmasstransfer.2009.01.028.
of the selective laser sintering process, Appl. Mech. Mater. (2017), https://doi. [210] L. Xin, M. Boutaous, S. Xin, D.A. Siginer, Multiphysical modeling of the heating
org/10.4028/www.scientific.net/amm.869.69. phase in the polymer powder bed fusion process, Addit. Manuf. (2017), https://
doi.org/10.1016/j.addma.2017.10.006.
27

Additive Manufacturing: Balaji Soundararajan, Daniele Sofia, Diego Barletta, Massimo Poletto

Uploaded by

Copyright:

Available Formats

Additive Manufacturing: Balaji Soundararajan, Daniele Sofia, Diego Barletta, Massimo Poletto

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Additive Manufacturing: Balaji Soundararajan, Daniele Sofia, Diego Barletta, Massimo Poletto

Uploaded by

Copyright:

Available Formats

Additive Manufacturing 47 (2021) 102336

Contents lists available at ScienceDirect

Review on modeling techniques for powder bed fusion processes based on

of the part. As a result, roughness and poor surface characteristics

During the PBF process, the powder bed is heated by a Laser or an

only (Fig. 5).

where A is the surface absorptivity of the powder bed, P is the power of

bed porosity and surface areas [38].

depends on the level of penetration among the particles. According to

2.4. Material properties

Accurate information on material properties is important in the

material property which indicates the ability of the material to expand

3. Mesh based approaches

The quality of the simulation results in mesh based approaches very

where α is the co-efficient of thermal expansion of material and Tref is the

3.3. FEA: simulation results

3.3.1. Temperature fields, residual stresses and distortions

3.3.2. Melt pool

FVM method [75,76,151]. Also a study considering shrinkage of volume

3.5. Phase field method

4. Mesh free methods

4.1. Discrete element modeling

The DEM is a numerical approximation method capturing the way

Fig. 27. Spatial discretization with the OTM method [178].

Fig. 28. Coupled DEM-FDM working flowchart [180].

which is a dual-wavelength pyrometer. In the simulation, the authors

Additive Manufacturing 47 (2021) 102336

You might also like