Selective laser melting 3D printing of Ni-based superalloy ...iam.nuaa.edu.cn/_upload/article/files/ba/ca/7e46818f4ebf90a64027f0f83526/b6b8709f-47f6...Keywords Selective laser melting

Artic le Materials Science

Selective laser melting 3D printing of Ni-based superalloy:understanding thermodynamic mechanisms

Mujian Xia • Dongdong Gu • Guanqun Yu •

Donghua Dai • Hongyu Chen • Qimin Shi

Received: 6 March 2016 / Revised: 21 April 2016 / Accepted: 27 April 2016 / Published online: 20 May 2016

� Science China Press and Springer-Verlag Berlin Heidelberg 2016

Abstract A mesoscopic model has been established to

investigate the thermodynamic mechanisms and densifica-

tion behavior of nickel-based superalloy during additive

manufacturing/three-dimensional (3D) printing (AM/3DP)

by numerical simulation, using a finite volume method

(FVM). The influence of the applied linear energy density

(LED) on dimensions of the molten pool, thermodynamic

mechanisms within the pool, bubbles migration and

resultant densification behavior of AM/3DP-processed

superalloy has been discussed. It reveals that the center of

the molten pool slightly shifts with a lagging of 4 lmtowards the center of the moving laser beam. The Mar-

angoni convection, which has various flow patterns, plays a

crucial role in intensifying the convective heat and mass

transfer, which is responsible for the bubbles migration and

densification behavior of AM/3DP-processed parts. At an

optimized LED of 221.5 J/m, the outward convection

favors the numerous bubbles to escape from the molten

pool easily and the resultant considerably high relative

density of 98.9 % is achieved. However, as the applied

LED further increases over 249.5 J/m, the convection

pattern is apparently intensified with the formation of

vortexes and the bubbles tend to be entrapped by the

rotating flow within the molten pool, resulting in a large

amount of residual porosity and a sharp reduction in

densification of the superalloy. The change rules of the

relative density and the corresponding distribution of

porosity obtained by experiments are in accordance with

the simulation results.

Keywords Selective laser melting � 3D printing �Mesoscopic simulation � Thermodynamics �Densification � Porosity

1 Introduction

Nickel-based superalloys are known as an attractive can-

didate for many industrial applications, e.g., gas turbine

disks, rocket motors, spacecraft, etc., due to their excellent

hot resistance to oxidation, high strength and wear resis-

tance [1–4]. However, since superalloys have the high

abilities of self-hardening and retaining superior mechani-

cal properties at high temperatures, they are difficult to be

manufactured by conventional processing methods, caused

by severe tool wear damage, low material removal rate,

poor thermal property and surface integrity [5–8].

The newly developed selective laser melting (SLM)

additive manufacturing/3D printing (AM/3DP) technology,

due to the possibility to fabricate the geometrically complex

components by user-defined computer aided design (CAD)

data files without tools or molds, has been proved to be an

effective and economical method for processing Ni-based

superalloys [9–11]. The as-built components having high

dimensional precision, perfect surface quality and out-

standing performance can be achieved precisely byAM/3DP

[12–15]. A number of previous attempts concerning thewear

resistance, high-temperature oxidation and mechanical

properties of AM/3DP Ni-based superalloys have been

investigated systematically [16–18]. Some typical defects

M. Xia � D. Gu (&) � G. Yu � D. Dai � H. Chen � Q. ShiCollege of Materials Science and Technology,

Nanjing University of Aeronautics and Astronautics,

Nanjing 210016, China

e-mail: [email protected]

M. Xia � D. Gu � G. Yu � D. Dai � H. Chen � Q. ShiInstitute of Additive Manufacturing (3D Printing),

Nanjing University of Aeronautics and Astronautics,

Nanjing 210016, China

123

Sci. Bull. (2016) 61(13):1013–1022 www.scibull.com

DOI 10.1007/s11434-016-1098-7 www.springer.com/scp

such as pore [19] and residual stress [20] have been observed

in as-fabricated components, limiting its further applications

in the practical industrial fields. Porosity is regarded as a

commonly existed processing defect that is usually observed

in amajority of themetallic parts processed byAM/3DP [21,

22], which results in the poor densification level and/or other

mechanical properties [19, 23]. Many previous studies have

ascribed the porosity formation to the following phenomena

including powder denudation [24], collapse of key holes

[25], gaseous bubbles entrapment [26], incomplete re-melt-

ing of some local sites [27] and splash [28]. Actually, for the

bubbles migration and porosity formation, it is difficult to be

monitored or characterized within the dynamic molten pool

during AM/3DP. So far, only few previous researches have

been reported on this issue. For instance, Dai and Gu [29]

have established a newmodel to investigate the densification

behavior of Cu-based composites by considering the

migration and escaping of bubbles. Nevertheless, there still

lacks of a relatively clear and comprehensive understanding

of the physical mechanisms for AM/3DP of Ni-based

superalloys. It is therefore necessary to give a thermody-

namic investigation on porosity evolution and densification

behavior of AM/3DP-processed superalloys.

In this paper, the influence of linear energy density

(LED) of laser on the thermodynamic behavior, bubbles

migration behavior and densification mechanism of AM/

3DP processed Ni-based superalloy was numerically

studied by applying commercially computational fluid

dynamics (CFD) software. To further validate the accuracy

of the newly developed mesoscopic model and to acquire

the optimal SLM AM/3DP processing parameters to fab-

ricate components with higher densification, the relative

density of the parts predicted by the numerical simulation

was compared with the experimental results.

2 Modeling approach and experimental procedures

2.1 Physical model

During SLM processing, the interaction between laser

beam and powder in the shield gas ambient is extremely

complex [30, 31]. Regarding the migration of gaseous

bubbles and attendant densification mechanism of powder,

a schematic of SLM physical model is depicted in Fig. 1a.

Recently, the powder-scale model, i.e., the mesoscopic

model, has been established to study the porosity evolution

and surface morphology of titanium alloy during SLM

[23]. In our present study, the powder system, processed in

3D size of (400 9 180 9 50) lm3 (X-, Y-, Z-axis), contains

a mixture of Ni-based superalloy powder particles and

argon gaseous bubbles, as shown in Fig. 1b. A number of

powder particles (average diameter of 30 lm) were closely

packed in powder bed with a relative packed density of

powder up to 47 %, which is close to the theoretical packed

density of 55 % of the spherical powder particles [32].

2.2 Governing equations

Generally, the movement of melt fluid is followed by the

mass, momentum and energy conservation, which can be

written as follows [33]:

oqot

þ oðquÞoX

þ oðqvÞoY

þ oðqwÞoZ

¼ 0; ð1Þ

qoV~

otþ V~ � rV~

!¼ lr2V~�rpV~þMs � V~ þ F, ð2Þ

qoT

otþ V~ � rT

� �¼ r � ðj rTÞ þ SH; ð3Þ

where q, j, l and p represent density, thermal conductivity,

dynamic viscosity and pressure, respectively. rT is the

temperature gradient distributed along the 3D coordinates,V~

is the motion velocity of the melt,Ms is the mass source, and

F~ refers to the body force. SH is the source term of the energy

equation in the X, Y and Z axis and can be defined by

SH ¼ �qo

otDH þr � ðV~DHÞ

� �; ð4Þ

where DH is the latent heat of phase transformation.

Based on the volume of fluid (VOF) model applied in

this simulation, the volume fraction equation for i phase is

[34]

oaiot

þ v~ � rai ¼Saiqi

; ð5Þ

wherePn

i¼1 ai ¼ 1, ai represents the volume fraction of

i phase, n is the total number of phases.

2.3 Boundary conditions

The transient spatial temperature distribution T(x, y, z, t) is

time-dependent. Prior to SLM process, the initial condition

of the uniform temperature distribution throughout the

powder bed at t = 0 can be settled as

Fig. 1 (Color online) Schematic of SLM physical model (a) and the

3D model established for simulating SLM process (b)

1014 Sci. Bull. (2016) 61(13):1013–1022

123

T x; y; z; tð Þ=t¼0 ¼ T0; x; y; zð Þ 2 Sn; ð6Þ

where T0, which represents the ambient temperature, is

293 K.

The natural thermal conductivity of the boundaries sat-

isfies the following equation [35]:

KoT

onþ hcðT � T0Þ þ reðT4 � T4

0 Þ ¼ q, ð7Þ

where n is the normal vector of the surface, hc is the heat

transfer coefficient of natural thermal convection, e is the

emissivity and r is the Stefan–Boltzmann constant.

A laser source with a volumetric Gaussian distribution

can be written as [36]

q ¼ 9PA

R2pH 1� 1e3

� � exp �9ðx2 þ y2ÞR2 log H

z

� �0@

1A; ð8Þ

where P is the laser power, R represents the diameter of

laser beam, H is the penetrated depth of heat source, and

A is the effective laser energy absorption of the material

[37]. The value of A is settled as 0.8 in the earlier study

[38].

The nonlinear behavior of thermal conductivity and

specific heat due to temperature change and phase trans-

formation should be taken into consideration, as shown in

Fig. 2.

The numerical simulation is carried out by the Fluent

commercial finite volume method (FVM) package and the

applied parameters are given in Table 1.

2.4 Experimental procedures

The gas atomized pre-alloy powder of Inconel 718 Ni-

based superalloy, with a purity of 99.7 %, was used as the

raw material. The SLM AM/3DP apparatus consisted of an

YLR-500-SM Ytterbium fiber laser with a power of

*500 W and a spot size of 70 lm (IPG Laser GmbH). The

processing parameters used in the experiments were in

accordance with that applied in the numerical simulation. V

was settled as a constant of 400 mm/s, and P were preset at

77.5, 88.6, 99.8 and 110 W, respectively. Thus, four dif-

ferent LEDs of 193.7, 221.5, 249.5 and 275 J/m were

changed to vary the SLM processing conditions, and were

defined by

LED ¼ P=v. ð9Þ

The SLM-processed samples for metallographic

examinations were cut, ground and polished according to

the standard procedures, and then etched with an acid

solution containing H2O2 (3 mL) and HCl (10 mL) for 10 s.

An Olympus PMG3 optical microscope (OM) was used to

observe the microstructural features of the SLM-processed

specimens. The relative density was defined as the

ratio between the density of SLM-processed Inconel 718

Ni-based alloy and the one of bulk-form Inconel 718 alloy

(8,200 kg/m3). Based on the Archimedes’ principle, the

density of SLM-processed Inconel 718 Ni-based alloy could

be calculated by the ratio between its qualitymeasured by the

electronic balance and volume of displacing water. To

ensure the accuracy of the measurement, each test was

repeated at least three times and the average value was taken

as the final value of the relative density.

3 Results and discussion

3.1 Temperature distribution and molten pool

dimensions

The temperature evolutions along X-axis (Y = 0,

Z = 20 lm) at various LEDs are shown in Fig. 3a. The

profiles of temperature distributions demonstrated that the

temperature generally increased with enhancing LEDs, but

the magnitude of temperature enhancement relied on the

input LEDs. Especially, the temperatures, which were

Fig. 2 (Color online) Thermal conductivity and special heat capacity

of Ni-based superalloy (Inconel 718) at various operating

temperatures

Table 1 The as-used material properties and SLM processing

parameters [38]

Parameters Value

Density, q (kg/m3) 8200

Absorption of superalloy powder, A 0.8

Ambient temperature, T0 (K) 293

Powder layer thickness, d (lm) 50

Radius of laser beam, D (lm) 35

Hatch spacing, s (lm) 50

Laser power, P (W) 77.5, 88.6, 99.8, 110

Scanning speed, v (mm/s) 400

Sci. Bull. (2016) 61(13):1013–1022 1015

123

distributed along the X-axis value ranging from 150 to

290 lm in the vicinity of the laser beam center, elevated

obviously with LEDs. The higher temperature gradients in

the front side of the moving laser beam at a larger LED

could be clearly observed. This was attributed to the fact

that the powder particles absorbed more energy under the

action of a higher LED, leading to more thermal accumu-

lation accordingly. It was noted that the temperature evo-

lutions of all applied LEDs presented a sharp reduction

(dashed line ellipse marked in Fig. 3a) at the location in X-

axis of 262 lm. This position was filled with gas phase

without the sufficient feeding of the molten material

between the adjacent melted powder particles, resulting in

the abrupt temperature difference due to the variation of

their absorption and thermal conductivity.

The enlarged characterization is carried out to reveal the

maximum temperature along X-axis on the specific position

ranging from 180 to 300 lm, as displayed in Fig. 3b. It was

found that the maximum temperature at various LEDs

presented at the location of 276 lm, i.e., the center of the

molten pool. This position was not consistent with the real

center of the laser beam (X = 280 lm), and shifted slightly

towards the negative direction of X-axis with a lagging of

4 lm. The similar phenomenon has also been found in

SLM of Cu-based composites [29], which resulted from the

combinational effect of thermal accumulation and variation

of thermal conductivity caused by the rapid transition from

powder to melt [39]. In this instance, the as-irradiated

zones located in the rear of the moving laser beam were

subjected to heat frequently. The slopes of the temperature

distribution curves (i.e., representing the temperature gra-

dients) in the rear side of the moving laser beam were much

steeper than those in the front side of the laser beam, owing

to the more significant thermal accumulation of the pre-

viously solidified regions of the track compared to that of

the powder particles. Meanwhile, in the front side of the

scanning laser beam, the maximum temperature and tem-

perature gradient were enhanced with increasing the LED,

which eventually increased to the maximum values at an

elevated LED of 275 J/m. The commensurate thermal

stresses and thermal cracks tended to be formed in the as-

built parts due to the release of stresses [13]. Therefore, the

SLM processing with a reasonable LED favors the pres-

ence of the more homogeneous temperature gradients and

the lower residual thermal stress.

The distance above the liquid line along scanning

direction, which was considered as the length of the molten

pool, was enhanced with the increase of LEDs (varying

from 12.2 lm at 193.7 J/m to 94.5 lm at 275 J/m)

(Fig. 3b). This was ascribed to the fact that an increase of

LEDs resulted in a higher temperature during AM/3DP

process and accordingly an enlarged molten pool length,

demonstrating the significance of high laser power for

thermal accumulation effect [40].

With respect to the width of the molten pool, it is

defined by the temperature counters of the pool. The width

of the molten pool is strengthened by increasing the LED,

as depicted in Fig. 4. The minimum and maximum of

width of the molten pool were 31.2 lm (at 193.7 J/m) and

67.9 lm (at 275 J/m), respectively, caused by the thermo-

capillary flow (Marangoni convection) induced by the

temperature gradient [41]. The applied laser source with a

Guassian energy distribution is responsible for the

appearance of the large thermal gradients in the molten

pool. The surface tension gradient and Marangoni con-

vection are accordingly generated, which tend to pull the

molten materials from the center to the periphery of the

pool. At the edge of the pool, the molten materials sink

down under the gravity force and, subsequently, the fluid at

the bottom floats up to the surface again owing to the

buoyancy force caused by density difference [40, 42].

Therefore, the outward Marangoni convection presents in

the molten pool, which favors the transfer of the melt from

Fig. 3 (Color online) a Temperature distributions at various LEDs on

the scanning path along X-axis (Y = 0, Z = 20 lm), t = 700 ls.b The local enlargement of (a) from 180 to 300 lm

1016 Sci. Bull. (2016) 61(13):1013–1022

123

the center to the edge of the pool. The higher input LED is

applied, the more intensive outward Marangoni convection

is obtained. In this situation, the width of the molten pool is

enhanced with the increase of LEDs. Therefore, it is

apparent that the applied LEDs play a crucial role in

determining the width of the molten pool that is elevated

with increasing the LEDs.

As shown in Fig. 4, it is further observed that the

surface morphology of the molten pool varied with the

different LEDs. The temperature gradient exerted from

the surface (with a maximum temperature) to the inner

region (with a minimum temperature) of powder particles

at a relative low LED of 193.7 J/m. Meanwhile, due to the

high viscosity of the molten pool, it was difficult for the

melt to spread outwards fully, thereby producing the

rough surface with a limited integrity [40]. In this case,

although the maximum temperature of the particle surface

reached 1,905 K that was higher than the melting point of

Inconel 718 Ni-based alloy (1,609 K), the core zones of

powder particles might still melt insufficiently, producing

a limited amount of liquid and a discontinuous liquid

track (marked by dashed rectangle). As the LED increased

to 221.5 J/m, the powder particles melted completely and

the resultant lower viscosity was beneficial to the com-

plete spreading of the melt flow. Thus, the surface quality

was significantly enhanced. However, as the LED further

exceeded 249.5 J/m, the melt spread easily owing to the

sharply reduced viscosity. In this situation, the instability

of the molten pool was enhanced and the irregular-shaped

tracks tended to be developed on the top surface, thus

reducing the surface integrity significantly. On the other

hand, when a relatively higher LED was applied, the

material evaporation companying with melt splash would

occur. Under this condition, the distinct recoil pressure

tended to be generated due to material evaporation and

gas expansion, which also resulted in a poor surface

integrity [23].

Concerning the depth of the molten pool, it was ana-

lyzed quantitatively by the temperature counters on the

cross-sections of the molten pools as well as the definition

of molten pool width. From Fig. 5, it was observed that the

depth of the molten pool decreased with increasing the

LED. The outward Marangoni convection, as discussed in

the above section, favored to the transfer of the melt and

heat from the bottom to the top of the molten pool at a

higher LED. As a result, the depth of molten pool was

reduced from 28.3 lm at 193.7 J/m successively to the

minimum value of 23.6 lm at 275 J/m.

3.2 Thermodynamics behavior and bubble migration

The previous publications reported that the bubble

migration was significantly influenced by the melt flow

[29]. In the present study, the thermodynamic behavior

within the molten pool, which was caused by the tem-

perature gradient, is introduced to investigate the bubble

migration and the velocity vector plots of convection

under different LEDs are given in Fig. 6. The intensities of

convection were significantly enhanced on increasing the

LEDs, and the obvious thermo-capillary convection pat-

terns were present at the relatively high LEDs above

193.7 J/m. For a relative low LED of 193.7 J/m, the

temperature gradient was insufficient, accordingly weak-

ening the resultant intensity of convection with a speed of

1.2 m/s (Fig. 6a). In this situation, the bubble migration

was slowed down. A majority of the bubbles were retained

within the molten pool, caused by the lower speed of the

bubbles (Fig. 7a). The bubble migration and melt flow at a

relatively low LED were similar to those during SLM of

Cu-based composites in our previous study [29]. As the

LED increased to 221.5 J/m, the present larger tempera-

ture gradient and surface tension gradient were favorable

to the formation of Marangoni convection with a larger

velocity of 2.6 m/s (Fig. 6b). The outward Marangoni

convection not only accelerated the heat and mass transfer

within the molten pool [43], but also provided the possi-

bility for bubbles to arise from the bottom to the top

surface of the molten pool. In this case, most of the bub-

bles moved and escaped from the molten pool easily at a

higher migration speed. Therefore, a more remarkable

Fig. 4 (Color online) Temperature contours at different LEDs.

a 193.7 J/m, p = 77.5 W, v = 400 mm/s; b 221.5 J/m, p =

88.6 W, v = 400 mm/s; c 249.5 J/m, p = 99.8 W, v = 400 mm/s;

d 275 J/m, p = 110 W, v = 400 mm/s

Sci. Bull. (2016) 61(13):1013–1022 1017

123

densification response of the molten pool was achieved in

this situation.

On increasing LED to 249.5 J/m, the molten pool

absorbed more laser energy, and the resultant temperature

and surface tension gradient increased rapidly. The out-

ward Marangoni convection was presented in conjunction

with the small-scaled vortexes in the upper region of the

molten pool (Fig. 6c). Even though the speed of melt flow

was enhanced, the behavior of bubbles still tended to be

suppressed by the small-scaled vortexes in the upper

regions of the molten pool [44]. Consequently, a fraction of

bubbles would be entrapped in the pool (Fig. 7c), caused

by the rotational flow vortexes. The densification showed a

certain extent of decrease as compared with that obtained

at 221.5 J/m. Interestingly, as the LED further increased to

275 J/m, the higher temperature and surface tension gra-

dient at the bottom of laser-induced molten pool tended to

be formed with considerable laser energy input, leading to

the formation of the outward Marangoni convection toge-

ther with the outward and inward patterns of scaling-down

vortexes in the upper and bottom zones of the molten pool,

respectively (Fig. 6d). In this situation, most of gaseous

bubbles were entrapped in the center of the upper and

bottom zones of the molten pool by the outward scaling-

down vortexes and inward scaling-down vortexes

(Fig. 7d). On the other hand, there was a high possibility

that the gaseous bubbles in the initial powder bed were

trapped and dragged into the molten pool, thereby pro-

ducing the residual porosity in the ultimately SLM-pro-

cessed parts, if the dynamics of melt flow were enhanced at

a higher level [38]. Thus, the resultant density after solid-

ification was relatively low. It was inferred that under an

appropriate LED, the reasonable Marangoni convection

could be formed in the molten pool, which favored the

Fig. 5 (Color online) Depths of SLM-processed molten pools at different LEDs. a 193.7 J/m, b 221.5 J/m, c 249.5 J/m, d 275 J/m

Fig. 6 (Color online) Velocity vector plots and patterns of convec-

tion in the molten pool in X–Z cross-sectional view at various LEDs.

a 193.7 J/m, b 221.5 J/m, c 249.5 J/m, d 275 J/m

1018 Sci. Bull. (2016) 61(13):1013–1022

123

sufficient escape of most gaseous bubbles in the finally

solidified components.

3.3 Densification mechanism

In the two-phase VOF model, when the bubbles escaped

from the molten pool, the volume would be occupied by the

metallic phase and the resultant densification would be

accordingly enhanced. As depicted in Fig. 8, a lower tem-

perature (Fig. 3a) and attendant higher melt viscosity l,were present in the molten pool at a lower LED of

193.7 J/m. The bubbles were difficult to be removed from

the molten pool and thereafter a number of spherical or

ellipse-like porosity was formed in the bottom regions of

the molten pool (Fig. 8a), inducing a lower relative density

of 95.2 % (Fig. 9a). As the LED increased to 221.5 J/m, the

significantly decreased viscosity was beneficial to the for-

mation of the lower viscous drag forces, favoring the bub-

bles to escape from the molten pool easily. Therefore, the

molten pool was nearly free of porosity (Fig. 8b), achieving

a higher relative density up to 98.9 % (Fig. 9b). As the LED

further increased to 249.5 J/m, the bubbles tended to arise

easily and then trapped by the present vortexes. In this

situation, the bubbles collided with each other within the

vortexes. A fraction of small bubbles, which were trapped

by vortexes with a lower speed, was caught and entrapped

Fig. 7 (Color online) Schematics of gaseous bubbles movement within the molten pool at various LEDs. a 193.7 J/m, b 221.5 J/m, c 249.5 J/m,

d 275 J/m

Fig. 8 Residual porosity within the molten pool in the cross-sectional

view of X–Z at various LEDs. a 193.7 J/m, b 221.5 J/m, c 249.5 J/m,

d 275 J/m

Sci. Bull. (2016) 61(13):1013–1022 1019

123

by the larger bubbles (Fig. 8c), resulting in a lower relative

density of 93.8 % (Fig. 9c). When the LED reached

275 J/m, the bubbles were entrapped by vortexes at the

bottom and upper of the molten pool (Fig. 8d), respectively,

and the resultant densification was reduced sharply to

89.3 % (Fig. 9d), caused by the lower surface tension of

melt flow in this instance [45].

3.4 Experiment verification

The cross-sectional surface morphologies of AM/3DP

components under different LEDs were characterized by

OM, as illustrated in Fig. 10. It was obvious that the

microstructural features of the SLM-processed samples

exhibited distinct variations with the applied LEDs. At the

relative lower LED of 193.7 J/m, a huge amount of nearly

sphere-shaped porosity presented in the ultimate SLM-

processed Inconel 718 Ni-base alloy. The observed

porosities, which possessed with an average size of 25 lm,

were distributed uniformly in the most regions of the part

(Fig. 10a). Instead, as the LED increased to 221.5 J/m,

little porosity was observed from the SLM-processed part

(Fig. 10b). However, at an evaluated LED of 221.5 J/m,

the considerable irregular porosity with the average size of

140 lm clustered in some local regions (Fig. 10c). At an

even higher of LED of 275 J/m, the larger-sized porosity

with the average size of 300 lm was formed in the SLM-

processed component, indicating a clustering and growing

of the porosities during SLM processing (Fig. 10d).

The formation of the residual porosity is generally

ascribed to the convection pattern and velocity of the melt

flow within the pool. On increasing the applied LED from

193.7 to 221.5 J/m, the temperature gradient is

successively enhanced in the molten pool and the resultant

surface tension and associated Marangoni convection are

formed, favoring the bubbles to escape from the pool.

Thus, the SLM-processed part presents the better

microstructural features free of the residual porosity.

However, as the considerably high LED exceeding

221.5 J/m is used, the small-scale vortexes are formed in

some local regions of the pool, which tend to entrap the

bubbles in the pool. In this condition, the considerably

entrapped bubbles experience clustering, colliding, merg-

ing and coarsening within the vortexes. Accordingly, the

residual porosity presents with a larger size of 300 lmunder the elevated LED of 275 J/m.

The quantitative analysis of densification showed that

the relative density calculated by simulation had a good

agreement with that obtained by experiments, as depicted

in Fig. 11. As the LEDs increased from 193.7 to 221.5 J/m,

the relative densities of Inconel 718 Ni-based alloy

obtained by experiments and simulations were successively

strengthened to the maximum of 98.9 % ± 0.5 % and

96.0 % ± 0.5 %, respectively. Instead, a significant

decrease of both the relative densities was obtained as the

LED exceeded 221.5 J/m. Hence, it was reasonable to

conclude that the appropriate LED of 221.5 J/m was

favorable for SLM-processed Inconel 718 Ni-based alloy to

obtain a preferable relative density, owing to the escaping

of the bubbles. The further observation was that the relative

densities obtained from experiments were in accordance

with those of simulations.

4 Conclusions

In the present study, an effective 3D FVM physical model

was performed to investigate the thermodynamic behavior

Fig. 9 (Color online) Relative density contour plots of molten pool in

X–Z cross-sectional view at variable LEDs. a 193.7 J/m, b 221.5 J/m,

c 249.5 J/m, d 275 J/m

Fig. 10 Micrographs of SLM-processed nickel-based superalloy at

different LEDs. a 193.7 J/m, b 221.5 J/m, c 249.5 J/m, d 275 J/m

1020 Sci. Bull. (2016) 61(13):1013–1022

123

and bubble migration within the molten pool under various

LEDs during SLM of Ni-based superalloy. The densifica-

tion mechanism was discussed in detail. The simulation

results were validated by the experimental results that were

conducted under the same laser processing parameters as

those used in the simulation. The conclusions regarding the

temperature distributions, migration of bubbles and densi-

fication mechanism of the molten pool with different LEDs

were drawn as follows.

(1) The LEDs played a crucial role in determining the

temperature evolutions, dimensions and surface mor-

phologies of the SLM-processed Inconel 718 Ni-based

alloy. The temperature evolutions, length, width and sur-

face integrity of the molten pool were enhanced with the

increase of the LEDs. Instead, the depth of the molten pool

was decreased with enhancing LEDs. As the LED of

221.5 J/m was properly settled, the considerably high

surface integrity was obtained. (2) The bubble migration

behavior was significantly influenced by the applied LEDs.

Employing an insufficient LED of 193.7 J/m and attendant

high viscosity resulted in a low motion speed of melt flow,

which was unfavorable for the bubbles to remove from the

pool. In contrast, the bubbles floated up with a higher speed

and escaped from the molten pool at an optimal LED of

221.5 J/m. Unfortunately, although the bubbles floated up

at an even higher velocity with the LEDs exceeding

249.5 J/m, the considerable bubbles still were entrapped

within the molten pool by the present vortexes. (3) Under a

reasonable LED of 221.5 J/m, less residual porosity and a

considerably high relative density were obtained of the

SLM-processed Inconel 718 Ni-based alloy, due to the

escaping of numerous bubbles. (4) The evolution rules of

the relative density and the corresponding distribution of

residual porosity obtained by experiments were in accor-

dance with the simulation results.

Acknowledgments This work was supported by the National Natural

Science Foundation of China (51575267, 51322509), the Top-Notch

Young Talents Program of China, the Outstanding Youth Foundation

of Jiangsu Province of China (BK20130035), the Program for New

Century Excellent Talents in University (NCET-13-0854), the Sci-

ence and Technology Support Program (the Industrial Part), Jiangsu

Provincial Department of Science and Technology of China

(BE2014009-2), the 333 high-level talents training project

(BRA2015368), the Science and Technology Foundation for Selected

Overseas Chinese Scholar, Ministry of Human Resources and Social

Security of China, the Aeronautical Science Foundation of China

(2015ZE52051), the Shanghai Aerospace Science and Technology

Innovation Fund (SAST2015053), the Fundamental Research Funds

for the Central Universities (NE2013103, NP2015206 and

NZ2016108), and the Priority Academic Program Development of

Jiangsu Higher Education Institutions.

Conflict of interest The authors declare that they have no conflict of

interest.

References

1. Chen Y, Zhang K, Huang J et al (2016) Characterization of heat

affected zone liquation cracking in laser additive manufacturing

of Inconel 718. Mater Des 90:586–594

2. Liu YG, Zhang CP, Liu CF et al (2016) The effect of nitrogen

annealing on lithium ion intercalation in nickel-doped lithium

trivanadate. Sci Bull 61:587–593

3. Zhao Y, Zhang HY, Wei H et al (2014) Progress of phase-field

investigations of CO rafting in nickel-base single-crystal super-

alloys. Chin Sci Bull 59:1684–1695

4. Kuo CM, Yang YT, Bor HY et al (2009) Aging effects on the

microstructure and creep behavior of Inconel 718 superalloy.

Mater Sci Eng A 510:289–294

5. Altin A, Nalbant M, Taskesen A (2007) The effects of cutting

speed on tool wear and tool life when machining Inconel 718

with ceramic tools. Mater Des 28:2518–2522

6. Qi H, Azer M, Ritter A et al (2009) Studies of standard heat

treatment effects on microstructure and mechanical properties of

laser net shape manufactured Inconel 718. Metall Mater Trans A

40:2410–2422

7. Thakur DG, Ramamoorthy B, Vijayaraghavan L et al (2009)

Study on the machinability characteristics of superalloy Inconel

718 during high speed turning. Mater Des 30:1718–1725

8. Wang ZY, Rajurkar KP, Fan J et al (2003) Hybrid machining of

Inconel 718. Int J Mach Tool Manuf 43:1391–1396

9. Wang Z, Guan K, Gao M et al (2012) The microstructure and

mechanical properties of deposited-IN718 by selective laser

melting. J Alloys Compd 513:518–523

10. Cloots M, Uggowitzer PJ, Wegener K et al (2016) Investigations

on the microstructure and crack formation of IN738LC samples

processed by selective laser melting using Gaussian and doughnut

profiles. Mater Design 89:770–784

11. Marcu T, Todea M, Gligor I et al (2012) Effect of surface con-

ditioning on the flowability of Ti6Al7Nb powder for selective

laser melting applications. Appl Surf Sci 258:3276–3282

12. Zhao XL, Li SJ, Zhang M et al (2016) Comparison of the

microstructures and mechanical properties of Ti–6Al–4V fabri-

cated by selective laser melting and electron beam melting. Mater

Des 95:21–31

13. Gu DD, Meiners W, Wissenbach K et al (2012) Laser additive

manufacturing of metallic components: materials, processes and

mechanisms. Int Mater Rev 57:133–164

Fig. 11 (Color online) Comparison of the relative density obtained

by numerical simulation and experiments

Sci. Bull. (2016) 61(13):1013–1022 1021

123

14. Kimura T, Nakamoto T (2016) Microstructures and mechanical

properties of A356 (AlSi7Mg0.3) aluminum alloy fabricated by

selective laser melting. Mater Des 89:1294–1301

15. Trosch T, Stroßner J, Volkl R et al (2016) Microstructure and

mechanical properties of selective laser melted Inconel 718

compared to forging and casting. Mater Lett 164:428–431

16. Jia QB, Gu DD (2014) Selective laser melting additive manu-

factured Inconel 718 superalloy parts: high-temperature oxidation

property and its mechanisms. Opt Laser Technol 162:161–171

17. Chlebus E, Gruber K, Kuznicka B et al (2015) The microstructure

and mechanical properties of deposited-IN718 by selective laser

melting. Mater Sci Eng A 639:647–655

18. Amato KN, Gaytan SM, Murr LE et al (2012) Microstructures

and mechanical behavior of Inconel 718 fabricated by selective

laser melting. Acta Mater 60:2229–2239

19. Jia QB, Gu DD (2014) Selective laser melting additive manu-

facturing of Inconel 718 superalloy parts: densification,

microstructure and properties. J Alloys Compd 585:713–721

20. Lu YJ, Wu SQ, Gan YL et al (2015) Study on the microstructure,

mechanical property and residual stress of SLM Inconel-718

alloy manufactured by differing island scanning strategy. Opt

Laser Technol 75:197–206

21. Li RD, Shi YS, Wang ZG et al (2010) Densification behavior of

gas and water atomized 316L stainless steel powder during

selective laser melting. Appl Surf Sci 256:4350–4356

22. Millett PC, Tonks M (2011) Phase-field simulations of gas den-

sity within bubbles in metals under irradiation. Comput Mater Sci

50:2044–2050

23. Qiu CL, Panwisawas C, Ward M et al (2015) On the role of melt

flow into the surface structure and pore development during

selective laser melting. Acta Mater 96:72–79

24. Thijs L, Verhaeghe F, Craeghs T et al (2010) A study of the

microstructural evolution during selective laser melting of Ti–

6Al–4V. Acta Mater 58:3303–3312

25. Thijs L, Kempen K, Kruth JP et al (2013) Fine-structured alu-

minium products with controllable texture by selective laser

melting of pre-alloyed AlSi10Mg powder. Acta Mater

61:1809–1819

26. Du LF, Wang LL, Zheng B et al (2016) Phase-field modeling of

microstructure evolution during solidification in presence of gas

bubble. Comput Mater Sci 114:94–98

27. Qiu CL, Adkins NJE, Attallah MM (2013) Microstructure and

tensile properties of selectively laser-melted and of HIPed laser-

melted Ti–6Al–4V. Mater Sci Eng A 578:230–239

28. Qiu CL, Yue S, Adkins NJE et al (2015) Influence of processing

conditions on strut structure and compressive properties of cel-

lular lattice structures fabricated by selective laser melting. Mater

Sci Eng A 628:188–197

29. Dai DH, Gu DD (2014) Thermal behavior and densification

mechanism during selective laser melting of copper matrix

composites: simulation and experiments. Mater Des 55:482–491

30. KhairallahSA AndersonA (2014) Mesoscopic simulation model

of selective laser melting of stainless steel powder. J Mater

Process Technol 214:2627–2636

31. MasmoudiA Bolot R, Coddet C (2015) Investigation of the laser-

powder-atmosphere interaction zone during the selective laser

melting process. J Mater Process Technol 225:122–132

32. Das S (2003) Physical aspects of process control in selective laser

sintering of metals. Adv Eng Mater 10:701–711

33. Li SZ, Chen R, Wang H et al (2015) Numerical study on solid–

liquid phase change in paraffin as phasechange material for bat-

tery thermal management. Sci Bull 60:1911–1926

34. Malgarinos I, Nikolopoulos N, Gavaises M et al (2015) Coupling

a local adaptive grid refinement technique with an interface

sharpening scheme for the simulation of two-phase flow and free-

surface flows using VOF methodology. J Comput Phys

300:732–753

35. Dai K, Shaw L et al (2005) Finite element analysis of the effect of

volume shrinkage during laser densification. Acta Mater

53:4743–4754

36. Bag S, Trivedi A, De A (2009) Development of a finite element

based heat transfer model for conduction mode laser spot welding

process using an adaptive volumetric heat source. Int J Therm Sci

48:1923–1931

37. Foroozmehr A, Badrossamay M, Foroozmehr E et al (2016)

Finite element simulation of selective laser melting process

considering optical penetration depth of laser in powder. Mater

Des 89:255–263

38. Nie PL, Ojo OA, Li ZG (2014) Modeling analysis of laser

cladding of a nickel-based superalloy. Surf Coat Technol

258:1048–1059

39. Hussein A, Hao L, Yan CZ et al (2013) Finite element simulation

of the temperature and stress fields in single layers built without-

support in selective laser melting. Mater Des 52:638–647

40. Yin J, Zhu H, Ke L et al (2012) Simulation of temperature dis-

tribution in single metallic powder layer for laser micro-sintering.

Comput Mater Sci 53:333–339

41. Lei YP, Murakawa H, Shi YW et al (2001) Numerical analysis of

the competitive influence of Marangoni flow and evaporation on

heat surface temperature and molten pool shape in laser surface

remelting. Comput Mater Sci 21:276–290

42. Zhu ZQ, Liu QS (2010) Coupling of thermocapillary convection

and evaporation effect in a liquid layer when the evaporating

interface is open to air. Chin Sci Bull 55:233–238

43. Yuan PP, Gu DD (2015) Molten pool behaviour and its physical

mechanism during selective laser melting of TiC/AlSi10Mg

nanocomposites: simulation and experiments. J Phys D Appl

Phys 48:035303–035318

44. Kulkarni AA, Joshi JB (2005) Bubble formation and bubble rise

velocity in gas–liquid systems: a review. Ind Eng Chem Res

44:5873–5931

45. Zhang XB, Xiang SJ, Cao XL et al (2011) Effects of surface

tension on bubble growth in an extensive uniformly superheated

liquid. Chin Sci Bull 56:3191–3198

1022 Sci. Bull. (2016) 61(13):1013–1022

123