Materials and Design 202 (2021) 109525

Contents lists available at ScienceDirect

Materials and Design

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

Substrate design to minimize residual stresses in Directed Energy

Deposition AM processes
Xufei Lu a,b,c, Michele Chiumenti a,⁎, Miguel Cervera a, Junjie Li b,c, Xin Lin b,c,⁎⁎, Liang Ma b,c,
Guohao Zhang b,c, Enquan Liang d
a
International Center for Numerical Methods in Engineering, Universidad Politécnica de Cataluña, 08034 Barcelona, Spain
b
State Key Laboratory of Solidification Processing, Northwestern Polytechnical University, Xi'an, Shaanxi 710072, China
c
Key Laboratory of Metal High Performance Additive Manufacturing and Innovative Design, MIIT China, Northwestern Polytechnical University, Xi'an, Shaanxi 710072, China
d
Shanghai Aircraft Design and Research Institute, Commercial Aircraft Corporation of China, Shanghai 200232, China

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Substrate design optimization to reduce

the residual stresses in DED.
• Coupled thermomechanical analysis for
AM by an in-house calibrated software.
• Sensitivity of conformal and non-
conformal substrates to the lower built
surface.
• Sensitivity of substrate stiffness and
thermal resistance on residual stresses.

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

Article history: This paper proposes a strategy to optimize the design of the substrate structures used in Additive Manufacturing
Received 30 September 2020 (AM) by Directed Energy Deposition (DED) to minimize the residual stresses induced by this fabrication process.
Received in revised form 12 January 2021 To this end, several numerical analyses were performed to analyse different substrate designs in order: (i) to re-
Accepted 24 January 2021
duce the sensitivity to the initial non-steady stage when the first layers of material are deposited, (ii) to optimize
Available online 28 January 2021
the heat flux through the substrate to reduce the Maximum Temperature Gradients (MTG) and, (iii) to modify
Keywords:
the substrate stiffness and its mechanical constraining to the thermal deformations during the building process
Directed energy deposition (DED) and the cooling phase. To ensure the reliability of the numerical simulations, an in-house software is calibrated
Substrate design to allow for an accurate analysis of DED. Thus, an experimental setting is undergone to feed the numerical
Residual stresses model with suitable values of both material and process parameters through temperature and displacement
Thermo-mechanical analysis measurements and numerical fitting. Once calibrated, the software is used to evaluate the performance of several
substrate designs to mitigate the residual stresses induced by the DED process. A thin-walled rectangular part se-
lected as industrial demonstrator showed a significant reduction (up to 62%) of the maximum tensile stresses.
© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).

1. Introduction

Additive manufacturing (AM) is an advanced industrial technology

⁎ Corresponding author. for manufacturing high-performance metallic parts through a layer-
⁎⁎ Corresponding author at: State Key Laboratory of Solidification Processing,
Northwestern Polytechnical University, Xi'an, Shaanxi 710072, China.
by-layer material deposition [1,2]. In Directed Energy Deposition
E-mail addresses: [email protected] (M. Chiumenti), [email protected] (DED) the metal powder is blown into the melting pool generated by
(X. Lin). a coaxial high energy source (e.g. laser or electron beam). Despite

https://doi.org/10.1016/j.matdes.2021.109525
0264-1275/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

being one of the most versatile AM processes used either to fabricate transfer of the support structures [29–31]; these are necessary when
new components or to repair damaged parts [3], the DED technology printing overhangs in SLM and they affect the development of thermal
is characterized by high-energy sources which generate high tempera- stresses and deformations of the build [29–33]. Contrariwise, in DED
ture fields as well as high temperature gradients. As a consequence, processes the material is directly deposited on the substrate and the
both residual (tensile) stresses and plastic deformations generally affect use of supporting structure is generally avoided. Therefore, the overall
the final metallurgy as well as the structural performance of in service stiffness of the building structure depends on the design of the substrate
AM parts [3–5,38,39]. and its optimization is marginally studied in the literature. A thick sub-
To mitigate the residual stresses and distortions of AM components, strate fixed on the work-bench is generally used to prevent the develop-
several researchers have investigated the sensitivity of the DED process ment of undesired thermal deformations. However, a very rigid
to the scanning strategy, the process parameters and the geometry of substrate generates large residual stresses which can transform into
the built structures. It has been observed that the residual stresses can plastic deformations during the cooling process. Moreover, during the
be mitigated by reducing the local heat accumulation during the cutting process when the substrate is removed, the residual stresses re-
metal deposition. This can be achieved by reducing the scanning length laxation induces a further distortion of the component [17]. Thus, the
and/or by increasing the scanning speed [6–8]. The substrate distortion optimization of the substrate stiffness in order to mitigate the residual
can be alleviated by optimizing the process parameters (e.g. source stresses and the component distortion is the second objective of
power, scanning speed, feeding rate, etc.) [9–11]. Some other research this work.
works [12–15] revealed that the geometry of the component and, in
particular, its local stiffness, also plays an important role in the develop- 2. Numerical model
ment of residual stresses in the build. Mukherjee et al. [16,17] demon-
strated how the thermal distortion of AM parts depends on the The numerical framework used for the high-fidelity modelling of AM
material properties, the deposition strategy, the process parameters, processes is based on a coupled 3D thermo-mechanical FE module for
the geometry of the component, as well as the preheating and the transient analysis as part of the in-house software platform COMET
cooling conditions. Mitigation of residual stresses can be achieved by [34] developed at the International Centre for Numerical Methods in Engi-
preheating the substrate before the AM process, thus reducing of the neering (CIMNE). The thermo-mechanical coupling is handled by a stag-
large thermal gradients during the metal deposition of the first few gered solution: at each time-step, the transient heat transfer problem is
layers [18–20]. Li et al. [21] reported a remarkable reduction of both solved first. Next, the mechanical analysis is solved accounting for the
longitudinal bending and angular distortion by replacing continuous temperature-dependent material parameter data-base.
clamping condition with a more flexible multi-point support strategy.
Denlinger et al. [22] optimized the metal deposition on the base plate 2.1. Thermal analysis for AM
to mitigate the longitudinal bending effects induced by the thermal
field. Afazov et al. [23,24] proposed the geometry compensation strat- The balance of energy is the governing equation used to solve the
egy to modify the nominal geometry of AM components according to heat transfer analysis. The local (strong) form of this equation is:
the thermal deformation predicted by the numerical analysis. Finally,
Hönnige et al. [25,26] showed the advantages of the rolling techniques Ḣ ¼ −∇∙q þ þ Q̇ ð1Þ
in WAAM processes to reduce distortions and residual stresses.
:
Fig. 1 shows the transition from the “non-steady stage” to the “quasi-
where H is the enthalpy rate. The heat flux q is defined through Fourier's
steady stage” during the DED process of a Ti-6Al-4V rectangular frame
law:
[12]. During the entire non-steady stage, both the thickness and the
width of the deposited layers are affected by the heat absorption
q ¼ −k∇T ð2Þ
through the (initially) cold substrate. As a general conclusion, in DED
process net-shape (nominal) geometries can obtained by reducing the
Maximum Thermal Gradients (MTG). Such MTG generally appear where k is the temperature-dependent thermal conductivity.
:
when the first layers of metal are deposited on the cold substrate. There- The heat source (per unit of volume), Q , is defined in terms of the
:
fore, to enhance the DED process it is mandatory to reduce the MTG by total energy input (e.g. laser source), P , the heat absorption efficiency,
either pre-heating the substrate [18–20] or reducing the heat flux from ηp and the volume of the melt pool, VΔtpool, as:
the Heat Affected Zone (HAZ) to the workbench through the substrate
[6–9,25–28]. Thereby, the first objective of this work consists of opti- :
: ηp P
mizing the non-steady stage to obtain high quality DED parts. Q¼ ð3Þ
In Selected Laser Melting (SLM) the workbench is very thick and V Δt
pool
rigid. Therefore, the focus is placed on the stiffness and the heat flux

Fig. 1. The transition from the non-steady stage to quasi-steady stage of a Ti-6Al-4V part deposited by DED.

2
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

Hence, the heat source is uniformly spread inside the melt-pool be- rffiffiffi
cause of the lack of resolution of the FE meshes generally used for the 2 
Φðs, qh , T Þ ¼ ‖s‖− σ y ðT Þ−qh ð12Þ
AM analysis at global (component) level [35–37]. 3
The heat loss by convection is computed by using Newton's law:
where σy(T) is the temperature-dependent elastic limit controlling the
qconv ¼ hconv ðT−T env Þ ð4Þ thermal softening and qh is the stress-like variable controlling the iso-
tropic strain-hardening, expressed as:
where hconv is the temperature-dependent Heat Transfer Coefficient  h i
qh ðξ, T Þ ¼ − σ ∞ ðT Þ−σ y ðT Þ 1−e−δðT Þξ −hðT Þξ ð13Þ
(HTC) by convection, T is the temperature at the component surface
and Tenv is the environment temperature.
The heat dissipation by the radiation is defined by Stefan- where ξ is the isotropic strain-hardening variable, σ∞(T) is the
Boltzmann's law: temperature-dependent saturation flow stress, while δ(T) and h(T) are
the temperature-dependent parameters to control the exponential


qrad ¼ εrad σ rad T 4 −T 4env ð5Þ and linear hardening laws, respectively.
The deviatoric counterpart of Cauchy's stress tensor is computed as:

where εrad is the surface emissivity and σrad is the Stefan-Boltzmann s ¼ 2Gðe−evp Þ ð14Þ
constant.
Note that the heat dissipation by convection and radiation mecha- where G = G(T) is the temperature-dependent shear modulus, while
nism depends on the actual boundary condition (i.e. the current exter- e and evp are the total (deviatoric) and the visco-plastic strain, respec-
nal surfaces of the built). Thus, the heat flow is proportional to the tively. The former is extracted from the total strain tensor ε(u) = ∇sym
evolution of the building process [35–37]. (u), while the evolution laws of both the visco-plastic strain tensor
and the isotropic strain-hardening variable are obtained through the
2.2. Mechanical analysis principle of maximum plastic dissipation as:

The mechanical problem is determined by the balance of momen-

: : ∂Φðs, qh Þ : s :
e vp ¼ γ vp ¼ γ vp ¼ γ vpn ð15Þ
tum and the continuity equations: ∂s ‖s‖
rffiffiffi
: ∂Φðs, qh Þ
∇∙s þ ∇p þ b ¼ 0 ð6Þ : 2:
ξ¼ γ vp ¼ γ vp ð16Þ
∂qh 3
  p
∇∙u−eT − ¼ 0 ð7Þ :
K where n is the normal of the yield surface, and γ vp is the visco-plastic
multiplier, defined as:
where the Cauchy stress tensor, σ is split into its hydrostatic (pressure),
p and deviatoric parts s, respectively, as: : ∂Φðs, qh Þ 1
γ vp ¼ 〈 〉m ð17Þ
η
σ ¼ pI þ sðuÞ ð8Þ
where 〈∙〉 is the Macaulay bracket, while η = η(T) and m = m(T) are the
temperature-dependent plastic viscosity and rate sensitivity,
b are the body forces per unit of volume and K(T) is the temperature-
respectively.
dependent bulk modulus, controlling the material compressibility. The
Note that, when the temperature is approaching the annealing
thermal deformation, eT is expressed as:
value, σy(T) → 0. Thus, the deviatoric Cauchy stress reduces to:
eT ðT, f S Þ ¼ ecool ðT Þ þ epc ð f S Þ ð9Þ : m :
s ¼ η γ vp n ¼ ηeff e vp ð18Þ

ecool(T) and epc(fS) are the thermal expansion/contraction and ther- : m−1
where ηeff ¼ η γ vp is the effective viscosity. Thereby, above the
mal shrinkage during the liquid-to-solid phase-change, as a function annealing temperature the material is characterized by a purely viscous
of the initial temperature T0 and the solid fraction fS, respectively, and law [36,37]. A non-Newtonian behavior m > 1 is used for the mushy
defined as: phase (from the annealing to the liquid temperature), while a Newto-
nian law, m = 1, characterizes the liquid phase (temperature values
ecool ðT Þ ¼ αðT ÞðT−T 0 Þ ð10Þ above the melting point).

epc ð f S Þ ¼ β f S ð11Þ 2.3. FE modelling of the DED process

pc
where α = α(T) and β ¼ ΔV V 0 are the thermal expansion and the shrink- To simulate the DED process, a time discretization is required. A time
age coefficients, respectively. step, Δt = tn+1 − tn, characterizes the time-marching scheme. Thereby,
Note that the mechanical problem defined by Eqs. (6)–(7) depends the melt-pool is allowed to move step-by-step along the scanning pat-
on both the displacements, u, and the pressure field, p and is suitable tern from the actual position defined at time tn to that at time tn+1.
for both compressible and the fully incompressible (isochoric) material Within this interval, the heat is distributed to the elements belonging
behaviors. to the melt pool volume and, at the same time, the blown powder con-
In AM, the temperature field varies from the room temperature to forms the metal deposition which characterizes this AM process.
(and above) the melting point. Thereby, the material must be character- The numerical simulation of the DED process needs an ad-hoc proce-
ized in its solid, mushy and liquid phases. dure to classify the elements into: active, inactive and activated ele-
A J2-thermo-elasto-visco-plastic model is assumed in the solid ments. At each time-step, an octree-based searching algorithm is used
phase, from the room to the annealing temperature. All the material to find the elements belonging to the melt-pool (active) and to the
properties are assumed as temperature-dependent. The von-Mises new metal deposition (activated). Thus, the software reads the same
yield-surface is defined as: input file (Common Layer Interface – CLI format) following the actual

3
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

Fig. 2. 3D FE meshes used for analysing the DED process: (a) single-wall used to calibrate the AM model; (b) single-wall with rounded corners; (c) hollow structure design for the
substrate; (d) demonstrator with hollowed structure substrate.

Table 1
Temperature-dependent material properties of Ti-6Al-4V [12].

Temperature Density Thermal Heat Poisson's Thermal Expansion Young's Elastic

(°C) (kg/m3) Conductivity Capacity Ratio Coefficient Modulus Limit
(W/(m·°C)) (J/(kg·°C)) (μm/(m·°C)) (GPa) (MPa)

20 4420 7 546 0.345 8.78 110 850

205 4395 8.75 584 0.35 10 100 630
500 4350 12.6 651 0.37 11.2 76 470
995 4282 22.7 753 0.43 12.3 15 13
1100 4267 19.3 641 0.43 12.4 5 5
1200 4252 21 660 0.43 12.42 4 1
1600 4198 25.8 732 0.43 12.5 1 0.5
1650 3886 83.5 831 0.43 12.5 0.1 0.1
2000 3818 83.5 831 0.43 12.5 0.01 0.01

scanning sequence as used to inform the DED machine. The birth-and- 2.5. Material properties and boundary conditions
death-element technique is used to activate the elements belonging to
the new deposited layer. Both active and activated elements belong to The temperature-dependent thermal and mechanical properties of
the current computational domain, while the inactive elements are nei- Ti-6Al-4V adopted for the material characterization are listed in
ther computed nor assembled into the global matrix of the problem Table 1 [12]. Note that the thermal conductivity is increased up to
[35]. 83.5 W/(m·°C) to take into account the convective flow inside the melt-
ing pool when the temperature is above the melt point [41].
2.4. Geometry model and FE meshes The thermal boundary conditions adopted for all the numerical sim-
ulations account for the heat loss by convection and radiation through
The generation of the CAD geometry, the meshing operation as well the external surfaces of both the build and the substrate. The HTC by
as the post-processing of the results are carried out using the in-house convection and the emissivity are hconv = 8 W/(m2·°C) and εrad = 0.7,
pre-post-processor GiD [40]. Fig. 2 shows the 3D geometries and the respectively. The HTC by conduction controlling the heat flux at the con-
corresponding FE meshes of different builds. Fig. 2(a) shows the tact interface between the base-plate and the clamping system (or the
single-wall model used to calibrate the AM software: it consists of workbench) is set to hconv = 100 W/(m2·°C). During the AM process,
19,040 hexahedral elements and 24,285 nodes. The base plate is the ambient temperature was Tenv = 25 °C. The power absorption in
clamped on the left-hand side as a cantilever beam, while the other DED process when using Ti-6Al-4V is set to ηp = 0.3.
end is free. Fig. 2(b) shows the same geometry but including some These parameters have been calibrated using the experimental data
extra material to alleviate the stress concentrations at the corners obtained through the in situ measurements as described in the next
where the wall is attached to the substrate. Fig. 2(c) depicts a new sub- section.
strate design characterized by a hollow structure, proposed to reduce
the residual stresses and the sensitivity to the initial non-steady stage.
Finally, a more complex part is presented in Fig. 2(d) and it is used as 3. Experimental setting
the industrial demonstrator to validate the optimization strategy for
the substrate design. The corresponding mesh consists of 44,000 Fig. 3 shows the experimental setting used for the calibration of the
hexahedral elements and 56,400 nodes. software platform. The single-wall shown in Fig. 2(a) is used for this

4
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

±0.75% (or 2.2 °C) were used to measure the temperature history at
points CH1, CH2, and CH3 of the substrate. Additionally, a Graphtec
GL-900 high-speed eight-channel data-logger was used to record the
signals from the thermocouples and displacement sensors.
Spherical Ti-6Al-4V powder (45–325 μm) produced via a plasma ro-
tating electrode process is dried in a vacuum oven at 120 °C during 2.5 h
before the blown-powder process. The DED machine is equipped with a
YLS-3000 IPG Photonics fibre laser system having a wave-length rang-
ing between 960 and 1200 nm and a maximum output power of
3 kW. During the fabrication, the laser power used was 1.5 kW with
3 mm of spot size and a Gaussian distribution profile. The laser scanning
speed is 10 mm/s. The up-lift height is set to 0.5 mm with a feeding rate
of 12.0 g/min. The average penetration depth is 0.5 mm.
Fig. 3. In-situ measurement set-up for the calibration of the numerical simulation of the The metal deposition is made of 40 layers, approximately 20 mm
DED process: single-wall on a substrate fixed as a cantilever. high, 80 mm long and 4 mm wide (one single hatching). The whole
building process is carried out in a closed chamber with protective gas
purpose. An annealed Ti-6Al-4V substrate 140 mm long, 6 mm wide and (pure argon).
25 mm high is used for the DED process.
Fig. 4 shows the locations of the thermocouples and the displace- 4. Calibration of DED model
ment sensor. The displacement sensor (WXXY PM11-R1-20L) has an ac-
curacy of 0.02%, a maximum range of 20 mm and was employed for the The thermo-mechanical FE model as well as the material data used
in-situ measurement of the vertical displacement at point DS placed on in this work have been validated in previous works [3,12,19,35–37].
the bottom surface of the substrate during the entire DED process. Here, only the heat dissipation by convection and conduction are cali-
Omega GG-K-30 type thermocouples with a recording uncertainty of brated. The numerical fitting of the experimental measurements is

Fig. 4. Single-wall: dimensions and location of the thermocouples (CH1, CH2 and CH3) and the displacement sensor (DS).

Fig. 5. Experimental measurements and numerical results: (a) temperature evolution at CH1, CH2 and CH3; (b) vertical displacement history at DS point.

5
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

Fig. 6. Residual von Mises stress field of the single-wall.

shown in Fig. 5. In this figure, the temperature curves shown corre- stiffness. On the one hand, if the heat flux is reduced, the thermal gradi-
sponds to the measurements at the thermocouple locations CH1, CH2 ents during the initial non-steady stage of the metal deposition can be
and CH3, while the vertical displacement is obtained at point DS (see reduced also. On the other hand, by reducing the global stiffness of the
Fig. 4). The agreement between experiments and numerical results is substrate, it is possible to achieve a weaker mechanical constraining be-
remarkable. tween the AM-build and the substrate, reducing the formation of plastic
deformations. The drawback of the latter is the possibility of inducing a
larger distortion of the component.
5. Substrate optimization strategy

Fig. 6 shows the von Mises stress indicator corresponding to the re- 5.1. Minimization of the stress concentrations
sidual stress field at the end of the DED process. The highest values are
located at the basement of the single-wall deposition, as previously ob- In order to minimize the stress concentrations at the interface be-
served in [3,12,17]. This is mainly due to the large temperature gradi- tween the building structure and the substrate, two possibilities are in-
ents generated during the deposition of the first layers of the building vestigated: (i) rounding off the sharp angles where the part is attached
over the cold substrate during the initial non-steady stage. The contrac- to the substrate; and (ii) conforming the substrate to the bottom surface
tion of these layers is restrained by the substrate, thus generating large of the build.
residual stresses. The first idea is very well known and is actually applied in many fab-
The optimization of the substrate geometry aims both to enhance rication processes to avoid crack initiation at the stress concentration
the heat dissipation through the substrate and to reduce its global points. Therefore, all the corners are rounded off to avoid sharp angles.

Fig. 7. Predicted residual von Mises stress field: (a) by rounding off the corners using different arc radii (R). (b) by using different tangent angles (θ) with respect to the substrate.

6
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

Fig. 8. Comparison of the residual von Mises stress distributions along the top surface of the substrate: (a) by rounding off the corners using different arc radii (R); (b) by using different
tangent angles (θ) with respect to the substrate.

The effectiveness of this method in AM is tested in this work by taking 5.2. Size optimization of the substrate
the rounding angle θ between top surface of the substrate and the
single-wall as the design parameter. The sample with sharp angles is In this section, the optimization of the substrate size is investigated.
used as the reference. The distance, d, between the left-hand side of the substrate and the
Figs. 7(a) and 8(a) show the contour fills and the distributions of the single-wall is used as design parameter (see Fig. 4). This dimension is
residual von Mises stress using different rounding radii varying from 2 progressively reduced from 20 mm to 15 mm, 10 mm, 5 mm and
to 10 mm. In both figures, it can be observed that the maximum residual 0 mm, respectively.
stresses increase by expanding the arc radius. This is due to the fast The evolution of the vertical displacement recorded during the entire
cooling of the extra material used to print this small zone. However, DED process simulation is shown in Fig. 9(a) for all the test-cases. As the
the highest stress values are located outside the nominal geometry of wall of the AM deposition gets closer to the free-end of the substrate, the
the single-wall, and more precisely, at the external sides of the material vertical displacement of the substrate at point DS gradually decreases. It
used for the rounding off. Thus, the residual stresses are effectively min- is also relevant to observe that most of the substrate distortion occurs
imized in the actual build. during the non-steady stage, when the first 5 layers are deposited on
As an alternative, the material supply for the rounding off can be re- the cold substrate. If compared with the reference configuration (d =
duced assuming non-tangent configurations where the angle θ with re- 20 mm), when d = 0 mm (i.e. the substrate conforms to the lower sur-
spect to the substrate is varied from 15° to 75°. Also in this case, the face of the build), this vertical displacement is reduced by 30%. The lon-
values of the residual von Mises stress depicted in Fig. 7(b) and Fig. 8 gitudinal residual stress (σxx) distributions at the basement of the build
(b) are increasing when reducing θ but, again, the corresponding are plotted in Fig. 9(b) and in the contour-fills shown in Fig. 10. It can
highest values are moved far away from the nominal geometry of the be seen that the stress concentration on the right-hand side of the
wall. Taking into account the material waste and practical experience, single-wall gradually decreases according to the size of the substrate,
it is found that θ = 45° is the optimal choice for minimizing the stress to vanish when d = 0. This is because the mechanical constraining in-
concentration and preventing crack initiation. duced by the substrate is gradually reduced, allowing for the free thermal

Fig. 9. Optimization of the substrate size for the single-wall: (a) vertical displacement histories at point DS; (b) distribution of longitudinal residual stresses (σxx) on the top surface of the
substrate.

7
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

Fig. 10. Contour-fill of the longitudinal residual stress (σxx).

Fig. 11. Different substrates for the metal deposition part during DED process.

Fig. 12. Temperature evolution of point P: (a) solid substrates with increasing height; (b) hollow substrates.

expansion/contraction of the deposited materials. Thus, residual stresses (ii) hollow substrates with a fixed height of 15 mm but with different
during the AM process can be satisfactorily mitigated. hollow shapes, as shown in Fig. 11. These structures are suitable for in-
dustrial DED fabrications and respond to two objectives: (1) reducing
5.3. Optimization of the substrate design the heat flux through the substrate, and (2) reducing the substrate stiff-
ness. As a consequence, the initial non-steady phase can be minimized
In this section, the design optimization of the substrate structure is because of a faster substrate heating. Moreover, the mechanical
investigated. Two types of substrates are analyzed: (i) solid substrates constraining due to the substrate stiffness is also reduced and thus, the
with increasing height of 5 mm, 15 mm, and 25 mm, respectively; and residual stresses induced by the metal deposition process are mitigated.

8
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

Fig. 13. Norm of the displacements of the single-wall according to the different substrates analyzed: (a) solid substrates; (b) hollow substrates.

The workbench is assumed as rigid, so the displacements at the bot- Fig. 13 shows the norm of the displacement field at the end of the
tom surface of the substrate are fixed as in the actual DED process. DED process, for all the tested substrate structures. As shown in
Moreover, the size of the substrate is reduced to 80 mm, the same as Fig. 13(a), by increasing the height of the solid substrate, the angular
the single-wall. Therefore, the stress concentrations at both side of the distortion of the build is also incremented. Thus, the hot material depos-
single wall are avoided. ited during the fabrication of the single-wall is no longer able to contract
Fig. 12 shows the temperature evolution in the middle point of the top because of the higher constraining of the substrate. However, the in-
surface of the substrate for the three substrate thicknesses and for the creased stiffness reduces the formation of residual stresses, as shown
three different hollow substrates. On the one hand, the higher is the sub- in Fig. 14(a).
strate, the slower is the heat dissipation from the HAZ to the workbench A different behavior is appreciated when using the proposed hollow
through the substrate. Hence, the steady state conditions can be reached structures. In this case, the distortion is reduced by increasing the num-
faster. On the other hand, the average temperature when using the hol- ber of hollows without compromising the residual stress distribution as
low substrates is higher than in the cases of solid substrates. Therefore, shown in Figs. 13(b) and 14(b), respectively.
the proposed design is helpful to increase the substrate temperature Fig. 15 shows the predicted residual stresses at the interface be-
and to reduce the non-steady phase at the beginning of the DED process. tween the single-wall and the substrate, for all the analyzed substrates.

Fig. 14. Von Mises residual stress field at the end of the DED process: (a) solid substrates; (b) hollow substrates.

9
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

Fig. 15. Residual stress distributions on the wall-substrate interface for the different substrate structures: (a) von Mises stresses; (b) longitudinal stresses.

Fig. 16. Wire-cutting process: (a) cutting process; (b) evolution of the vertical displacement at the free end of the substrates as a function of the cutting time.

Fig. 17. Geometry compensation strategy: (a) required steps for the geometric compensation; (b) single-wall AM analysis without geometric compensation; (c) with geometric
compensation.

10
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

Both longitudinal and von Mises stresses are mitigated when adopting shown in Fig. 18(a), used as demonstrator. The AM thermo-
hollow substrates. mechanical analysis of this component was reported in a previous
To prove the effective mitigation of the residual stresses, both solid work [12]. Here, the objective is to investigate three new substrate de-
and hollow substrates are removed by wire-cutting, as shown in signs referred to as: solid, rectangular and hollow substrates, respec-
Fig. 16(a). The evolution of the vertical displacement at the free end of tively. Also in this case, the workbench is not simulated and the
the part is presented in Fig. 16(b) as a function of the cutting time. It bottom surface of all substrates is fixed.
can be observed that the stress release induced by the cutting process The temperature evolution at point T for all substrates is depicted
of the solid substrate generates a pronounced residual distortion of in Fig. 18(b), while the corresponding MTG (Gxx) along the scanning
the AM part. Contrarily, removing the hollow substrate only affects mar- direction are shown in Fig. 18(c). As for the single-wall, reducing the
ginally the distortion of the single-wall. Therefore, the amount of resid- mass of the substrate or/and defining a hollow structure induces
ual stresses accumulated in the AM part when using the solid substrate higher peak and average temperatures of the substrate. Hence, the
is larger than for the hollow substrate. substrate is forced to heat faster reducing the transient stage. Also in
As shown in Fig. 13, the drawback when reducing the substrate stiff- this case, the hollow substrate shows the fastest fall of the MTG
ness is the increase of thermal warpage. A possible solution to recover (Fig. 18 (c)), indicating the rapid establishment of the quasi-steady
the nominal geometry is using the geometry compensation technique stage and a better performance being the preferred design for its
[23,24,42]. Fig. 17(a) summarizes the necessary steps: (1) starting industrialization.
from the nominal geometry, the DED process is simulated to compute Fig. 19 compares the evolution of the longitudinal stress (σxx) at
the warpage due to the thermal deformation and material shrinkage; the mid-point of the YZ cross-section for all the proposed substrate
(2) this information is used to modify the nominal geometry generating structures. This figure shows that, also in this case, the maximum
the compensated model; (3) the process simulation is repeated to check longitudinal tensile stresses are produced at the interface between
if the resulting distortions do not exceed the fabrication tolerances. If the part and the substrate just after the deposition of the first
this is not the case, steps (2) and (3) are repeated until the desired geo- layer. By increasing the number of layers, the tensile stresses grad-
metrical accuracy is reached. Fig. 17(b) illustrates the residual part dis- ually reduce according to the reduction of the MTG (see Fig. 18(c)).
tortion without geometry compensation, at the end of step (1), and at Nevertheless, more flexible substrates produce lower tensile
the end of the compensation loop (see Fig. 17(c)). Hence, this strategy stresses in the build. The hollow substrate showed a 62% reduction
can be efficiently used to obtain high precision AM components [42]. of the maximum longitudinal tensile stress. Fig. 20 shows both the
final distortion of the part (displacements norm) and the residual
6. Validation of the substrate design strategy von Mises stress indicator for the different substrates analyzed.
Also for the selected demonstrator, more flexible substrates gener-
The validation of the proposed methodology to improve the AM fab- ate lower residual stresses even if the final distortion can be
rication by DED is carried out using the rectangular-shape geometry slightly higher.

Fig. 18. Substrate structure optimization: (a) CAD geometries of the different substrates analyzed; (b) thermal histories at point T; (c) MTG (Gxx) evolution as a function of the number of
deposited layers.

11
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

Fig. 19. Evolution of the longitudinal stresses (σxx) at the mid YZ cross section during the AM process: (a) reference substrate; (b) solid substrate; (c) rectangular substrate; (d) hollow
substrate.

Fig. 20. Mechanical responses of the demonstrator: (a) final distortions (displacements norm); (b) residual von Mises stresses.

12
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

7. Concluding remarks [6] W. Woo, D.K. Kim, E.J. Kingston, et al., Effect of interlayers and scanning strategies on
through-thickness residual stress distributions in additive manufactured ferritic-
austenitic steel structure[J], Mater. Sci. Eng. A 744 (2019) 618–629.
A 3D thermo-mechanical FE model is calibrated by in-situ experi- [7] K. Ren, Y. Chew, J.Y.H. Fuh, et al., Thermo-mechanical analyses for optimized path
mental measurements using the single-wall structure fabricated by planning in laser aided additive manufacturing processes[J], Mater. Des. 162
(2018) 80–93.
DED processes. This done, the software platform is used to analyse dif-
[8] Q. Wu, T. Mukherjee, C. Liu, et al., Residual stresses and distortion in the patterned
ferent substrate structures and some fabrication recommendations are printing of titanium and nickel alloys[J], Add. Manufact. 29 (2019) 100808.
drawn: [9] Y. Lee, Y. Bandari, P. Nandwana, et al., Effect of interlayer cooling time, constraint
and tool path strategy on deformation of large components made by laser metal de-
position with wire[J], Appl. Sci. 9 (23) (2019) 5115.
(1) Most of the residual stresses in DED processes are generated dur-
[10] F. Soffel, D. Eisenbarth, K. Wegener, Minimizing Substrate Distortion in Direct Metal
ing the deposition of the first layers, when the substrate is still Deposition (DMD) with Optimized Process Parameters and Buildup Strategies[C]//
cold and very high temperature gradients develop. When the 3rd International Symposium on Additive Manufacturing (ISAM 2019), 2019.
quasi-steady conditions are reached, much lower residual [11] Z. Wang, E. Denlinger, P. Michaleris, et al., Residual stress mapping in Inconel 625
fabricated through additive manufacturing: method for neutron diffraction mea-
stresses are generated during the metal deposition. surements to validate thermomechanical model predictions[J], Mater. Des. 113
(2) Rounding off sharp angles is a good practice to reduce the stress (2017) 169–177.
concentrations in AM as observed in many other fabrication pro- [12] X. Lu, X. Lin, M. Chiumenti, et al., Residual stress and distortion of rectangular and S-
shaped Ti-6Al-4V parts by directed energy deposition: modelling and experimental
cesses. Moreover, focussing on the size of the substrate, the best
calibration[J], Add. Manufact. 26 (2019) 166–179.
option is to consider a substrate conformal to the bottom surface [13] Q. Yang, P. Zhang, L. Cheng, et al., Finite element modeling and validation of
of the AM component, so that possible sharp angles are avoided thermomechanical behavior of Ti-6Al-4V in directed energy deposition additive
and the flexibility of the substrate is enhanced. manufacturing[J], Add. Manufact. 12 (2016) 169–177.
[14] A. Grégoire, L. Jakabčin, Taking into account thermal residual stresses in topology
(3) Reducing the size of the substrate is beneficial to mitigate resid- optimization of structures built by additive manufacturing[J], Math. Models
ual stresses because both the heat flux and the stiffness are re- Methods Appl. Sci. 28 (12) (2018) 2313–2366.
duced, allowing for a faster non-steady stage and a lower [15] R. Ranjan, Y. Yang, C. Ayas, et al., Controlling local overheating in topology optimiza-
mechanical constraining. tion for additive manufacturing[C], Dimensional Accuracy and Surface Finish in Ad-
ditive Manufacturing, 2017.
(4) Reducing the stiffness of the substrate can increase the final dis- [16] T. Mukherjee, V. Manvatkar, A. De, et al., Mitigation of thermal distortion during ad-
tortion and warpage of the AM-build. The geometric compensa- ditive manufacturing[J], Scr. Mater. 127 (2017) 79–83.
tion strategy can be employed to recover the nominal geometry [17] H.L. Wei, T. Mukherjee, W. Zhang, et al., Mechanistic models for additive
manufacturing of metallic components[J], Prog. Mater. Sci. 100703 (2020).
at the end of the fabrication.
[18] J. Cao, M.A. Gharghouri, P. Nash, Finite-element analysis and experimental valida-
(5) Hollow substrate structures reduce the transient stage of the DED tion of thermal residual stress and distortion in electron beam additive
fabrication. The heat flow through the substrate to the work- manufactured Ti-6Al-4V build plates[J], J. Mater. Process. Technol. 237 (2016)
bench is slowed. Hence, the substrate heats up much faster 409–419.
[19] D.J. Corbin, A.R. Nassar, E.W. Reutzel, et al., Effect of substrate thickness and
than in the case of a solid structures. preheating on the distortion of laser deposited Ti–6Al–4V[J], J. Manuf. Sci. Eng.
140 (6) (2018), 061009.
[20] X. Lu, X. Lin, M. Chiumenti, et al., Finite element analysis and experimental valida-
tion of the thermomechanical behavior in laser solid forming of Ti-6Al-4V[J], Add.
Manufact. 21 (2018) 30–40.
Declaration of Competing Interest [21] F. Li, S. Chen, J. Shi, et al., In-process control of distortion in wire and arc additive
manufacturing based on a flexible multi-point support fixture[J], Sci. Technol.
The authors declare no conflict of interest. Weld. Join. 24 (1) (2019) 36–42.
[22] E.R. Denlinger, P. Michaleris, Mitigation of distortion in large additive manufacturing
parts[J], Proc. Inst. Mech. Eng. B J. Eng. Manuf. 231 (6) (2017) 983–993.
Acknowledgements [23] S. Afazov, A. Okioga, A. Holloway, et al., A methodology for precision additive
manufacturing through compensation[J], Precis. Eng. 50 (2017) 269–274.
This research was funded by the National Key Technologies R & D [24] A. Yaghi, S. Ayvar-Soberanis, S. Moturu, et al., Design against distortion for additive
manufacturing[J], Add. Manufact. 27 (2019) 224–235.
Program (No. 2016YFB1100104) and the Natural Science Foundation [25] J.R. Hönnige, P.A. Colegrove, B. Ahmad, et al., Residual stress and texture control in
of Shaanxi Province of China (Grant No. 2018JM5115). Funding from Ti-6Al-4V wire + arc additively manufactured intersections by stress relief and
the Spanish Ministry of Economy and Competitivity (MINECO) under rolling[J], Mater. Des. 150 (2018) 193–205.
the ADaMANT project: Computational Framework for Additive [26] J.R. Hönnige, P.A. Colegrove, S. Ganguly, et al., Control of residual stress and distor-
tion in aluminium wire + arc additive manufacture with rolling[J], Add. Manufact.
Manufacturing of Titanium Alloy Components (ref: DPI2017-85998-P), 22 (2018) 775–783.
within the Excellency Programme for Knowledge Generation is grate- [27] C. Baykasoğlu, O. Akyildiz, M. Tunay, et al., A process-microstructure finite element
fully acknowledged. CIMNE is recipient of the Severo Ochoa Award of simulation framework for predicting phase transformations and microhardness for
directed energy deposition of Ti6Al4V[J], Add. Manufact. 35 (2020) 101252.
Excellence from MINECO. This work was also funded by the European
[28] C. Baykasoglu, O. Akyildiz, D. Candemir, et al., Predicting microstructure evolution
KYKLOS 4.0 project: An Advanced Circular and Agile Manufacturing Eco- during directed energy deposition additive manufacturing of Ti-6Al-4V[J], J.
system based on rapid reconfigurable manufacturing process and individu- Manuf. Sci. Eng. 140 (5) (2018), 051003.
alized consumer preferences (No. H2020-DT-2019-1). [29] F. Calignano, Design optimization of supports for overhanging structures in alumi-
num and titanium alloys by selective laser melting[J], Mater. Des. 64 (2014)
203–213.
References [30] M. Langelaar, Combined optimization of part topology, support structure layout and
build orientation for additive manufacturing[J], Struct. Multidiscip. Optim. 57 (5)
[1] D. Herzog, V. Seyda, E. Wycisk, et al., Additive manufacturing of metals[J], Acta (2018) 1985–2004.
Mater. 117 (2016) 371–392. [31] G. Allaire, B. Bogosel, Optimizing supports for additive manufacturing[J], Struct.
[2] H. Huang, N. Ma, J. Chen, et al., Toward large-scale simulation of residual stress and Multidiscip. Optim. 58 (6) (2018) 2493–2515.
distortion in wire and arc additive manufacturing[J], Add. Manufact. 34 (2020) [32] L. Cheng, X. Liang, J. Bai, et al., On utilizing topology optimization to design support
101248. structure to prevent residual stress induced build failure in laser powder bed metal
[3] X. Lu, X. Lin, M. Chiumenti, et al., In situ measurements and Thermo-mechanical additive manufacturing[J], Add. Manufact. 27 (2019) 290–304.
simulation of Ti-6Al-4V laser solid forming processes[J], Int. J. Mech. Sci. 153 [33] N. Gardan, A. Schneider, Topological optimization of internal patterns and support
(2019) 119–130. in additive manufacturing[J], J. Manuf. Syst. 37 (2015) 417–425.
[4] Q. Wang, J. Shi, L. Zhang, et al., Impacts of laser cladding residual stress and material [34] M. Cervera, C. Agelet de Saracibar, M. Chiumenti, COMET: Coupled mechanical and
properties of functionally graded layers on titanium alloy sheet[J], Add. Manufact. 35 thermal analysis, Data Input Manual, Version 5.0, Technical Report IT-308, 2002
(2020) 101303. http://www.cimne.com.
[5] J.M. Shi, N. Ma, L.X. Zhang, et al., Residual stress and fracture strength of brazed joint [35] M. Chiumenti, X. Lin, M. Cervera, et al., Numerical simulation and experimental cal-
of ceramic and titanium alloy with the aid of laser deposited functionally graded ibration of additive manufacturing by blown powder technology. Part I: thermal
material layers[J], J. Manuf. Process. 34 (2018) 495–502. analysis[J], Rapid Prototyp. J. 23 (2) (2017) 448–463.

13
X. Lu, M. Chiumenti, M. Cervera et al. Materials and Design 202 (2021) 109525

[36] M. Chiumenti, M. Cervera, A. Salmi, et al., Finite element modeling of multi-pass [40] T. Kawaoka, Introduction of the personal pre-post processor GiD[J], Proceed. Design
welding and shaped metal deposition processes[J], Comput. Methods Appl. Mech. Systems Conf. 12 (2002) 151–153.
Eng. 199 (37–40) (2010) 2343–2359. [41] V. Fallah, M. Alimardani, S.F. Corbin, et al., Temporal development of melt-pool mor-
[37] M. Chiumenti, M. Cervera, N. Dialami, et al., Numerical modeling of the electron phology and clad geometry in laser powder deposition[J], Comput. Mater. Sci. 50 (7)
beam welding and its experimental validation[J], Finite Elements Analysis Design (2011) 2124–2134.
121 (2016) 118–133. [42] S. Afazov, W.A.D. Denmark, B.L. Toralles, et al., Distortion prediction and compensa-
[38] Y. Ueda, H. Murakawa, N. Ma, Welding Deformation and Residual Stress Prevention, tion in selective laser melting[J], Add. Manufact. 17 (2017) 15–22.
Elsevier, 2012.
[39] Erik R. Denlinger, P. Michaleris, Effect of stress relaxation on distortion in additive
manufacturing process modeling[J], Add. Manufact. 12 (2016) 51–59.

14