micromachines
Article
Effect of Pore Shape on Mechanical Properties of Porous Shape
Memory Alloy
Bingfei Liu 1, * and Yaxuan Pan 2






Citation: Liu, B.; Pan, Y. Effect of
Pore Shape on Mechanical Properties
of Porous Shape Memory Alloy.
Micromachines 2022, 13, 566. https://
doi.org/10.3390/mi13040566
Academic Editor: Aiqun Liu
Received: 7 January 2022
Accepted: 23 February 2022
Published: 31 March 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Micromachines 2022, 13, 566. https://doi.org/10.3390/mi13040566
1 Science and Technology Innovation Research Institute, Civil Aviation University of China,
Tianjin 300300, China
2 Aviation Engineering Institute, Civil Aviation University of China, Tianjin 300300, China;
yaxuanpan2022@126.com
* Correspondence: bfliu@cauc.edu.cn; Tel.: +86-133-2336-7291
Abstract: Porous shape memory alloys (SMAs) have been widely used in the aerospace, military,
medical, and health fields due to its unique mechanical properties such as superelasticity, biocom-
patibility, and shape memory effect. In this work, the pore shape was considered in the constitutive
model of the porous SMAs by respectively introducing the parameter of aspect ratio and for different
pore shapes including oblate, sphere, and prolate shapes, so the expression of Young’s modulus for
the porous SMA can be derived. Then, the constitutive model for such a porous shape memory alloy
was established. When the porosity was zero, the model can be degenerated to the dense case. The
stress–strain curves for the porous SMA with a porosity of 13% with different aspect ratio are then
given. Numerical results showed good agreement with the published experimental data that proved
the validation of the model. Based on the proven constitutive model, the properties of porous SMA
with different porosity and pore shapes are discussed. The results showed that the pore shapes and
the porosities had a big effect on the stress–strain curves for the porous shape memory, while with
the increasing porosities, the Young’s modulus and the hysteresis both decreased. With the same
porosities, the Young’s modulus and hysteresis loop of SMA with round pores were the largest, while
the Young’s modulus and hysteresis loop were the smallest when r = 0.1, and they were greater
when r = 0.75 than when r = 10. It can be seen that the closer to the circle, the better the performance
of the material.
Keywords: shape memory alloy; aspect ratio; porosity; constitutive model
1. Introduction
Porous shape memory alloys have been widely used in various fields including
the aerospace, military, medical, and health fields [1–4] due to their unique mechanical
properties including superelasticity, shape memory effect, biocompatibility, low density,
high porosity, and high permeability [5–7].
The mechanical properties of porous SMA have been given much attention [8]. Re-
garding the experiments and simulation work on the porous SMA, the influence of pores
on the mechanical properties such as Young’s modulus and strength of the porous SMA has
been studied in recent years [9]. Xu et al., prepared SMA with different porosities by the mi-
crowave sintering method and studied the effects of porosity on microstructure, hardness,
compressive strength, bending strength, elastic modulus, phase transition temperature, and
superelasticity of porous SMA [10]. The results showed that with the increase in porosity,
the compressive strength, elastic modulus, bending strength, and superelasticity of porous
SMA decreased. Zhao et al., manufactured SMA with constant porosity and pore diameter
and found that its damping performance could be adjusted with the change in porosity
and pore diameters [11]. Zhang et al., successfully prepared porous SMA by combining
one-step spark plasma sintering technology with space retainer technology [12]. It was
https://www.mdpi.com/journal/micromachines
Micromachines 2022, 13, 566 2 of 12

found that the superelastic recovery strain ratio of porous SMA could be increased to more
than 90%, but with the further increase in training periods, the curve presents an obvious
stepped stress platform, indicating the collapse of pores. In addition, with the increase in
porosities and pore sizes, the elastic modulus and compressive strength of porous SMA
decrease. Gur et al., compared the simulation results of nano-porous NiTi with different
pore structures with those of dense NiTi alloys. The simulation results showed that the
martensite phase fraction and phase transition temperature increased significantly with
the increase in porosities; moreover, the stress–strain response changed significantly, and
residual strain and hysteretic energy dissipation capacity increased significantly with the in-
crease in porosity [13]. Kim et al. studied the effect of porosity on the mechanical properties
of the porous stent by the compression test [14]. Galimzyanov et al., used non-equilibrium
molecular dynamic simulation methods to study the influence of the degree of porosity of
porous amorphous titanium nickel ester on mechanical properties under uniaxial tension,
uniaxial compression, and uniform shear loads [15].
Regarding the theoretical work of the porous SMA, Zhao et al. established a macro-
scopic compression behavior model of porous SMA and predicted the effective elastic
and superelastic behavior of porous shape memory alloys based on the assumption of the
stress–strain curve by using the Eshelby heterogeneous inclusion method [16]. In addition,
they studied two types of pore connectivity and compared the stress–strain curves of
porous shape memory alloys with spherical pores and elliptical pores [17]. Based on the
micromechanical method and thermodynamic theory, a constitutive model of macroscopic
mechanical behavior of porous shape memory alloys was established [18]. According to
the phase transition function of dense SMA, considering the hydrostatic stress of porous
SMA, the phase transition function of porous SMA is given. Olsen et al., proposed a new
constitutive model of porous SMA based on the Gurson–Tvergaard–Needleman formula.
The main results showed that the stress of phase transformation and plastic yield will
decrease, even if the pores are relatively small [19]. In addition, the existence of micropores
leads to a reduction in the area of the stress–strain hysteresis curve, thus affecting the energy
dissipation in hyperelastic cycle. Considering Gibbs free energy, Xu et al. established a phe-
nomenological constitutive model that can be used to predict the mechanical behavior of
FGP-SMA by using thermodynamic theory and a new transformation function considering
the influence of hydrostatic stress was proposed [3]. Zheng et al., studied the effect of a
surface modified porous titanium implant and different porosity and aperture [20].
However, most of the above studies are spherical porous SMA. There is no detailed
theory about the effect of pore shape on the mechanical behavior of porous SMA, especially
in the case of different pore ratios. Although the above practical and theoretical research
work on porous SMA have been carried out, at present, no scholars have considered the
influence of pore shape on the properties of porous SMA, and the constitutive relationship
considering pore shape has not been established. Qiu found that the influence of pore
shape on metal materials is very important and demonstrated the great influence of pore
shape on metal materials [21]. Therefore, it is necessary to study the influence of pore shape
on the properties of SMA materials. In light of this, under the assumption that all pore
shapes in the whole material are consistent and evenly distributed in the material, not only
is the influence of porosity on modulus of porous SMA discussed, but also the influence of
pore shape on the modulus and the stress–strain relationship of porous SMA is established
in this work.

2. Theoretical Model of Porous SMA with Different Pore Shapes

2.1. Modulus
For porous SMA, it is regarded as a uniform porous material with uniform pore shape
and uniform orientation, as shown in Figure 1a, where the pores are evenly distributed in
space, as shown in Figure 1b. In light of this, when analyzing its mechanical properties, the
representative volume unit (RVE), as shown in Figure 1c, can be selected, where the center
of the circle is defined as O, the horizontal axis is the x axis, the vertical direction is the z
 2.1. Modulus
For porous SMA, it is regarded as a uniform porous material with uniform pore
shape and uniform orientation, as shown in Figure 1a, where the pores are evenly distrib-
uted in space, as shown in Figure 1b. In light of this, when analyzing its mechanical prop-
Micromachines 2022, 13, 566
erties, the representative volume unit (RVE), as shown in Figure 1c, can be selected, where
3 of 12
the center of the circle is defined as O , the horizontal axis is the x axis, the vertical di-
rection is the z axis, and Lx , Lz represent the diameters of the x axis and z axis of
axis,
the and Lrespectively,
ellipse, x , Lz represent then r = Lx / Lzof, called
the diameters the x axis
the and z axis
aspect of the
ratio, canellipse, respectively,
be defined to de-
then r = L x /Lz , called the aspect ratio, can be defined to describe the change in pores,
scribe the Figure
from the change1d in oblate
pores,shape
from (Lthe >
Figure
Lz ) to1d oblate
the Figureshape ( Lx >shape
1e sphere
Lz ) to(Lthe=Figure
Lz ) to 1e
the
x x
Figureshape
sphere ( Lx =
1f prolate Lz ) to
shape (Lthe Lz ). 1f prolate shape ( Lx < Lz ).
x < Figure

Schematicdiagram
Figure1.1.Schematic
Figure diagramofofthe
thepore
poreshape.
shape.

Theporous
The porousSMA SMAisisassumed
assumedtotobe bethe
thecomposite
compositeofofthe
theSMA SMAmatrix
matrixand
andpore
poreinclu-
inclu-
sions, which can be shown in Figure 1. It
sions, which can be shown in Figure 1. It is assumed is assumed that the matrix and inclusion
that the matrix and inclusion are are
linearly isotropic, and the elastic modulus is C0ijkl 0 and C1 , respectively. The subscript 1
1
linearly isotropic, and the elastic modulus is Cijkl and Cijkl ijkl , respectively. The subscript
indicates the pore inclusion phase, and the subscript 0 indicates the SMA matrix phase. In
1order
indicates the porethe
to facilitate inclusion phase,
analysis, and the composed
the material subscript 0only
indicates
of thethe SMAwith
matrix matrix
thephase.
elastic
Inmodulus
order toCfacilitate the analysis,asthe
0 was introduced thematerial
reference.composed only
Therefore, the of the matrix
equivalent with the
applied elastic
stress σij
ijkl
0
modulus C
and the strain was introduced as the reference. Therefore,
0 the equivalent
ijkl of composite ε ij and reference material ε are expressed by
ij
applied stress
σ ij and the strain of composite ε ij and reference material ε ij0 are expressed by
σij = Cijkl ε kl
0 ε0
(1)
σij = Cijkl kl

where σij and ε0ij represent the average stress and strain of the matrix, respectively.
When the inclusion with volume fraction f v is inserted into the reference material of
the pure matrix, the average stress of matrix will have a disturbance of e σij , resulting in the
strain disturbance of e 0
ε ij on the basis of ε ij , which can be expressed by

0
σij + e
σij = Cijkl (ε0kl + eε kl ) (2)
Micromachines 2022, 13, 566 4 of 12

For the average stress of pore inclusions, because the specific orientation of pores will
pt pt
produce a certain amount of stress–strain disturbance σij and ε ij , it can be obtained by
Eshelby’s equivalence principle
pt
σij1 = σij + e
σij + σij
1 ( ε0 + e pt
= Cijkl kl ε kl + ε kl ) (3)
pt
= 0 ( ε0
Cijkl kl + eε kl + ε kl − ε∗kl )

where ε∗kl is the equivalent transformation strain of the pore inclusion.

pt
From Equations (2) and (3), σij can be expressed as

pt pt
0
σij = Cijkl (ε kl − ε∗kl ) (4)

The stress of thee matrix and pore inclusion shall be balanced with the external stress
σij , that is
pt
σij = − f v e
e σij (5)

In addition, e
σij also satisfies the following equation

0
σij = Cijkl
e ε kl
e (6)

where e
ε ij is given by
pt
ε ij = − f v (ε ij − ε∗ij )
e (7)

which has also been derived by Takao et al. [22]. Then, it could be obtained that
pt
ε ij + f v ε ij = ε0ij + f v ε∗ij
ε ij = ε0ij + e (8)

Which is substituted into Equation (1) to obtain the following relation

Cijkl (ε0kl + f v ε∗kl ) = Cijkl

0 0
ε kl (9)

Since the matrix and inclusion are isotropic, their elastic modulus can be expressed by
the bulk modulus, shear modulus, and Kronecker delta δij , as follows

0 = κ δ δ + µ ( δ δ + δ δ − 2δ δ /3)
Cijkl 0 ij kl 0 ik jl il jk ij kl
(10)
1 = κ δ δ + µ ( δ δ + δ δ − 2δ δ /3)
Cijkl 1 ij kl 1 ik jl il jk ij kl

Then, combine Equation (7) to obtain

ε∗11 = a1 (ε011 + e
ε 11 ) − a2 (ε022 + e
ε 22 + ε033 + e
 
ε 33 ) /a
ε∗22 = 2a3 (ε011 + eε 11 ) + ( a4 + a5 a)(ε022 + eε 22 ) + ( a4 − a5 a)(ε033 + e
 
ε 33 ) /2a
ε∗22 = 2a3 (ε011 + eε 11 ) + ( a4 − a5 a)(ε022 + eε 22 ) + ( a4 + a5 a)(ε033 + e
 
ε 33 ) /2a
ε012 +e (11)
ε∗12 = − 2S
ε 12
1212 + µ0 / ( µ1 − µ0 )

ε023 +e
ε∗23 = − 2S
ε 23
2323 + µ0 / ( µ1 − µ0 )

ε013 +e
ε 13
ε∗13 = − 2S
1313 + µ0 / ( µ1 − µ0 )
Micromachines 2022, 13, 566 5 of 12

where components of Eshelby’s Sijkl are

n h i o
1 3r2 −1 2
S1111 = 2(1−ν0 )
1 − 2ν0 + − 1 − 2ν0 + r23r−1 g
r 2 −1
h i
r2
S2222 = S3333 3
= 8(1− ν0 ) r2 −1
+ 4(1−g ν ) 1 − 2ν0 − 4(r29−1)
0
n 2 h i o
r
S2233 = S3322 1
= 4(1−ν ) 2r2 −1 − 1 − 2ν0 + 4(r23−1) g
0
2
n 2 o
S2211 = S3311 = − 2(1−ν ) r2r−1 + 4(1−g ν ) r23r−1 − (1 − 2ν0 )
1
(12)
0 0
n h i h io
S1122 = S1133 = 2(1−ν ) g 1 − 2ν0 + 2(r23−1) ) − 1 − 2ν0 + r2 1−1
1
0
n 2 h io
S2323 = S3232 = 4(1−ν ) 2(r2r −1) + g 1 − 2ν0 − 4(r23−1)
1
0

3(r 2 +1)
n 2 1
h io
g
S1212 = S1313 = 4(1−ν ) 1 − 2ν0 − rr2 +
1
−1
− 2 1 − 2ν0 − 2
r −1
0

where v0 is the Poisson’s ratio of the SMA matrix, and g corresponding to r is given by
 
1

r 2
3 r (r − 1) − arccoshr
2

 r<1
(r 2 −1) 2





2

g= r=1 (13)


 3
  
 1
 r 3 arccosr − r (1 − r2 ) 2 r>1


(1−r 2 ) 2

The constants a, a1 , a2, a3 , a4 , a5 , depending on κ1 , κ0 , µ1 , µ0 , which are the bulk modulus

and shear modulus of pore inclusion and matrix, respectively,

a1 = 6(κ1 − κ0 )(µ1 − µ0 )(S2222 + S2233 − 1) − 2(κ0 µ1 − κ1 µ0 ) + 6κ1 (µ1 − µ0 )

a2 = 6(κ1 − κ0 )(µ1 − µ0 )S1133 + 2(κ0 µ1 − κ1 µ0 )
a3 = −6(κ1 − κ0 )(µ1 − µ0 )S3311 − 2(κ0 µ1 − κ1 µ0 )
a4 = 6(κ1 − κ0 )(µ1 − µ0 )(S1111 − 1) + 2(κ0 µ1 − κ1 µ0 ) + 6µ1 (κ1 − κ0 )
(14)
a5 = 1/[S3322 − S3333 + 1 − µ1 /(µ1 − µ0 )]
a = 6(κ1 − κ0 )(µ1 − µ0 )[2S1133 S3311 − (S1111 − 1)(S3322 + S3333 − 1)]
+2(κ0 µ1 − κ1 µ0 )[2(S1133 + S3311 ) + (S1111 − S3322 − S3333 )]
−6(κ1 − κ0 )(µ1 − µ0 )(S1111 − 1) − 6µ1 (κ1 − κ0 )(S2222 + S2233 − 1) − 6κ1 µ1

When inclusions are uniformly distributed in three-dimensional space, it is assumed

that the composite, as a whole, is macroscopically isotropic. The stress or strain can be
decomposed into hydrostatic and deviatoric parts, resulting in the effective bulk modulus
and shear modulus, respectively.
From Equation (7), the hydrostatic and deviatoric parts can be found by

ε kk = (1/p1 − 1)ε0kk
e
(15)
ε 12 = (1/q1 − 1)ε012
e

Then
ε∗kk = pε0kk
(16)
ε∗12 = qε012
Micromachines 2022, 13, 566 6 of 12

where
p = p2 /p1
q = q2 /q1  
2(S1122 + S2222 + S2233 − 1)( a3 + a4 )
p1 = 1 + f v  /3a
+(S1111 + 2S2211 − 1)( a1 − 2a2 )

p2 = [ a1 − 2( a2 − a3 − a4 )]/3a
2S1212 −1 2S2323 −1
 
2 1
5 2S1212 +µ0 /(µ1 −µ0 ) + 3 2S2323 +µ0 /(µ1 −µ0 )

 



 


  (17)
(S1122 − S2233 )(2a3 − a4 + a5 a)

 

 
q1 = 1 − f v  



1 
− 15a  +2(S1111 − S2211 − 1)( a1 + a2 )  
 


   

+(S1122 − S2222 + 1)(2a3 − a4 − a5 a)

 

 

q2 = − 25 2S 1
− 1 1
3 2S2323 +µ0 /(µ1 −µ0 )
1212 + µ0 / ( µ1 − µ0 )
1
+ 15a[2(a
1 + a2 − a3 )+ a4 + a5 a ]

Combined with Equation (12), the average disturbance stress in the matrix is obtained by

σkk = (1/p1 − 1)σkk

e
(18)
σ12 = (1/q1 − 1)σ12
e

Then, the effective bulk modulus κ and shear modulus µ of the porous composite can
be obtained from Equation (9)
κ 1
κ 0 = 1+ f v p
(19)
µ 1
µ0 = 1+ f v q

κ0 and µ0 of the SMA matrix can be determined, respectively, by

0
ESMA
κ0 = (20)
3(1 − 2ν0 )
0
ESMA
µ0 = (21)
2(1 + ν0 )
0
where v0 is the Poisson’s ratio of the SMA matrix and ESMA is the Young’s modulus of the
SMA matrix. For the porous SMA with different specified pore shape, its Young’s modulus
can be obtained from the following expression,

p 9κµ
ESMA = (22)
3κ + µ

2.2. Constitutive Model

For the uniaxial loading case, the one-dimensional unified constitutive model de-
graded from three-dimensional model of Lagoudas is applied here [23],
. . . .
ε = σij : M p + Λij ξ + α p ∆ T (23)

−1 0 .
3 p
2H σij σ ,ξ > 0

Λij = . (24)
 Hp −1 t
εt


ε ij ,ξ < 0
Micromachines 2022, 13, 566 7 of 12

p −1
where H p is the maximum phase transformation strain and M p = ( ESMA ) is the elastic
p
flexibility tensor; and ESMA is the Young’s modulus of porous SMA, which is expressed as
follows
p p p p
ESMA = E A + ξ p ( E M − E A ) (25)
where ξ p is the martensitic volume fraction of porous SMA, which is related to the critical
stresses of porous SMA. 
p

 0 σe ≤ σs
 σ −σ p

p p
e
ξ p = σ p −σsp σs ≤ σe ≤ σ f
s
(26)
 f
 p
 1 σ f ≤ σe


where σe is the effective stress, which is expressed as follows:

p
σe = 3J2 (27)

where J2 represents the invariant of the second deviation of stress, which is expressed by

1 0 0
J2 = σ :σ (28)
2
1
σ0 = σij − σkk (29)
3
For the uniaxial loading,
 
σ11 0 0
σij =  0 0 0  (30)
0 0 0

α p is the thermal expansion coefficient, which can be expressed as follows:

p p p
α p = α A + ξ p (α M − α A )
p p
where α M and α A represent the thermal expansion coefficients of martensite and austenite,
. p p
respectively. ∆ T is the increment of temperature difference, σs , σ f (the lower corner
marks s and f represent the starting point and the ending point of phase transformation,
respectively.)
The critical stress of porous SMA is affected by porosity, f v , can be assumed as a linear
relationship with the critical stress of solid SMA, and σsD or σ Df , which could be assumed
and expressed as follows
p
σs = (1 − 2 f v )σsD (31)
p
σ f = (1 − 2 f v ) σ D
f (32)

3. Numerical Analysis Results

3.1. Young’s Modulus of SMA with Different Pore Shapes and Porosities
In order to analyze the influence of pores including the volume fraction and aspect
ratio on the mechanical properties of SMA, the mechanical parameters of solid SMA in the
references are quoted here [17], as shown in Table 1. It is assumed that the temperature and
the Poisson’s ratio of the material remains constant (i.e., νM = νA = ν = 0.3).

Table 1. Mechanical parameters of solid SMA.

EsM EsA s
σMs s
σAs s
σMf s
σAf H v
31 75 400 600 920 140 0.02 0.3
 Table 1. Mechanical parameters of solid SMA.
s
EM 2022, 13, 566
Micromachines EAs σ Ms
s
σ As
s
σ Mf
s
σ Afs H 8vof 12
31 75 400 600 920 140 0.02 0.3

The
The influence
influence ofof pore
pore aspect
aspect ratio
ratio on
on the
the Young’s modulus of
Young’s modulus of porous
porous SMA
SMA under
under
different
different porosity is shown in Figure 2. Figure 2a–e show the influence of pore aspect ratio
porosity is shown in Figure 2. Figure 2a–e show the influence of pore aspect ratio
on the Young’s
on the Young’s modulus
modulus of of porous
porous SMA
SMA at at different
different phase
phase transformation
transformation stages.
stages. It can
It can
be seen from the figure that when the aspect ratio was between 0 and 0.25, the Young’s
modulus changed greatly with the increase in aspect ratio, while when the aspect ratio
exceeded 0.25, the aspect ratio had little effect on the Young’s modulus and the modulus
was the
the largest
largestwhen
whenr = r= 0.75
0.75. . For
For the same
the same aspect
aspect ratio,ratio, the Young’s
the Young’s modulusmodulus de-
decreased
creased
with thewith the increase
increase in porosity.
in porosity.

(a) 90 (b)
70
80

70 60
f=0, f=0, ξ=0.2
60 f=13% 50 f=13%, ξ=0.2
f=25% f=25%, ξ=0.2

E/Gpa
E/Gpa

50 40
40
30
30
20
20
10
10

0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
r r

(c) 60 (d) 45 (e) 40

40 35
50
f=0, ξ=0.5 35 f=0, ξ =0.8 30
f=0, ξ=1
40 f=13%, ξ=0.5 30 f=13%, ξ=0.8
25 f=13%, ξ=1
f=25%, ξ=0.5 f=25%, ξ=0.8
E/Gpa
E/Gpa

E/Gpa

25 f=25%, ξ=1
30 20
20
15
20 15
10
10
10
5
5
0 0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 r
r r

Figure 2. Relationship between Young’s modulus in different phase states and the aspect ratio of
porous SMA
Figure 2. at porosities
Relationship of 0%, Young’s
between 13%, andmodulus
25%: (a) pure austenitic
in different phase,
phase (b)and
states mixed
the martensite
aspect ratioand
of
porous SMA
austenite at at
phase porosities
ξ = 20%,of(c)0%, 13%,martensite
mixed and 25%: and
(a) pure austenitic
austenite phasephase, (b) mixed
at ξ = 50%, martensite
(d) mixed and
martensite
austenite phase
and austenite at ξat
phase =20%
ξ = ,80%,
(c) mixed martensite
and (e) phase. phase at ξ =50% , (d) mixed marten-
and austenite
pure martensite
site and austenite phase at ξ =80% , and (e) pure martensite phase.
Then, similar to Figure 2, when the aspect ratio was constant, the relationship between
the Young’s modulus
Then, similar and porosity
to Figure 2, when in the
different
aspectphase
ratio states were plotted
was constant, in Figure 3. For
the relationship be-
porous SMA in any phase transformation state, the Young’s modulus
tween the Young’s modulus and porosity in different phase states were plotted in Figure of porous SMA with
round
3. pores (rSMA
For porous = 1) inwasany always
phasethe largest, followed
transformation state,by = 0.75, r modulus
ther Young’s = 10, r =of0.1, and
porous
rSMA with round pores ( r = 1 ) was always the largest, followed by r = 0.75 , r = 10 ,
= 0.01. When r = 0.1 and r = 10, the Young’s modulus was close. It can be seen that
r =closer
the to the
0.1 , and r=round
0.01hole,
. When = 0.1 the
the rgreater r = 10modulus
andYoung’s of themodulus
, the Young’s material.wasIn addition,
close. It
when
can betheseen porosity
that thewascloserbetween
to the 0round
and 0.1,
hole, the Young’s
the greatermodulus
the Young’s decreased
modulus rapidly
of thewith
ma-
the increase in porosity f v , while when f v exceeded 0.1, the
terial. In addition, when the porosity was between 0 and 0.1, the Young’s modulus de-effect of porosity on Young’s
r =when0.1 and rexceeded
modulusrapidly
creased decreased.
with theFurthermore, for the case, of
increase in porosity f v while f = 10, which actually
0.1, the effect
had the same shape in appearance but were situated in differentv orientations in uniaxial
of porosity ontest,
compression Young’s
it can modulus
be observed decreased.
that theFurthermore,
Young’s modulus for theof case
porous r = 0.1
of SMA for and
the
r = 10 , which actually had
former was smaller than the latter. the same shape in appearance but were situated in different
orientations
The change in uniaxial
in Young’scompression
modulustest, it can beSMA
of porous observed
underthat the Young’s
different loads ismodulus
shown in of
porous
Figure 4.SMAIt canfor
bethe
seenformer was smaller
from Figure 4 that inthanthethe latter.
phase transition stage, the Young’s modulus
decreased with the increase in load, but remained unchanged in the complete elastic stage.
For different pore shapes, the Young’s modulus was the largest when r = 1 was the largest,
followed by r = 0.75, r = 0.1, r = 10, and the smallest when r = 0.01.
 (a)80 (b) 70

70 60

60 r r
50
Micromachines 2022,
Micromachines 2022, 13,
13, 566
x FOR PEER REVIEW
r r 109 of 13
of 12
50
r r

E/Gpa
40

E/Gpa
40 r r
30 r
r
30

(a)2080 (b)2070

70 10
10 60

60
0 r=1, ξ=0 0 r=1, ξ=0.2
50
0.0 0.2 0.4 0.6 0.8 ξ=0.2 1.0
0.0 0.2 0.4 0.6 0.8 ξ=0 1.0
r=0.75, f r=0.75,
50 f r=10, ξ=0.2
r=10, ξ=0

E/Gpa
40
E/Gpa

40 r=0.1, ξ=0 r=0.1, ξ=0.2

(c) 60 （d） 30 (e) 35 r=0.01, ξ=0.2
r=0.01, ξ=0
30
40
50 20 30
20
r r , 
10 25
40 10 r=0.75, =0.5 30 r 
r r 
E/Gpa

E/Gpa
0 20

E/Gpa
0
30 0.0 0.2 r
0.4 0.6 0.8 1.0 r 0.0 0.2 0.4 0.6 0.8 
1.0
f
f
r 20
r 15 
20
10
(c) 60 （ d） (e) 35
10
10 5
40
50 30
0 r=1, ξ=0.5 0 r=1, ξ=0.8 0 α=1, ξ=1
0.0 0.2 0.4 0.6 0.8 25
0.0 0.2 0.4 0.6 0.8 1.0
40 f r=0.75, ξ=0.51.0 0.0
30 0.2 0.4 f 0.6 0.8 ξ=0.8 1.0
r=0.75, f α=0.75, ξ=1
r=10, ξ=0.5 r=10, ξ=0.8 α=10, ξ=1
E/Gpa

E/Gpa
20
E/Gpa

30 r=0.1, ξ=0.5 r=0.1, ξ=0.8 α=0.1, ξ=1

r=0.01, ξ=0.5 20 15 α=0.01, ξ=1
r=0.01, ξ=0.8
20
Figure 3. Relationship between Young’s modulus in different phase states and the porosity of po-
rous SMA with the aspect ratio of 0.01, 0.1, 0.75, 1, and 10: (a)10pure austenitic phase, (b) mixed mar-
10
10 tensite and austenite phase at  =20% , (c) mixed martensite5 and austenite phase at  =50% , (d)
0 mixed martensite0 and austenite phase at  =80% , and (e) pure0 martensite phase.
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
f f f

The change in Young’s modulus of porous SMA under different loads is shown in
Figure 3. Relationship between Young’s modulus in different phase states and the porosity of porous
Figure
Figure 4. It can be seen
3. Relationship from Figure
between Young’s 4 that in the
modulus in phase transition
different stage,
phase states andthe
theYoung’s
porositymod-
of po-
SMA with the aspect ratio of 0.01, 0.1, 0.75, 1, and 10: (a) pure austenitic phase, (b) mixed martensite
ulus
rous decreased
SMA with the with the ratio
aspect increase in load,
of 0.01, but1,remained
0.1, 0.75, and 10: (a)unchanged in the
pure austenitic complete
phase, elastic
(b) mixed mar-
and austenite phase at ξ = 20%, (c) mixed martensite and austenite phase at ξ = 50%, (d) mixed
stage.
tensiteFor
anddifferent
austenite pore
phaseshapes, the, Young’s
at ξ =20% (c) mixed modulus
martensitewas
andthe whenat rξ 
largestphase
austenite was
1 , (d)
=50%
martensite and austenite phase at ξ = 80%, and (e) pure martensite phase.
the largest, followed by r  0.75 , r  0.1 , r  10 , and the smallest when r  0.01 .
mixed martensite and austenite phase at ξ =80% , and (e) pure martensite phase.

(a) 80 The change

(b) in Young’s modulus of porous SMA (c) 50 under different loads is shown in

Figure 4. It can60be seen from Figure 4 that

f=13% r=1
in the
f=13% r=0.01
phase f=25% r=1 mod-
transition stage, the Young’s
f=25% r=0.01
70 f=13% r=0.1 40
ulus
f=0 decreased 50 with the increase in load, but remained unchanged in the complete
f=13% r=0.75
elastic
f=25% r=0.1
f=25% r=0.75

60
stage. For different
40
f=13% r=10
pore shapes, the Young’s modulus 30
was the largest whenf=25%r =r=101 was
the largest, followed by r = 0.75 , r = 0.1 , r = 10 , and the smallest when r = 0.01 .
E/GPa

E/GPa

E/GPa

30
50 20
(a) 80 (b) 20 (c) 50
40 60 f=13% r=1 10 f=25% r=1
10 f=13% r=0.01 f=25% r=0.01
70 40 f=25% r=0.1
f=0 50 f=13% r=0.1
30 0 f=13% r=0.75 0 f=25% r=0.75
400 500 600 700 800 900 250 300 350 400 450 500 550f=13%
600 r=10
650 700 200 250 300 350 400 r=10 450
f=25%
60 30
/MPa 40 Pa /MPa
E/GPa

E/GPa
E/GPa

30
50
Figure 4. Comparison of the relationship between the Young’s20modulus of five types of porous SMA
40
Figure 4. Comparison
with effective 20
stress atofdifferent
the relationship between
porosities of (a) 0,the
(b)Young’s
13%, andmodulus
10
(c) 25%.of five types of porous SMA
with effective stress
10 at different porosities of (a) 0, (b) 13%, and (c) 25%.
30
3.2. Stress–Strain Relationship of SMA with Different Pore Shapes
0
and Porosities
400 500 600 700 800 900 0 200 250 300 350 400 450
σ/MPa 3.2. Stress–Strain Relationship
In order to250verify of SMA
the400correctness
300 350 450 with
500 550 Different
of600the Pore Shapes
650theoretical
700 model and Porosities
in this σ/MPa
paper, it is necessary
σ/ΜPa
to compare
In order to verify the correctness of the theoretical model in this paper, it is
the theoretical results with the experimental results. However, it is difficult
necessary to
prepare porous SMA with different pore shapes, and there have been
to compare the theoretical results with the experimental results. However, it is difficult to few experimental
reports
Figure 4.on the mechanical behavior ofbetween
porousthe SMA withmodulus
differentofpore shapes. Due toSMAthe
prepare porous SMAofwith
Comparison different pore shapes,
the relationship and there
Young’s have been few experimental
five types of porous
lack of experimental data, it is difficult to compare the theoretical curve of stress and strain
reports on thestress
with effective mechanical behavior
at different of porous
porosities of (a) 0, SMA with
(b) 13%, anddifferent
(c) 25%. pore shapes. Due to the
of porous SMA with the different pore shapes in the experimental results. Therefore, in
lack of experimental data, it is difficult to compare the theoretical curve of stress and strain
order
3.2. to verify the
Stress–Strain correctnessofof
Relationship thiswith
SMA model, the constitutive
Different Shapesmodel
Poreexperimental of porous SMA with
and Porosities
of porous SMA with the different pore shapes in the results. Therefore, in
spherical pores in this paper was compared with the experimental data of porous SMA
In order to verify the correctness of the theoretical model in this paper, it is necessary
with spherical pores [17], as shown in Figure 5. The solid line represents the published
to compare the theoretical results with the experimental results. However, it is difficult to
experimental result, and the dotted line represents the simulation result in this paper.
prepare porous SMA with different pore shapes, and there have been few experimental
Obviously, the comparison results of the two cases showed a certain degree of consistency
reports on the mechanical behavior of porous SMA with different pore shapes. Due to the
between the theory and the experiment, which verifies the correctness of the model in this
lack of experimental data, it is difficult to compare the theoretical curve of stress and strain
of porous SMA with the different pore shapes in the experimental results. Therefore, in
 viously, the comparison results of the two cases showed a certain degree of consistency
order tothe
between verify the correctness
theory of this model,
and the experiment, the constitutive
which verifies the model of porous
correctness SMA
of the within this
model
spherical pores in this paper was compared with the experimental data of porous
paper. In contrast, the warping angle at the top of the hysteretic curve depends on the SMA
with spherical
selected pores
theoretical [17], as shown in Figure 5. The solid line represents the published
model.
Micromachines 2022, 13, 566 experimental result, and the dotted line represents the simulation result in this paper.
10 ofOb-
12
viously, the comparison results of the two cases showed a certain degree of consistency
(a) (b)which verifies the correctness of the model in this
between the theory and the experiment,
1200 Experimental result Experimental result
paper. In contrast, the warping angle at 1000
the top of the hysteretic curve depends on the
Theoretical result paper. In contrast, the warping angle at the top of Theoretical result
the hysteretic curve depends on the
1000
selected theoretical model.
selected theoretical model.
800

800(a) (b)
1200 Experimental result 600 Experimental result

σ
σ

1000
600 Theoretical result Theoretical result
1000
800400
400
800
600200
200

σ
σ

600
400 0
0
0 400 1 2 3 4 5 0 1 2 3 4 5
ε ε
200
200

0
Figure 5. Comparisons of the theoretical and
0 experimental results of (a) solid SMA and (b) porous
0 1 2SMA 3 4 5 13%. 0 1 2 3 4 5
ε with a porosity of ε

Figure 5. Comparisons of the theoretical and experimental results of (a) solid SMA and (b) porous
Next,
Figure the influence
5. Comparisons oftheoretical
of the various andporeexperimental
shapes onresults
the mechanical properties
of (a) solid SMA of porous
and (b) porous
SMA with a porosity of 13%.
SMA with a porosity of 13%.
SMA is further discussed. The stress–strain curves can be obtained when the porosity is
13% andNext,25%, respectively,
the influence as shown
of various pore shapesin Figure 6. It can be
on the mechanical seen from
properties FigureSMA
of porous 6 that the
Next, the influence of various pore shapes on the mechanical properties of porous
hysteresis curve was the largest when r = 1 , the smallest when r = 0.75 , and the maxi-
is further discussed. The stress–strain curves can be obtained when the porosity is 13% and
SMA is further discussed. inThe stress–strain curves can be obtained when the porosity is
25%,
mum respectively,
strain was the
13%theand
as shown
25%,when
largest
respectively,
Figure
when 6.
r
assmallest
=
It
shown in
0.1
can be seen
, but
Figure
from
there Figure
was 6 that
little the hysteresis
difference in thecurve
other three
was
cases. largest
Thus, the pore = 1, the
rshape had a great when
impactr = 6. It can
0.75,
on and
the
be
theseen from Figure
maximum
stress–strain strain6was
relationship
thatthe
the
of the po-
hysteresis curve was the largest when r = 1 , the smallest when r = 0.75 ,
largest when r = 0.1, but there was little difference in the other three cases. Thus, the poreand the maxi-
rous
mumSMA.strain was the largest when r = 0.1 , but there was little difference in the other three
shape had a great impact on the stress–strain relationship of the porous SMA.
cases. Thus, the pore shape had a great impact on the stress–strain relationship of the po-
rous SMA.
(a) (b) 500
800

700 400
(a) (b) 500
600800
700 400 300
500
σ/MPa

σ/MPa

600
400 r=1 r=1
500 r=0.75 300 200
σ/MPa

r=0.75
σ/MPa

300 r=10
400 r=1 r=10
r=0.1 r=1
200 r=0.75 200
100 r=0.75 r=0.1
300 r=10 r=10
100 r=0.1 r=0.1
200
100
0100 0
0 1 2 3 4 5 6 7 0 2 4 6 8 10
0 ε 0 ε
0 1 2 3 4 5 6 7 0 2 4 6 8 10
ε ε
Figure 6. Stress–strain curves of different aspect ratios with porosities of (a) 13% and (b) 25%.

FigureWhen
6. Stress–strain curvesisof
the pore shape different aspect
determined, ratios with porosities
the stress–strain curves ofof (a) 13%
SMA withand (b) 25%.
different
Figure 6. Stress–strain
porosities are shown in curves of different
Figure 7. It canaspect ratiosfrom
be seen with Figure
porosities of (a)with
7 that 13% and
the (b) 25%. in
increase
When
porosity, thethe poretransformation
phase shape is determined,
stresses ofthe
thestress–strain curves ofdecreased,
porous SMA gradually SMA withthe different
When
hysteretic the pore
curve shapemoved
gradually is determined,
down theshrank,
and stress–strain
and curves
the degreeof of
SMA withwas
change different
the
porosities are shown in Figure 7. It can be seen from Figure 7 that with the increase in
porosities are=shown in Figure 7. It can be seen from Figure 7 that with the increase in
largest
porosity, the rphase
when 0.1. transformation stresses of the porous SMA gradually decreased, the
porosity, the phase transformation stresses of the porous SMA gradually decreased, the
hysteretic
hystereticcurve
curve gradually moveddown
gradually moved downand and shrank,
shrank, andand
the the degree
degree of change
of change was the
was the
largest when r = 0.1
largest when r = 0.1 . .
Micromachines
Micromachines2022,
2022,13,
13,x566
FOR PEER REVIEW 12 11
ofof1312

(a) (b)1000
1000
solid SMA solid SMA
r=0.1 f=13% r=0.75 f=13%
800 r=0.1 f=25% 800 r=0.75 f=25%

600 600

σ/MPa
σ/MPa
400 400

200 200

0 0
0 2 4 6 8 10 0 1 2 3 4 5
ε ε

（c）
1000 (d)1000
solid SMA solid SMA
r=1 f=13% r=10 f=13%
800 r=1 f=25% 800 r=10 f=25%

600 600
σ/MPa

σ/MPa
400 400

200 200

0 0
0 1 2 3 4 5 0 1 2 3 4 5 6
ε ε

Figure 7. Stress–strain curves of different porosities with aspect ratios of (a) r = 0.1, (b) r = 0.75,
Figure 7. Stress–strain curves of different porosities with aspect ratios of (a) r = 0.1 , (b) r = 0.75 ,
(c) rr =
(c) = 11,, and (d) rr== 10.
and (d) 10 .
4. Conclusions
4. Conclusions
1. The Young’s modulus of porous SMA is related to porosity and pore shape. With
1. The theYoung’s
increasemodulus of porous
in porosity f v , the SMA is related
Young’s modulusto porosity
decreased; andwhen
pore shape.
f v wasWith the
between
increase in porosity
0 and 0.1, the influence f , the Young’s modulus decreased; when
v degree was the largest, and when f v was between f v
was between
0.1 and01,
it had
and 0.1,little impact ondegree
the influence the Young’s
was themodulus. When
largest, and the pore
when aspect
f v was ratio r0.1
between gradually
and 1,
it had little impact on the Young’s modulus. When the pore aspect ratio r graduallyr
increased in the range of 0–0.25, the Young’s modulus increased significantly. When
exceededin0.25
increased the and increased
range of 0–0.25, gradually, the modulus
the Young’s Young’s modulus
increasedfirst increased When
significantly. slowly
and then decreased slowly, reaching the maximum at about
r exceeded 0.25 and increased gradually, the Young’s modulus first increased r = 0.75.
2. slowly
The poreandshape and porosity
then decreased hadreaching
slowly, a great influence on theatstress–strain
the maximum about r = 0.75 relationship
.
of porous SMA, and the pore shape had an irregular influence on the stress–strain
2. The pore shape and porosity had a great influence on the stress–strain relationship
relationship. With the increase in porosity, the critical stresses decreased and the area
of porous SMA, and the pore shape had an irregular influence on the stress–strain
of the hysteretic curve of the stress–strain relationship decreased.
relationship. With the increase in porosity, the critical stresses decreased and the area
of the hysteretic curve of the stress–strain relationship decreased.
Author Contributions: Conceptualization, B.L.; methodology, Y.P.; validation, B.L. and Y.P.; data cu-
ration, B.L.;
Author writing—original
Contributions: draft preparation,
Conceptualization, L.B.; Y.P.; writing—review
methodology, and editing,
P.Y.; validation, B.L.;
L.B. andvisualization,
P.Y.; data
Y.P.; supervision,
curation, B.L.; project administration,
L.B.; writing—original draft preparation,B.L.;P.funding acquisition, B.L.
Y.; writing—review and All authors
editing, L.B.;have read
visuali-
and agreed
zation, P.Y.; to the published
supervision, version
L.B.; projectofadministration,
the manuscript.L.B.; funding acquisition, L.B. All authors
have read and agreed to the published version of the manuscript.
Funding: This research was funded by the National Natural Science Foundation of China
(No. 11502284),
Funding: Key deployment
This research was fundedprojects of the
by the Chinese
National Academy
Natural of Sciences
Science (KFZD-SW-435),
Foundation of China (No.and
the Fundamental
11502284), Research Funds
Key deployment forofthe
projects theCentral
Chinese Universities
Academy of of Sciences
China (3122020077).
(KFZD-SW-435), and the
Fundamental Research Funds for the Central Universities of China (3122020077).
Institutional Review Board Statement: “Not applicable” for studies not involving humans or animals.
Institutional Review Board Statement: “Not applicable” for studies not involving humans or ani-
Acknowledgments: The authors acknowledge the financial support of the National Natural Science
mals.
Foundation of China (No. 11502284), Key deployment projects of the Chinese Academy of Sciences
(KFZD-SW-435), the Fundamental
Acknowledgements: Research Funds
The authors acknowledge thefor the Central
financial Universities
support of ChinaNatural
of the National (3122020077)
Sci-
and Foundation
ence the general project of(No.
of China Tianjin Natural Key
11502284), Science Foundation:
deployment "Fatigue
projects Damage
of the Chineseand Life Prediction
Academy of Sci-
Analysis
ences in Jet Engine Chevron".
(KFZD-SW-435), the Fundamental Research Funds for the Central Universities of China
(3122020077) and the general project of Tianjin Natural Science Foundation: “Fatigue Damage and
Conflicts of Interest: The authors declare no conflict of interest.
Life Prediction Analysis of Deformed Teeth of Aeroengine”.
Conflicts of Interest: The authors declare no conflict of interest.
Micromachines 2022, 13, 566 12 of 12

References
1. Kaya, M.; Çakmak, Ö.; Gülenç, B.; Atlı, K.C. Thermomechanical cyclic stability of porous NiTi shape memory alloy. Mater. Res.
Bull. 2017, 95, 243–247. [CrossRef]
2. Xu, J.L.; Jin, X.F.; Luo, J.M.; Zhong, Z.C. Fabrication and properties of porous NiTi alloys by microwave sintering for biomedical
applications. Mater. Lett. 2014, 124, 110–112. [CrossRef]
3. Li, S.; Kim, Y.; Choi, M.; Nam, T. Highly porous Ni-free Ti-based scaffolds with large recoverable strain for biomedical applications.
Intermetallics 2020, 116, 106657. [CrossRef]
4. Abidi, I.H.; Khalid, F.A.; Farooq, M.U.; Hussain, M.A.; Maqbool, A. Tailoring the pore morphology of porous nitinol with suitable
mechanical properties for biomedical applications. Mater. Lett. 2015, 154, 17–20. [CrossRef]
5. Farber, E.; Zhu, J.N.; Popovich, A.; Popovich, V. A review of NiTi shape memory alloy as a smart material produced by additive
manufacturing. Mater. Today Proc. 2019, 30, 761–767. [CrossRef]
6. Liu, S.; Liu, J.; Wang, L.; Ma, R.L.W.; Zhong, Y.; Lu, W.; Zhang, L.C. Superelastic behavior of in-situ eutectic-reaction manufactured
high strength 3D porous NiTi-Nb scaffold. Scr. Mater. 2020, 181, 121–126. [CrossRef]
7. Zhao, M.; Qing, H.; Wang, Y.; Liang, J.; Zhao, M.; Geng, Y.; Liang, J.; Lu, B. Superelastic behaviors of additively manufactured
porous NiTi shape memory alloys designed with Menger sponge-like fractal structures. Mater. Des. 2021, 200, 109448. [CrossRef]
8. Peng, W.; Liu, K.; Shah, B.A.; Yuan, B.; Gao, Y.; Zhu, M. Enhanced internal friction and specific strength of porous TiNi shape
memory alloy composite by the synergistic effect of pore and Ti2Ni. J. Alloys Compd. 2020, 816, 152578. [CrossRef]
9. Khalifehzadeh, R.; Forouzan, S.; Arami, H.; Sadrnezhaad, S.K. Prediction of the effect of vacuum sintering conditions on porosity
and hardness of porous NiTi shape memory alloy using ANFIS. Comput. Mater. Sci. 2007, 40, 359–365. [CrossRef]
10. Xu, J.L.; Bao, L.Z.; Liu, A.H.; Jin, X.F.; Luo, J.M.; Zhong, Z.C.; Zheng, Y.F. Effect of pore sizes on the microstructure and properties
of the biomedical porous NiTi alloys prepared by microwave sintering. J. Alloys Compd. 2015, 645, 137–142. [CrossRef]
11. Zhao, M.; Shao, Y.; Zheng, W.; Luo, Y.; Qiao, J.; Wu, S.; Yan, Y.; Guo, W. Tailoring the damping and mechanical properties of
porous NiTi by a phase leaching process. J. Alloys Compd. 2021, 855, 157471. [CrossRef]
12. Zhang, L.; Zhang, Y.Q.; Jiang, Y.H.; Zhou, R. Mechanical behaviors of porous Ti with high porosity and large pore size prepared
by one-step spark plasma sintering technique. Vacuum 2015, 122, 187–194. [CrossRef]
13. Gur, S.; Frantziskonis, G.N.; Muralidharan, K. Atomistic simulation of shape memory effect (SME) and superelasticity (SE) in
nano-porous NiTi shape memory alloy (SMA). Comput. Mater. Sci. 2018, 152, 28–37. [CrossRef]
14. Kim, Y. Martensitic transformation behavior and mechanical properties of highly porous Ti-Ni-Mo scaffolds. J. Alloys Compd.
2020, 821, 153220. [CrossRef]
15. Galimzyanov, B.N.; Mokshin, A.V. Mechanical response of mesoporous amorphous NiTi alloy to external deformations. Int. J.
Solids Struct. 2021, 224, 111047. [CrossRef]
16. Zhao, Y.; Taya, M.; Kang, Y.; Kawasaki, A. Compression behavior of porous NiTi shape memory alloy. Acta Mater. 2005, 53,
337–343. [CrossRef]
17. Zhao, Y.; Taya, M. Analytical modeling for stress-strain curve of a porous NiTi. J. Appl. Mech. Trans. ASME 2007, 74, 291–297.
[CrossRef]
18. Zhu, Y.; Dui, G. A model considering hydrostatic stress of porous NiTi shape memory alloy. Acta Mech. Solida Sin. 2011, 24,
289–298. [CrossRef]
19. Olsen, J.S.; Zhang, Z.L. Effect of spherical micro-voids in shape memory alloys subjected to uniaxial loading. Int. J. Solids Struct.
2021, 49, 1947–1960. [CrossRef]
20. Zheng, J.; Chen, L.; Chen, D.; Shao, C.; Yi, M.; Zhang, B. Effects of pore size and porosity of surface-modified porous titanium
implants on bone tissue ingrowth. Trans. Nonferrous Met. Soc. China 2019, 29, 2534–2545. [CrossRef]
21. Qiu, Y.P.; Weng, G.J. A theory of plasticity for porous materials and particle-reinforced composites. J. Appl. Mech. Trans. ASME
1992, 59, 261–268. [CrossRef]
22. Takao, Y.; Taya, M.; Chou, T.W. Effects of fiber-end cracks on the stiffness of aligned short-fiber composites. Int. J. Solids Struct.
1982, 18, 723–728. [CrossRef]
23. Lagoudas, D.C.; Bo, Z.; Qidwai, M.A. A unified thermodynamic constitutive model for sm a and finite element analysis of active
metal matrix composites. Mech. Compos. Mater. Struct. 1996, 3, 153–179. [CrossRef]