International Journal of Plasticity 134 (2020) 102790

Contents lists available at ScienceDirect

International Journal of Plasticity

journal homepage: http://www.elsevier.com/locate/ijplas

Modeling and simulation of coupled phase transformation and

stress evolution in thermal barrier coatings
Ferit Sait a, b, Ercan Gurses b, Ozgur Aslan a, *
a
Atilim University, Department of Mechanical Engineering, 06830, Ankara, Turkey
b
Middle East Technical University, Department of Aerospace Engineering, 06800, Ankara, Turkey

A R T I C L E I N F O A B S T R A C T

Keywords: The thermally grown oxide layer is known to be responsible for the failure of coating systems due
Thermal barrier coating to the generation of severely high stresses. In this work, oxidation induced stresses generated in
Oxidation thermal barrier coating (TBC) systems are investigated for high temperature isothermal oxidation.
Phase field
In that sense, a comprehensive model, where phase transformation is coupled with mechanics is
FEM
Finite strain plasticity
developed for the life-time estimation of TBC systems and a modified version of the Allen-Cahn
Coupled analysis type phase field approach is adopted in order to model the generation of thermally grown
Phase interface oxide (TGO) in finite strain constitutive framework. The top-coat material behavior is modeled
using a rate-dependent Gurson type plasticity for porous materials which also accounts for creep.
The results for the isothermal phase transformation analysis and the model validation using
experimental results are demonstrated. The capability of the model in predicting the local stresses
which is the main variable in the analysis of possible delaminations and accurate lifetime esti­
mation of TBC systems is shown.

1. Introduction

Thermal barrier coating systems are serving as protective layers against high temperatures of structural components. Since the
efficiency of the gas turbines improves by increasing inlet temperature, a significant amount of research has been conducted focusing
on the development of more temperature tolerant systems (see Jones (1996)). TBC systems are composed of three layers, a protective
top-coat, a layer bonding the top-coat material to the metallic substrate layer and a substrate layer usually made of a superalloy (see
DeMasi-Marcin et al. (1990)). The top-coat (TC) material is usually a ceramic based material deposited on the bond-coat (BC), which
plays the thermal isolator role in the system. In this work, the top-coat is assumed to be produced by the electron beam physical vapor
deposition (EB-PVD) method, which is one of the widely used techniques for the deposition of coating layers. This technique has the
advantage of providing a structurally more flexible protective ceramic layer due to the porous and columnar microstructure. In the case
of harsh service conditions, high strain compliance improves the resistance of the coatings against failure and spallation (see Vai­
dyanathan et al. (2000)). Metallic bond-coats are known to decrease the stresses induced by the material mismatch in the case of
thermal expansion. The diffusion of oxygen through the top-coat leads to the oxidation of the bond-coat and the development of
thermally grown oxide (TGO). Oxidation of the bond-coat occurs due to inward transportation of oxygen anions and outward
transportation of Aluminum cations. For bond-coat alloys containing yttrium, diffusion of aluminum is affected by this reactive
component, which decelerates diffusion of Aluminum cations, thus oxide components meet in the regions closer to substrate material

* Corresponding author.
E-mail address: ozgur.aslan@atilim.edu.tr (O. Aslan).

https://doi.org/10.1016/j.ijplas.2020.102790
Received 28 January 2020; Received in revised form 12 May 2020; Accepted 22 May 2020
Available online 18 June 2020
0749-6419/© 2020 Elsevier Ltd. All rights reserved.
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

Fig. 1. Gas turbine blade with thermal barrier coating system and thermally grown oxide developed in the structure.

(see Balmain et al. (1997)). Consequently, the presence of yttrium leads to the formation of oxide mainly on the oxide-bond coat
interface (Fig. 1). Due to the significant change in the volume of the oxidized region, high stresses are generated between the bond-coat
and the top-coat, which can be considered as the prime failure mechanism of the TBC systems (see Miller (1995)).
In the past few decades, simulation of TBC systems has been a challenge due to its multiphysics nature involving coupling between
mechanical, chemical and thermal processes. Modeling the growth of the TGO as a new phase in the system has solely led to numerous
efforts in the literature. Chang et al. (1987) used a numerical method to calculate isothermal stress mismatches between different
layers of TBC in a cylindrical model coated with plasma-sprayed YSZ (Yttria Stabilized Zirconia). Oxidation was modeled by changing
the properties of bond-coat elements with oxide elements in time and assigning a higher thermal expansion coefficient. A finite element
model for oxidation induced failure in plasma-sprayed TBC was also developed by Freborg et al. (1998). They focused on the creep
behavior and the oxidation of the bond-coat. Incorporating the same method for modeling swelling due to oxidation, Cheng et al.
(1998) used an elastic-perfectly plastic material model in the system components in addition to previous elastic material models.
Material property changes by temperature and thermal expansion mismatches between the TGO and the ceramic layer were also taken
into account in the analysis. A coupled thermo-mechanical model was used by He et al. (2000, 2003) and Xu et al. (2003) to analyze the
durability of TBC systems by investigating the fracture process and crack initiation due to thermal cyclic loading. Oxidation growth
was modeled by applying a definite amount of growth strain at every step of analysis in oxidized elements. Caliez et al. (2002)
developed a weakly coupled model, in which the mechanical response at one step is passed to a second step, where the Fick’s law of
diffusion is solved. Busso and Qian (2006) presented a model which additionally included the solidification process of the top-coat. A
staggered scheme for oxide generation is considered, in which oxygen concentration was calculated prior to themo-mechanical
analysis. Hille et al. (2009) incorporated the chemical reaction into the classical Fick’s law of diffusion to model TGO growth with
a sharp oxide phase front propagating in the metal. This model has been the motivation for the latest studies for coupled
chemo-themo-mechanic models by Loeffel and Anand (2011) and Chai et al. (2016). Following Hill’s work, Loeffel et al. (2013) and
Al-Athel et al. (2013) conducted a chemo-thermo-mechanical analysis in finite deformation framework. Later they also took the
top-coat creep into account in addition to the top-coat/TGO interface failure, which was modeled using cohesive zone.
Phase field approach is used in modeling a variety of microstructural heterogeneities caused by phase transformations in polymers,
metals, alloys and ceramics. Precipitation of new phases, phase growth, grain growth, dendrite formation, solidification, competitive
particle ripening, spinodal decomposition and crack propagation are among physical and metallurgical phenomena that have been
simulated using phase field models (Chen and Khachaturyan (1991); Chen et al. (1991); Wang and Young (1993); Gurtin and Voorhees
(1996); Wang and Dai (2010); Levitas et al. (2010); Brassart et al. (2012); Klusemann and Yalçinkaya (2013); Miehe et al. (2016);
Esfahani et al. (2018); Levitas et al. (2018)).
Major contributions have been made in the past decade in chemo-thermo-mechanical modeling of oxidation of metals and phase
precipitation induced swelling in two essential ways:

1. Numerical tools have been used to simulate the new phase growth and temporal evolution of the mechanical properties of the
oxided domain in the bulk material (eg. Evans et al. (2008); Busso et al. (2010); Frachon (2013)).
2. Standard (eg. Ammar (2010); Chester and Anand (2011)) or non-standard phase field (Hille et al. (2011); Loeffel et al. (2013))
approaches have been coupled with mechanics to simulate isotropic and anisotropic swelling/shrinking in the material undergoing
phase change process.

However to be best knowledge of authors a model and the corresponding simulation utilizing standard phase field theories coupled
with mechanics in large deformation framwork for formulating anisotropic swelling/shrinking is lacking. In this study, the phase field
model for the diffuse interface is coupled with mechanics to simulate the growing oxide in a thermodynamically consistent manner.
The model can be considered as a modified version of the approach proposed by Allen-Cahn (see Allen and Cahn (1979)). The
modification is based on the anisotropic phase interface diffusion of the growing oxide which is coupled with mechanics. A parametric
study of the model and the effect of the top-coat creep on the oxidation induced stresses are also presented. Moreover, a rate-dependent
Gurson-type plasticity theory is utilized for the modeling of inelastic behavior of the porous ceramic coating. The results for the stresses
in the system reveals the necessity of the incorporation of a creep model in the top-coat ceramic for the overall relaxation in the system.

2
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

Table 1
Definition of variables.
ϵ Internal energy per unit reference volume φ Phase field variable
p
ψ Helmholtz free energy per unit reference volume ϵ̇ Equivalent plastic strain rate
η Entropy per unit reference volume Me0 Deviatoric Mandel stress
q Heat flux per unit reference area P Auxiliary field variable
q Heat flux with external source per unit reference volume F Deformation gradient
j Species flux per unit reference area D Stretching tensor
ν Absolute temperature W Rotation tensor
μ Chemical potential Ce Right Cauchy-Green tensor
℘ Energy conjugate for φ (scalar) J Jacobian
ℰ Energy conjugate for ∇φ (vector) Np Plastic flow direction
st
TR 1 Piola-Krichhoff stress σ Equivalent stress
Te 2nd Piola-Krichhoff stress σv Equivalent volumetric stress
T Cauchy stress σm Matrix equivalent stress
Me Mandel stress ϵ̇m
p
Matrix equivalent plastic strain rate
v Velocity ϕ Viscoplastic potential
n Unit normal vector of the surface F Amplifying function
Ee Elastic strain tensor X Triaxiality
c Normalized concentration of species

The Gurson-type inelastic model ensures the relaxation in the porous top-coat and better consistency with the experimental results.

2. Kinematics

Considering a smooth function x = χ (X, t) mapping the motion of a point of body B in the reference configuration denoted by X to
the spatial point x at time t, the deformation gradient, the velocity, the velocity gradient and the Jacobian are given by

F = ∇χ , v = χ̇ , L = grad(v) = ḞF− 1 , J = det F (1)

Note that vectors and second order tensors are shown in bold (see Table 1). The Kröner’s decomposition (see Kröner (1959)) is used
for the deformation gradient,

F = Fe Fi (2)

where Fe is the elastic distortion which represents the stretch and rotation and Fi is the inelastic distortion. Then the velocity gradient is
decomposed in the following form.

L = Le + Fe Li Fe − 1
(3)

where,

(4)
e i − 1
Le = Ḟ Fe− 1 , Li = Ḟ Fi
e
The right polar decomposition of F yields,
Fe = Re Ue (5)

in terms of the right stretch tensor Ue and the rotation tensor Re. The right elastic Cauchy-Green tensor is defined as,

Ce = (Ue )2 = FeT Fe (6)

The Hencky’s strain,
Ee = ln(Ue ) (7)

is chosen as the strain measure. Following the decomposition given in equation (2), the Jacobian can also be decomposed as,

J = JeJi , J e = det Fe > 0, J i = det Fi > 0. (8)

The elastic and inelastic stretching and spin tensors are,

De = symLe , Di = symLi (9)

We = skewLe , Wi = skewLi (10)

Assuming irrotational plastic flow i.e.,

3
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

Wi = 0 (11)

the rate of inelastic deformation gradient would be,

(12)
i
Ḟ = Di Fi .
The total stretching tensor is decomposed in an additive manner (see Loeffel et al. (2013)),

D = De + Di . (13)
From here on, the equations defining the plastic response of the different layers are formulated separately for the top-coat ceramic
material and the bond-coat/TGO two-phase material.
Inelastic stretching tensor for a multiphase region (the bond-coat and the TGO) is defined as,

Di = Ds + Dp (14)

where Ds is the swelling stretching and Dp is due to the viscoplastic deformation. The swelling stretching Ds due to oxidation is defined
as below,
Ds = φ̇S (15)

where φ is the phase field variable defining the content of the new phase in the region, and S is the swelling strain tensor defined using
βl (normal to interface), βt (in-plane) growth parameters (see Chunhua and Wei (2000); Saeidi et al. (2018)) and n is the surface normal
to the interface profile as follows.
S = βl n ⊗ n + βt (1 − n ⊗ n) (16)
p
D is assumed to be the volume average of inelastic stretching of the bond-coat and the TGO, i.e.,
Dp = (1 − φ)Dpbc + φDpox (17)

where Dpbc and Dpox are the viscoplastic flow rate in the bond-coat and the TGO respectively. For both the oxide (α = ox) and the bond-
coat material (α = bc) the plastic stretching is
√̅̅̅
3 p p
Dpα = N ϵ̇α (18)
2
p
where ϵ̇α is the plastic strain rate. The plastic flow direction is expressed as,
Dpα
Np = (19)
dpα

where dpα is

dpα = |Dpα | (20)

3. Balance equations

Balance equations solved through the numerical process are presented below. Neglecting inertial effects in the body, the balance of
linear momentum can be expressed as,
Div(TR ) + b = 0 (21)

where TR and b are the 1st Piola-Kirchhoff stress tensor and the body force vector respectively. The first Piola stress then reads,

TR = JTF− T
(22)
The Mandel stress is defined by,
Me = Ce Te (23)
The second Piola stress can be expressed as,

Te = Fe− 1 Fe− T Me (24)

Thus, the Cauchy stress can be computed using the equation below.

T = J − 1 Fe Me Fe T (25)
Under isotropy assumption the Cauchy stress may be also expressed as (see Loeffel and Anand (2011)),

4
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

T = J − 1 Re Me ReT (26)
In this study, a continuous model proposed by Allen-Cahn, capable of describing the kinetics of the moving phase interface using
non-coherent field variables in the system is used. The model is an extension of the Cahn-Hilliard model (see Cahn and Hilliard (1958))
used for the description of two-phase domains, which includes the effect of phase interface free energy (see Raabe (1998)). The
isotropic version which is also known as time dependent Ginzburg-Landau model is used previously by Ammar et al. (2009) for the
modeling of oxide growth in a bulk metal. In contrast to bulk metal oxidation, the TGO/bond-coat interface profile would not undergo
shape smoothing evolution. The isotropic model is then incapable of maintaining the interface profile shape, due to the interface
smoothing solution of the phase kinetic equations. Thus, equations are modified to maintain the shape of the initial oxide/bond-coat
interface. The standard equations for the normalized concentration of species c (0 ≤ c ≤ 1) and the modified phase field φ (0 ≤ φ ≤ 1)
balance proposed in Saeidi et al. (2018) are as follows,
( )
∂ψ (c, φ, Ee )
ċ = Div ℳ∇( 0 ) (27)
∂c

∂ψ 0
βφ̇ = Div(α∇φ) − (28)
∂φ

where ψ 0, ℳ, α and β represent the free energy density, the mobility tensor, a tensorial and a scalar phase field parameter, respectively.
Furthermore, following Di Leo (2015) an auxiliary equation is defined with a scalar parameter π, providing the gradient of tr(Ee) term
needed in numerical implementation of the equations coupling diffusion and mechanical solutions.
P = πtr(Ee ) (29)
e
In the equation above, P is the auxiliary field variable, and E is the elastic strain tensor.
Following Gurtin et al. (2013); Gurtin (1996) neglecting the kinetic energy, and assuming that the variation of any independent
variable in the system will cause an energy change, the balance of energy in the reference configuration can be written as,
∫ ∫ ∫ ∫ ∫ ∫
d
ϵ dV = (TR n) · v dS + b · v dV − q · n dS + q dV − μj · n dS
dt V
(30)
S V S V S
∫ ∫
− ℘ φ̇ dV + ℰ · ∇φ̇ dV
V V

where ϵ, j, ℘ and ℰ are internal energy, species flux, energy conjugate for φ and energy conjugate for ∇φ respectively. μ, q and q
represent the chemical potential, the heat flux and the external heat supply.
The term with species flux j in the equation above refers to the amount of energy carried into the system by diffusing species.
Application of divergence theorem leads to
∫
dϵ
[ − [Div(TR ) + b] · v − TR : Ḟ + Div(q) − q + μDiv(j) + j · ∇μ + ℘ φ̇ − ℰ · ∇φ̇] dV = 0. (31)
V dt

Knowing the local balance of diffusion equation,

ċ = − Div(j), (32)
equation (31) can be written as,

ϵ̇ = TR : Ḟ − Div(q) + q + μċ − j · ∇μ − ℘ φ̇ + ℰ · ∇φ̇ (33)

The second law of thermodynamics for the imbalance of entropy (η) reads,
∫ ∫ ∫
d q q·n
η dV ≥ dV − dS (34)
dt V Vν S ν

where ν is the temperature. Using the divergence theorem, the energy balance equation and also keeping in mind that qν (Entropy flow)
is present, whenever the heat flux is available and neither of them can vanish without the other one Gurtin et al. (2013), one can write,
q q
η̇ ≥ − Div( ) (35)
ν ν
Then using equations (33) and (35) one obtains,
1
(ϵ̇ − νη̇) − TR : Ḟ − μċ + q · ∇ν + j · ∇μ + ℘ φ̇ − ℰ · ∇φ̇ ≤ 0 (36)
ν
Having introduced a specific free energy,
ψ = ϵ − νη (37)
The following local imbalance of energy can be written,

5
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

1
ψ̇ + ην̇ − TR : Ḟ − μċ + q · ∇ν + j · ∇μ + ℘ φ̇ − ℰ · ∇φ̇ ≤ 0 (38)
ν

4. Constitutive equations

Considering the balance of energy found in the previous section, following forms are assumed for the free energy, the stress, the
chemical potential and the entropy.
⎧
⎪
⎪ ψ = ψ (Φ)
⎪
⎪
⎪
⎪ Te = T(Φ)
⎨
μ = μ(Φ)
(39)
⎪
⎪ η = η(Φ)
⎪
⎪
⎪ ℘ = ℘(Φ)
⎪
⎩
ℰ = ℰ(Φ)

where Φ represents the set of variables

(Φ) = (Ce , ν, c, φ, ∇φ) (40)
The stress power for a multiphase material can be written as,
1
(41)
e
TR : Ḟ = Te : Ċ + (1 − φ)Me : Dpbc + φMe : Dpox + φ̇Me : S
2
Then the free energy imbalance equation (38) can be written as,
1 e e
ψ̇ + ην̇ − T : Ċ − (1 − φ)Me : Dpbc − φMe : Dpox − φ̇Me : S
2
(42)
1
− μċ + q · ∇ν + j · ∇μ + ℘ φ̇ − ℰ · ∇φ̇ ≤ 0
ν
Considering the dependence of the free energy given in equation (39), it is possible to reformulate equation (42) as,
( )
∂ψ (Φ) 1 e e ∂ψ (Φ) ∂ψ (Φ)
− T : Ċ + ν̇( + η) − (1 − φ)Me : Dpbc − φMe : Dpox − φ̇(Me : S − + ℘)
∂Ce 2 ∂ν ∂φ
(43)
∂ψ (Φ) ∂ψ (Φ) 1
− (μ − )ċ + ( − ℰ)∇φ̇ + q · ∇ν + j · ∇μ ≤ 0
∂c ∂∇φ ν
This equation must hold for all values of Φ. Therefore, the following equations can be inferred,
∂ψ (Φ) ∂ψ (Φ) ∂ψ (Φ) ∂ψ (Φ) ∂ψ (Φ)
Te = 2 , η=− , μ= , ℰ= , ℘= − Me : S (44)
∂Ce ∂ν ∂c ∂∇φ ∂φ
With thermodynamic force ℘ and energetic constitutive response function Ã,
∂ψ (Φ)
Ã(Φ) = − (45)
∂φ
Then the inequality simplifies to the following form,
1
(1 − φ)Me : Dpbc + φMe : Dpox − q · ∇ν − j · ∇μ ≥ 0 (46)
ν
Using the heat conduction equation q = − K∇ν, diffusion of species (27), and defining K as the thermal conductivity tensor and ℳ
as mobility tensor inequality (46) can be rewritten as,
1
(1 − φ)Me : Dpbc + φMe : Dpox + K∇ν · ∇ν + ℳ∇μ · ∇μ ≥ 0 (47)
ν

5. Specialization of the constitutive equations

5.1. Bond-coat

Considering the Allen-Cahn definition of total free energy density, ψ is given by,
1
ψ (Φ) = ψ 0 (Φ) + ∇φ.(α∇φ) (48)
2
As a consequence of isotropy of the material, the free energy density, given in terms of Φ(Ce, ν, c, φ, ∇φ), can also be defined in
terms of Φ(Ee, ν, c, φ, ∇φ) (see Loeffel and Anand (2011)). A specific form of free energy density for isothermal analysis can be written

6
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

as,

ψ 0 (Φ) = ψ M (φ, Ee ) + ψ φ (φ, c) + ψ P (φ, Ee , c) (49)

with mechanical part of the free energy density defined as,

1 2
ψ M (φ, Ee ) = G(φ)|Ee |2 + [K(φ) − G(φ)]tr(Ee )2 (50)
2 3

where G(φ) and K(φ) denote the shear and the bulk modulus of the two phase material, respectively. Material properties which have
phase field variable dependency are defined as a linear function of phase field.
K(φ) = Kbc (1 − φ) + Kox φ (51)

G(φ) = Gbc (1 − φ) + Gox φ (52)

The phase field part is adapted from Kim et al. (1998),

ψ φ (φ, c) = φ2 (3 − 2φ)ψ 1 (c) + [1 − φ2 (3 − 2φ)]ψ 2 (c) + χ φ2 (1 − φ)2 (53)

1
ψ i (c) = ki (c − ci )2 (i = 1, 2) (54)
2

where k and χ are material phase parameters. c1 and c2 are the maximum and minimum normalized concentration of species in the
domain. The coupling part of the free energy density is formulated as,

ψ P (φ, Ee , c) = − cmax K(φ)P(c − c0 ) (55)

where cmax is the maximum oxygen concentration in the alumina (see Grimes and Lagerlo (1998)).
Using (24), (44), (49) and (48), the Mandel Stress Me, thermodynamic force Ã(Φ), ℰ and ℘ can derived in the following form,
Me = 2GEe + λtr(Ee )1 (56)

Ã(Φ) = (3φ − 2)(ψ 1 − ψ 2 ) − 2χ φ(1 − φ)(1 − 2φ) − (Kox − Kbc )(c − c0 )cmax P (57)

ℰ = 2(α∇φ) (58)

℘ = Me : S + Ã(Φ) (59)

where, G and λ are the shear modulus and the lamé constant, respectively and 1 is the identity tensor. The chemical potential μ can be
calculated using equation (49).

μ = k1 (c − c1 )φ2 (3 − 2φ) + k2 (c − c2 )(1 − φ2 (3 − 2φ)) − cmax K(φ)P (60)

Calculating the deviatoric part of the Mandel Stress,
1
Me0 = Me − tr(Me )1 (61)
3
The plastic flow direction Np is assumed to be colinear to the Mandel stress Me and codirectional for both phases.
√̅̅̅ e
3 M0
Np = (62)
2 σ

where σ is the equivalent stress. The equivalent tensile stress is defined as,
√̅̅̅
3 e
σ= |M | (63)
2 0
The microforce balance including a power-law rate dependence reads,
p m
ϵ̇
(64)
p p
σ = g(ϵ̇α , Sα ) = Sα ( α ) when ϵ̇α > 0
ϵ̇0
Parameter Sα is known as the material or deformation saturation resistance. In equation (64), g is a power law function, m is the rate
sensitivity parameter and ϵ̇0 is the reference plastic flow rate.

7
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

5.2. Top-coat

Creep resistant bond-coats are more desirable due to less interface oxidation roughening under severe thermo-mechanical con­
ditions (see Vaidyanathan et al. (2000)). However the top-coat creep is believed to be one of the major mechanisms extending coating
life via lowering stress levels in the TBC system. Thus, in our modeling attempt, the top-coat creep is considered and a rate-dependent
Gurson type plasticity (see Zavaliangos (1992)) for porous material is adopted to the overall model. In addition to volumetric plasticity,
the model can also describe creep relaxation and plastic flow due to deviatoric stresses, (see Haghi and Anand (1992)). For the
top-coat, the pure mechanical free energy density presented in equation (50) is considered by taking φ = 0. The plastic stretching
tensor is defined as
√̅̅̅
3 p ∂ϕ 1 ∂ϕ
Dp = N + 1 (65)
2 ∂σ 3 ∂σ v

where definitions for σ and Np are given in (62) and (63). σ v is the equivalent volumetric stress and defined as
1
σv = tr(Me ) (66)
3
In equation (65) ϕ represents the viscoplastic potential. In order to include creep in the model the volumetric plastic deformation
rate tensor is also added to the plastic deformation rate tensor (see Zavaliangos and Anand (1993)).
The effective equivalent plastic strain rate of the dense material was described using an evolution function formulated for creep at
elevated temperatures (see Anand (1982)).
σm
(67)
p 1
ϵ̇m = ϵ̇0 ( )m
S
Expressing a rate-dependent behavior for the porous material, the viscoplastic potential function ϕ and the effective equivalent
stress in the matrix (material without pores) σ m are formulated as,
ϵ̇0 S σm
ϕ = F(X, m)[ ( )1/m+1 ] (68)
1/m + 1 S

(69)
m
σm = κ(F)m+1 · σ

where, F is a function amplifying plastic flow due to presence of the voids via amplifying potential. κ, m and X represent a porosity
parameter, the rate sensitivity parameter and the triaxiality respectively. F and X are given as,

F = A1 + (A2 X 2 + A3 )(1/m+1)/2 (70)

σv
X= (71)
σ

where A1, A2 and A3 parameters defined in terms of rate sensitivity parameter m and triaxiality parameter X. For simplicity these
parameters are adopted in the specific form introduced by Haghi and Anand (1992) and presented as constant values. The evolution of
deformation resistance for top-coat material is defined as,
S q
(72)
p
Ṡ = hϵ̇m (1 − )
Ssat

where S0, Ssat, q and h respectively denote the initial deformation resistance, the saturation resistance and the hardening parameters.

6. Finite element implementation

Implementation of the model is carried out by using User Element (UEL) in ABAQUS software, where ABAQUS is basically used as a
solver and a visualiser. Two-dimensional quadrilateral elements with four linear nodes and bi-linear shape functions are developed for
the plane-strain analysis. Since ABAQUS does not support graphical output of the user elements, for the visualization of the results, a
mock mesh is created using dummy CPE4T plane strain elements already defined in ABAQUS. These elements are attached to the user
elements to share the outputs. A total of five degrees of freedom are defined for the elements. In addition to the displacement degrees of
freedom (u1 and u2), the normalized concentration of species (c), the phase field variable (φ) and the mechanical coupling term P
defined in equation (29) are defined as nodal degrees of freedom of the finite element formulation. The extra degree of freedom (P)
provides ease in the calculation of its gradient through shape functions. This gradient term is used in the calculation of the residual
vector given in equation (76) for the normalized concentration, Rc.
We define,

a = k1 φ2 (3 − 2φ) + k2 (1 − φ2 (3 − 2φ)) (73)

8
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

Fig. 2. Mechanical boundary condition and TGO evolution in time after 10 h, 100 h and 200 h.

b = k1 (c − c1 ) − k2 (c − c2 ) (74)

for a compact representation. Let NA be the shape function for the interpolation of node A, then the residuals for the displacement u,
phase field φ, normalized concentration c and the mechanical coupling term P for each element are expressed as,
∫ ∫
Ru = (NA b − ∇NA TR ) + NA t dS (75)
e ∂e

∫ ∫
Rc = (NA ċ − cmax K(φ)ℳ∇NA ∇P + ℳa∇NA ∇c + 6ℳbφ(1 − φ)∇NA ∇φ) dV − NA jR · n dS = 0 (76)
e ∂e

∫ ∫
Rφ = (NA (βφ̇ + 3φ(1 − φ)𝒞 + χ 𝒟) + α∇NA ∇φ) dV + NA ℰ · n dS (77)
e ∂e

∫
Re = NA (P − πtr(Ee )) dV (78)
e

where the scalar values 𝒞 and 𝒟 are defined as,

𝒞 = k1 (c − c1 )2 − k2 (c − c2 )2 (79)

𝒟 = 4φ3 − 6φ2 + 2φ (80)

Mobility of the species ℳ and parameter α are defined to be second order tensors. The components of these tensors are chosen such
that they prevent lateral propagation of the oxide supporting the experimental results claiming negligible lateral growth against rapid
growth in the direction perpendicular to the interface. All the elements of these tensors other than the diagonal element that defines the
mobility perpendicular to the interface are assumed to be sufficiently small. This modification results in a controllable oxidation
interface profile shape in agreement with the experiments. Neglecting body forces, non-zero elements of the stiffness matrix are
derived in the following form,

9
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

Table 2
Material and analysis parameters.
Definition TC (EB-PVD) BC (FeCrAlY) TGO (Al2O3) Unit

G Shear modulus 15.68 × 103 92.31 × 103 132.0 × 103 MPa

K Bulk modulus 19.27 × 103 200.0 × 103 273.0 × 103 MPa
Sα Saturation resistance 7.0 × 101 5.0 × 102 1.0 × 103 MPa
S0 Initial saturation 2.0 × 101 – – MPa
h Hardening parameter 9.0 × 103 – – MPa
m Rate sensitivity parameter 0.25 0.4 0.4 –
ϵ˙0 Reference strain rate 0.5 × 10− 10 1 × 10− 4 1 × 10− 4 –
17 17 17
ℳ22 Mobility 2 × 10− 2 × 10− 2 × 10− m5 ⋅ s ⋅ mol/J
α22 Phase field parameter 3.62 × 10− 7 3.62 × 10− 7 3.62 × 10− 7 J/m
β Phase field parameter 0.95 × 106 0.95 × 106 0.95 × 106 J/m3
χ Phase field parameter 3.916 × 10− 1 3.916 × 10− 1 3.916 × 10− 1 –
k1, k2 Phase field parameter 6.0 × 103 6.0 × 103 6.0 × 103 J ⋅/mol ⋅ m3
c1 Min. norm. concentration 0.14 0.14 0.14 –
c2 Max. norm. concentration 0.90 0.90 0.90 –
βl Growth parameter – – 1.4 × 10− 1 –
βt Growth parameter – – 1.6 × 10− 3 –
cmax Max. oxygen concentration – – 0.8 × 105 mol/m3
π Coupling parameter – – 1 × 10− 1 –
q Creep parameter 1.3 – – –
A1 Creep parameter 1.0 – – –
A2 Creep parameter 20.3 × 105 – – –
A3 Creep parameter 1.3 – – –
κ Creep parameter 1.02 – – –

∫ ∫
∂TR ∂t
KAB
uu = ∇NA ∇NB dV − NA NB dS (81)
e ∂F ∂e ∂u
∫ ( )) ∫
NA NB ∂b ∂j
KAB
cc = − + aℳ∇NA ∇NB + 6ℳφ(1 − φ) NB ∇NA ∇φ dV − NA NB · n dS (82)
e Δt ∂c ∂e ∂c
∫ ( ) ∫
NA NB ∂𝒟 ∂ℰ
KAB
φφ = − β + α∇NA ∇NB + 3NA NB (1 − 2φ)𝒞 + NA NB χ dV − NA NB · n dS (83)
e Δt ∂φ ∂e ∂φ
∫
KAB
ee = − (NA NB ) dV (84)
e

∫ ( )
∂K(φ)
KAB
cφ = − NA NB + 6bℳ∇NA ∇NB dV (85)
e ∂φ
∫
KAB
ce = − (3NA NB K(φ)) dV (86)
e

∫
KAB
φc = − (2NA NB k1 φ(1 − φ)(c − c1 )) dV (87)
e

∫
∂TR
KAB
uφ = ∇NA N dV (88)
e ∂φ B

7. Results and discussion

For the simulation of a TBC system a two-layer geometric model is created with a 150 μm top-coat and a bond-coat with equal
thicknesses (see Fig. 2). The problem is assumed to be isothermal at 1200oC. No species flux boundary condition is applied on the left
and right edges of the medium. The phase field and the concentration variables are set as φ = 1 and c = 0.92 at the top edge and φ =
0 and c = 0.12 at the bottom edge. The initial conditions for the phase field and the concentration variables are consistent with the
boundary conditions and follow a hyperbolic tangent transition function at the phase interface. This function provides a smooth
transition of the order parameter from φ = 1 at the top of the interface to φ = 0 at the bottom. Although the physical length of the
transition region is not clearly defined in the literature for a growing oxide in a TBC system, 2 μm transition length is chosen for the
hyperbolic tangent transition function. This length is in accordance with the mesh sizes in the interface region of our finite element
model (see Kim et al. (1998) and Ammar et al. (2009)). Displacement of the left edge is constrained in the horizontal direction and
vertical displacement is constrained at the bottom boundary as depicted in Fig. 2. The right boundary is constrained to remain vertical
to mimic the continuity of the sinusoidal pattern in the system. The mesh used has a minimum size of 0.5 μm as a result of the mesh

10
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

Fig. 3. Calibration of parameter π and creep strain during TGO growth.

Fig. 4. TGO thickness versus time for the increasing values of phase field parameter β and the mobility parameter ℳ.

sensitivity test in the previous work (see Saeidi et al. (2018)). The material parameters used in the analysis are presented in Table 2.
Mechanical properties are taken from previous works by He et al. (2003); Draper et al. (2017); Patsias et al. (2006); Spriggs and
Brissette (1962) and Morrell (1987).
Defining the ℳ22 parameter of the ℳ tensor as the only non-zero component, leads to the TGO growth in the transvers direction.
For distinct geometries and undulations this tensor has to be calibrated with the growth direction by defining remaining elements of
the tensor. The top-coat/bond-coat interface is magnified in Fig. 2, where the progression of the oxide front in the bond-coat can be
clearly observed. There is no experiment in the literature showing the effect of volumetric pressure on the species diffusion in porous
materials to the best knowledge of authors, but numerical and physical validation of the hypothesis can be observed in the result
presented in Fig. 3. The results show that for higher values of the parameter π the TGO growth speed is lower as expected due to the
increase in the scaled compressive pressure. The parameter π is calibrated in accordance with the experimental data provided by
Tolpygo and Clarke (1998).
Due to the lack of creep data for EB-PVD coatings, experimental results provided for APS ceramic tested at 1050oC are used for the
calibration of creep parameters q and A3 (see Echsler et al. (2004)). The remaining creep parameters A1, A2 and κ are taken from Haghi
and Anand (1992). The numerical results showing creep strain for one element medium under different compressive loads are pre­
sented in Fig. 3.
Having identified all the possible solution parameters an increase in the mobility of the species ℳ and a decrease in the phase
parameter β are expected to accelerate the growth of TGO in the region, thus effects of those parameters on TGO growth are presented
in Fig. 4.
The contour plots showing the distribution of σ11, σ 22 and the equivalent deviatoric stresses and the plastic strain are shown in the
bond-coat and the top-coat (Fig. 5 and Fig. 6). During the phase transformation, large plastic strains due to high compressive stresses
are generated in the multiphase region, which later transforms into a single oxide phase. Fig. 7 shows the evolution of σ 11 and σ 22 at a
point in the vicinity of the interface in the TGO region and Fig. 8 shows evolution of those stresses in the vicinity of the interface in the

11
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

Fig. 5. Stress (MPa) and equivalent plastic strain results in the top-coat after 50 h.

Fig. 6. Stress (MPa) and strain results for bond-coat layer after 50 h of isothermal service.

top-coat region. The stress magnitudes generated in the oxide region is in good agreement with the experimental results provided by
Tortorelii et al. (2003) for a similar system, which reports a maximum of 800 MPa compressive stress (see Figs. 7 and 8). Fig. 9 shows
two stress distributions in the bond-coat for the purely elastic top-coat and plasticity-creep included top-coat material models. It can be
seen from the figure that the maximum tensile and compressive stresses in the TGO and the bond-coat are significantly reduced. Fig. 9

12
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

Fig. 7. Evolution of stresses in TGO, in 22 (left) and 11 (right) directions.

Fig. 8. Evolution of stresses in 11 and 22 directions in top-coat in downhill (left) and uphill of the interface (right).

Fig. 9. Stress distribution for purely elastic top-coat (left) and plasticity-creep included top-coat (right) conditions.

13
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

shows two stress distributions in the bond-coat for the purely elastic top-coat and plasticity-creep included top-coat material models. It
can be seen from the figure that the maximum tensile and compressive stresses in the TGO and the bond-coat are significantly reduced.

8. Concluding remarks

A finite element based phase-field model is developed in the large deformation framework for oxidation of TBC systems at high
temperatures. The phase field approach and mechanics are fully coupled within the model, where the thermodynamic consistency is
also ensured.
A numerical analysis is carried out for the progression of TGO in a TBC system at 1200oC. The residual stresses achieved by the
numerical simulation in the bond-coat are shown to be in good agreement with the experimental data.
The parametric study of the model is also conducted to investigate the effect of individual phase field and coupling parameters on
the oxide growth rate. The results related to the calibration of the mechanical coupling parameter π demonstrate that the coupling
parameter strongly connects the volumetric stress to the TGO growth rate.
The effect of creep on the oxidation induced stresses in the top-coat is also studied by including a Gurson-type inelasticity model in
the ceramic coating. A considerable creep strain is observed in the ceramic coating, which leads to significant stress relaxation in the
system in contrast to the models, where the top-coat is assumed to be elastic. The assumption of pure elastic behavior increases the
stresses generated in the bond-coat, which results in shorter lifetime estimations.
Coupling the Allen-Cahn phase field approach with mechanics, the presented model shows a great potential for predicting lifetime
of TBC systems, where the calculation of stresses and plastic strains due to TGO growth is possible. This approach can be further
extended to implement thermal coupling for the consideration of cyclic thermal loads. Furthermore, a continuum level strain based
damage can also be introduced in the model for a better lifetime estimation.

CRediT authorship contribution statement

Ferit Sait: Conceptualization, Methodology, Software, Validation, Writing - original draft. Ercan Gurses: Writing - review &
editing. Ozgur Aslan: Conceptualization, Methodology, Funding acquisition, Supervision, Writing - review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Acknowledgements

The authors gratefully acknowledge the support from the Scientific and Technological Research Council of Turkey under project
number 315M138.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijplas.2020.102790.

References

Al-Athel, K., Loeffel, K., Liu, H., Anand, L., 2013. Modeling decohesion of a top-coat from a thermally-growing oxide in a thermal barrier coating. Surf. Coating.
Technol. 222, 68–78. https://doi.org/10.1016/j.surfcoat.2013.02.005.
Allen, S.M., Cahn, J.W., 1979. A microscopic theory for antiphase boundary motion and its application to antiphase domain coarsening. Acta Metall. 27, 1085–1095.
https://doi.org/10.1016/0001-6160(79)90196-2 arXiv:9809069v1.
Ammar, K., 2010. Modelling and simulation of phase transformation-mechanics coupling using a phase field method Kais Ammar l ’ École nationale supérieure des
mines de Paris Spécialité ≪ Sciences et Génie des Matériaux ≫ Modelling and Simulation of Phase Transformation-Mec.
Ammar, K., Appolaire, B., Cailletaud, G., Feyel, F., Forest, S., 2009. Finite element formulation of a phase field model based on the concept of generalized stresses.
Comput. Mater. Sci. 45, 800–805. https://doi.org/10.1016/j.commatsci.2008.09.015, 10.1016/j.commatsci.2008.09.015.
Anand, L., 1982. Constitutive equations for the rate-dependent deformation of metals at elevated temperatures. J. Eng. Mater. Technol. 104, 12–17. https://doi.org/
10.1115/1.3225028.
Balmain, J., Huntz, A.M., Philibert, J., 1997. Atomic transport properties in alumina scales and calculation of the oxidation parabolic rate constant. Defect Diffusion
Forum 143–147, 1189–1194. https://doi.org/10.4028/www.scientific.net/DDF.143-147.1189.
Brassart, L., Stainier, L., Doghri, I., Delannay, L., 2012. Homogenization of elasto-(visco) plastic composites based on an incremental variational principle. Int. J. Plast.
36, 86–112. https://doi.org/10.1016/j.ijplas.2012.03.010, 10.1016/j.ijplas.2012.03.010.
Busso, E.P., Evans, H.E., Qian, Z.Q., Taylor, M.P., 2010. Effects of breakaway oxidation on local stresses in thermal barrier coatings. Acta Mater. 58, 1242–1251.
https://doi.org/10.1016/j.actamat.2009.10.028, 10.1016/j.actamat.2009.10.028.
Busso, E.P., Qian, Z.Q., 2006. A mechanistic study of microcracking in transversely isotropic ceramic-metal systems. Acta Mater. https://doi.org/10.1016/j.
actamat.2005.09.003.
Cahn, J.W., Hilliard, J.E., 1958. Free energy of a nonuniform system. I. Interfacial free energy. J. Chem. Phys. 28, 258–267. https://doi.org/10.1063/1.1744102
arXiv:9809069v1. http://aip.scitation.org/doi/10.1063/1.1744102.

14
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

Caliez, M., Feyel, F., Kruch, S., Chaboche, J.L., 2002. Oxidation induced stress fields in an EB-PVD thermal barrier coating. Surf. Coating. Technol. 157, 103–110.
https://doi.org/10.1016/S0257-8972(02)00167-6.
Chai, Y., Lin, C., Wang, X., Li, Y., 2016. Study on stress development in the phase transition layer of thermal barrier coatings. Materials 9, 773. https://doi.org/
10.3390/ma9090773. http://www.mdpi.com/1996-1944/9/9/773.
Chang, G.C., Phucharoen, W., Miller, R.A., 1987. Behavior of thermal barrier coatings for advanced gas turbine blades. Surf. Coating. Technol. 30, 13–28. https://doi.
org/10.1016/0257-8972(87)90004-1.
Chen, L.Q., Khachaturyan, A.G., 1991. Computer simulation of structural transformations during precipitation of an ordered intermetallic phase. Acta Metall. Mater.
39, 2533–2551. https://doi.org/10.1016/0956-7151(91)90069-D.
Chen, L.Q., Wang, Y., Khachaturyan, A.G., 1991. Transformation-induced elastic strain effect on the precipitation kinetics of ordered intermetallics. Phil. Mag. Lett.
64, 241–251. https://doi.org/10.1080/09500839108214618.
Cheng, J., Jordan, E., Barber, B., Gell, M., 1998. Thermal/residual stress in an electron beam physical vapor deposited thermal barrier coating system. Acta Mater. 46,
5839–5850. https://doi.org/10.1016/S1359-6454(98)00230-4.
Chester, S.A., Anand, L., 2011. A Thermo-Mechanically Coupled Theory for Fluid Permeation in Elastomeric Materials: Application to Thermally Responsive Gels.
https://doi.org/10.1016/j.jmps.2011.07.005.
Chunhua, X., Wei, G., 2000. Pilling-Bedworth ratio for oxidation of alloys. Mat Res Innovat 3, 231–235. https://doi.org/10.1007/s100190050008.
DeMasi-Marcin, J.T., Sheffler, K.D., Bose, S., 1990. Mechanisms of degradation and failure in a plasma-deposited thermal barrier coating. J. Eng. Gas Turbines Power
112, 521–526. https://doi.org/10.1115/1.2906198.
Di Leo, C.V., 2015. Chemo-Mechanics of Lithium-Ion Battery Electrodes (Ph.D. thesis).
Draper, S.L., AIKlNB, J., Eldridge, J., 2017. Tensile Behavior of AI 2 0 3/FeAI Composites + Band AI 2 0iFeCrAIY. Technical Report October 1995.
Echsler, H., Renusch, D., Schütze, M., 2004. Mechanical behaviour of as sprayed and sintered air plasma sprayed partially stabilised zirconia. Mater. Sci. Technol. 20,
869–876. https://doi.org/10.1179/026708304225017346.
Esfahani, S.E., Ghamarian, I., Levitas, V.I., Collins, P.C., 2018. Microscale phase field modeling of the martensitic transformation during cyclic loading of NiTi single
crystal. Int. J. Solid Struct. 146, 80–96. https://doi.org/10.1016/j.ijsolstr.2018.03.022, 10.1016/j.ijsolstr.2018.03.022.
Evans, A.G., Clarke, D.R., Levi, C.G., 2008. The influence of oxides on the performance of advanced gas turbines. J. Eur. Ceram. Soc. 28, 1405–1419. https://doi.org/
10.1016/j.jeurceramsoc.2007.12.023.
Frachon, J., 2013. Multiscale Approach to Predict the Lifetime of EB-PVD Thermal Barrier Coatings. Thesis. Ecole Nationale Superieure des Mines de Paris.
Freborg, A.M., Ferguson, B.L., Brindley, W.J., Petrus, G.J., 1998. Modeling oxidation induced stresses in thermal barrier coatings. Mater. Sci. Eng. 245, 182–190.
https://doi.org/10.1016/S0921-5093(97)00849-6.
Grimes, R.W., Lagerlo, K.P.D., 1998. THE DEFECT CHEMISTRY OF SAPPHIRE (a -Al 2 O 3), 46, pp. 5689–5700.
Gurtin, M., Fried, E., Anand, L., 2013. The Mechanics and Thermodynamics of Continua, 53. https://doi.org/10.1017/CBO9781107415324.004 arXiv:arXiv:
1011.1669vol. 3.
Gurtin, M.E., 1996. Generalized Ginzburg-Landau and Cahn-Hilliard equations based on a microforce balance. Physica D 96, 178–192. https://doi.org/10.1155/
S1110757X03204083.
Gurtin, M.E., Voorhees, P.W., 1996. The thermodynamics of evolving interfaces far from equilibrium. Acta Mater. 44, 235–247. https://doi.org/10.1016/1359-6454
(95)00139-X.
Haghi, M., Anand, L., 1992. A constitutive model for isotropic, porous, elastic-viscoplastic metals. Mech. Mater. 13, 37–53. https://doi.org/10.1016/0167-6636(92)
90034-B.
He, M.Y., Evans, A.G., Hutchinson, J.W., 2000. Ratcheting of compressed thermally grown thin films on ductile substrates. Acta Mater. 48, 2593–2601. https://doi.
org/10.1016/S1359-6454(00)00053-7.
He, M.Y., Hutchinson, J.W., Evans, A.G., 2003. Simulation of stresses and delamination in a plasma-sprayed thermal barrier system upon thermal cycling. Mater. Sci.
Eng. 345, 172–178. https://doi.org/10.1016/S0921-5093(02)00458-6.
Hille, T.S., Nijdam, T.J., Suiker, A.S.J., Turteltaub, S., Sloof, W.G., 2009. Damage growth triggered by interface irregularities in thermal barrier coatings. Acta Mater.
57, 2624–2630. https://doi.org/10.1016/j.actamat.2009.01.022, 10.1016/j.actamat.2009.01.022.
Hille, T.S., Turteltaub, S., Suiker, A.S., 2011. Oxide growth and damage evolution in thermal barrier coatings. Eng. Fract. Mech. 78, 2139–2152. https://doi.org/
10.1016/j.engfracmech.2011.04.003, 10.1016/j.engfracmech.2011.04.003.
Jones, R., 1996. Thermal barrier coatings. In: Stern, K.H. (Ed.), Metallurgical and Ceramic Protective Coatings. Chapman & Hall, London, p. 194. https://doi.org/
10.1533/9780857090829.2.115.
Kim, S.G., Kim, W.T., Suzuki, T., 1998. Interfacial compositions of solid and liquid in a phase-field model with finite interface thickness for isothermal solidification in
binary alloys. Phys. Rev. E 58, 3316–3323. https://doi.org/10.1103/PhysRevE.58.3316.
Klusemann, B., Yalçinkaya, T., 2013. Plastic deformation induced microstructure evolution through gradient enhanced crystal plasticity based on a non-convex
Helmholtz energy. Int. J. Plast. 48, 168–188. https://doi.org/10.1016/j.ijplas.2013.02.012, 10.1016/j.ijplas.2013.02.012.
Kröner, E., 1959. Allgemeine Kontinuumstheorie der Versetzungen und Eigenspannungen. Arch. Ration. Mech. Anal. 4, 273–334. https://doi.org/10.1007/
BF00281393.
Levitas, V.I., Jafarzadeh, H., Farrahi, G.H., Javanbakht, M., 2018. Thermodynamically Consistent and Scale-dependent Phase Field Approach for Crack Propagation
Allowing for Surface Stresses, 111. Elsevier. https://doi.org/10.1016/j.ijplas.2018.07.005, 10.1016/j.ijplas.2018.07.005.
Levitas, V.I., Lee, D.W., Preston, D.L., 2010. Interface propagation and microstructure evolution in phase field models of stress-induced martensitic phase
transformations. Int. J. Plast. 26, 395–422. https://doi.org/10.1016/j.ijplas.2009.08.003, 10.1016/j.ijplas.2009.08.003.
Loeffel, K., Anand, L., 2011. A chemo-thermo-mechanically coupled theory for elastic-viscoplastic deformation, diffusion, and volumetric swelling due to a chemical
reaction. Int. J. Plast. 27, 1409–1431. https://doi.org/10.1016/j.ijplas.2011.04.001.
Loeffel, K., Anand, L., Gasem, Z.M., 2013. On modeling the oxidation of high-temperature alloys. Acta Mater. 61, 399–424. https://doi.org/10.1016/j.
actamat.2012.07.067.
Miehe, C., Aldakheel, F., Raina, A., 2016. Phase field modeling of ductile fracture at finite strains: a variational gradient-extended plasticity-damage theory. Int. J.
Plast. 84, 1–32. https://doi.org/10.1016/j.ijplas.2016.04.011, 10.1016/j.ijplas.2016.04.011.
Miller, R.a., 1995. Thermal barrier coatings for aircraft engines: history and directions. J. Therm. Spray Technol. 6, 35–42. https://doi.org/10.1007/BF02646310.
Morrell, R., 1987. Handbook of properties of technical & engineering ceramics. In: Number Pt. 2 in Handbook of Properties of Technical & Engineering Ceramics, H.
M.S.O. https://books.google.com.tr/books?id=7LxRAAAAMAAJ.
Patsias, S., Tassini, N., Lambrinou, K., 2006. Ceramic coatings: effect of deposition method on damping and modulus of elasticity for yttria-stabilized zirconia. Mater.
Sci. Eng. 442, 504–508. https://doi.org/10.1016/j.msea.2006.04.129.
Raabe, D., 1998. Computational Material Science. Wiley-VCH. https://doi.org/10.15713/ins.mmj.3.
Saeidi, F., Gurses, E., Aslan, O., 2018. A numerical approach to simulate oxidation in thermal barrier coatings. In: Reference Module in Materials Science and
Materials Engineering. Elsevier Ltd., pp. 1–7. https://doi.org/10.1016/B978-0-12-803581-8.11194-4. https://linkinghub.elsevier.com/retrieve/pii/
B9780128035818111944
Spriggs, R.M., Brissette, L.A., 1962. Expressions for shear modulus and Poisson’s ratio of porous refractory oxides. J. Am. Ceram. Soc. 45, 198–199. https://doi.org/
10.1111/j.1151-2916.1962.tb11121.x.
Tolpygo, V.K., Clarke, D.R., 1998. Wrinkling of [alpha]-alumina films grown by thermal oxidation–I. Quantitative studies on single crystals of Fe-Cr-Al alloy. Acta
Mater. 46, 5153–5166. https://doi.org/10.1016/s1359-6454(98)00133-5. http://www.sciencedirect.com/science/article/B6TW8-3TYG6HT-13/2/
aba152d5194d0e3dfa7104b1217462d6.
Tortorelii, P.F., More, K.L., Specht, E.D., Pint, B.A., Zschack, P., 2003. Growth stress - microstructure relationships for alumina scales. Mater. A. T. High. Temp. 20,
303–310. https://doi.org/10.1179/mht.2003.036.

15
F. Sait et al. International Journal of Plasticity 134 (2020) 102790

Vaidyanathan, K., Gell, M., Jordan, E., 2000. Mechanisms of spallation of electron beam physical vapor deposited thermal barrier coatings with and without platinum
aluminide bond coat ridges. Surf. Coating. Technol. 133–134, 28–34. https://doi.org/10.1016/S0257-8972(00)00891-4.
Wang, J., Dai, H.H., 2010. Phase transitions induced by extension in a slender SMA cylinder: analytical solutions for the hysteresis loop based on a quasi-3D
continuum model. Int. J. Plast. 26, 467–487. https://doi.org/10.1016/j.ijplas.2009.08.004, 10.1016/j.ijplas.2009.08.004.
Wang, J., Young, A.P., 1993. Monte Carlo study of the six-dimensional Ising spin glass. J. Phys. Math. Gen. 26, 1063. http://stacks.iop.org/0305-4470/26/i=5/
a=025.
Xu, T., He, M.Y., Evans, A.G., 2003. A numerical assessment of the durability of thermal barrier systems that fail by ratcheting of the thermally grown oxide. Acta
Mater. 51, 3807–3820. https://doi.org/10.1016/S1359-6454(03)00194-0.
Zavaliangos, A.I., 1992. zavaliangos.pdf. Ph.D. thesis. Massachusetts institute of technology.
Zavaliangos, A.I., Anand, L., 1993. Thermo-elasto-viscoplasticity of isotropic porous metals. J. Mech. Phys. Solid. 41, 1087–1118. https://doi.org/10.1016/0022-
5096(93)90056-L.

16