A Lattice-misfit-dependent Micromechanical Approach in Ni-based Single Crystal Superalloys

https://doi.org/10.1016/j.ijmecsci.2020.106229Get rights and content

Highlights

  • Phase-field-generated microstructures are used to conduct crystal-plasticity finite-element simulations

  • A lattice-misfit-dependent isotropic hardening is formulated depending on the phase and the microstructural state

  • A natural lattice misfit of -0.3% and a 60%-completed rafting state produce the smallest stress triaxiality at the microstructure level

  • The more rafted microstructures are, the higher local stress triaxialities

  • Microstructures with larger natural lattice misfit produce smaller stress triaxiality

Abstract

In Ni-based single crystal superalloys, the macroscopic properties are intertwined with their complex bi-phased γ (NiAl)/γ’ (Ni3Al) microstructure. This study aims to better understand the coupled effect of lattice misfit and microstructural state on the macro-scale performance by utilizing a faithful representation of the microstructure, via a phase-field model, as a starting point for a micromechanical model embedded in a finite-element crystal plasticity framework. We outlined a micromechanical model with a lattice-misfit-dependent isotropic hardening that takes a unique description depending on the phase: an Eshelby tensor (sphere or penny) for the precipitates (γ’ phase) and a continuous medium for the matrix (γ phase). Finite-element simulations of strain-controlled monotonic tensile and alternate cyclic tests were performed on a set of realistic 3D phase-field microstructures (cuboidal states and directionally coarsened rafted states) defined by their natural (undeformed) or constrained (deformed) lattice misfits and the width of the γ matrix channels. As the magnitude of the natural lattice misfit gets larger for the cuboidal microstructural states, the tensile test simulations predict smaller macroscopic yield stress and strain hardening as well as smaller stress triaxiality at the microstructure level. The model also predicts that having a more pronounced rafted microstructure yields a smaller stress triaxiality at the microscale. The cyclic simulations showed that the averaged stress triaxiality at the microscale during five simulated cycles is the smallest for a natural lattice misfit of -0.3% and a 60% completed rafting process. These insights allow to better understand the experimental studies that investigated the effect of microstructural states on the mechanical response during monotonic and cyclic loading.

Introduction

Ni-based single-crystal superalloys are the materials of choice for a wide range of temperature/stress creep regimes [1,2]. It is due to the precipitation of a high volume fraction (close to 70%) of the long-range ordered L12 γ’ phase that comes as cubes coherently embedded in a face-centered cubic (fcc) solid solution γ matrix. For temperatures higher than 800°C in the low-stress regimes, the γ’ precipitates display a directional coarsening, also known as γ’ rafting [3,4]. If the natural lattice misfit, i.e., the lattice misfit before deformation, is negative/positive, the rafts will be perpendicular/parallel to the load axis. This microstructure evolution happens with an increase in the γ channel width and a modification of the lattice misfit defined as δ = 2(aγ′ − aγ)/(aγ′ + aγ) where aγ′ and aγ are the lattice parameter in the deformed state for the γ’ and γ phases, respectively. The natural lattice misfit increases in absolute value with the temperature due to a difference in the thermal coefficient between γ and γ’ and also because some alloying elements from the γ’ phase, such as Al, Ti, and Ta, contribute to the γ phase when γ’ dissolves which increases its lattice parameter [5,6]. The constrained lattice misfit, viz. the misfit during deformation, was shown to evolve during constant-stress [7,8], variable-stress [9], [10], [11], and non-isothermal [12], [13], [14] creep. Diologent et al. [7] tracked in-situ the evolution of the constrained lattice misfit during isothermal creep. They found that it increases in absolute value during the primary creep stage, which corresponds to the increase of γ/γ’ interfacial dislocations, then decreases in absolute value during the secondary creep stage due to relaxation. So, even if a fundamental understanding of the lattice misfit evolution during creep seems to have been reached, most of the modeling efforts consider either a constant lattice misfit [15], [16], [17], [18] or an incomplete description of its evolution [19]. Unfortunately, the way the lattice misfit evolves during cyclic loading and strain-controlled experiments is not available due to a lack of experimental data.

Nonetheless, it was shown, as for creep [20], [21], [22], [23]; low cycle fatigue (LCF) [24,25]; and thermo-mechanical loading [26,27], that rafting and, therefore, lattice misfit affect monotonic and cyclic loading [28], [29], [30]. Mackay et al. [28] showed on a single-crystal Ni-Al-Mo-Ta alloy that rafted microstructures have smaller yield (at 0.2%) and rupture stresses in tension for temperatures ranging from 930 to 1040°C. The same qualitative results were obtained by Pessah-Simonetti et al. on the first generation Ni-based single crystal superalloy MC2 for a 650-950°C range [29]. The same temperature range was also investigated by Espié [30] to determine the effect of the microstructure on the monotonic and cyclic behavior for the AM1 alloy. If 650°C revealed no effect because of the shearing of the γ’ particles independently of their shape, 950°C showed modifications. Espié tested the mechanical response of creep-obtained and cyclic-obtained (Δε=2%) rafted microstructures to monotonic strain-controlled tensile tests at 9. 10−4 s−1. Both rafted microstructures revealed smaller hardening at 2% than the cuboidal one. Contrary to Espié, who found no difference between creep-obtained and cyclic-obtained rafts in terms of softening, Gaubert [31] noticed that rafts obtained by cycling hardening lead to a more considerable softening compared to rafts obtained by creep. The difference might be because Espié used samples aged at 950°C while Gaubert's were aged at 1050°C leading to a significant difference in the γ’ volume fraction (70 vs. 60%) and γ channel widths. Therefore, Is the widening of the γ channels or, the modification of the lattice misfit, or perhaps a mix of these two factors responsible for the above-mentioned discrepancy in the softening behavior? The softening observed in the monotonic and cyclic mechanical responses for creep-rafted and cyclic-rafted microstructures is often interpreted as a consequence of the decrease in the Orowan stress, i.e., an increase of the γ channel widths. However, relaxation tests at 1050°C/εt = 1.2% performed on an initially undeformed and on a creep-rafted (1050°C/150 MPa for 72h and 0.65% deformation) microstructure showed that the asymptotic value, i.e. the remaining internal stress, is 1.6 times higher for the initially undeformed specimen (120 MPa vs. 75 MPa) whereas the two tests started at the same stress magnitude [31]. One should expect that a deformed specimen would lead to a higher value of the internal stresses. If the difference in the γ channel widths (110 nm for undeformed vs 300 nm for rafted) is used to explain the smaller internal stress value of the rafted specimen, then it is important to note that the decrease in the Orowan stress should only be around 35 MPa at 1050°C. It is clear from these experiments that something else is at stake and that the difference in the lattice misfit is something that should be considered in the numerical calculations. Moreover, interpreting the changes in mechanical behavior with just the Orowan stress might be erroneous and deserves more detailed studies. Such studies would provide a better understanding on how internal stresses evolve and, therefore, on the evolution of the hardening variables employed in constitutive modeling, which could, for instance, allow for better predictions of size effects [32] that was attributed to surface relaxation effects on the lattice misfit during experiments on micro-pillars [33].

Micro-mechanical modeling is a way to answer the overarching question asked in the previous paragraph. However, except the work by the authors [34], most micromechanical finite element simulations were performed with only one γ’ particle or idealized γ’ precipitate shapes [35], [36], [37], [38]. This does not provide full confidence in the results knowing that the shape of the γ’ precipitates depends on the value of the natural lattice misfit [39] and that precipitates will have different shapes between dendrite and interdendrite regions due to micro-segregations [40]. Therefore, it seems a dead end to study the effect of lattice misfit for certain microstructural states without having faithful representations of intricate γ/γ’ microstructures in all their details to account for realistic internal stress and strain distributions. The best approach to address this issue would be coupled phase-field crystal-plasticity models [41], [42], [43]. However, if such models can well predict rafting and to simulate macroscopic responses in certain conditions, they consider γ’ as either elastic or deforming elastically due to a very large critical resolved shear stress, which seems in contradiction with what was experimentally observed [44]. In addition, these studies do not control equilibrium concentrations in the presence of plasticity, which is essential in the computation of the chemical free energy density, and consider constant misfit strains.

A much more reliable and straight forward approach for short thermo-mechanical loadings, i.e, loading conditions that do not trigger microstructure evolutions, would be to improve the formulation of the recent multi-scale approach employed by the authors in [34], namely perform macro-scale finite-element calculations on realistic statistically volume elements (SVEs) obtained by phase-field simulations. On one hand, this approach leverages the advantages of both phase-field and finite-element crystal plasticity models and, on the other hand, gives the flexibility to build and calibrate a robust micromechanical model connecting the two. Furthermore, averaging or homogenization-based approaches [15,45,46] are sensitive to the volume fraction of the precipitates, but not to the shape differences in γ’ precipitates owing to different lattice misfits considered in this study. It is worth pointing out that the approach of utilizing phase-field SVEs in the finite-element crystal plasticity model has one major drawback: only short duration tests can be simulated while the microstructure is essentially non-evolving. However, this drawback is not a limitation to the aim of the present study, which is to bring forth the importance of considering the effect of misfit strains to better understand the effect of microstructural state on the macro-scale performance of Ni-based single-crystal superalloys. The study will, therefore, use realistic 3D phase-field microstructures (cuboidal and rafted) with the experimentally-consistent γ’ volume fractions as well as shapes and distributions of the γ’ precipitates at the envisioned temperature/stress conditions to simulate strain-controlled tensile tests up to 2% total strain and strain-controlled cyclic tests with Δεt = 2% both at a strain rate of ε˙=103s1.

It is important to hereby point out that a large range of natural lattice misfits, viz. from -0.2% to -0.5%, is investigated to be relevant with the large modifications of the lattice misfit during heat treatments, as shown in [47], and help the design of materials with improved performance. Thus, the changes in the natural lattice misfits are neither due to different chemical compositions nor temperature variations [48], which would, otherwise, make the set of parameters discussed in section 2.2.2 invalid. The authors want to point out that unlike the previous study [34], the present phase-field informed micro-mechanical model built into a crystal plasticity framework is sensitive to lattice misfit strain, the details of which are explained in detail in section 2.2.1. Furthermore, the authors also want to emphasize that the lattice misfit values (natural and constrained) will be considered constant during the calibration and the FE simulations, even though it is probably incorrect when such simulations last for a long time (>1300s), like in Fig. 2b, and Fig. 3b and d. It is the reason why the FE simulations performed in this study are for short tests: 20s and 200s for the monotonic and cyclic loadings, respectively. Considering that the lattice misfit does not evolve during short tests is twofold: 1) there is not enough time to trigger microstructural evolutions as well as to allow the γ/γ’ interfacial dislocations rearranging to form a network and 2) the simulated thermo-mechanical loadings do not include softening. It was shown that γ/γ’ interfacial dislocation networks are rapidly formed when softening occurs [49] and that hexagonal networks are responsible for relieving misfit strains at {111} interfaces when square networks are for {100} interfaces [50,51]. It is also consistent with what Carroll et al. [52] found, which is that the interfacial dislocation networks were not formed during creep deformation in an alloy with a natural lattice misfit that was close to zero. Therefore, for fast thermo-mechanical loading, there is not enough time for the γ/γ’ interfacial dislocations to rearrange into dislocation networks to accommodate the misfit.

Section snippets

Phase-field microstructures

Phase-field microstructures are utilized for the micromechanical approach. The phase-field framework was already described in the two previous articles by the authors [53, 54]. For the sake of completeness, the model is revisited here. This section provides the basic formulation and information regarding the parameters to generate the bi-phased γ/γ' systems similar to the microstructural states commonly encountered in Ni-based single crystal superalloys: cuboidal microstructural states

Results and Discussion

The phase-field SVEs, imported into the FE crystal plasticity framework, are enforced with boundary conditions that are prescribed on individual volume element V in a similar fashion as what can be found in [83,84].

  • kinematic uniform boundary conditions: the displacement u̲ is imposed at point x̲ that belongs to the boundary ∂V such that:ui=EijxjxjV

with E̲̲ is a symmetrical second-rank tensor that does not depend on x̲: E̲̲=1VVε̲̲dV
  • uniform traction boundary condition is applied, i.e. ti=σijon

Conclusion

Ni-based single crystal superalloys are defined by their complex bi-phased microstructures that take different shapes depending on the heat-treatment and thermo-mechanical conditions. In this study, a novel multi-scale multi-model numerical strategy is devised to understand the effect of a microstructural state of material on their macroscale performance by utilizing:

  • 1

    Thermodynamically consistent and experimentally verified microstructural configurations derived from phase-field SVE realizations

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.

Acknowledgments

The simulations were performed using the computing resources from Laboratory for Molecular Simulation (LMS) and High Performance Research Computing (HPRC) at Texas A&M University. The authors are grateful to Mr. James Fillerup and the financial support from AFOSR through Award No.: FA9550-17-1-0233 to carry out this study. We also acknowledge Dr. Adrian Loghin and GE Global Research for their support and interest in our research.

References (101)

  • C. Schwalbe et al.

    In situ measurement of the precipitate volume fraction and interfacial lattice misfit during non-isothermal creep in the superalloy CMSX-4

    Materials Science and Engineering: A

    (2019)
  • Y.-N. Fan et al.

    Constitutive modeling of creep behavior in single crystal superalloys: Effects of rafting at high temperatures

    Materials Science and Engineering: A

    (2015)
  • M.S. Titus et al.

    Creep-induced planar defects in L12-containing Co- and CoNi-base single-crystal superalloys

    Acta Materialia

    (2015)
  • K. Xu et al.

    Creep behavior and a deformation mechanism based creep rate model under high temperature and low stress condition for single crystal superalloy DD5

    Materials Science and Engineering: A

    (2020)
  • A. Ma et al.

    A model for the creep deformation behaviour of single-crystal superalloy CMSX-4

    Acta Materialia

    (2008)
  • R.C. Reed et al.

    Creep of CMSX-4 superalloy single crystals: effects of rafting at high temperature

    Acta Materialia

    (1999)
  • M.V. Nathal et al.

    The stability of lamellar γ-γ′ structures

    Materials Science and Engineering

    (1987)
  • M. Ott et al.

    Dependence of the high-temperature low-cycle fatigue behaviour of the monocrystalline nickel-base superalloys CMSX-4 and CMSX-6 on the γ/γ′-morphology

    Materials Science and Engineering: A

    (1999)
  • A. Srivastava et al.

    Phenomenological modeling of the effect of specimen thickness on the creep response of Ni-based superalloy single crystals

    Acta Materialia

    (2013)
  • K. Arora et al.

    Effects of lattice misfit on plastic deformation behavior of single-crystalline micropillars of Ni-based superalloys

    Acta Materialia

    (2017)
  • T.M. Pollock et al.

    Creep resistance of CMSX-3 nickel base superalloy single crystals

    Acta Metallurgica et Materialia

    (1992)
  • L. Müller et al.

    Calculation of the internal stresses and strains in the microstructure of a single crystal nickel-base superalloy during creep

    Acta Metallurgica et Materialia

    (1993)
  • T. Tinga et al.

    Time-incremental creep-fatigue damage rule for single crystal Ni-base superalloys

    Materials Science & Engineering A

    (2009)
  • S.-Y. Li et al.

    An anisotropic micromechanics model for predicting the rafting direction in Ni-based single crystal superalloys

    Acta Mechanica Sinica

    (2016)
  • J.S. Van Sluytman et al.

    Optimal precipitate shapes in nickel-base γ-γ' alloys

    Acta Materialia

    (2012)
  • M. Cottura et al.

    Coupling the Phase Field Method for diffusive transformations with dislocation density-based crystal plasticity: Application to Ni-based superalloys

    Journal of the Mechanics and Physics of Solids

    (2016)
  • C. Wang et al.

    Combined phase-field crystal plasticity simulation of P- and N-type rafting in Co-based superalloys

    Acta Materialia

    (2019)
  • J.-B. le Graverend et al.

    A microstructure-sensitive constitutive modeling of the inelastic behavior of single crystal nickel-based superalloys at very high temperature

    International Journal of Plasticity

    (2014)
  • B. Fedelich et al.

    Constitutive modelling of creep degradation due to rafting in single-crystalline Ni-base superalloys

    Materials Science & Engineering A

    (2009)
  • E. Wu et al.

    A neutron diffraction study of lattice distortion, mismatch and misorientation in a single-crystal superalloy after different heat treatments

    Acta Materialia

    (2013)
  • S. Ma et al.

    CRSS of γ/γ' phases from in situ neutron diffraction of a directionally solidified superalloy tension tested at 900 ˚C

    Acta Materiala

    (2008)
  • T.P. Gabb et al.

    The high temperature deformation in cyclic loading of a single crystal nickel-base superalloy

    Acta Metallurgica

    (1989)
  • R. Harikrishnan et al.

    A creep-damage phase-field model: Predicting topological inversion in Ni-based single crystal superalloys

    Materials & Design

    (2018)
  • M. Cottura et al.

    Role of elastic inhomogeneity in the development of cuboidal microstructures in Ni-based superalloys

    Acta Materialia

    (2015)
  • I. Steinbach et al.

    A phase field concept for multiphase systems

    Physica D: Nonlinear Phenomena

    (1996)
  • G. Boussinot et al.

    Phase-field simulations with inhomogeneous elasticity: Comparison with an atomic-scale method and application to superalloys

    Acta Materialia

    (2010)
  • J.W. Cahn

    On spinodal decomposition

    Acta Metallurgica

    (1961)
  • I. Steinbach et al.

    Multi phase field model for solid state transformation with elastic strain

    Physica D: Nonlinear Phenomena

    (2006)
  • K. Serin et al.

    On the influence of stress state, stress level and temperature on γ-channel widening in the singe crystal superalloy CMSX-4

    Materials Science & Engineering A

    (2004)
  • X.P. Tan et al.

    Effect of ruthenium on high-temperature creep rupture life of a single crystal nickel-based superalloy

    Materials Science and Engineering: A

    (2011)
  • R.J. Asaro

    Micromechanics of Crystals and Polycrystals

    Advances in Applied Mechanics

    (1983)
  • J. Cormier et al.

    Constitutive modeling of the creep behavior of single crystal superalloys under non-isothermal conditions inducing phase transformations

    Materials Science and Engineering: A

    (2010)
  • P. Franciosi

    The concepts of latent hardening and strain hardening in metallic single crystals

    Acta Metallurgica

    (1985)
  • S. Gao et al.

    Influence of misfit stresses on dislocation glide in single crystal superalloys: A three-dimensional discrete dislocation dynamics study

    Journal of the Mechanics and Physics of Solids

    (2015)
  • A. Royer et al.

    A non-destructive technique for the microstructural investigation of nickel base single crystal superalloys: high resolution diffraction of high energy synchrotron radiation

    Scripta Materialia

    (1997)
  • J. Gayda et al.

    Analysis of γ' shape changes in a single crystal Ni-Base superalloy

    Scripta Metallurgica

    (1989)
  • U. Glatzel et al.

    Calculations of internal stresses in the γ/γ' microstructure of a nickel-base superalloy with high volume fraction of γ'-phase

    Scripta Metall

    (1989)
  • J.-B. le Graverend et al.

    Strengthening Behavior in Non-Isothermal Monotonic and Cyclic Loading in a Ni-Based Single Crystal Superalloy

    International Journal of Fatigue

    (2016)
  • D.W. MacLachlan et al.

    Modelling the uniaxial creep anisotropy of nickel base single crystal superalloys CMSX-4 and RR2000 at 1023 K using a slip system based finite element approach

    Computational Materials Science

    (2002)
  • T. Kanit et al.

    Determination of the size of the representative volume element for random composites: statistical and numerical approach

    International Journal of Solids and Structures

    (2003)
  • Cited by (11)

    • Effect of heat treatment on microstructure and mechanical properties of GH4099 superalloy fabricated by selective laser melting

      2023, Journal of Alloys and Compounds
      Citation Excerpt :

      Owing to the precipitation of γ' phase during aging treatment, substantial strengthening elements are consumed. The contribution of the solid solution strengthening in the γ matrix will decrease with the occurrence of γ' [39–42]. In addition, the solid solution strengthening is insufficient effect on the overall strength of the precipitation-strengthened superalloy under different heat treatment conditions [39,44].

    • Tensile behavior of single-crystal superalloy with different structured cooling holes

      2022, International Journal of Mechanical Sciences
      Citation Excerpt :

      The chemical compositions of DD6 material are listed in Table 1. The microstructure of the material consists of cuboidal γ’ precipitates embedded in the γ matrix, as shown in Fig. 2 [36]. All the test specimens were fabricated from a single casted cylindrical bar with fully heat-treated conditions; i.e., 1290 °C for 1 h, 1300 °C for 2 h, 1315 °C for 4 h followed by air cooling, 1120 °C for 4 h followed by air cooling, and 870 °C for 32 h with air cooling.

    • A microstructural rafting state-based constitutive model for single crystal Ni-based superalloys at elevated temperature

      2022, International Journal of Mechanical Sciences
      Citation Excerpt :

      To date, attempts have been made for microstructure-sensitive constitutive modelling of Ni-based superalloys for tensile, cyclic [28,63], fatigue-creep interactions [57] with a stable microstructure. The models that account for the effects of microstructure evolution on the constitutive response mainly concentrate on the isothermal/non-isothermal creep deformation [11,18,23,27,38,43,36,37,44]. Three approaches that link microstructure rafting with the constitutive theory have been adopted now:

    View all citing articles on Scopus
    View full text