Fatigue in Materials: Fundamentals, Multiscale Characterizations and Computational Modeling: Multiscale Modeling Approaches to Improve Fatigue Predictions
Sponsored by: TMS Structural Materials Division, TMS Materials Processing and Manufacturing Division, TMS: Additive Manufacturing Committee, TMS: Advanced Characterization, Testing, and Simulation Committee, TMS: Computational Materials Science and Engineering Committee, TMS: Integrated Computational Materials Engineering Committee, TMS: Mechanical Behavior of Materials Committee
Program Organizers: Garrett Pataky, Clemson University; Ashley Spear, University of Utah; Antonios Kontsos, Drexel University; Brian Wisner, Ohio University; J.C. Stinville, University of Illinois Urbana-Champaign
Thursday 8:30 AM
March 18, 2021
Room: RM 34
Location: TMS2021 Virtual
Session Chair: Antonios Kontsos, Drexel University
8:30 AM
Experimental Analysis and Numerical Simulation of Cyclic Deformation and Fatigue Behavior of AZ31 Mg Alloy: Abbas Jamali1; Meijuan Zhang2; Anxin Ma2; Javier Llorca1; 1IMDEA Materials Institute & Technical University of Madrid; 2IMDEA Materials Institute
The mechanical response and the low cycle fatigue behavior under fully-reversed cyclic deformation was determined in different orientations in an extruded AZ31 Mg alloy for different cyclic strain amplitudes. The shape of the cyclic stress-strain curves depended on the orientation due to the strong texture of the alloy and to the development of twinning under certain combinations of load and orientation. The mechanical response of the material under cyclic deformation was simulated by means of computational homogenization of a representative volume element of the microstructure of the alloy, The behavior of the Mg crystals followed a phenomenological crystal plasticity model which included all the different slip systems of Mg (basal, prismatic and pyramidal) as well as twinning and detwinning. The simulation results (in terms of the cyclic stress-strain curve and fraction of twinned material) were used to understand the micromechanisms controlling the fatigue life of Mg alloys.
8:50 AM
PRISMS-fatigue: A General Framework for Fatigue Analysis in Polycrystalline Metals and Alloys Using the Crystal Plasticity Finite Element Method: Mohammadreza Yaghoobi1; Krzysztof S. Stopka2; Aaditya Lakshmanan1; John E. Allison1; Veera Sundararaghavan1; David L. McDowell2; 1University of Michigan; 2Georgia Institute of Technology
A novel open source framework that enables simulation-based comparisons of microstructures with regard to fatigue resistance is presented here for polycrystalline metals and alloys. The framework uses the crystal plasticity finite element software PRISMS-Plasticity as its microstructural analysis tool. This framework provides a highly efficient, scalable, easily modified, and easy-to-use ICME community platform. The performance and flexibility of this framework are demonstrated with various examples, including effects of crystallographic texture, grain morphology, strain state, free surface, and choice of FIP on the driving forces for fatigue crack formation. The results show that the multilevel parallelism scheme of PRISMS-Fatigue framework is more efficient and scalable than ABAQUS for microstructure instantiations having over one million degrees of freedom. The links between the PRISMS-Fatigue and experimental characterization techniques and virtual microstructure generators are elaborated. PRISMS-Fatigue is also linked to the information repository of Materials Commons to store and share inputs and results.
9:10 AM
Propagation of Microstructure-induced Fatigue Variability onto Stress Concentrations: Gustavo Castelluccio1; Farhan Ashraf1; 1Cranfield University
Crystal plasticity computational approaches have demonstrated good predictive capabilities to quantify fatigue crack nucleation variability in smooth samples. However, the origin of such variability is poorly understood due to the multiplicity of operative mechanisms. Moreover, the evaluation of stress concentrations usually requires full microstructure-sensitive computational analyses that are sometimes unfeasible or impractical. This work explores the mechanisms that can be attributed to the intra-granular fatigue crack response predicted by microstructure-sensitive models. Furthermore, we explore engineering approaches to propagate microstructure sensitivity computed from smooth samples onto different notch geometries. We employ a physics-based crystal plasticity model in Abaqus for aluminium to quantify fatigue indicator parameters that correlate with transgranular crack nucleation in smooth samples and notches. Our results estimate the crack evolution within grains and predict fatigue notch sensitivity.
9:30 AM
Origin of Long-range Internal Stress with Heterogeneous Dislocation Distributions: Yejun Gu1; Jaafar El-Awady1; 1Johns Hopkins University
We present a new theory based on three-dimensional discrete dislocation dynamics simulations for the sources of long-range internal stresses (LRIS) associated with heterogeneous dislocation distributions in metals, to develop an understanding of material fatigue. Earlier works (e.g. Mughrabi, Physica Status Solidi (a)104.1, 1987) attribute the LRIS to geometrically necessary dislocations (GNDs) at the interface between “hard” regions having high dislocation density and “soft” regions having low dislocation densities. However, the current theory shows that the ratio between the vacancy loop concentration and interstitial loop concentration predominantly determines the LRIS. When the “hard” region mainly consists of vacancy loops, it is in tension, while the “soft” region is in compression. The opposite is also true. This theory has been successfully applied to the cases of PSBs and dislocation cell structures.
9:50 AM
A Simplified Formula to Estimate the Size of the Cyclic Plastic Zone in Metals Containing Elastic Particles: Tito Andriollo1; Varvara Kouznetsova2; 1Technical University of Denmark; 2Eindhoven University of Technology
The applicability of the well-known Irwin’s estimate to metals containing particles becomes dubious when the plastic zone (PZ) size is comparable to the particle spacing. To shed light on this, a multiscale finite element model is used to investigate the crack tip plasticity in a cyclically loaded elastic-perfectly plastic matrix containing a distribution of spherical particles generated using X-ray tomography. Homogenization and dimensional analysis demonstrate that the PZ shape differs from what expected based on von Mises plasticity. The scaling of the PZ size with the maximum stress intensity factor squared is confirmed, but the proportionality factor is strongly affected by the particle volume fraction. Consequently, a new formula to estimate the PZ size is proposed, which holds for a wide range of material properties, spanning from metal matrix composites to cast irons, and is robust to minor particle deviations from the spherical shape.