4th World Congress on Integrated Computational Materials Engineering (ICME 2017): Phase Field Modeling - I
Program Organizers: Paul Mason, Thermo-Calc Software Inc.; Michele Manuel, University of Florida; Alejandro Strachan, Purdue University; Ryan Glamm, Boeing Research and Technology; Georg J. Schmitz, Micress/Aachen; Amarendra Singh, IIT Kanpur; Charles Fisher, Naval Surface Warfare Center

Tuesday 8:00 AM
May 23, 2017
Room: Salon IV
Location: Ann Arbor Marriott Ypsilanti at Eagle Crest

8:00 AM  
Multiscale Simulation of α-Mg dendrite growth via 3D Phase Field Modeling and Ab-initio First Principle Calculations: Jinglian Du1; Zhipeng Guo1; Manhong Yang1; Shoumei Xiong1; 1Tsinghua University
    Based on synchrotron X-ray tomography and electron backscatterd diffraction techniques, recent studies reveal that the α-Mg dendrite exhibits a morphology of 18 primary branches in 3D, of which six grow along <11-20> on the basal plane, whereas the other twelve along <11-23> on the non-basal plane. To describe this growth behaviour and simulate the morphology of the α-Mg dendrite in 3D, an anisotropy function based on cubic harmonics was developed and coupled into a 3D phase field model previously developed by the current authors. Results showed that this anisotropy funtion together with the phase field model could perfectly describe the 18-branch dendrite morphology for the magnesium alloys. The growth tendency or orientation selection of the 18-branch morphology was further investigated by performing ab-initio first principle calculations based on the hexagonal symmetry structure. It was showed that those crystallographic planes normal to the preferred growth directions of α-Mg dendrite were characterized by higher surface energy than these of others, i.e. coinciding with the 18-branch dendritic morphology. Apart from agreement with experiment results and providing great insights in understanding dendrite growth behaviour, such multiscale computing scheme could also be employed as a standard tool for studying general pattern formation behaviours in solidification.

8:20 AM  
Macro- and micro-simulation and experiment study on microstructure and mechanical properties of squeeze casting wheel of magnesium alloy: Shan Shang1; Bin Hu2; Zhiqiang Han1; Weihua Sun3; Alan Luo3; 1Tsinghua University; 2 General Motors China Science Laboratory; 3The Ohio State University
    The macro- and micro-simulation based on a coupled thermo-mechanical simulation method using ANSYS® and phase field modeling with pressure effects were carried out for squeeze casting wheel of AT72 alloy. The mechanical properties at different positions of the wheel and under different pressures were analyzed by the macro- and micro-simulation and experimental results, and the corresponding strengthening mechanism was discussed. Firstly, the mechanical properties in spoke are better than those in rim due to higher integrity associated with more forced feeding including more liquid flow feeding and almost all of the plastic deformation feeding in spoke. Furthermore, the mechanical properties increase with pressure due to the enhanced forced feeding shown by the macro-simulation results and the more developed dendrite arms, finer dendrites and more solutes in dendrites under higher pressure indicated by the micro-simulation and experimental results. As analyzed, the mechanical properties are improved by applied pressure according to the strengthening mechanism, including strengthening associated with high integrity, fine-grain strengthening and solution strengthening.

8:40 AM  
Solidification Simulation of Fe-Cr-Ni-Mo-C Duplex Stainless Steel using CALPHAD-coupled Multi-phase-field Model with Finite Interface Dissipation: Sukeharu Nomoto1; Kazuki Mori1; Masahito Segawa2; Akinori Yamanaka2; 1ITOCHU Techno-Solutions Corporation; 2Tokyo University of Agriculture and Technology
    A multi-phase-field (MPF) model with finite interface dissipation proposed by Steinbach et al. is applied to simulate dendritic solidification in Fe-Cr-Ni-Mo-C duplex stainless steel. This MPF model does not require the equal diffusion potential assumption and can take into account strong non-equilibrium interfacial condition. We develop the MPF code for coupling with CALPHAD thermodynamic database in order to simulate two-dimensional microstructure evolutions in multi-component alloys by using TQ-interface of Thermo-Calc. MPI parallelization technique is adapted to the program code development to reduce computational elapse time. It is confirmed that the developed MPF code can give highly stabilized calculation in some different initial composition conditions of the continuous casting process. The reliability of the obtained microstructure evolutions will be qualitatively discussed by comparing with the recent experimental observation.

9:00 AM  
A Cellular-automata Model for Dynamical Recrystallization with Using the Cell-orientated Grain Boundary Velocities (P-1): Daliya Aflyatunova1; 1The University of Sheffield
     The recrystallization and the grain growth in metals strongly influences on the microstructure and final properties, but it is difficult to study by experimental and statistical methods. Last decade CA technique was successfully applied [1] for recrystallization and grain growth [2] using cellular automata. However the used distance variable which corresponds to the grain boundary velocity was not considered as orientation depended and was the same for all direction of moving grain boundary interface. The current work suggests an extended CA technique which includes into consideration of the grain boundary velocity for each of the neighboring cell, thereby changing switching probability depending on the direction of the grain boundary movement. The results showed that the method is robust and is in agreement with the previously developed models, nevertheless is more physically correct. 1. G. Kugler, R. Turk. Modeling the dynamic recrystallization under multi-stage hot deformation. Acta Materialia, 2004.2. Ye. Vertyagina, M. Mahfouf, and X. Xu. 3d modelling of ferrite and austenite grain coarsening using real-valued cellular automata based on transition function. Journal of Materials Science, 48(16): 5517- 5527, August 2013.

9:20 AM  
Effect of Mn Diffusivity on the Austenite-to-ferrite Transformation Behavior in Fe-C-Mn Ternary Alloy: A Multi-phase-field Study: Takahiko Kohtake1; Akinori Yamanaka2; Yoshihiro Suwa1; 1Nippon Steel & Sumitomo Metal Corporation; 2Tokyo University of Agriculture and Technology
    The austenite-to-ferrite (γ-to-α) transformation in a Fe-C-Mn alloy is one of the important solid-solid phase transformations for steel-making processes. However, because there is a large difference in diffusion coefficient between Mn and C atoms in the alloy, the transformation behavior strongly depends on the chemical composition and temperature. Recently, the multi-phase-field (MPF) method has attracted much attention as a tool for understanding the transformation behavior. However, to the best of the authors’ knowledge, few quantitative investigations of the interfacial migration and the diffusion of the alloying elements during the γ-to-α transformation by the MPF method have been reported. Therefore, we have investigated the Mn diffusivity’s effect on the γ-to-α transformation by the one-dimensional MPF simulation coupled with the thermodynamic database based on the CALPHAD method. In this study, in order to further clarify the Mn diffusivity’s effect on the transformation kinetics, we perform the MPF simulation of the γ-to-α transformation using the various ratio of the Mn diffusion coefficient to that of C from 0 to 1 under the condition of the fixed C diffusion coefficient. The results clarify that when the Mn diffusivity is sufficiently high, Mn atoms are partitioned between γ and α during the transformation. On the other hand, when the Mn diffusivity is low, the γ-to-α transformation proceeds without Mn partitioning. Furthermore, it is interestingly found that, for a case of intermediate Mn diffusion coefficient, the γ-to-α transformation stagnates after the transformation without Mn partitioning. In this presentation, several factors for this stagnation will be discussed.

9:40 AM  Cancelled
Full Field Simulation of the Recrystallization during Annealing of AZ31 Mg Alloy Taking into Account Plastic Deformation and Twinning: Alvaro Ridruejo1; Raul Sanchez-Martin2; Efim Borukhovich3; Reza Darvishi Kamachali4; Oleg Shchyglo4; Ingo Steinbach4; Javier Segurado1; Javier LLorca2; 1Technical University of Madrid; 2Imdea-Materials Institute; 3KTH Royal Institute of Technology; 4ICAMS, Ruhr-Universität Bochum
    Microstructural optimization of metallic alloys is carried out by means of thermo-mechanical processes. The microstructural changes are controlled by a number of factors that include the details of the initial microstructure and the driving forces and kinetics processes associated with temperature (diffusion, phase transitions) as well as mechanical deformation (elastic strain energy, twinning, etc.). In this work, a full-field approach capable of taking into account these phenomena is presented. In this approach, the phase field model of microstructure evolution is coupled with a dislocation-based crystal plasticity model to account for the mechanical effects. In addition to conventional dislocation glide, the model is able to account for plastic deformation by mechanical twinning, which is simulated as a mechanically-driven solid-state phase transformation. The full field model is used to simulate the effect of prior mechanical deformation on the recrystallization and texture evolution during high temperature annealing of an AZ31 Mg alloy sheet. Our findings suggest that the inhomogeneous distribution of energy stored within the grains after deformation leads to a notable reduction of the initial basal texture during subsequent recrystallization processes. These results are validated by experiments where the texture evolution during annealing of an AZ31 magnesium sheet, previously deformed by in-plane compression, was measured. Finally, it is confirmed that the full field model can be used to simulate the complete sequence of deformation and recrystallization steps during sheet metal manufacturing.

10:00 AM Break

10:30 AM  
Using the PRISMS-PF Phase Field Code in an ICME Framework to Examine Precipitates in Mg-Nd Alloys: Stephen DeWitt1; Shiva Rudraraju1; Katsuyo Thornton1; Anton Van der Ven2; John Allison1; 1University of Michigan - Ann Arbor; 2University of California - Santa Barbara
    We introduce PRISMS-PF, a new, advanced open source phase field modeling code, and its application to simulate precipitate microstructures. PRISMS-PF provides a simple, flexible interface for massively parallel phase field simulations. Leveraging the deal.II finite element library and its matrix-free framework, improved performance over finite difference codes and traditional finite element codes is demonstrated. Using PRISMS-PF, we investigate precipitate formation in magnesium-rare earth alloys, which have garnered substantial interest as a structural material for automotive and aerospace applications. In one magnesium-rare earth alloy, magnesium-neodymium, recent experimental and first-principles work indicates that the alloy’s high strength is due to β’’’ precipitates. We conducted phase field simulations of these precipitates, examining the morphology of isolated precipitates and the evolution of interacting precipitates, with inputs determined from first-principles calculations. Simulations of isolated precipitates predict the habit plane seen experimentally, (100), but predict that the precipitates are longer in the [010] direction than the [001] direction, in contrast to observations. Systematic variation of the model parameters within the first principles calculation uncertainty indicates that the discrepancy may not be fully explained by uncertainty in the inputs. Simulations of small clusters of precipitates demonstrate that effect of interactions between precipitates also affects their aspect ratio. Accounting for these interactions, parameters within the first-principles uncertainty and consistent with observations are determined and used in a simulation of the nucleation and growth of many β’’’ precipitates. The predicted volume fraction, number density, and shape distribution as a function of time are compared to experimental observations.

10:50 AM  
Phase Field Modeling of Precipitate-Free Zones Around Grain Boundaries: David Montiel1; Jason Luce1; Katsuyo Thornton1; 1University of Michigan
    We employ a coupled conserved-nonconserved phase field model with a stochastic nucleation method to simulate the nucleation and subsequent growth of precipitates. This model was implemented into the PRISMS-PF code, an open-source finite element simulation code developed within the PRredictive Integrated Structural Materials Science (PRISMS) Center at the University of Michigan. Using this model, we study the formation of precipitate free zones (PFZs) around grain boundaries during aging. In particular, we focus on the effect of preferential nucleation along a grain boundary and the resulting solute depletion on the formation of the PFZ. In addition, we study the effect of solute diffusivity and of the initial solute supersaturation. We compare our results to experiments including PFZs formed during aging of Mg-RE alloys. We also explore a simplified semi-analytical model that uses the Zener’s approximation to describe the solute concentration profile around precipitates.

11:10 AM  
Aspects of Microstructure Simulation in ICME: A Virtual Process Chain for Diffusion Brazing of Alloy 24 : Bernd Böttger1; 1Access eV
     Different aspects of microstructure simulation within an ICME setting are discussed with focus on the multiphase-field software MICRESS®. In the first step, microstructure modelling of alloys often needs related input from external sources: Concentration distributions may be read in from experiments, prior simulation runs or other software tools. Particularly in the case of multicomponent systems, thermodynamic data typically is provided by Computational Thermodynamics tools using Calphad databases. During simulation, the microstructure tool must be able to treat phase transformations like melting, solidification or precipitation, different types of diffusion of chemical elements (chemical diffusion, cross diffusion, interface diffusion, infinite diffusion, far-field diffusion, etc.) as well as redistribution of the elements between phases (including solute trapping, para-equilibrium, LENP). Finally, concentration distributions or derived quantities like average compositions or a segregation index can be stored using standardized formats (e.g. HDF5 or VTK) and be handed over to other tools for simulation of subsequent process steps.During the talk, a virtual process chain for diffusion brazing of the Ni-based superalloy M247 is presented which spans the fields between the sub-µ and the macro-scale, between thermodynamic data and properties and between casting and the product life time. The challenges are discussed and the importance of a correct and complete handling of the relevant microstructural quantities in ICME is illustrated.

11:30 AM  
Phase-field Modeling of θ’ Precipitation Kinetics in W319 Alloys: Yanzhou Ji1; Bita Ghaffari2; Mei Li2; Long-Qing Chen3; 1The Pennsylvania State University, University Park; 2Ford Motor Company; 3The Pennsylvania State University, University Park
    Understanding and predicting the morphology, kinetics and hardening effects of θ’ precipitates are critical in improving the mechanical properties of Al-Cu-based alloys through controlling the temperature and duration of any isothermal aging process. In this work, we present a comprehensive phase-field framework for simulating the kinetics of θ’ precipitates in W319 alloys, integrating the thermodynamic and diffusion mobility databases of the system, the key anisotropic energy contributions from literature and first-principles calculations, as well as a nucleation model based on the classical nucleation theory. By systematically performing phase-field simulations, we optimize the model parameters to obtain the best possible match to the peak number density, average diameters and volume fractions of θ’ precipitates from experimental measurements at 190°C, 230°C and 260°C. With these parameters available, the phase-field simulations can be performed at other aging temperatures. The possible extensions of the current phase-field model for more accurate prediction of the precipitate behaviors in W319 alloys will also be discussed.

11:50 AM  
Phase-field Modeling of Cu-Mn-Ni Rich Precipitate Behavior in Reactor Pressure Vessel (RPV) Steel: Kunok Chang1; Junhyun Kwon1; 1Korea Atomic Energy Research Institute
    The late blooming phase (late-stage precipitate) is a limiting factor for the long term operation of the nuclear power plants. The precipitated gamma precipitate cause the late-stage hardening and it reduced the fracture toughness of the RPV steel. The elements consist of the gamma precipitate is Cu, Mn, Ni, Si and other minor alloying elements and we choose three major components (Cu, Mn, Ni) for simplicity. We simulated the behavior of the gamma precipitate (Cu-Mn-Ni rich) in the alpha-Fe matrix. We investigated a role of the dislocation loop on the behavior of the gamma precipitate in the 3-D system. To conduct the simulations, we adopted the thermodynamic database of Fe-Cu-Mn-Ni quaternary system assessed by Koyama et al.. It is demonstrated that the interstitial loop reduce the stability of the gamma precipitate and also it alters the morphology of the precipitate in 3D space.

12:10 PM Break