Computational Thermodynamics and Kinetics: Thermodynamics and Alloy Design
Sponsored by: TMS: Chemistry and Physics of Materials Committee, TMS: Computational Materials Science and Engineering Committee
Program Organizers: Niaz Abdolrahim, University of Rochester; Stephen Foiles, Sandia National Laboratories; James Morris, Oak Ridge National Laboratory; Raymundo Arroyave, Texas A & M University
Monday 2:00 PM
February 27, 2017
Location: San Diego Convention Ctr
Session Chair: Amit Shyam, Oak Ridge National Laboratory; Fadi Abdeljawad, Sandia National Laboratories
2:00 PM Invited
Computational Discovery of Novel Structural and Functional Heusler Compounds: Chris Wolverton1; 1Northwestern University
Heusler compounds are being widely studied for their potential usage in spintronics, shape-memory devices, superconductors, thermoelectrics, topological insulators, etc. The crystal structure and its variants are ubiquitous, with more than 1000 Heusler compounds being reported. However, the phase space for possible Heusler compounds is orders of magnitude larger, raising the real possibility that many new, stable Heusler compounds are still awaiting discovery. We demonstrate a high-throughput computational DFT screening approach for ~200,000 potential Heusler compounds, and use this method to predict hundreds of new stable and metastable Heusler compounds, which we further examine for interesting functional properties. We highlight three distinct examples of computational discovery of Heusler compounds, demonstrating the extraordinary diversity of properties possible with this single structure type: efficient thermoelectrics, off-stoichiometric semiconducting Heuslers, and strengthening precipitates in bcc metals.
Computational Design and Optimization of Shape Memory Alloys for Solid State Cooling and Refrigeration: Brian Blankenau1; 1University of Illinois
In the search for energy-efficient refrigeration technologies, one promising avenue is a solid-state magnetic cooling approach exhibited by Heusler shape memory alloys such as Ni2MnIn and Ni2FeGa. These alloys undergo a martensitic phase transformation upon the adiabatic application of a magnetic field, and the latent heat of the transformation induces a change in temperature. This so-called giant magnetocaloric effect arises from a coupling between the structural phase transition and the magnetism. The response of these complex magnetostructural materials and the temperature change achievable is very sensitive to changes in composition. In order to accurately predict the thermodynamic properties of these alloys across the phase transition and for all alloy compositions, we propose a hybrid model consisting of a cluster expansion, spin model, and Blume-Emery-Griffiths model. In our approach all parameters are obtained from density functional theory. This will provide a computational tool for the discovery and investigation of new alloys.
2:50 PM Invited
High Temperature Aluminum Alloy Development: Computational Thermodynamics and Kinetics: Amit Shyam1; Dongwon Shin1; Shibayan Roy1; Lawrence Allard1; Yukinori Yamamoto1; James Haynes1; 1Oak Ridge National Laboratory
Steadily increasing U.S. fuel economy standards for vehicles have created an urgent societal demand for cast aluminum alloys with significantly higher temperature (300oC) capacity to enable higher efficiency passenger car engines. For the common precipitation-hardened Al-Cu alloys strengthened by Al2Cu intermetallic precipitates, the upper limit of use is related to the ability to suppress transformation of the 𝜃′ phase to the thermodynamically stable 𝜃 phase. A new approach to substantially increase the thermal stability of 𝜃′ precipitates in Al-Cu alloys will be outlined. Our approach included performing first principles calculations to determine the propensity of microalloying elements to stabilize the critical interfaces in relevant aluminum alloys. A supercell approach was applied within a supercomputing framework and the benefits of a high throughput approach to model interfacial stability will be described. It will be demonstrated that alloys that possess excellent mechanical properties up to 350oC can be developed with this approach.
3:20 PM Break
Development of a Thermodynamic Database for a Co Based Superalloy for GT Vanes to Predict the Service Induced fcc-hcp Martensitic Transformation: Erica Vacchieri1; Gabriele Cacciamani2; Giacomo Roncallo2; Alessio Costa1; 1Ansaldo Sviluppo Energia S.p.A.; 2Chemistry Department, University of Genoa
The thermodynamic calculations are very important to predict and assess material behaviour during service. In GT power plant, old generation Co based superalloys are employed as structural material of vanes because of their good conductibility, TMF resistance and repairability. The vanes after service for an F class plant show several cracks and an increased fragility of the material due to the presence of unexpected hcp phase in the alloy microstructure. An in-house database was developed, starting from literature data. The main 6 element system (C, Co, Cr, Ni, Ta and W) were considered and the 15 binaries and 9 ternaries were assessed looking at all the stable and metastable phases that can be formed in the whole range of composition of each element. The developed database predicts the fcc to hcp phase transformation at a temperature range compatible with service condition and it was validated through laboratory tests.
Thermodynamic Models for the Design of Stable Nanocrystalline Alloys: Jason Trelewicz1; Heather Murdoch2; Fadi Abdeljawad3; 1Stony Brook University; 2Army Research Laboratory; 3Sandia National Laboratories
Stability in nanocrystalline materials is dependent on grain boundary/interface characteristics, whether through a thermodynamic approach – lowering the grain boundary/interfacial energy – and/or a kinetic approach – limiting grain boundary mobility. Several models have been developed to address the conditions necessary for stability; we will briefly review the evolution of existing models and stabilization criteria, followed by a detailed analysis of selected models. We specifically focus on the Regular Nanocrystalline Solution (RNS) model and the Diffuse-Interface Phase Field model, and select binary alloys to explore with analytical, phase field, and Monte Carlo treatments. Using consistent system parameters across the three approaches, a direct comparison of the model predictions of grain boundary segregation behavior, grain size, and nanostructure stability was achieved. The similarities and differences across the models are highlighted and future directions discussed for modeling of grain boundary segregation behavior and its impact on nanocrystalline (and other) alloy design.
Surface Stability of Austenitic Stainless Steel Alloys under Pressurized Water Reactor (PWR) Conditions: Zsolt Rak1; Donald Brenner1; 1NCSU
Corrosion of steam generator tubes and reactor vessel in a PWRs are the main sources of corrosion products that deposit on fuel rods. The materials used in PWRs for reactor vessels and steam generator tubes are stainless steel alloys (304, 308, 309, and 316 in the SAE steel grades system) and nickel-chromium superalloys (INCONEL 600, 690, and 800). Using ab initio atomistic modeling that takes into account the chemical and magnetic disorder of the random alloys, the stability of low index surfaces is established. The role of magnetic frustration on Cr and Ni surface segregation of is discussed. Ab initio results are combined with experimental thermodynamic data to evaluate the stability of alloy surfaces under conditions of operating PWR (i. e. as a function of temperature, pressure and pH). The initial stages of oxidation and its effect on surface stability are evaluated.
Metropolis-Hastings Algorithm for Bayesian Uncertainty Analysis of CALPHAD Model: Thien Duong1; Pejman Honarmandi1; Raymundo Arroyave1; 1Texas A&M University
Uncertainty analysis of thermodynamic models holds an important role in the field of computational thermodynamics. In the current work, we propose the use of Metropolis-Hastings algorithm in the framework of Markov Chain Monte Carlo as a mean to assess the uncertainty of CALPHAD thermodynamic models. This algorithm is used to infer Bayesian statistics which better describes the uncertainty of the model given some prior knowledge. Such prior knowledge about the model can be achieved from previous CALPHAD thermodynamic assessments. For the demonstration of the Metropolis-Hastings algorithm, the uncertainty of the thermodynamic models previously assessed for U-Nb system is analyzed.