Accelerated Discovery and Insertion of Next Generation Structural Materials: Accelerated Discovery of Structural Alloys
Sponsored by: TMS Materials Processing and Manufacturing Division, TMS: Phase Transformations Committee, TMS: Integrated Computational Materials Engineering Committee
Program Organizers: Soumya Nag, Oak Ridge National Laboratory; Andrew Bobel, General Motors Corporation; Bharat Gwalani, North Carolina State Universtiy; Jonah Klemm-Toole, Colorado School of Mines; Antonio Ramirez, Ohio State University; Matthew Steiner, University of Cincinnati
Wednesday 8:30 AM
March 22, 2023
Room: Sapphire M
Location: Hilton
Session Chair: Soumya Nag, ORNL; Bharat Gwalani, PNNL; Antonio Ramirez, Ohio State University
8:30 AM
Design of a Compact Morphology Cobalt-based Superalloy for Additive Manufacturing: Krista Biggs1; Brandon Snow1; Benjamin Graybill1; Christopher Kiehl1; Gregory Olson1; 1Massachusetts Institute of Technology
There is significant potential for the use of additively manufactured cobalt-based superalloys in turbine engine blades because they are capable of higher operating temperatures than their nickel-based counterparts. We have designed a novel cobalt-based superalloy to operate at 900 ⁰C (approximately 200 ⁰C hotter than printable nickel-based superalloys), inspired by Cozar and Pineau’s nickel-based compact morphology alloy. The design incorporates multiple reinforcement phases for improved printability and resistance to coarsening compared to traditional γ-γ’ microstructures. The structural and elastic considerations for selecting appropriate phases are discussed, considering the shear field which surrounds a precipitating γ’ particle due to its lattice mismatch with the matrix phase. Additionally, the relative phase fractions of the reinforcements are tuned using finite element analysis to minimize the strain energy of the system resulting from the precipitates’ formation. The design’s prototype and the kinetics of forming its resulting microstructure are then assessed.
8:50 AM
High-throughput Prediction of Fracture and Brittle to Ductile Transition in Tungsten using Variable Temperature Nanoindentation: Kevin Schmalbach1; Radhika Laxminarayana1; Douglas Stauffer1; William Gerberich2; Nathan Mara2; 1Bruker Nano; 2University of Minnesota
Methods to predict material fracture frequently rely on large experimental datasets tuned to the properties of one material or are based on computationally expensive modeling. Development of analytical models with easily measured physically meaningful parameters are key to alleviating bottlenecks in new materials development. Here, I describe the use of nanoindentation strain rate jump tests, applied at low temperature (-100 °C) and high temperature (50-300 °C), to measure the effective stress and activation volume as a function of temperature. These activation parameters, in combination with an analytical model for the strain energy release rate, accurately predict the brittle-ductile transition temperature along particular fracture systems in single crystal tungsten. Activation parameters measured from both nanoindentation and bulk tension of single crystal tungsten accurately predict the fracture toughness and brittle-ductile transition in macroscale tungsten single crystals.
9:10 AM
Computational Design of an Ultra-strong High-entropy Alloy: Mauricio Ponga1; 1The University of British Columbia
We present a combined experimental and computational investigation of the mechanical properties of a CoCrFe0.75NiMo0.3Nb0.125 high-entropy alloy additively manufactured via cold spray. Under compression, the sprayed alloy exhibits extraordinary mechanical properties, reaching yield stress of ~1745 MPa, ultimate stress of ~2622 MPa, and a maximum strain at failure of ~9%. These exceptional mechanical properties are the result of four independent hardening mechanisms. Using a novel design condition, an optimal solid solution and precipitation-strengthening alloy are obtained from ab-initio simulations. We show how the microstructure can be tailored to develop optimal mechanical strength using additive manufacturing. These subtle atomic and microstructural features result in outstanding experimentally evaluated yield and ultimate stresses compared to other high-entropy alloys with similar compositions.
9:30 AM
Computational Design of High Entropy Alloy Hardmetals: Joshua Berry1; Robert Snell1; Magnus Anderson1; Olivier Messe2; Iain Todd1; Katerina Christofidou1; 1University of Sheffield; 2Oerlikon AM Europe GmbH
High Entropy Alloys (HEAs) present an opportunity for the design and development of new wear resistant hardmetals, to replace the conventional WC-Co cemented carbides, used in demanding metal forming applications. However, the vast compositional space occupied by HEAs, results in unguided experimental searches being unfeasible. Here, a random forest machine learning architecture, in conjunction with the CALPHAD method, is trained from experimental HEA databases, to perform high-throughput phase formation and hardness predictions. Nine of the hardest predicted FCC solid solution forming HEA compositions from the machine learning model were selected and fabricated. Mechanical and thermal assessments of these selected alloys will demonstrate their potential suitability for WC-Co replacement, while simultaneously enabling comparison and verification of the machine learning methodology and providing further data for future model development. This work was supported by Oerlikon AM Europe GmbH, Engineering and Physical Sciences Research Council UK [EP/S022635/1] and Science Foundation Ireland [18/EPSRC-CDT/3584].
9:50 AM
Machine Learning-assisted Discovery of Novel High Temperature Ni-rich NiTiHfZr Multi-component Shape Memory Alloys: John Broucek1; Daniel Salas1; William Trehern1; Ibrahim Karaman1; 1Texas A&M University
Machine learning (ML) and artificial intelligence (AI) are emerging in material science and engineering as methods to discover novel materials for various applications. AI/ML methods can survey small and unexplored regions of a material design space with desired shape memory characteristics, increasing the rate of material discovery by limiting the number of experiments needed to demonstrate the properties of interest. In this work, we utilized ML to predict the martensitic transformation characteristics of Ni-rich NiTiHfZr multi-component shape memory alloys intended for high-temperature aerospace applications. These alloys are characterized by having multiple principal elements and typically demonstrate martensitic transformation between B2 austenite and B19’ martensite occurring at temperatures as high as 600°C. Experimental examination of thermal properties such as transformation temperatures, and enthalpy of transformation was conducted to validate predictions through direct comparison. Active learning was used during data collection to further enhance the accuracy of predictions.
10:10 AM Break
10:30 AM Cancelled
Data Efficient Bayesian ICME Workflow for the Design of Targeted Mechanical Properties of Structural Materials: Anssi Laukkanen1; Tatu Pinomaa1; Matti Lindroos1; Sicong Ren1; Abhishek Biswas1; Napat Vajragupta1; Tom Andersson1; Tomi Suhonen1; 1VTT Technical Research Center of Finland
Case specific design of targeted properties of structural materials across material processing workflows can be greatly improved by way of Bayesian machine and deep learning methods. The role of "traditional" ICME tools is crucial in integrating to experiments, characterization, and in providing access to baseline synthetic data pools. We demonstrate such a workflow for high strength steels and high entropy alloys as use cases with respect to their micromechanical properties related to strengthening, and especially, cyclic response to fatigue. We integrate an ICME workflow to active learning and Bayesian approaches utilizing machine learning and recurrent deep learning to provide a data efficient workflow for either analyzing or obtaining targeted properties by way of engineering material microstructure and chemistry.
10:50 AM
Accelerated Discovery of Novel Titanium Alloys using High-throughput Manufacturing, Characterization and Testing: Dian Li1; Sydney Fields1; Yufeng Zheng1; 1University of Nevada-Reno
Due to the combination of great properties, such as high strength, low density and excellent corrosion resistance, titanium alloys have become a good candidate for structural materials in aerospace, chemical and bio-medical industries. However, the relatively high manufacturing cost still limits the wide application of titanium alloys. Thus, it still requires significant effort in designing novel low-cost titanium alloys with great performance. In this work, the inter-relationship between composition, microstructure and properties in Ti-X (X=Mo or Fe) alloys were rapidly mapping using the high-throughput combinatorial methods. The compositionally graded sample manufactured using different techniques were firstly studied using SEM EDS. Nanostructures including omega phase, O’ phase and incommensurate domains were characterized using diffraction contrast TEM. The properties, such as hardness and modulus, for different compositions were evaluated using high-throughput nanoindentation technique. This work is supported by the National Science Foundation, grant CMMI-2122272 and DMR-2145844.
11:10 AM
A Diffusion Couple Approach to β-Ti Alloy Development: Evaluating the Oxidation Performance of Ti-Fe-X+ Alloys: Paraic O'Kelly1; Alexander Knowles1; 1University of Birmingham
Materials discovery has greatly benefited from ICME utilising thermodynamic databases, however the predictive capability of ICME can be poor when extrapolating to new materials systems away from current understanding, precisely where the largest potential innovations may lie. Recently in the Ti-Fe system, ordered-bcc precipitate-reinforced refractory-metal-based alloys has demonstrated the possibility of β –β’ bcc-superalloys as a new class of high temperature materials. However, the introduction of scale-forming elements is required. The current research applies a diffusion couple approach to generate compositional and microstructural gradients in the Ti-Fe-Al-Cr composition space. As such, the oxidation characteristics of the graded alloy are assessed which offers key insights into composition-microstructure-property relations and a move away from the iterative ingot-by-ingot approach. The experimental results enable a significant down selection of promising discrete quaternary alloys for further study to optimise chemistry, microstructure, environmental and mechanical response.
11:30 AM
Using Machine Intuitive Learning to Predict Advanced Steel Properties: Krista Limmer1; Andrew Garza1; Heather Murdoch1; Benjamin Szajewski1; Daniel Field1; Christopher Rinderspacher1; Levi McClenny1; Mulugeta Haile1; 1DEVCOM Army Research Laboratory
Recent advances in data science and high-throughput materials simulations are being evaluated to accelerate advanced steel alloy development. Here we use machine learning (ML) to take advantage of the large amounts of historic data available for martensitic steels. A series of models and diagrams using varying amounts of data are used to develop predictive ML models. Multiple approaches are used to assess the degree of information required to predict toughness as a function of composition and processing parameters. The first approach directly minimizes the composition and processing variables using Gaussian process regression. The latter approaches incorporate various differing neural networks, such as multilayer perceptron and recurrent neural networks, to predict toughness based on intermediate variables. These intermediate variables are synthetic microstructures and thermodynamic properties generated using high-throughput CALPHAD simulations.
11:50 AM
Rapid Characterisation of Active Slip Systems in Titanium Ordered-bcc Compounds using an Algorithm for Automated Indentation Slip Trace Analysis.: Vincent Gagneur1; Alexander Knowles1; 1University of Birmingham
Increasing the operating temperature of aerospace gas turbines is a key means to improve their fuel efficiency, which is currently limited by the capability of materials employed. Our work focusses on new beta-titanium ‘bcc-superalloys’, harnessing the high melting points of Mo and Nb, paired with the low-density of Ti. Inspired by widely used Nickel superalloys, these bcc-superalloys exploit a combination of a β matrix with ordered-bcc β’ precipitates. However, many bcc-superalloy systems are too brittle for commercial applications, with one reason being the absence of slip transfer between the bcc Ti matrix and the ordered-bcc TiFe precipitates, due partly to different slip directions being favoured, ½[111] vs [100] respectively. In this work, attempts to modify the favoured slip direction in TiFe through alloying are presented. Various B2 bulk compositions & gradient samples are characterised using micro-indentation slip trace analysis assisted by a novel high-throughput automated slip trace detection algorithm.