Proceedings of the 3rd World Congress on Integrated Computational Materials Engineering (ICME 2015)

Steel has been used since long in the past, but there still remain many unexplored possibilities. In order to draw out this latent potential, the concept of materials integration is garnering attention in view of discontinuously improving the function and performance of steel products, while reducing the necessary development period. In this paper, considering the utilization of materials integration, the evolution of the microstructure during plastic deformation and phase transformation as well as controlling hydrogen embrittlement are discussed in terms of heterogeneity. In addition, a brief outlook on the future of materials integration is presented.

ICME will play a major role in reducing the lead time in development of a new product or component. One of the areas where ICME is likely to play a crucial role is process scale-up of mill products. Process scale-up of a mill product from laboratory to production stage is largely done through hit-and-trial and is a non-trivial exercise. It involves plant level trials which are expensive and time consuming. Use of ICME can significantly narrow down the design search space thereby reducing need for experimentation or plant trial, which in turn will lead to bringing down the cost and time of development. However many challenges need to be addressed to realize the full potential of ICME at an industrial scale. Manufacturing any product/ component involves a host of unit operations and the properties of the end product are intrinsically linked with final as well as intermediate processing steps. To link the material-processing-structure-performance matrix, there is a need to enhance models across various unit operations through multi-scale/multi-phase modelling and integration of models at various length scales. This allows for the information flow across various unit operations and thereby ensures horizontal integration of each process to simulate the entire manufacturing chain. This step is crucial in designing set points and quantifying the influence of various unit operations on end product performance. In this paper, we illustrate the vertical-horizontal integration of models through an example

The complete manufacturing process of cold drawn welded steel tubing was simulated using a history dependent internal state variable plasticity-damage model. In order to implement the history dependent model, the Internal State Variables (ISV) Plasticity-Damage (DMG) model that has been developed at Sandia National Labs and Mississippi State University was employed for a low carbon steel AISI 1010, which is the alloy used by the partnering tube manufacturing company. Even though the model is performed at the structural scale, the history variables (internal state variables) which are driven by microstructural changes—e.g., dislocation density and grain size—were tracked through the whole sequences, and the history model captured real thermal and mechanical behaviors of the target material. Once the internal state variables were extracted from the simulation, the model showed critical deformation histories that one could not notice with the naked eye, and must therefore be considered for future design of the tubing process.

Ab-initio density functional theory calculations have been performed to determine the effect of solutes including Cr, Ni, Mo, and Mn on the boundary energy for a variety of coincident site lattice grain boundaries of FCC iron (austenite). The boundaries investigated were of tilt character and both symmetric and asymmetric boundary planes were investigated. Boundary energies were determined for boundaries in pure Fe and for boundaries with single solute atoms at a variety of sites in each boundary. The results are compared to Arrhenius type equations developed from experimental work in the literature and used to hypothesize a mechanistic model for the effects of solutes on boundary mobility based upon the thermodynamic and kinetic effects of solutes at austenite grain boundaries. The predictive capabilities of atomic configurations and bond structures are evaluated and areas for future work are identified. This work provides a new framework for understanding the effects of solutes on atomic scale grain boundary energies and solute drag effects on boundaries.

In the design of high strength steels, bainite structures gain importance because of their excellent balance of strength and toughness. However, bainite still remains the least understood austenite decomposition reaction. Following the philosophy of Integrated Computational Materials Engineering (ICME), we combine various simulation and advanced characterization approaches at different length scales in order to improve the scientific understanding of this solid phase transformation. In the present work, bainite transformation in 100Cr6 steel with nano-sized cementite (θ) precipitation within bainitic ferrite (α_B) is analyzed. The bainite transformation was introduced and investigated by TEM, atom probe tomography (APT), phase-field simulations and ab initio calculations. APT shows that in lower bainite isothermally held at 260 °C both ε and θ precipitate adopting plate-like shapes and precipitate under para-equilibrium mode. With the help of atom probe concentration data input and para-equilibrium phase diagram calculation using ThermoCalc software, isothermal bainite formation was simulated by means of phase field approach. To simulate the very fine cementite, ultra-small grid spacing, i.e. 2 nm, was applied in the simulation. A faceted model has been coupled in the simulation for both bainitic ferrite (α_B) and cementite (θ) growth. APT showed the high amount of carbon trapped within bainitic ferrite. It is of significant importance to take into account both carbon partitioning and carbon trapping within bainite ferrite in bainite transformation simulations.

This contribution reports the results of the collaborative research unit SFB 761 “Steel — ab initio”, a cooperative project between RWTH Aachen University and the Max-Planck-Institute for Iron Research in Düsseldorf (MPIE) financed by the German Research Foundation (DFG). For the first time, in a structural measure it is exploited how ab initio approaches may lead to a detailed understanding and thus to a specific improvement of the design process for a structural steel. Here, the challenge lies in the combination of Abstract natural science theories with rather engineering-like established concepts. Aiming at the technological target of the development of new type of structural materials based Fe-Mn-C and Fe-Mn-Al-C alloys, the combination of ab initio and engineering methods is new, but could be followed quite successfully. Three major topics are treated in this research unit: a) development of a new method for material- and process-development based on ab initio calculations; b) design of a new class of structural materials with extraordinary property combinations; c) acceleration of development time and reduction of experimental efforts and complexity for material- and process-development. In the present work, the general concept of new material design is presented together with some exemplarily results showing the advantage of this combined scale bridging approach.

Magnesium alloys are one of the lightest structural materials. The mechanical properties of magnesium alloys are often affected by the dendritic microstructures. With hexagonal close-packed structure, the preferred growth direction of α-Mg dendrite was believed to be <math display='block'> <mrow> <mrow><mo>&#x2329;</mo> <mrow> <mn>11</mn><mover accent='true'> <mn>2</mn> <mo>&#x00AF;</mo> </mover> <mn>0</mn> </mrow> <mo>&#x0232A;</mo></mrow> </mrow> </math>$$\left\langle {11\bar 20} \right\rangle$$, and different 3-D growth models were proposed. However, these growth models for α-Mg dendrites were divergent and not yet generally accepted. Recently, Mg-20 wt.% Y alloy was studied by synchrotron X-ray tomography to establish the three-dimensional dendritic morphology of magnesium alloy. According to the reconstructed results, the α-Mg dendrites grew along eighteen branches with six branches along <math display='block'> <mrow> <mrow><mo>&#x2329;</mo> <mrow> <mn>11</mn><mover accent='true'> <mn>2</mn> <mo>&#x00AF;</mo> </mover> <mn>0</mn> </mrow> <mo>&#x0232A;</mo></mrow> </mrow> </math>$$\left\langle {11\bar 20} \right\rangle$$ in the {0001} basal plane, and twelve along <math display='block'> <mrow> <mrow><mo>&#x2329;</mo> <mrow> <mn>11</mn><mover accent='true'> <mn>2</mn> <mo>&#x00AF;</mo> </mover> <mn>3</mn> </mrow> <mo>&#x0232A;</mo></mrow> </mrow> </math>$$\left\langle {11\bar 23} \right\rangle$$ in the non-basal plane. Based on the three-dimensional morphology of α-Mg (Y) dendrite, a cellular automaton model was developed to simulate the dendritic growth of magnesium alloy. Both the experimental and simulation results offer a deep insight in understanding the dendritic growth evolution of magnesium alloys during solidification.

The phase field dislocation model has been used to compute and simulate interactions between basal α-type dislocations and coherent β₁ precipitate plates in a Mg-3wt.%Nd alloy that is strengthened exclusively by the β₁ plates. The computed increments of the critical resolved shear stress (ΔCRSS) for samples aged for 10 hours at 523 K agree well with those calculated from the existing strengthening equation for plate-shaped particles. The phase field simulations further indicate that the ΔCRSS value increases with an increase in plate aspect ratio and number density, and that the change of ΔCRSS is not sensitive to the variation of β₁ plate diameter distribution when the average diameter of β₁ plates is fixed. When the volume fraction of β₁ plates is constant, the ΔCRSS value for a random spatial distribution of the β₁ plates is approximately 0.78 times of that for a regular spatial distribution.

Lightweight metal castings are increasingly used in critical structural applications which are often subjected to cyclic loading during service. Fatigue property of the lightweight metal castings has become a critical design criterion. Fatigue performance of lightweight metal castings strongly depends upon the presence of casting flaws and characteristics of microstructural constituents. The existence of casting flaws significantly reduces fatigue crack initiation life. In the absence of casting flaws, however, crack initiation occurs at the fatigue-sensitive multi-scale microstructural constituents. This paper discusses the recently-developed multi-scale fatigue (MSF) models and their applications in lightweight aluminum and magnesium castings.

The actual volume fraction of the solid in semi-solid metal processing is critical to achieving the most effective semi-solid metal processing. The semi-solid slurry with different solid fractions shows different ability to fill molds. To determine suitable processing conditions for semi-solid sterling silver (92.5 wt.%), thermal analysis (TA), experimental verification in actual investment casting machine as well as thermodynamics models were performed. Classical Gibbs free energy calculation and Scheil-Gulliver modeling were used to estimate the relationship between solid fraction and temperature. In order to obtain the relationship between temperature and solid fraction, comparisons were made amongst different techniques. The investigated 925 silver alloy was heated to 950 °C, and gas bubbling apparatus was used to introduce argon gas into the melt. Cooling curves for the metal were recorded with two thermocouples, one at the center of the melt volume and one beside the containing crucible wall. Different cooling curves and temperature profiles are presented from various tests to compare solid fractions. It was found that the latent heat of fusion strongly affected the solid fraction in addition to thermodynamic contribution. To adequately express the solid fraction relationship, the thermodynamics modeling needed modifications that took into account the cooling effects as well as temperature gradient in the crucible.

In this paper, the validity of some formulations allowing the determination of the fracture energy of PAN/phenolic-based carbon/carbon composites has been verified. The G _IIC is chosen as fracture characterizing parameter which is experimentally determined by considering a multiplying form that is numerically evaluated using a finite element method. The numerical results are compared to the experimental data, and a good agreement has been observed. The specimen geometry was used to determine the mode II delamination fracture energy.

Recent advances in the simulation of the quench, cold-work and machining processes for large aluminum forgings are opening the way for a new paradigm in the design, manufacture and sustainment of aircraft structures. The use of large forgings permits the unitization of smaller parts (brackets, fittings, lugs, etc.) with primary structural components like spars and bulkheads. This is being done in order to reduce part count, which in turn leads to significant reductions in manufacturing cost. Unitization can also translate into weight reduction / avoidance when comparing against built-up structure, but it raises a number of issues for structural durability and damage tolerance, notably reduced repair / replace capability and reduced crack arrest capability. The viability of the unitization concept is dependent not only on the availability of material systems that retain their mechanical properties in very thick sections, but also on the designer’s ability to retain durability and damage tolerance, and to understand and mitigate the effects of residual stresses.

Investigation of diffusion process in Ni-based alloy is a problem of high relevance in the area of understanding corrosion behavior. We explored the use of a combined approach consisting of density functional theory to compute migration barriers and Kinetic Monte Carlo method to evaluate hard to measure tracer diffusion coefficients. A major challenge in the implementation is the need to find one by one the rate constants for each diffusion process that can occur in the alloy. To overcome this, a pair interaction model was utilized to evaluate the influence of local environment on the kinetic parameters. Previous application of this approach yielded self-diffusion coefficients in pure Ni and the tracer diffusivity of dilute Al in the Ni host that are in good agreement with available experiments. The method was extended to examine oxygen diffusivity in pure Ni and the results show good agreement with values obtained using electrochemical and potentiometric techniques. The presence of Al was found to have a dragging effect on the mobility of oxygen.

The undercooled supersaturated aluminium alloy EN AW-6082 has been investigated during in-situ quenching experiments by means of differential scanning calorimetry and thermomechanical analysis. Calorimetry resulted in continuous time-temperature-precipitation diagrams and thermomechanical analysis in stress/strain curves. In this work, the above results have been correlated. Supersaturation of the undercooled aluminium alloy has been determined from calorimetric data. Next, a solid solution strengthening model has been used to calculate yield strength. The model has further been extended by a temperature dependent term considering other strengthening mechanisms. Yield strength has been computed for a broad range of quenching rates and quenching temperatures. Results have been compared with experimental data from thermomechanical analysis and found to correlate well.

Stainless steel 440C with high content of C and Cr is the desired candidate in many engineering component for bearing purpose due to the extremely high yield strength and resistance to corrosion and erosion. However, high content of C and Cr usually results in carbides precipitates during cooling process. In this work, deformation mechanisms and the interplay between martensitic matrix and carbides precipices are explored using crystal plasticity finite element method (CPFEM). Two deformation stages are clearly revealed correlated to the yielding of matrix and precipitate respectively. When the “softer” phase, matrix is approaching the yielding point, lattice strains of matrix cease increasing and experience the stable stage, while precipitates carry more stresses. When sample is further deformed and precipitates are yielding, lattice strains in matrix exhibit elastic relaxation.

In molecular dynamics (MD) simulation, atomic interaction is characterized by the interatomic potential energy as the input of simulation models. The interatomic potentials are derived experimentally or from first-principles calculations. Therefore they are inherently imprecise because of the measurement or model-form errors. In this work, a reliable molecular dynamics (R-MD) mechanism is developed to extend the predictive capability of MD given the input uncertainty. In R-MD, the locations and velocities of particles are not assumed to be precisely known as in traditional MD. Instead, they are represented as intervals in order to capture the input uncertainty associated with the atomistic model. The advantage of the new mechanism is the significant reduction of computational cost from traditional sensitivity analysis when assessing the effects of input uncertainty. A formalism of generalized interval is incorporated in R-MD, as an intrusive uncertainty quantification method, to model the propagation of uncertainty during the simulation. Error generating functions associated interatomic potentials are developed to capture the bounds of input variations.

Atomistic scale simulation can predict state transition processes such as adsorption, diffusion, and reaction. The challenge of the accurate prediction is to obtain a global view of many local minima and saddle points on the potential energy surface (PES) of a material system. The transition conditions are determined by the saddle points on the PES with the minimum energy barrier between local minima. In this paper, a new algorithm is developed to exhaustively search local minima and saddle points within a region of PES in order to provide a global view of the energy landscape. Unlike the existing saddle point search methods, the algorithm represents a transition path by a parametric Bézier curve with control points. It uses multiple groups of such curves, each of which represents a multistage transition path. During the searching process, each group of curves communicates with others to maintain cohesion and avoid collision based on a collective potential model. The algorithm is integrated with density functional theory calculation and demonstrated by diffusion of hydrogen atoms in the FeTiH system.

A method based on the kinetics of crystal growth has been developed and applied to the computation of 3D microstrucuture in polycrystaline austenite-matensite steels. The detailed crystallography of the transformation and the effect of austenite grain size on the martensite-start temperature are employed to simulate a realistic martensitic microstrucuture. The interaction energy based on the plastic work model of Patel and Cohen is taken into account to model the variant selection under external system of applied stresses. The method has been integrated to the homogeneous deformation theory for computation of microstructure evolution in thermomechanical processing.

Our project aims at providing the materials engineering fraternity with a simple and effective interface using ipython to operate Quantum ESPRESSO (QE), an open source code for materials simulation. QE is a first principles code using density functional theory, plane waves and pseudo potentials; it has ability to predict material properties. Ipython notebook interface uses the scope of the following libraries; Atomic Simulation Environment (ASE), matplotlib, scipy, numpy, pyspglib, elastic and newly developed library: QE-nipy-advanced to predict the properties. QE-nipy-advanced is the latest version of QE-nipy. The latest version incorporates features that can take care of all the input parameters supported by PWscf and PHonon packages of Quantum ESPRESSO. Thermo-mechanical properties of some nuclear materials with different magnetic and metallic behavior has been studied using the QE-nipy-advanced, but in here we demonstrate the thermo-mechanical properties of non-magnetic insulator like silicon carbide and thoria, which are materials for future nuclear reactor applications.

The importance of microstructure simulations in ICME settings is discussed with respect to their added value provided to macroscopic process simulations and their contribution to the prediction of materials properties. Their role in integrating the scales from component/process scale down to atomistic scales and also in integrating the experimental and virtual worlds will be highlighted. Practical implications for coupling a heterogeneous variety of codes and tools to microstructure simulations will be discussed using the example of the commercial multi-phase-field software MICRESS®. The paper concludes with some conceptual thoughts about a future standardized format for the description of digital microstructures.

Corrosion costs the United States over $1 trillion annually, yet is typically not analyzed at a detailed level during the product design phase. Our vision is to develop robust corrosion performance modeling tools that will ultimately enable us to predict corrosion behavior and generate composition-processing-structure-performance relationships that can be integrated into the product design standard work for product corrosion liability, and reduction of both the design cycle duration and cost. Our approach links modeling and experiments with verification and validation by coupling thermodynamic and kinetic modeling of alloy processing, phase-specific electrochemical characterization, and a multi-physics model description of the localized electrochemical properties such as corrosion current density. Material composition, processing conditions, and environmental exposure will serve as inputs to the predictive corrosion modeling. An initial application of this methodology on aluminum alloy will be presented and exemplary guidance on how the heterogeneity of the intermetallic (IM) particles within the alloy influence the alloy electrochemical response will be discussed. By understanding the phase-specific electrochemical response through measurements and density functional theory calculations, the impact of variations in alloy composition and processing can be related to the corrosion performance. Combining the intermetallic phase descriptions with the electrochemical response into a multi-physics model will then allow us to interrogate effects such as IM particles composition, size and distribution on the electrochemical response to guide alloy modification for corrosion risk mitigation.

In 2014 TimkenSteel commissioned a state of the art jumbo bloom vertical continuous caster at the Faircrest Steel Plant. This $200 million investment provides for additional capacity, manufacturing flexibility, superior cleanness for strand cast products, and broader capability to support higher value SBQ and seamless mechanical tube markets. In preparation for specification of caster design and material quality, computational modeling tools were applied to evaluate the anticipated product quality and variation from the existing process path. Process modeling included computational fluid dynamics (CFD) modeling for design and operation of the tundish (alongside physical water modeling), deformation modeling of soft reduction roll effects on porosity and segregation, thermal modeling to evaluate transformation and thermal stress effects during cooling and reheating, and process modeling to ensure the identified metallurgical solutions fit within the practical manufacturing time envelope.

Integrated computational materials engineering (ICME) is an emerging discipline that can accelerate product development and unify design and manufacturing. This paper summarizes the ICME applications for optimizing welding and thermal-forming processes. In the first application, numerical analyses were conducted to control weld distortion by using intermittent welding to replace full-length double-sided fillet welds and adapting new low heat-input processes such as the hybrid laser-arc welding process. In the second application, an existing software tool was validated and applied to automate thermal plate forming, a process which has been extensively used in the shipbuilding industry to form curved plates for several decades.

The main driving force for replacing Co as the binder in cemented carbides is due to possible restrictions in the European regulations regarding Co. The first systematical investigations of alternative binder systems based on Fe and Ni were performed in the 1980’s but in absence of a strong driving force to replace Co the efforts did not lead to any new material. More recently different promising alternatives have been presented which could lead to reasonable solutions for the cemented carbide industry. Nevertheless, the important sector of metal cutting still suffers from a lack of suitable binder phase replacement. The present work identifies the crucial points for the materials developing focusing on metal cutting applications. A computational materials design approach has been applied by using thermodynamic and kinetic models involving DFT, DICTRA and CALPHAD-type of calculations.

Turbine blades are the key parts of aero-engines and industrial gas turbines. Liquid metal cooling (LMC) method has been used in the directional solidification process to get higher temperature gradient and produce the large size turbine blade casting. To reduce the developing cycle and cost, numerical simulation technology is used to optimize the directional solidification process by liquid metal cooling.In this paper, mathematical models of directional solidification assisted by liquid metal cooling were established, including the heat transfer model which considered the heat convection between mold shell and cooling liquid metal, nucleation and grain growth model, etc. The dynamic change of boundary conditions at different parts of the turbine blade casting in LMC process was identified and described mathematically.Experiments have been done to validate the proposed models for LMC directional solidification. The cooling curves and the grains morphology obtained in the experiments agree well with the numerical results. Compared with conventional Bridgman method, LMC led to narrower and more stabilized mushy zone, higher allowable withdrawal rate and better microstructure. LMC process of a real industrial gas turbine blades was optimized. The morphology of mushy zone in the beginning stage was calculated under the initial and improved processes. The grain structure in the blade body was also simulated. By the optimized process, the grains in the DS blade can be significantly improved and become parallel.

Rapid manufacturing technologies are lately gaining interest as alternative manufacturing method. Due to the large parameter sets applicable in these manufacturing methods and their impact on achievable material properties and quality, support of the manufacturing process development by the use of simulation is highly attractive. This is especially true for aerospace applications with their high quality demands and controlled scatter in the resulting material properties. The applicable simulation techniques to these manufacturing methods are manifold.The paper will start with the 2D melt pool simulation for SLM (selective laser melting) process. The mapping to a 3D structure using equivalent heat sources is demonstrated and validated. The precipitation kinetics are calculated for the temperature-time history of the process and subsequent heat treatment using CALPHAD based methods. In a last step the concept for the prediction of the material strength based on the precipitation condition is presented.

Multiscale modeling of material systems demands novel solution strategies to simulating physical phenomena that occur in a hierarchy of length scales. Majority of the current approaches involve one way coupling such that the information is transferred from a lower length scale to a higher length scale. To enable bi-directional scale-bridging, a new data-driven framework called Materials Knowledge System (MKS) has been developed recently. The remarkable advantages of MKS in establishing computationally efficient localization linkages (e.g., spatial distribution of a field in lower length scale for an imposed loading condition in higher length scale) have been demonstrated in prior work. In these prior MKS studies, the effort was focused on composite materials that had a finite number of discrete local states. As a major extension, in this work, the MKS framework has been extended for polycrystalline aggregates which need to incorporate crystal lattice orientation as a continuous local state. This extension of the MKS framework for elastic deformation of polycrystals is achieved by employing compact Fourier representations of functions defined in the crystal orientation space. The viability of this new formulation will be presented for case studies involving single and multi-phase polycrystals.

Fundamental changes to aero-vehicle management require the utilization of automated health monitoring of vehicle structural components. A novel method is the use of self-sensing materials, which contain embedded sensory particles (SP). SPs are micron-sized pieces of shape-memory alloy that undergo transformation when the local strain reaches a prescribed threshold. The transformation is a result of a spontaneous rearrangement of the atoms in the crystal lattice under intensified stress near damaged locations, generating acoustic waves of a specific spectrum that can be detected by a suitably placed sensor. The sensitivity of the method depends on the strength of the emitted signal and its propagation through the material. To study the transition behavior of the sensory particle inside a metal matrix under load, a simulation approach based on a coupled atomistic-continuum model is used. The simulation results indicate a strong dependence of the particle’s pseudoelastic response on its crystallographic orientation with respect to the loading direction and suggest possible ways of optimizing particle sensitivity. The technology of embedded sensory particles will serve as the key element in an autonomous structural health monitoring system that will constantly monitor for damage initiation in service, which will enable quick detection of unforeseen damage initiation and progression in real-time and during on-ground inspections.

Due to non-uniform solidification, internal void defects (like porosity and shrinkage) often exist in the large ingot and behave as sources of damage in the materials performance. Multi-stage forging process, like stretching swaging and upsetting, is often used to eliminate the voids. However, the material with voids inside often undertakes loading from different direction in different forging stage, which may results in closure-reopen effects of the voids. Therefore, modeling the 3-d deformation behavior of the voids is of great significance in multi-hit deformation of the ingot. A 3-d voids evolution modeling is presented by using mesoscale representative volume element (RVE). In the RVE, the deformation of the void is expressed as a function of the remote stress, remote effective strain-rate and the void aspect parameter. The coefficients of the model are determined from the finite element (FE) calculations of the RVE. In this model, the change of void radius is influenced by the mean stress and the deviatoric stress in the corresponding direction. The relationship between the void radius deformation rate and the void aspect parameter in the corresponding direction are heuristically established as inverse proportional functions. As a consequence, the volume and shape of the voids can be obtained by integrating the material deformation history. By combining the void evolution model with macroscopic finite element simulation of the multi-stage forging process, the voids evolution can be efficiently predicted and the condition for closing the voids can be obtained. This model is validated by laboratory experiments and applied in industry.

The PW-8: Nickel Base Superalloy Residual Stress Foundational Engineering Problem (FEP) is a program funded by the United States Air Force through the Metals Affordability Initiative (MAI) to address bulk residual stresses in Nickel-base superalloy engine disk components. These stresses can be induced during various manufacturing stages such as the heat treatment process or the forging process. Bulk residual stresses can be a problem and result in component distortion during the machining process and/or during elevated temperature service. Bulk residual stresses in aeroengine disks components are considered a Foundational Engineering Problem that affects both suppliers and Original Equipment Manufacturers (OEMs) and is an issue that must be addressed with a cross-functional team. The FEP addresses this problem by developing the infrastructure and tools needed to predict and incorporate bulk residual stress into the design and development of a turbine disk. In doing so, the FEP answers the challenge given in a 2008 report issued by the National Research Council in which the authors commented that addressing FEPs are an essential means to help establish the infrastructure needed to make Integrated Computational Materials Engineering (ICME) a reality. This paper will report on a key aspect of the ICME infrastructure; namely the infrastructure needed to manage physical and model data. In addition to discussing the infrastructure developed, this paper will document the lessons learned which can be applied to the ICME community as a whole.

The use of physics-based simulations of manufacturing processes for the prediction of material properties and defects is increasingly widespread in industry. Such simulation tools help answer “what-if” questions, and a materials engineer may have to conduct a number of simulations for decision making. In practice the engineer is often seeking a solution to an “inverse problem”, i.e., prediction of inputs/process parameters for a desired outcome. Such an inverse problem is often solved by formulating it as a constrained-optimization problem. Extensive simulation in the input-parameter space when performing the optimization is avoided by approximate response surfaces iteratively constructed using simulations executed while traversing the design space. In this paper, we present a case study on the application of machine learning techniques to address such inverse problems. Specifically, using data from physics-based simulations, we explore the use of two different kinds of models constructed by machine learning. The first approach constructs a “generative” model (a Bayesian network), from which input values can be obtained directly from output values, without the need of an optimization step. It does however need additional knowledge in the form of conditional (in)-dependences between process parameters, intermediate state variables, and outputs. The second is a purely predictive machine learning model capturing complex non-linearity followed by the use of optimization methods (simulated annealing) for inverse prediction. We present results for modelling of a heat treatment process chain involving carburization, quenching and tempering. Our findings are as follows: For the range of output-values we examined, the predictive model performs better than the generative model. However the generative model has the ability to discover multiple solutions to the inverse problem, unlike in the traditional response-surface-based design of experiments. Thus the generative approach may prove more useful for exploratory industrial practice in the long run.

http://nanoHUB.org is an open-access cyberinfrastructure supported by the US National Science Foundation. It provides a powerful and versatile platform for delivering computational resources to support the implementation of Integrated Computational Materials Engineering (ICME) solutions and associated educational efforts. nanoHUB provides free cloud scientific computing, where users can access simulation tools for research and education using a web-browser or iPad, without the need to install software or have access to local computing resources. The tools have friendly, fully-interactive graphical user interfaces, meaningful default values, output of both numerical data and visualizations, and extensive support material that provides an accessible pathway towards advanced materials simulations. On the back end, nanoHUB computational resources include high performance computing clusters that enable research-quality simulation.The nanoHUB platform can empower ICME researchers and educators in two distinct ways: by using existing tools or by creating and publishing new, customized nanoHUB tools. Existing nanoHUB tools are currently used to introduce ICME simulations in courses in a variety of areas such as electronic structures, diffusion kinetics, and mechanics of materials. These tools were created by installing the available open source code into nanoHUB, then building custom graphical user interfaces (GUIs) using the Rappture toolkit (Rapid Application Infrastructure; http://rappture.org), which is integrated into the nanoHUB workspace. In a similar manner, instructors can install other simulation tools into nanoHUB using Rappture to easily customize their tools to meet their learning objectives. This paper introduces existing nanoHUB ICME tools, additional resources for materials science education, and the procedure for customizing and publishing new simulation tools on nanoHUB.

Access to digital data and the ability to use it in an integrated computational materials engineering framework are critical needs for materials and manufacturing innovation. Since the start of the Materials Genome Initiative in 2011, several efforts have been launched to develop the resources and protocols necessary for a robust materials data infrastructure. These projects have yielded promising results, while helping to uncover and clarify the challenges that remain. ASM International’s Computational Materials Data Network is active in several of these initiatives, and closely tracks others with similar goals. This presentation reports on the status of CMD Network projects and gives a brief update on many others. It also suggests ways to consolidate efforts and accelerate the development of an infrastructure for ICME.

Creating an environment to enable the seamless integration of experiment, computation, and data within a laboratory environment is essential to enabling the practice of Integrated Computational Materials Engineering. Such an environment depends on the connection of experimental equipment and high performance computing resources to a collaborative software environment that supports research teams through simulation tool sharing and archival data management in a secure manner. Key functions of such a system include project management, workflow management, tool staging, data provenance tracking, and user authentication. An overview will be provided on efforts to establish such an integrated collaborative environment in a research laboratory involved in material and process discovery and development in both structural and functional materials.

Material selection is always driven by multiple constraints. Tools available to a designer such as Ashby plots can be used to study and relate properties, such as density and stiffness, of various material systems. One class of materials, cast iron, has a rich history of property characterization over the last 200 years. In the cast iron system, the resultant microstructure plays an important role in the final properties. In addition to chemistry, the microstructure of cast iron develops from process variables such as addition sequence, spherodizing treatment, inoculation treatment, hold time, solidification rate, cooling rate, etc. Full computational modeling of this material is elusive because of these many process variables and spatial variations. Numerous works have been done in the past to control the microstructure to achieve a desired grade of cast iron. The purpose of this review is to compile and compare the historical experimental results. Important cast iron properties such as ultimate tensile strength, thermal conductivity, and ultrasonic velocity are catalogued into a database from the reported properties. This work should help those gaps in experimental and modeling work to be more easily identified for future research programs.

A general simulation technology was provided as an integration of finite element simulation with a void closure model, a thermal crack model and microstructural evolution models. This integrated simulation method is suitable for hot forging process in which the quality control was still a tough task. To validate the proposed technology, simulation results were compared with experiments data in terms of void closure extent, crack happening position and condition, and grain size. Furthermore, a forming of head of nuclear power vessel was simulated to test its applicability in multi-step forming process.

To meet requirements emanating from environmental, safety, and competition, auto manufacturers are demanding improved performance and reduced defect levels from steel makers. The defects are dependent on the design of unit operations, and the processing conditions in the ladle, tundish, and the caster. To improve performance and reduce defect levels, steel makers need to design the process considering multiple unit operations. In this paper, we present a method to determine the design set points of ladle, tundish and casting operation to meet the desired properties of a cast slab, for a given input of molten steel to the ladle. The decisions associated with ladle, tundish and slab continuous casting unit operations are modeled using the compromise Decision Support Problem (cDSP) construct. Within the cDSPs, the required properties and tolerable defect levels for the continuously cast slab and available process window for ladle, tundish and continuous casting are specified. An inductive approach (upstream-downstream design method) is adopted for exploring the solution spaces of the three unit operations as an integrated whole. In this paper, the design set points obtained using this method for ladle, tundish and continuous casting of slab for different set of requirements is presented. The primary advantage of the proposed method is that it enables rapid exploration of the steel slab production process space. The proposed method is extensible and other downstream processes involved in manufacturing of a finished product from steel will also be integrated together.

While the materials community within ICME is focused on model development and integration across scales, the primary challenge from a design standpoint is the utilization of these models within product development. There are various types of uncertainties associated with the input parameters and the models themselves. These uncertainties are propagated as we integrate and generate process chains. Typically performance estimation at the end of such integrated process chains may have high degree of uncertainty. One challenge in ICME is to integrate the process models while managing various types of associated uncertainties such as input, parameter, model and propagated uncertainties, and to develop process chains which can be used in product development. In this paper, we introduce the Inductive Design Exploration Method (IDEM), that facilitates the exploration of the design space in an inductive manner and provides designers with a set of solutions relatively insensitive to uncertainty. We capture the effect of process parameters on evolving microstructure and the properties at intermediate stages, and link it to the final properties such as surface and core hardness and case depth. This information from all stages of the process including intermediate states predicted by the simulation is used in exploration of the design space. In this paper, we discuss the salient features of IDEM and illustrate its efficacy in developing heat treatment process chain for designing an automotive gear to the desired design specifications.

In order to describe continuous optimization tasks for the efficient design of materials and production processes from a reasonable data sample size, we propose an integrated surrogate modeling approach. We show the proof of concept by application to a draw bending simulation that describes the relation between the process parameters and the spring-back as the process result. The introduced concept can also be directly applied to experimental data while taking into account the process noise as uncertainty (e.g. for process control). The integrated approach combines three components: Design of Experiments, surrogate process modeling (based on function approximation by regression, e.g. Artificial Neural Networks) and optimization of process or material parameters. The identified parameters enable to rapidly find the optimal operating conditions for real experiments or to constrain them for further detailed simulation studies. Future work involves applications to more complex experiments or simulations to efficiently determine the optimal process or material parameters by sparse and adaptive data samples.

We contend that ICME is not limited to selecting an available material from a database; instead, ICME includes actually tailoring material structure at various levels of hierarchy (atoms, microstructure, etc.) via associated processing paths to achieve properties and performance levels that are customized for a particular application. Accordingly, from a systems design perspective we view ICME as the top-down driven, simulation-supported, decision-based design of material hierarchy to satisfy a ranged set of product-level performance requirements. George Box is reputed to have said that all models are wrong but only some of them are useful. Form a systems design perspective we observe that in model-based realization of engineered systems, the decision maker must be able to work constructively with decision models and analysis models that are typically incomplete and inaccurate. Hence, the need to manage uncertainty.In this paper, we highlight our approach for managing uncertainty in the realization of engineered materials and components and a key development that is necessary to institutionalize ICME in industry, namely, a computational platform.

In this paper we show how we explore the performance-property-structure solution space for friction stir welding (FSW). There are several factors that influence weldability in FSW, such as processing conditions and processing parameters like tool shape and size, axial force, and rotational speed. The heat generation due to friction is a major factor in producing distortion problems and energy consumption during the FSW process. Therefore, we focus on improving the strength and efficiency of the process by tailoring tool shape and size in the structure, property and performance space.

Temperature field in metal cutting process is one of the most important phenomena in machining process. Temperature rise in machining directly or indirectly determines other cutting parameters such as tool life, tool wear, thermal deformation, surface quality and mechanics of chip formation. The variation in temperature of a cutting tool in end milling is more complicated than any other machining operation especially in high speed machining. It is therefore very important to investigate the temperature distribution on the cutting tool—work piece interface in end milling operation. The determination of the temperature field is carried out by the analysis of heat transfer in metal cutting zone. Most studies previously carried out on the temperature distribution model analysis were based on analytical model and with the used of conventional machining that is continuous cutting in nature. The limitations discovered in the models and validated experiments include the oversimplified assumptions which affect the accuracy of the models. In metal cutting process, thermo-mechanical coupling is required and to carry out any temperature field determination successfully, there is need to address the issue of various forces acting during cutting and the frictional effect on the tool-work piece interface. Most previous studies on the temperature field either neglected the effect of friction or assumed it to be constant. The friction model at the tool-work interface and tool-chip interface in metal cutting play a vital role in influencing the modelling process and the accuracy of predicted cutting forces, stress, and temperature distribution. In this work, mechanistic model was adopted to establish the cutting forces and also a new coefficient of friction was also established. This can be used to simulate the cutting process in order to enhance the machining quality especially surface finish and monitor the wear of tool.