4th World Congress on Integrated Computational Materials Engineering (ICME 2017): Additive Manufacturing - II
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
Monday 2:00 PM
May 22, 2017
Room: Salon II, III
Location: Ann Arbor Marriott Ypsilanti at Eagle Crest
2:00 PM Invited
Validation of Integrated Computational Materials Engineering Principles for Optimization of Fatigue Properties of Ti-6Al-4V Alloys Made by Laser Direct Energy Deposition: Sudarsanam Babu1; A. Prabhu1; N. Sridharan2; K. Makiewicz3; W. Zhang4; A. Chaudhary5; 1The University of Tennessee, Knoxville; 2Oak Ridge National Laboratory; 3Formerly from The Ohio State University; 4The Ohio State University; 5Applied Optimization Inc.
In early 2000, Kobryn and Semiatin published a classic paper that showed the challenges associated with laser additive manufacturing of Ti6Al4V alloys in terms of anisotropic fatigue properties. The above deficiencies were attributed to tendency for the formation of physical defects that include porosity and lack of fusion, as well as, microstructural heterogeneity. In 2005, Kelly et al attributed the microstructural heterogeneities such as “white band” formation to cyclic phase transformations within the + phase field. Although, these challenges were addressed through post-process hot isostatic pressing, the need for in-process control of microstructure was realized for complex geometries. In this talk, the step-by-step approach for developing computational tools for describing thermal gyrations and concurrent microstructure (solidification and solid-state transformation) evolution during laser additive manufacturing will be discussed. This integrated model was used in a “numerical trial-and-error experimentation” and optimum conditions that led to homogeneous microstructure was selected. The processing conditions, which involved processing above transus, showed that it is indeed possible to arrive at homogeneous basketweave microstructure. Further process optimizations lead to improved fatigue properties without any post process heat treatments. In addition, the experimental data also showed uncertainties in ensuring fatigue properties with this ideal process parameter set. Such uncertainties were attributed to spatial location of defects and variability in performance of auxiliary equipments, such as powder flow. The feasibility of extending these tools for nickel base alloys was also explored. These results demonstrate that the ICME tools provide a viable pathway for rapid qualification of AM components.
Mesoscale Multi-Physics Simulation of Solidification in Selective Laser Melting Process: Dehao Liu1; Yan Wang1; 1Georgia Institute of Technology
Selective laser melting (SLM) is a powder bed based additive manufacturing process by melting fine-grained metallic powders with a laser heating source. Understanding the solidification of alloys during SLM process is of importance for accurate prediction of microstructures and properties for process optimization. In this study, a multi-physics based phase field model is developed to simulate evolution of alloy microstructure during solidification, which incorporates heat transfer, fluid dynamics, and kinetics of phase transformations and grain growth. The effect of cooling rate is investigated through simulation. The simulation results are compared with the morphology of dendrite from experimental measurement.
Investigating the Role of Porosity in Additively Manufactured IN718 by Crystal Plasticity Modeling and Tomography Characterization: Veerappan Prithivirajan1; Todd Book2; Alexander Finch1; Michael Sangid1; 1Purdue University; 2US Military Academy
Direct metal laser sintering (DMLS) can be used for aerospace applications to realize numerous advantages, which have been well documented. However, prior to their use in safety critical components, the failure mechanisms have to be well understood, especially for the unique defects in DMLS materials, especially porosity. In our work, we study the critical pore size and volume fraction relative to the microstructure of a Ni-based superalloy, IN718, produced by DMLS via crystal plasticity finite element (CPFE) simulations and compare with synchrotron micro-tomography characterization results. 3D virtual microstructures with varying pore sizes and volume fractions are developed and subsequent cyclic simulations are carried out using a CP-FE framework. The critical size of the pore is defined as the size beyond which the crack nucleation transitions from crystallographic planes/GBs to the vicinity of the void. The crack nucleation site is influenced by many factors such as large grains, orientation distribution, grain neighbor interactions, twin boundaries, void size, and void interactions. Finally, the CPFE simulation results are compared with the porosity observed in DMLS IN718 via synchrotron micro-tomography characterization results.
Precipitate Kinetics in Inconel 718 Additive Manufactured Components: Magnus Anderson1; Chinnapat Paniwisawas1; Yogesh Solvani1; Richard Turner1; Jeffery Brooks1; Hector Basoalto1; 1University of Birmingham
A mean field model of the precipitation kinetics of the intermetallic particle phases in IN718 has been developed and applied to predict nucleation, growth, coarsening and dissolution behaviour during additive manufacture. The engineering properties of this alloy depend on the formation of a complex distribution of both metastable and equilibrium phases. Consequently, a simulation tool is needed to assist in the design of heat treatments to avoid excessive precipitation of the δ phase and understand how the heat affected zone develops during additive manufacture. Svoboda et al’s (2004) multi-component mean field particle coarsening model has been implemented. The approach is similar to that described by Zickler et al, 2010. The γ’ particles are approximated to be spherical, whilst Kozeschnik et al’s shape factors are used to describe the γ” and δ phases by cylindrical discs. The interfacial energy and nucleation site density were calibrated to predict the precipitation kinetics of the γ” and δ presented by Fisk et al (2014), Azadian et al (2004) and Beaubois et al (2004). The thermal history occurring within an ALM component during manufacture has been simulated using a Finite element model. The kinetics of precipitation resulting from the thermal loading have been simulated at a cross section of the component. The model predicts changes in particle dispersion across the heat affected zone, and also the repeated nucleation and dissolution of precipitates due to the repetitive passing of the heat source over the deposited material.
3:30 PM Break
Large Scale Phase Field Simulations of Microstructure Evolution in POlycrystalline Ti-6Al-4V during Multiple Thermal Cycling: Bala Radhakrishnan1; Sarma Gorti1; Suresh Babu1; 1Oak Ridge National Laboratory
The phase field method has been established as a standard ICME tool for simulating microstructural evolution during thermo-mechanical processing of structural alloys. The objective of this study is to perform large scale, high spatial resolution phase field simulations of the beta to alpha transformation in polycrystalline Ti-6Al-4V alloy during multiple thermal cycling, and to validate the simulation results using experimental microstructures generated during additive manufacturing of the alloy. The proposed simulations will be based on an extension of our recent work on isothermal simulations in single crystals using a novel composite nucleation model that demonstrated an intragranular mechanism for the formation of lamellar alpha in Ti-6Al-4V based on the interaction between the nucleation rate and the accommodation of strain energy associated with the beta to alpha transformation. The proposed simulations will address the combined effects of grain boundary nucleation and variant selection of alpha and intragranular nucleation of alpha on the morphology transition from basket-weave to lamellar alpha within the layer bands formed during additive manufacturing. Research at Oak Ridge National Laboratory is managed by UT-Battelle, LLC, under contract DE-AC05- 00OR22725 for the U.S. Department of Energy.
A Phase Field-based Microstructure Evolution Model for Selective Laser Melting of Ti-6Al-4V: Mehdi Amiri1; Kai Wing Kelvin Leung1; Nagaraja Iyyer1; 1Technical Data Analysis Inc.
Selective laser melting (SLM) is an innovative additive manufacturing (AM) technique which offers an opportunity to control local microstructure and mechanical properties of a part, by adjusting process parameters such as scan speed and power of laser. Optimization of such process parameters is important in order to ensure the mechanical properties of parts fabricated by SLM process meet design requirements. In this work, we present an integrated multi-scale computational framework for predicting thermal history and evolution of microstructure of Ti-6Al-4V using a phase field modeling approach. Thermal model includes transient heat transfer analysis of a single layer multi-track SLM process using finite element techniques while microstructure model utilizes thermal history as an input to simulate alpha phase formation and its dissolution for the given thermal history. Phase field equations are implemented to account for alpha phase dissolution with diffusion equations and alpha phase formation by a Johnson-Mehl-Avrami-Kolmorgorov nucleation and growth model. This combined approach is able to quantify the effect of thermal cycling on the as-deposited microstructure and predict the evolution of alpha phase fraction, which ultimately controls mechanical performances. Results for different laser scan speed and laser power are presented and discussed.
A Coupled Framework for the Spatial Design of Shape Memory Functionality in NiTi Based Additive Manufacturing: Luke Johnson1; Kubra Karayagiz1; Ji Ma1; Brian Franco1; Gustavo Tapia1; Alaa Elwany1; Ibrahim Karaman1; Raymundo Arroyave1; 1Texas A&M University
The ability to precisely control part geometry and functional grading are the two most common examples of fabrication freedom granted by additive manufacturing that are not commonly attained through other more conventional manufacturing approaches. However, there are more dimensions in the design space that remain to be explored. This talk will describe a framework for precipitation design in NiTi shape-memory alloy systems fabricated via SLM. The framework couples finite-element thermal modeling simulations with a thermodynamic and kinetic based precipitate evolution model to predict precipitate volume fraction in a fabricated part. The thermal model was calibrated against experimental pyrometer data and the precipitation model was calibrated against experimental data from the literature using Bayesian calibration techniques. While originally developed to help explain within-part spatial variations in shape-memory behavior, this model can be inverted to provide top-down spatial design of shape-memory functionality.
Measurement and Calculation of Liquid Ti Alloy Properties with Application to 3D Printing: Jonathan Raush1; Brian Novak2; Xiaoman Zhang2; Dorel Moldovan2; Wenjin Meng2; Shengmin Guo2; 1University of Louisiana at Lafayette; 2Louisiana State University
Accurate high-temperature thermophysical property data for liquid metals and alloys are important for the development of realistic simulations of laser-based 3D printing processes. We are using an approach combining electrostatic levitation (ESL), molecular simulation, and CALPHAD calculations to obtain such data for Ti alloys. We performed vacuum ESL measurements with a container-less oscillating drop technique at NASA MSFC on molten elemental Ti, Ti-xAl binaries (x = 0-10% wt.), Ti-6Al-4V, and Ti-6Al-4V-10Mo which showed improved mechanical properties compared with traditional β Ti alloys. We used classical molecular simulations to obtain densities, pair distribution functions, diffusivities, viscosities, surface tensions, and vapor pressures for Ti-xAl. The viscosity and surface tension results for pure Ti agree well with the ESL data while the Ti-xAl results show the same trends as the ESL data, but not always quantitative agreement. Chemical activity and Gibbs free energy of Ti-10Al were generated through the CALPHAD technique and compared to experimental values.