Additive Manufacturing of Metals: Applications of Solidification Fundamentals: Physics-based and Data-based Modeling II
Sponsored by: TMS Materials Processing and Manufacturing Division, TMS: Additive Manufacturing Committee, TMS: Solidification Committee
Program Organizers: Wenda Tan, The University of Michigan; Alex Plotkowski, Oak Ridge National Laboratory; Lang Yuan, University of South Carolina; Lianyi Chen, University of Wisconsin-Madison

Tuesday 2:30 PM
March 21, 2023
Room: 21
Location: SDCC

Session Chair: Lang Yuan, University of South Carolina


2:30 PM  
Simulation of Microstructures Produced by Laser Powder Bed Fusion: Theophile Camus1; Daniel Maisonnette2; Oriane Baulin2; Oriane Senninger1; Gildas Guillemot1; Charles-Andre Gandin1; 1PSL University; 2CETIM
    In laser powder bed fusion, the solidification microstructure is formed due to both epitaxial grain growth and nucleation. The structure has a strong influence on the final mechanical properties of parts and is influenced by process parameters such as the laser power and velocity as well as the scanning strategy. The prediction of this microstructure is of prime interest considering the distribution of size, crystallographic orientation and shape of the grains. The cellular automaton method is used to investigate microstructure formation applied to a nickel-base superalloy. The steady state thermohydraulic behavior at the scale of the melt pool is analyzed and used to compute the development of the grain structure. This structure is computed at the scale of a representative volume for several scanning strategies including meander with interlayer rotation 0°, 67° and 90° as well as stripes. Results are intended as an input to the simulation of thermomechanical behavior.

2:50 PM  
Prediction of Large-scale 3D Solidification Microstructure Evolution during Metal Additive Manufacturing with High Efficiency and Resolution: Shunyu Liu1; Yung Shin2; 1Clemson University; 2Purdue University
    Prediction of solidification microstructure during fusion-based metal additive manufacturing should be on large-scale with a high resolution of microstructural features to capture the influence of the 3D temporal and spatial thermal history and non-equilibrium metallurgical feature. This talk will present a comparative study between a probabilistic 3D cellular automata (CA) model and a newly developed convergent 3D cellular automata-phase field (CA-PF) model. The 3D CA model has a simple algorithm and high efficiency, and can predict the grain growth pattern in response to the thermal history but fails to capture the microstructural details resulting from the thermodynamic effects. To address this limitation, the deterministic PF model is used to accurately calculate the solidification kinetics at the solid-liquid interface. Informed by solidification kinetics, the 3D CA model could capture the microstructural details and composition evolution during solidification. As a result, the 3D CA-PF model balances efficiency, accuracy, and resolution.

3:10 PM  
Investigation of Scan Rotation Effects in Additive Manufacturing Using Cellular Automata-based Microstructure Modeling: Matthew Rolchigo1; John Coleman1; Gerry Knapp1; Jamie Stump1; 1Oak Ridge National Laboratory
    The strength and mechanical anisotropy of additively manufactured parts are strongly linked to microstructural features such as grain size and texture. One strategy previously used to successfully control microstructure through careful selection of processing conditions is the application of layer-wise rotations to standard raster scan patterns. This study demonstrates the ability of ExaCA, a cellular automata-based model of the liquid-solid phase transformation, to rapidly generate explicit microstructure predictions in stainless steel builds for scan patterns utilizing layer-wise rotations. Model accuracy will be assessed using pole figure and EBSD mapping data from test builds. The use of simulations to understand process-microstructure relationships associated with scan rotation effects will enable the code to assist in future scan pattern design, addressing the challenges posed by the many possible rotation permutations and expensive build and characterization processes. Work performed under the auspices of the U.S. Department of Energy under contract DE-AC05-00OR22725 for ORNL.

3:30 PM  
Testing Analytic Models and Heuristics for Microstructure Evolution with 3D, Dendrite-resolved Phase-field Simulations of Entire Spot Melts: Stephen DeWitt1; Christopher Newman2; Stephen Nichols1; Jean-Luc Fattebert1; Balasubramaniam Radhakrishnan1; John Coleman1; Gerry Knapp1; James Belak3; John Turner1; 1Oak Ridge National Laboratory; 2Los Alamos National Laboratory; 3Lawrence Livermore National Laboratory
     During additive manufacturing, microstructure evolution at the grain and dendrite scales is complicated by rapid changes in solidification conditions and the melt pool geometry. Analytical models and heuristics based on fundamental solidification theory are commonly applied in additive manufacturing contexts; however, the limits of their applicability are not fully understood. Here, we present a series of 3D, full melt pool, dendrite-resolved phase-field simulations of spot melts to examine the accuracy of such analytical models and heuristics. These simulations provide a full 3D history of microstructure evolution where the thermal conditions are precisely known, which facilitates direct comparisons that cannot be obtained using current experimental methods. In the sub-grain solidification microstructure, we focus on examining planar-cellular-dendritic transitions and primary spacings. In the grain structure, we focus on columnar-equiaxed transitions and grain selection.This abstract has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy.

3:50 PM  
Switching of Controlling Mechanisms during the Rapid Solidification of a Melt Pool in Additive Manufacturing: Yijia Gu1; 1Missouri University of Science and Technology
    Fusion-based metal additive manufacturing (AM) is a disruptive technology that can be employed to fabricate metallic component of near-net-shape with an unprecedented combination of superior properties. However, the interrelationship between AM processing and the resulting microstructures is still not well understood. This poses a grand challenge in controlling the development of microstructures to achieve desired properties. Here we study the microstructure development of a single melt pool, the building block of AM-fabricated metallic component, using a phase-field model specifically developed for the rapid solidification of AM. It is found that during the rapid solidification of the melt pool, the solid-liquid interface is initially controlled by solute diffusion followed by a thermal diffusion-controlled stage with an undercooling larger than the freezing range. This switching of controlling mechanisms leads to the sudden changes in interfacial velocity, solute concentration, and temperature, which perfectly explains the formation of various heterogeneous microstructures observed in AM.

4:10 PM Break

4:25 PM  
Prediction of Solidification Cracking in Rene 80 Superalloy during the Directed Energy Deposition Process: Hamedreza Hosseinzadeh1; Lang Yuan1; Luke Mohr2; Lee Kerwin2; Anindya Bhaduri3; Arushi Dhakad2; Chen Shen3; Shenyan Huang3; Changjie Sun3; Alexander Kitt2; 1University of South Carolina; 2EWI; 3GE Research
    Avoiding solidification cracks is one of the requirements to successfully manufacture high gramma prime superalloys by Additive Manufacturing. In this research, an analytical model based on the RDG model was developed to predict the susceptibility of solidification cracks in Rene 80 via the directed energy deposition process. Taking into account the solidification path calculated by ThermoCalc, local thermal history via the process model for DED, and thermal strain via finite element models, the analytical model shows a promising capability to estimate the crack formation in Rene 80. This predictive model was examined over a wide range of print parameters, including laser power, speed, spot size, and feed rate, to account for different local cooling rates and thermal gradients. Thin wall samples of Rene 80 were fabricated to calibrate and validate the proposed model. This development provides a practical and physics-based method to evaluate the solidification cracking in additively manufactured alloys.

4:45 PM  
The Effects of Non-equilibrium Interfaces and Partial Solute Drag on Morphological Stability: Christopher Hareland1; Gildas Guillemot2; Charles-André Gandin2; Peter Voorhees1; 1Northwestern University; 2Mines Paris - PSL University
    The properties of additively manufactured (AM) components are controlled by the solidification structures that form during processing. The morphologies (i.e., planar or dendritic) and length scales for the growth of these structures can be estimated by examining the stability of a planar solid-liquid interface. Due to the large interfacial velocities, AM processes can lead to significant deviations from local equilibrium at the solid-liquid interface, and novel phenomena such as oscillatory instabilities and banding have been previously observed. Herein, an interfacial stability analysis for a concentrated multicomponent alloy in a system with a planar interface that accounts for the effects of interfacial non-equilibrium, latent heat, and partial solute drag is discussed. The effects of interfacial non-equilibrium are described with a set of interfacial response functions developed for non-equilibrium phase transformations in concentrated multicomponent alloys. Comparisons with existing theories and applications of the results to multicomponent dendritic growth will be discussed.

5:05 PM  
Predicting Phase and Morphology for Use in Site Specific Control of Microstructures in L-PBF Stainless Steel: Michael Haines1; Maxwell Moyle1; Nima Haghdadi1; Sophie Primig1; 1University of New South Wales
    Much interest in metal additive manufacturing (AM) is related to the many opportunities open by production of geometries with unlimited complexity that cannot be produced through traditional methods such as casting and forging. However, despite AM’s benefits, process variations make it difficult to predict final microstructures and mechanical properties. 17-4PH stainless steel is one such alloy that is notably sensitive to process variations. While traditionally known to be a martensitic precipitation hardened stainless steel, under AM, there have been differing reports on the formation of δ-ferrite, γ-austenite, and α'-martensite depending on the processing conditions utilized. This study uses interface response functions (IRF) in connection with heat transfer simulations to predict the resulting phases in laser powder bed fusion (L-PBF) 17-4PH. The simulation is shown to agree with experimental results. This study will allow for the development of parameter sets that permit site specific control of phases within a given geometry.

5:25 PM  
Eliminating Hot Tearing in Laser Powder Bed Fusion of High Strength Aluminium Alloy 2139 Through Parameter Optimisation and Grain Refinement: Joe Elambasseril1; Michael Benoit2; Suming Zhu1; Mark Easton1; Edward Lui1; Craig Brice3; Ma Qian1; Milan Brandt1; 1RMIT University; 2University of British Columbia; 3Colarado School of Mines
    In this work, a systematic study has been conducted to eliminate hot tearing of Al2139 alloy in laser powder bed fusion (L-PBF). In general, hot tearing of Al2139 alloy in L-PBF was reduced or eliminated by increasing ‎energy density. Grain refiner, through the addition of a conventional Al5Ti1B master alloy to Al2139, at a level of 0.2wt%, prior to powder manufacture, has a moderate effect in reducing hot tearing ‎as the refined grains remain predominantly columnar rather than transitioning to equiaxed.‎ Thermomechanical finite element simulation showed that increasing energy density leads to a decrease in thermal stress during L-PBF correlating with the observation of a reduction in hot tearing.‎ The influence of the loss of Mg at high energy densities on hot tearing of Al2139 was evaluated by thermodynamic simulation and is predicted to reduce hot tearing of Al2139 by up to 10 %, compared to the initial powder composition.