Additive Manufacturing of Metals: Applications of Solidification Fundamentals: Continuum Scale Modeling and Experiments
Sponsored by: TMS Materials Processing and Manufacturing Division, TMS: Additive Manufacturing Committee, TMS: Solidification Committee
Program Organizers: Alex Plotkowski, Oak Ridge National Laboratory; Lang Yuan, University of South Carolina; Kevin Chaput, Northrop Grumman; Mohsen Asle Zaeem, Colorado School of Mines; Wenda Tan, The University of Michigan; Lianyi Chen, University of Wisconsin-Madison
Monday 8:30 AM
March 15, 2021
Room: RM 4
Location: TMS2021 Virtual
Session Chair: Alex Plotkowski, ORNL; Matt Rolchigo, LLNL
8:30 AM
Statistical Modeling of Microstructure Signatures in Laser Powder Bed Fusion: Supriyo Ghosh1; Raiyan Seede1; Jaylen James1; Ibrahim Karaman1; Alaa Elwany1; Douglas Allaire1; Raymundo Arroyave1; 1Texas A&M University
Laser powder bed fusion solidification of metal alloys often results in cellular morphology in which the microstructure signatures, such as solute microsegregation, were determined using experiments and simulations, and the data obtained were utilized to explore the predictive capability of microsegregation models. The experimental ground truth was compared with high-fidelity phase-field simulations as well as with analytical model predictions. Supervised statistical analyses, including linear regression, polynomial regression, and model reification, were employed to understand the merit of these approaches toward microsegregation estimation. The bias-variance and accuracy-interpretability trade-off limits were considered in the data analysis that was consistent with our experimental findings.
8:50 AM
Solidification Behavior of Martensitic Precipitation-hardenable Stainless Steels Produced via Additive Manufacturing: Eric Lass1; 1University of Tennessee, Knoxville
The extreme solidification conditions found in additive manufacturing processing of metallic alloys often lead to significant compositional heterogeneity, non-equilibrium phase formation, and microstructure that look and behavior nothing like those found in their wrought counterparts. Nowhere is this more apparent than in steels, specifically martensitic precipitation-hardenable stainless steels, where small variations in the solidification conditions can lead to changes in primary solidification phase (austenite or -ferrite), a transition from columnar to equiaxed grain morphology, and wildly different solid-state transformations during the cyclic heating/cooling cycles and ultimate cooling to room temperature. This work probes the effects of laser power and scan speed on solidification behavior in laser powder-bed-produces 17-4 PH and 15-5 PH stainless steels. Small changes in composition are find to have profound changes on microstructure for a given set of parameters. Optimization of chemistry and processing parameters to produce tailored microstructures (i.e. homogeneous and fine-grained) is also explored.
9:10 AM
3D Characterisation of Cracks Formed in AA2024 and Implications for Alloy Design: Giuseppe Del Guercio1; Marco Simonelli1; Nesma Aboulkhair1; Graham McCartney1; Chris Tuck1; 1University of Nottingham
Solidification cracking is one of the major issues in the development of Laser-Powder Bed Fusion (L-PBF) of high strength Al-alloys. Weldability and cracking indices developed for processes where single heating/cooling profiles are imposed present shortcomings in predicting cracking behaviour during L-PBF - where materials experience multiple thermal cycles. In this talk, we investigate the origins of the cracking phenomena in the aluminium alloy AA2024, to expose the merits and limitations of commonly used solidification cracking indices in the context of L-PBF. The cracks’ formation in printed AA2024 with standard ASM and custom alloying elements’ range is characterised using a pFIB-FEGSEM to understand the effects of chemical composition, precipitation of second phases and crystallographic texture on the cracking phenomena. Such fundamental understand is of paramount importance and will be discussed in relation to a newly proposed alloy design strategy for high strength aluminium alloys for L-PBF.
9:30 AM
Quantification and Propagation of Aleatoric Uncertainty Through Numerical Simulation of Laser Powder Bed Fusion Process for IN625: Scott Wells1; 1Purdue University
Defects and build inconsistencies pose a significant hindrance to widespread adoption of laser powder bed fusion. Due to the complexity of the process numerical models are widely used to study L-PBF, but can be computationally cumbersome and fail to account for the many sources of uncertainty. Aleatoric uncertainty arising from thermophysical properties of the alloy, processing parameters, thermodynamically driven solidification models, and parameters driving fluid mechanics within the melt pool significantly contribute to uncertainty in simulation predictions. Using a computational fluid dynamics model for melt pool predictions of Inconel 625, and applying sparse grids and interpolation techniques, uncertainty in the high-dimensional problem can be modeled via polynomial surrogate models. Monte Carlo sampling of the constructed surrogate model provides insight into the propagation of aleatoric input uncertainty to model predictions of melt pool dimensions and solidification time.
9:50 AM
Quantifying Impact of Fluid Flow on Melt Pool Model Predictions Across AM Processing Regimes: Gerry Knapp1; Matthew Rolchigo2; Tarasankar DebRoy1; Jim Belak2; Alex Plotkowski3; 1The Pennsylvania State University; 2Lawrence Livermore National Laboratory; 3Oak Ridge National Laboratory
During additive manufacturing (AM) the molten pool dynamics govern key processing outcomes, importantly the solidification conditions and the resulting microstructure. Flow driven by thermocapillary forces is known to contribute significantly to heat transfer that controls solidification conditions under many conditions. However, computational fluid dynamics calculations are computationally expensive compared to simpler approximate heat transfer calculations. Through nondimensionalization of process and material characteristics, here a dimensionless parameter is developed to correlate with the effects of fluid flow on molten pool geometry and microstructure across a range of processing conditions. Modeling results are used to prove correlation and discuss physical interpretation. Selected cases coupled with cellular automata calculations demonstrate how important fluid flow is to grain structure under different additive manufacturing processing regimes.
10:10 AM
Alternative Scan Strategies for Laser Powder Bed Additive Manufacturing to Expand Process Space: Elizabeth Chang-Davidson1; Nicholas Jones1; Jack Beuth1; 1Carnegie Mellon University
Metals additive manufacturing is an emerging field in manufacturing, in which one commonly used technology is laser powder bed fusion (L-PBF). Melt pool sizes in L-PBF are closely tied to printed part material properties, but are currently limited by keyholing porosity flaws or by machine limits on laser power and velocity. For carefully selected process parameters, much larger than typical melt pools were created by rapidly scanning the laser back and forth across a constant width. A systematic way to apply this technique was mapped across laser power and velocity using semi-analytical simulation software. Sample single stripes and cubes were printed using parameters selected to span process space in power and velocity. These experimental results were used to calibrate the simulations and demonstrate viability of the technique. This systematically applicable technique increases the range of melt pool sizes and therefore range of material properties possible to print using L-PBF machines.
10:30 AM
Microstructure Control with Advanced Scan Strategies Developed via Fast Analytic Thermal Modeling of Additive Processes: Benjamin Stump1; Patxi Fernandez1; Matt Rolchigo2; Alex Plotkowski1; Jim Belak2; 1ORNL; 2LLNL
Advanced microstructure control is a major objective in additive manufacturing (AM). Since microstructure is related to macroscopic properties, enabling control could result in built parts that are both lighter and stronger than before. Microstructure in AM is dependent on both material parameters and processing conditions. Many studies have been done tweaking various processing conditions, such as power, beam velocity, and dwell time, and seeing the resulting effect on the microstructure. Even fever studies look at adjusting the scan path itself due to the problem being of substantially higher complexity. This talk delves into using a fast, analytic heat transfer solution for this purpose and some of the resulting scan strategies. These scan strategies came from thought experiments, optimizations, and neural networks and provide different insights into the problem. CA results and actual microstructure data from the resulting build are also presented.
10:50 AM
Consistent Coupling between Melt Pool Heat Transfer and Grain-scale CA Calculations for Additive Manufacturing: John Coleman1; Alex Plotkowski1; Matt Rolchigo2; 1Oak Ridge National Laboratory; 2Lawrence Livermore National Laboratory
Grain structure predictions in CA models rely on spatiotemporal thermal information from melt pool heat transfer calculations. At the microscale, the high cooling rates in additive manufacturing cause significant thermal undercooling ahead of the dendrite tips. CA models use this undercooling to approximate dendrite tip velocities. However, resolution of such phenomena by the melt pool heat transfer calculations is computationally prohibitive. Therefore, volume averaging is used to approximate the relative amounts of thermodynamic phases present in a control volume based on local thermal conditions. To provide conceptually consistent thermal information with the interpretations of dendrite behavior, consideration is given to the thermodynamic and kinetic models used to approximate phase fractions in melt pool calculations. Further consideration is given to the calculation and interpolation of spatiotemporal thermal information down to the grain scale. Various methods are compared in terms of melt pool-scale and grain-scale predictions.