-
Machine Learning for Materials Developments in Metals Additive Manufacturing
Authors:
N. S. Johnson,
P. S. Vulimiri,
A. C. To,
X. Zhang,
C. A. Brice,
B. B. Kappes,
A. P. Stebner
Abstract:
In metals additive manufacturing (AM), materials and components are concurrently made in a single process as layers of metal are fabricated on top of each other in the near-final topology required for the end-use product. Consequently, tens to hundreds of materials and part design degrees of freedom must be simultaneously controlled and understood; hence, metals AM is a highly interdisciplinary te…
▽ More
In metals additive manufacturing (AM), materials and components are concurrently made in a single process as layers of metal are fabricated on top of each other in the near-final topology required for the end-use product. Consequently, tens to hundreds of materials and part design degrees of freedom must be simultaneously controlled and understood; hence, metals AM is a highly interdisciplinary technology that requires synchronized consideration of physics, chemistry, materials science, physical metallurgy, computer science, electrical engineering, and mechanical engineering. The use of modern machine learning approaches to model these degrees of freedom can reduce the time and cost to elucidate the science of metals AM and to optimize the engineering of these complex, multidisciplinary processes. New machine learning techniques are not needed for most metals AM development; those used in other sects of materials science will also work for AM. Most prolifically, the density functional theory (DFT) community has used many of them since the early 2000s for evaluating numerous combinations of elements and crystal structures to discover new materials. This materials technologies-focused review introduces the basic mathematics and terminology of machine learning through the lens of metals AM, and then examines potential uses of machine learning to advance metals AM, highlighting the many parallels to previous efforts in materials science and manufacturing while also discussing new challenges and adaptations specific to metals AM.
△ Less
Submitted 11 May, 2020;
originally announced May 2020.
-
Physics-informed machine learning for composition-process-property alloy design: shape memory alloy demonstration
Authors:
Sen Liu,
Branden B. Kappes,
Behnam Amin-ahmadi,
Othmane Benafan,
Xiaoli Zhang,
Aaron P. Stebner
Abstract:
Machine learning (ML) is shown to predict new alloys and their performances in a high dimensional, multiple-target-property design space that considers chemistry, multi-step processing routes, and characterization methodology variations. A physics-informed featured engineering approach is shown to enable otherwise poorly performing ML models to perform well with the same data. Specifically, previo…
▽ More
Machine learning (ML) is shown to predict new alloys and their performances in a high dimensional, multiple-target-property design space that considers chemistry, multi-step processing routes, and characterization methodology variations. A physics-informed featured engineering approach is shown to enable otherwise poorly performing ML models to perform well with the same data. Specifically, previously engineered elemental features based on alloy chemistries are combined with newly engineered heat treatment process features. The new features result from first transforming the heat treatment parameter data as it was previously recorded using nonlinear mathematical relationships known to describe the thermodynamics and kinetics of phase transformations in alloys. The ability of the ML model to be used for predictive design is validated using blind predictions. Composition - process - property relationships for thermal hysteresis of shape memory alloys (SMAs) with complex microstructures created via multiple melting-homogenization-solutionization-precipitation processing stage variations are captured, in addition to the mean transformation temperatures of the SMAs. The quantitative models of hysteresis exhibited by such highly processed alloys demonstrate the ability for ML models to design for physical complexities that have challenged physics-based modeling approaches for decades.
△ Less
Submitted 8 October, 2020; v1 submitted 3 March, 2020;
originally announced March 2020.
-
Knowledge of Process-Structure-Property Relationships to Engineer Better Heat Treatments for Laser Powder Bed Fusion Additive Manufactured Inconel 718
Authors:
Thomas G. Gallmeyer,
Senthamilaruvi Moorthy,
Branden B. Kappes,
Michael J. Mills,
Aaron P. Stebner,
Behnam Aminahmadi
Abstract:
Dislocation structures, chemical segregation, {γ^{\prime}, {γ^{\prime \prime}}, δ precipitates and Laves phase were quantified within the microstructures of Inconel 718 (IN718) produced by laser powder bed fusion additive manufacturing (AM) and subjected to standard, direct aging, and modified multi-step heat treatments. Additionally, heat-treated samples still attached to the build plates vs. tho…
▽ More
Dislocation structures, chemical segregation, {γ^{\prime}, {γ^{\prime \prime}}, δ precipitates and Laves phase were quantified within the microstructures of Inconel 718 (IN718) produced by laser powder bed fusion additive manufacturing (AM) and subjected to standard, direct aging, and modified multi-step heat treatments. Additionally, heat-treated samples still attached to the build plates vs. those removed were also documented for a standard heat treatment. The effects of the different resulting microstructures on room temperature strengths and elongations to failure is revealed. Knowledge derived from these process structure property relationships was used to engineer a super solvus solution anneal at 1020 degC for 15 minutes, followed by aging at 720 degC for 24 hours heat treatment for AM-IN718 that eliminates Laves and δ phases, preserves AM specific dislocation cells that are shown to be stabilized by MC carbide particles, and precipitates dense {γ^{\prime} and {γ^{\prime \prime}} nanoparticle populations. This 'optimized for AM-IN718 heat treatment' results in superior properties relative to wrought/additively manufactured, then industry standard heat treated IN718: relative increases of 7/10 percent in yield strength, 2/7 percent in ultimate strength, and 23/57 percent in elongation to failure are realized, respectively, regardless of as-built vs. machined surface finishes.
△ Less
Submitted 24 October, 2019; v1 submitted 26 September, 2019;
originally announced September 2019.
-
The interactions of same-row oxygen vacancies on rutile TiO$_2$(110)
Authors:
B. B. Kappes,
W. B. Maddox,
D. P. Acharya,
P. Sutter,
C. V. Ciobanu
Abstract:
Based on a dipolar-elastic model for oxygen vacancies on rutile (110), we evaluated analytically the overall energy of a periodic array of two vacancies and extracted the interaction parameters from total-energy density functional theory (DFT) calculations. Our calculations show that the dipole model holds for next-nearest neighbor vacancies and beyond. The elastic-dipolar interaction vanishes for…
▽ More
Based on a dipolar-elastic model for oxygen vacancies on rutile (110), we evaluated analytically the overall energy of a periodic array of two vacancies and extracted the interaction parameters from total-energy density functional theory (DFT) calculations. Our calculations show that the dipole model holds for next-nearest neighbor vacancies and beyond. The elastic-dipolar interaction vanishes for adjacent vacancies, but they still experience an electrostatic repulsion. The proposed interaction model predicts a vacancy separation distribution that agrees well with that determined in our ultra-high vacuum scanning tunneling microscopy experiments, and provides a perspective for understanding earlier DFT reports.
△ Less
Submitted 13 January, 2013;
originally announced January 2013.
-
Growth Structure and Work Function of Bilayer Graphene on Pd(111)
Authors:
Y. Murata,
S. Nie,
A. Ebnonnasir,
E. Starodub,
B. B. Kappes,
K. F. McCarty,
C. V. Ciobanu,
S. Kodambaka
Abstract:
Using in situ low-energy electron microscopy and density functional theory, we studied the growth structure and work function of bilayer graphene on Pd(111). Low-energy electron diffraction analysis established that the two graphene layers have multiple rotational orientations relative to each other and the substrate plane. We observed heterogeneous nucleation and simultaneous growth of multiple,…
▽ More
Using in situ low-energy electron microscopy and density functional theory, we studied the growth structure and work function of bilayer graphene on Pd(111). Low-energy electron diffraction analysis established that the two graphene layers have multiple rotational orientations relative to each other and the substrate plane. We observed heterogeneous nucleation and simultaneous growth of multiple, faceted layers prior to the completion of second layer. We propose that the facetted shapes are due to the zigzag-terminated edges bounding graphene layers growing under the larger overlying layers. We also found that the work functions of bilayer graphene domains are higher than those of monolayer graphene, and depend sensitively on the orientations of both layers with respect to the substrate. Based on first-principles simulations, we attribute this behavior to oppositely oriented electrostatic dipoles at the graphene/Pd and graphene/graphene interfaces, whose strengths depend on the orientations of the two graphene layers.
△ Less
Submitted 14 May, 2012;
originally announced May 2012.
-
Orientation-dependent binding energy of graphene on palladium
Authors:
B. B. Kappes,
A. Ebnonnasir,
S. Kodambaka,
C. V. Ciobanu
Abstract:
Using density functional theory calculations, we show that the binding strength of a graphene monolayer on Pd(111) can vary between physisorption and chemisorption depending on its orientation. By studying the interfacial charge transfer, we have identified a specific four-atom carbon cluster that is responsible for the local bonding of graphene to Pd(111). The areal density of such clusters varie…
▽ More
Using density functional theory calculations, we show that the binding strength of a graphene monolayer on Pd(111) can vary between physisorption and chemisorption depending on its orientation. By studying the interfacial charge transfer, we have identified a specific four-atom carbon cluster that is responsible for the local bonding of graphene to Pd(111). The areal density of such clusters varies with the in-plane orientation of graphene, causing the binding energy to change accordingly. Similar investigations can also apply to other metal substrates, and suggests that physical, chemical, and mechanical properties of graphene may be controlled by changing its orientation.
△ Less
Submitted 6 August, 2012; v1 submitted 28 January, 2012;
originally announced January 2012.
-
Reactive force field for lithium-aluminum silicates with applications to eucryptite phases
Authors:
Badri Narayanan,
Adri C. T. van Duin,
Branden B. Kappes,
Ivar E. Reimanis,
Cristian V. Ciobanu
Abstract:
We have parameterized a reactive force field (ReaxFF) for lithium aluminum silicates using density functional theory (DFT) calculations of structural properties of a number of bulk phase oxides, silicates, and aluminates, as well as of several representative clusters. The force field parameters optimized in this study were found to predict lattice parameters and heats of formation of selected cond…
▽ More
We have parameterized a reactive force field (ReaxFF) for lithium aluminum silicates using density functional theory (DFT) calculations of structural properties of a number of bulk phase oxides, silicates, and aluminates, as well as of several representative clusters. The force field parameters optimized in this study were found to predict lattice parameters and heats of formation of selected condensed phases in excellent agreement with previous DFT calculations and with experiments. We have used the newly developed force-field to study the eucryptite phases in terms of their thermodynamic stability and their elastic properties. We have found that (a) these ReaxFF parameters predict the correct order of stability of the three crystalline polymorphs of eucryptite, α, β, and γ, and (b) that upon indentation, a new phase appears at applied pressures \geq 7 GPa. The high pressure phase obtained upon indentation is amorphous, as illustrated by the radial distribution functions calculated for different pairs of elements. In terms of elastic properties analysis, we have determined the elements of the stiffness tensor for α- and β- eucryptite at the level of ReaxFF, and discussed the elastic anisotropy of these two polymorphs. Polycrystalline average properties of these eucryptite phases are also reported to serve as ReaxFF predictions of their elastic moduli (in the case of α-eucryptite), or as tests against values known from experiments or DFT calculations (β- eucrypite). The ReaxFF potential reported here can also describe well single-species systems (e.g., Li-metal, Al-metal, and condensed phases of silicon), which makes it suitable for investigating struc- ture and properties of suboxides, atomic-scale mechanisms responsible for phase transformations, as well as oxidation-reduction reactions.
△ Less
Submitted 16 June, 2011;
originally announced June 2011.
-
Bandgap opening in metallic carbon nanotubes due to silicon impurities
Authors:
Branden B. Kappes,
Cristian V. Ciobanu
Abstract:
Controlling the bandgap of carbon nanostructures is key to the development and mainstream application of carbon-based nanoelectronic devices. We report density functional theory calculations of the effect of silicon impurities on the electronic properties of carbon nanotubes (CNTs). We have found that Si adatoms open up a bandgap in intrinsically metallic CNTs even when the linear density of Si…
▽ More
Controlling the bandgap of carbon nanostructures is key to the development and mainstream application of carbon-based nanoelectronic devices. We report density functional theory calculations of the effect of silicon impurities on the electronic properties of carbon nanotubes (CNTs). We have found that Si adatoms open up a bandgap in intrinsically metallic CNTs even when the linear density of Si atoms is low enough that they do not create a bonded adatom chain. The bandgap opened in metallic CNTs can range between 0.10 eV and 0.47 eV, depending on adsorption site, linear density of Si adatoms, and CNT chirality.
△ Less
Submitted 4 January, 2010; v1 submitted 21 December, 2009;
originally announced December 2009.
