Skip to main content
Erschienen in:

Open Access 17.07.2024 | Thematic Section: Additive Manufacturing Benchmarks 2022

Outcomes and Conclusions from the 2022 AM Bench Measurements, Challenge Problems, Modeling Submissions, and Conference

verfasst von: Lyle Levine, Brandon Lane, Chandler Becker, James Belak, Robert Carson, David Deisenroth, Edward Glaessgen, Thomas Gnaupel-Herold, Michael Gorelik, Gretchen Greene, Saadi Habib, Callie Higgins, Michael Hill, Nik Hrabe, Jason Killgore, Jai Won Kim, Gerard Lemson, Kalman Migler, Shawn Moylan, Darren Pagan, Thien Phan, Maxwell Praniewicz, David Rowenhorst, Edwin Schwalbach, Jonathan Seppala, Brian Simonds, Mark Stoudt, Jordan Weaver, Ho Yeung, Fan Zhang

Erschienen in: Integrating Materials and Manufacturing Innovation | Ausgabe 3/2024

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config
loading …

Abstract

The Additive Manufacturing Benchmark Test Series (AM Bench) provides rigorous measurement data for validating additive manufacturing (AM) simulations for a broad range of AM technologies and material systems. AM Bench includes extensive in situ and ex situ measurements, simulation challenges for the AM modeling community, and a corresponding conference series. In 2022, the second round of AM Bench measurements, challenge problems, and conference were completed, focusing primarily upon laser powder bed fusion (LPBF) processing of metals, and both material extrusion processing and vat photopolymerization of polymers. In all, more than 100 people from 10 National Institute of Standards and Technology (NIST) divisions and 21 additional organizations were directly involved in the AM Bench 2022 measurements, data management, and conference organization. The international AM community submitted 138 sets of blind modeling simulations for comparison with the in situ and ex situ measurements, up from 46 submissions for the first round of AM Bench in 2018. Analysis of these submissions provides valuable insight into current AM modeling capabilities. The AM Bench data are permanently archived and freely accessible online. The AM Bench conference also hosted an embedded workshop on qualification and certification of AM materials and components.
Hinweise
The views, opinions, and/or findings contained in this paper are those of the authors and should not be interpreted as representing the official views or policies, either expressed or implied, of the Air Force Research Laboratory, the Air Force, the Department of Defense or the Federal Aviation Administration.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Introduction

Additive manufacturing (AM) is a broad set of transformative manufacturing technologies that produce components directly from three-dimensional (3D) digital files using a wide range of materials, including metals, polymers, ceramics, biomaterials, and composites. In addition to being used for mass customization, AM can produce component geometries that are too costly, difficult, or in some cases, impossible to produce using traditional manufacturing processes. However, challenges persist regarding throughput, reproducibility, reliability, and properties of the printed parts [1]. For example, metal alloy components produced using laser powder bed fusion (LPBF) have cooling rates that vary with location due to the interplay between the laser scan path and the local part geometry. This thermal history variation can produce differences in local microstructures and mechanical behaviors. Although post-build thermal processing can reduce these variations, heat treatments developed for wrought or cast materials often yield unexpected results such as the growth of deleterious phases [2, 3]. Quantitative modeling is critical for mitigating all these challenges, but broad model validation requires community access to extensive and rigorous AM benchmark measurement data.
In response to this need, the National Institute of Standards and Technology (NIST) proposed a framework for developing international benchmark measurements for the AM community at an AM workshop at the National Academies of Sciences, Engineering, and Medicine (NASEM) in Washington D.C. on October 7, 2015 [4]. The response was highly positive, with researchers from dozens of organizations contacting NIST and volunteering their assistance, leading to the development of the Additive Manufacturing Benchmark Test Series (AM Bench). In 2018, the first round of benchmark measurements was completed, along with challenge problems for the AM modeling community and an international conference, AM Bench 2018, which was held at NIST’s Gaithersburg, MD campus. Measurement data and metadata from this effort are freely available. Descriptions and data links may be found in the AM Bench 2018 measurement papers [514] and on the AM Bench website [15].
AM Bench has now completed the second round of benchmark measurements, challenge problems, and conference. In all, more than 100 people directly contributed to the AM Bench measurements, data management, and conference organization, with participation from 10 NIST divisions and 21 external organizations. The contributing organizations are shown in Table 1. We sincerely thank these people and organizations for contributing their time, effort, and resources to helping AM Bench to support the broader AM community by providing and disseminating rigorous measurement test data for model guidance and validation. We also thank the members of the AM Bench Steering Committee, 2022 Organizing Committee, and Scientific Advisory Committee for their invaluable advice and contributions. In addition, AM Bench provides no centralized funding so all measurement collaborators were responsible for obtaining their own support and funding for this work. We thank all these funding sources for contributing to the AM Bench success. In particular, we thank the Department of Energy (DOE)-funded Exascale Computing Project’s ExaAM effort that has been a long-standing partner to AM Bench, providing both partial funding and participation in our measurement efforts. We also thank The Minerals, Metals & Materials Society (TMS) for their invaluable assistance in organizing the logistics for the AM Bench 2022 conference. Lastly, we thank the many companies, universities, and research organizations who submitted simulation results for the AM Bench 2022 challenge problems. In all, AM Bench received 138 challenge problem submissions for the metals and polymers benchmarks, which is a significant increase in participation from 2018 when we received a total of 46 challenge problem submissions. For details about the AM Bench 2018 measurements, challenge problem submissions, and conference, please see the AM Bench 2018 overview paper [5].
Table 1
Organizations that contributed to the AM Bench 2022 measurements, data management, and conference organization
National Institute of Standards and Technology
Los Alamos National Laboratory
Air Force Research Laboratory
US Naval Research Laboratory
Argonne National Laboratory (Advanced Photon Source)
Oak Ridge National Laboratory
Carnegie Mellon University
Oak Ridge National Laboratory (High Flux Isotope Reactor)
Colorado University–Boulder
National Aeronautics and Space Administration–Langley
Cornell High Energy Synchrotron Source
Northwestern University
Eindhoven University of Technology
Penn State University
Federal Aviation Administration
The Minerals, Metals, & Materials Society
Hill Engineering
University of California Davis
Johns Hopkins University
University of North Florida
Lawrence Livermore National Laboratory
University of Virginia
The AM Bench 2022 conference was held in Bethesda, MD, from August 14–18, 2022. The conference included:
  • Numerous presentations on the AM Bench 2022 measurement results
  • Introduction of the new AM Bench data management systems
  • Plenary, invited, and contributed talks from the AM scientific community
  • Discussion sessions
  • An embedded workshop on qualification and certification (Q&C) of AM components and materials
  • An awards ceremony for the winners of the AM Bench 2022 challenge problems.
In the following sections, we will 1) describe the AM Bench scope, 2) provide an overview of the 2022 measurements and modeling challenges, 3) give a general description of the blind modeling challenge submissions we received, 4) make general comparisons between the measurements and model submissions, 5) describe the new AM Bench data management systems, 6) summarize input we received during the AM Bench 2022 discussion sessions, and 7) summarize the results from the Q&C embedded workshop. We will then give our conclusions and our vision for future benchmark measurements for AM.

AM Bench Scope and Benchmark Measurement Selection Criteria

AM Bench was established to provide benchmark measurement data for all AM methods and material classes. However, constrained resources and the requirement for highly controlled experimental conditions limit the current scope. The primary selection criteria AM Bench uses to determine which sets of benchmark measurements are pursued are presented in Table 2.
Table 2
Selection criteria for sets of AM Bench measurements
1
Interest and availability of measurement teams
2
Availability of resources
3
Ability to measure key factors for simulations
4
Relevance to industrial applications
5
Expansion of AM Bench scope
6
Extensions to existing AM Bench datasets
7
Interest by the community
The first three criteria are requirements for pursuing a given set of benchmarks. The first criterion is interest and availability of measurement teams, which stems from the fact that AM Bench works with many internal and external measurement teams. The second criterion, availability of resources, stems largely from the ability of a given measurement team to acquire the resources needed to plan, execute, and disseminate benchmark measurements that meet the stringent AM Bench quality requirements. The third criterion is the ability to measure key factors (e.g., inputs, outputs, and boundary conditions) for the targeted simulation areas with quantified uncertainties.
The remaining four criteria are not requirements, but instead are additional factors that must be considered in the selection process. The fourth criterion, relevance to industrial applications, is particularly significant. The AM Bench mission is to promote US innovation and industrial competitiveness in AM by providing open and accessible benchmark measurement data for guiding and validating predictive AM simulations across the full range of industrially relevant AM processes and materials. Thus, industrial relevance is core to our mission, and we strive to maintain regular interaction with our industrial stakeholders. It is recognized, however, that AM modeling and simulation development is strongly propelled by academic research, and we aim to help bridge the gap from fundamental research to industrial applications. The fifth criterion is expansion of the AM Bench scope. Sets of benchmark measurements that expand the AM Bench datasets to new materials or processes are generally favored, while considering the fourth goal of industrial relevance. The sixth criterion refers to direct extensions to existing AM Bench datasets. For example, benchmark AMB2018-01 from the original round of benchmarks included 3D builds of nickel alloy 625, with extensive in situ characterization during the build, residual elastic strain measurements, part deflection, and microstructural characterization. At that time, there were insufficient resources and time to complete comprehensive mechanical property measurements, so a build plate of specimens was reserved for the next round of AM Bench. For AM Bench 2022, a set of mechanical-properties benchmark measurements was conducted that included the reserved 2018 specimens, producing an integrated set of processing–structure–properties datasets for these builds. The final criterion in Table 2 is interest by the AM Bench community. Interest is explored using several different mechanisms, including queries to the AM Bench scientific advisory committee, direct interaction with the AM modeling community, and feedback acquired during the question-and-answer sessions at the AM Bench conferences.
Most AM Bench benchmarks follow a nominal three-year schedule. AM Bench 2022 was originally scheduled for summer of 2021 but was delayed by a year due to the COVID-19 pandemic. The next round of regular benchmark measurements is scheduled for summer of 2025. In addition to these regularly scheduled sets of benchmarks, AM Bench now supports asynchronous benchmarks that are not tied to the regular benchmark schedule. Such asynchronous benchmarks provide increased flexibility in supporting the needs of the modeling community. The first asynchronous benchmark included simultaneous time-resolved laser absorptivity and high-speed radiography of the laser melt pool for both Ti alloy and Al alloy targets. This set of benchmarks was completed in 2022.
Any research groups that are interested in participating in future AM Bench measurements are strongly encouraged to contact the AM Bench organizers who are listed on the AM Bench website [15].

Overview of the 2022 Measurements and Modeling Challenges

AM Bench 2022 included five sets of metals benchmark measurements and two sets of polymers benchmarks as part of the regular three-year schedule along with one asynchronous set of metals benchmarks also completed in 2022. The eight sets of benchmarks along with information on the materials used and the AM build methods targeted are listed in Table 3.
Table 3
Materials and targeted build methods for 2022 benchmarks
Benchmark designation
Material class/materials
Targeted build method
AMB2022-01
Metal alloy/nickel alloy 718
LPBF
AMB2022-02
Metal alloy/nickel alloy 718
LPBF
AMB2022-03
Metal alloy/nickel alloy 718
LPBF
AMB2022-04
Metal alloy/nickel alloy 625
LPBF
AMB2022-05
Metal alloy/nickel alloy 625
LPBF
AMB2022-06
Polymers/polycarbonate
Thermoplastic material extrusion
AMB2022-07
Polymers/methacrylate- and acrylate-based resins
Vat Photopolymerization
A-AMB2022-01
Metal alloy/Ti-6Al-4 V /AA5182
LPBF
For each set of benchmarks, challenge problems were released to the AM modeling community along with any experimental conditions and calibration data necessary to develop the corresponding models. Details of the challenge problems can be found on the AM Bench website [15]. In addition, question-and-answer webinars for all of the sets of challenge problems were held and recorded. Links to the recorded webinars are provided in Table 4.

AMB2022-01, AMB2022-02, and AMB2022-03

The three sets of benchmarks AMB2022-01, AMB2022-02, and AMB2022-03 are closely integrated sets of metal LPBF benchmark measurements that directly connect 3D builds (AMB2022-01 and AMB2022-02) to measurements on bare metal plates using single laser tracks and two-dimensional (2D) scan patterns (pads, AMB2022-03) that match the laser scan patterns used for the 3D builds. The intent is to provide a highly quantitative dataset that seamlessly spans the full range from feedstock characterization to microstructure of the as-built and heat-treated final components. The material for all three sets of benchmark measurements is nickel alloy 718. All these AM builds and in situ build measurements used the NIST additive manufacturing metrology testbed (AMMT) [16] which underwent substantial upgrades in preparation for these measurements.

AMB2022-01

AMB2022-01 uses a geometry similar to the earlier AMB2018-01 set of benchmarks with the addition of thinner wall regions internal to leg L10 (see Fig. 1 and Fig. 2). AMB2022-01 includes feedstock characterization, 3D builds with in situ process monitoring, residual elastic strain measurements, deflection measurements after partial cutting off the build plate, a two-step heat treatment, and extensive microstructural characterization of the as-built and heat-treated parts. Similar measurements were conducted for the earlier AMB2018-01 benchmarks using nickel alloy 625, but major improvements were made in almost all facets of the present work. Follow-up mechanical property measurements are planned for the next iteration of AM Bench, providing data spanning the full processing–structure–properties range.
The geometry of the AMB2022-01 bridge artifacts built using LPBF on the AMMT is shown in Fig. 1, the build parameters that were used are shown in Table 5, and an expanded view of the scan strategy that was used to build the thin walls within leg L10 is shown in Fig. 2. The laser diameter is specified as D4σ which is four standard deviations of the laser power distribution function. Please note that the part orientation on the base plate is rotated by 180° about the z-axis with respect to that used for AMB2018-01. The field of view (FOV) for the in situ thermography measurements is shown in red in Fig. 1 and Fig. 2. Each build plate included four bridge artifacts and the FOV only included one of these artifacts. The geometry of the build plates along with the part numbering system and a photograph of build plate #7 are shown in Fig. 3. Full details about the AMB2022-01 builds, scan strategy, and in situ measurements may be found in the corresponding AM Bench 2022 measurement paper [17].
Table 5
AMMT build settings used for AMB2022-01
AMMT build settings
Laser power
285 W
Laser speed
960 mm/s
Hatch spacing
110 µm
Layer thickness
40 µm
Scan pattern
X–Y rotation between odd and even layers, segmented
Laser diameter
75 µm D4σ
A list of the primary sets of measurements that were conducted as part of AMB2022-01, the sample conditions, the corresponding challenge problems, and the associated measurement papers are shown in Table 6. The challenge problems for AMB2022-01 are:
  • Time Above Melting Temperature (CHAL-AMB2022-01-TAM): Time above the midpoint between the solidus and liquidus temperatures for the longest part of the melt pool at specified locations within the build volume. This metric is closely related to melt pool length but is explicitly location specific.
  • Solid Cooling Rate (CHAL-AMB2022-01-SCR): Cooling rate immediately following complete solidification (below solidus) at specified locations within the build.
  • Residual Elastic Strains (CHAL-AMB2022-01-RS): Residual elastic strain components at select locations internal to the bridge structure, corresponding to synchrotron X-ray diffraction measurements.
  • Part Deflection (CHAL-AMB2022-01-PD): Deflection of the as-built (no heat treatment) bridge structure after it is partially separated from the build plate.
  • Microstructure (CHAL-AMB2022-01-MS): Histograms of direction-specific grain sizes from specified regions within as-built and heat-treated samples.
  • Phase Evolution (CHAL-AMB2022-01-PE): Formation and evolution of phases and phase fractions, including major precipitates, as a function of time for heat treatments of nickel alloy 718 from a 2.5-mm leg.
Table 6
Primary measurements conducted as part of AMB2022-01
Measurements
Sample condition
Challenge problem(s)
Measurement paper
In situ thermography
During build process
CHAL-AMB2022-01-TAM
[17]
CHAL-AMB2022-01-SCR
Build plate and chamber temperatures
During build process
Not part of challenge problem
[17]
Synchrotron X-ray diffraction residual elastic strain
As-built condition
CHAL-AMB2022-01-RS
[18]
Neutron diffraction residual elastic strain
As-built condition
Not part of challenge problem
[19]
Contour method residual stress
As-built condition
Not part of challenge problem
[19]
Part deflection for partial cut off build plate
As-built condition
CHAL-AMB2022-01-PD
[20]
Large area 2D electron backscatter diffraction (EBSD)
As-built condition
CHAL-AMB2022-01-MS
[21]
Large area 2D EBSD
Fully heat-treated condition
CHAL-AMB2022-01-MS
[21]
Transmission electron microscopy (TEM) imaging and selected area diffraction
As-built condition
CHAL-AMB2022-01-PE
[22]
TEM imaging and selected area diffraction
Fully heat-treated condition
CHAL-AMB2022-01-PE
[22]
Synchrotron small-angle X-ray scattering and diffraction
In situ during two stage heat treatment
CHAL-AMB2022-01-PE
[22]
3D serial sectioning with EBSD
As-built condition
Not part of challenge problem
[23]
X-ray computed tomography
As-built condition
Not part of challenge problem
[23]
These challenge problems represent only a small fraction of the available measurement results. For example, residual strain/stress measurements were conducted using three independent measurement approaches, including synchrotron X-ray diffraction, neutron diffraction, and mechanical release (referred to here as the contour method), but only the X-ray diffraction measurement results were used for the challenge problems. All the measurement data may be found in the referenced publications and on the AM Bench data management systems that are described in a later section. Full details of the AMB2022-01 challenges may be found on the AM Bench website [15].

AMB2022-02

AMB2022-02 includes additional LPBF builds using the same geometry as AMB2022-01. This set of measurements explores the effect of different laser scan patterns on the local melt pool behavior. Another set of measurements that is being considered for the next iteration of AM Bench includes residual elastic strain characterization and part deflection after partial cutting off the base plate of the parts built in AMB2022-02.
The different scan strategies for the three AMB2022-02 3D builds are outlined in Table 7. ‘Alternate scan stacking,’ used in builds V6 through V8, has the laser scan direction rotate 90° between layers similar to AMB2022-01 builds, but additionally offsets each scan vector by ½ the hatch spacing so as not to directly overlap scan vectors in subsequent layers. Furthermore, V7 tests the possibility of pre-sintering the powder layer by exposing every other scan vector with lower laser power, and V8 tests pre-sintering by exposing the surface with diagonal scans at lower laser power and spot size, followed by the standard parameters. Full details about the AMB2022-02 builds, scan strategies, and in situ measurements may be found in the corresponding AM Bench 2022 measurement paper [17].
Table 7
Scan strategy descriptions for the three builds V6, V7, and V8 conducted as part of AMB2022-02
Build #
Scan strategy description
V6
Constant laser power, alternate scan stacking
V7
‘Interleaved’ scans with alternating laser power of 171 W and 285 W between tracks
V8
Pre-sintering (diagonal scans) at 85.5 W and 230 µm spot size, followed by standard 285 W
A list of the primary sets of measurements that were conducted as part of AMB2022-02, the sample conditions, the corresponding challenge problems, and the associated measurement papers are shown in Table 8. The challenge problems for AMB2022-02 are:
  • Time Above Melting Temperature (CHAL-AMB2022-02-TAM): Time above the midpoint between the solidus and liquidus temperatures for the longest part of the melt pool at specified locations within the build volume for the different laser scan patterns. This metric is closely related to melt pool length but is explicitly location specific.
  • Solid Cooling Rate (CHAL-AMB2022-02-SCR): Cooling rate immediately following complete solidification (below solidus) at specified locations within the build for the different laser scan patterns.
  • Part Deflection (CHAL-AMB2022-02-PD): Deflection of the as-built (no heat treatment) bridge structure after it is partially separated from the build plate. NOTE: These measurements were not completed.
Table 8
Primary measurements conducted as part of AMB2022-02
Measurements
Sample condition
Challenge problem(s)
Measurement paper
In situ thermography
During build process
CHAL-AMB2022-02-TAM
[17]
CHAL-AMB2022-02-SCR
Full details of the AMB2022-02 challenges may be found on the AM Bench website [15].

AMB2022-03

The AMB2022-03 benchmarks explore a range of individual and overlapping melt pool behaviors using individual laser tracks and 2D arrays of laser tracks on solid metal nickel alloy 718 plates. For the individual laser tracks, a range of laser parameters were used, with variations in laser power, speed, and spot diameter. All processing and in situ measurements were conducted using the AMMT. In situ measurements include time-resolved laser coupling, location-specific liquid and solid cooling rates, and location-specific time above melting. Ex situ measurements include 3D topography of the solidified laser tracks, cross-sectional geometry, and microstructure measurements. For the 2D pads, scan patterns and timing match those used for the even layers (X-pads) and odd layers (Y-pads) corresponding to Leg 9 (L9) of the AMB2022-01 3D builds. The same sets of in situ and ex situ measurements used for the single laser tracks were used for the laser pad studies.
Examples of bare plate samples with different laser scan patterns and orientations are shown in Fig. 4. Samples were mounted in groups of six within the AMMT on a custom holder and each sample had a k-type thermocouple that contacted the center underside.
The baseline parameters for the single laser tracks utilized 285 W laser power, 960 mm/s scan speed, and 67 μm D4σ spot size. Many other detailed factors pertaining to the gas flow, environment, surface preparation, etc., are discussed in [24]. Additionally, substrate temperatures were measured during laser scanning of both tracks and pads. The offset laser parameters for all seven parameter sets for the single laser tracks are listed in Table 9.
Table 9
Offset laser parameters for AMB2022-03 individual laser track thermography measurements with varying laser power, spot size, and scan speed. Each case is repeated three times for a total of 21 laser scan tracks. Volumetric energy density (VED) based on 1σ laser beam size, VEDσ, is also provided
 
Case number
Laser power [W]
Scan speed [mm/s]
Spot size, D4σ [μm]
VEDσ = P/v/σ2 [J/mm3]
VED/VEDbase
Baseline
0
285
960
67
1058
1.00
Change spot
1.1
285
960
49
1978
1.87
1.2
285
960
82
706
0.67
Change speed
2.1
285
1200
67
847
0.80
2.2
285
800
67
1270
1.20
Change power
3.1
325
960
67
1207
1.14
3.2
245
960
67
910
0.86
A list of the primary sets of measurements that were conducted as part of AMB2022-03, the sample conditions, the corresponding challenge problems, and the associated measurement papers are shown in Table 10. Full details of the AMB2022-03 challenges may be found on the AM Bench website [15]. The challenge problems for AMB2022-03 are:
  • Track Solid Cooling Rate (CHAL-AMB2022-03-TSCR): Cooling rate immediately following complete solidification (below solidus) at the center of each track for all processing conditions.
  • Track Liquid Cooling Rate (CHAL-AMB2022-03-TLCR): Liquid cooling rate immediately before start of solidification (above liquidus) at the center of each track for all processing conditions.
  • Track Time Above Melt (CHAL-AMB2022-03-TAM): Time above the midpoint between the solidus and liquidus temperatures for single laser-scanned tracks.
  • Track Melt Pool Geometry (CHAL-AMB2022-03-TMPG): The laser track width and depth near the center of each track for all processing conditions.
  • Pad Solid Cooling Rate (CHAL-AMB2022-03-PSCR): Cooling rate immediately following complete solidification (below solidus) at the center of each track for all processing conditions.
  • Pad Time Above Melting (CHAL-AMB2022-03-PTAM): Time above the midpoint between solidus and liquidus temperature for specified locations of the X-Pads and Y-Pads.
  • Pad Melt Pool Geometry (CHAL-AMB2022-03-PMPG): Laser track depth and geometrical measurements describing the overlapping laser tracks near the center and near the edge of both X-Pads and Y-Pads.
Table 10
Primary measurements conducted as part of AMB2022-03
Measurements
Sample condition
Challenge problem(s)
Measurement paper
In situ thermography
During laser scans
CHAL-AMB2022-03-TSCR
[24]
CHAL-AMB2022-03-TLCR
CHAL-AMB2022-03-TAM
CHAL-AMB2022-03-PSCR
CHAL-AMB2022-03-PTAM
Plate temperature
During laser scans
Not part of challenge problem
[24]
Optical microscopy
Cross sections
CHAL-AMB2022-03-TMPG
[25, 26]
CHAL-AMB2022-03-PMPG
Large area 2D EBSD
Cross sections
Not part of challenge problem
[26]
Large area 2D EDS
Cross sections
Not part of challenge problem
[26]

AMB2022-04

AMB2022-04 extends the earlier AMB2018-01 nickel alloy 625 measurements by providing mechanical property data for the as-built material. When the original AMB2018-01 specimens were produced, a full build plate of four bridge specimens was reserved for future use. For AMB2022-04, a set of compression measurements was conducted from test specimens extracted from the front section of a bridge specimen (see Fig. 5). Additional samples were obtained from a 5-mm leg for in situ diffraction measurements at the Cornell High Energy Synchrotron Source (CHESS) during compression testing. Two additional build plates of parts designed for mechanical testing were fabricated using the same build machine as AMB2018-01, the same alloy (different powder lot numbers), and the same (and related) bulk material scan pattern. These parts were used for macroscopic and mesoscopic mechanical testing, with additional characterization provided by X-ray computed tomography (XRCT) and scanning electron microscopy (SEM). All the mechanical test specimens were measured in the as-built state, with no residual stress anneal.
A list of the primary sets of measurements that were conducted as part of AMB2022-04, the sample conditions, the corresponding challenge problems, and the associated measurement papers are shown in Table 11. Full details of the AMB2022-04 challenges may be found on the AM Bench website [15]. The challenge problems for AMB2022-04 are:
  • Subcontinuum Mesoscale Tensile Test (CHAL-AMB2022-04-MeTT): Predict subcontinuum stress strain behavior, fracture location, and width reduction of as-built nickel alloy 625 mesoscale specimens.
  • Macroscale Tensile Tests at Different Orientations (CHAL-AMB2022-04-MaTTO): Predict bulk/continuum stress strain behavior of as-built nickel alloy 625 tensile specimens at different specimen tensile axis orientations with respect to the build direction.
  • Macroscale Compression at Different Temperatures and Orientations (CHAL-AMB2022-04-MaCTO): Predict bulk/continuum stress strain behavior of as-built nickel alloy 625 compression specimens at different specimen compression axis orientations with respect to the build at three temperatures.
Table 11
Primary measurements conducted as part of AMB2022-04
Measurements
Sample condition
Challenge problem(s)
Measurement paper
Compression testing at different temperatures and strain rates
As-built AMB2018-01 specimens
CHAL-AMB2022-04-MaCTO
[27]
Compression testing at different temperatures and strain rates with in situ diffraction
As-built AMB2018-01 specimens
Not part of challenge problem
[27]
ASTM-compliant tensile testing with digital image correlation
As-built 2022 specimens
Not part of challenge problem
[28]
Tensile tests, electron backscatter imaging, EBSD, high energy X-ray diffraction, XRCT
As-built 2022 specimens
CHAL-AMB2022-04-MeTT
[29]
Tensile tests, electron backscatter imaging, EBSD, high energy X-ray diffraction, XRCT
As-built 2022 specimens
CHAL-AMB2022-04-MaTTO
[30]

AMB2022-05

AMB2022-05 is another direct extension to the measurement data provided by AMB2018-01. Although the 2018 set of benchmarks included extensive measurements of the as-built microstructure, interactions with the AM modeling community made it clear that additional microstructure characterization data would prove useful. Here, we extend the previous microstructure data to include three additional datasets: 1) large-area SEM characterization of the midplane of the bridge specimen, including part of the baseplate, of a 0.5-mm leg, and one side of a 5.0-mm leg, 2) multiple SEM cross sections of a complete test artifact parallel to the baseplate, and 3) a 3D microstructure measurement obtained through serial sectioning with optical and SEM imaging. The region of interest (ROI) for the 3D microstructure measurement measures approximately 500 µm × 500 µm × 750 µm along the build X, Y, Z directions, respectively, including one corner of a 2.5-mm leg extending into the baseplate. This is the same ROI used for the 3D microstructure measurements included in AMB2022-01. As described previously for AMB2022-04, these measurements used nickel alloy 625 bridge specimens that were reserved from AMB2018-01.
A list of the primary sets of measurements that were conducted as part of AMB2022-05, the sample conditions, the corresponding challenge problems, and the associated measurement papers are shown in Table 12. Because the earlier set of benchmarks, AMB2018-01, already included microstructure challenges from this same set of samples, AMB2022-05 only includes one challenge problem that uses the new measurement data. Full details of the AMB2022-05 challenge may be found on the AM Bench website [15]. The challenge problem for AMB2022-05 is:
  • Microstructure (CHAL-AMB2022-05-MS): Histograms of direction-specific grain sizes from specified regions within an as-built specimen.
Table 12
Primary measurements conducted as part of AMB2022-05
Measurements
Sample condition
Challenge problem(s)
Measurement paper
Large area 2D EBSD of sample midplane
As-built AMB2018-01 specimens
CHAL-AMB2022-05-MS
[27]
Large area 2D EBSD of slices parallel to baseplate
As-built AMB2018-01 specimens
Not part of challenge problem
[27]
3D serial sectioning with EBSD
As-built AMB2018-01 specimens
Not part of challenge problem
[23]

AMB2022-06

The goal of this thermoplastic material extrusion (MatEx) AMB2022-06 set of benchmark measurements is to elucidate the relationship between material properties (viscoelasticity), print parameters (temperature and flow rate), and the resultant single-layer part dimensions (width and cross section) of an amorphous polymer. The primary objectives of the corresponding challenge problems are to predict the width and cross-sectional shape of the printed material. Measurement data for model calibration and challenge comparison are provided through material characterization (linear rheology) and system calibration (extrudate temperature).
Full details for AMB2022-06 may be found on the AM Bench website [15]. Briefly, commercial bisphenol-A-polycarbonate filament (Tg = 154 °C, Mn = 30.1 kg/mol, Mw = 68.7 kg/mol) was used to prepare the samples. Here, Mn is the number average relative molecular mass and Mw is the mass average relative molecular mass. The filament is nominally 2.85 mm, with actual dimensions of 2.84 ± 0.03 mm. The sample is two layers high, one extrudate wide, and 150 mm long, with one layer high, 1 cm by 1 cm adhesion feet at either end. Figure 6 shows a diagram of the two-layer structure. The build plate is borosilicate glass painted with a thin layer of polyvinyl alcohol water-soluble adhesive. Print head translation speeds are 10 mm/s and 100 mm/s. The bed set point is 145 °C with a measured surface temperature of 118.3 °C ± 0.1 °C. Extrusion setpoints are 240 °C, 260 °C, 280 °C, 300 °C, 320 °C, and 340 °C. However, the actual extrusion temperature is 50 °C–60 °C lower. Measurement and stability of the volumetric flow rate proved unreliable, so publication of the AMB2022-06 results is not currently planned, and work is underway to correct the deficiencies.
A list of the primary sets of measurements that were conducted as part of AMB2022-06, the sample conditions, and the corresponding challenge problems are shown in Table 13. Full details of the AMB2022-06 challenges may be found on the AM Bench website [15]. The challenge problems for AMB2022-06 are:
  • Road width (CHAL-AMB2022-06-EW): The maximum and minimum horizontal width of a single-printed road.
  • Road Dimensionless Cross Section (CHAL-AMB2022-06-DCS): Dimensionless shape of the cross section of a single road.
  • Road Cross Section (CHAL-AMB2022-06-CS): Shape of the cross section with dimensions.
Table 13
Primary measurements conducted as part of AMB2022-06
Measurements
Sample condition
Challenge problem(s)
Measurement paper
XRCT
Cooled polycarbonate extrudates
CHAL-AMB2022-06-EW
None
CHAL-AMB2022-06-DCS
CHAL-AMB2022-06-CS
Linear rheology
As received filament
Not part of challenge problem
None

AMB2022-07

AMB2022-07 includes vat photopolymerization measurements of cure depth and print fidelity as a function of exposure duration, photopattern dimensions, and resin characteristics [31]. These measurements allow users to validate models for the relationship between photopattern exposure duration and resultant single-layer part dimension (i.e., fidelity and cure depth) for four resins, which serve to orthogonally probe the relationship between resin reactivity, k, and viscosity, ν. The primary objectives are to determine how the dimensions of a photomask and resin viscosity and reactivity affect cure depth and print fidelity. Ultimately, the improved understanding and prediction of these relations will foster enhanced print resolution and part performance. Experimental data for model calibration and challenge comparison is provided through cure profile measurements using laser confocal scanning microscopy in conjunction with resin characterization (i.e., Fourier transform infrared spectroscopy, oscillatory rheometry) and system calibration (e.g., photomask dimensions, beam profilometry, radiometry). Released calibration measurements were performed on four, open-source resins to serve as representative examples for resins in the field. A high-level look at the measurement results is given in Fig. 7 where the measured cure depth is shown as a function of photomask dimensions and exposure for all four resins.
A list of the primary sets of measurements that were conducted as part of AMB2022-07, the sample conditions, the corresponding challenge problems, and the associated measurement papers are shown in Table 14. Full details of the AMB2022-07 challenges may be found on the AM Bench website [15]. The challenge problems for AMB2022-07 are:
  • Optical profile at print plane (CHAL-AMB2022-07-OP): Light intensity profile at the print plane.
  • Cure depth dependence on photomask dimensions (CHAL-AMB2022-07-CD): Cure depth as a function of photomask linewidth and resin characteristics.
  • Print profile (CHAL-AMB2022-07-PP): Solid cross section profile of the printed patterns as a function of photomask linewidth and resin characteristics.
Table 14
Primary measurements conducted as part of AMB2022-07
Measurements
Sample condition
Challenge problem(s)
Measurement paper
Beam profilometry and radiometry
N/A, optical source prediction
CHAL-AMB2022-07-OP
[31]
Laser scanning confocal microscopy
Post-processed cured resin
CHAL-AMB2022-07-CD
[31]
CHAL-AMB2022-07-PP

A-AMB2022-01

This asynchronous set of benchmark measurements provides simultaneous time-resolved laser coupling (laser absorptivity) measurements and high-speed radiography of the melt pool geometry during high-power laser interactions with a bare metal surface. The primary objective is to correlate the laser coupling and melt pool geometry for both a stationary and linearly scanned laser beam on samples of Ti-6Al-4 V and aluminum alloy 5182 (AA5182). These experiments were carried out at the Advanced Photon Source (APS) of Argonne National Laboratory. A diagram of the experiment is shown in Fig. 8. The Ti-6Al-4 V results were provided as calibration data for challenge problems that were focused on the AA5182 data. Some of these data have been published previously [32]. For general information about the X-ray imaging technique, please see publication [33].
A list of the primary sets of measurements that were conducted as part of A-AMB2022-01, the corresponding challenge problems, and the associated measurement papers are shown in Table 15. Full details of the A-AMB2022-01 challenges may be found on the AM Bench website [15]. Unlike the other challenge problems described in this section, the A-AMB2022-01 challenge problems were grouped into categories, with separate awards provided for each group. The challenge problems for A-AMB2022-01 are:
  • Laser coupling for a stationary laser spot
    o
    Spot Time-Dependent Absorption (CHAL-A-AMB2022-01-Spot-TDA): Predict the amount of laser light absorbed during laser irradiation for a laser spot as a function of time.
     
    o
    Spot Average Absorption (CHAL-A-AMB2022-01-Spot-AA): Predict the average absorption for a laser spot both before and during keyhole formation.
     
  • Melt pool geometry for a stationary laser spot
    o
    Spot Time-Dependent Width (CHAL-A-AMB2022-01-Spot-TDW): Predict the melt pool width for a laser spot as a function of time.
     
    o
    Spot Average Solidification Rate (CHAL-A-AMB2022-01-Spot-ASR): Predict the average solidification rate of the melt pool after laser off.
     
  • Laser coupling for a scanning laser spot
    o
    Scan Time-Dependent Absorption (CHAL-A-AMB2022-01-Scan-TDA): Predict the amount of absorbed laser light as a function of time during a linear laser scan.
     
    o
    Scan Average Absorption (CHAL-A-AMB2022-01-Scan-AA): Predict the average absorption during a scanned laser both before and during keyhole formation.
     
  • Melt pool geometry for a scanning laser spot
    o
    Scan Maximum Width and Depth (CHAL-A-AMB2022-01-Scan-MWD): Predict the maximum melt pool width and depth during a linear laser scan.
     
    o
    Average Solidification Rate (CHAL-A-AMB2022-01-Scan-ASR): Predict the average solidification rate after a linear laser scan has finished.
     
Table 15
Primary measurements conducted as part of A-AMB2022-01
Measurements
Challenge problem(s)
Measurement paper
Laser coupling
CHAL-A-AMB2022-01-Spot-TDA
[34]
CHAL-A-AMB2022-01-Spot-AA
CHAL-A-AMB2022-01-Scan-TDA
CHAL-A-AMB2022-01-Scan-AA
High-speed radiography
CHAL-A-AMB2022-01-Spot-TDW
[34]
CHAL-A-AMB2022-01-Spot-ASR
CHAL-A-AMB2022-01-Scan-MWD
CHAL-A-AMB2022-01-Scan-ASR

AM Bench Impact and 2022 Modeling Challenge Problem Submissions

The most significant long-term impact of AM Bench lies in the extensive measurement data that are permanently archived by NIST and freely available online. These datasets are being used to guide and validate AM simulations worldwide. In addition, the presence of these data directly impacts AM research directions. For example, ExaAM is a large multi-institution effort to develop high performance computing (HPC) codes1 for simulating LPBF processing, microstructure evolution, and mechanical behavior. The entire collaboration is designed around the AM Bench 2018 metals benchmarks because these were the most rigorous datasets available for validating the ExaAM simulations [35]. AM Bench is also working with other outside AM efforts on the idea of expanding the AM Bench datasets to include additional high-impact data that our customers may find useful for guiding and validating their AM simulations.
The AM Bench challenge problems provide both short- and long-term impacts to the AM community. The original idea behind the challenge problems was to provide snapshots in time of the AM modeling community’s ability to model critical aspects of AM processing, structure, properties, and performance. For example, comparisons between the 2018 and 2022 challenge problem submissions show significant progress in many areas and these will be highlighted in the next section. A longer-term impact of the AM Bench challenge problems is that the AM modeling community is using these challenges to help guide their research directions. This impact was clearly highlighted in the AM Bench 2022 discussion sessions and work is ongoing to modify the structure of the challenge problems to provide participants with additional time to develop new capabilities before the submission deadlines.
The AM Bench conference also provides impact to the AM community by bringing together a broad range of AM modeling and measurement researchers to exchange information and views on what models and measurements are most needed and to bring these often-disparate communities together. Close interaction between modeling and measurement efforts is needed for all AM material classes, build methods, and stages of the build and post-build processing.
Including the seven sets of 2022 benchmarks that were part of the regular three-year schedule and the asynchronous benchmark that concluded a few months earlier, AM Bench received a total of 138 challenge problem submissions from 19 groups around the world, which is a significant increase in the number of submissions when compared with the 48 submissions received in 2018. The geographical distribution of the groups, as determined by the home institution of the submitting author, included 15 from North America, one from Europe, and three from Asia. In all, 41 awards were presented during the AM Bench 2022 awards ceremony held on August 17, 2022, at the AM Bench 2022 conference.
The number of submissions received for each of the 34 modeling challenges in 2022 are shown in Fig. 9. For each set of benchmark measurements, the designation and short description is given in blue for the metal benchmarks and green for the polymer benchmarks. The first column of the figure is a graphic indicating the general topic of each set of benchmarks. The individual challenges are grouped within these sets of benchmark measurements. The second column of the figure is a brief description of each challenge problem, and the third column gives the corresponding identification code. The number of modeling submissions received for each challenge problem is shown in column four. Some of the sets of challenge problems received significantly more submissions than others, so we summed the more popular groups, and the results are shown in red in column four. Adding all the challenge problem submissions in Fig. 9 comes to 137 instead of the stated number of 138. This is because one of the groups submitted residual elastic strain results for AMB2022-02 even though the strain challenge was only issued for AMB2022-01.
An obvious feature in these data is the large difference in modeling challenge submissions between metals and polymers: 135 submissions for metals and just 2 for polymers. This difference mirrors a similar observation in 2018, where 45 submissions were received for metal challenge problems and just one for polymers. It is true that the scientific community continues to invest much more heavily in AM modeling research for metals than for polymers, but the magnitude of the submission difference is surprising. After AM Bench 2018, we retargeted the 2022 polymers benchmarks to address more lower-level processes and attempted to improve our outreach to that community, but these efforts did not affect the number of submissions. Moving forward, we will continue to work with the polymers AM community to ensure our efforts provide value. It is also worth reiterating an observation we made in 2018. The low number of polymer AM challenge submissions does not mean that these benchmarks are not valuable. Instead, as theoretical and computational work expands in this area, these benchmarks can provide the rigorous measurement data needed to guide and validate such models. It is also possible that the availability of the AM Bench polymer AM datasets will stimulate additional work on understanding and simulating polymer AM processes [5].
Looking at the metal challenges, we will start by examining the sets of benchmarks that received the fewest submissions: AMB2022-05 and AMB2022-02. As mentioned previously, AMB2022-05 is a direct extension to microstructure measurements conducted as part of AM Bench 2018 and it was included because of strong requests from the modeling community. However, the corresponding single challenge problem was considerably more difficult than any of the earlier microstructure challenges since it required the modeling teams to simulate and compare the location-specific microstructures at widely different positions within the build, with quantitative metrics. Modeling teams that contributed to the AM Bench 2018 microstructure challenges typically used highly simplified models assuming a single region of interest. Although the number of submissions for the 2022 challenge was low (just two), this is not indicative of the high value of these community-requested data for guiding and validating such simulations moving forward; rather, this reflects the complexity of real-world microstructures relative to the simplifications and assumptions often required in computational models.
AMB2022-02 also requires considerable modeling sophistication. Two of the challenges required participants to model details of the melt pool thermal behavior from specified sample locations at eight different heights (Z axis) during the build. Although similar requirements existed for the AMB2022-01 set of benchmarks, these challenges required quantitative comparisons between many such simulations for a range of different laser parameters and scan patterns. Given that solutions must be submitted just three months after the challenge problems are presented to the AM community, it is not surprising that the number of submissions was relatively low. As mentioned in the AM Bench Discussion Summary section, several modelers strongly requested that this three-month interval be significantly increased. This issue will be addressed for future AM Bench challenge problems.
AMB2022-04 is another direct extension to measurements conducted as part of AM Bench 2018, focusing on mechanical performance of 3D builds of nickel alloy 625 test specimens. When combined with the earlier 2018 measurements, AMB2022-04 completes a comprehensive set of measurements for a single system that spans the full range of processing–structure–properties. Three of the challenge problem submissions focus on the Macroscale Compression at Different Temperatures and Orientations set of measurements that used samples reserved for future use in 2018. However, more submissions focused on stand-alone challenges that used new AM builds produced with the same build machine and alloy. Looking to the future, a similar mechanical-properties extension to the 2022 AMB2022-01 measurements is planned, and a full build plate of bridge structures has been reserved for that work.
AMB2022-01 is the primary set of benchmarks focusing on 3D builds of nickel alloy 718 and the number of challenge problem submissions ranged from 2 to 4 for each of the six challenges. The same observations made previously for AMB2022-02 and AMB2022-05 are also true here. Many of the included challenge problems are considerably more difficult than those that were posed for the 3D builds in AMB2018-01, requiring quantitative predictions from widely separated regions of the as-built parts. The two-step heat treatment further complicated the microstructure-based challenge problems. We are highly encouraged that multiple groups submitted challenge results for all of these challenge problems.
The large number of challenge problem submissions for AMB2022-03 (40 submissions) and A-AMB2022-01 (56 submissions) is in line with an observation made for AM Bench 2018 [5] that there was strong interest in low-level processes. Both of these popular sets of 2022 benchmarks and challenge problems deal with the interaction of a laser with bare metal plates. This popularity suggests that the AM modeling community believes that understanding the underlying melt pool physics is a prerequisite for reliable modeling of complete parts.

Comparisons Between AM Bench Measurements and Challenge Problem Submissions

Given the large number of challenge problems and challenge problem submissions, it is not feasible to provide a comprehensive comparison between AM Bench 2022 challenge problem submissions and corresponding measurement results in this summary document. Instead, more complete comparisons will be included in the individual measurement papers and the invited modeling papers that are part of this special issue of Integrating Materials and Manufacturing Innovation (IMMI). Some general observations are provided in Table 16.
Table 16
General observations about the AM Bench 2018 and 2022 challenge problem submissions
AM Bench 2018
AM Bench 2022
Strong interest in low-level processes
Strong interest in low-level processes
For 2/3 of the metal challenges, at least one submission was qualitatively close to the measured solution (trends and general shape)
Nearly all metal challenges had submissions that had good quantitative agreement with the measured solution
Wide range of predictions
Close competition with most simulations reasonably close to the measured solution
Simulations incorporating the most physics tended to be most successful
Simulations incorporating the most physics tended to be most successful
Table 16 includes corresponding general comments about the submissions for the 2018 and 2022 sets of metal challenge problems. The first observation has already been mentioned; there was strong interest in modeling the low-level processes. This interest is particularly evident for the AMB2022-03 set of benchmarks that included single laser tracks on bare metal plates along with arrays of adjacent laser tracks (pads) on bare metal plates. The single laser tracks can be thought of as the smallest problems that incorporate most of the basic physics of the laser-substrate interaction (a unitary process). The pad challenge problems are just one step closer to real 3D builds by including the overlaps between adjacent laser tracks along with possible interactions of the laser with plumes from previous tracks. Looking at the number of submissions in Fig. 9, we see that the simpler single-track challenges received more than twice the average number of submissions as the pad challenges. This difference may reflect the modeling community’s areas of interest, but another possibility is that the short 3-month time frame between the announcement of the challenge problems and the submission deadline limited the number of groups that could complete the pad simulations on time. This issue will be discussed in the Conclusions and Future Directions section.
The second general observation about the challenge problem submissions is that the solutions submitted in 2022 were significantly improved over those received in 2018. To put these improvements in perspective, some of the challenges cannot be directly compared since they involved completely new topical areas such as mechanical performance (AMB2022-04) and melt pool geometry with laser coupling (A-AMB2022-01). Other 2022 challenges were directly comparable to those from 2018, such as the residual strain and part deflection challenges, while others included additional levels of difficulty, such as the laser pads on bare metal plates described previously. Another difference between the 2018 and 2022 challenge problems is the introduction of quantitative metrics with prescribed submission templates that allowed more rigorous comparison between the simulation and measurement results.
For all cases where direct, one-to-one comparisons can be made between the 2018 and 2022 challenges, the newer results were substantially improved. For example, the best residual strain predictions from 2018 show good qualitative agreement with the measured strain distributions. In contrast, the best 2022 predictions were nearly indistinguishable from the measured results. The measured elastic strain in the vertical (Z) direction (top) compared with submitted 1st place (middle) and 2nd place (bottom) simulations is shown in Fig. 10. The submitted results were compared to the energy dispersive synchrotron X-ray diffraction measured XX and ZZ elastic strain maps. The modeled submissions were evaluated using the following metrics: difference in mean, root-mean-squared error (RMSE), and qualitative visual inspection. The mean elastic strain values were calculated and compared between the measured and modeled results for the XX and ZZ components. These delta-mean values were then used to calculate the mean-corrected modeled results. This way, only the variations around the mean were used to calculate the RMSE. The delta-mean and RMSE between the top two submissions were extremely close with only a 10% difference between the two. Visually, these model submissions captured the spatial variations in elastic strains extremely well compared with the measured results.
More generally, although all but one of the 2018 metal challenge problems received submissions, only 2/3 of these challenges received at least one submission judged to be qualitatively close to the measured solution (trends and general shape). The submitted solutions were also widely distributed, making selection of winners fairly simple. In contrast, many of the submitted solutions for the 2022 challenges were more tightly distributed and close to the measured solutions; quantitative metrics were often the only way of determining first, second, and honorable mention awards. The significant improvement in the simulation results from 2018 to 2022 demonstrates that rapid progress has been made in simulating these AM processes.
A final observation again matches one made in 2018. Models that incorporate the most physics tend to be most successful. This result is certainly not surprising, but there are exceptions. For example, the team that won first place for the mechanical property challenge on Macroscale Tensile Tests at Different Orientations (CHAL-AMB2022-04-MaTO) used coupled physics-based analytical solutions with parameters obtained from extensive experience with nickel-based superalloys. All of the other submitting teams used sophisticated crystal plasticity codes. While both methods build upon assumptions that are physically motivated, the crystal plasticity codes, in principle, can better match the measurements using a single parameter set as they account for the underlying microstructure and crystallographic texture. In this case, deep familiarity with the material system enabled the analytical approach to produce better predictions than the crystal plasticity approach.

AM Bench Data Management Systems

AM Bench provides multiple systems and pathways for users to access, download, search, and analyze AM Bench data and metadata. The AM Bench data management systems were completely redesigned for AM Bench 2022. The older AM Bench 2018 systems are still being maintained, but ultimately, the 2018 data and metadata will be transferred to the new systems.
AM Bench 2022 data and metadata access is being supported through a suite of integrated tools providing several different capabilities that target different user needs:
  • AM Bench Website—best source for information and data links concerning the AM Bench measurements, data, challenge problems, and conference series.
  • NIST Public Data Repository (PDR)—primary access to all public AM Bench measurement data, including processed and calibrated results.
  • Traditional Journal Articles—published in the journal Integrating Materials and Manufacturing Innovation, within the thematic section: AM Bench 2022.
  • Measurement Catalog—searchable sample and measurement characterization metadata with linked access to associated measurement datasets.
  • SciServer—public platform for users to provision a workspace with compute and storage resources for running supplied data analysis and user-developed codes with direct access to a full mirror volume of AM Bench measurement datasets.
  • AM Bench GitHub—AM Bench users will be able to share models and codes that can run on the AM Bench SciServer or at their home institution.

AM Bench Website

This website [15] includes detailed background material on AM Bench, descriptions of all AM Bench measurements and challenge problems along with informational videos from the measurement teams, links to all public AM Bench data and metadata, schedules for upcoming AM Bench events, lists of the 2018 and 2022 AM Bench challenge problem award winners, and descriptions of the data management systems.

NIST Public Data Repository

All AM Bench public data are stored on the NIST Public Data Repository (PDR) and may be accessed through the NIST Science Data Portal < https://​data.​nist.​gov > , which provides a user-friendly discovery and exploration tool for publicly available datasets at NIST. This portal is designed with FAIR principles (Findable, Accessible, Interoperable, and Reusable) [36] and best practice for Federal Data Strategy. For FAIR use of these data, please include the citation provided on the PDR homepage in any published works, including the digital object identifier (DOI). NIST DOIs are registered with the DataCite organization and provide globally unique persistent identifiers. The DOI also serves as a direct link to data homepages giving access to the full research publication description and underlying datasets. Direct links to the various AM Bench datasets in the PDR are provided on the data access pages of the AM Bench website.

Measurement Catalog

Measurement data are incomplete without the critical metadata describing the measurement instrument, instrument configuration, calibration, sample details, analysis methods, and many other factors. The AM Bench metadata are curated using the NIST Configurable Data Curation System (CDCS) [37]. The CDCS provides a method for capturing, sharing, and transforming unstructured data into a structured format based on the Extensible Markup Language (XML). Data and metadata are organized using AM Bench-developed templates encoded in XML schema to create searchable data documents that are saved in a non-relational (NoSQL) document database.
AM Bench users can access the AM Bench CDCS at < https://​ambench2022.​nist.​gov > . Documentation on using the CDCS may be found through the project website < https://​www.​nist.​gov/​programs-projects/​configurable-data-curation-system-cdcs > .

SciServer

Some of the AM Bench datasets are large (> 1 TB) and may require processing to extract desired quantities. Since it is impractical to require all AM Bench users to download such large datasets and to develop all their own codes for extracting meaningful results, AM Bench provides server-side processing through SciServer. SciServer is operated by the Institute for Data Intensive Engineering and Science (IDIES) at Johns Hopkins University (JHU) and is funded by the National Science Foundation through its Data Infrastructure Building Blocks (DIBBs) program. AM Bench users can apply for free virtual machines that include Jupyter notebooks and pre-installed software packages for AM Bench data analysis. A mirror of the AM Bench public measurement data on the PDR is maintained on the SciServer platform and search features are available.
AM Bench users can register for a SciServer account at < https://​sciserver.​org > . General help may be found at < https://​www.​sciserver.​org/​support > . Once a user has registered, they will be able to access a data volume containing AM Bench data through the Manufacturing Science Domain. For detailed instructions, please see data management descriptions on the AM Bench website [15].

AM Bench GitHub

As mentioned previously, many of the AM Bench datasets can be processed and analyzed to obtain valuable information for validating model predictions and exploring connections between disparate phenomena. Although some pre-written analysis codes are provided, it is expected that some AM Bench users will need to develop their own codes and algorithms for exploring these datasets. We are providing a public AM Bench GitHub where the AM Bench users can share codes, strategies, and results. Information on accessing this GitHub may be found on the AM Bench website [15].

AM Bench Discussion Summary

Discussion sessions were an invaluable part of the AM Benchmark 2022 conference. In addition to two discussion sessions held as part of the main AM Bench 2022 conference, a third discussion session was included in the embedded workshop on Q&C. The main conference discussion sessions will be described here, and the Q&C discussion will be described in the Q&C section.
The main conference discussion sessions were held to solicit lessons learned from the 2022 test series and stimulate ideas for future benchmark measurements. Each session was chaired by discussion panels selected by the organizing committee, and questions were provided to guide the discussion among the conference participants. The first session focused on the role of simulation in additive manufacturing while the second session focused on the direction for future benchmark measurements. Both sessions had ample audience participation and provided excellent feedback to the organizing committee. Please note that the following is a summary of the comments made by conference participants during the discussion session and does not necessarily reflect the views of the authors or the AM Bench 2022 Organizing Committee.

Discussion I: Simulations for Additive Manufacturing

Discussion in this first session focused on the general understanding and roles of different types of models for additive manufacturing, appropriate modeling scales, what properties and performance metrics should be modeled, and requirements for model accuracy, speed, and uncertainty. A large portion of this discussion focused on how to determine the appropriate model length scale to capture the important phenomena within AM processes. While high fidelity models can provide detailed insight into the physics of the process, they may not be appropriate or practical for Q&C due to the required speed of rapid qualification and cost of model development. The development of high fidelity models requires long computing times and model development times (coding) to incorporate the appropriate physics into the model. Thus, there is interest in reduced order, low fidelity, or generalized models. Lower fidelity models are more easily linked to in-process monitoring systems and reduce complexity in the real-time analysis of process monitoring data. However, the transfer of information between different modeling scales (coarse, mesoscale, fine scale) is a continuing issue. This information transfer could potentially be accomplished by the integration of artificial intelligence (AI) into modeling but limited high fidelity modeling results reduces the potential training capacity of these models. Computational models can make a great impact on generating process maps, transferability between machines, and reducing the need for design of experiments for every new machine, material, or process.
The simulated properties and performance of models were also a major topic of discussion in this session. The simulation of mechanical properties needs to be expanded to cover the properties required for Q&C (e.g., fatigue resistance). There is currently a major gap in fatigue modeling, specifically in the modeling of dislocations and local microstructures. To help overcome this, there are plans for fatigue and fracture challenges to be added to AM Bench in the future. However, a consensus on which fatigue and fracture experiments should be given the highest priority is needed (high cycle, low cycle, etc.). It may be best to broadly survey the AM community to determine what property challenges/needs would be helpful to increase the usefulness of models. An important point was made that the Metallic Materials Properties Development and Standardization (MMPDS) handbook has quantified uncertainties based on statistically derived properties. To achieve a similar level of success, models need to produce quantified uncertainty values. There should be additional emphasis on propagating process and model uncertainty through the model into the final evaluated properties. However, iteration of the model with different input parameters may be challenging for determining the uncertainty, and thus limiting their scope. This approach should be discussed further. A potential challenge for the future could be to predict not only the value of a mechanical property, but also the distribution of property values. Well-quantified thermo-physical properties are also either missing or hard to find. Researchers spend considerable time combing through previous literature to find any data and must take time to determine the quality/consensus of the data. It may be helpful to provide an authoritative source of well-quantified data for these values. Another important point raised in this session was related to the samples tested. Concern was raised that tensile coupons are not representative of industrial components and that limited transferability between the modeling of ‘samples’ vs. components may not be beneficial to industry. There was suggestion for a potential challenge where tensile coupon data are provided, but that the final properties of an industrial component must be modeled.

Discussion II: Directions for Future Benchmark Challenges

Discussion in this second session focused on the AM Bench challenges themselves. The first portion of the discussion focused on what did and did not work in the 2022 challenges. There was much input on what could be done to maximize the impact of the conference, data, and models. It was proposed to hold focus meetings during the AM Bench conference for submitters to get together and discuss modeling approaches. The hope in these sessions would be for people to make presentations showing what they learned during the process and how they improved and could further improve their models. It was also suggested that AM Bench host post-mortem sessions for submitters to discuss what worked and what did not work in modeling these processes (possibly hosted online, allowing voluntary involvement). Such post-mortem sessions would enable modelers to learn from each other and would help to improve the modeling community overall. There was also a proposal to create a formal steering committee to help guide what data is part of the calibration for challenges and what challenges/details are most important. It may be good to involve committee members (or volunteers in general) from a wide breadth of technology readiness levels (TRLs) of AM within their industries and members of their respective regulatory agencies. General comments on the communication and timeline were positive, but there were several ideas for improvement. The initial online discussion sessions after the release of the challenges were found to be helpful, but greater advertisement of these sessions would be appreciated. It was discussed that the time frame between the challenge problem release and submission deadline was too short. A four-to-six-month delay would be appreciated for the metals problems while the challenges for polymers would benefit from additional time (possibly one year). It was also suggested that even if there is not data ready, that general descriptions on the types of problems posed should be presented. Such descriptions would allow modelers additional time to research the literature and both prepare and focus the models. There was also a suggestion to create a notification system (such as an official AM Bench social media account) to alert participants about changes to templates and measurements, when date changes occur, or when data are updated.
Questions and comments were solicited regarding the polymer specific AM challenges, which received a relatively low number of submissions given the larger share of polymer AM in the market. The consensus was to continue issuing the challenges. There is interest in the mechanical performance of polymers. However, the polymer challenges should be limited to engineering polymers utilized in industry (e.g., PEEK—polyether ether ketones, PEK—polyether ketones), not the commonly used consumer polymers like ABS (acrylonitrile butadiene styrene) and PLA (polylactic acid). There is also interest in direct-write bio printing, as the high investment in the materials necessitates printing it right the first time. However, these complex materials systems would require longer lead times to model, as previously mentioned. There was interest from the community into expanding into other additive processes (e.g., directed energy deposition (DED) and binder jet). Wire-arc additive of Ti6Al4V was suggested due to the prevalence of research on this process/material combination within the aerospace industry. While extension into these domains is largely needed by the community, this requires significant extension and reliance on outside (non-NIST) sources for building parts. The pedigree of the instruments and calibrations would need to be sufficiently scrutinized and proven to ensure viable challenge data. Comments were also solicited from the audience on the sample geometry, namely the bridge structure used for both AM Benchmark 2018 and 2022. Should the geometry change, and if so, how? Most comments focused on how to better relate the sample geometry to industrial applications. Complex geometries such as curved/freeform surfaces or lattice structures would greatly benefit industry, but new geometries should be easily accessible/practical for many people to measure using commonly applied methods. The prints could be a combination of simpler problematic geometries common within complex components. An example of this could be holes with varying size printed at varying orientations. More complex geometries could also be investigated (e.g., tubular geometries, heat exchangers) if the measurand/process was well defined.

Embedded Workshop on Q&C

The AM Bench conference concluded with an embedded workshop on Computational Materials-Informed Q&C, primarily focused on the aviation industry. The workshop included a series of opening remarks by the National Aeronautics and Space Administration (NASA) Marshall Space Flight Center, NASA Langley Research Center, Federal Aviation Administration (FAA), Lockheed Martin, and Pratt & Whitney to set the foundation for a NIST-chaired panel discussion including representatives of the listed organizations and NIST. Although the discussion was focused on the aviation industry, many of the identified challenges and opportunities are common to other industrial segments such as health care and the civilian nuclear industry.
The opening remarks highlighted the need to bridge the gaps between the low-to-mid-TRL focus of the AM Benchmark community and the high-TRL requirements for Q&C while recognizing that requirements for testing, including full scale testing, will continue longer-term. Despite the increased development and maturation of computational materials capabilities, their acceptance in the Q&C domain requires systematic development and rigorous validation of the computation models on specimens and components that accurately represent critical features in flight hardware. Additionally, maturation of computational materials capabilities in non-regulatory domains should be continued.
Topics of discussion during the panel session included opportunities for computational materials in the Q&C domain, key elements needed for maturation of low-TRL research to high-TRL Q&C capabilities, the appropriate balance between simulation and testing in a fully mature computational materials framework for next-generation Q&C, the role of traditional verification and validation (V&V) processes [38, 39] in the maturation of computation materials capabilities, and examples of components that have gone through the Q&C process. The panel’s discussion of these topics will be summarized here. Please note that the following is a summary of the comments made by panel and conference participants and does not necessarily reflect the views of the authors, their organizations, or the AM Bench 2022 Organizing Committee.
Opportunities for computational materials in the Q&C domain span simulation of AM processing, including solidification, microstructure/defect evolution, and formation of as-built surface roughness, to understanding of the effects of each of these features on structural and fatigue performance. The need to mature computational materials-informed capabilities to address Q&C of flight hardware, including flight hardware having complex processing history, rather than coupons produced under nominal processing conditions, was widely recognized by the panel. Similarly, the design of test coupons that reflect the processing history, microstructures, defect populations, and surface roughness representative of flight hardware was discussed.
The key elements needed for maturation of low-TRL research to high-TRL Q&C capabilities include:
  • Determination of the justification, benefit, purpose, intended use, and path toward maturation of computational materials simulations
  • Development of AM Benchmarks and other high-quality datasets, including those based on flight hardware, needed to guide model development
  • Meaningful V&V and focused development of dedicated pathways to bridge from low-TRL research to high-TRL application, including suitable business cases that motivate original equipment manufacturer (OEM) engagement
  • Development of non-proprietary exemplar problems that are foundational for gaining confidence in the use of computational materials on flight hardware.
Hence, systematic development of the capabilities required for computational materials-informed Q&C requires benchmarking on increasingly realistic specimen configurations that span from idealized test coupons to flight components. Outcomes of traditional Q&C on flight components may be considered; however, the means of compliance and data associated with flight hardware are often company proprietary. As a result, an important intermediate step is benchmarking on non-proprietary or exemplar components.
Discussion of the appropriate balance between simulation and testing in a fully mature computational materials framework for next-generation Q&C began with recognition that the insertion of computational materials in Q&C is a long-term and incremental process. The process begins with gradual insertion of simulation in lower-criticality applications to understand benefits, assess risk, and gain confidence in these methods. Once adequate confidence is gained in lower-criticality applications, the gradual insertion of simulation to supplement traditional empirical methods can begin in safety–critical applications. Realistically, the promise of simulation is to supplement testing while providing increased understanding of the data from testing, rather than to replace testing altogether. The optimal balance of simulation and testing will be dependent on the specific component, loading, environment, and other requirements of the OEMs and regulators.

Conclusions and Future Directions

Identifying suitable quantitative metrics for comparing measurement results to their simulation output counterparts continues to be a challenge, particularly for complex results such as microstructure, 3D build thermography, and fracture locations. For the 2022 AM Bench challenge problems, the selected metrics were fairly simplistic, making it possible for the AM Bench organizers to rapidly assess and compare the results from multiple challenge submissions. Nevertheless, it was still necessary to incorporate some qualitative assessments in both 2018 and 2022. Identifying ‘proper’ quantitative comparison metrics remains a major gap in enabling AM model V&V and uncertainty quantification. Developing such metrics is essential for AM modeling and simulation to be of value in real-world industrial applications, where statistical assessment of the model’s predictive accuracy and precision is paramount. AM Bench is the largest source of public ‘ground-truth’ or reference measurement data for AM model development and the organizers and measurement teams will continue to investigate the design and evaluation of quantitative model-to-measurement comparison metrics. We also implore AM Bench participants to do the same, particularly for complex measurement and modeling data.
The AM Bench 2022 discussion sessions provided valuable input and much of it is already being acted upon. For example, we have been holding a series of online AM Bench-moderated, topically focused workshops that allow the various challenge problem participants to share their modeling approaches and experiences with one another and discuss the lessons learned. This targeted discussion will lead to improved AM modeling capabilities and better understanding by the AM Bench organizers of the relevant modeling difficulties and needs. Another request that the AM Bench organizers are responding to is the repeated call for a longer time frame between the challenge problem release and submission dates, currently three months. AM Bench participants clearly stated that they want to use the AM Bench challenge problems as a mechanism for developing and testing new AM modeling capabilities and three months is inadequate for new model development. The AM Bench organizers have agreed to accommodate this request, but any new schedules must also consider the long benchmark measurement timeframes. One possible solution would be to announce the general topics of the challenge problems one year before the submission deadline with the full details and calibration data being released at the six-month point.
Another important conclusion that was reached by the AM Bench organizers is that the connection between AM Bench and the Q&C community should be maintained and strengthened. After the conference, multiple participants thanked the AM Bench organizers for hosting the embedded Q&C workshop and stated this was one of the conference highlights. Currently, there exists a considerable gap between much of the basic research community (low TRL) and the application-focused needs of industry (high TRL). Bridging this gap has long been recognized as critical for advancing technological applications. For 2018 and 2022, the AM Bench organizing committee largely focused the benchmark measurements on low-TRL and mid-TRL needs. Moving forward, the organizers plan to develop benchmarks that address more high-TRL modeling topics. It is also probable that the next AM Bench conference will incorporate another embedded workshop emphasizing requirements for high-TRL applications.

Declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creativecommons.​org/​licenses/​by/​4.​0/​.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Fußnoten
1
Certain equipment, instruments, software, or materials are identified in this paper in order to specify the experimental procedure adequately. Such identification is not intended to imply recommendation or endorsement of any product or service by NIST, nor is it intended to imply that the materials or equipment identified are necessarily the best available for the purpose.
 
Literatur
3.
4.
Zurück zum Zitat Levine LE (2016) Software architecture, database development, and model validation: toward a computational benchmark in additive manufacturing. In: Michelle Schwalbe (ed) National Academies of Sciences, Engineering, and Medicine 2016. Predictive theoretical and computational approaches for additive manufacturing: proceedings of a workshop. Washington, National Academies Press pp. 86–88 Levine LE (2016) Software architecture, database development, and model validation: toward a computational benchmark in additive manufacturing. In: Michelle Schwalbe (ed) National Academies of Sciences, Engineering, and Medicine 2016. Predictive theoretical and computational approaches for additive manufacturing: proceedings of a workshop. Washington, National Academies Press pp. 86–88
5.
Zurück zum Zitat Levine LE, Lane BM, Heigel JC, Migler KB, Stoudt MR, Phan TQ, Ricker RE, Strantza M, Hill MR, Zhang F, Seppala JE, Garboczi EJ, Bain ED, Cole DP, Allen AJ, Fox J, Campbell CE (2020) Outcomes and conclusions from the 2018 AM-Bench measurements, challenge problems, modeling submissions, and conference. Integr Mater Manuf Innov 9:1–15. https://doi.org/10.1007/s40192-019-00164-1CrossRef Levine LE, Lane BM, Heigel JC, Migler KB, Stoudt MR, Phan TQ, Ricker RE, Strantza M, Hill MR, Zhang F, Seppala JE, Garboczi EJ, Bain ED, Cole DP, Allen AJ, Fox J, Campbell CE (2020) Outcomes and conclusions from the 2018 AM-Bench measurements, challenge problems, modeling submissions, and conference. Integr Mater Manuf Innov 9:1–15. https://​doi.​org/​10.​1007/​s40192-019-00164-1CrossRef
8.
9.
Zurück zum Zitat Phan TQ, Strantza M, Hill M, Gnaeupel-Herold G-H, Heigel JC, E’Elia C, DeWald AT, Clausen B, Pagan DC, Ko P, Brown DW, Levine LE (2019) Elastic residual strain and stress measurements and corresponding part deflections of 3D additive manufacturing builds of in625 am-bench artifacts using neutron diffraction, synchrotron X-ray diffraction, and contour method. Integr Mater Manuf Innov 8:318–334. https://doi.org/10.1007/s40192-019-00149-0CrossRef Phan TQ, Strantza M, Hill M, Gnaeupel-Herold G-H, Heigel JC, E’Elia C, DeWald AT, Clausen B, Pagan DC, Ko P, Brown DW, Levine LE (2019) Elastic residual strain and stress measurements and corresponding part deflections of 3D additive manufacturing builds of in625 am-bench artifacts using neutron diffraction, synchrotron X-ray diffraction, and contour method. Integr Mater Manuf Innov 8:318–334. https://​doi.​org/​10.​1007/​s40192-019-00149-0CrossRef
11.
13.
Zurück zum Zitat Cole DP, Gardea F, Henry TC, Seppala JE, Garboczi EJ, Migler KD, Shumeyko CM, Westrich JR, Orski SV, Gair JL (2020) AMB2018-03: benchmark physical property measurements for material extrusion additive manufacturing of polycarbonate. Integr Mater Manuf Innov 9:358–375. https://doi.org/10.1007/s40192-020-00188-yCrossRef Cole DP, Gardea F, Henry TC, Seppala JE, Garboczi EJ, Migler KD, Shumeyko CM, Westrich JR, Orski SV, Gair JL (2020) AMB2018-03: benchmark physical property measurements for material extrusion additive manufacturing of polycarbonate. Integr Mater Manuf Innov 9:358–375. https://​doi.​org/​10.​1007/​s40192-020-00188-yCrossRef
16.
Zurück zum Zitat Lane BM, Mekhontsev S, Grantham S, Vlasea M, Whiting J, Yeung H, Fox J, Zarobila C, Neira J, McGlauflin M (2016) Design, Developments, and results from the nist additive manufacturing metrology testbed (AMMT). In proceedings of the 26th Annual International Solid Freeform Fabrication Symposium, Austin, pp. 1145–1160. Lane BM, Mekhontsev S, Grantham S, Vlasea M, Whiting J, Yeung H, Fox J, Zarobila C, Neira J, McGlauflin M (2016) Design, Developments, and results from the nist additive manufacturing metrology testbed (AMMT). In proceedings of the 26th Annual International Solid Freeform Fabrication Symposium, Austin, pp. 1145–1160.
17.
Zurück zum Zitat Lane BM, Deisenroth D, Yeung H, Levine LE, Mekhontsev S, Weaver JS (2024) In-Situ Thermography laser powder bed fusion IN718 3D builds for the 2022 additive manufacturing benchmark challenges. For submission to Integr Mater Manuf Innov Lane BM, Deisenroth D, Yeung H, Levine LE, Mekhontsev S, Weaver JS (2024) In-Situ Thermography laser powder bed fusion IN718 3D builds for the 2022 additive manufacturing benchmark challenges. For submission to Integr Mater Manuf Innov
18.
Zurück zum Zitat Phan T, Şeren MH, Das A, Ko P, Nygren K, Levine LE (2024) Elastic residual strain measurements of 3D additively manufactured builds of nickel alloy 718 AM bench 2022 artifacts using energy dispersive synchrotron X-ray diffraction. For submission to Integr Mater Manuf Innov Phan T, Şeren MH, Das A, Ko P, Nygren K, Levine LE (2024) Elastic residual strain measurements of 3D additively manufactured builds of nickel alloy 718 AM bench 2022 artifacts using energy dispersive synchrotron X-ray diffraction. For submission to Integr Mater Manuf Innov
19.
Zurück zum Zitat Christopher R. D’Elia, Nicholas A. Bachus, Thomas Gnaupel-Herold, Thien Phan, Amlan Das, Brandon Lane, Lyle Levine, Michael R. Hill, (2024) Mapping of bulk residual stress in the AM-Bench IN718 artifact using diverse measurement techniques. For submission to Integr Mater Manuf Innov Christopher R. D’Elia, Nicholas A. Bachus, Thomas Gnaupel-Herold, Thien Phan, Amlan Das, Brandon Lane, Lyle Levine, Michael R. Hill, (2024) Mapping of bulk residual stress in the AM-Bench IN718 artifact using diverse measurement techniques. For submission to Integr Mater Manuf Innov
21.
Zurück zum Zitat Levine LE, Williams ME, Creuziger A, Stoudt MR, Young SA, Moon KW, Lane BM (2024) Location-specific microstructure characterization within AM Bench 2022 nickel alloy 718 3D Builds, submitted to Integr Mater Manuf Innov Levine LE, Williams ME, Creuziger A, Stoudt MR, Young SA, Moon KW, Lane BM (2024) Location-specific microstructure characterization within AM Bench 2022 nickel alloy 718 3D Builds, submitted to Integr Mater Manuf Innov
22.
23.
Zurück zum Zitat Schwalbach EJ, Chapman MG, Shah MN, Uchic MD, Levine LE, Hrabe N, Kafka OL, Moser NH, Carson R, Belak J (2024) Laser powder bed fusion alloy 625 and alloy 718 3D microstructures for AMB2022–01 and AMB2022–05. For submission to submission to IMMI Schwalbach EJ, Chapman MG, Shah MN, Uchic MD, Levine LE, Hrabe N, Kafka OL, Moser NH, Carson R, Belak J (2024) Laser powder bed fusion alloy 625 and alloy 718 3D microstructures for AMB2022–01 and AMB2022–05. For submission to submission to IMMI
24.
Zurück zum Zitat Deisenroth D, Weaver JS, Yeung H, Mekhontsev S, Lane BM, Levine LE, Laser-scanned tracks and pads and In-situ thermography for the 2022 additive manufacturing benchmark challenges. For submission to Integr Mater Manuf Innov Deisenroth D, Weaver JS, Yeung H, Mekhontsev S, Lane BM, Levine LE, Laser-scanned tracks and pads and In-situ thermography for the 2022 additive manufacturing benchmark challenges. For submission to Integr Mater Manuf Innov
25.
Zurück zum Zitat Weaver JS, Deisenroth D, Mekhontsev S, Lane BM, Levine LE (2024) Cross-sectional melt pool geometry of laser scanned tracks and pads on nickel alloy 718 for the 2022 additive manufacturing benchmark challenges. Integr Mater Manuf Innov 13:363–379. https://doi.org/10.1007/s40192-024-00355-5 Weaver JS, Deisenroth D, Mekhontsev S, Lane BM, Levine LE (2024) Cross-sectional melt pool geometry of laser scanned tracks and pads on nickel alloy 718 for the 2022 additive manufacturing benchmark challenges. Integr Mater Manuf Innov 13:363–379. https://​doi.​org/​10.​1007/​s40192-024-00355-5
26.
Zurück zum Zitat Levine LE, Williams ME, Stoudt MR, Weaver JS, Young SA, Deisenroth D, Lane BM (2024) Location-specific microstructure characterization within AM bench 2022 laser tracks on bare nickel alloy 718 plates. Integr Mater Manuf Innov 13:380–395. https://doi.org/10.1007/s40192-024-00361-7 Levine LE, Williams ME, Stoudt MR, Weaver JS, Young SA, Deisenroth D, Lane BM (2024) Location-specific microstructure characterization within AM bench 2022 laser tracks on bare nickel alloy 718 plates. Integr Mater Manuf Innov 13:380–395. https://​doi.​org/​10.​1007/​s40192-024-00361-7
27.
Zurück zum Zitat Carson R, Maich A, Knox R, Raghavan N, Shanks K, Matthew Rolchigo M, John Coleman J, Darren Pagan D, Rowenhorst D, Creuziger A, Schwalbach E, Levine L, Turner J, Belak J, Bement M (2024) Microstructural characterization and macro/mesoscale mechanical compression measurements of the inconel 625 AMBench 2018–01 Bridge Part. For submission to Integr Mater Manuf Innov Carson R, Maich A, Knox R, Raghavan N, Shanks K, Matthew Rolchigo M, John Coleman J, Darren Pagan D, Rowenhorst D, Creuziger A, Schwalbach E, Levine L, Turner J, Belak J, Bement M (2024) Microstructural characterization and macro/mesoscale mechanical compression measurements of the inconel 625 AMBench 2018–01 Bridge Part. For submission to Integr Mater Manuf Innov
28.
Zurück zum Zitat Habib SA, Zuback JS, Weaver JS, Mates SP, Levine LE, Microstructural and mechanical characterization of as-built IN625 at different strain rates. For submission to Integr Mater Manuf Innov Habib SA, Zuback JS, Weaver JS, Mates SP, Levine LE, Microstructural and mechanical characterization of as-built IN625 at different strain rates. For submission to Integr Mater Manuf Innov
31.
Zurück zum Zitat Higgins C, Kehe GM, Caplins B, Miller C, Stansbury JW, Zong Y, Killgore JP (2024) AM Bench 2022 Vat Photopolymerization Challenge Data: Resin properties, light engine calibration, and calibration print dimensions (AMB2022–07). For submission to Integr Mater Manuf Innov Higgins C, Kehe GM, Caplins B, Miller C, Stansbury JW, Zong Y, Killgore JP (2024) AM Bench 2022 Vat Photopolymerization Challenge Data: Resin properties, light engine calibration, and calibration print dimensions (AMB2022–07). For submission to Integr Mater Manuf Innov
34.
35.
Zurück zum Zitat Turner JA, Belak J, Barton N, Bement M, Carlson N, Carson R, DeWitt S, Fattebert JL, Hodge N, Jibben Z, King W, Levine LE, Newman C, Plotkowski A, Radhakrishnan B, Rolchigo M, Sabau A, Slattery S (2022) Metal additive manufacturing simulation at the fidelity of the microstructure. Int J High Perform Comput Appl 36:13–39. https://doi.org/10.1177/10943420211042558CrossRef Turner JA, Belak J, Barton N, Bement M, Carlson N, Carson R, DeWitt S, Fattebert JL, Hodge N, Jibben Z, King W, Levine LE, Newman C, Plotkowski A, Radhakrishnan B, Rolchigo M, Sabau A, Slattery S (2022) Metal additive manufacturing simulation at the fidelity of the microstructure. Int J High Perform Comput Appl 36:13–39. https://​doi.​org/​10.​1177/​1094342021104255​8CrossRef
38.
Zurück zum Zitat The American Society of Mechanical Engineers (2020) Standard for verification and validation in computational solid mechanics. ASME, New York The American Society of Mechanical Engineers (2020) Standard for verification and validation in computational solid mechanics. ASME, New York
39.
Zurück zum Zitat The Minerals, Metals & Materials Society (TMS) (2020) Accelerating the broad implementation of verification and validation in computational models of the mechanics of materials and structures. TMS, Pittsburgh, PA. www.tms.org/VVaccelerator The Minerals, Metals & Materials Society (TMS) (2020) Accelerating the broad implementation of verification and validation in computational models of the mechanics of materials and structures. TMS, Pittsburgh, PA. www.​tms.​org/​VVaccelerator
Metadaten
Titel
Outcomes and Conclusions from the 2022 AM Bench Measurements, Challenge Problems, Modeling Submissions, and Conference
verfasst von
Lyle Levine
Brandon Lane
Chandler Becker
James Belak
Robert Carson
David Deisenroth
Edward Glaessgen
Thomas Gnaupel-Herold
Michael Gorelik
Gretchen Greene
Saadi Habib
Callie Higgins
Michael Hill
Nik Hrabe
Jason Killgore
Jai Won Kim
Gerard Lemson
Kalman Migler
Shawn Moylan
Darren Pagan
Thien Phan
Maxwell Praniewicz
David Rowenhorst
Edwin Schwalbach
Jonathan Seppala
Brian Simonds
Mark Stoudt
Jordan Weaver
Ho Yeung
Fan Zhang
Publikationsdatum
17.07.2024
Verlag
Springer International Publishing
Erschienen in
Integrating Materials and Manufacturing Innovation / Ausgabe 3/2024
Print ISSN: 2193-9764
Elektronische ISSN: 2193-9772
DOI
https://doi.org/10.1007/s40192-024-00372-4

Weitere Artikel der Ausgabe 3/2024

Integrating Materials and Manufacturing Innovation 3/2024 Zur Ausgabe

    Marktübersichten

    Die im Laufe eines Jahres in der „adhäsion“ veröffentlichten Marktübersichten helfen Anwendern verschiedenster Branchen, sich einen gezielten Überblick über Lieferantenangebote zu verschaffen.