Fundamental Measurements for Metal Additive Manufacturing

Summary

Critical parts for defense, aerospace, and medical applications must be formally qualified prior to use. Currently, extensive empirical testing consisting of thousands of individual tests, costing millions of dollars and necessitating several years to complete, is required to qualify parts built through additive manufacturing (AM) and minor changes to the process requires complete re-qualification. Additionally, complex surface topographies and features, internal defects, and anisotropic/location specific properties challenge current state of the art measurement techniques and testing protocols. All of this hinders the manufacturers’ and users’ ability to benefit from the key advantages that AM has to offer (i.e., complexity of design and reduced development time over traditional manufacturing methods). To address this need, the Fundamental Measurements for Metal Additive Manufacturing (FMMAM) project will develop the measurement science that creates robust measurements and testing to develop the correlations, enable qualification of AM parts, processes, and feedstocks by manufacturers, and reduce the need for complete re-qualification.

Description

Objective
To develop and deploy test methods and protocols, standard test artifacts, exemplar data, data processing tools, automation tools, and advanced characterization methods that address the complexity of AM, create robust measurements, and instill confidence in the metrology used in the qualification of metal AM feedstock materials, processes, and parts.

Technical Idea
To instill confidence and aid adoption of AM as a viable technology for production in critical applications, a strong understanding of how to measure, characterize and qualify AM parts, processes, and feedstock materials is required. The complex physical interactions in the process and wide range of process variables inhibit our understanding and development of PSPP relationships. Furthermore, the flexibility in design and customizability of parts available to AM technologies has created new and unique measurement challenges that must be addressed.

The three main thrusts are:
Feedstock material
In-process
Post-process

Feedstock Material Characterizations

• There is a need to experimentally verify the extent to which flowability and spreadability of AM powders are related, if at all, and to enable more robust relationships to be used in feedstock characterization. Methods for characterizing metal feedstock powders, their spreadability, and their effect on the quality of the powder bed or the performance of the part will be developed to enable qualification. This project will utilize the powder spreading testbed as well as commercial and experimental test equipment to elucidate the relationship between flowability and spreadability.

In-process Characterizations

• The spread powder layer is a key factor in the success of a build. This work will investigate the powder packing density and powder denudation in the spread layer. 
  • Powder packing density has been linked to part quality, but the extent to which these factors are correlated is still unclear. This work will build upon the work developing spread layer quality metrics, extend analysis from the powder spreading testbed to commercially available equipment, and will seek to determine correlations between spread layer quality metrics and part quality.
  • Powder denudation is a characteristic of the LPBF process where the flow of powder is caused by the vapor jet created by the laser melting process. This project will address the two key gaps that limit the use of powder denudation measurements for improving the LPBF process: 1) a lack of metrics to quantify denudation, and 2) a lack of understanding of the repeatability and sensitivity of powder denudation.
• There remains insufficient understanding of the optical phenomena that occur during laser-matter interaction, which limits the accuracy of non-contact measurements for multiphysics model validation applications. Process byproducts (including metal vapor, condensate, and ejecta) can interact with the light propagating to (and from) the melt pool, which can attenuate, scatter, or otherwise distort the power density distribution that is delivered to the melting surface.  Quantification of the total laser power reflected from the laser-matter interaction (used to approximate laser absorption), as well as to visualize the directional distribution of reflected laser power are of great importance and highly relevant to multiphysics model validation and process monitoring applications. This project will utilize the systems being developed by the Metrology for AM Model Validation project to develop the methods for characterization and understanding of laser matter interactions in the LPBF process.

Post-process Characterizations

• The detection and elimination of defects (e.g., pores) are critical for qualification. To understand the limitations of a XCT system, well characterized reference artifacts are required. This project will build upon the existing prototype developed by NIST to refine the artifact design and fabrication process. The new artifact will be characterized to establish ground truth and used to evaluate and compare probability of detection (POD) among different XCT defect detection algorithms used in qualifying AM parts.
• Melt pool morphology measurements are used for process development, process health, and model validation. Experiments for melt pool measurements vary and may lead to a misunderstanding of a representative melt pool for a given LPBF process. Experiments also lack a focus on off-nominal scenarios, and therefore the performance of processes and models in these scenarios is unknown. This work will build upon the current body of knowledge to characterize melt pool geometry in nominal and off-nominal scenarios. The effect on the nominal melt pool dimensions and the variability in the melt pool dimensions will be studied and characterized, and will be compared with non-destructive evaluation techniques (e.g., XCT, optical metrology) and in-situ metrology (e.g., coaxial images) to develop methods for accurate and repeatable measurement of melt pool dimensions, and aiding the development of relationships between melt pool geometry and performance predictions. 
• Mechanical property measurements of standard coupons do not necessarily represent the properties of complex parts or other part geometries because the part geometry can shift the process. Furthermore, AM parts may contain intentional or unintentional gradients that create challenges with typical test methods. Location specific or feature specific mechanical properties are needed, but test methods that can be applied because they test a small volume (e.g., indentation and small punch) may or may not correlate with traditional test methods.  This project seeks to compare indentation testing with tensile testing for AM metals and small punch testing with tensile and fracture properties to identify and develop correlations that may exist. Furthermore, the knowledge gained from that work will be used to develop test coupons that represent location or component specific material.
• Complex surface topography (e.g., steep slopes and re-entrant features), lattice structures, internal geometries, and topology optimized parts, challenge the current state of the art for dimensional characterization (e.g., geometric dimensions and accuracy, form, surface finish). Additionally, black/grey box algorithms create challenges for determining the metrological characteristics of a system . Intercomparisons between systems where a strong understanding of the metrological characteristics exists and grey/black box systems where these characterizations are much more difficult to determine will be performed. Furthermore, reference artifacts (new and existing) will be designed (where applicable) and analyzed to improve uncertainty analysis in the dimensional measurement of AM parts.

These challenges exist both in measurement of AM feedstocks and parts, but also in the connections from design to production. While robust characterizations and measurements build confidence in the correlations being developed, advances in this field hold greater importance when they can be integrated into the qualification of the full AM production process.  

Finally, the metrology equipment used to characterize AM is vast and not all systems are created equally. Each system has limitations and advantages and, in some instances, may not be comparable at all (e.g., surface metrology equipment that measure across different and separate spatial bandwidths). As such, particular attention will be paid to the measurement uncertainty, and the methods to compare measurements on different metrology equipment and in different phases of the AM lifecycle (e.g., variations from part-to-part, user-to-user, system-to-system, etc.).

Research Plan
The scope of the FMMAM project is vast. Thus, there are many research tasks that will be performed simultaneously to successfully accomplish the goals set forth by the project. Equipment ranging from commercially available to novel and experimental, will be used to accomplish the following tasks:

• The relationship between flowability and spreadability, as well as the effect of powder layer density on part quality will be assessed experimentally. Commercial and experimental powder characterization equipment, the powder spreadability testbed developed at NIST, optical microscopy, and melt pool size characterizations will be used in the analysis to support correlation development and the advancement of feedstock characterization methods. Methodologies and results of this work will be used to develop correlations between these factors and will be disseminated to the community through research publications and standards development.
• Idealized experiments with simple conditions (e.g., bare plate, single tracks) and real process conditions will be performed to understand variability of melt pool geometry, define measurement methodologies that improve repeatability of measurements, and assess the quality of such information for engineering decision and process development.
• Test methods will be developed and executed to determine powder and process performance under denudation-like conditions using the new understandings developed to control denudation, and in collaboration with the Advanced Machines, Monitoring and Control (AMMC) project.. These experiments will also be contrasted with feedstock characterization analyses to support feedstock qualification metrics.
• Spatial and temporal resolution of the reflected laser power metrology equipment developed through other MSAM projects, and in collaboration with the AMMC and MAMMV projects, will be refined and well characterized to provide cross-comparisons with process monitoring data and strengthen our understanding of this phenomena.
• Defect artifacts that are representative of the defects seen in LPBF will be developed. Techniques such as AM, FIB, Laser micromachining, and projection photolithography have been demonstrated, and maskless lithography, and nanoindentation will be further investigated for different substrate materials and defect characteristics. Different assembly techniques and reference measurement methods for created features will be investigated. The findings of the demonstration studies will be shared to community for guidance (i.e., via journal publications, standards development), and allowing industries to identify the best techniques for their applications. Prototype artifacts will be developed, and demonstration studies on POD, probability of sizing (POS), and/or automated detection algorithms will be performed using the artifacts.
• Benchmark XCT defect data sets will be developed and shared to community to evaluate various defect detection algorithms. Data sets will be prepared through XCT simulation and from measurements of physical artifacts under development. A computational framework is being developed to generate realistic simulation data. Various detection algorithms including those based on AI/ML will be investigated using different evaluation metrics to help with designing challenge problems.
• XCT; optical and tactile surface, form, and coordinate metrology; and other non-destructive testing (NDT) systems will be used to develop more detailed and quantitative characterization of part dimensions, form, surface finish, defect morphology, and defect locations. Test artifacts based on the unique dimensional characteristics of AM parts will be developed to better understand how part and surface complexity affect a metrological system’s performance. 
• This work will continue the case study in the development and realization of surface topography measurands that is already underway. Experiments and analysis will be performed to better understand the repeatability of various measurement relevant to the AM process. This will also provide a basis for understanding how these measurands vary from machine-to-machine, operator-to-operator, etc.

Where and to the extent possible, the output of the above tasks will be integrated into the holistic fabrication and qualification process. Successful methodologies will be disseminated to the wider AM community and utilized as a model for incorporating additional outputs. Challenges will be documented and studied to identify future research tasks. Finally, this project will support the development of measurement science to support equivalence-based qualification method and work closely with other AM projects in the MSAM program to develop use cases/guidelines for model-based qualification.