In This Issue
Winter Bridge on Frontiers of Engineering
December 18, 2019 Volume 49 Issue 4
The winter issue of The Bridge is focused on the 2019 Frontiers of Engineering symposium.

Computational Materials for the Design and Qualification of Additively Manufactured Components

Wednesday, December 18, 2019

Author: Christapher G. Lang

NASA is developing next-generation computational materials capabilities to support the qualification of additively manufactured metallic -structural components for aerospace applications. The quality of these parts directly depends on a wide range of process parameters, including build conditions and feedstock properties. Computational materials research aims to develop a fundamental understanding of the dependence of the part properties and performance on the process parameters and to apply that understanding to efficient qualification practices.

Integrated multiscale modeling methods allow prediction of the process-structure-property relationships, including the effect of defects. This paper primarily focuses on the powder bed fusion process and its application to aerospace flight systems, with discussion of in situ monitoring, process-to-microstructure linkages including residual stress, and microstructure-to-performance linkages. Computational materials research for additive manufacturing (AM) processes will enable efficient and accurate design, manufacture, and certification of future aerospace flight systems.

Introduction

Although AM technology has recently experienced considerable growth and publicity for its potential to significantly transform the manufacturing industry, its promise is limited in application because of a lack of confidence in part quality. Improvements in material properties, consistency, and process control are necessary for AM to realize the advertised potential of enhanced performance, reduced cost, and increased manufacturing speed; for example, the application of AM to fracture-critical flight components requires extensive qualification efforts.

Additive manufacturing encompasses a variety of materials (e.g., metals, polymers, and ceramics) and processes (e.g., powder bed, blown powder, wire fed, laser, and electron beam). Part quality and consistency depend on numerous process-specific parameters that are selected or adjusted for each component.

Laser Powder Bed Fusion

I focus on the laser powder bed fusion (LPBF) process for metallic AM, although many of the approaches are applicable to a wide range of materials and manufacturing processes. The LPBF parameter space consists of laser power, scan speed, laser spot size, scanning strategy, feedstock, part geometry, and machine conditions. The selection of process parameters determines the resulting microstructure and component properties.

Various libraries of process parameters for a given machine and material have been determined through physical testing by AM suppliers or individual laboratories, with additional testing required for each new part geometry or powder supply. An integrated computational materials engineering (ICME) approach reduces the amount of physical testing and informs design engineers about detrimental performance expected for specific process parameters (Turner et al. 2015).

NASA is developing AM rocket engine components for human spaceflight. To address the immediate need for a consistent framework specific to the production and evaluation of LPBF processes, standards have been released by Marshall Space Flight Center (MSFC 2017a,b) for materials, process control, personnel training, inspection, and acceptance requirements. Concurrently, an ICME approach to the design and qualification of aerospace AM materials and their components is being developed at NASA and provides a path toward rapid manufacturing and qualification.

Improved control and understanding of the AM process offer improved consistency and more complex design such as multiple alloys and functionally graded material components. When combined with in situ process monitoring, computational modeling enables the development and integration of manufacturing process capabilities and constraints as well as qualification considerations such as inspection requirements in the component design.

Computational Modeling of the AM Process

Process modeling is used to develop an understanding of the relationship between the process parameters, feedstock, microstructural and porosity evolutions, and resulting mechanical properties by solving the governing equations for the physics of the process. Determination of the temperature history, deformations due to residual stress, microstructure evolution, and porosity are among the goals of current process simulation efforts.

Physics

Modeling of the AM process requires a multiscale approach to accurately account for the physics at multiple length scales from microstructure to component. An accurate temperature history and melt pool geometry are necessary to understand the microstructure, defect formation, and residual stress formation. The temperature history is predicted by numerical models at different levels of fidelity. Various physics—melting, evaporation, fluid flow, recoil pressure, powder packing density, and surface tension—are incorporated to improve the model accuracy. To accommodate accuracy and computational resource requirements, thermal models are generally restricted to a low number of scan tracks and powder layers.

Simulation of residual stress formation requires a scale-up to efficiently account for the numerous layers in an AM build. A promising approach for predicting residual stress is the modified inherent strain method, which computes the strain at the scan track scale and imposes the strains in a layer-by-layer fashion to a part scale mechanical analysis (Liang et al. 2018). Phase-field and kinetic Monte Carlo models are used to simulate grain structures dependent on feedstock and temperature history.

Porosity

Two sources of porosity during the LPBF process are lack of fusion and keyholing. The melt pool transitions from conduction mode to keyhole mode for increased laser power and reduced scan speed. Keyhole mode occurs when a vapor cavity forms with a high aspect ratio of depth to width as compared to conduction mode (Trapp et al. 2017). In contrast, lack of fusion porosity occurs when insufficient power and overlap of successive melt pools are applied to fully melt the powder. A balance for avoiding lack of fusion and keyhole porosity is determined by the selected process parameters (Tang et al. 2017).

Porosity cannot be completely avoided, and its impact on part performance is application dependent. Micromechanical simulations quantitatively characterize the influence of porosity and other heterogeneities in the microstructure on the mechanical behavior of parts produced by LPBF. Porosity is embedded in process-specific microstructure models, and the heterogeneous strain localization in the vicinity of the porosity is solved as a function of the pore shape, size, density, and proximity to the free surface.

In Situ Process Data

For the design and qualification of AM components, experimental data are required to capture critical events and behavior during the manufacturing process. To that end,

  • Powder bed systems are being equipped with sensors and measuring devices to record data during the manufacturing process.
  • System monitoring provides critical data necessary for understanding process events, performing feedback control, diagnosing machine operation, and validating computational models.
  • Key process measurements include temperature history, melt pool dimensions, and defect formation.

Collection of in situ data provides a component build history that can be used to identify critical events during the process that may affect part quality.

Dynamic x-ray radiography (DXR) at the Argonne National Laboratory Advanced Photon Source -provides high-speed cross-section videos of the LPBF process (Zhao et al. 2017). The real-time imaging yields data relative to the laser position, including melt pool dimensions, keyhole behavior, solidification rate, and porosity formation. DXR data help characterize the melt pool and solidification behavior for various feedstock compositions and baseplate material as well as varying laser parameters.

Summary

Computational modeling supports the qualification efforts necessary to realize the full potential of additive manufacturing for designing and manufacturing aerospace components. A large design space exists for AM, and an ICME approach to process and component design will support qualification efforts through improved process understanding and control for -application- and material-specific needs. Simulation tools that assist in choosing parameters for process control and designing AM-specific components will lead to microstructures that help attain and even exceed design specifications.

Micromechanical simulations characterize part performance for process-specific microstructures including the effect of defects. Integrated computational modeling and in situ process monitoring efforts provide a path toward accelerated design and qualification of aerospace components.

References

Liang X, Cheng L, Chen Q, Yang Q, To AC. 2018. A modified method for estimating inherent strains from detailed process simulation for fast residual distortion prediction of single-walled structures fabricated by directed energy deposition. Additive Manufacturing 23:471–86.

MSFC [Marshall Space Flight Center]. 2017a. Standard for additively manufactured spaceflight hardware by laser powder bed fusion in metals. Technical Standard -MSFC-STD-3716. Huntsville AL.

MSFC. 2017b. Specification for control and qualification of laser powder bed fusion metallurgical processes. Technical Standard MSFC-SPEC-3717. Huntsville AL.

Tang M, Pistorius PC, Beuth JL. 2017. Prediction of lack-of-fusion porosity for powder bed fusion. Additive Manufacturing 14:39–48.

Trapp J, Rubenchik AM, Guss G, Matthews MJ. 2017. In situ absorptivity measurements of metallic powders during laser powder-bed fusion additive manufacturing. Applied -Materials Today 9:341–49.

Turner JA, Babu SS, Blue C. 2015. Advanced simulation for additive manufacturing: Meeting challenges through collaboration. ORNL/TM-2015/324. Oak Ridge TN: Oak Ridge National Laboratory.

Zhao C, Fezzaa K, Cunningham RW, Wen H, De Carlo F, Chen L, Rollett AD, Sun T. 2017. Real-time monitoring of laser powder bed fusion process using high-speed X-ray imaging and diffraction. Scientific Reports 7(1):3602.

About the Author:Christapher Lang is an aerospace engineer at NASA Langley Research Center.