Predicting geometric influences in metal additive manufacturing

https://doi.org/10.1016/j.mtcomm.2020.101174Get rights and content

Abstract

Metal additive manufacturing (AM) has evolved from scientific curiosity to a potential paradigm shift in materials manufacturing owing to its ability to control microstructure and fabricate complex geometries. However, geometry, often ignored as a variable, can have a significant impact on microstructure evolution, thereby indirectly limiting the design flexibility offered by AM. Currently, the process of retaining microstructure with changing geometry relies on trial and error-based processing followed by microstructural evaluation that places significant demands on time and costs associated with experimentation and characterization. We demonstrate that lightweight thermal models can be effectively used to predict microstructural changes with geometry. These models can in turn be used as precursors to developing geometry independent scan strategies and achieve microstructure control in various alloy systems. Even if the thermal signatures predicted by the model are not accurate, scan strategies can be developed to mimic the thermal conditions during initial process development. We present a case study on one of the most studied AM alloys: Ti-6Al-4V.

Introduction

Additive manufacturing (AM) is a family of processes for layer-by-layer deposition of material guided by a computer aided design to fabricate near net shaped components. AM allows for production of complex geometries that cannot be manufactured with conventional subtractive processes, and without the expense and time associated with obtaining tooling. These advantages are having significant impacts in the aerospace, biomedical, automotive, and tool and die industries through production of complex parts that provide improved performance and reduced lead times [[1], [2], [3], [4], [5], [6], [7], [8], [9], [10]]. Further, the site-specific nature of the thermal conditions in AM enables control over the spatial distribution of microstructure through manipulation of process variables and scan strategies. Previous research has shown the ability to control a variety of microstructural features, such as the grain size, morphology and crystallographic texture in Ni-based superalloys [[11], [12], [13], [14]] and electrical steels [15], non-equilibrium phase selection in Al alloys [16], and defects in Ti-6Al-4V [[17], [18], [19]].

Despite these advances, current design optimization is largely separated from process development, and both are treated independent of microstructure evolution. As an example, Antonysamy et al. published on the effects of section thickness on β grain and texture evolution in Ti-6Al-4V while Yoder et al. published on topology optimization of Ti-6Al-4V brackets where they reported unexpected failure due to porosity developed in optimized geometries [20,21]. In another study, Yoder et al. demonstrated that subtle changes in build layout and changes in the height of support structures can have a noticeable impact on the tensile behavior of Ti-6Al-4V processed using electron beam powder bed fusion (EBPF) [22]. Frederick et al. also demonstrated variations in the grains morphology and texture of Rene-N5 as a function of local geometric features [23]. These differences in defects and microstructure can be attributed to changes in thermal conditions experienced by the material over the course of the build in response to the interactions between the process conditions, scan pattern, and geometric features. While there have been studies on in-situ monitoring to better understand the defect structures and thermal signatures during the process, in the current state, in-situ monitoring can be used to qualify the material with regards to the defects but can provide only limited insights towards microstructure evolution [18,21,22,24,25].

The true design potential of AM can be realized only if the interplay between processing conditions, geometry, and microstructure can be understood and controlled at will. However, to realize this vision it is important to select and design materials systems in which modifications within the available process space may be leveraged to produce relevant changes in microstructural features and associated properties. It is important to understand that useful property control is not possible in all situations and is heavily dependent on both material and process. As an example, while IN718 may be processed to select either columnar or equiaxed grains in electron beam powder bed fusion, the same cannot be accomplished for Ti-6Al-4V due to limited constitutional supercooling [26,27]. A generalization of this example is shown conceptually in Fig. 1a. Meaningful microstructural design is only possible where manipulation of relevant phase transformations overlaps the practical process window for a given material system. For equiaxed grain nucleation, there is significant overlap for IN718, but none for Ti-6Al-4V. This highlights the need for either developing new machines capable of reaching the required thermal conditions or developing a new Ti-base alloy that has properties comparable to Ti-6Al-4V but solidifies with greater constitutional supercooling [28].

The above discussion on the impact of processing techniques and parameters on the resulting microstructure and their interplay impacting the process optimization for arbitrary geometries is schematically summarized in Fig. 1b. An effective approach must first define the relevant process space for defects and microstructure to determine fundamental correlations between thermal conditions and relevant microstructural features, and between material state and properties. The former requires a thermal process model that reasonably captures the relevant process dynamics, including the effects of geometry, while the latter requires either constitutive relationships or suitable experimental correlations. Most numerical process models have sought to capture the complexity of the relevant physical phenomena, but the resulting simulations are prohibitively expensive and limited to only small length and time scales [12,14,[29], [30], [31], [32]]. Consequently, several researchers have instead developed simplified semi-analytical approaches that retain only the core physical attributes of heat conduction in AM in exchange for greatly reduced computational expense [[33], [34], [35], [36], [37]]. In some cases, correlations with microstructure and properties exist in enough detail to enable appropriate design, in others, additional data will be required. An optimization algorithm may then be implemented and iterated to define the scan pattern that achieves the desired outcome, i.e., the desired geometry-independent microstructure distribution, without excessive build defects.

This approach to design for AM requires a systematic understanding of process correlations with microstructure and defects, as well as scalable modeling and optimization tools. The understanding of phase transformations has developed to a reasonable extent for some materials, but not for others. Recent developments in process modeling and optimization show promise for successfully applying this approach to a wide variety of materials, but still requires validation. The purpose of this work is to demonstrate the use of a semi-analytical model to explain and quantify the microstructural evolution during solidification as well as during the solid-solid phase transformations in Ti-6Al-4V that encompasses a small but significant space in the overall complex process space described in Fig. 1. Here we apply a simple process model to predict the influence of geometry on the microstructure in Ti-6Al-4V produced by electron beam powder bed fusion, where the solid-state transformation from β to α plays a critical role in governing material properties. We use a spot-melting scheme owing to its promise in allowing for an unprecedented level of microstructure control since the thermal gradients can be managed more precisely by controlling the order in which points are melted. Our goal is to validate the use of thermal models for scan pattern design and prediction of resulting microstructures with the long-term aim of developing geometrically independent process conditions.

Section snippets

Scan strategy and additive manufacturing

The purpose of this work is to investigate the influence of geometry on microstructure for a fixed scan pattern and set of process conditions. The scan pattern used was a spot melt pattern (Fig. 2a), which enables a more flexible distribution of heat input than conventional raster patterns. The scan strategy relies on a uniform triangular grid of points that are overlaid on the desired geometry. The first point to be melted is at the top left corner (Spot 1 in Fig. 2a(1)). The beam then melts

Results and discussion

Although intuitively, it can be hypothesized that the difference in beam return time resulting from the geometric changes will cause fluctuations in the microstructure, it is important to quantify the associated thermal profiles to achieve the longer-term goal of geometry independent microstructure design. While single point melt pool calculations can be effective in understanding the role of process parameters on the melt pool [27,44], they cannot factor in the geometric effects. Thus, the

Conclusions

In summary, we demonstrated that the use of lightweight thermal models in tandem with the development of new scan strategies are critical to unlocking the true design freedom of AM. It is important to be cognizant of the limits of the process window of the AM technology in consideration and whether the technology is capable of achieving the thermal profiles to deliver meaningful microstructural changes. We highlight that while process parameters can be developed for specific geometries to

CRediT authorship contribution statement

Peeyush Nandwana: Conceptualization, Methodology, Investigation, Writing - original draft. Alex Plotkowski: Methodology, Software, Writing - original draft. Rangasayee Kannan: Investigation, Writing - review & editing. Sean Yoder: Investigation. Ryan Dehoff: Funding acquisition.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

Research was performed at the U.S. Department of Energy’s Manufacturing Demonstration Facility, located at Oak Ridge National Laboratory. Research was sponsored by the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, Advanced Manufacturing Office, under Contract DE-AC05-00OR22725 with UT-Battelle, LLC.

References (47)

  • S. Yoder

    Approach to qualification using E-PBF in-situ process monitoring in Ti-6Al-4V

    Addit. Manuf.

    (2019)
  • M.J. Bermingham et al.

    Controlling the microstructure and properties of wire arc additive manufactured Ti-6Al-4V with trace boron additions

    Acta Mater.

    (2015)
  • N. Raghavan

    Numerical modeling of heat-transfer and the influence of process parameters on tailoring the grain morphology of IN718 in electron beam additive manufacturing

    Acta Mater.

    (2016)
  • S.A. Khairallah et al.

    Laser powder-bed fusion additive manufacturing: physics of complex melt flow and formation mechanisms of pores, spatter, and denudation zones

    Acta Mater.

    (2016)
  • M.J. Matthews

    Denudation of metal powder layers in laser powder bed fusion processes

    Acta Mater.

    (2016)
  • C. Panwisawas

    Keyhole formation and thermal fluid flow-induced porosity during laser fusion welding in titanium alloys: experimental and modelling

    Acta Mater.

    (2017)
  • R. Forslund et al.

    Analytical solution for heat conduction due to a moving Gaussian heat flux with piecewise constant parameters

    Appl. Math. Model.

    (2019)
  • E.J. Schwalbach et al.

    A discrete source model of powder bed fusion additive manufacturing thermal history

    Addit. Manuf.

    (2019)
  • A. Plotkowski et al.

    Verification and validation of a rapid heat transfer calculation methodology for transient melt pool solidification conditions in powder bed metal additive manufacturing

    Addit. Manuf.

    (2017)
  • H. Galarraga et al.

    Effects of the microstructure and porosity on properties of Ti-6Al-4V ELI alloy fabricated by electron beam melting (EBM)

    Addit. Manuf.

    (2016)
  • P.R. Carriere et al.

    Energy absorption during pulsed electron beam spot melting of 304 stainless steel : Monte–Carlo simulations and in-situ temperature measurements

    Vaccum

    (2017)
  • P. Nandwana et al.

    Influence of scan strategy on porosity and microstructure of Ti-6Al-4V fabricated by electron beam powder bed fusion

    Mater. Today Commun.

    (2020)
  • G. Lütjering

    Influence of processing on microstructure and mechanical properties of (α+β) titanium alloys

    Mater. Sci. Eng. A

    (1998)
  • Cited by (15)

    • Geometric feature reproducibility for laser powder bed fusion (L-PBF) additive manufacturing with Inconel 718

      2021, Additive Manufacturing
      Citation Excerpt :

      From a systems-engineering perspective, all metal AM methods must be understood in full as their products are susceptible to process inputs from design through implementation. To successfully apply AM to practical applications for aerospace, industrial, automotive, and medical fields, the critical aspects are the combination of microstructure and resulting properties and the geometry [1,11,12]. The end-use performance is derived from the understanding and control of the geometry and microstructure to meet the functional design in addition to the structural margins of an AM-produced part [13].

    • Solidification texture, variant selection, and phase fraction in a spot-melt electron-beam powder bed fusion processed Ti-6Al-4V

      2021, Additive Manufacturing
      Citation Excerpt :

      Recently, however, several studies [22,34–37] reported the effect of spot-melt strategy on the microstructure development in the E-PBF Ti-64. For example, Nandwana et al. [66] reported that the spot-melt strategy offers a flexibility in the control of heat distribution imparted onto the build due to the availability of numerous fill patterns that can be implemented. This could potentially translate into a method of localized microstructural control in AM components.

    • Application of Data Science and Engineering

      2021, Encyclopedia of Materials: Metals and Alloys
    • A multivariate meltpool stability criterion for fabrication of complex geometries in electron beam powder bed fusion

      2021, Additive Manufacturing
      Citation Excerpt :

      This could be achieved by tailoring the melt geometry through advanced process and scanning strategies. One possible alternative scan strategy that is able achieve homogeneous meltpool geometries independent of part geometry is the dot printing strategy [27]. With appropriate processing parameters, each meltpool solidifies before the adjacent point is melted and the meltpool geometry will not exceed the meltpool stability limit.

    View all citing articles on Scopus
    View full text