Elsevier

Additive Manufacturing

Volume 50, February 2022, 102578
Additive Manufacturing

Role of scan strategies and heat treatment on grain structure evolution in Fe-Si soft magnetic alloys made by laser-powder bed fusion

https://doi.org/10.1016/j.addma.2021.102578Get rights and content

Highlights

  • Different scan strategies result in different microstructures for thin wall samples printed in Soft magnetic Fe-3Si Steel.

  • Differences in resulting microstructure between samples can be attributed to differences in the resulting thermal gradient.

  • Annealing of samples show that the sample with a higher number of equiaxed grains saw faster grain growth.

  • Similar geometric necessary dislocation (GND) densities were noted for the as-built samples but not the annealed samples.

  • Differences in annealing response is attributed to differences in grain size and density of high angle grain boundaries.

Abstract

A major goal in printing soft magnetic Fe-Si steels using additive manufacturing is to take advantage of the potential for complex geometric designs and site-specific grain control. One major step in the processing of these alloys is understanding how processing parameters might impact how the as-built microstructure responds to annealing (i.e. the annealing response). The impact of scan strategy on the annealing response for thin wall geometries is specifically explored. Two scan strategies were explored for a thin wall geometry that produced a strongly columnar grain structure and equiaxed grain structure. Samples from both scan strategies annealed at 1200 °C showed a marked difference in annealing response with the more equiaxed sample seeing full recrystallization and grain growth, while the more columnar grain structure saw little change in microstructure. After analysis through characterization techniques and thermal-mechanical simulations Differences in internal energy within the grains were ruled out because calculated GND density values were similar for both samples. The formation of secondary particles was ruled out as a contributing factor due to the type of oxide formations and their size. It was concluded that the contributing factor to the difference in the annealing response were a difference in the resulting grain size and the density of high angle grain boundaries. These two differences were largely attributed to differences in the thermal gradient conditions due to grains preferentially growing in the direction of the steepest thermal gradient.

Introduction

Metal additive manufacturing (AM) processes have transitioned from prototype to structural applications because of their ability to construct complex and light weight geometries with tailored microstructures designed for complex loading conditions. A next step is to extend the same ability to impart site-specific functional properties based on electrical and magnetic phenomena. If proven, the ability to impact functional and mechanical properties in complex geometries opens up new avenues for the adoption of AM. Recent publications have considered various ferromagnetic alloys including Fe-Co [1], [2], [3], [4] and Fe-Ni [2], [5], [6] due to their importance in electrical transformer cores and electrical motor applications. Soft-magnetic Fe-Si alloys, the focus of current research, should exhibit high magnetic permeability and electrical resistivity which is desirable for reducing power losses [7]. Generally, on increasing of the silicon content up to 6.5 wt% we can increase magnetic permeability and electrical resistivity [8]. The fundamental questions explored in this paper: can we produce site-specific microstructures in Fe-Si alloys through modification of laser scanning strategy during the laser powder bed fusion (L-PBF) process and subsequent heat treatment. Second, what is the role of solidification microstructure, solid-state phase transformations and evolution of plastic strain gradients due to spatially and temporally varying thermo-mechanical signatures during AM?

For soft magnetic steels, reduction of power losses usually involves the minimization of hysteresis and eddy current losses [8]. Hysteresis losses are dependent on crystallographic texture, grain size and material properties. To decrease the hysteresis losses, materials will be processed to have a crystallographic texture aligned in the < 001 > direction known as the easy magnetization direction [7]. Hysteresis losses are also decreased with increasing grain size, however as Eddy current losses also increase with increasing grain size, an optimum grain diameter is reported to be 1 mm [7]. Eddy current losses, resulting from induced electrical currents, are related to electrical resistivity and device geometry [7]. To minimize eddy current losses, thin wall geometries are traditionally utilized. Electrical resistivity which decreases eddy current losses is maximized by picking an optimized composition such as Fe-6.5Si wt%.

Current research under AM have looked at various methods for the reduction of power losses [9], [10], [11], [12], [13] within soft magnetic steels. For instance, the potential for the use of alloys with higher silicon content between 6 and 7 wt% that are traditionally too brittle for conventional deformation processes due to the formation of ordered B2 and D03 phases [14], [15], [16]. However, these alloys have been successfully processed using AM [12], [17]. The use of specific geometries such as Hilbert curve geometries within transformer cores have been demonstrated to further reduce eddy currents [9], [10]. Hybrid mesostructures have also been developed to reduce eddy currents further and improve AC power losses [11], [18]. Further, power losses within soft magnetic steels have been shown to be impacted by crystallographic texture along the < 001 > direction. It has been shown that through tailoring of the crystallographic texture by manipulation of AM processing conditions the performance of soft magnetic steels can be changed [13]. More specifically, it has been shown that under L-PBF conditions the Fe-Si alloy’s grain texture is strongly influenced by the processing parameters. Garibaldi et. al. [12] showed it is possible to change the < 001 > aligned crystallographic texture to a cubic texture by increasing the laser power along the build direction. Plotkowski et. al. [9] showed texture could be controlled through different scan strategies with a single rotating raster scan and double rotating raster scan. The single rotating raster scan strategy was described as having a chaotic structure with many small grains while the double rotating scan strategy was noted to have larger grains with a strong < 001 > texture oriented along the build direction. Overall, the development of soft magnetic steels, utilizing the unique possibilities of AM to potentially reduce power losses, have been demonstrated by Plotkowski et. al. [10]. These authors developed a transformer core using a unique Hilbert curve-based print geometry with Fe-6Si and Fe-3Si alloys. This is indeed remarkable since the Fe-6Si alloy is traditionally known to be too brittle to process with deformation processes. These authors also showed power losses lower than conventional non-oriented sheet, but higher power losses then Goss oriented steel.

As mentioned previously, the crystallographic texture is important in limiting power losses. This phenomenon is rationalized based on the characterization of the magnetic properties from samples processed with L-PBF. Within AM builds, one could use the published columnar to equiaxed transition (CET) solidification theories to control the crystallographic texture. The paper by Kustas et. al. [3] questioned the generality of this theory based on their observation of equiaxed grain formation in regions where columnar solidification was predicted during direct energy deposition (DED) processing. These discrepancies could be associated with recrystallization and grain growth that may be triggered during extended high-temperature excursions during the build process. This instability of solidification grain structure can be potentially leveraged to induce grain growth[19]. Interestingly, under conventional manufacturing processes for soft magnetic steels a multi-step production method is utilized to encourage the formation and abnormal grain growth (AGG) with Goss oriented grains along the rolling direction for optimal magnetic properties. In general, traditional processing has shown that recrystallization and grain growth within soft magnetic steels to be highly sensitive to initial microstructure. Within AM there are no published work on the influence of as-built microstructure on the annealing response and final microstructure. It has also not been confirmed whether AGG or preferred grain growth of the Goss texture can occur during AM. In this research, we hypothesize that the final Fe-Si microstructure evolution will be sensitive to the processing parameters and scan strategy, because these changes are expected to impact spatial and temporal variations of solidification, solid-state phase transformation and accumulated plastic strains.

In our literature review, we found only a limited number of publications on the annealing of Fe-Si additive produced parts [9], [10], [11], [13], [19]. Most of these papers considered compositions with high silicon content [11], [13], [19]. Of the research reviewed here, only papers by Plotkowski et. al. [9], [10] have looked at the annealing of Fe-3Si. Their works have reported the use of two annealing regimes, one at 1150 °C for 1 h and one at 1200 °C for 4 h. Both papers have reported large grain growth for samples annealed at 1150 °C. Interestingly, despite the more aggressive annealing regime they reported very little grain growth for samples annealed at 1200 °C for 4 h. These publications, however, do not characterize the impact of processing parameters on the annealing response. There has been some work for L-PBF for other alloys that have shown processing parameters can have an influence on the annealing response, similar to the research performed by Gao et. al. [20] for 316 L. Therefore, in this research we produced thin-walled Fe-3Si wt% alloy builds using L-PBF with different scan strategies and characterized the potential impact on the microstructure after annealing. The results were rationalized based on integrated computational models capable of predicting thermal gradients, liquid-solid interface velocity, thermal stresses, plastic deformation, and residual stresses with the expressed goal of understanding how processing parameters impact the annealing response within Fe-Si alloys.

Section snippets

Material composition

A custom gas atomized Fe-Si powder produced by Praxair Surface Technologies, was used in this research. The powders were screened for a 15–44 µm particle size range. The nominal composition is given in Table 1 and was measured according to the ASTM B215 standard. A calculation using a Scheil-Gulliver solidification simulation (Fig. 1) with ThermoCalc® software and TCFE9 database showed that the alloy under non-equilibrium solidification conditions will undergo the following sequence of phase

Geometrical conformity of the builds

Build results are presented in Fig. 5. No builds were able to meet the 30 mm build height due to the occurrence of delamination during the build. Of note is the occurrence of delamination at consistent levels dependent on scan strategy. The transverse scan strategy showed the lowest mean delamination height of 19.0 mm, followed by the rotated scan strategy at 23.1 mm and the longitudinal scan strategy with the highest at 27.3 mm. The lack of successful builds here is not so much a

Discussion

It is clear from Fig. 6 and Fig. 9 that the annealing response of initial microstructure can be attributed to the different scan strategies and their role on recrystallization and grain growth. Various attributes within the as-built microstructure can influence the driving force for recrystallization and growth. Some of these attributes include grain size, grain orientation, grain misorientation, grain boundary pinning from secondary particles, and dislocation density. Specifically, a higher

Conclusion

This research successfully demonstrated the feasibility of using longitudinal and transverse scan strategies with respect to the sample geometry to modulate the BCC ferrite microstructure in the as built and annealed condition. The recrystallized grain size and texture evolution were rationalized based on detailed characterization and modeling tools and published theories of texture evolution during solidification. The results indicate that the recrystallization texture is predominantly

CRediT authorship contribution statement

MP Haines: Conceptualization, Methodology, Investigation, Formal analysis, Writing – original draft, Writing – review & editing, Visualization F List III: Investigation, Validation. K Carver: Investigation, Validation. DN Leonard: Investigation, Formal analysis, Resources. A Plotkowski: Supervision, Conceptualization, Methodology. CM Fancher: Supervision, Conceptualization, Methodology, Investigation, Formal analysis. RR Dehoff: Supervision SS Babu: Supervision, Project administration, Writing

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 sponsored by the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, USA and U.S. Department of Energy Advanced Manufacturing Office, USA under contract DE-AC05-00OR22725 with UT-Battelle, LLC. SSB’s contribution this research is partially supported from the US Department of the Navy, Office of Naval Research, USA under ONR award number N00014-18-1-2794. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the

References (68)

  • J.S. Shin et al.

    Ordering-disordering phenomena and micro-hardness characteristics of B2 phase in Fe-(5-6.5%)Si alloys

    Mater. Sci. Eng. A

    (2005)
  • B. Viala et al.

    Study of the brittle behaviour of annealed Fe-6.5 wt%Si ribbons produced by planar flow casting

    Mater. Sci. Eng. A

    (1996)
  • J.N. Lemke et al.

    Calorimetric study and microstructure analysis of the order-disorder phase transformation in silicon steel built by SLM

    J. Alloy. Compd.

    (2017)
  • B. Kocsis et al.

    Metallographic and magnetic analysis of direct laser sintered soft magnetic composites

    J. Magn. Magn. Mater.

    (2020)
  • M. Garibaldi et al.

    Effect of annealing on the microstructure and magnetic properties of soft magnetic Fe-Si produced via laser additive manufacturing

    Scr. Mater.

    (2018)
  • S. Gao et al.

    Recrystallization-based grain boundary engineering of 316L stainless steel produced via selective laser melting

    Acta Mater.

    (2020)
  • Y. Hayakawa et al.

    A new model of Goss texture development during secondary recrystallization of electrical steel

    Acta Mater.

    (1997)
  • R. Shimizu et al.

    Coincidence grain boundary and texture evolution in Fe-3%Si

    Acta Met.

    (1989)
  • E.J. Schwalbach et al.

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

    Addit. Manuf.

    (2019)
  • A. Rai et al.

    A coupled Cellular Automaton–Lattice Boltzmann model for grain structure simulation during additive manufacturing

    Comput. Mater. Sci.

    (2016)
  • 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)
  • P. Prabhakar et al.

    Computational modeling of residual stress formation during the electron beam melting process for Inconel 718

    Addit. Manuf.

    (2015)
  • X. Tan et al.

    An experimental and simulation study on build thickness dependent microstructure for electron beam melted Ti-6Al-4V

    J. Alloy. Compd.

    (2015)
  • H. Ali et al.

    Effect of scanning strategies on residual stress and mechanical properties of Selective Laser Melted Ti6Al4V

    Mater. Sci. Eng. A.

    (2018)
  • B. Cheng et al.

    Stress and deformation evaluations of scanning strategy effect in selective laser melting

    Addit. Manuf.

    (2016)
  • E. Assuncao et al.

    Comparison of continuous wave and pulsed wave laser welding effects

    Opt. Lasers Eng.

    (2013)
  • X. Ding et al.

    Comparison study of numerical analysis for heat transfer and fluid flow under two different laser scan pattern during selective laser melting

    Optics

    (2016)
  • S. Li et al.

    Melt-pool motion, temperature variation and dendritic morphology of Inconel 718 during pulsed- and continuous-wave laser additive manufacturing: a comparative study

    Mater. Des.

    (2017)
  • J.-P.A. Immarigeon et al.

    The deformation of armco iron and silicon steel in the vicinity of the curie temperature

    Acta Met.

    (1974)
  • W. Pantleon

    Resolving the geometrically necessary dislocation content by conventional electron backscattering diffraction

    Scr. Mater.

    (2008)
  • S. Yoder et al.

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

    Addit. Manuf.

    (2019)
  • X. Lou et al.

    Oxide inclusions in laser additive manufactured stainless steel and their effects on impact toughness and stress corrosion cracking behavior

    J. Nucl. Mater.

    (2018)
  • F. Yan et al.

    Characterization of nano-scale oxides in austenitic stainless steel processed by powder bed fusion

    Scr. Mater.

    (2018)
  • Y.S. Lee et al.

    Correlations of cracking with scan strategy and build geometry in electron beam powder bed additive manufacturing

    Addit. Manuf.

    (2020)
  • Cited by (11)

    • 3D printed metal-insulator layered structure

      2022, Journal of Magnetism and Magnetic Materials
      Citation Excerpt :

      A new branch of powder technology is the selective laser melting (SLM) based on laser powder bed fusion technology (l-PBF) [1].

    View all citing articles on Scopus
    View full text