Coupled CFD-DEM simulation of interfacial fluid–particle interaction during binder jet 3D printing

Coupled CFD-DEM simulation of interfacial fluid–particle interaction during binder jet 3D printing

바인더 제트 3D 프린팅 중 계면 유체-입자 상호 작용에 대한 CFD-DEM 결합 시뮬레이션

Joshua J. Wagner, C. Fred Higgs III

https://doi.org/10.1016/j.cma.2024.116747

Abstract

The coupled dynamics of interfacial fluid phases and unconstrained solid particles during the binder jet 3D printing process govern the final quality and performance of the resulting components. The present work proposes a computational fluid dynamics (CFD) and discrete element method (DEM) framework capable of simulating the complex interfacial fluid–particle interaction that occurs when binder microdroplets are deposited into a powder bed. The CFD solver uses a volume-of-fluid (VOF) method for capturing liquid–gas multifluid flows and relies on block-structured adaptive mesh refinement (AMR) to localize grid refinement around evolving fluid–fluid interfaces. The DEM module resolves six degrees of freedom particle motion and accounts for particle contact, cohesion, and rolling resistance. Fully-resolved CFD-DEM coupling is achieved through a fictitious domain immersed boundary (IB) approach. An improved method for enforcing three-phase contact lines with a VOF-IB extension technique is introduced. We present several simulations of binder jet primitive formation using realistic process parameters and material properties. The DEM particle systems are experimentally calibrated to reproduce the cohesion behavior of physical nickel alloy powder feedstocks. We demonstrate the proposed model’s ability to resolve the interdependent fluid and particle dynamics underlying the process by directly comparing simulated primitive granules with one-to-one experimental counterparts obtained from an in-house validation apparatus. This computational framework provides unprecedented insight into the fundamental mechanisms of binder jet 3D printing and presents a versatile new approach for process parameter optimization and defect mitigation that avoids the inherent challenges of experiments.

바인더 젯 3D 프린팅 공정 중 계면 유체 상과 구속되지 않은 고체 입자의 결합 역학이 결과 구성 요소의 최종 품질과 성능을 좌우합니다. 본 연구는 바인더 미세액적이 분말층에 증착될 때 발생하는 복잡한 계면 유체-입자 상호작용을 시뮬레이션할 수 있는 전산유체역학(CFD) 및 이산요소법(DEM) 프레임워크를 제안합니다.

CFD 솔버는 액체-가스 다중유체 흐름을 포착하기 위해 VOF(유체량) 방법을 사용하고 블록 구조 적응형 메쉬 세분화(AMR)를 사용하여 진화하는 유체-유체 인터페이스 주위의 그리드 세분화를 국지화합니다. DEM 모듈은 6개의 자유도 입자 운동을 해결하고 입자 접촉, 응집력 및 구름 저항을 설명합니다.

완전 분해된 CFD-DEM 결합은 가상 도메인 침지 경계(IB) 접근 방식을 통해 달성됩니다. VOF-IB 확장 기술을 사용하여 3상 접촉 라인을 강화하는 향상된 방법이 도입되었습니다. 현실적인 공정 매개변수와 재료 특성을 사용하여 바인더 제트 기본 형성에 대한 여러 시뮬레이션을 제시합니다.

DEM 입자 시스템은 물리적 니켈 합금 분말 공급원료의 응집 거동을 재현하기 위해 실험적으로 보정되었습니다. 우리는 시뮬레이션된 기본 과립과 내부 검증 장치에서 얻은 일대일 실험 대응물을 직접 비교하여 프로세스의 기본이 되는 상호 의존적인 유체 및 입자 역학을 해결하는 제안된 모델의 능력을 보여줍니다.

이 계산 프레임워크는 바인더 제트 3D 프린팅의 기본 메커니즘에 대한 전례 없는 통찰력을 제공하고 실험에 내재된 문제를 피하는 공정 매개변수 최적화 및 결함 완화를 위한 다용도의 새로운 접근 방식을 제시합니다.

Introduction

Binder jet 3D printing (BJ3DP) is a powder bed additive manufacturing (AM) technology capable of fabricating geometrically complex components from advanced engineering materials, such as metallic superalloys and ultra-high temperature ceramics [1], [2]. As illustrated in Fig. 1(a), the process is comprised of many repetitive print cycles, each contributing a new cross-sectional layer on top of a preceding one to form a 3D CAD-specified geometry. The feedstock material is first delivered from a hopper to a build plate and then spread into a thin layer by a counter-rotating roller. After powder spreading, a print head containing many individual inkjet nozzles traverses over the powder bed while precisely jetting binder microdroplets onto select regions of the spread layer. Following binder deposition, the build plate lowers by a specified layer thickness, leaving a thin void space at the top of the job box that the subsequent powder layer will occupy. This cycle repeats until the full geometries are formed layer by layer. Powder bed fusion (PBF) methods follow a similar procedure, except they instead use a laser or electron beam to selectively melt and fuse the powder material. Compared to PBF, binder jetting offers several distinct advantages, including faster build rates, enhanced scalability for large production volumes, reduced machine and operational costs, and a wider selection of suitable feedstock materials [2]. However, binder jetted parts generally possess inferior mechanical properties and reduced dimensional accuracy [3]. As a result, widescale adoption of BJ3DP to fabricate high-performance, mission-critical components, such as those common to the aerospace and defense sectors, is contingent on novel process improvements and innovations [4].

A major obstacle hindering the advancement of BJ3DP is our limited understanding of how various printing parameters and material properties collectively influence the underlying physical mechanisms of the process and their effect on the resulting components. To date, the vast majority of research efforts to uncover these relationships have relied mainly on experimental approaches [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], which are often expensive and time-consuming and have inherent physical restrictions on what can be measured and observed. For these reasons, there is a rapidly growing interest in using computational models to circumvent the challenges of experimental investigations and facilitate a deeper understanding of the process’s fundamental phenomena. While significant progress has been made in developing and deploying numerical frameworks aimed at powder spreading [20], [21], [22], [23], [24], [25], [26], [27] and sintering [28], [29], [30], [31], [32], simulating the interfacial fluid–particle interaction (IFPI) in the binder deposition stage is still in its infancy. In their exhaustive review, Mostafaei et al. [2] point out the lack of computational models capable of resolving the coupled fluid and particle dynamics associated with binder jetting and suggest that the development of such tools is critical to further improving the process and enhancing the quality of its end-use components.

We define IFPI as a multiphase flow regime characterized by immiscible fluid phases separated by dynamic interfaces that intersect the surfaces of moving solid particles. As illustrated in Fig. 1(b), an elaborate IFPI occurs when a binder droplet impacts the powder bed in BJ3DP. The momentum transferred from the impacting droplet may cause powder compaction, cratering, and particle ejection. These ballistic disturbances can have deleterious effects on surface texture and lead to the formation of large void spaces inside the part [5], [13]. After impact, the droplet spreads laterally on the bed surface and vertically into the pore network, driven initially by inertial impact forces and then solely by capillary action [33]. Attractive capillary forces exerted on mutually wetted particles tend to draw them inward towards each other, forming a packed cluster of bound particles referred to as a primitive [34]. A single-drop primitive is the most fundamental building element of a BJ3DP part, and the interaction leading to its formation has important implications on the final part characteristics, such as its mechanical properties, resolution, and dimensional accuracy. Generally, binder droplets are deposited successively as the print head traverses over the powder bed. The traversal speed and jetting frequency are set such that consecutive droplets coalesce in the bed, creating a multi-drop primitive line instead of a single-drop primitive granule. The binder must be jetted with sufficient velocity to penetrate the powder bed deep enough to provide adequate interlayer binding; however, a higher impact velocity leads to more pronounced ballistic effects.

A computational framework equipped to simulate the interdependent fluid and particle dynamics in BJ3DP would allow for unprecedented observational and measurement capability at temporal and spatial resolutions not currently achievable by state-of-the-art imaging technology, namely synchrotron X-ray imaging [13], [14], [18], [19]. Unfortunately, BJ3DP presents significant numerical challenges that have slowed the development of suitable modeling frameworks; the most significant of which are as follows:

  • 1.Incorporating dynamic fluid–fluid interfaces with complex topological features remains a nontrivial task for standard mesh-based CFD codes. There are two broad categories encompassing the methods used to handle interfacial flows: interface tracking and interface capturing [35]. Interface capturing techniques, such as the popular volume-of-fluid (VOF) [36] and level-set methods [37], [38], are better suited for problems with interfaces that become heavily distorted or when coalescence and fragmentation occur frequently; however, they are less accurate in resolving surface tension and boundary layer effects compared to interface tracking methods like front-tracking [39], arbitrary Lagrangian–Eulerian [40], and space–time finite element formulations [41]. Since interfacial forces become increasingly dominant at decreasing length scales, inaccurate surface tension calculations can significantly deteriorate the fidelity of IFPI simulations involving <100 μm droplets and particles.
  • 2.Dynamic powder systems are often modeled using the discrete element method (DEM) introduced by Cundall and Strack [42]. For IFPI problems, a CFD-DEM coupling scheme is required to exchange information between the fluid and particle solvers. Fully-resolved CFD-DEM coupling suggests that the flow field around individual particle surfaces is resolved on the CFD mesh [43], [44]. In contrast, unresolved coupling volume averages the effect of the dispersed solid phase on the continuous fluid phases [45], [46], [47], [48]. Comparatively, the former is computationally expensive but provides detailed information about the IFPI in question and is more appropriate when contact line dynamics are significant. However, since the pore structure of a powder bed is convoluted and evolves with time, resolving such solid–fluid interfaces on a computational mesh presents similar challenges as fluid–fluid interfaces discussed in the previous point. Although various algorithms have been developed to deform unstructured meshes to accommodate moving solid surfaces (see Bazilevs et al. [49] for an overview of such methods), they can be prohibitively expensive when frequent topology changes require mesh regeneration rather than just modification through nodal displacement. The pore network in a powder bed undergoes many topology changes as particles come in and out of contact with each other, constantly closing and opening new flow channels. Non-body-conforming structured grid approaches that rely on immersed boundary (IB) methods to embed the particles in the flow field can be better suited for such cases [50]. Nevertheless, accurately representing these complex pore geometries on Cartesian grids requires extremely high mesh resolutions, which can impose significant computational costs.
  • 3.Capillary effects depend on the contact angle at solid–liquid–gas intersections. Since mesh nodes do not coincide with a particle surface when using an IB method on structured grids, imposing contact angle boundary conditions at three-phase contact lines is not straightforward.

While these issues also pertain to PBF process modeling, resolving particle motion is generally less crucial for analyzing melt pool dynamics compared to primitive formation in BJ3DP. Therefore, at present, the vast majority of computational process models of PBF assume static powder beds and avoid many of the complications described above, see, e.g., [51], [52], [53], [54], [55], [56], [57], [58], [59]. Li et al. [60] presented the first 2D fully-resolved CFD-DEM simulations of the interaction between the melt pool, powder particles, surrounding gas, and metal vapor in PBF. Following this work, Yu and Zhao [61], [62] published similar melt pool IFPI simulations in 3D; however, contact line dynamics and capillary forces were not considered. Compared to PBF, relatively little work has been published regarding the computational modeling of binder deposition in BJ3DP. Employing the open-source VOF code Gerris [63], Tan [33] first simulated droplet impact on a powder bed with appropriate binder jet parameters, namely droplet size and impact velocity. However, similar to most PBF melt pool simulations described in the current literature, the powder bed was fixed in place and not allowed to respond to the interacting fluid phases. Furthermore, a simple face-centered cubic packing of non-contacting, monosized particles was considered, which does not provide a realistic pore structure for AM powder beds. Building upon this approach, we presented a framework to simulate droplet impact on static powder beds with more practical particle size distributions and packing arrangements [64]. In a study similar to [33], [64], Deng et al. [65] used the VOF capability in Ansys Fluent to examine the lateral and vertical spreading of a binder droplet impacting a fixed bimodal powder bed with body-centered packing. Li et al. [66] also adopted Fluent to conduct 2D simulations of a 100 μm diameter droplet impacting substrates with spherical roughness patterns meant to represent the surface of a simplified powder bed with monosized particles. The commercial VOF-based software FLOW-3D offers an AM module centered on process modeling of various AM technologies, including BJ3DP. However, like the above studies, particle motion is still not considered in this codebase. Ur Rehman et al. [67] employed FLOW-3D to examine microdroplet impact on a fixed stainless steel powder bed. Using OpenFOAM, Erhard et al. [68] presented simulations of different droplet impact spacings and patterns on static sand particles.

Recently, Fuchs et al. [69] introduced an impressive multipurpose smoothed particle hydrodynamics (SPH) framework capable of resolving IFPI in various AM methods, including both PBF and BJ3DP. In contrast to a combined CFD-DEM approach, this model relies entirely on SPH meshfree discretization of both the fluid and solid governing equations. The authors performed several prototype simulations demonstrating an 80 μm diameter droplet impacting an unconstrained powder bed at different speeds. While the powder bed responds to the hydrodynamic forces imparted by the impacting droplet, the particle motion is inconsistent with experimental time-resolved observations of the process [13]. Specifically, the ballistic effects, such as particle ejection and bed deformation, were drastically subdued, even in simulations using a droplet velocity ∼ 5× that of typical jetting conditions. This behavior could be caused by excessive damping in the inter-particle contact force computations within their SPH framework. Moreover, the wetted particles did not appear to be significantly influenced by the strong capillary forces exerted by the binder as no primitive agglomeration occurred. The authors mention that the objective of these simulations was to demonstrate their codebase’s broad capabilities and that some unrealistic process parameters were used to improve computational efficiency and stability, which could explain the deviations from experimental observations.

In the present paper, we develop a novel 3D CFD-DEM numerical framework for simulating fully-resolved IFPI during binder jetting with realistic material properties and process parameters. The CFD module is based on the VOF method for capturing binder–air interfaces. Surface tension effects are realized through the continuum surface force (CSF) method with height function calculations of interface curvature. Central to our fluid solver is a proprietary block-structured AMR library with hierarchical octree grid nesting to focus enhanced grid resolution near fluid–fluid interfaces. The GPU-accelerated DEM module considers six degrees of freedom particle motion and includes models based on Hertz-Mindlin contact, van der Waals cohesion, and viscoelastic rolling resistance. The CFD and DEM modules are coupled to achieve fully-resolved IFPI using an IB approach in which Lagrangian solid particles are mapped to the underlying Eulerian fluid mesh through a solid volume fraction field. An improved VOF-IB extension algorithm is introduced to enforce the contact angle at three-phase intersections. This provides robust capillary flow behavior and accurate computations of the fluid-induced forces and torques acting on individual wetted particles in densely packed powder beds.

We deploy our integrated codebase for direct numerical simulations of single-drop primitive formation with powder beds whose particle size distributions are generated from corresponding laboratory samples. These simulations use jetting parameters similar to those employed in current BJ3DP machines, fluid properties that match commonly used aqueous polymeric binders, and powder properties specific to nickel alloy feedstocks. The cohesion behavior of the DEM powder is calibrated based on the angle of repose of the laboratory powder systems. The resulting primitive granules are compared with those obtained from one-to-one experiments conducted using a dedicated in-house test apparatus. Finally, we demonstrate how the proposed framework can simulate more complex and realistic printing operations involving multi-drop primitive lines.

Section snippets

Mathematical description of interfacial fluid–particle interaction

This section briefly describes the governing equations of fluid and particle dynamics underlying the CFD and DEM solvers. Our unified framework follows an Eulerian–Lagrangian approach, wherein the Navier–Stokes equations of incompressible flow are discretized on an Eulerian grid to describe the motion of the binder liquid and surrounding gas, and the Newton–Euler equations account for the positions and orientations of the Lagrangian powder particles. The mathematical foundation for

CFD solver for incompressible flow with multifluid interfaces

This section details the numerical methodology used in our CFD module to solve the Navier–Stokes equations of incompressible flow. First, we introduce the VOF method for capturing the interfaces between the binder and air phases. This approach allows us to solve the fluid dynamics equations considering only a single continuum field with spatial and temporal variations in fluid properties. Next, we describe the time integration procedure using a fractional-step projection algorithm for

DEM solver for solid particle dynamics

This section covers the numerical procedure for tracking the motion of individual powder particles with DEM. The Newton–Euler equations (Eqs. (10), (11)) are ordinary differential equations (ODEs) for which many established numerical integrators are available. In general, the most challenging aspects of DEM involve processing particle collisions in a computationally efficient manner and dealing with small time step constraints that result from stiff materials, such as metallic AM powders. The

Unified CFD-DEM solver

The preceding sections have introduced the CFD and DEM solution algorithms separately. Here, we discuss the integrated CFD-DEM solution algorithm and related details.

Binder jet process modeling and validation experiments

In this section, we deploy our CFD-DEM framework to simulate the IFPI occurring during the binder droplet deposition stage of the BJ3DP process. The first simulations attempt to reproduce experimental single-drop primitive granules extracted from four nickel alloy powder samples with varying particle size distributions. The experiments are conducted with a dedicated in-house test apparatus that allows for the precision deposition of individual binder microdroplets into a powder bed sample. The

Conclusions

This paper introduces a coupled CFD-DEM framework capable of fully-resolved simulation of the interfacial fluid–particle interaction occurring in the binder jet 3D printing process. The interfacial flow of binder and surrounding air is captured with the VOF method and surface tension effects are incorporated using the CSF technique augmented by height function curvature calculations. Block-structured AMR is employed to provide localized grid refinement around the evolving liquid–gas interface.

CRediT authorship contribution statement

Joshua J. Wagner: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing. C. Fred Higgs III: Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing.

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.

Acknowledgments

This work was supported by a NASA Space Technology Research Fellowship, United States of America, Grant No. 80NSSC19K1171. Partial support was also provided through an AIAA Foundation Orville, USA and Wilbur Wright Graduate Award, USA . The authors would like to gratefully acknowledge Dr. Craig Smith of NASA Glenn Research Center for the valuable input he provided on this project.

References (155)

Influences of the Powder Size and Process Parameters on the Quasi-Stability of Molten Pool Shape in Powder Bed Fusion-Laser Beam of Molybdenum

Influences of the Powder Size and Process Parameters on the Quasi-Stability of Molten Pool Shape in Powder Bed Fusion-Laser Beam of Molybdenum

몰리브덴 분말층 융합-레이저 빔의 용융 풀 형태의 준안정성에 대한 분말 크기 및 공정 매개변수의 영향

Abstract

Formation of a quasi-steady molten pool is one of the necessary conditions for achieving excellent quality in many laser processes. The influences of distribution characteristics of powder sizes on quasi-stability of the molten pool shape during single-track powder bed fusion-laser beam (PBF-LB) of molybdenum and the underlying mechanism were investigated.

The feasibility of improving quasi-stability of the molten pool shape by increasing the laser energy conduction effect and preheating was explored. Results show that an increase in the range of powder sizes does not significantly influence the average laser energy conduction effect in PBF-LB process. Whereas, it intensifies fluctuations of the transient laser energy conduction effect.

It also leads to fluctuations of the replenishment rate of metals, difficulty in formation of the quasi-steady molten pool, and increased probability of incomplete fusion and pores defects. As the laser power rises, the laser energy conduction effect increases, which improves the quasi-stability of the molten pool shape. When increasing the laser scanning speed, the laser energy conduction effect grows.

However, because the molten pool size reduces due to the decreased heat input, the replenishment rate of metals of the molten pool fluctuates more obviously and the quasi-stability of the molten pool shape gets worse. On the whole, the laser energy conduction effect in the PBF-LB process of Mo is low (20-40%). The main factor that affects quasi-stability of the molten pool shape is the amount of energy input per unit length of the scanning path, rather than the laser energy conduction effect.

Moreover, substrate preheating can not only enlarge the molten pool size, particularly the length, but also reduce non-uniformity and discontinuity of surface morphologies of clad metals and inhibit incomplete fusion and pores defects.

준안정 용융 풀의 형성은 많은 레이저 공정에서 우수한 품질을 달성하는 데 필요한 조건 중 하나입니다. 몰리브덴의 단일 트랙 분말층 융합 레이저 빔(PBF-LB) 동안 용융 풀 형태의 준안정성에 대한 분말 크기 분포 특성의 영향과 그 기본 메커니즘을 조사했습니다.

레이저 에너지 전도 효과와 예열을 증가시켜 용융 풀 형태의 준안정성을 향상시키는 타당성을 조사했습니다. 결과는 분말 크기 범위의 증가가 PBF-LB 공정의 평균 레이저 에너지 전도 효과에 큰 영향을 미치지 않음을 보여줍니다. 반면, 과도 레이저 에너지 전도 효과의 변동이 강화됩니다.

이는 또한 금속 보충 속도의 변동, 준안정 용융 풀 형성의 어려움, 불완전 융합 및 기공 결함 가능성 증가로 이어집니다. 레이저 출력이 증가함에 따라 레이저 에너지 전도 효과가 증가하여 용융 풀 모양의 준 안정성이 향상됩니다. 레이저 스캐닝 속도를 높이면 레이저 에너지 전도 효과가 커집니다.

그러나 열 입력 감소로 인해 용융 풀 크기가 줄어들기 때문에 용융 풀의 금속 보충 속도의 변동이 더욱 뚜렷해지고 용융 풀 형태의 준안정성이 악화됩니다.

전체적으로 Mo의 PBF-LB 공정에서 레이저 에너지 전도 효과는 낮다(20~40%). 용융 풀 형상의 준안정성에 영향을 미치는 주요 요인은 레이저 에너지 전도 효과보다는 스캐닝 경로의 단위 길이당 입력되는 에너지의 양입니다.

또한 기판 예열은 용융 풀 크기, 특히 길이를 확대할 수 있을 뿐만 아니라 클래드 금속 표면 형태의 불균일성과 불연속성을 줄이고 불완전한 융합 및 기공 결함을 억제합니다.

References

  1. M. Sharifitabar, F.O. Sadeq, and M.S. Afarani, Synthesis and Kinetic Study of Mo (Si, Al)2 Coatings on the Surface of Molybdenum Through Hot Dipping into a Commercial Al-12 wt.% Si Alloy Melt, Surf. Interfaces, 2021, 24, p 101044.Article CAS Google Scholar 
  2. Z. Zhang, X. Li, and H. Dong, Response of a Molybdenum Alloy to Plasma Nitriding, Int. J. Refract. Met. Hard Mater., 2018, 72, p 388–395.Article CAS Google Scholar 
  3. C. Tan, K. Zhou, M. Kuang, W. Ma, and T. Kuang, Microstructural Characterization and Properties of Selective Laser Melted Maraging Steel with Different Build Directions, Sci. Technol. Adv. Mater., 2018, 19(1), p 746–758.Article CAS Google Scholar 
  4. C. Tan, F. Weng, S. Sui, Y. Chew, and G. Bi, Progress and Perspectives in Laser Additive Manufacturing of Key Aeroengine Materials, Int. J. Mach. Tools Manuf, 2021, 170, p 103804.Article Google Scholar 
  5. S.A. Khairallah and A. Anderson, Mesoscopic Simulation Model of Selective Laser Melting of Stainless Steel Powder, J. Mater. Process. Technol., 2014, 214(11), p 2627–2636.Article CAS Google Scholar 
  6. S.A. Khairallah, A.T. Anderson, A. Rubenchik, and W.E. King, Laser Powder-Bed Fusion Additive Manufacturing: Physics of Complex Melt Flow and Formation Mechanisms of Pores, Spatter, and Denudation Zones, Acta Mater., 2016, 108, p 36–45.Article CAS ADS Google Scholar 
  7. K.Q. Le, C. Tang, and C.H. Wong, On the Study of Keyhole-Mode Melting in Selective Laser Melting Process, Int. J. Therm. Sci., 2019, 145, p 105992.Article Google Scholar 
  8. M. Bayat, A. Thanki, S. Mohanty, A. Witvrouw, S. Yang, J. Thorborg, N.S. Tiedje, and J.H. Hattel, Keyhole-Induced Porosities in Laser-Based Powder Bed Fusion (L-PBF) of Ti6Al4V: High-Fidelity Modelling and Experimental Validation, Addit. Manuf., 2019, 30, p 100835.CAS Google Scholar 
  9. B. Liu, G. Fang, L. Lei, and X. Yan, Predicting the Porosity Defects in Selective Laser Melting (SLM) by Molten Pool Geometry, Int. J. Mech. Sci., 2022, 228, p 107478.Article Google Scholar 
  10. W. Ge, J.Y.H. Fuh, and S.J. Na, Numerical Modelling of Keyhole Formation in Selective Laser Melting of Ti6Al4V, J. Manuf. Process., 2021, 62, p 646–654.Article Google Scholar 
  11. W. Ge, S. Han, S.J. Na, and J.Y.H. Fuh, Numerical Modelling of Surface Morphology in Selective Laser Melting, Comput. Mater. Sci., 2021, 186, p 110062.Article Google Scholar 
  12. Y.-C. Wu, C.-H. San, C.-H. Chang, H.-J. Lin, R. Marwan, S. Baba, and W.-S. Hwang, Numerical Modeling of Melt-Pool Behavior In Selective Laser Melting with Random Powder Distribution and Experimental Validation, J. Mater. Process. Technol., 2018, 254, p 72–78.Article Google Scholar 
  13. C. Tang, J.L. Tan, and C.H. Wong, A Numerical Investigation on the Physical Mechanisms of Single Track Defects in Selective Laser Melting, Int. J. Heat Mass Transf., 2018, 126, p 957–968.Article CAS Google Scholar 
  14. X. Zhou, X. Liu, D. Zhang, Z. Shen, and W. Liu, Balling Phenomena in Selective Laser Melted Tungsten, J. Mater. Process. Technol., 2015, 222, p 33–42.Article CAS Google Scholar 
  15. J.D.K. Monroy and J. Ciurana, Study of the Pore Formation on CoCrMo Alloys by Selective Laser Melting Manufacturing Process, Procedia Eng., 2013, 63, p 361–369.Article CAS Google Scholar 
  16. L. Kaserer, J. Braun, J. Stajkovic, K.H. Leitz, B. Tabernig, P. Singer, I. Letofsky-Papst, H. Kestler, and G. Leichtfried, Fully Dense and Crack Free Molybdenum Manufactured by Selective Laser Melting Through Alloying with Carbon, Int. J. Refract. Met. Hard Mater., 2019, 84, p 105000.Article CAS Google Scholar 
  17. T.B.T. Majumdar, E.M.C. Ribeiro, J.E. Frith, and N. Birbilis, Understanding the Effects of PBF Process Parameter Interplay on Ti-6Al-4V Surface Properties, PLoS ONE, 2019, 14, p e0221198.Article CAS PubMed PubMed Central Google Scholar 
  18. A.K.J.-R. Poulin, P. Terriault, and V. Brailovski, Long Fatigue Crack Propagation Behavior of Laser Powder Bed-Fused Inconel 625 with Intentionally- Seeded Porosity, Int. J. Fatigue, 2019, 127, p 144–156.Article CAS Google Scholar 
  19. P. Rebesan, M. Ballan, M. Bonesso, A. Campagnolo, S. Corradetti, R. Dima, C. Gennari, G.A. Longo, S. Mancin, M. Manzolaro, G. Meneghetti, A. Pepato, E. Visconti, and M. Vedani, Pure Molybdenum Manufactured by Laser Powder Bed Fusion: Thermal and Mechanical Characterization at Room and High Temperature, Addit. Manuf., 2021, 47, p 102277.CAS Google Scholar 
  20. D. Wang, C. Yu, J. Ma, W. Liu, and Z. Shen, Densification and Crack Suppression in Selective Laser Melting of Pure Molybdenum, Mater. Des., 2017, 129, p 44–52.Article CAS Google Scholar 
  21. K.-H. Leitz, P. Singer, A. Plankensteiner, B. Tabernig, H. Kestler, and L.S. Sigl, Multi-physical Simulation of Selective Laser Melting, Met. Powder Rep., 2017, 72, p 331–338.Article Google Scholar 
  22. D.G.J. Zhang, Y. Yang, H. Zhang, H. Chen, D. Dai, and K. Lin, Influence of Particle Size on Laser Absorption and Scanning Track Formation Mechanisms of Pure Tungsten Powder During Selective Laser Melting, Engineering, 2019, 5, p 736–745.Article CAS Google Scholar 
  23. L. Caprio, A.G. Demir, and B. Previtali, Influence of Pulsed and Continuous Wave Emission on Melting Efficiency in Selective Laser Melting, J. Mater. Process. Technol., 2019, 266, p 429–441.Article CAS Google Scholar 
  24. D. Gu, M. Xia, and D. Dai, On the Role of Powder Flow Behavior in Fluid Thermodynamics and Laser Processability of Ni-based Composites by Selective Laser Melting, Int. J. Mach. Tools Manuf, 2018, 137, p 67–78.Article Google Scholar 
  25. W.-I. Cho, S.-J. Na, C. Thomy, and F. Vollertsen, Numerical Simulation of Molten Pool Dynamics in High Power Disk Laser Welding, J. Mater. Process. Technol., 2012, 212(1), p 262–275.Article CAS Google Scholar 
  26. S.W. Han, J. Ahn, and S.J. Na, A Study on Ray Tracing Method for CFD Simulations of Laser Keyhole Welding: Progressive Search Method, Weld. World, 2016, 60, p 247–258.Article CAS Google Scholar 
  27. W. Ge, S. Han, Y. Fang, J. Cheon, and S.J. Na, Mechanism of Surface Morphology in Electron Beam Melting of Ti6Al4V Based on Computational Flow Patterns, Appl. Surf. Sci., 2017, 419, p 150–158.Article CAS ADS Google Scholar 
  28. W.-I. Cho, S.-J. Na, C. Thomy, and F. Vollertsen, Numerical Simulation of Molten Pool Dynamics in High Power Disk Laser Welding, J. Mater. Process. Technol., 2012, 212, p 262–275.Article CAS Google Scholar 
  29. W. Ma, J. Ning, L.-J. Zhang, and S.-J. Na, Regulation of Microstructures and Properties of Molybdenum-Silicon-Boron Alloy Subjected to Selective Laser Melting, J. Manuf. Process., 2021, 69, p 593–601.Article Google Scholar 
  30. S. Haeri, Y. Wang, O. Ghita, and J. Sun, Discrete Element Simulation and Experimental Study of Powder Spreading Process in Additive Manufacturing, Powder Technol., 2016, 306, p 45–54.Article Google Scholar 
  31. D. Yao, X. Liu, J. Wang, W. Fan, M. Li, H. Fu, H. Zhang, X. Yang, Q. Zou, and X. An, Numerical Insights on the Spreading of Practical 316 L Stainless Steel Powder in SLM Additive Manufacturing, Powder Technol., 2021, 390, p 197–208.Article CAS Google Scholar 
  32. S. Vock, B. Klöden, A. Kirchner, T. Weißgärber, and B. Kieback, Powders for Powder Bed Fusion: A Review, Prog. Addit. Manuf., 2019, 4, p 383–397.Article Google Scholar 
  33. X. Luo, C. Yang, Z.Q. Fu, L.H. Liu, H.Z. Lu, H.W. Ma, Z. Wang, D.D. Li, L.C. Zhang, and Y.Y. Li, Achieving Ultrahigh-Strength in Beta-Type Titanium Alloy by Controlling the Melt Pool Mode in Selective Laser Melting, Mater. Sci. Eng. A, 2021, 823, p 141731.Article CAS Google Scholar 
  34. J. Braun, L. Kaserer, J. Stajkovic, K.-H. Leitz, B. Tabernig, P. Singer, P. Leibenguth, C. Gspan, H. Kestler, and G. Leichtfried, Molybdenum and Tungsten Manufactured by Selective Laser Melting: Analysis of Defect Structure and Solidification Mechanisms, Int. J. Refract. Met. Hard Mater., 2019, 84, p 104999.Article CAS Google Scholar 
  35. L. Kaserera, J. Brauna, J. Stajkovica, K.-H. Leitzb, B. Tabernigb, P. Singerb, I. Letofsky-Papstc, H. Kestlerb, and G. Leichtfried, Fully Dense and Crack Free Molybdenum Manufactured by Selective Laser Melting Through Alloying with Carbon, Int. J. Refract Metal Hard Mater., 2019, 84, p 105000.Article Google Scholar 
Figure 3. The simulated temperature distribution and single-layer multi-track isothermograms of LPBF Hastelloy X, located at the bottom of the powder bed, are presented for various laser energy densities. (a) depicts the single-point temperature distribution at the bottom of the powder bed, followed by the isothermograms corresponding to laser energy densities of (b) 31 J/mm3 , (c) 43 J/mm3 , (d) 53 J/mm3 , (e) 67 J/mm3 , and (f) 91 J/mm3 .

An integrated multiscale simulation guiding the processing optimisation for additively manufactured nickel-based superalloys

적층 가공된 니켈 기반 초합금의 가공 최적화를 안내하는 통합 멀티스케일 시뮬레이션

Xing He, Bing Yang, Decheng Kong, Kunjie Dai, Xiaoqing Ni, Zhanghua Chen
& Chaofang Dong

ABSTRACT

Microstructural defects in laser powder bed fusion (LPBF) metallic materials are correlated with processing parameters. A multi-physics model and a crystal plasticity framework are employed to predict microstructure growth in molten pools and assess the impact of manufacturing defects on plastic damage parameters. Criteria for optimising the LPBF process are identified, addressing microstructural defects and tensile properties of LPBF Hastelloy X at various volumetric energy densities (VED). The results show that higher VED levels foster a specific Goss texture {110} <001>, with irregular lack of fusion defects significantly affecting plastic damage, especially near the material surface. A critical threshold emerges between manufacturing defects and grain sizes in plastic strain accumulation. The optimal processing window for superior Hastelloy X mechanical properties ranges from 43 to 53 J/mm3 . This work accelerates the development of superior strengthductility alloys via LPBF, streamlining the trial-and-error process and reducing associated costs.

Figure 3. The simulated temperature distribution and single-layer multi-track isothermograms of LPBF Hastelloy X, located at the bottom of the powder bed, are presented for various laser energy densities. (a) depicts the single-point temperature distribution at the bottom of the powder bed, followed by the isothermograms corresponding to laser energy densities of (b) 31 J/mm3 , (c) 43 J/mm3 , (d) 53 J/mm3 , (e) 67 J/mm3 , and (f) 91 J/mm3 .
Figure 3. The simulated temperature distribution and single-layer multi-track isothermograms of LPBF Hastelloy X, located at the bottom of the powder bed, are presented for various laser energy densities. (a) depicts the single-point temperature distribution at the bottom of the powder bed, followed by the isothermograms corresponding to laser energy densities of (b) 31 J/mm3 , (c) 43 J/mm3 , (d) 53 J/mm3 , (e) 67 J/mm3 , and (f) 91 J/mm3 .

References
[1] DebRoy T, Wei HL, Zuback JS, et al. Additive manufacturing of metallic components – process, structure and properties. Prog Mater Sci. 2018;92:112–224. doi:10.
1016/j.pmatsci.2017.10.001
[2] Mostafaei A, Ghiaasiaan R, Ho IT, et al. Additive manufacturing of nickel-based superalloys: A state-of-the-art
review on process-structure-defect-property relationship.
Prog Mater Sci. 2023;136:101108. doi:10.1016/j.pmatsci.
2023.101108
[3] Akande IG, Oluwole OO, Fayomi OSI, et al. Overview of
mechanical, microstructural, oxidation properties and
high-temperature applications of superalloys. Mater
Today Proc. 2021;43:2222–2231. doi:10.1016/j.matpr.
2020.12.523
[4] Sanchez S, Smith P, Xu Z, et al. Powder bed fusion of
nickel-based superalloys: a review. Int J Machine Tools
Manuf. 2021;165:103729. doi:10.1016/j.ijmachtools.2021.
103729
[5] Xie Y, Teng Q, Shen M, et al. The role of overlap region
width in multi-laser powder bed fusion of Hastelloy X
superalloy. Virtual Phys Prototyp. 2023;18(1):e2142802.
doi:10.1080/17452759.2022.2142802
[6] Yuan W, Chen H, Cheng T, et al. Effects of laser scanning
speeds on different states of the molten pool during
selective laser melting: simulation and experiment.
Mater Des. 2020;189:108542. doi:10.1016/j.matdes.2020.
108542
[7] He X, Kong D, Zhou Y, et al. Powder recycling effects on
porosity development and mechanical properties of
Hastelloy X alloy during laser powder bed fusion
process. Addit Manuf. 2022;55:102840. doi:10.1016/j.
addma.2022.102840
[8] Sanaei N, Fatemi A. Defects in additive manufactured
metals and their effect on fatigue performance: a stateof-the-art review. Prog Mater Sci. 2021;117:100724.
doi:10.1016/j.pmatsci.2020.100724
[9] Pourbabak S, Montero-Sistiaga ML, Schryvers D, et al.
Microscopic investigation of as built and hot isostatic
pressed Hastelloy X processed by selective laser
melting. Mater Charact. 2019;153:366–371. doi:10.1016/j.
matchar.2019.05.024
[10] He X, Wang L, Kong D, et al. Recrystallization effect on
surface passivation of Hastelloy X alloy fabricated by
laser powder bed fusion. J Mater Sci Technol.
2023;163:245–258. doi:https://doi.org/10.1016j.jmst.
2023.06.003.
[11] Sabzi HE, Maeng S, Liang X, et al. Controlling crack formation and porosity in laser powder bed fusion: alloy
design and process optimisation. Addit Manuf.
2020;34:101360. doi:10.1016/j.addma.2020.101360
[12] Yu C, Chen N, Li R, et al. Selective laser melting of GH3536
superalloy: microstructure, mechanical properties, and
hydrocyclone manufacturing. Adv Powder Mater. 2023:

doi:10.1016/j.apmate.2023.100134
[13] Ye C, Zhang C, Zhao J, et al. Effects of post-processing on
the surface finish, porosity, residual stresses, and fatigue
performance of additive manufactured metals: a review.
J Mater Eng Perform. 2021;30(9):6407–6425. doi:10.
1007/s11665-021-06021-7
[14] Zhang W, Zheng Y, Liu F, et al. Effect of solution temperature on the microstructure and mechanical properties of
Hastelloy X superalloy fabricated by laser directed energy
deposition. Mater Sci Eng A. 2021;820:141537. doi:10.
1016/j.msea.2021.141537
[15] Lehmann T, Rose D, Ranjbar E, et al. Large-scale metal
additive manufacturing: a holistic review of the state of the art and challenges. Int Mater Rev. 2021;67(4):410–459. doi:10.1080/09506608.2021.1971427

[16] Wu S, Hu Y, Yang B, et al. Review on defect characterization and structural integrity assessment method of additively manufactured materials. J Mech Eng. 2021;57 (22):3–34. doi:10.3901/JME.2021.22.003

[17] Keller C, Mokhtari M, Vieille B, et al. Influence of a rescanning strategy with different laser powers on the microstructure and mechanical properties of Hastelloy X elaborated by powder bed fusion. Mater Sci Eng A. 2021;803:140474. doi:10.1016/j.msea.2020.140474

[18] Keshavarzkermani A, Marzbanrad E, Esmaeilizadeh R,et al. An investigation into the effect of process parameters on melt pool geometry, cell spacing, and grain refinement during laser powder bed fusion. Optics & Laser Technol. 2019;116:83–91. doi:10.1016/j.optlastec. 2019.03.012

[19] Watring DS, Benzing JT, Hrabe N, et al. Effects of laserenergy density and build orientation on the structureproperty relationships in as-built Inconel 718 manufactured by laser powder bed fusion. Addit Manuf. 2020;36:101425. doi:10.1016/j.addma.2020.101425

[20] Xiao H, Liu X, Xiao W, et al. Influence of molten-pool cooling rate on solidification structure and mechanical property of laser additive manufactured Inconel 718. J Mater Res Technol. 2022;19:4404–4416. doi:10.1016/j. jmrt.2022.06.162

[21] Wang J, Zhu R, Liu Y, et al. Understanding melt pool characteristics in laser powder bed fusion: An overview of single- and multi-track melt pools for process optimization. Adv Powder Mater. 2023;2(4):100137. doi:10.1016/j. apmate.2023.100137

[22] Li Z, Deng Y, Yao B, et al. Effect of laser scan speed on pool size and densification of selective laser melted CoCr alloy under constant laser energy density. Laser Optoelectronics Progress. 2022;59(7):0736001. doi:10. 3788/LOP202259.0736001

[23] Zhang J, Yuan W, Song B, et al. Towards understanding metallurgical defect formation of selective laser melted wrought aluminum alloys. Adv Powder Mater. 2022;1 (4):100035. doi:10.1016/j.apmate.2022.100035

[24] Rui H, Meiping W, Chen C, et al. Effects of laser energy density on microstructure and corrosion resistance of FeCrNiMnAl high entropy alloy coating. Optics & Laser Technol. 2022;152:108188. doi:https://doi.org/10.1016j. optlastec.2022.108188.

[25] Zhao Y, Sun W, Wang Q, et al. Effect of beam energy density characteristics on microstructure and mechanical properties of nickel-based alloys manufactured by laser directed energy deposition. J Mater Process Technol. 2023;319:118074. doi:10.1016/j.jmatprotec.2023.118074

[26] Tan P, Zhou M, Tang C, et al. Multiphysics modelling of powder bed fusion for polymers. Virtual Phys Prototyp. 2023;18(1):e2257191. doi:10.1080/17452759.2023. 2257191

[27] Tan P, Shen F, Shian Tey W, et al. A numerical study on the packing quality of fibre/polymer composite powder for powder bed fusion additive manufacturing. Virtual Phys Prototyp. 2021;16(sup1):S1–S18. doi:10.1080/17452759. 2021.1922965

[28] Kusano M, Watanabe M. Microstructure control of Hastelloy X by geometry-induced elevation of sample temperature during a laser powder bed fusion process. Mater Des. 2022;222:111016. doi:10.1016/j.matdes.2022. 111016

[29] Lee YS, Zhang W. Modeling of heat transfer, fluid flow and solidification microstructure of nickel-base superalloy fabricated by laser powder bed fusion. Addit Manuf. 2016;12:178–188. doi:10.1016/j.addma.2016.05.003

[30] Lv F, Liang HX, Xie DQ, et al. On the role of laser in situ remelting into pore elimination of Ti-6Al-4V components fabricated by selective laser melting. J Alloys Compd. 2021;854:156866. doi:10.1016/j.jallcom.2020.156866

[31] Prithivirajan V, Sangid MD. The role of defects and critical pore size analysis in the fatigue response of additively manufactured IN718 via crystal plasticity. Mater Des. 2018;150:139–153. doi:10.1016/j.matdes.2018.04.022

[32] Huang Y. A user-material subroutine incroporating single crystal plasticity in the ABAQUS finite element program. Cambridge: Harvard University Press; 1991.

[33] Pilgar CM, Fernandez AM, Lucarini S, et al. Effect of printing direction and thickness on the mechanical behavior of SLM fabricated Hastelloy-X. Int J Plasticity. 2022;153:103250. doi:10.1016/j.ijplas.2022.103250

[34] Garlea E, Choo H, Sluss CC, et al. Variation of elastic mechanical properties with texture, porosity, and defect characteristics in laser powder bed fusion 316L stainless steel. Mater Sci Eng A. 2019;763:138032. doi:10.1016/j. msea.2019.138032

[35] Sanchez-Mata O, Wang X, Muñiz-Lerma JA, et al. Dependence of mechanical properties on crystallographic orientation in nickel-based superalloy Hastelloy X fabricated by laser powder bed fusion. J Alloys Compd. 2021;865:158868. doi:10.1016/j.jallcom.2021. 158868

[36] Gu H, Wei C, Li L, et al. Multi-physics modelling of molten

pool development and track formation in multi-track, multi-layer and multi-material selective laser melting. Int J Heat Mass Transf. 2020;151:119458. doi:10.1016/j. ijheatmasstransfer.2020.119458

[37] Johnson L, Mahmoudi M, Zhang B, et al. Assessing printability maps in additive manufacturing of metal alloys. Acta Mater. 2019;176:199–210. doi:10.1016/j.actamat. 2019.07.005

[38] Wang S, Ning J, Zhu L, et al. Role of porosity defects in metal 3D printing: formation mechanisms, impacts on properties and mitigation strategies. Mater Today. 2022;59:133–160. doi:10.1016/j.mattod.2022.08.014

[39] Guo Y, Collins DM, Tarleton E, et al. Measurements of stress fields near a grain boundary: exploring blocked arrays of dislocations in 3D. Acta Mater. 2015;96:229–doi:10.1016/j.actamat.2015.05.041
[40] Kong D, Dong C, Ni X, et al. Hetero-deformation-induced
stress in additively manufactured 316L stainless steel.
Mater Res Lett. 2020;8(10):390–397. doi:10.1080/
21663831.2020.1775149

Figure 1. Experimental setup and materials. (a) Schematic of the DED process, where three types of base materials were adopted—B1 (IN718), B2 (IN625), and B3 (SS316L), and two types of powder materials were adopted—P1 (IN718) and P2 (SS316L). (b) In situ high-speed imaging of powder flow and the SEM images of IN718 and SS316L powder particle. (c) Powder size statistics, and (d) element composition of powder IN718 (P1) and SS316L (P2).

Printability disparities in heterogeneous materialcombinations via laser directed energy deposition:a comparative stud

Jinsheng Ning1,6, Lida Zhu1,6,∗, Shuhao Wang2, Zhichao Yang1, Peihua Xu1,Pengsheng Xue3, Hao Lu1, Miao Yu1, Yunhang Zhao1, Jiachen Li4, Susmita Bose5 and Amit Bandyopadhyay5,∗

Abstract

적층 제조는 바이메탈 및 다중 재료 구조의 제작 가능성을 제공합니다. 그러나 재료 호환성과 접착성은 부품의 성형성과 최종 품질에 직접적인 영향을 미칩니다. 적합한 프로세스를 기반으로 다양한 재료 조합의 기본 인쇄 가능성을 이해하는 것이 중요합니다.

여기에서는 두 가지 일반적이고 매력적인 재료 조합(니켈 및 철 기반 합금)의 인쇄 적성 차이가 레이저 지향 에너지 증착(DED)을 통해 거시적 및 미시적 수준에서 평가됩니다.

증착 프로세스는 현장 고속 이미징을 사용하여 캡처되었으며, 용융 풀 특징 및 트랙 형태의 차이점은 특정 프로세스 창 내에서 정량적으로 조사되었습니다. 더욱이, 다양한 재료 쌍으로 처리된 트랙과 블록의 미세 구조 다양성이 비교적 정교해졌고, 유익한 다중 물리 모델링을 통해 이종 재료 쌍 사이에 제시된 기계적 특성(미세 경도)의 불균일성이 합리화되었습니다.

재료 쌍의 서로 다른 열물리적 특성에 의해 유발된 용융 흐름의 차이와 응고 중 결과적인 요소 혼합 및 국부적인 재합금은 재료 조합 간의 인쇄 적성에 나타난 차이점을 지배합니다.

이 작업은 서로 다른 재료의 증착에서 현상학적 차이에 대한 심층적인 이해를 제공하고 바이메탈 부품의 보다 안정적인 DED 성형을 안내하는 것을 목표로 합니다.

Additive manufacturing provides achievability for the fabrication of bimetallic and
multi-material structures; however, the material compatibility and bondability directly affect the
parts’ formability and final quality. It is essential to understand the underlying printability of
different material combinations based on an adapted process. Here, the printability disparities of
two common and attractive material combinations (nickel- and iron-based alloys) are evaluated
at the macro and micro levels via laser directed energy deposition (DED). The deposition
processes were captured using in situ high-speed imaging, and the dissimilarities in melt pool
features and track morphology were quantitatively investigated within specific process
windows. Moreover, the microstructure diversity of the tracks and blocks processed with varied
material pairs was comparatively elaborated and, complemented with the informative
multi-physics modeling, the presented non-uniformity in mechanical properties (microhardness)
among the heterogeneous material pairs was rationalized. The differences in melt flow induced
by the unlike thermophysical properties of the material pairs and the resulting element
intermixing and localized re-alloying during solidification dominate the presented dissimilarity
in printability among the material combinations. This work provides an in-depth understanding
of the phenomenological differences in the deposition of dissimilar materials and aims to guide
more reliable DED forming of bimetallic parts.

Figure 1. Experimental setup and materials. (a) Schematic of the DED process, where three types of base materials were adopted—B1
(IN718), B2 (IN625), and B3 (SS316L), and two types of powder materials were adopted—P1 (IN718) and P2 (SS316L). (b) In situ
high-speed imaging of powder flow and the SEM images of IN718 and SS316L powder particle. (c) Powder size statistics, and (d) element
composition of powder IN718 (P1) and SS316L (P2).
Figure 1. Experimental setup and materials. (a) Schematic of the DED process, where three types of base materials were adopted—B1 (IN718), B2 (IN625), and B3 (SS316L), and two types of powder materials were adopted—P1 (IN718) and P2 (SS316L). (b) In situ high-speed imaging of powder flow and the SEM images of IN718 and SS316L powder particle. (c) Powder size statistics, and (d) element composition of powder IN718 (P1) and SS316L (P2).
Figure 2. Deposition process and the track morphology. (a)–(c) Display the in situ captured tableaux of melt propagation and some physical
features during depositing for P1B1, P1B2, and P1B3, respectively. (d) The profiles of the melt pool at a frame of (t0 + 1) ms, and the flow
streamlines in the molten pool of each case. (e) The outer surface of the formed tracks, in which the colored arrows mark the scanning
direction. (f) Cross-section of the tracks. The parameter set used for in situ imaging was P-1000 W, S-600 mm·min–1, F-18 g·min–1. All the
scale bars are 2 mm.
Figure 2. Deposition process and the track morphology. (a)–(c) Display the in situ captured tableaux of melt propagation and some physical features during depositing for P1B1, P1B2, and P1B3, respectively. (d) The profiles of the melt pool at a frame of (t0 + 1) ms, and the flow streamlines in the molten pool of each case. (e) The outer surface of the formed tracks, in which the colored arrows mark the scanning direction. (f) Cross-section of the tracks. The parameter set used for in situ imaging was P-1000 W, S-600 mm·min–1, F-18 g·min–1. All the scale bars are 2 mm.

References

[1] Tan C L, Weng F, Sui S, Chew Y and Bi G J 2021 Progress and perspectives in laser additive manufacturing of key aeroengine materials Int. J. Mach. Tools Manuf. 170 103804
[2] Bandyopadhyay A, Traxel K D, Lang M, Juhasz M, Eliaz N and Bose S 2022 Alloy design via additive manufacturing: advantages, challenges, applications and perspectives Mater. Today 52 207–24
[3] Sui S, Chew Y, Weng F, Tan C L, Du Z L and Bi G J 2022 Study of the intrinsic mechanisms of nickel additive for grain refinement and strength enhancement of laser aided additively manufactured Ti–6Al–4V Int. J. Extrem. Manuf. 4 035102
[4] Xue P S, Zhu L D, Xu P H, Ren Y, Xin B, Meng G R, Yang Z C and Liu Z 2021 Research on process optimization and microstructure of CrCoNi medium-entropy alloy formed by laser metal deposition Opt. Laser Technol. 142 107167
[5] Bandyopadhyay A, Traxel K D and Bose S 2021 Nature-inspired materials and structures using 3D printing Mater. Sci. Eng. R 145 100609
[6] Zuback J S, Palmer T A and DebRoy T 2019 Additive manufacturing of functionally graded transition joints between ferritic and austenitic alloys J. Alloys Compd. 770 995–1003
[7] Feenstra D R, Banerjee R, Fraser H L, Huang A, Molotnikov A and Birbilis N 2021 Critical review of the state of the art in multi-material fabrication via directed energy deposition Curr. Opin. Solid State Mater. Sci. 25 100924
[8] Wei C, Zhang Z Z, Cheng D X, Sun Z, Zhu M H and Li L 2021 An overview of laser-based multiple metallic material additive manufacturing: from macro- to micro-scales Int. J. Extrem. Manuf. 3 012003
[9] Gu D D, Shi X Y, Poprawe R, Bourell D L, Setchi R and Zhu J H 2021 Material-structure-performance integrated laser-metal additive manufacturing Science 372 eabg1487
[10] Bandyopadhyay A and Heer B 2018 Additive manufacturing of multi-material structures Mater. Sci. Eng. R 129 1–16
[11] Tammas-Williams S and Todd I 2017 Design for additive manufacturing with site-specific properties in metals and alloys Scr. Mater. 135 105–10
[12] Chen W, Gu D D, Yang J K, Yang Q, Chen J and Shen X F 2022 Compressive mechanical properties and shape memory effect of NiTi gradient lattice structures fabricated by laser powder bed fusion Int. J. Extrem. Manuf. 4 045002
[13] Svetlizky D, Das M, Zheng B L, Vyatskikh A L, Bose S, Bandyopadhyay A, Schoenung J M, Lavernia E J and Eliaz N 2021 Directed energy deposition (DED) additive manufacturing: physical characteristics, defects, challenges and applications Mater. Today 49 271–95
[14] Panwisawas C, Tang Y T and Reed R C 2020 Metal 3D printing as a disruptive technology for superalloys Nat. Commun. 11 2327
[15] Wang S H, Ning J S, Zhu L D, Yang Z C, Yan W T, Dun Y C, Xue P S, Xu P H, Bose S and Bandyopadhyay A 2022 Role of porosity defects in metal 3D printing: formation mechanisms, impacts on properties and mitigation strategies Mater. Today 59 133–60
[16] DebRoy T, Mukherjee T, Milewski J O, Elmer J W, Ribic B, Blecher J J and Zhang W 2019 Scientific, technological and economic issues in metal printing and their solutions Nat. Mater. 18 1026–32
[17] Afrouzian A, Groden C J, Field D P, Bose S and Bandyopadhyay A 2022 Additive manufacturing of Ti-Ni bimetallic structures Mater. Des. 215 110461
[18] Bandyopadhyay A, Zhang Y N and Onuike B 2022 Additive manufacturing of bimetallic structures Virtual Phys. Prototyp. 17 256–94
[19] Onuike B, Heer B and Bandyopadhyay A 2018 Additive manufacturing of Inconel 718—copper alloy bimetallic structure using laser engineered net shaping (LENSTM) Addit. Manuf. 21 133–40
[20] Sahasrabudhe H, Harrison R, Carpenter C and Bandyopadhyay A 2015 Stainless steel to titanium bimetallic structure using LENSTM Addit. Manuf. 5 1–8
[21] Li B Y, Han C J, Lim C W J and Zhou K 2022 Interface formation and deformation behaviors of an additively manufactured nickel-aluminum-bronze/15-5 PH multimaterial via laser-powder directed energy deposition Mater. Sci. Eng. A 829 142101
[22] Zhang X C, Pan T, Chen Y T, Li L, Zhang Y L and Liou F 2021 Additive manufacturing of copper-stainless steel hybrid components using laser-aided directed energy deposition J. Mater. Sci. Technol. 80 100–16
[23] Shinjo J and Panwisawas C 2022 Chemical species mixing during direct energy deposition of bimetallic systems using titanium and dissimilar refractory metals for repair and biomedical applications Addit. Manuf. 51 102654
[24] Wang D et al 2022 Recent progress on additive manufacturing of multi-material structures with laser powder bed fusion Virtual Phys. Prototyp. 17 329–65
[25] Lin X, Yue T M, Yang H O and Huang W D 2005 Laser rapid forming of SS316L/Rene88DT graded material Mater. Sci. Eng. A 391 325–36
[26] Melzer D, Dˇzugan J, Koukolíková M, Rzepa S and Vavˇrík J 2021 Structural integrity and mechanical properties of the functionally graded material based on 316L/IN718 processed by DED technology Mater. Sci. Eng. A 811 141038
[27] Melzer D, Dˇzugan J, Koukolíková M, Rzepa S, Dlouh´y J, Brázda M and Bucki T 2022 Fracture characterisation of vertically build functionally graded 316L stainless steel with Inconel 718 deposited by directed energy deposition process Virtual Phys. Prototyp. 17 821–40
[28] Zhang Y N and Bandyopadhyay A 2018 Direct fabrication of compositionally graded Ti-Al2O3 multi-material structures using laser engineered net shaping Addit. Manuf. 21 104–11
[29] Ben-Artzy A, Reichardt A, Borgonia P J, Dillon R P, McEnerney B, Shapiro A A and Hosemann P 2021 Compositionally graded SS316 to C300 maraging steel using additive manufacturing Mater. Des. 201 109500
[30] Tan C L, Liu Y C, Weng F, Ng F L, Su J L, Xu Z K, Ngai X D and Chew Y 2022 Additive manufacturing of voxelized heterostructured materials with hierarchical phases Addit. Manuf. 54 102775
[31] Chen J, Yang Y Q, Song C H, Zhang M K, Wu S B and Wang D 2019 Interfacial microstructure and mechanical properties of 316L/CuSn10 multi-material bimetallic structure fabricated by selective laser melting Mater. Sci. Eng. A 752 75–85
[32] Wei C, Gu H, Gu Y C, Liu L C, Huang Y H, Cheng D X, Li Z Q and Li L 2022 Abnormal interfacial bonding mechanisms of multi-material additive-manufactured tungsten–stainless steel sandwich structure Int. J. Extrem. Manuf. 4 025002
[33] Zhang Y N and Bandyopadhyay A 2021 Influence of compositionally graded interface on microstructure and compressive deformation of 316L stainless steel to Al12Si aluminum alloy bimetallic structures ACS Appl. Mater. Interfaces 13 9174–85
[34] Wei C et al 2022 Cu10Sn to Ti6Al4V bonding mechanisms in laser-based powder bed fusion multiple material additive 15 Int. J. Extrem. Manuf. 6 (2024) 025001 J Ning et al manufacturing with different build strategies Addit. Manuf. 51 102588
[35] Li W, Karnati S, Kriewall C, Liou F, Newkirk J, Brown Taminger K M and Seufzer W J 2017 Fabrication and characterization of a functionally graded material from Ti-6Al-4V to SS316 by laser metal deposition Addit. Manuf. 14 95–104
[36] Shi Q M, Zhong G Y, Sun Y, Politis C and Yang S F 2021 Effects of laser melting+remelting on interfacial macrosegregation and resulting microstructure and microhardness of laser additive manufactured H13/IN625 bimetals J. Manuf. Process. 71 345–55
[37] Zhang W X, Hou W Y, Deike L and Arnold C 2022 Understanding the Rayleigh instability in humping phenomenon during laser powder bed fusion process Int. J. Extrem. Manuf. 4 015201
[38] Chen Y W, Zhang X, Li M M, Xu R Q, Zhao C and Sun T 2020 Laser powder bed fusion of Inconel 718 on 316 stainless steel Addit. Manuf. 36 101500
[39] Yang Z C, Wang S H, Zhu L D, Ning J S, Xin B, Dun Y C and Yan W T 2022 Manipulating molten pool dynamics during metal 3D printing by ultrasound Appl. Phys. Rev. 9 021416
[40] Hofmann D C, Roberts S, Otis R, Kolodziejska J, Dillon R P, Suh J O, Shapiro A A, Liu Z K and Borgonia J P 2014 Developing gradient metal alloys through radial deposition additive manufacturing Sci. Rep. 4 5357
[41] Tumkur T U et al 2021 Nondiffractive beam shaping for enhanced optothermal control in metal additive manufacturing Sci. Adv. 7 eabg9358
[42] Scipioni Bertoli U, Guss G, Wu S, Matthews M J and Schoenung J M 2017 In-situ characterization of laser-powder interaction and cooling rates through high-speed imaging of powder bed fusion additive manufacturing Mater. Des. 135 385–96
[43] Siva Prasad H, Brueckner F and Kaplan A F H 2020 Powder incorporation and spatter formation in high deposition rate blown powder directed energy deposition Addit. Manuf. 35 101413
[44] Ebrahimi A, Kleijn C R and Richardson I M 2021 Numerical study of molten metal melt pool behaviour during conduction-mode laser spot melting J. Appl. Phys. 54 105304
[45] Mumtaz K A and Hopkinson N 2010 Selective laser melting of thin wall parts using pulse shaping J. Mater. Process. Technol. 210 279–87
[46] Sikandar Iquebal A, Yadav A, Botcha B, Krishna Gorthi R and Bukkapatnam S 2022 Tracking and quantifying spatter characteristics in a laser directed energy deposition process using Kalman filter Manuf. Lett. 33 692–700
[47] Criales L E, Arısoy Y M, Lane B, Moylan S, Donmez A and Özel T 2017 Laser powder bed fusion of nickel alloy 625: experimental investigations of effects of process parameters on melt pool size and shape with spatter analysis Int. J. Mach. Tools Manuf. 121 22–36
[48] Coen V, Goossens L and van Hooreweder B 2022 Methodology and experimental validation of analytical melt pool models for laser powder bed fusion J. Mater. Process. Technol. 304 117547
[49] Zhao C, Shi B, Chen S L, Du D, Sun T, Simonds B J, Fezzaa K and Rollett A D 2022 Laser melting modes in metal powder bed fusion additive manufacturing Rev. Mod. Phys. 94 045002
[50] Wang J H, Han F Z, Chen S F and Ying W S 2019 A novel model of laser energy attenuation by powder particles for laser solid forming Int. J. Mach. Tools Manuf. 145 103440
[51] Haley J C, Schoenung J M and Lavernia E J 2018 Observations of particle-melt pool impact events in directed energy deposition Addit. Manuf. 22 368–74
[52] Chen Y H et al 2021 Correlative synchrotron x-ray imaging and diffraction of directed energy deposition additive manufacturing Acta Mater. 209 116777
[53] Khorasani M, Ghasemi A, Leary M, Cordova L, Sharabian E, Farabi E, Gibson I, Brandt M and Rolfe B 2022 A comprehensive study on meltpool depth in laser-based powder bed fusion of Inconel 718 Int. J. Adv. Manuf. Technol. 120 2345–62
[54] Shamsaei N, Yadollahi A, Bian L and Thompson S M 2015 An overview of direct laser deposition for additive manufacturing; part II: mechanical behavior, process parameter optimization and control Addit. Manuf. 8 12–35
[55] Ghanavati R, Naffakh-Moosavy H, Moradi M and Eshraghi M 2022 Printability and microstructure of directed energy deposited SS316l-IN718 multi-material: numerical modeling and experimental analysis Sci. Rep. 12 16600
[56] Galbusera F, Demir A G, Platl J, Turk C, Schnitzer R and Previtali B 2022 Processability and cracking behaviour of novel high-alloyed tool steels processed by laser powder bed fusion J. Mater. Process. Technol. 302 117435
[57] Wang A et al 2023 Effects of processing parameters on pore defects in blue laser directed energy deposition of aluminum by in and ex situ observation J. Mater. Process. Technol. 319 118068
[58] Hinojos A, Mireles J, Reichardt A, Frigola P, Hosemann P, Murr L E and Wicker R B 2016 Joining of Inconel 718 and 316 stainless steel using electron beam melting additive manufacturing technology Mater. Des. 94 17–27
[59] Yang Z C, Zhu L D, Wang S H, Ning J S, Dun Y C, Meng G R, Xue P S, Xu P H and Xin B 2021 Effects of ultrasound on multilayer forming mechanism of Inconel 718 in directed energy deposition Addit. Manuf. 48 102462
[60] Yao L M, Huang S, Ramamurty U and Xiao Z M 2021 On the formation of “Fish-scale” morphology with curved grain interfacial microstructures during selective laser melting of dissimilar alloys Acta Mater. 220 117331
[61] Ghanavati R, Naffakh-Moosavy H and Moradi M 2021 Additive manufacturing of thin-walled SS316L-IN718 functionally graded materials by direct laser metal deposition J. Mater. Res. Technol. 15 2673–85
[62] Chen N N, Khan H A, Wan Z X, Lippert J, Sun H, Shang S L, Liu Z K and Li J J 2020 Microstructural characteristics and crack formation in additively manufactured bimetal material of 316L stainless steel and Inconel 625 Addit. Manuf. 32 101037
[63] Xiao Y H, Wan Z X, Liu P W, Wang Z, Li J J and Chen L 2022 Quantitative simulations of grain nucleation and growth at additively manufactured bimetallic interfaces of SS316L and IN625 J. Mater. Process. Technol. 302 117506
[64] Mukherjee T, DebRoy T, Lienert T J, Maloy S A and Hosemann P 2021 Spatial and temporal variation of hardness of a printed steel part Acta Mater. 209 116775
[65] Dinda G P, Dasgupta A K and Mazumder J 2021 Texture control during laser deposition of nickel-based superalloy Scr. Mater. 67 503–6
[66] Tan Z E, Pang J H L, Kaminski J and Pepin H 2019 Characterisation of porosity, density, and microstructure of directed energy deposited stainless steel AISI 316L Addit. Manuf. 25 286–96
[67] Wolff S J, Gan Z T, Lin S, Bennett J L, Yan W T, Hyatt G, Ehmann K F, Wagner G J, Liu W K and Cao J 2019 Experimentally validated predictions of thermal history and microhardness in laser-deposited Inconel 718 on carbon steel Addit. Manuf. 27 540–51 16 Int. J. Extrem. Manuf. 6 (2024) 025001 J Ning et al
[68] Zhang L, Wen M, Imade M, Fukuyama S and Yokogawa K 2008 Effect of nickel equivalent on hydrogen gas embrittlement of austenitic stainless steels based on type 316 at low temperatures Acta Mater. 56 3414–21
[69] Zuback J S and DebRoy T 2018 The hardness of additively manufactured alloys Materials 11 2070
[70] Adomako N K, Lewandowski J J, Arkhurst B M, Choi H, Chang H J and Kim J H 2022 Microstructures and mechanical properties of multi-layered materials composed of Ti-6Al-4V, vanadium, and 17–4PH stainless steel produced by directed energy deposition Addit. Manuf. 59 103174

Fig. 3. (a–c) Snapshots of the CtFD simulation of laser-beam irradiation: (a) Top, (b) longitudinal vertical cross-sectional, and (c) transversal vertical cross-sectional views. (d) z-position of the solid/liquid interface during melting and solidification.

Solute segregation in a rapidly solidified Hastelloy-X Ni-based superalloy during laser powder bed fusion investigated by phase-field simulations and computational thermal-fluid dynamics

Masayuki Okugawa ab, Kenji Saito a, Haruki Yoshima a, Katsuhiko Sawaizumi a, Sukeharu Nomoto c, Makoto Watanabe c, Takayoshi Nakano ab, Yuichiro Koizumi abShow moreAdd to MendeleyShareCite

https://doi.org/10.1016/j.addma.2024.104079

Get rights and content Under a Creative Commons license open access

Abstract

Solute segregation significantly affects material properties and is a critical issue in the laser powder-bed fusion (LPBF) additive manufacturing (AM) of Ni-based superalloys. To the best of our knowledge, this is the first study to demonstrate a computational thermal-fluid dynamics (CtFD) simulation coupled multi-phase-field (MPF) simulation with a multicomponent-composition model of Ni-based superalloy to predict solute segregation under solidification conditions in LPBF. The MPF simulation of the Hastelloy-X superalloy reproduced the experimentally observed submicron-sized cell structure. Significant solute segregations were formed within interdendritic regions during solidification at high cooling rates of up to 10K s-1, a characteristic feature of LPBF. Solute segregation caused a decrease in the solidus temperature (TS), with a reduction of up to 30.4 K, which increases the risk of liquation cracks during LPBF. In addition, the segregation triggers the formation of carbide phases, which increases the susceptibility to ductility dip cracking. Conversely, we found that the decrease in TS is suppressed at the melt-pool boundary regions, where re-remelting occurs during the stacking of the layer above. Controlling the re-remelting behavior is deemed to be crucial for designing crack-free alloys. Thus, we demonstrated that solute segregation at the various interfacial regions of Ni-based multicomponent alloys can be predicted by the conventional MPF simulation. The design of crack-free Ni-based superalloys can be expedited by MPF simulations of a broad range of element combinations and their concentrations in multicomponent Ni-based superalloys.

Graphical abstract

Keywords

Laser powder-bed fusion, Hastelloy-X Nickel-based superalloy, solute element segregation, computational thermal-fluid dynamics simulation, phase-field method

1. Introduction

Additive manufacturing (AM) technologies have attracted considerable attention as they allow us to easily build three-dimensional (3D) parts with complex geometries. Among the wide range of available AM techniques, laser powder-bed fusion (LPBF) has emerged as a preferred technique for metal AM [1][2][3][4][5]. In LPBF, metal products are built layer-by-layer by scanning laser, which fuse metal powder particles into bulk solids.

Significant attempts have been made to integrate LPBF techniques within the aerospace industry, with a particular focus on weldable Ni-based superalloys, such as IN718 [6][7][8], IN625 [9][10], and Hastelloy-X (HX) [11][12][13][14]. Non-weldable alloys, such as IN738LC [15][16] and CMSX-4 [1][17] are also suitable for their sufficient creep resistance under higher temperature conditions. However, non-weldable alloys are difficult to build using LPBF because of their susceptibility to cracking during the process. In general, a macro solute-segregation during solidification is suppressed by the rapid cooling conditions (up to 108 K s-1) unique to the LPBF process [18]. However, the solute segregation still occurs in the interdendritic regions that are smaller than the micrometer scale [5][19][20][21]; these regions are suggested to be related to the hot cracks in LPBF-fabricated parts. Therefore, an understanding of solute segregation is essential for the fabrication of reliable LPBF-fabricated parts while avoiding cracks.

The multiphase-field (MPF) method has gained popularity for modeling the microstructure evolution and solute segregation under rapid cooling conditions [5][20][21][22][23][24][25][26][27][28]. Moreover, quantifiable predictions have been achieved by combining the MPF method with temperature distribution analysis methods such as the finite-element method (FEM) [20] and computational thermal-fluid dynamics (CtFD) simulations [28]. These aforementioned studies have used binary-approximated multicomponent systems, such as Ni–Nb binary alloys, to simulate IN718 alloys. While MPF simulations using binary alloy systems can effectively reproduce microstructure formations and segregation behaviors, the binary approximation might be affected by the chemical interactions between the removed solute elements in the target multicomponent alloy. The limit of absolute stability predicted by the Mullins-Sekerka theory [29] is also crucial because the limit velocity is close to the solidification rate in the LPBF process and is different in multicomponent and binary-approximated systems. The difference between the solidus and liquidus temperatures, ΔT0, directly determines the absolute stability according to the Mullins-Sekerka theory. For example, the ΔT0 values of IN718 and its binary-approximated Ni–5 wt.%Nb alloy are 134 K [28] and 71 K [30], respectively. The solidification rate compared to the limit of absolute stability, i.e., the relative non-equilibrium of solidification, changes by simplification of the system. It is therefore important to use the composition of the actual multicomponent system in such simulations. However, to the best of our knowledge, there is no MPF simulation using a multicomponent model coupled with a temperature analysis simulation to predict solute segregation in a Ni-based superalloy.

In this study, we demonstrate that the conventional MPF model can reproduce experimentally observed dendritic structures by performing a phase-field simulation using the temperature distribution obtained by a CtFD simulation of a multicomponent Ni-based alloy (conventional solid-solution hardening-type HX). The MPF simulation revealed that the segregation behavior of solute elements largely depends on the regions of the melt pool, such as the cell boundary, the interior of the melt-pool boundary, and heat-affected regions. The sensitivities of the various interfaces to liquation and solidification cracks are compared based on the predicted concentration distributions. Moreover, the feasibility of using the conventional MPF model for LPBF is discussed in terms of the absolute stability limit.

2. Methods

2.1. Laser-beam irradiation experiments

Rolled and recrystallized HX ingots with dimensions of 20 × 50 × 10 mm were used as the specimens for laser-irradiation experiments. The specimens were irradiated with a laser beam scanned along straight lines of 10 mm in length using a laser AM machine (EOS 290 M, EOS) equipped with a 400 W Yb-fiber laser. Irradiation was performed with a beam power of P = 300 W and a scanning speed of V = 600 mm s-1, which are the conditions generally used in the LPBF fabrication of Ni-based superalloy [8]. The corresponding line energy was 0.5 J mm-1. The samples were cut perpendicular to the beam-scanning direction for cross-sectional observation using a field-emission scanning electron microscope (FE-SEM, JEOL JSM 6500). Crystal orientation analysis was performed by electron backscatter diffraction (EBSD). The sizes of each crystal grain and their aspect ratios were evaluated by analyzing the EBSD data.

2.2. CtFD simulation

CtFD simulations of the laser-beam irradiation of HX were performed using a 3D thermo-fluid analysis software (Flow Science FLOW-3D® with Flow-3D Weld module). A Gaussian heat source model was used, in which the irradiation intensity distribution of the beam is regarded as a symmetrical Gaussian distribution over the entire beam. The distribution of the beam irradiation intensity is expressed by the following equation.(1)q̇=2ηPπR2exp−2r2R2.

Here, P is the power, R is the effective beam radius, r is the actual beam radius, and η is the beam absorption rate of the substrate. To improve the accuracy of the model, η was calculated by assuming multiple reflections using the Fresnel equation:(2)�=1−121+1−�cos�21+1+�cos�2+�2−2�cos�+2cos2��2+2�cos�+2cos2�.

ε is the Fresnel coefficient and θ is the incident angle of the laser. A local laser melt causes the vaporization of the material and results in a high vapor pressure. This vapor pressure acts as a recoil pressure on the surface, pushing the weld pool down. The recoil pressure is reproduced using the following equation.(3)precoil=Ap0exp∆HLVRTV1−TVT.

Here, p0 is the atmospheric pressure, ∆HLV is the latent heat of vaporization, R is the gas constant, and TV is the boiling point at the saturated vapor pressure. A is a ratio coefficient that is generally assumed to be 0.54, indicating that the recoil pressure due to evaporation is 54% of the vapor pressure at equilibrium on the liquid surface.

Table 1 shows the parameters used in the simulations. Most parameters were evaluated using an alloy physical property calculation software (Sente software JMatPro v11). The values in a previously published study [31] were used for the emissivity and the Stefan–Boltzmann constant, and the values for pure Ni [32] were used for the heat of vaporization and vaporization temperatures. The Fresnel coefficient, which determines the beam absorption efficiency, was used as a fitting parameter to reproduce the morphology of the experimentally observed melt region, and a Fresnel coefficient of 0.12 was used in this study.

Table 1. Parameters used in the CtFD simulations.

ParameterSymbolValueReference
Density at 298.15 Kρ8.24 g cm-3[]
Liquidus temperatureTL1628.15 K[]
Solidus temperatureTS1533.15 K[]
Viscosity at TLη6.8 g m-1 s-1[]
Specific heat at 298.15 KCP0.439 J g-1 K-1[]
Thermal conductivity at 298.15 Kλ10.3 W m-1 K-1[]
Surface tension at TLγL1.85 J m-2[]
Temperature coefficient of surface tensiondγL/dT–2.5 × 10−4 J m-2 K-1[]
EmissivityΕ0.27[31]
Stefan–Boltzmann constantσ5.67 × 10-8 W m-2 K-4[31]
Heat of fusionΔHSL2.76 × 102 J g-1[32]
Heat of vaporizationΔHLV4.29 × 10J g-1[32]
Vaporization temperatureTV3110 K[32]

Calculated using JMatPro v11.

The dimensions of the computational domain of the numerical model were 4.0 mm in the beam-scanning direction, 0.4 mm in width, and 0.3 mm in height. A uniform mesh size of 10 μm was applied throughout the computational domain. The boundary condition of continuity was applied to all boundaries except for the top surface. The temperature was initially set to 300 K. P and V were set to their experimental values, i.e., 300 W and 600 mm s-1, respectively. Solidification conditions based on the temperature gradient, G, the solidification rate, R, and the cooling rate were evaluated, and the obtained temperature distribution was used in the MPF simulations.

2.3. MPF simulation

Two-dimensional MPF simulations weakly coupled with the CtFD simulation were performed using the Microstructure Evolution Simulation Software (MICRESS) [33][34][35][36][37] with the TQ-Interface for Thermo-Calc [38]. A simplified HX alloy composition of Ni-21.4Cr-17.6Fe-0.46Mn-8.80Mo-0.39Si-0.50W-1.10Co-0.08 C (mass %) was used in this study. The Gibbs free energy and diffusion coefficient of the system were calculated using the TCNI9 thermodynamic database [39] and the MOBNi5 mobility database [40]. Τhe equilibrium phase diagram calculated using Thermo-Calc indicates that the face-centered cubic (FCC) and σ phases appear as the equilibrium solid phases [19]. However, according to the time-temperature-transformation (TTT) diagram [41], the phases are formed after the sample is maintained for tens of hours in a temperature range of 1073 to 1173 K. Therefore, only the liquid and FCC phases were assumed to appear in the MPF simulations. The simulation domain was 5 × 100 μm, and the grid size Δx and interface width were set to 0.025 and 0.1 µm, respectively. The interfacial mobility between the solid and liquid phases was set to 1.0 × 10-8 m4 J-1 s-1. Initially, one crystalline nucleus with a [100] crystal orientation was placed at the left bottom of the simulation domain, with the liquid phase occupying the remainder of the domain. The model was solidified under the temperature field distribution obtained by the CtFD simulation. The concentration distribution and crystal orientation of the solidified model were examined. The primary dendrite arm space (PDAS) was compared to the experimental PDAS measured by the cross-sectional SEM observation.

In an actual LPBF process, solidified layers are remelted and resolidified during the stacking of the one layer above, thereby greatly affecting solute element distributions in those regions. Therefore, remelting and resolidification simulations were performed to examine the effect of remelting on solute segregation. The solidified model was remelted and resolidified by applying a time-dependent temperature field shifted by 60 μm in the height direction, assuming reheating during the stacking of the upper layer (i.e., the upper 40 μm region of the simulation box was remelted and resolidified). The changes in the composition distribution and formed microstructure were investigated.

3. Results

3.1. Experimental observation of melt pool

Fig. 1 shows a cross-sectional optical microscopy image and corresponding inverse pole figure (IPF) orientation maps obtained from the laser-melted region of HX. The dashed line indicates the fusion line. A deep melted region was formed by keyhole-mode melting due to the vaporization of the metal and resultant recoil pressure. Epitaxial growth from the unmelted region was observed. Columnar crystal grains with an average diameter of 5.46 ± 0.32 μm and an aspect ratio of 3.61 ± 0.13 appeared at the melt regions (Figs. 1b–1d). In addition, crystal grains growing in the z direction could be observed in the lower center.

Fig. 1

Fig. 2a shows a cross-sectional backscattering electron image (BEI) obtained from the laser-melted region indicated by the black square in Fig. 1a. The bright particles with a diameter of approximately 2 μm observed outside the melt pool. It is well known that M6C, M23C6, σ, and μ precipitate phases are formed in Hastelloy-X [41]. These precipitates mainly consisted of Mo, Cr, Fe, and Ni; The μ and M6C phases are rich in Mo, while the σ and M23C6 phases are rich in Cr. The SEM energy dispersive X-ray spectroscopy analysis suggested that the bright particles are the stable precipitates as shown in Fig. S2 and Table S1. Conversely, there are no carbides in the melt pool. This suggests that the cooling rate is extremely high during LPBF, which prevents the formation of a stable carbide during solidification. Figs. 2b–2f show magnified BEI images at different height positions indicated in Fig. 2a. Bright regions are observed between the cells, which become fragmentary at the center of the melt pool, as indicated by the yellow arrow heads in Figs. 2e and 2f.

Fig. 2

3.2. CtFD simulation

Figs. 3a–3c show snapshots of the CtFD simulation of HX at 2.72 ms, with the temperature indicated in color. A melt pool with an elongated teardrop shape formed and keyhole-mode melting was observed at the front of the melt region. The cooling rate, temperature gradient (G), and solidification rate (R) were evaluated from the temporal change in the temperature distribution of the CtFD simulation results. The z-position of the solid/liquid interface during the melting and solidification processes is shown in Fig. 3d. The interface goes down rapidly during melting and then rises during solidification. The MPF simulation of the microstructure formation during solidification was performed using the temperature distribution. Moreover, the microstructure formation process during the fabrication of the upper layer was investigated by remelting and resolidifying the solidified layer using the same temperature distribution with a 60 μm upward shift, corresponding to the layer thickness commonly used in the LPBF of Ni-based superalloys.

Fig. 3

Figs. 4a–4c show the changes in the cooling rate, temperature gradient, and solidification rate in the center line of the melt pool parallel to the z direction. To output the solidification conditions at the solid/liquid interface in the melt pool, only the data of the mesh where the solid phase ratio was close to 0.5 were plotted. Solidification occurred where the cooling rate was in the range of 2.1 × 105–1.6 × 10K s-1G was in the range of 3.6 × 105–1.9 × 10K m-1, and R was in the range of 8.2 × 10−2–6.3 × 10−1 m s-1. The cooling rate was the highest near the fusion line and decreased as the interface approached the center of the melt region (Fig. 4a). G also exhibited the highest value in the regions near the fusion line and decreased throughout the solid/liquid interface toward the center of the melt pool (Fig. 4b). R had the lowest value near the fusion line and increased as the interface approached the center of the melt region (Fig. 4c).

Fig. 4

3.3. MPF simulations coupled with CtFD simulation

MPF simulations of solidification, remelting, and resolidification were performed using the temperature-time distribution obtained by the CtFD simulation. Fig. 5 shows the MPF solidified models colored by phase and Mo concentration. All the computational domains show the FCC phase after the solidification (Fig. 5a). Dendrites grew parallel to the heat flow direction, and solute segregations were observed in the interdendritic regions. At the bottom of the melt pool (Fig. 5d), planar interface growth occurred before the formation of primary dendrites. The bottom of the melt pool is the turning point of the solid/liquid interface from the downward motion in melting to the upward motion in solidification. Thus, the solidification rate at the boundary is zero, and is extremely low immediately above the molt-pool boundary. Here, the lower limit of the solidification rate (R) for dendritic growth can be represented by the constitutional supercooling criterion [29]Vcs = (G × DL) / ΔT, and planar interface growth occurs at R < VcsDL and ΔT denote the diffusion coefficient in the liquid and the equilibrium freezing range, respectively. The results suggest that planar interface growth occurs at the bottom of the melt pool, resulting in a dark region with a different solute element distribution. Some of the primary dendrites were diminished by competition with other dendrites. In addition, secondary dendrite arms could be seen in the upper regions (Fig. 5c), where solidification occurred at a lower cooling rate. The fragmentation of the solute segregation near the secondary dendrite arms is similar to that observed in the experimental melt pool shown in Figs. 2e and 2f, and the secondary dendrite arms are suggested to have appeared at the center of the melt region. Fig. 6 shows the PDASs measured from the MPF simulation models, compared to the experimental PDASs measured by the cross-sectional SEM observation of the laser-melted regions (Fig. 2). The PDAS obtained by the MPF simulation become larger as the solidification progress. Ghosh et al. [21] evident by the phase-field method that the PDAS decreases as the cooling rate increases under the rapid cooling conditions obtained by the finite element analysis. In this study, the cooling rate was decreased as the interface approached the center of the melt region (Fig. 4a), and the trends in PDAS changes with respect to cooling rate is same as the reported trend [21]. The simulated trends of the PDAS with the position in the melt pool agreed well with the experimental trends. However, all PDASs in the simulation were larger than those observed in the experiment at the same positions. Ode et al. [42] reported that PDAS differences between 2D and 3D MPF simulations can be represented by PDAS2D = 1.12 × PDAS3D owing to differences in the effects of the interfacial energy and diffusivity. We also performed 2D and 3D MPF simulations under the solidification conditions of G = 1.94 × 10K m-1 and R = 0.82 m s-1 (Fig. S1), and found that the PDAS from the 2D MPF simulation was 1.26 times larger than that from the 3D simulation. Therefore, the cell structure obtained by the CtFD simulation coupled with the 2D MPF simulation agreed well with the experimental results over the entire melt pool region considering the dimensional effects.

Fig. 5
Fig. 6

Fig. 7b1 and 7c1 show the concentration profiles of the solidified model along the growth direction indicated by dashed lines in Fig. 7a. The differences in concentrations from the alloy composition are also shown in Fig. 7b2 and 7c2. Cr, Mo, C, Mn, and W were segregated to the interdendritic regions, while Si, Fe, and Co were depressed. The solute segregation behavior agrees with the experimentally observation [43] and the prediction by the Scheil-Gulliver simulation [19]. Segregation occurred to the highest degree in Mo, while the ratio of segregation to the alloy composition was remarkable in C. The concentration fluctuations correlated with the position in the melt pool and decreased at the center of the melt pool, which was suggested to correspond to the lower cooling rate in this region. Conversely, droplets that appeared between secondary dendrite arms in the upper regions of the simulation domain exhibited a locally high segregation of solute elements, with the same amount of segregation as that at the bottom of the melt pool.

Fig. 7

3.4. Remelting and resolidification simulation

The solidified model was subjected to remelting and resolidification conditions by shifting the temperature profile upward by 60 µm to reveal the effect of reheating on the solute segregation behavior. Figs. 8a and 8b shows the simulation domains of the HX model after resolidification, colored by phase and Mo concentration. The magnified MPF models during the resolidification of the regions indicated by rectangles in Figs. 8a and 8b are also shown as Figs. 8c and 8d. Dendrites grew from the bottom of the remelted region, with the segregation of solute elements occurring in the interdendritic regions. The entire domain become the FCC phase after the resolidification, as shown in Fig. 8a. The bottom of the remelted regions exhibited a different microstructure, and Mo was depressed at the remelted regions, rather than the interdendritic regions. The different solute segregation behavior [44] and the microstructure formation [45] at the melt pool boundary is also observed in LPBF manufactured 316 L stainless steel. We found that this microstructure was formed by further remelting during the resolidification process, which is shown in Fig. 9. Here, the solidified HX model was heated, and the interdendritic regions were preferentially melted while concentration fluctuations were maintained (Fig. 9a1 and 9a2). Subsequently, planer interface growth occurs near the melt pool boundary where the solidification rate is almost zero, and the dendrites outside of the boundary are grown epitaxially (Fig. 9b1 and 9b2). However, these remelted again because of the temperature rise (Fig. 9c1 and 9c2, and the temperature-time profile shown in Fig. 9e). The remelted regions then cooled and solidified with the abnormal solute segregations (Fig. 9d1 and 9d2). Then, dendrite grows from amplified fluctuations under the solidification rate larger than the criterion of constitutional supercooling (Fig. 9d1, 9d2, and Fig. 8d). It has been reported [46][47] that temperature rising owning to latent heat affects microstructure formation: phase-field simulations of a Ni–Al binary alloy suggest that the release of latent heat during solidification increases the average temperature of the system [46] and strongly influences the solidification conditions [47]. In this study, the release of latent heat during solidification is considered in CtFD simulations for calculating the temperature distribution, and the temperature increase is suggested to have also occurred due to the release of latent heat.

Fig. 8
Fig. 9

Fig. 10b1 and 10c1 show the solute element concentration line profiles of the resolidified model along the growth direction indicated by dashed lines in Fig. 10a. Fig. 10b2 and 10c2 show the corresponding differences in concentration from the alloy composition. The segregation behavior of solute elements at the interdendritic regions (Fig. 10b1 and 10b2) was the same as that in the solidified model (Figs. 7b1 and 7b2). Here, Cr, Mo, C, Mn, and W were segregated to the interdendritic regions, while Si, Fe, and Co were depressed. However, the concentration fluctuations at the interdendritic regions were larger than those in the solidified model. Moreover, the segregation of the outside of the melt pool, i.e., the heat-affected zone, was remarkable throughout remelting and resolidification. Different segregation behaviors were observed in the re-remelted region: Mo, Si, Mn, and W were segregated, while Ni, Fe, and Co were depressed. These solute segregations caused by remelting are expected to heavily influence the crack behavior.

Fig. 10

4. Discussion

4.1. Effect of segregation of solute elements on liquation cracking susceptibility

Strong solute segregation was observed between the interdendritic regions of the solidified alloy (Fig. 7). In addition, the solute segregation behavior was significantly affected by remelting and resolidification and varied across the alloy. Solute segregation can be categorized by the regions shown in Fig. 11a1–11a4, namely the cell boundary (Fig. 11a1), interior of the melt-pool boundary (Fig. 11a2), re-remelted regions (Fig. 11a3), and heat-affected regions (Fig. 11a4). The concentration profiles of these regions are shown in Fig. 11b1–11b4. Solute segregation was the highest in the cell boundary region. The solute segregation in the heat-affected region was almost the same as that in the cell boundary region, but seemed to have been attenuated by reheating during remelting and resolidification. The interior of the melt-pool boundary region also had the same tendency for solute segregation. However, the amount of Cr segregation was smaller than that of Mo. A decrease in the Cr concentration was also mitigated, and the concentration remained the same as that in the alloy composition. Fig. 11c1–11c4 show the chemical potentials of the solute elements for the FCC phase at 1073 K calculated using the compositions of those interfacial regions. All the interfacial regions showed non-constant chemical potentials for each element along the perpendicular direction, but the fluctuations of the chemical potentials differed by the type of interfaces. In particular, the fluctuation of the chemical potential of C at the cell boundary region was the largest, suggesting it can be relaxed easily by heat treatment. On the other hand, the fluctuations of the other elements in all the regions were small. The solute segregations are most likely to remain after the heat treatment and are supposed to affect the cracking susceptibilities.

Fig. 11

The solidus temperatures TS, the difference between the liquidus and solidus temperatures (i.e., the brittle temperature range (BTR)), and the fractions of the equilibrium precipitate phases at 1073 K of the interfacial regions were calculated as the liquation, solidification, and ductility dip cracking susceptibilities, respectively. At the cell boundary (Fig. 12a1), interior of the melt-pool boundary (Fig. 12a1), and heat-affected regions (Fig. 12a1), the internal and interfacial regions exhibited higher and lower TS compared to that of the alloy composition, respectively. The lowest Ts was obtained with the composition at the cell boundary region, which is the largest solute-segregated region. It has been suggested that strong segregations of solute elements in LPBF lead to liquation cracks [16]. This study also supports this suggestion, and liquation cracks are more likely to occur at the interfacial regions indicated by predicting the solute segregation behavior using the MPF model. Additionally, the BTRs of the cell boundary, interior of the melt-pool boundary, and heat-affected regions were wider at the interdendritic regions, and solidification cracks were also likely to occur in these regions. Moreover, within the solute segregation regions, the fraction of the precipitate phases in these interfacial regions was larger than that calculated using the alloy composition (Fig. 12c1, 12c2, and 12c4). This indicates that ductility dip cracking is also likely to occur at the cell boundary, interior of the melt-pool boundary, and in heat-affected regions. Contrarily, we found that the re-remelted region exhibited a higher TS and smaller BTR even in the interfacial region (Fig. 12a3 and 12b3), where the solute segregation behavior was different from that of the other regions. In addition, the re-remelting region exhibited less precipitation compared with the other segregated regions (Fig. 12c3). The re-remelting caused by the latent heat can attenuate solute segregation, prevent Ts from decreasing, decrease the BTR, and decrease the amount of precipitate phases. Alloys with a large amount of latent heat are expected to increase the re-remelting region, thereby decreasing the susceptibility to liquation and ductility dip cracks due to solute element segregation. This can be a guide for designing alloys for the LPBF process. As mentioned in Section 3.4, the microstructure [45] and the solute segregation behavior [44] at the melt pool boundary of LPBF-manufactured 316 L stainless steel are observed, and they are different from that of the interdendritic regions. Experimental observations of the solute segregation behavior in the LPBF-fabricated Ni-based alloys are currently underway.

Fig. 12

4.2. Applicability of the conventional MPF simulation to microstructure formation under LPBF

As the solidification growth rate increases, segregation coefficients approach 1, and the fluctuation of the solid/liquid interface is suppressed by the interfacial tension. The interface growth occurs in a flat fashion instead of having a cellular morphology at a velocity above the absolute stability limit, Ras, predicted by the Mullins-Sekerka theory [29]Ras = (ΔT0 DL) / (k Γ) where ΔT0DLk, and Γ are the difference between the liquidus and solidus temperatures, equilibrium segregation coefficient, the diffusivity of liquid, and the Gibbs-Thomson coefficient, respectively.

The Ras of HX was calculated using the equation and the thermodynamic parameters obtained by the TCNI9 thermodynamic database [39]. The calculated Ras of HX was 3.9 m s-1 and is ten times larger than that of the Ni–Nb alloy (approximately 0.4 m s-1[20]. The HX alloy was solidified under R values in the range of 8.2 × 10−2–6.3 × 10−1 m s-1. The theoretically calculated criterion is larger than the evaluated R, and is in agreement with the experiment in which dendritic growth is observed in the melt pool (Fig. 5). In contrast, Karayagiz et al. [20] reported that the R of the Ni–Nb binary alloy under LPBF was as high as approximately 2 m s-1, and planar interface growth was observed to be predominant under the high-growth-rate conditions. These experimentally observed microstructures agree well with the prediction by the Mullins-Sekerka theory about the relationship between the morphology and solidification rates.

In this study, the solidification microstructure formed by the laser-beam irradiation of an HX multicomponent Ni-based superalloy was reproduced by a conventional MPF simulation, in which the system was assumed to be in a quasi-equilibrium condition. Boussinot et al. [24] also suggested that the conventional phase-field model can be applied to simulate the microstructure of an IN718 multicomponent Ni-based superalloy in LPBF. In contrast, Kagayaski et al. [20] suggested that the conventional MPF simulation cannot be applied to the solidification of the Ni-Nb binary alloy system and that the finite interface dissipation model proposed by Steinbach et al. [48][49] is necessary to simulate the high solidification rates observed in LPBF. The difference in the applicability of the conventional MPF method to HX and Ni–Nb binary alloys is presumed to arise from the differences in the non-equilibrium degree of these systems under the high solidification rates of LPBF. The results suggest that Ras can be used as a simple index to apply the conventional MPF model for solidification in LPBF. Solidification becomes a non-equilibrium process as the solidification rate approaches the limit of absolute stability, Ras. In this study, the solidification of the HX multicomponent system occurred under a relatively low solidification rate compared to Ras, and the microstructure of the conventional MPF model was successfully reproduced in the physical experiment. However, note that the limit of absolute stability predicted by the Mullins-Sekerka theory was originally proposed for solidification in a binary alloy system, and further investigation is required to consider its applicability to multicomponent alloy systems. Moreover, the fast solidification, such as in the LPBF process, causes segregation coefficient approaching a value of 1 [20][21][25] corresponds to a diffusion length that is on the order of the atomic interface thickness. When the segregation coefficient approaches 1, solute undercooling disappears; hence, there is no driving force to amplify fluctuations regardless of whether interfacial tension is present. This phenomenon should be further investigated in future studies.

5. Conclusions

We simulated solute segregation in a multicomponent HX alloy under the LPBF process by an MPF simulation using the temperature distributions obtained by a CtFD simulation. We set the parameters of the CtFD simulation to match the melt pool shape formed in the laser-irradiation experiment and found that solidification occurred under high cooling rates of up to 1.6 × 10K s-1.

MPF simulations using the temperature distributions from CtFD simulation could reproduce the experimentally observed PDAS and revealed that significant solute segregation occurred at the interdendritic regions. Equilibrium thermodynamic calculations using the alloy compositions of the segregated regions when considering crack sensitivities suggested a decrease in the solidus temperature and an increase in the amount of carbide precipitation, thereby increasing the susceptibility to liquation and ductility dip cracks in these regions. Notably, these changes were suppressed at the melt-pool boundary region, where re-remelting occurred during the stacking of the layer above. This effect can be used to achieve a novel in-process segregation attenuation.

Our study revealed that a conventional MPF simulation weakly coupled with a CtFD simulation can be used to study the solidification of multicomponent alloys in LPBF, contrary to the cases of binary alloys investigated in previous studies. We discussed the applicability of the conventional MPF model to the LPBF process in terms of the limit of absolute stability, Ras, and suggested that alloys with a high limit velocity, i.e., multicomponent alloys, can be simulated using the conventional MPF model even under the high solidification velocity conditions of LPBF.

CRediT authorship contribution statement

Masayuki Okugawa: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Takayoshi Nakano: Writing – review & editing, Validation, Supervision, Funding acquisition. Yuichiro Koizumi: Writing – review & editing, Visualization, Validation, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Sukeharu Nomoto: Writing – review & editing, Validation, Investigation. Makoto Watanabe: Writing – review & editing, Validation, Supervision, Funding acquisition. Katsuhiko Sawaizumi: Validation, Software, Investigation, Formal analysis, Data curation. Kenji Saito: Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Data curation. Haruki Yoshima: Visualization, Validation, Software, Investigation, Formal analysis, Data curation.

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

Acknowledgments

This work was partly supported by the Cabinet Office, Government of Japan, Cross-ministerial Strategic Innovation Promotion Program (SIP), “Materials Integration for Revolutionary Design System of Structural Materials,” (funding agency: The Japan Science and Technology Agency), by JSPS KAKENHI Grant Numbers 21H05018 and 21H05193, and by CREST Nanomechanics: Elucidation of macroscale mechanical properties based on understanding nanoscale dynamics for innovative mechanical materials (Grant Number: JPMJCR2194) from the Japan Science and Technology Agency (JST). The authors would like to thank Mr. H. Kawabata and Mr. K. Kimura for their technical support with the sample preparations and laser beam irradiation experiments.

Appendix A. Supplementary material

Download : Download Word document (654KB)

Supplementary material.

Data availability

Data will be made available on request.

References

Figure 5. Simulation of the molten pool under low-speed scanning (1.06 m/s). (a) Sequential solidification of the molten pool at the end of the melt track for laser powers of 190 and 340 W, respectively. (b) Recoil pressure on the molten pool at the keyhole for laser powers of 190 and 340 W, respectively. (c) The force diagram of the melt at the back of the keyhole at t = 750 μs in case B. (d) Temperature gradient at the solid–liquid interface of the molten pool at the moment the laser is deactivated in case A. (e) Temperature gradient at the solid–liquid interface of the molten pool at the moment the laser is deactivated in case B.

Revealing formation mechanism of end of processdepression in laser powder bed fusion by multiphysics meso-scale simulation

다중물리 메조 규모 시뮬레이션을 통해 레이저 분말층 융합에서 공정 종료의 함몰 형성 메커니즘 공개

Haodong Chen a,b, Xin Lin a,b,c, Yajing Sund, Shuhao Wanga,b, Kunpeng Zhu a,b,c and Binbin Dana,b

To link to this article: https://doi.org/10.1080/17452759.2024.2326599

ABSTRACT

Unintended end-of-process depression (EOPD) commonly occurs in laser powder bed fusion (LPBF), leading to poor surface quality and lower fatigue strength, especially for many implants. In this study, a high-fidelity multi-physics meso-scale simulation model is developed to uncover the forming mechanism of this defect. A defect-process map of the EOPD phenomenon is obtained using this simulation model. It is found that the EOPD formation mechanisms are different under distinct regions of process parameters. At low scanning speeds in keyhole mode, the long-lasting recoil pressure and the large temperature gradient easily induce EOPD. While at high scanning speeds in keyhole mode, the shallow molten pool morphology and the large solidification rate allow the keyhole to evolve into an EOPD quickly. Nevertheless, in the conduction mode, the Marangoni effects along with a faster solidification rate induce EOPD. Finally, a ‘step’ variable power strategy is proposed to optimise the EOPD defects for the case with high volumetric energy density at low scanning speeds. This work provides a profound understanding and valuable insights into the quality control of LPBF fabrication.

의도하지 않은 공정 종료 후 함몰(EOPD)은 LPBF(레이저 분말층 융합)에서 흔히 발생하며, 특히 많은 임플란트의 경우 표면 품질이 떨어지고 피로 강도가 낮아집니다. 본 연구에서는 이 결함의 형성 메커니즘을 밝히기 위해 충실도가 높은 다중 물리학 메조 규모 시뮬레이션 모델을 개발했습니다.

이 시뮬레이션 모델을 사용하여 EOPD 현상의 결함 프로세스 맵을 얻습니다. EOPD 형성 메커니즘은 공정 매개변수의 별개 영역에서 서로 다른 것으로 밝혀졌습니다.

키홀 모드의 낮은 스캔 속도에서는 오래 지속되는 반동 압력과 큰 온도 구배로 인해 EOPD가 쉽게 유발됩니다. 키홀 모드에서 높은 스캐닝 속도를 유지하는 동안 얕은 용융 풀 형태와 큰 응고 속도로 인해 키홀이 EOPD로 빠르게 진화할 수 있습니다.

그럼에도 불구하고 전도 모드에서는 더 빠른 응고 속도와 함께 마랑고니 효과가 EOPD를 유발합니다. 마지막으로, 낮은 스캐닝 속도에서 높은 체적 에너지 밀도를 갖는 경우에 대해 EOPD 결함을 최적화하기 위한 ‘단계’ 가변 전력 전략이 제안되었습니다.

이 작업은 LPBF 제조의 품질 관리에 대한 심오한 이해와 귀중한 통찰력을 제공합니다.

Figure 5. Simulation of the molten pool under low-speed scanning (1.06 m/s). (a) Sequential solidification of the molten pool at the
end of the melt track for laser powers of 190 and 340 W, respectively. (b) Recoil pressure on the molten pool at the keyhole for laser
powers of 190 and 340 W, respectively. (c) The force diagram of the melt at the back of the keyhole at t = 750 μs in case B. (d) Temperature gradient at the solid–liquid interface of the molten pool at the moment the laser is deactivated in case A. (e) Temperature
gradient at the solid–liquid interface of the molten pool at the moment the laser is deactivated in case B.
Figure 5. Simulation of the molten pool under low-speed scanning (1.06 m/s). (a) Sequential solidification of the molten pool at the end of the melt track for laser powers of 190 and 340 W, respectively. (b) Recoil pressure on the molten pool at the keyhole for laser powers of 190 and 340 W, respectively. (c) The force diagram of the melt at the back of the keyhole at t = 750 μs in case B. (d) Temperature gradient at the solid–liquid interface of the molten pool at the moment the laser is deactivated in case A. (e) Temperature gradient at the solid–liquid interface of the molten pool at the moment the laser is deactivated in case B.

References

[1] Zhang C, Li Z, Zhang J, et al. Additive manufacturing of magnesium matrix composites: comprehensive review of recent progress and research perspectives. J Mag
Alloys. 2023. doi:10.1016/j.jma.2023.02.005
[2] Webster S, Lin H, Carter III FM, et al. Physical mechanisms in hybrid additive manufacturing: a process design framework. J Mater Process Technol. 2022;291:117048. doi:10. 1016/j.jmatprotec.2021.117048
[3] Wang S, Ning J, Zhu L, et al. Role of porosity defects in metal 3D printing: formation mechanisms, impacts on properties and mitigation strategies. Mater Today. 2022. doi:10.1016/j.mattod.2022.08.014
[4] Wei C, Li L. Recent progress and scientific challenges in multi-material additive manufacturing via laser-based powder bed fusion. Virtual Phys Prototyp. 2021;16 (3):347–371. doi:10.1080/17452759.2021.1928520
[5] Lin X, Wang Q, Fuh JYH, et al. Motion feature based melt pool monitoring for selective laser melting process. J Mater Process Technol. 2022;303:117523. doi:10.1016/j. jmatprotec.2022.117523
[6] Gockel J, Sheridan L, Koerper B, et al. The influence of additive manufacturing processing parameters on surface roughness and fatigue life. Int J Fatigue. 2019;124:380–388. doi:10.1016/j.ijfatigue.2019.03.025
[7] Nicoletto G. Influence of rough as-built surfaces on smooth and notched fatigue behavior of L-PBF AlSi10Mg. Addit Manuf. 2020;34:101251. doi:10.1016/j. addma.2020.101251
[8] Spece H, Yu T, Law AW, et al. 3D printed porous PEEK created via fused filament fabrication for osteoconductive orthopaedic surfaces. J Mech Behav Biomed Mater. 2020;109:103850. doi:10.1115/1.0004270v
[9] Andrukhov O, Huber R, Shi B, et al. Proliferation, behavior, and differentiation of osteoblasts on surfaces of different microroughness. Dent Mater. 2016;32(11):1374–1384. doi:10.1016/j.dental.2016.08.217
[10] Dai N, Zhang LC, Zhang J, et al. Corrosion behavior of selective laser melted Ti-6Al-4 V alloy in NaCl solution. Corros Sci. 2016;102:484–489. doi:10.1016/j.corsci.2015. 10.041
[11] Li EL, Wang L, Yu AB, et al. A three-phase model for simulation of heat transfer and melt pool behaviour in laser powder bed fusion process. Powder Technol. 2021;381:298–312. doi:10.1016/j.powtec.2020.11.061
[12] Liao B, Xia RF, Li W, et al. 3D-printed ti6al4v scaffolds with graded triply periodic minimal surface structure for bone tissue engineering. J Mater Eng Perform. 2021;30:4993– 5004. doi:10.1007/s11665-021-05580-z
[13] Li E, Zhou Z, Wang L, et al. Melt pool dynamics and pores formation in multi-track studies in laser powder bed fusion process. Powder Technol. 2022;405:117533. doi:10.1016/j.powtec.2022.117533
[14] Guo L, Geng S, Gao X, et al. Numerical simulation of heat transfer and fluid flow during nanosecond pulsed laser processing of Fe78Si9B13 amorphous alloys. Int J Heat Mass Transfer. 2021;170:121003. doi:10.1016/j.ijheatma sstransfer.2021.121003
[15] Guo L, Li Y, Geng S, et al. Numerical and experimental analysis for morphology evolution of 6061 aluminum alloy during nanosecond pulsed laser cleaning. Surf Coat Technol. 2022;432:128056. doi:10.1016/j.surfcoat. 2021.128056
[16] Li S, Liu D, Mi H, et al. Numerical simulation on evolution process of molten pool and solidification characteristics of melt track in selective laser melting of ceramic powder. Ceram Int. 2022;48(13):18302–18315. doi:10. 1016/j.ceramint.2022.03.089
[17] Aboulkhair NT, Maskery I, Tuck C, et al. On the formation of AlSi10Mg single tracks and layers in selective laser melting: microstructure and nano-mechanical properties. J Mater Process Technol. 2016;230:88–98. doi:10.1016/j. jmatprotec.2015.11.016
[18] Thijs L, Kempen K, Kruth JP, et al. Fine-structured aluminium products with controllable texture by selective laser melting of pre-alloyed AlSi10Mg powder. Acta Mater. 2013;61(5):1809–1819. doi:10.1016/j.actamat.2012.11.052
[19] Qiu C, Adkins NJE, Attallah MM. Microstructure and tensile properties of selectively laser-melted and of HIPed laser-melted Ti–6Al–4 V. Mater Sci Eng A. 2013;578:230–239. doi:10.1016/j.msea.2013.04.099
[20] Kazemi Z, Soleimani M, Rokhgireh H, et al. Melting pool simulation of 316L samples manufactured by selective laser melting method, comparison with experimental results. Int J Therm Sci. 2022;176:107538. doi:10.1016/j. ijthermalsci.2022.107538
[21] Cao L. Workpiece-scale numerical simulations of SLM molten pool dynamic behavior of 316L stainless steel. Comput Math Appl. 2021;96:209–228. doi:10.1016/j. camwa.2020.04.020
[22] Liu B, Fang G, Lei L, et al. Predicting the porosity defects in selective laser melting (SLM) by molten pool geometry. Int J Mech Sci. 2022;228:107478. doi:10.1016/j.ijmecsci. 2022.107478
[23] Ur Rehman A, Pitir F, Salamci MU. Full-field mapping and flow quantification of melt pool dynamics in laser powder bed fusion of SS316L. Materials. 2021;14(21):6264. doi:10. 3390/ma14216264
[24] Chia HY, Wang L, Yan W. Influence of oxygen content on melt pool dynamics in metal additive manufacturing: high-fidelity modeling with experimental validation. Acta Mater. 2023;249:118824. doi:10.1016/j.actamat. 2023.118824
[25] Cheng B, Loeber L, Willeck H, et al. Computational investigation of melt pool process dynamics and pore formation in laser powder bed fusion. J Mater Eng Perform. 2019;28:6565–6578. doi:10.1007/s11665-019- 04435-y
[26] Li X, Guo Q, Chen L, et al. Quantitative investigation of gas flow, powder-gas interaction, and powder behavior under different ambient pressure levels in laser powder bed fusion. Int J Mach Tools Manuf. 2021;170:103797. doi:10.1016/j.ijmachtools.2021.103797
[27] Wu Y, Li M, Wang J, et al. Powder-bed-fusion additive manufacturing of molybdenum: process simulation, optimization, and property prediction. Addit Manuf. 2022;58:103069. doi:10.1016/j.addma.2022.103069
[28] Wu S, Yang Y, Huang Y, et al. Study on powder particle behavior in powder spreading with discrete element method and its critical implications for binder jetting additive manufacturing processes. Virtual Phys Prototyp. 2023;18(1):e2158877. doi:10.1080/17452759.2022.2158877
[29] Klassen A, Schakowsky T, Kerner C. Evaporation model for beam based additive manufacturing using free surface lattice Boltzmann methods. J Phys D Appl Phys. 2014;47 (27):275303. doi:10.1088/0022-3727/47/27/275303
[30] Cao L. Mesoscopic-scale numerical simulation including the influence of process parameters on slm single-layer multi-pass formation. Metall Mater Trans A. 2020;51:4130–4145. doi:10.1007/s11661-020-05831-z
[31] Zhuang JR, Lee YT, Hsieh WH, et al. Determination of melt pool dimensions using DOE-FEM and RSM with process window during SLM of Ti6Al4V powder. Opt Laser Technol. 2018;103:59–76. doi:10.1016/j.optlastec.2018. 01.013
[32] Li Y, Gu D. Thermal behavior during selective laser melting of commercially pure titanium powder: numerical simulation and experimental study. Addit Manuf. 2014;1–4:99–109. doi:10.1016/j.addma.2014.09.001
[33] Dai D, Gu D. Thermal behavior and densification mechanism during selective laser melting of copper matrix composites: simulation and experiments. Mater Des. 2014;55 (0):482–491. doi:10.1016/j.matdes.2013.10.006
[34] Wang S, Zhu L, Dun Y, et al. Multi-physics modeling of direct energy deposition process of thin-walled structures: defect analysis. Comput Mech. 2021;67:c1229– c1242. doi:10.1007/s00466-021-01992-9
[35] Wu J, Zheng J, Zhou H, et al. Molten pool behavior and its mechanism during selective laser melting of polyamide 6 powder: single track simulation and experiments. Mater Res Express. 2019;6. doi:10.1088/2053-1591/ab2747
[36] Cho JH, Farson DF, Milewski JO, et al. Weld pool flows during initial stages of keyhole formation in laser welding. J Phys D Appl Phys. 2009;42. doi:10.1088/0022- 3727/42/17/175502
[37] Sinha KN. Identification of a suitable volumetric heat source for modelling of selective laser melting of Ti6Al4V powder using numerical and experimental validation approach. Int J Adv Manuf Technol. 2018;99:2257–2270. doi:10.1007/s00170-018-2631-4
[38] Fu CH, Guo YB. Three-dimensional temperature gradient mechanism in selective laser melting of Ti-6Al-4V. J Manuf Sci Eng. 2014;136(6):061004. doi:10.1115/1.4028539
[39] Ansari P, Rehman AU, Pitir F, et al. Selective laser melting of 316 l austenitic stainless steel: detailed process understanding using multiphysics simulation and experimentation. Metals. 2021;11(7):1076. doi:10.3390/met11071076
[40] Zhao C, Shi B, Chen S, et al. Laser melting modes in metal powder bed fusion additive manufacturing. Rev Mod Phys. 2022;94(4):045002. doi:10.1103/revmodphys.94. 045002
[41] Bertoli US, Wolfer AJ, Matthews MJ, et al. On the limitations of volumetric energy density as a design parameter for selective laser melting. Mater Des. 2017;113:331–340. doi:10.1016/j.matdes.2016.10.037
[42] Dash A, Kamaraj A. Prediction of the shift in melting mode during additive manufacturing of 316L stainless steel. Mater Today Commun. 2023: 107238. doi:10.1016/j. mtcomm.2023.107238
[43] Majeed M, Khan HM, Rasheed I. Finite element analysis of melt pool thermal characteristics with passing laser in SLM process. Optik. 2019;194:163068. doi:10.1016/j.ijleo. 2019.163068

Predicting solid-state phase transformations during metal additive manufacturing: A case study on electron-beam powder bed fusion of Inconel-738

Predicting solid-state phase transformations during metal additive manufacturing: A case study on electron-beam powder bed fusion of Inconel-738

금속 적층 제조 중 고체 상 변형 예측: Inconel-738의 전자빔 분말층 융합에 대한 사례 연구

Nana Kwabena Adomako a, Nima Haghdadi a, James F.L. Dingle bc, Ernst Kozeschnik d, Xiaozhou Liao bc, Simon P. Ringer bc, Sophie Primig a

Abstract

Metal additive manufacturing (AM) has now become the perhaps most desirable technique for producing complex shaped engineering parts. However, to truly take advantage of its capabilities, advanced control of AM microstructures and properties is required, and this is often enabled via modeling. The current work presents a computational modeling approach to studying the solid-state phase transformation kinetics and the microstructural evolution during AM. Our approach combines thermal and thermo-kinetic modelling. A semi-analytical heat transfer model is employed to simulate the thermal history throughout AM builds. Thermal profiles of individual layers are then used as input for the MatCalc thermo-kinetic software. The microstructural evolution (e.g., fractions, morphology, and composition of individual phases) for any region of interest throughout the build is predicted by MatCalc. The simulation is applied to an IN738 part produced by electron beam powder bed fusion to provide insights into how γ′ precipitates evolve during thermal cycling. Our simulations show qualitative agreement with our experimental results in predicting the size distribution of γ′ along the build height, its multimodal size character, as well as the volume fraction of MC carbides. Our findings indicate that our method is suitable for a range of AM processes and alloys, to predict and engineer their microstructures and properties.

Graphical Abstract

ga1

Keywords

Additive manufacturing, Simulation, Thermal cycles, γ′ phase, IN738

1. Introduction

Additive manufacturing (AM) is an advanced manufacturing method that enables engineering parts with intricate shapes to be fabricated with high efficiency and minimal materials waste. AM involves building up 3D components layer-by-layer from feedstocks such as powder [1]. Various alloys, including steel, Ti, Al, and Ni-based superalloys, have been produced using different AM techniques. These techniques include directed energy deposition (DED), electron- and laser powder bed fusion (E-PBF and L-PBF), and have found applications in a variety of industries such as aerospace and power generation [2][3][4]. Despite the growing interest, certain challenges limit broader applications of AM fabricated components in these industries and others. One of such limitations is obtaining a suitable and reproducible microstructure that offers the desired mechanical properties consistently. In fact, the AM as-built microstructure is highly complex and considerably distinctive from its conventionally processed counterparts owing to the complicated thermal cycles arising from the deposition of several layers upon each other [5][6].

Several studies have reported that the solid-state phases and solidification microstructure of AM processed alloys such as CMSX-4, CoCr [7][8], Ti-6Al-4V [9][10][11]IN738 [6]304L stainless steel [12], and IN718 [13][14] exhibit considerable variations along the build direction. For instance, references [9][10] have reported that there is a variation in the distribution of α and β phases along the build direction in Ti-alloys. Similarly, the microstructure of an L-PBF fabricated martensitic steel exhibits variations in the fraction of martensite [15]. Furthermore, some of the present authors and others [6][16][17][18][19][20] have recently reviewed and reported that there is a difference in the morphology and fraction of nanoscale precipitates as a function of build height in Ni-based superalloys. These non-uniformities in the as-built microstructure result in an undesired heterogeneity in mechanical and other important properties such as corrosion and oxidation [19][21][22][23]. To obtain the desired microstructure and properties, additional processing treatments are utilized, but this incurs extra costs and may lead to precipitation of detrimental phases and grain coarsening. Therefore, a through-process understanding of the microstructure evolution under repeated heating and cooling is now needed to further advance 3D printed microstructure and property control.

It is now commonly understood that the microstructure evolution during printing is complex, and most AM studies concentrate on the microstructure and mechanical properties of the final build only. Post-printing studies of microstructure characteristics at room temperature miss crucial information on how they evolve. In-situ measurements and modelling approaches are required to better understand the complex microstructural evolution under repeated heating and cooling. Most in-situ measurements in AM focus on monitoring the microstructural changes, such as phase transformations and melt pool dynamics during fabrication using X-ray scattering and high-speed X-ray imaging [24][25][26][27]. For example, Zhao et al. [25] measured the rate of solidification and described the α/β phase transformation during L-PBF of Ti-6Al-4V in-situ. Also, Wahlmann et al. [21] recently used an L-PBF machine coupled with X-ray scattering to investigate the changes in CMSX-4 phase during successive melting processes. Although these techniques provide significant understanding of the basic principles of AM, they are not widely accessible. This is due to the great cost of the instrument, competitive application process, and complexities in terms of the experimental set-up, data collection, and analysis [26][28].

Computational modeling techniques are promising and more widely accessible tools that enable advanced understanding, prediction, and engineering of microstructures and properties during AM. So far, the majority of computational studies have concentrated on physics based process models for metal AM, with the goal of predicting the temperature profile, heat transfer, powder dynamics, and defect formation (e.g., porosity) [29][30]. In recent times, there have been efforts in modeling of the AM microstructure evolution using approaches such as phase-field [31], Monte Carlo (MC) [32], and cellular automata (CA) [33], coupled with finite element simulations for temperature profiles. However, these techniques are often restricted to simulating the evolution of solidification microstructures (e.g., grain and dendrite structure) and defects (e.g., porosity). For example, Zinovieva et al. [33] predicted the grain structure of L-PBF Ti-6Al-4V using finite difference and cellular automata methods. However, studies on the computational modelling of the solid-state phase transformations, which largely determine the resulting properties, remain limited. This can be attributed to the multi-component and multi-phase nature of most engineering alloys in AM, along with the complex transformation kinetics during thermal cycling. This kind of research involves predictions of the thermal cycle in AM builds, and connecting it to essential thermodynamic and kinetic data as inputs for the model. Based on the information provided, the thermokinetic model predicts the history of solid-state phase microstructure evolution during deposition as output. For example, a multi-phase, multi-component mean-field model has been developed to simulate the intermetallic precipitation kinetics in IN718 [34] and IN625 [35] during AM. Also, Basoalto et al. [36] employed a computational framework to examine the contrasting distributions of process-induced microvoids and precipitates in two Ni-based superalloys, namely IN718 and CM247LC. Furthermore, McNamara et al. [37] established a computational model based on the Johnson-Mehl-Avrami model for non-isothermal conditions to predict solid-state phase transformation kinetics in L-PBF IN718 and DED Ti-6Al-4V. These models successfully predicted the size and volume fraction of individual phases and captured the repeated nucleation and dissolution of precipitates that occur during AM.

In the current study, we propose a modeling approach with appreciably short computational time to investigate the detailed microstructural evolution during metal AM. This may include obtaining more detailed information on the morphologies of phases, such as size distribution, phase fraction, dissolution and nucleation kinetics, as well as chemistry during thermal cycling and final cooling to room temperature. We utilize the combination of the MatCalc thermo-kinetic simulator and a semi-analytical heat conduction model. MatCalc is a software suite for simulation of phase transformations, microstructure evolution and certain mechanical properties in engineering alloys. It has successfully been employed to simulate solid-state phase transformations in Ni-based superalloys [38][39], steels [40], and Al alloys [41] during complex thermo-mechanical processes. MatCalc uses the classical nucleation theory as well as the so-called Svoboda-Fischer-Fratzl-Kozeschnik (SFFK) growth model as the basis for simulating precipitation kinetics [42]. Although MatCalc was originally developed for conventional thermo-mechanical processes, we will show that it is also applicable for AM if the detailed time-temperature profile of the AM build is known. The semi-analytical heat transfer code developed by Stump and Plotkowski [43] is used to simulate these profile throughout the AM build.

1.1. Application to IN738

Inconel-738 (IN738) is a precipitation hardening Ni-based superalloy mainly employed in high-temperature components, e.g. in gas turbines and aero-engines owing to its exceptional mechanical properties at temperatures up to 980 °C, coupled with high resistance to oxidation and corrosion [44]. Its superior high-temperature strength (∼1090 MPa tensile strength) is provided by the L12 ordered Ni3(Al,Ti) γ′ phase that precipitates in a face-centered cubic (FCC) γ matrix [45][46]. Despite offering great properties, IN738, like most superalloys with high γ′ fractions, is challenging to process owing to its propensity to hot cracking [47][48]. Further, machining of such alloys is challenging because of their high strength and work-hardening rates. It is therefore difficult to fabricate complex INC738 parts using traditional manufacturing techniques like casting, welding, and forging.

The emergence of AM has now made it possible to fabricate such parts from IN738 and other superalloys. Some of the current authors’ recent research successfully applied E-PBF to fabricate defect-free IN738 containing γ′ throughout the build [16][17]. The precipitated γ′ were heterogeneously distributed. In particular, Haghdadi et al. [16] studied the origin of the multimodal size distribution of γ′, while Lim et al. [17] investigated the gradient in γ′ character with build height and its correlation to mechanical properties. Based on these results, the present study aims to extend the understanding of the complex and site-specific microstructural evolution in E-PBF IN738 by using a computational modelling approach. New experimental evidence (e.g., micrographs not published previously) is presented here to support the computational results.

2. Materials and Methods

2.1. Materials preparation

IN738 Ni-based superalloy (59.61Ni-8.48Co-7.00Al-17.47Cr-3.96Ti-1.01Mo-0.81W-0.56Ta-0.49Nb-0.47C-0.09Zr-0.05B, at%) gas-atomized powder was used as feedstock. The powders, with average size of 60 ± 7 µm, were manufactured by Praxair and distributed by Astro Alloys Inc. An Arcam Q10 machine by GE Additive with an acceleration voltage of 60 kV was used to fabricate a 15 × 15 × 25 mm3 block (XYZ, Z: build direction) on a 316 stainless steel substrate. The block was 3D-printed using a ‘random’ spot melt pattern. The random spot melt pattern involves randomly selecting points in any given layer, with an equal chance of each point being melted. Each spot melt experienced a dwell time of 0.3 ms, and the layer thickness was 50 µm. Some of the current authors have previously characterized the microstructure of the very same and similar builds in more detail [16][17]. A preheat temperature of ∼1000 °C was set and kept during printing to reduce temperature gradients and, in turn, thermal stresses [49][50][51]. Following printing, the build was separated from the substrate through electrical discharge machining. It should be noted that this sample was simultaneously printed with the one used in [17] during the same build process and on the same build plate, under identical conditions.

2.2. Microstructural characterization

The printed sample was longitudinally cut in the direction of the build using a Struers Accutom-50, ground, and then polished to 0.25 µm suspension via standard techniques. The polished x-z surface was electropolished and etched using Struers A2 solution (perchloric acid in ethanol). Specimens for image analysis were polished using a 0.06 µm colloidal silica. Microstructure analyses were carried out across the height of the build using optical microscopy (OM) and scanning electron microscopy (SEM) with focus on the microstructure evolution (γ′ precipitates) in individual layers. The position of each layer being analyzed was determined by multiplying the layer number by the layer thickness (50 µm). It should be noted that the position of the first layer starts where the thermal profile is tracked (in this case, 2 mm from the bottom). SEM images were acquired using a JEOL 7001 field emission microscope. The brightness and contrast settings, acceleration voltage of 15 kV, working distance of 10 mm, and other SEM imaging parameters were all held constant for analysis of the entire build. The ImageJ software was used for automated image analysis to determine the phase fraction and size of γ′ precipitates and carbides. A 2-pixel radius Gaussian blur, following a greyscale thresholding and watershed segmentation was used [52]. Primary γ′ sizes (>50 nm), were measured using equivalent spherical diameters. The phase fractions were considered equal to the measured area fraction. Secondary γ′ particles (<50 nm) were not considered here. The γ′ size in the following refers to the diameter of a precipitate.

2.3. Hardness testing

A Struers DuraScan tester was utilized for Vickers hardness mapping on a polished x-z surface, from top to bottom under a maximum load of 100 mN and 10 s dwell time. 30 micro-indentations were performed per row. According to the ASTM standard [53], the indentations were sufficiently distant (∼500 µm) to assure that strain-hardened areas did not interfere with one another.

2.4. Computational simulation of E-PBF IN738 build

2.4.1. Thermal profile modeling

The thermal history was generated using the semi-analytical heat transfer code (also known as the 3DThesis code) developed by Stump and Plotkowski [43]. This code is an open-source C++ program which provides a way to quickly simulate the conductive heat transfer found in welding and AM. The key use case for the code is the simulation of larger domains than is practicable with Computational Fluid Dynamics/Finite Element Analysis programs like FLOW-3D AM. Although simulating conductive heat transfer will not be an appropriate simplification for some investigations (for example the modelling of keyholding or pore formation), the 3DThesis code does provide fast estimates of temperature, thermal gradient, and solidification rate which can be useful for elucidating microstructure formation across entire layers of an AM build. The mathematics involved in the code is as follows:

In transient thermal conduction during welding and AM, with uniform and constant thermophysical properties and without considering fluid convection and latent heat effects, energy conservation can be expressed as:(1)��∂�∂�=�∇2�+�̇where � is density, � specific heat, � temperature, � time, � thermal conductivity, and �̇ a volumetric heat source. By assuming a semi-infinite domain, Eq. 1 can be analytically solved. The solution for temperature at a given time (t) using a volumetric Gaussian heat source is presented as:(2)��,�,�,�−�0=33�����32∫0�1������exp−3�′�′2��+�′�′2��+�′�′2����′(3)and��=12��−�′+��2for�=�,�,�(4)and�′�′=�−���′Where � is the vector �,�,� and �� is the location of the heat source.

The numerical integration scheme used is an adaptive Gaussian quadrature method based on the following nondimensionalization:(5)�=��xy2�,�′=��xy2�′,�=��xy,�=��xy,�=��xy,�=���xy

A more detailed explanation of the mathematics can be found in reference [43].

The main source of the thermal cycling present within a powder-bed fusion process is the fusion of subsequent layers. Therefore, regions near the top of a build are expected to undergo fewer thermal cycles than those closer to the bottom. For this purpose, data from the single scan’s thermal influence on multiple layers was spliced to represent the thermal cycles experienced at a single location caused by multiple subsequent layers being fused.

The cross-sectional area simulated by this model was kept constant at 1 × 1 mm2, and the depth was dependent on the build location modelled with MatCalc. For a build location 2 mm from the bottom, the maximum number of layers to simulate is 460. Fig. 1a shows a stitched overview OM image of the entire build indicating the region where this thermal cycle is simulated and tracked. To increase similarity with the conditions of the physical build, each thermal history was constructed from the results of two simulations generated with different versions of a random scan path. The parameters used for these thermal simulations can be found in Table 1. It should be noted that the main purpose of the thermal profile modelling was to demonstrate how the conditions at different locations of the build change relative to each other. Accurately predicting the absolute temperature during the build would require validation via a temperature sensor measurement during the build process which is beyond the scope of the study. Nonetheless, to establish the viability of the heat source as a suitable approximation for this study, an additional sensitivity analysis was conducted. This analysis focused on the influence of energy input on γ′ precipitation behavior, the central aim of this paper. This was achieved by employing varying beam absorption energies (0.76, 0.82 – the values utilized in the simulation, and 0.9). The direct impact of beam absorption efficiency on energy input into the material was investigated. Specifically, the initial 20 layers of the build were simulated and subsequently compared to experimental data derived from SEM. While phase fractions were found to be consistent across all conditions, disparities emerged in the mean size of γ′ precipitates. An absorption efficiency of 0.76 yielded a mean size of approximately 70 nm. Conversely, absorption efficiencies of 0.82 and 0.9 exhibited remarkably similar mean sizes of around 130 nm, aligning closely with the outcomes of the experiments.

Fig. 1

Table 1. A list of parameters used in thermal simulation of E-PBF.

ParameterValue
Spatial resolution5 µm
Time step0.5 s
Beam diameter200 µm
Beam penetration depth1 µm
Beam power1200 W
Beam absorption efficiency0.82
Thermal conductivity25.37 W/(m⋅K)
Chamber temperature1000 °C
Specific heat711.756 J/(kg⋅K)
Density8110 kg/m3

2.4.2. Thermo-kinetic simulation

The numerical analyses of the evolution of precipitates was performed using MatCalc version 6.04 (rel 0.011). The thermodynamic (‘mc_ni.tdb’, version 2.034) and diffusion (‘mc_ni.ddb’, version 2.007) databases were used. MatCalc’s basic principles are elaborated as follows:

The nucleation kinetics of precipitates are computed using a computational technique based on a classical nucleation theory [54] that has been modified for systems with multiple components [42][55]. Accordingly, the transient nucleation rate (�), which expresses the rate at which nuclei are formed per unit volume and time, is calculated as:(6)�=�0��*∙�xp−�*�∙�∙exp−��where �0 denotes the number of active nucleation sites, �* the rate of atomic attachment, � the Boltzmann constant, � the temperature, �* the critical energy for nucleus formation, τ the incubation time, and t the time. � (Zeldovich factor) takes into consideration that thermal excitation destabilizes the nucleus as opposed to its inactive state [54]. Z is defined as follows:(7)�=−12�kT∂2∆�∂�2�*12where ∆� is the overall change in free energy due to the formation of a nucleus and n is the nucleus’ number of atoms. ∆�’s derivative is evaluated at n* (critical nucleus size). �* accounts for the long-range diffusion of atoms required for nucleation, provided that the matrix’ and precipitates’ composition differ. Svoboda et al. [42] developed an appropriate multi-component equation for �*, which is given by:(8)�*=4��*2�4�∑�=1��ki−�0�2�0��0�−1where �* denotes the critical radius for nucleation, � represents atomic distance, and � is the molar volume. �ki and �0� represent the concentration of elements in the precipitate and matrix, respectively. The parameter �0� denotes the rate of diffusion of the ith element within the matrix. The expression for the incubation time � is expressed as [54]:(9)�=12�*�2

and �*, which represents the critical energy for nucleation:(10)�*=16�3�3∆�vol2where � is the interfacial energy, and ∆Gvol the change in the volume free energy. The critical nucleus’ composition is similar to the γ′ phase’s equilibrium composition at the same temperature. � is computed based on the precipitate and matrix compositions, using a generalized nearest neighbor broken bond model, with the assumption of interfaces being planar, sharp, and coherent [56][57][58].

In Eq. 7, it is worth noting that �* represents the fundamental variable in the nucleation theory. It contains �3/∆�vol2 and is in the exponent of the nucleation rate. Therefore, even small variations in γ and/or ∆�vol can result in notable changes in �, especially if �* is in the order of �∙�. This is demonstrated in [38] for UDIMET 720 Li during continuous cooling, where these quantities change steadily during precipitation due to their dependence on matrix’ and precipitate’s temperature and composition. In the current work, these changes will be even more significant as the system is exposed to multiple cycles of rapid cooling and heating.

Once nucleated, the growth of a precipitate is assessed using the radius and composition evolution equations developed by Svoboda et al. [42] with a mean-field method that employs the thermodynamic extremal principle. The expression for the total Gibbs free energy of a thermodynamic system G, which consists of n components and m precipitates, is given as follows:(11)�=∑���0��0�+∑�=1�4���33��+∑�=1��ki�ki+∑�=1�4���2��.

The chemical potential of component � in the matrix is denoted as �0�(�=1,…,�), while the chemical potential of component � in the precipitate is represented by �ki(�=1,…,�,�=1,…,�). These chemical potentials are defined as functions of the concentrations �ki(�=1,…,�,�=1,…,�). The interface energy density is denoted as �, and �� incorporates the effects of elastic energy and plastic work resulting from the volume change of each precipitate.

Eq. (12) establishes that the total free energy of the system in its current state relies on the independent state variables: the sizes (radii) of the precipitates �� and the concentrations of each component �ki. The remaining variables can be determined by applying the law of mass conservation to each component �. This can be represented by the equation:(12)��=�0�+∑�=1�4���33�ki,

Furthermore, the global mass conservation can be expressed by equation:(13)�=∑�=1���When a thermodynamic system transitions to a more stable state, the energy difference between the initial and final stages is dissipated. This model considers three distinct forms of dissipation effects [42]. These include dissipations caused by the movement of interfaces, diffusion within the precipitate and diffusion within the matrix.

Consequently, �̇� (growth rate) and �̇ki (chemical composition’s rate of change) of the precipitate with index � are derived from the linear system of equation system:(14)�ij��=��where �� symbolizes the rates �̇� and �̇ki [42]. Index i contains variables for precipitate radius, chemical composition, and stoichiometric boundary conditions suggested by the precipitate’s crystal structure. Eq. (10) is computed separately for every precipitate �. For a more detailed description of the formulae for the coefficients �ij and �� employed in this work please refer to [59].

The MatCalc software was used to perform the numerical time integration of �̇� and �̇ki of precipitates based on the classical numerical method by Kampmann and Wagner [60]. Detailed information on this method can be found in [61]. Using this computational method, calculations for E-PBF thermal cycles (cyclic heating and cooling) were computed and compared to experimental data. The simulation took approximately 2–4 hrs to complete on a standard laptop.

3. Results

3.1. Microstructure

Fig. 1 displays a stitched overview image and selected SEM micrographs of various γ′ morphologies and carbides after observations of the X-Z surface of the build from the top to 2 mm above the bottom. Fig. 2 depicts a graph that charts the average size and phase fraction of the primary γ′, as it changes with distance from the top to the bottom of the build. The SEM micrographs show widespread primary γ′ precipitation throughout the entire build, with the size increasing in the top to bottom direction. Particularly, at the topmost height, representing the 460th layer (Z = 22.95 mm), as seen in Fig. 1b, the average size of γ′ is 110 ± 4 nm, exhibiting spherical shapes. This is representative of the microstructure after it solidifies and cools to room temperature, without experiencing additional thermal cycles. The γ′ size slightly increases to 147 ± 6 nm below this layer and remains constant until 0.4 mm (∼453rd layer) from the top. At this position, the microstructure still closely resembles that of the 460th layer. After the 453rd layer, the γ′ size grows rapidly to ∼503 ± 19 nm until reaching the 437th layer (1.2 mm from top). The γ′ particles here have a cuboidal shape, and a small fraction is coarser than 600 nm. γ′ continue to grow steadily from this position to the bottom (23 mm from the top). A small fraction of γ′ is > 800 nm.

Fig. 2

Besides primary γ′, secondary γ′ with sizes ranging from 5 to 50 nm were also found. These secondary γ′ precipitates, as seen in Fig. 1f, were present only in the bottom and middle regions. A detailed analysis of the multimodal size distribution of γ′ can be found in [16]. There is no significant variation in the phase fraction of the γ′ along the build. The phase fraction is ∼ 52%, as displayed in Fig. 2. It is worth mentioning that the total phase fraction of γ′ was estimated based on the primary γ′ phase fraction because of the small size of secondary γ′. Spherical MC carbides with sizes ranging from 50 to 400 nm and a phase fraction of 0.8% were also observed throughout the build. The carbides are the light grey precipitates in Fig. 1g. The light grey shade of carbides in the SEM images is due to their composition and crystal structure [52]. These carbides are not visible in Fig. 1b-e because they were dissolved during electro-etching carried out after electropolishing. In Fig. 1g, however, the sample was examined directly after electropolishing, without electro-etching.

Table 2 shows the nominal and measured composition of γ′ precipitates throughout the build by atom probe microscopy as determined in our previous study [17]. No build height-dependent composition difference was observed in either of the γ′ precipitate populations. However, there was a slight disparity between the composition of primary and secondary γ′. Among the main γ′ forming elements, the primary γ′ has a high Ti concentration while secondary γ′ has a high Al concentration. A detailed description of the atom distribution maps and the proxigrams of the constituent elements of γ′ throughout the build can be found in [17].

Table 2. Bulk IN738 composition determined using inductively coupled plasma atomic emission spectroscopy (ICP-AES). Compositions of γ, primary γ′, and secondary γ′ at various locations in the build measured by APT. This information is reproduced from data in Ref. [17] with permission.

at%NiCrCoAlMoWTiNbCBZrTaOthers
Bulk59.1217.478.487.001.010.813.960.490.470.050.090.560.46
γ matrix
Top50.4832.9111.591.941.390.820.440.80.030.030.020.24
Mid50.3732.6111.931.791.540.890.440.10.030.020.020.010.23
Bot48.1034.5712.082.141.430.880.480.080.040.030.010.12
Primary γ′
Top72.172.513.4412.710.250.397.780.560.030.020.050.08
Mid71.602.573.2813.550.420.687.040.730.010.030.040.04
Bot72.342.473.8612.500.260.447.460.500.050.020.020.030.04
Secondary γ′
Mid70.424.203.2314.190.631.035.340.790.030.040.040.05
Bot69.914.063.6814.320.811.045.220.650.050.100.020.11

3.2. Hardness

Fig. 3a shows the Vickers hardness mapping performed along the entire X-Z surface, while Fig. 3b shows the plot of average hardness at different build heights. This hardness distribution is consistent with the γ′ precipitate size gradient across the build direction in Fig. 1Fig. 2. The maximum hardness of ∼530 HV1 is found at ∼0.5 mm away from the top surface (Z = 22.5), where γ′ particles exhibit the smallest observed size in Fig. 2b. Further down the build (∼ 2 mm from the top), the hardness drops to the 440–490 HV1 range. This represents the region where γ′ begins to coarsen. The hardness drops further to 380–430 HV1 at the bottom of the build.

Fig. 3

3.3. Modeling of the microstructural evolution during E-PBF

3.3.1. Thermal profile modeling

Fig. 4 shows the simulated thermal profile of the E-PBF build at a location of 23 mm from the top of the build, using a semi-analytical heat conduction model. This profile consists of the time taken to deposit 460 layers until final cooling, as shown in Fig. 4a. Fig. 4b-d show the magnified regions of Fig. 4a and reveal the first 20 layers from the top, a single layer (first layer from the top), and the time taken for the build to cool after the last layer deposition, respectively.

Fig. 4

The peak temperatures experienced by previous layers decrease progressively as the number of layers increases but never fall below the build preheat temperature (1000 °C). Our simulated thermal cycle may not completely capture the complexity of the actual thermal cycle utilized in the E-PBF build. For instance, the top layer (Fig. 4c), also representing the first deposit’s thermal profile without additional cycles (from powder heating, melting, to solidification), recorded the highest peak temperature of 1390 °C. Although this temperature is above the melting range of the alloy (1230–1360 °C) [62], we believe a much higher temperature was produced by the electron beam to melt the powder. Nevertheless, the solidification temperature and dynamics are outside the scope of this study as our focus is on the solid-state phase transformations during deposition. It takes ∼25 s for each layer to be deposited and cooled to the build temperature. The interlayer dwell time is 125 s. The time taken for the build to cool to room temperature (RT) after final layer deposition is ∼4.7 hrs (17,000 s).

3.3.2. MatCalc simulation

During the MatCalc simulation, the matrix phase is defined as γ. γ′, and MC carbide are included as possible precipitates. The domain of these precipitates is set to be the matrix (γ), and nucleation is assumed to be homogenous. In homogeneous nucleation, all atoms of the unit volume are assumed to be potential nucleation sitesTable 3 shows the computational parameters used in the simulation. All other parameters were set at default values as recommended in the version 6.04.0011 of MatCalc. The values for the interfacial energies are automatically calculated according to the generalized nearest neighbor broken bond model and is one of the most outstanding features in MatCalc [56][57][58]. It should be noted that the elastic misfit strain was not included in the calculation. The output of MatCalc includes phase fraction, size, nucleation rate, and composition of the precipitates. The phase fraction in MatCalc is the volume fraction. Although the experimental phase fraction is the measured area fraction, it is relatively similar to the volume fraction. This is because of the generally larger precipitate size and similar morphology at the various locations along the build [63]. A reliable phase fraction comparison between experiment and simulation can therefore be made.

Table 3. Computational parameters used in the simulation.

Precipitation domainγ
Nucleation site γ′Bulk (homogenous)
Nucleation site MC carbideBulk (Homogenous)
Precipitates class size250
Regular solution critical temperature γ′2500 K[64]
Calculated interfacial energyγ′ = 0.080–0.140 J/m2 and MC carbide = 0.410–0.430 J/m2
3.3.2.1. Precipitate phase fraction

Fig. 5a shows the simulated phase fraction of γ′ and MC carbide during thermal cycling. Fig. 5b is a magnified view of 5a showing the simulated phase fraction at the center points of the top 70 layers, whereas Fig. 5c corresponds to the first two layers from the top. As mentioned earlier, the top layer (460th layer) represents the microstructure after solidification. The microstructure of the layers below is determined by the number of thermal cycles, which increases with distance to the top. For example, layers 459, 458, 457, up to layer 1 (region of interest) experience 1, 2, 3 and 459 thermal cycles, respectively. In the top layer in Fig. 5c, the volume fraction of γ′ and carbides increases with temperature. For γ′, it decreases to zero when the temperature is above the solvus temperature after a few seconds. Carbides, however, remain constant in their volume fraction reaching equilibrium (phase fraction ∼ 0.9%) in a short time. The topmost layer can be compared to the first deposit, and the peak in temperature symbolizes the stage where the electron beam heats the powder until melting. This means γ′ and carbide precipitation might have started in the powder particles during heating from the build temperature and electron beam until the onset of melting, where γ′ dissolves, but carbides remain stable [28].

Fig. 5

During cooling after deposition, γ′ reprecipitates at a temperature of 1085 °C, which is below its solvus temperature. As cooling progresses, the phase fraction increases steadily to ∼27% and remains constant at 1000 °C (elevated build temperature). The calculated equilibrium fraction of phases by MatCalc is used to show the complex precipitation characteristics in this alloy. Fig. 6 shows that MC carbides form during solidification at 1320 °C, followed by γ′, which precipitate when the solidified layer cools to 1140 °C. This indicates that all deposited layers might contain a negligible amount of these precipitates before subsequent layer deposition, while being at the 1000 °C build temperature or during cooling to RT. The phase diagram also shows that the equilibrium fraction of the γ′ increases as temperature decreases. For instance, at 1000, 900, and 800 °C, the phase fractions are ∼30%, 38%, and 42%, respectively.

Fig. 6

Deposition of subsequent layers causes previous layers to undergo phase transformations as they are exposed to several thermal cycles with different peak temperatures. In Fig. 5c, as the subsequent layer is being deposited, γ′ in the previous layer (459th layer) begins to dissolve as the temperature crosses the solvus temperature. This is witnessed by the reduction of the γ′ phase fraction. This graph also shows how this phase dissolves during heating. However, the phase fraction of MC carbide remains stable at high temperatures and no dissolution is seen during thermal cycling. Upon cooling, the γ′ that was dissolved during heating reprecipitates with a surge in the phase fraction until 1000 °C, after which it remains constant. This microstructure is similar to the solidification microstructure (layer 460), with a similar γ′ phase fraction (∼27%).

The complete dissolution and reprecipitation of γ′ continue for several cycles until the 50th layer from the top (layer 411), where the phase fraction does not reach zero during heating to the peak temperature (see Fig. 5d). This indicates the ‘partial’ dissolution of γ′, which continues progressively with additional layers. It should be noted that the peak temperatures for layers that underwent complete dissolution were much higher (1170–1300 °C) than the γ′ solvus.

The dissolution and reprecipitation of γ′ during thermal cycling are further confirmed in Fig. 7, which summarizes the nucleation rate, phase fraction, and concentration of major elements that form γ′ in the matrix. Fig. 7b magnifies a single layer (3rd layer from top) within the full dissolution region in Fig. 7a to help identify the nucleation and growth mechanisms. From Fig. 7b, γ′ nucleation begins during cooling whereby the nucleation rate increases to reach a maximum value of approximately 1 × 1020 m−3s−1. This fast kinetics implies that some rearrangement of atoms is required for γ′ precipitates to form in the matrix [65][66]. The matrix at this stage is in a non-equilibrium condition. Its composition is similar to the nominal composition and remains unchanged. The phase fraction remains insignificant at this stage although nucleation has started. The nucleation rate starts declining upon reaching the peak value. Simultaneously, diffusion-controlled growth of existing nuclei occurs, depleting the matrix of γ′ forming elements (Al and Ti). Thus, from (7)(11), ∆�vol continuously decreases until nucleation ceases. The growth of nuclei is witnessed by the increase in phase fraction until a constant level is reached at 27% upon cooling to and holding at build temperature. This nucleation event is repeated several times.

Fig. 7

At the onset of partial dissolution, the nucleation rate jumps to 1 × 1021 m−3s−1, and then reduces sharply at the middle stage of partial dissolution. The nucleation rate reaches 0 at a later stage. Supplementary Fig. S1 shows a magnified view of the nucleation rate, phase fraction, and thermal profile, underpinning this trend. The jump in nucleation rate at the onset is followed by a progressive reduction in the solute content of the matrix. The peak temperatures (∼1130–1160 °C) are lower than those in complete dissolution regions but still above or close to the γ′ solvus. The maximum phase fraction (∼27%) is similar to that of the complete dissolution regions. At the middle stage, the reduction in nucleation rate is accompanied by a sharp drop in the matrix composition. The γ′ fraction drops to ∼24%, where the peak temperatures of the layers are just below or at γ′ solvus. The phase fraction then increases progressively through the later stage of partial dissolution to ∼30% towards the end of thermal cycling. The matrix solute content continues to drop although no nucleation event is seen. The peak temperatures are then far below the γ′ solvus. It should be noted that the matrix concentration after complete dissolution remains constant. Upon cooling to RT after final layer deposition, the nucleation rate increases again, indicating new nucleation events. The phase fraction reaches ∼40%, with a further depletion of the matrix in major γ′ forming elements.

3.3.2.2. γ′ size distribution

Fig. 8 shows histograms of the γ′ precipitate size distributions (PSD) along the build height during deposition. These PSDs are predicted at the end of each layer of interest just before final cooling to room temperature, to separate the role of thermal cycles from final cooling on the evolution of γ′. The PSD for the top layer (layer 460) is shown in Fig. 8a (last solidified region with solidification microstructure). The γ′ size ranges from 120 to 230 nm and is similar to the 44 layers below (2.2 mm from the top).

Fig. 8

Further down the build, γ′ begins to coarsen after layer 417 (44th layer from top). Fig. 8c shows the PSD after the 44th layer, where the γ′ size exhibits two peaks at ∼120–230 and ∼300 nm, with most of the population being in the former range. This is the onset of partial dissolution where simultaneously with the reprecipitation and growth of fresh γ′, the undissolved γ′ grows rapidly through diffusive transport of atoms to the precipitates. This is shown in Fig. 8c, where the precipitate class sizes between 250 and 350 represent the growth of undissolved γ′. Although this continues in the 416th layer, the phase fractions plot indicates that the onset of partial dissolution begins after the 411th layer. This implies that partial dissolution started early, but the fraction of undissolved γ′ was too low to impact the phase fraction. The reprecipitated γ′ are mostly in the 100–220 nm class range and similar to those observed during full dissolution.

As the number of layers increases, coarsening intensifies with continued growth of more undissolved γ′, and reprecipitation and growth of partially dissolved ones. Fig. 8d, e, and f show this sequence. Further down the build, coarsening progresses rapidly, as shown in Figs. 8d, 8e, and 8f. The γ′ size ranges from 120 to 1100 nm, with the peaks at 160, 180, and 220 nm in Figs. 8d, 8e, and 8f, respectively. Coarsening continues until nucleation ends during dissolution, where only the already formed γ′ precipitates continue to grow during further thermal cycling. The γ′ size at this point is much larger, as observed in layers 361 and 261, and continues to increase steadily towards the bottom (layer 1). Two populations in the ranges of ∼380–700 and ∼750–1100 nm, respectively, can be seen. The steady growth of γ′ towards the bottom is confirmed by the gradual decrease in the concentration of solute elements in the matrix (Fig. 7a). It should be noted that for each layer, the γ′ class with the largest size originates from continuous growth of the earliest set of the undissolved precipitates.

Fig. 9Fig. 10 and supplementary Figs. S2 and S3 show the γ′ size evolution during heating and cooling of a single layer in the full dissolution region, and early, middle stages, and later stages of partial dissolution, respectively. In all, the size of γ′ reduces during layer heating. Depending on the peak temperature of the layer which varies with build height, γ′ are either fully or partially dissolved as mentioned earlier. Upon cooling, the dissolved γ′ reprecipitate.

Fig. 9
Fig. 10

In Fig. 9, those layers that underwent complete dissolution (top layers) were held above γ′ solvus temperature for longer. In Fig. 10, layers at the early stage of partial dissolution spend less time in the γ′ solvus temperature region during heating, leading to incomplete dissolution. In such conditions, smaller precipitates are fully dissolved while larger ones shrink [67]. Layers in the middle stages of partial dissolution have peak temperatures just below or at γ′ solvus, not sufficient to achieve significant γ′ dissolution. As seen in supplementary Fig. S2, only a few smaller γ′ are dissolved back into the matrix during heating, i.e., growth of precipitates is more significant than dissolution. This explains the sharp decrease in concentration of Al and Ti in the matrix in this layer.

The previous sections indicate various phenomena such as an increase in phase fraction, further depletion of matrix composition, and new nucleation bursts during cooling. Analysis of the PSD after the final cooling of the build to room temperature allows a direct comparison to post-printing microstructural characterization. Fig. 11 shows the γ′ size distribution of layer 1 (460th layer from the top) after final cooling to room temperature. Precipitation of secondary γ′ is observed, leading to the multimodal size distribution of secondary and primary γ′. The secondary γ′ size falls within the 10–80 nm range. As expected, a further growth of the existing primary γ′ is also observed during cooling.

Fig. 11
3.3.2.3. γ′ chemistry after deposition

Fig. 12 shows the concentration of the major elements that form γ′ (Al, Ti, and Ni) in the primary and secondary γ′ at the bottom of the build, as calculated by MatCalc. The secondary γ′ has a higher Al content (13.5–14.5 at% Al), compared to 13 at% Al in the primary γ′. Additionally, within the secondary γ′, the smallest particles (∼10 nm) have higher Al contents than larger ones (∼70 nm). In contrast, for the primary γ′, there is no significant variation in the Al content as a function of their size. The Ni concentration in secondary γ′ (71.1–72 at%) is also higher in comparison to the primary γ′ (70 at%). The smallest secondary γ′ (∼10 nm) have higher Ni contents than larger ones (∼70 nm), whereas there is no substantial change in the Ni content of primary γ′, based on their size. As expected, Ti shows an opposite size-dependent variation. It ranges from ∼ 7.7–8.7 at% Ti in secondary γ′ to ∼9.2 at% in primary γ′. Similarly, within the secondary γ′, the smallest (∼10 nm) have lower Al contents than the larger ones (∼70 nm). No significant variation is observed for Ti content in primary γ′.

Fig. 12

4. Discussion

A combined modelling method is utilized to study the microstructural evolution during E-PBF of IN738. The presented results are discussed by examining the precipitation and dissolution mechanism of γ′ during thermal cycling. This is followed by a discussion on the phase fraction and size evolution of γ′ during thermal cycling and after final cooling. A brief discussion on carbide morphology is also made. Finally, a comparison is made between the simulation and experimental results to assess their agreement.

4.1. γ′ morphology as a function of build height

4.1.1. Nucleation of γ′

The fast precipitation kinetics of the γ′ phase enables formation of γ′ upon quenching from higher temperatures (above solvus) during thermal cycling [66]. In Fig. 7b, for a single layer in the full dissolution region, during cooling, the initial increase in nucleation rate signifies the first formation of nuclei. The slight increase in nucleation rate during partial dissolution, despite a decrease in the concentration of γ′ forming elements, may be explained by the nucleation kinetics. During partial dissolution and as the precipitates shrink, it is assumed that the regions at the vicinity of partially dissolved precipitates are enriched in γ′ forming elements [68][69]. This differs from the full dissolution region, in which case the chemical composition is evenly distributed in the matrix. Several authors have attributed the solute supersaturation of the matrix around primary γ′ to partial dissolution during isothermal ageing [69][70][71][72]. The enhanced supersaturation in the regions close to the precipitates results in a much higher driving force for nucleation, leading to a higher nucleation rate upon cooling. This phenomenon can be closely related to the several nucleation bursts upon continuous cooling of Ni-based superalloys, where second nucleation bursts exhibit higher nucleation rates [38][68][73][74].

At middle stages of partial dissolution, the reduction in the nucleation rate indicates that the existing composition and low supersaturation did not trigger nucleation as the matrix was closer to the equilibrium state. The end of a nucleation burst means that the supersaturation of Al and Ti has reached a low level, incapable of providing sufficient driving force during cooling to or holding at 1000 °C for further nucleation [73]. Earlier studies on Ni-based superalloys have reported the same phenomenon during ageing or continuous cooling from the solvus temperature to RT [38][73][74].

4.1.2. Dissolution of γ′ during thermal cycling

γ′ dissolution kinetics during heating are fast when compared to nucleation due to exponential increase in phase transformation and diffusion activities with temperature [65]. As shown in Fig. 9Fig. 10, and supplementary Figs. S2 and S3, the reduction in γ′ phase fraction and size during heating indicates γ′ dissolution. This is also revealed in Fig. 5 where phase fraction decreases upon heating. The extent of γ′ dissolution mostly depends on the temperature, time spent above γ′ solvus, and precipitate size [75][76][77]. Smaller γ′ precipitates are first to be dissolved [67][77][78]. This is mainly because more solute elements need to be transported away from large γ′ precipitates than from smaller ones [79]. Also, a high temperature above γ′ solvus temperature leads to a faster dissolution rate [80]. The equilibrium solvus temperature of γ′ in IN738 in our MatCalc simulation (Fig. 6) and as reported by Ojo et al. [47] is 1140 °C and 1130–1180 °C, respectively. This means the peak temperature experienced by previous layers decreases progressively from γ′ supersolvus to subsolvus, near-solvus, and far from solvus as the number of subsequent layers increases. Based on the above, it can be inferred that the degree of dissolution of γ′ contributes to the gradient in precipitate distribution.

Although the peak temperatures during later stages of partial dissolution are much lower than the equilibrium γ′ solvus, γ′ dissolution still occurs but at a significantly lower rate (supplementary Fig. S3). Wahlmann et al. [28] also reported a similar case where they observed the rapid dissolution of γ′ in CMSX-4 during fast heating and cooling cycles at temperatures below the γ′ solvus. They attributed this to the γ′ phase transformation process taking place in conditions far from the equilibrium. While the same reasoning may be valid for our study, we further believe that the greater surface area to volume ratio of the small γ′ precipitates contributed to this. This ratio means a larger area is available for solute atoms to diffuse into the matrix even at temperatures much below the solvus [81].

4.2. γ′ phase fraction and size evolution

4.2.1. During thermal cycling

In the first layer, the steep increase in γ′ phase fraction during heating (Fig. 5), which also represents γ′ precipitation in the powder before melting, has qualitatively been validated in [28]. The maximum phase fraction of 27% during the first few layers of thermal cycling indicates that IN738 theoretically could reach the equilibrium state (∼30%), but the short interlayer time at the build temperature counteracts this. The drop in phase fraction at middle stages of partial dissolution is due to the low number of γ′ nucleation sites [73]. It has been reported that a reduction of γ′ nucleation sites leads to a delay in obtaining the final volume fraction as more time is required for γ′ precipitates to grow and reach equilibrium [82]. This explains why even upon holding for 150 s before subsequent layer deposition, the phase fraction does not increase to those values that were observed in the previous full γ′ dissolution regions. Towards the end of deposition, the increase in phase fraction to the equilibrium value of 30% is as a result of the longer holding at build temperature or close to it [83].

During thermal cycling, γ′ particles begin to grow immediately after they first precipitate upon cooling. This is reflected in the rapid increase in phase fraction and size during cooling in Fig. 5 and supplementary Fig. S2, respectively. The rapid growth is due to the fast diffusion of solute elements at high temperatures [84]. The similar size of γ′ for the first 44 layers from the top can be attributed to the fact that all layers underwent complete dissolution and hence, experienced the same nucleation event and growth during deposition. This corresponds with the findings by Balikci et al. [85], who reported that the degree of γ′ precipitation in IN738LC does not change when a solution heat treatment is conducted above a certain critical temperature.

The increase in coarsening rate (Fig. 8) during thermal cycling can first be ascribed to the high peak temperature of the layers [86]. The coarsening rate of γ′ is known to increase rapidly with temperature due to the exponential growth of diffusion activity. Also, the simultaneous dissolution with coarsening could be another reason for the high coarsening rate, as γ′ coarsening is a diffusion-driven process where large particles grow by consuming smaller ones [78][84][86][87]. The steady growth of γ′ towards the bottom of the build is due to the much lower layer peak temperature, which is almost close to the build temperature, and reduced dissolution activity, as is seen in the much lower solute concentration in γ′ compared to those in the full and partial dissolution regions.

4.2.2. During cooling

The much higher phase fraction of ∼40% upon cooling signifies the tendency of γ′ to reach equilibrium at lower temperatures (Fig. 4). This is due to the precipitation of secondary γ′ and a further increase in the size of existing primary γ′, which leads to a multimodal size distribution of γ′ after cooling [38][73][88][89][90]. The reason for secondary γ′ formation during cooling is as follows: As cooling progresses, it becomes increasingly challenging to redistribute solute elements in the matrix owing to their lower mobility [38][73]. A higher supersaturation level in regions away from or free of the existing γ′ precipitates is achieved, making them suitable sites for additional nucleation bursts. More cooling leads to the growth of these secondary γ′ precipitates, but as the temperature and in turn, the solute diffusivity is low, growth remains slow.

4.3. Carbides

MC carbides in IN738 are known to have a significant impact on the high-temperature strength. They can also act as effective hardening particles and improve the creep resistance [91]. Precipitation of MC carbides in IN738 and several other superalloys is known to occur during solidification or thermal treatments (e.g., hot isostatic pressing) [92]. In our case, this means that the MC carbides within the E-PBF build formed because of the thermal exposure from the E-PBF thermal cycle in addition to initial solidification. Our simulation confirms this as MC carbides appear during layer heating (Fig. 5). The constant and stable phase fraction of MC carbides during thermal cycling can be attributed to their high melting point (∼1360 °C) and the short holding time at peak temperatures [75][93][94]. The solvus temperature for most MC carbides exceeds most of the peak temperatures observed in our simulation, and carbide dissolution kinetics at temperatures above the solvus are known to be comparably slow [95]. The stable phase fraction and random distribution of MC carbides signifies the slight influence on the gradient in hardness.

4.4. Comparison of simulations and experiments

4.4.1. Precipitate phase fraction and morphology as a function of build height

A qualitative agreement is observed for the phase fraction of carbides, i.e. ∼0.8% in the experiment and ∼0.9% in the simulation. The phase fraction of γ′ differs, with the experiment reporting a value of ∼51% and the simulation, 40%. Despite this, the size distribution of primary γ′ along the build shows remarkable consistency between experimental and computational analyses. It is worth noting that the primary γ′ morphology in the experimental analysis is observed in the as-fabricated state, whereas the simulation (Fig. 8) captures it during deposition process. The primary γ′ size in the experiment is expected to experience additional growth during the cooling phase. Regardless, both show similar trends in primary γ′ size increments from the top to the bottom of the build. The larger primary γ’ size in the simulation versus the experiment can be attributed to the fact that experimental and simulation results are based on 2D and 3D data, respectively. The absence of stereological considerations [96] in our analysis could have led to an underestimation of the precipitate sizes from SEM measurements. The early starts of coarsening (8th layer) in the experiment compared to the simulation (45th layer) can be attributed to a higher actual γ′ solvus temperature than considered in our simulation [47]. The solvus temperature of γ′ in a Ni-based superalloy is mainly determined by the detailed composition. A high amount of Cr and Co are known to reduce the solvus temperature, whereas Ta and Mo will increase it [97][98][99]. The elemental composition from our experimental work was used for the simulation except for Ta. It should be noted that Ta is not included in the thermodynamic database in MatCalc used, and this may have reduced the solvus temperature. This could also explain the relatively higher γ′ phase fraction in the experiment than in simulation, as a higher γ′ solvus temperature will cause more γ′ to precipitate and grow early during cooling [99][100].

Another possible cause of this deviation can be attributed to the extent of γ′ dissolution, which is mainly determined by the peak temperature. It can be speculated that individual peak temperatures at different layers in the simulation may have been over-predicted. However, one needs to consider that the true thermal profile is likely more complicated in the actual E-PBF process [101]. For example, the current model assumes that the thermophysical properties of the material are temperature-independent, which is not realistic. Many materials, including IN738, exhibit temperature-dependent properties such as thermal conductivityspecific heat capacity, and density [102]. This means that heat transfer simulations may underestimate or overestimate the temperature gradients and cooling rates within the powder bed and the solidified part. Additionally, the model does not account for the reduced thermal diffusivity through unmelted powder, where gas separating the powder acts as insulation, impeding the heat flow [1]. In E-PBF, the unmelted powder regions with trapped gas have lower thermal diffusivity compared to the fully melted regions, leading to localized temperature variations, and altered solidification behavior. These limitations can impact the predictions, particularly in relation to the carbide dissolution, as the peak temperatures may be underestimated.

While acknowledging these limitations, it is worth emphasizing that achieving a detailed and accurate representation of each layer’s heat source would impose tough computational challenges. Given the substantial layer count in E-PBF, our decision to employ a semi-analytical approximation strikes a balance between computational feasibility and the capture of essential trends in thermal profiles across diverse build layers. In future work, a dual-calibration strategy is proposed to further reduce simulation-experiment disparities. By refining temperature-independent thermophysical property approximations and absorptivity in the heat source model, and by optimizing interfacial energy descriptions in the kinetic model, the predictive precision could be enhanced. Further refining the simulation controls, such as adjusting the precipitate class size may enhance quantitative comparisons between modeling outcomes and experimental data in future work.

4.4.2. Multimodal size distribution of γ′ and concentration

Another interesting feature that sees qualitative agreement between the simulation and the experiment is the multimodal size distribution of γ′. The formation of secondary γ′ particles in the experiment and most E-PBF Ni-based superalloys is suggested to occur at low temperatures, during final cooling to RT [16][73][90]. However, so far, this conclusion has been based on findings from various continuous cooling experiments, as the study of the evolution during AM would require an in-situ approach. Our simulation unambiguously confirms this in an AM context by providing evidence for secondary γ′ precipitation during slow cooling to RT. Additionally, it is possible to speculate that the chemical segregation occurring during solidification, due to the preferential partitioning of certain elements between the solid and liquid phases, can contribute to the multimodal size distribution during deposition [51]. This is because chemical segregation can result in variations in the local composition of superalloys, which subsequently affects the nucleation and growth of γ′. Regions with higher concentrations of alloying elements will encourage the formation of larger γ′ particles, while regions with lower concentrations may favor the nucleation of smaller precipitates. However, it is important to acknowledge that the elevated temperature during the E-PBF process will largely homogenize these compositional differences [103][104].

A good correlation is also shown in the composition of major γ′ forming elements (Al and Ti) in primary and secondary γ′. Both experiment and simulation show an increasing trend for Al content and a decreasing trend for Ti content from primary to secondary γ′. The slight composition differences between primary and secondary γ′ particles are due to the different diffusivity of γ′ stabilizers at different thermal conditions [105][106]. As the formation of multimodal γ′ particles with different sizes occurs over a broad temperature range, the phase chemistry of γ′ will be highly size dependent. The changes in the chemistry of various γ′ (primary, secondary, and tertiary) have received significant attention since they have a direct influence on the performance [68][105][107][108][109]. Chen et al. [108][109], reported a high Al content in the smallest γ′ precipitates compared to the largest, while Ti showed an opposite trend during continuous cooling in a RR1000 Ni-based superalloy. This was attributed to the temperature and cooling rate at which the γ′ precipitates were formed. The smallest precipitates formed last, at the lowest temperature and cooling rate. A comparable observation is evident in the present investigation, where the secondary γ′ forms at a low temperature and cooling rate in comparison to the primary. The temperature dependence of γ′ chemical composition is further evidenced in supplementary Fig. S4, which shows the equilibrium chemical composition of γ′ as a function of temperature.

5. Conclusions

A correlative modelling approach capable of predicting solid-state phase transformations kinetics in metal AM was developed. This approach involves computational simulations with a semi-analytical heat transfer model and the MatCalc thermo-kinetic software. The method was used to predict the phase transformation kinetics and detailed morphology and chemistry of γ′ and MC during E-PBF of IN738 Ni-based superalloy. The main conclusions are:

  • 1.The computational simulations are in qualitative agreement with the experimental observations. This is particularly true for the γ′ size distribution along the build height, the multimodal size distribution of particles, and the phase fraction of MC carbides.
  • 2.The deviations between simulation and experiment in terms of γ′ phase fraction and location in the build are most likely attributed to a higher γ′ solvus temperature during the experiment than in the simulation, which is argued to be related to the absence of Ta in the MatCalc database.
  • 3.The dissolution and precipitation of γ′ occur fast and under non-equilibrium conditions. The level of γ′ dissolution determines the gradient in γ′ size distribution along the build. After thermal cycling, the final cooling to room temperature has further significant impacts on the final γ′ size, morphology, and distribution.
  • 4.A negligible amount of γ′ forms in the first deposited layer before subsequent layer deposition, and a small amount of γ′ may also form in the powder induced by the 1000 °C elevated build temperature before melting.

Our findings confirm the suitability of MatCalc to predict the microstructural evolution at various positions throughout a build in a Ni-based superalloy during E-PBF. It also showcases the suitability of a tool which was originally developed for traditional thermo-mechanical processing of alloys to the new additive manufacturing context. Our simulation capabilities are likely extendable to other alloy systems that undergo solid-state phase transformations implemented in MatCalc (various steels, Ni-based superalloys, and Al-alloys amongst others) as well as other AM processes such as L-DED and L-PBF which have different thermal cycle characteristics. New tools to predict the microstructural evolution and properties during metal AM are important as they provide new insights into the complexities of AM. This will enable control and design of AM microstructures towards advanced materials properties and performances.

CRediT authorship contribution statement

Primig Sophie: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization. Adomako Nana Kwabena: Writing – original draft, Writing – review & editing, Visualization, Software, Investigation, Formal analysis, Conceptualization. Haghdadi Nima: Writing – review & editing, Supervision, Project administration, Methodology, Conceptualization. Dingle James F.L.: Methodology, Conceptualization, Software, Writing – review & editing, Visualization. Kozeschnik Ernst: Writing – review & editing, Software, Methodology. Liao Xiaozhou: Writing – review & editing, Project administration, Funding acquisition. Ringer Simon P: Writing – review & editing, Project administration, 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

This research was sponsored by the Department of Industry, Innovation, and Science under the auspices of the AUSMURI program – which is a part of the Commonwealth’s Next Generation Technologies Fund. The authors acknowledge the facilities and the scientific and technical assistance at the Electron Microscope Unit (EMU) within the Mark Wainwright Analytical Centre (MWAC) at UNSW Sydney and Microscopy Australia. Nana Adomako is supported by a UNSW Scientia PhD scholarship. Michael Haines’ (UNSW Sydney) contribution to the revised version of the original manuscript is thankfully acknowledged.

Appendix A. Supplementary material

Download : Download Word document (462KB)

Supplementary material.

Data Availability

Data will be made available on request.

References

Thermo-fluid modeling of influence of attenuated laser beam intensity profile on melt pool behavior in laser-assisted powder-based direct energy deposition

레이저 보조 분말 기반 직접 에너지 증착에서 용융 풀 거동에 대한 감쇠 레이저 빔 강도 프로파일의 영향에 대한 열유체 모델링

Thermo-fluid modeling of influence of attenuated laser beam intensity profile on melt pool behavior in laser-assisted powder-based direct energy deposition

Mohammad Sattari, Amin Ebrahimi, Martin Luckabauer, Gert-willem R.B.E. Römer

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Professional

5Downloads (Pure)

Abstract

A numerical framework based on computational fluid dynamics (CFD), using the finite volume method (FVM) and volume of fluid (VOF) technique is presented to investigate the effect of the laser beam intensity profile on melt pool behavior in laser-assisted powder-based directed energy deposition (L-DED). L-DED is an additive manufacturing (AM) process that utilizes a laser beam to fuse metal powder particles. To assure high-fidelity modeling, it was found that it is crucial to accurately model the interaction between the powder stream and the laser beam in the gas region above the substrate. The proposed model considers various phenomena including laser energy attenuation and absorption, multiple reflections of the laser rays, powder particle stream, particle-fluid interaction, temperature-dependent properties, buoyancy effects, thermal expansion, solidification shrinkage and drag, and Marangoni flow. The latter is induced by temperature and element-dependent surface tension. The model is validated using experimental results and highlights the importance of considering laser energy attenuation. Furthermore, the study investigates how the laser beam intensity profile affects melt pool size and shape, influencing the solidification microstructure and mechanical properties of the deposited material. The proposed model has the potential to optimize the L-DED process for a variety of materials and provides insights into the capability of numerical modeling for additive manufacturing optimization.

Original languageEnglish
Title of host publicationFlow-3D World Users Conference
Publication statusPublished – 2023
EventFlow-3D World User Conference – Strasbourg, France
Duration: 5 Jun 2023 → 7 Jun 2023

Conference

ConferenceFlow-3D World User Conference
Country/TerritoryFrance
CityStrasbourg
Period5/06/23 → 7/06/23
Figure 14. Defects: (a) Unmelt defects(Scheme NO.4);(b) Pores defects(Scheme NO.1); (c); Spattering defect (Scheme NO.3); (d) Low overlapping rate defects(Scheme NO.5).

Molten pool structure, temperature and velocity
flow in selective laser melting AlCu5MnCdVA alloy

용융 풀 구조, 선택적 온도 및 속도 흐름 레이저 용융 AlCu5MnCdVA 합금

Pan Lu1 , Zhang Cheng-Lin2,6,Wang Liang3, Liu Tong4 and Liu Jiang-lin5
1 Aviation and Materials College, Anhui Technical College of Mechanical and Electrical Engineering, Wuhu Anhui 241000, People’s
Republic of China 2 School of Engineering Science, University of Science and Technology of China, Hefei Anhui 230026, People’s Republic of China 3 Anhui Top Additive Manufacturing Technology Co., Ltd., Wuhu Anhui 241300, People’s Republic of China 4 Anhui Chungu 3D Printing Institute of Intelligent Equipment and Industrial Technology, Anhui 241300, People’s Republic of China 5 School of Mechanical and Transportation Engineering, Taiyuan University of Technology, Taiyuan Shanxi 030024, People’s Republic of
China 6 Author to whom any correspondence should be addressed.
E-mail: ahjdpanlu@126.com, jiao__zg@126.com, ahjdjxx001@126.com,tongliu1988@126.com and liujianglin@tyut.edu.cn

Keywords

SLM, molten pool, AlCu5MnCdVA alloy, heat flow, velocity flow, numerical simulation

Abstract

선택적 레이저 용융(SLM)은 열 전달, 용융, 상전이, 기화 및 물질 전달을 포함하는 복잡한 동적 비평형 프로세스인 금속 적층 제조(MAM)에서 가장 유망한 기술 중 하나가 되었습니다. 용융 풀의 특성(구조, 온도 흐름 및 속도 흐름)은 SLM의 최종 성형 품질에 결정적인 영향을 미칩니다. 이 연구에서는 선택적 레이저 용융 AlCu5MnCdVA 합금의 용융 풀 구조, 온도 흐름 및 속도장을 연구하기 위해 수치 시뮬레이션과 실험을 모두 사용했습니다.

그 결과 용융풀의 구조는 다양한 형태(깊은 오목 구조, 이중 오목 구조, 평면 구조, 돌출 구조 및 이상적인 평면 구조)를 나타냈으며, 용융 풀의 크기는 약 132 μm × 107 μm × 50 μm였습니다. : 용융풀은 초기에는 여러 구동력에 의해 깊이 15μm의 깊은 오목형상이었으나, 성형 후기에는 장력구배에 의해 높이 10μm의 돌출형상이 되었다. 용융 풀 내부의 금속 흐름은 주로 레이저 충격력, 금속 액체 중력, 표면 장력 및 반동 압력에 의해 구동되었습니다.

AlCu5MnCdVA 합금의 경우, 금속 액체 응고 속도가 매우 빠르며(3.5 × 10-4 S), 가열 속도 및 냉각 속도는 각각 6.5 × 107 K S-1 및 1.6 × 106 K S-1 에 도달했습니다. 시각적 표준으로 표면 거칠기를 선택하고, 낮은 레이저 에너지 AlCu5MnCdVA 합금 최적 공정 매개변수 창을 수치 시뮬레이션으로 얻었습니다: 레이저 출력 250W, 부화 공간 0.11mm, 층 두께 0.03mm, 레이저 스캔 속도 1.5m s-1 .

또한, 실험 프린팅과 수치 시뮬레이션과 비교할 때, 용융 풀의 폭은 각각 약 205um 및 약 210um이었고, 인접한 두 용융 트랙 사이의 중첩은 모두 약 65um이었다. 결과는 수치 시뮬레이션 결과가 실험 인쇄 결과와 기본적으로 일치함을 보여 수치 시뮬레이션 모델의 정확성을 입증했습니다.

Selective Laser Melting (SLM) has become one of the most promising technologies in Metal Additive Manufacturing (MAM), which is a complex dynamic non-equilibrium process involving heat transfer, melting, phase transition, vaporization and mass transfer. The characteristics of the molten pool (structure, temperature flow and velocity flow) have a decisive influence on the final forming quality of SLM. In this study, both numerical simulation and experiments were employed to study molten pool structure, temperature flow and velocity field in Selective Laser Melting AlCu5MnCdVA alloy. The results showed the structure of molten pool showed different forms(deep-concave structure, double-concave structure, plane structure, protruding structure and ideal planar structure), and the size of the molten pool was approximately 132 μm × 107 μm × 50 μm: in the early stage, molten pool was in a state of deep-concave shape with a depth of 15 μm due to multiple driving forces, while a protruding shape with a height of 10 μm duo to tension gradient in the later stages of forming. The metal flow inside the molten pool was mainly driven by laser impact force, metal liquid gravity, surface tension and recoil pressure. For AlCu5MnCdVA alloy, metal liquid solidification speed was extremely fast(3.5 × 10−4 S), the heating rate and cooling rate reached 6.5 × 107 K S−1 and 1.6 × 106 K S−1 , respectively. Choosing surface roughness as a visual standard, low-laser energy AlCu5MnCdVA alloy optimum process parameters window was obtained by numerical simulation: laser power 250 W, hatching space 0.11 mm, layer thickness 0.03 mm, laser scanning velocity 1.5 m s−1 . In addition, compared with experimental printing and numerical simulation, the width of the molten pool was about 205 um and about 210 um, respectively, and overlapping between two adjacent molten tracks was all about 65 um. The results showed that the numerical simulation results were basically consistent with the experimental print results, which proved the correctness of the numerical simulation model.

Figure 1. AlCu5MnCdVA powder particle size distribution.
Figure 1. AlCu5MnCdVA powder particle size distribution.
Figure 2. AlCu5MnCdVA powder
Figure 2. AlCu5MnCdVA powder
Figure 3. Finite element model and calculation domains of SLM.
Figure 3. Finite element model and calculation domains of SLM.
Figure 4. SLM heat transfer process.
Figure 4. SLM heat transfer process.
Figure 14. Defects: (a) Unmelt defects(Scheme NO.4);(b) Pores defects(Scheme NO.1); (c); Spattering defect (Scheme NO.3); (d) Low
overlapping rate defects(Scheme NO.5).
Figure 17. Two-pass molten tracks overlapping for Scheme NO.2.
Figure 17. Two-pass molten tracks overlapping for Scheme NO.2.

References

[1] Cuiyun H 2008 Phase diagram determination and thermodynamic study of Al–Cu–Mn, Al–Cu–Si, Al–Mg–Ni and Ni–Ti–Si systems Central South University
[2] Zhanfei Z 2017 Study on theta phase segregation and room temperature properties of high strength cast Al–Cu–Mn alloy Lanzhou University of Technology
[3] Nie X et al 2018 Analysis of processing parameters and characteristics of selective laser melted high strength Al–Cu–Mg alloys: from single tracks to cubic samplesJ. Mater. Process. Technol. 256 69–77
[4] Shenping Y et al 2017 Laser absorptance measurement of commonly used metal materials in laser additive manufacturing technology Aviation Manufacturing Technology 12 23–9
[5] Wenqing W 2007 Relationship between cooling rate and grain size of AlCu5MnCdVA alloy Harbin University of Technology
[6] Majeed M, Vural M, Raja S and Bilal Naim Shaikh M 2019 Finite element analysis of thermal behavior in maraging steel during SLM process Optik 208 113–24
[7] Khairallah S A, Anderson A T, Rubenchik A and King W E 2016 Laser powder-bed fusion additive manufacturing: physics of complex melt flow and formation mechanisms of pores, spatter, and denudation zones Acta Mater. 108 36–45
[8] Bo C, Zhiyu X, Quanquan Z, Yuanbiao W, Liping W and Jin C 2020 Process optimization and microstructure and properties of SLM forming Cu6AlNiSnInCe imitation gold alloy Chin. J. Nonferr. Met. 30 372–82
[9] Li W 2012 Research on performance of metal parts formed by selective laser melting Huazhong University of Science and Technology
[10] Yu Q 2013 The influence of different laser heat sources on the surface shape of the molten pool in laser cladding Surf. Technol. 42 40–3

[11] Xianfeng J, Xiangchen M, Rongwei S, Xigen Y and Ming Y 2015 Research on the influence of material state change on temperature field
in SLM processing Applied Laser 35 155–9
[12] Körner C, Attar E and Heinl P 2011 Mesoscopic simulation of selective beam melting processesJ. Mater. Process. Technol. 211 978–87
[13] Yadroitsev I, Gusarov A, Yadroitsava I and Smurov I 2010 Single track formation in selective laser melting of metal powdersJ. Mater.
Process. Technol. 210 1624–31
[14] King W, Anderson A T, Ferencz R M, Hodge N E, Kamath C and Khairallah S A 2014 Overview of modelling and simulation of metal
powder bed fusion process at Lawrence Livermore National Laboratory Mater. Sci. Technol. 31 957–68
[15] Hussein A, Hao L, Yan C and Everson R 2013 Finite element simulation of the temperature and stress fields in single layers built
without-support in selective laser melting Materials & Design (1980–2015) 52 638–47
[16] Qiu C, Panwisawas C, Ward M, Basoalto H C, Brooks J W and Attallah M M 2015 On the role of melt flow into the surface structure and
porosity development during selective laser melting Acta Mater. 96 72–9
[17] Weihao Y, Hui C and Qingsong W 2020 Thermodynamic behavior of laser selective melting molten pool under the action of recoil
pressure Journal of Mechanical Engineering 56 213–9
[18] Weijuan Y 2019 Numerical simulation of melt pool temperature field and morphology evolution during laser selective melting process
Xi’an University of Technology
[19] Genwang W 2017 Research on the establishment of laser heat source model based on energy distribution and its simulation application
Harbin Institute of Technology
[20] FLOW-3D 2017 User Manual (USA: FLOW SCIENCE)
[21] Hirt C and Nichols B 1981 Volume of fluid (VOF) method for the dynamics of free boundariesJ. Comput. Phys. 39 201–25
[22] Hu Z, Zhang H, Zhu H, Xiao Z, Nie X and Zeng X 2019 Microstructure, mechanical properties and strengthening mechanisms of
AlCu5MnCdVA aluminum alloy fabricated by selective laser melting Materials Science and Engineering: A 759 154–66
[23] Ketai H, Liu Z and Lechang Y 2020 Simulation of temperature field, microstructure and mechanical properties of 316L stainless steel in
selected laser melting Progress in Laser and Optoelectronics 9 1–18
[24] Cao L 2020 Workpiece-scale numerical simulations of SLM molten pool dynamic behavior of 316L stainless steel Comput. Math. Appl.
4 22–34
[25] Dening Z, Yongping L, Tinglu H and Junyi S 2000 Numerical study of fluid flow and heat transfer in molten pool under the condition of
moving heat source J. Met. 4 387–90
[26] Chengyun C, Cui F and Wenlong Z 2018 The effect of Marangoni flow on the thermal behavior and melt flow behavior of laser cladding
Applied Laser 38 409–16
[27] Peiying B and Enhuai Y 2020 The effect of laser power on the morphology and residual stress of the molten pool of metal laser selective
melting Progress in Laser and Optoelectronics 7 1–12 http://kns.cnki.net/kcms/detail/31.1690.TN.20190717.0933.032.html
[28] Zhen L, Dongyun Z, Zhe F and Chengjie W 2017 Numerical simulation of the influence of overlap rate on the forming quality of
Inconel 718 alloy by selective laser melting processing Applied Laser 37 187–93
[29] Wei W, Qi L, Guang Y, Lanyun Q and Xiong X 2015 Numerical simulation of electromagnetic field, temperature field and flowfield of
laser melting pool under the action of electromagnetic stirring China Laser 42 48–55
[30] Hu Y, He X, Yu G and Zhao S 2016 Capillary convection in pulsed—butt welding of miscible dissimilar couple Proc. Inst. Mech. Eng.
Part C J. Mech. Eng. Sci. 231 2429–40
[31] Li R 2010 Research on the key basic problems of selective laser melting forming of metal powder Huazhong University of Science and
Technology
[32] Zijue T, Weiwei L, Zhaorui Y, Hao W and Hongchao Z 2019 Study on the shape evolution behavior of metal laser melting deposition
based on molten pool dynamic characteristicsJournal of Mechanical Engineering 55 39–47
[33] Pan L, Cheng-Lin Z, Hai-Yi L, Liang W and Tong L 2020 A new two-step selective laser remelting of 316L stainless steel: process,
density, surface roughness, mechanical properties, microstructure Mater. Res. Express 7 056503
[34] Pan L, Cheng-Lin Z, Hai-Yi L, Jiang H, Tong L and Liang W 2019 The influence and optimization of forming process parameters of
316L stainless steel prepared by laser melting on the density Forging Technology 44 103–9

이종 금속 인터커넥트의 펄스 레이저 용접을 위한 가공 매개변수 최적화

Optimization of processing parameters for pulsed laser welding of dissimilar metal interconnects

본 논문은 독자의 편의를 위해 기계번역된 내용이어서 자세한 내용은 원문을 참고하시기 바랍니다.

NguyenThi TienaYu-LungLoabM.Mohsin RazaaCheng-YenChencChi-PinChiuc

aNational Cheng Kung University, Department of Mechanical Engineering, Tainan, Taiwan

bNational Cheng Kung University, Academy of Innovative Semiconductor and Sustainable Manufacturing, Tainan, Taiwan

cJum-bo Co., Ltd, Xinshi District, Tainan, Taiwan

Abstract

워블 전략이 포함된 펄스 레이저 용접(PLW) 방법을 사용하여 알루미늄 및 구리 이종 랩 조인트의 제조를 위한 최적의 가공 매개변수에 대해 실험 및 수치 조사가 수행됩니다. 피크 레이저 출력과 접선 용접 속도의 대표적인 조합 43개를 선택하기 위해 원형 패킹 설계 알고리즘이 먼저 사용됩니다.

선택한 매개변수는 PLW 프로세스의 전산유체역학(CFD) 모델에 제공되어 용융 풀 형상(즉, 인터페이스 폭 및 침투 깊이) 및 구리 농도를 예측합니다. 시뮬레이션 결과는 설계 공간 내에서 PLW 매개변수의 모든 조합에 대한 용융 풀 형상 및 구리 농도를 예측하기 위해 3개의 대리 모델을 교육하는 데 사용됩니다.

마지막으로, 대체 모델을 사용하여 구성된 처리 맵은 용융 영역에 균열이나 기공이 없고 향상된 기계적 및 전기적 특성이 있는 이종 조인트를 생성하는 PLW 매개변수를 결정하기 위해 세 가지 품질 기준에 따라 필터링됩니다.

제안된 최적화 접근법의 타당성은 최적의 용접 매개변수를 사용하여 생성된 실험 샘플의 전단 강도, 금속간 화합물(IMC) 형성 및 전기 접촉 저항을 평가하여 입증됩니다.

결과는 최적의 매개변수가 1209N의 높은 전단 강도와 86µΩ의 낮은 전기 접촉 저항을 생성함을 확인합니다. 또한 용융 영역에는 균열 및 기공과 같은 결함이 없습니다.

An experimental and numerical investigation is performed into the optimal processing parameters for the fabrication of aluminum and copper dissimilar lap joints using a pulsed laser welding (PLW) method with a wobble strategy. A circle packing design algorithm is first employed to select 43 representative combinations of the peak laser power and tangential welding speed. The selected parameters are then supplied to a computational fluidic dynamics (CFD) model of the PLW process to predict the melt pool geometry (i.e., interface width and penetration depth) and copper concentration. The simulation results are used to train three surrogate models to predict the melt pool geometry and copper concentration for any combination of the PLW parameters within the design space. Finally, the processing maps constructed using the surrogate models are filtered in accordance with three quality criteria to determine the PLW parameters that produce dissimilar joints with no cracks or pores in the fusion zone and enhanced mechanical and electrical properties. The validity of the proposed optimization approach is demonstrated by evaluating the shear strength, intermetallic compound (IMC) formation, and electrical contact resistance of experimental samples produced using the optimal welding parameters. The results confirm that the optimal parameters yield a high shear strength of 1209 N and a low electrical contact resistance of 86 µΩ. Moreover, the fusion zone is free of defects, such as cracks and pores.

Fig. 1. Schematic illustration of Al-Cu lap-joint arrangement
Fig. 1. Schematic illustration of Al-Cu lap-joint arrangement
Fig. 2. Machine setup (MFQS-150W_1500W
Fig. 2. Machine setup (MFQS-150W_1500W
Fig. 5. Lap-shear mechanical tests: (a) experimental setup and specimen dimensions, and (b) two different failures of lap-joint welding.
N. Thi Tien et al.
Fig. 5. Lap-shear mechanical tests: (a) experimental setup and specimen dimensions, and (b) two different failures of lap-joint welding. N. Thi Tien et al.
Fig. 9. Simulation and experimental results for melt pool profile. (a) Simulation results for melt pool cross-section, and (b) OM image of melt pool cross-section.
(Note that laser processing parameter of 830 W and 565 mm/s is chosen.).
Fig. 9. Simulation and experimental results for melt pool profile. (a) Simulation results for melt pool cross-section, and (b) OM image of melt pool cross-section. (Note that laser processing parameter of 830 W and 565 mm/s is chosen.).

References

[1]

G. Santos

Road transport and CO2 emissions: What are the challenges?

Transport Policy, 59 (2017), pp. 71-74

ArticleDownload PDFView Record in ScopusGoogle Scholar[2]

A. Das, D. Li, D. Williams, D. Greenwood

Joining technologies for automotive battery systems manufacturing

World Electric Veh. J., 9 (2) (2018), p. 22 View PDF

CrossRefGoogle Scholar[3]

M. Zwicker, M. Moghadam, W. Zhang, C. Nielsen

Automotive battery pack manufacturing–a review of battery to tab joining

J. Adv. Joining Process., 1 (2020), Article 100017

ArticleDownload PDFView Record in ScopusGoogle Scholar[4]

T. Mai, A. Spowage

Characterisation of dissimilar joints in laser welding of steel–kovar, copper–steel and copper–aluminium

Mater. Sci. Eng. A, 374 (1–2) (2004), pp. 224-233

ArticleDownload PDFView Record in ScopusGoogle Scholar[5]

S.S. Lee, T.H. Kim, S.J. Hu, W.W. Cai, J. Li, J.A. Abell

Characterization of joint quality in ultrasonic welding of battery tabs

International Manufacturing Science and Engineering Conference, vol. 54990, American Society of Mechanical Engineers (2012), pp. 249-261

Google Scholar[6]

Y. Zhou, P. Gorman, W. Tan, K. Ely

Weldability of thin sheet metals during small-scale resistance spot welding using an alternating-current power supply

J. Electron. Mater., 29 (9) (2000), pp. 1090-1099 View PDF

CrossRefView Record in ScopusGoogle Scholar[7]

S. Katayama

Handbook of laser welding technologies

Elsevier (2013)

Google Scholar[8]

A. Sadeghian, N. Iqbal

A review on dissimilar laser welding of steel-copper, steel-aluminum, aluminum-copper, and steel-nickel for electric vehicle battery manufacturing

Opt. Laser Technol., 146 (2022), Article 107595

ArticleDownload PDFView Record in ScopusGoogle Scholar[9]

M.J. Brand, P.A. Schmidt, M.F. Zaeh, A. Jossen

Welding techniques for battery cells and resulting electrical contact resistances

J. Storage Mater., 1 (2015), pp. 7-14

ArticleDownload PDFView Record in ScopusGoogle Scholar[10]

M. Jarwitz, F. Fetzer, R. Weber, T. Graf

Weld seam geometry and electrical resistance of laser-welded, aluminum-copper dissimilar joints produced with spatial beam oscillation

Metals, 8 (7) (2018), p. 510 View PDF

CrossRefView Record in ScopusGoogle Scholar[11]S. Smith, J. Blackburn, M. Gittos, P. de Bono, and P. Hilton, “Welding of dissimilar metallic materials using a scanned laser beam,” in International Congress on Applications of Lasers & Electro-Optics, 2013, vol. 2013, no. 1: Laser Institute of America, pp. 493-502.

Google Scholar[12]

P. Schmitz, J.B. Habedank, M.F. Zaeh

Spike laser welding for the electrical connection of cylindrical lithium-ion batteries

J. Laser Appl., 30 (1) (2018), Article 012004 View PDF

CrossRefView Record in ScopusGoogle Scholar[13]

P. Kah, C. Vimalraj, J. Martikainen, R. Suoranta

Factors influencing Al-Cu weld properties by intermetallic compound formation

Int. J. Mech. Mater. Eng., 10 (1) (2015), pp. 1-13

Google Scholar[14]

Z. Lei, X. Zhang, J. Liu, P. Li

Interfacial microstructure and reaction mechanism with various weld fillers on laser welding-brazing of Al/Cu lap joint

J. Manuf. Process., 67 (2021), pp. 226-240

ArticleDownload PDFView Record in ScopusGoogle Scholar[15]

T. Solchenbach, P. Plapper

Mechanical characteristics of laser braze-welded aluminium–copper connections

Opt. Laser Technol., 54 (2013), pp. 249-256

ArticleDownload PDFView Record in ScopusGoogle Scholar[16]

T. Solchenbach, P. Plapper, W. Cai

Electrical performance of laser braze-welded aluminum–copper interconnects

J. Manuf. Process., 16 (2) (2014), pp. 183-189

ArticleDownload PDFView Record in ScopusGoogle Scholar[17]

S.J. Lee, H. Nakamura, Y. Kawahito, S. Katayama

Effect of welding speed on microstructural and mechanical properties of laser lap weld joints in dissimilar Al and Cu sheets

Sci. Technol. Weld. Join., 19 (2) (2014), pp. 111-118

Google Scholar[18]

Z. Xue, S. Hu, D. Zuo, W. Cai, D. Lee, K.-A. Elijah Jr

Molten pool characterization of laser lap welded copper and aluminum

J. Phys. D Appl. Phys., 46 (49) (2013), Article 495501 View PDF

CrossRefView Record in ScopusGoogle Scholar[19]

S. Zhao, G. Yu, X. He, Y. Zhang, W. Ning

Numerical simulation and experimental investigation of laser overlap welding of Ti6Al4V and 42CrMo

J. Mater. Process. Technol., 211 (3) (2011), pp. 530-537

ArticleDownload PDFView Record in ScopusGoogle Scholar[20]

W. Huang, H. Wang, T. Rinker, W. Tan

Investigation of metal mixing in laser keyhole welding of dissimilar metals

Mater. Des., 195 (2020), Article 109056

ArticleDownload PDFView Record in ScopusGoogle Scholar[21]

E. Kaiser, G. Ambrosy, E. Papastathopoulos

Welding strategies for joining copper and aluminum by fast oscillating, high quality laser beam

High-Power Laser Materials Processing: Applications, Diagnostics, and Systems IX, vol. 11273, International Society for Optics and Photonics (2020), p. 112730C

View Record in ScopusGoogle Scholar[22]

V. Dimatteo, A. Ascari, A. Fortunato

Continuous laser welding with spatial beam oscillation of dissimilar thin sheet materials (Al-Cu and Cu-Al): Process optimization and characterization

J. Manuf. Process., 44 (2019), pp. 158-165

ArticleDownload PDFView Record in ScopusGoogle Scholar[23]

V. Dimatteo, A. Ascari, E. Liverani, A. Fortunato

Experimental investigation on the effect of spot diameter on continuous-wave laser welding of copper and aluminum thin sheets for battery manufacturing

Opt. Laser Technol., 145 (2022), Article 107495

ArticleDownload PDFView Record in ScopusGoogle Scholar[24]

D. Wu, X. Hua, F. Li, L. Huang

Understanding of spatter formation in fiber laser welding of 5083 aluminum alloy

Int. J. Heat Mass Transf., 113 (2017), pp. 730-740

ArticleDownload PDFView Record in ScopusGoogle Scholar[25]

R. Ducharme, K. Williams, P. Kapadia, J. Dowden, B. Steen, M. Glowacki

The laser welding of thin metal sheets: an integrated keyhole and weld pool model with supporting experiments

J. Phys. D Appl. Phys., 27 (8) (1994), p. 1619 View PDF

CrossRefView Record in ScopusGoogle Scholar[26]

C.W. Hirt, B.D. Nichols

Volume of fluid (VOF) method for the dynamics of free boundaries

J. Comput. Phys., 39 (1) (1981), pp. 201-225

ArticleDownload PDFGoogle Scholar[27]

W. Piekarska, M. Kubiak

Three-dimensional model for numerical analysis of thermal phenomena in laser–arc hybrid welding process

Int. J. Heat Mass Transf., 54 (23–24) (2011), pp. 4966-4974

ArticleDownload PDFView Record in ScopusGoogle Scholar[28]J. Zhou, H.-L. Tsai, and P.-C. Wang, “Transport phenomena and keyhole dynamics during pulsed laser welding,” 2006.

Google Scholar[29]

D. Harrison, D. Yan, S. Blairs

The surface tension of liquid copper

J. Chem. Thermodyn., 9 (12) (1977), pp. 1111-1119

ArticleDownload PDFView Record in ScopusGoogle Scholar[30]

M. Leitner, T. Leitner, A. Schmon, K. Aziz, G. Pottlacher

Thermophysical properties of liquid aluminum

Metall. Mater. Trans. A, 48 (6) (2017), pp. 3036-3045 View PDF

This article is free to access.

CrossRefView Record in ScopusGoogle Scholar[31]

H.-C. Tran, Y.-L. Lo

Systematic approach for determining optimal processing parameters to produce parts with high density in selective laser melting process

Int. J. Adv. Manuf. Technol., 105 (10) (2019), pp. 4443-4460 View PDF

CrossRefView Record in ScopusGoogle Scholar[32]A. Ascari, A. Fortunato, E. Liverani, and A. Lutey, “Application of different pulsed laser sources to dissimilar welding of Cu and Al alloys,” in Proceedings of Lasers in Manufacturing Conference (LIM), 2019.

Google Scholar[33]

A. Fortunato, A. Ascari

Laser welding of thin copper and aluminum sheets: feasibility and challenges in continuous-wave welding of dissimilar metals

Lasers in Manufacturing and Materials Processing, 6 (2) (2019), pp. 136-157 View PDF

CrossRefView Record in ScopusGoogle Scholar[34]

A. Boucherit, M.-N. Avettand-Fènoël, R. Taillard

Effect of a Zn interlayer on dissimilar FSSW of Al and Cu

Mater. Des., 124 (2017), pp. 87-99

ArticleDownload PDFView Record in ScopusGoogle Scholar[35]

N. Kumar, I. Masters, A. Das

In-depth evaluation of laser-welded similar and dissimilar material tab-to-busbar electrical interconnects for electric vehicle battery pack

J. Manuf. Process., 70 (2021), pp. 78-96

ArticleDownload PDFView Record in ScopusGoogle Scholar[36]

M. Abbasi, A.K. Taheri, M. Salehi

Growth rate of intermetallic compounds in Al/Cu bimetal produced by cold roll welding process

J. Alloy. Compd., 319 (1–2) (2001), pp. 233-241

ArticleDownload PDFGoogle Scholar[37]

D. Zuo, S. Hu, J. Shen, Z. Xue

Intermediate layer characterization and fracture behavior of laser-welded copper/aluminum metal joints

Mater. Des., 58 (2014), pp. 357-362

ArticleDownload PDFView Record in ScopusGoogle Scholar[38]

S. Yan, Y. Shi

Influence of Ni interlayer on microstructure and mechanical properties of laser welded joint of Al/Cu bimetal

J. Manuf. Process., 59 (2020), pp. 343-354

ArticleDownload PDFView Record in ScopusGoogle Scholar[39]

S. Yan, Y. Shi

Influence of laser power on microstructure and mechanical property of laser-welded Al/Cu dissimilar lap joints

J. Manuf. Process., 45 (2019), pp. 312-321

ArticleDownload PDFView Record in ScopusGoogle Scholar

Fig. 8 Distribution of solidification properties on the yz cross section at the maximum width of the melt pool.(a) thermal gradient G, (b) solidification velocity vT, (c) cooling rate G×vT, and (d) morphology factor G/vT. These profiles are calculated with a laser power 300 W and velocity 400 mm/s using (a1 through d1) analytical Rosenthal simulation and (a2 through d2) high-fidelity CFD simulation. The laser is moving out of the page from the upper left corner of each color map (Color figure online)

Quantifying Equiaxed vs Epitaxial Solidification in Laser Melting of CMSX-4 Single Crystal Superalloy

CMSX -4 단결정 초합금의 레이저 용융에서 등축 응고와 에피택셜 응고 정량화

본 논문은 독자의 편의를 위해 기계번역된 내용이어서 자세한 내용은 원문을 참고하시기 바랍니다.

Abstract

에피택셜 과 등축 응고 사이의 경쟁은 적층 제조에서 실행되는 레이저 용융 동안 CMSX-4 단결정 초합금에서 조사되었습니다. 단일 트랙 레이저 스캔은 레이저 출력과 스캐닝 속도의 여러 조합으로 방향성 응고된 CMSX-4 합금의 분말 없는 표면에서 수행되었습니다. EBSD(Electron Backscattered Diffraction) 매핑은 새로운 방향의 식별을 용이하게 합니다. 영역 분율 및 공간 분포와 함께 융합 영역 내에서 핵을 형성한 “스트레이 그레인”은 충실도가 높은 전산 유체 역학 시뮬레이션을 사용하여 용융 풀 내의 온도 및 유체 속도 필드를 모두 추정했습니다. 이 정보를 핵 생성 모델과 결합하여 용융 풀에서 핵 생성이 발생할 확률이 가장 높은 위치를 결정했습니다. 금속 적층 가공의 일반적인 경험에 따라 레이저 용융 트랙의 응고된 미세 구조는 에피택셜 입자 성장에 의해 지배됩니다. 더 높은 레이저 스캐닝 속도와 더 낮은 출력이 일반적으로 흩어진 입자 감소에 도움이 되지만,그럼에도 불구하고 길쭉한 용융 풀에서 흩어진 입자가 분명했습니다.

The competition between epitaxial vs. equiaxed solidification has been investigated in CMSX-4 single crystal superalloy during laser melting as practiced in additive manufacturing. Single-track laser scans were performed on a powder-free surface of directionally solidified CMSX-4 alloy with several combinations of laser power and scanning velocity. Electron backscattered diffraction (EBSD) mapping facilitated identification of new orientations, i.e., “stray grains” that nucleated within the fusion zone along with their area fraction and spatial distribution. Using high-fidelity computational fluid dynamics simulations, both the temperature and fluid velocity fields within the melt pool were estimated. This information was combined with a nucleation model to determine locations where nucleation has the highest probability to occur in melt pools. In conformance with general experience in metals additive manufacturing, the as-solidified microstructure of the laser-melted tracks is dominated by epitaxial grain growth; nevertheless, stray grains were evident in elongated melt pools. It was found that, though a higher laser scanning velocity and lower power are generally helpful in the reduction of stray grains, the combination of a stable keyhole and minimal fluid velocity further mitigates stray grains in laser single tracks.

Introduction

니켈 기반 초합금은 고온에서 긴 노출 시간 동안 높은 인장 강도, 낮은 산화 및 우수한 크리프 저항성을 포함하는 우수한 특성의 고유한 조합으로 인해 가스 터빈 엔진 응용 분야에서 광범위하게 사용됩니다. CMSX-4는 특히 장기 크리프 거동과 관련하여 초고강도의 2세대 레늄 함유 니켈 기반 단결정 초합금입니다. 1 , 2 ]입계의 존재가 크리프를 가속화한다는 인식은 가스 터빈 엔진의 고온 단계를 위한 단결정 블레이드를 개발하게 하여 작동 온도를 높이고 효율을 높이는 데 기여했습니다. 이러한 구성 요소는 사용 중 마모될 수 있습니다. 즉, 구성 요소의 무결성을 복원하고 단결정 미세 구조를 유지하는 수리 방법을 개발하기 위한 지속적인 작업이 있었습니다. 3 , 4 , 5 ]

적층 제조(AM)가 등장하기 전에는 다양한 용접 공정을 통해 단결정 초합금에 대한 수리 시도가 수행되었습니다. 균열 [ 6 , 7 ] 및 흩어진 입자 8 , 9 ] 와 같은 심각한 결함 이 이 수리 중에 자주 발생합니다. 일반적으로 “스트레이 그레인”이라고 하는 응고 중 모재의 방향과 다른 결정학적 방향을 가진 새로운 그레인의 형성은 니켈 기반 단결정 초합금의 수리 중 유해한 영향으로 인해 중요한 관심 대상입니다. 3 , 10 ]결과적으로 재료의 단결정 구조가 손실되고 원래 구성 요소에 비해 기계적 특성이 손상됩니다. 이러한 흩어진 입자는 특정 조건에서 에피택셜 성장을 대체하는 등축 응고의 시작에 해당합니다.

떠돌이 결정립 형성을 완화하기 위해 이전 작업은 용융 영역(FZ) 내에서 응고하는 동안 떠돌이 결정립 형성에 영향을 미치는 수지상 응고 거동 및 처리 조건을 이해하는 데 중점을 두었습니다. 11 , 12 , 13 , 14 ] 연구원들은 단결정 합금의 용접 중에 표류 결정립 형성에 대한 몇 가지 가능한 메커니즘을 제안했습니다. 12 , 13 , 14 , 15 ]응고 전단에 앞서 국부적인 구성 과냉각은 이질적인 핵 생성 및 등축 결정립의 성장을 유발할 수 있습니다. 또한 용융 풀에서 활발한 유체 흐름으로 인해 발생하는 덴드라이트 조각화는 용융 풀 경계 근처에서 새로운 결정립을 형성할 수도 있습니다. 두 메커니즘 모두에서, 표류 결정립 형성은 핵 생성 위치에 의존하며, 차이점은 수상 돌기 조각화는 수상 돌기 조각이 핵 생성 위치로 작용한다는 것을 의미하는 반면 다른 메커니즘은 재료,  를 들어 산화물 입자에서 발견되는 다른 유형의 핵 생성 위치를 사용한다는 것을 의미합니다. 잘 알려진 바와 같이, 많은 주물에 대한 반대 접근법은 TiB와 같은 핵제의 도입을 통해 등축 응고를 촉진하는 것입니다.22알루미늄 합금에서.

헌법적 과냉 메커니즘에서 Hunt 11 ] 는 정상 상태 조건에서 기둥에서 등축으로의 전이(CET)를 설명하는 모델을 개발했습니다. Gaumann과 Kurz는 Hunt의 모델을 수정하여 단결정이 응고되는 동안 떠돌이 결정립이 핵을 생성하고 성장할 수 있는 정도를 설명했습니다. 12 , 14 ] 이후 연구에서 Vitek은 Gaumann의 모델을 개선하고 출력 및 스캐닝 속도와 같은 용접 조건의 영향에 대한 보다 자세한 분석을 포함했습니다. Vitek은 또한 실험 및 모델링 기술을 통해 표류 입자 형성에 대한 기판 방향의 영향을 포함했습니다. 3 , 10 ]일반적으로 높은 용접 속도와 낮은 출력은 표류 입자의 양을 최소화하고 레이저 용접 공정 중 에피택셜 단결정 성장을 최대화하는 것으로 나타났습니다. 3,10 ] 그러나 Vitek은 덴드라이트 조각화를 고려하지 않았으며 그의 연구는 불균질 핵형성이 레이저 용접된 CMSX -4 단결정 합금에서 표류 결정립 형성을 이끄는 주요 메커니즘임을 나타냅니다. 현재 작업에서 Vitek의 수치적 방법이 채택되고 금속 AM의 급속한 특성의 더 높은 속도와 더 낮은 전력 특성으로 확장됩니다.

AM을 통한 금속 부품 제조 는 지난 10년 동안 급격한 인기 증가를 목격했습니다. 16 ] EBM(Electron Beam Melting)에 의한 CMSX-4의 제작 가능성은 자주 조사되었으나 17 , 18 , 19 , 20 , 21 ] CMSX의 제조 및 수리에 대한 조사는 매우 제한적이었다. – 4개의 단결정 구성요소는 레이저 분말 베드 융합(LPBF)을 사용하며, AM의 인기 있는 하위 집합으로, 특히 표류 입자 형성을 완화하는 메커니즘과 관련이 있습니다. 22 ]이러한 조사 부족은 주로 이러한 합금 시스템과 관련된 처리 문제로 인해 발생합니다. 2 , 19 , 22 , 23 , 24 ] 공정 매개변수( 예: 열원 전력, 스캐닝 속도, 스폿 크기, 예열 온도 및 스캔 전략)의 엄격한 제어는 완전히 조밀한 부품을 만들고 유지 관리할 수 있도록 하는 데 필수적입니다. 단결정 미세구조. 25 ] EBM을 사용하여 단결정 합금의 균열 없는 수리가 현재 가능하지만 19 , 24 ] 표류 입자를 생성하지 않는 수리는 쉽게 달성할 수 없습니다.23 , 26 ]

이 작업에서 LPBF를 대표하는 조건으로 레이저 용융을 사용하여 단결정 CMSX-4에서 표류 입자 완화를 조사했습니다. LPBF는 스캐닝 레이저 빔을 사용하여 금속 분말의 얇은 층을 기판에 녹이고 융합합니다. 층별 증착에서 레이저 빔의 사용은 급격한 온도 구배, 빠른 가열/냉각 주기 및 격렬한 유체 흐름을 경험하는 용융 풀을 생성 합니다 이것은 일반적으로 부품에 결함을 일으킬 수 있는 매우 동적인 물리적 현상으로 이어집니다. 28 , 29 , 30 ] 레이저 유도 키홀의 동역학( 예:, 기화 유발 반동 압력으로 인한 위상 함몰) 및 열유체 흐름은 AM 공정에서 응고 결함과 강하게 결합되고 관련됩니다. 31 , 32 , 33 , 34 ] 기하 구조의 급격한 변화가 발생하기 쉬운 불안정한 키홀은 다공성, 볼링, 스패터 형성 및 흔하지 않은 미세 구조 상을 포함하는 유해한 물리적 결함을 유발할 수 있습니다. 그러나 키홀 진화와 유체 흐름은 자연적으로 다음을 통해 포착 하기 어렵 습니다 .전통적인 사후 특성화 기술. 고충실도 수치 모델링을 활용하기 위해 이 연구에서는 전산유체역학(CFD)을 적용하여 표면 아래의 레이저-물질 상호 작용을 명확히 했습니다. 36 ] 이것은 응고된 용융물 풀의 단면에 대한 오랫동안 확립된 사후 특성화와 비교하여 키홀 및 용융물 풀 유체 흐름 정량화를 실행합니다.

CMSX-4 구성 요소의 레이저 기반 AM 수리 및 제조를 위한 적절한 절차를 개발하기 위해 적절한 공정 창을 설정하고 응고 중 표류 입자 형성 경향에 대한 예측 기능을 개발하는 것부터 시작합니다. 다중 합금에 대한 단일 트랙 증착은 분말 층이 있거나 없는 AM 공정에서 용융 풀 형상 및 미세 구조의 정확한 분석을 제공하는 것으로 나타났습니다. 37 , 38 , 39 ]따라서 본 연구에서는 CMSX-4의 응고 거동을 알아보기 위해 분말을 사용하지 않는 단일 트랙 레이저 스캔 실험을 사용하였다. 이는 CMSX-4 단결정의 LPBF 제조를 위한 예비 실험 지침을 제공합니다. 또한 응고 모델링은 기존 용접에서 LPBF와 관련된 급속 용접으로 확장되어 표류 입자 감소를 위한 최적의 레이저 용융 조건을 식별했습니다. 가공 매개변수 최적화를 위한 추가 지침을 제공하기 위해 용융물 풀의 매우 동적인 유체 흐름을 모델링했습니다.

재료 및 방법

단일 트랙 실험

방전 가공(EDM)을 사용하여 CMSX-4 방향성 응고 단결정 잉곳으로부터 샘플을 제작했습니다. 샘플의 최종 기하학은 치수 20의 직육면체 형태였습니다.××20××6mm. 6개 중 하나⟨ 001 ⟩⟨001⟩잉곳의 결정학적 방향은 레이저 트랙이 이 바람직한 성장 방향을 따라 스캔되도록 절단 표면에 수직으로 위치했습니다. 단일 레이저 용융 트랙은 EOS M290 기계를 사용하여 분말이 없는 샘플 표면에 만들어졌습니다. 이 기계는 최대 출력 400W, 가우시안 빔 직경 100의 이터븀 파이버 레이저가 장착된 LPBF 시스템입니다. μμ초점에서 m. 실험 중에 직사각형 샘플을 LPBF 기계용 맞춤형 샘플 홀더의 포켓에 끼워 표면을 동일한 높이로 유지했습니다. 이 맞춤형 샘플 홀더에 대한 자세한 내용은 다른 곳에서 설명합니다. 실험 은 아르곤 퍼지 분위기에서 수행되었으며 예열은 적용되지 않았습니다 단일 트랙 레이저 용융 실험은 다양한 레이저 출력(200~370W)과 스캔 속도(0.4~1.4m/s)에서 수행되었습니다.

성격 묘사

레이저 스캐닝 후, 레이저 빔 스캐닝 방향에 수직인 평면에서 FZ를 통해 다이아몬드 톱을 사용하여 샘플을 절단했습니다. 그 후, 샘플을 장착하고 220 그릿 SiC 페이퍼로 시작하여 콜로이드 실리카 현탁액 광택제로 마무리하여 자동 연마했습니다. 결정학적 특성화는 20kV의 가속 전압에서 TESCAN MIRA 3XMH 전계 방출 주사 전자 현미경(SEM)에서 수행되었습니다. EBSD 지도는0.4μm _0.4μ미디엄단계 크기. Bruker 시스템을 사용하여 EBSD 데이터를 정리하고 분석했습니다. EBSD 클린업은 그레인을 접촉시키기 위한 그레인 확장 루틴으로 시작한 다음 인덱스되지 않은 회절 패턴과 관련된 검은색 픽셀을 해결하기 위해 이웃 방향 클린업 루틴으로 이어졌습니다. 용융 풀 형태를 분석하기 위해 단면을 광학 현미경으로 분석했습니다. 광학 특성화의 대비를 향상시키기 위해 10g CuSO로 구성된 Marbles 시약의 변형으로 샘플을 에칭했습니다.44, 50mL HCl 및 70mL H22영형.

응고 모델링

구조적 과냉 기준에 기반한 응고 모델링을 수행하여 표유 입자의 성향 및 분포에 대한 가공 매개변수의 영향을 평가했습니다. 이 분석 모델링 접근 방식에 대한 자세한 내용은 이전 작업에서 제공됩니다. 3 , 10 ] 참고문헌 3 에 기술된 바와 같이 , 기본 재료의 결정학적 배향을 가진 용융 풀에서 총 표유 입자 면적 분율의 변화는 최소이므로 기본 재료 배향의 영향은 이 작업에서 고려되지 않았습니다. 우리의 LPBF 결과를 이전 작업과 비교하기 위해 Vitek의 작업에서 사용된 수학적으로 간단한 Rosenthal 방정식 3 ]또한 레이저 매개변수의 함수로 용융 풀의 모양과 FZ의 열 조건을 계산하기 위한 기준으로 여기에서 채택되었습니다. Rosenthal 솔루션은 열이 일정한 재료 특성을 가진 반무한 판의 정상 상태 점원을 통해서만 전도를 통해 전달된다고 가정하며 일반적으로 다음과 같이 표현 됩니다 40 , 41 ] .

티=티0+η피2 파이케이엑스2+와이2+지2———-√경험치[- 브이(엑스2+와이2+지2———-√− 엑스 )2α _] ,티=티0+η피2파이케이엑스2+와이2+지2경험치⁡[-V(엑스2+와이2+지2-엑스)2α],(1)

여기서 T 는 온도,티0티0본 연구에서 313K(  , EOS 기계 챔버 온도)로 설정된 주변 온도, P 는 레이저 빔 파워, V 는 레이저 빔 스캐닝 속도,ηη는 레이저 흡수율, k 는 열전도율,αα베이스 합금의 열확산율입니다. x , y , z 는 각각 레이저 스캐닝 방향, 가로 방향 및 세로 방향의 반대 방향과 정렬된 방향입니다 . 이 직교 좌표는 참조 3 의 그림 1에 있는 시스템을 따랐습니다 . CMSX-4에 대한 고상선 온도(1603K)와 액상선 온도(1669K)의 등온선 평균으로 응고 프런트( 즉 , 고체-액체 계면)를 정의했습니다. 42 , 43 , 44 ] 시뮬레이션에 사용된 열물리적 특성은 표 I 에 나열되어 있습니다.표 I CMSX-4의 응고 모델링에 사용된 열물리적 특성

풀 사이즈 테이블

열 구배는 외부 열 흐름에 의해 결정되었습니다.∇ 티∇티45 ] 에 의해 주어진 바와 같이 :

지 = | ∇ 티| =∣∣∣∂티∂엑스나^^+∂티∂와이제이^^+∂티∂지케이^^∣∣∣=(∂티∂엑스)2+(∂티∂와이)2+(∂티∂지)2————————√,G=|∇티|=|∂티∂엑스나^^+∂티∂와이제이^^+∂티∂지케이^^|=(∂티∂엑스)2+(∂티∂와이)2+(∂티∂지)2,(2)

어디나^^나^^,제이^^제이^^, 그리고케이^^케이^^는 각각 x , y 및 z 방향 을 따른 단위 벡터 입니다. 응고 등온선 속도,V티V티는 다음 관계에 의해 레이저 빔 스캐닝 속도 V 와 기하학적으로 관련됩니다.

V티= V코사인θ =V∂티∂엑스(∂티∂엑스)2+(∂티∂와이)2+(∂티∂지)2——————-√,V티=V코사인⁡θ=V∂티∂엑스(∂티∂엑스)2+(∂티∂와이)2+(∂티∂지)2,(삼)

어디θθ는 스캔 방향과 응고 전면의 법선 방향(  , 최대 열 흐름 방향) 사이의 각도입니다. 이 연구의 용접 조건과 같은 제한된 성장에서 수지상 응고 전면은 고체-액체 등온선의 속도로 성장하도록 강제됩니다.V티V티. 46 ]

응고 전선이 진행되기 전에 새로 핵 생성된 입자의 국지적 비율ΦΦ, 액체 온도 구배 G 에 의해 결정 , 응고 선단 속도V티V티및 핵 밀도N0N0. 고정된 임계 과냉각에서 모든 입자가 핵형성된다고 가정함으로써△티N△티N, 등축 결정립의 반경은 결정립이 핵 생성을 시작하는 시점부터 주상 전선이 결정립에 도달하는 시간까지의 성장 속도를 통합하여 얻습니다. 과냉각으로 대체 시간d (ΔT_) / dt = – _V티G디(△티)/디티=-V티G, 열 구배 G 사이의 다음 관계 , 등축 입자의 국부적 부피 분율ΦΦ, 수상 돌기 팁 과냉각ΔT _△티, 핵 밀도N0N0, 재료 매개변수 n 및 핵생성 과냉각△티N△티N, Gäumann 외 여러분 에 의해 파생되었습니다 . 12 , 14 ] Hunt의 모델 11 ] 의 수정에 기반함 :

지 =1엔 + 1- 4π _N03 인치( 1 − Φ )———√삼ΔT _( 1 -△티엔 + 1N△티엔 + 1) .G=1N+1-4파이N0삼인⁡(1-Φ)삼△티(1-△티NN+1△티N+1).(4)

계산을 단순화하기 위해 덴드라이트 팁 과냉각을 전적으로 구성 과냉각의 것으로 추정합니다.△티씨△티씨, 멱법칙 형식으로 근사화할 수 있습니다.△티씨= ( _V티)1 / 엔△티씨=(ㅏV티)1/N, 여기서 a 와 n 은 재료 종속 상수입니다. CMSX-4의 경우 이 값은a = 1.25 ×106ㅏ=1.25×106 s K 3.4m− 1-1,엔 = 3.4N=3.4, 그리고N0= 2 ×1015N0=2×1015미디엄− 3,-삼,참고문헌 3 에 의해 보고된 바와 같이 .△티N△티N2.5K이며 보다 큰 냉각 속도에서 응고에 대해 무시할 수 있습니다.106106 K/s. 에 대한 표현ΦΦ위의 방정식을 재배열하여 해결됩니다.

Φ= 1 -이자형에스\ 여기서\  S=- 4π _N0삼(1( 엔 + 1 ) (GN/ 아V티)1 / 엔)삼=−2.356×1019(vTG3.4)33.4.Φ=1−eS\ where\ S=−4πN03(1(n+1)(Gn/avT)1/n)3=−2.356×1019(vTG3.4)33.4.

(5)

As proposed by Hunt,[11] a value of Φ≤0.66Φ≤0.66 pct represents fully columnar epitaxial growth condition, and, conversely, a value of Φ≥49Φ≥49 pct indicates that the initial single crystal microstructure is fully replaced by an equiaxed microstructure. To calculate the overall stray grain area fraction, we followed Vitek’s method by dividing the FZ into roughly 19 to 28 discrete parts (depending on the length of the melt pool) of equal length from the point of maximum width to the end of melt pool along the x direction. The values of G and vTvT were determined at the center on the melt pool boundary of each section and these values were used to represent the entire section. The area-weighted average of ΦΦ over these discrete sections along the length of melt pool is designated as Φ¯¯¯¯Φ¯, and is given by:

Φ¯¯¯¯=∑kAkΦk∑kAk,Φ¯=∑kAkΦk∑kAk,

(6)

where k is the index for each subsection, and AkAk and ΦkΦk are the areas and ΦΦ values for each subsection. The summation is taken over all the sections along the melt pool. Vitek’s improved model allows the calculation of stray grain area fraction by considering the melt pool geometry and variations of G and vTvT around the tail end of the pool.

수년에 걸쳐 용융 풀 현상 모델링의 정확도를 개선하기 위해 많은 고급 수치 방법이 개발되었습니다. 우리는 FLOW-3D와 함께 고충실도 CFD를 사용했습니다. FLOW-3D는 여러 물리 모델을 통합하는 상용 FVM(Finite Volume Method)입니다. 47 , 48 ] CFD는 유체 운동과 열 전달을 수치적으로 시뮬레이션하며 여기서 사용된 기본 물리 모델은 레이저 및 표면력 모델이었습니다. 레이저 모델에서는 레이 트레이싱 기법을 통해 다중 반사와 프레넬 흡수를 구현합니다. 36 ]먼저, 레이저 빔은 레이저 빔에 의해 조명되는 각 그리드 셀을 기준으로 여러 개의 광선으로 이산화됩니다. 그런 다음 각 입사 광선에 대해 입사 벡터가 입사 위치에서 금속 표면의 법선 벡터와 정렬될 때 에너지의 일부가 금속에 의해 흡수됩니다. 흡수율은 Fresnel 방정식을 사용하여 추정됩니다. 나머지 에너지는 반사광선 에 의해 유지되며 , 반사광선은 재료 표면에 부딪히면 새로운 입사광선으로 처리됩니다. 두 가지 주요 힘이 액체 금속 표면에 작용하여 자유 표면을 변형시킵니다. 금속의 증발에 의해 생성된 반동 압력은 증기 억제를 일으키는 주요 힘입니다. 본 연구에서 사용된 반동 압력 모델은피아르 자형= 특급 _{ B ( 1- _티V/ 티) }피아르 자형=ㅏ경험치⁡{비(1-티V/티)}, 어디피아르 자형피아르 자형는 반동압력, A 와 B 는 재료의 물성에 관련된 계수로 각각 75와 15이다.티V티V는 포화 온도이고 T 는 키홀 벽의 온도입니다. 표면 흐름 및 키홀 형성의 다른 원동력은 표면 장력입니다. 표면 장력 계수는 Marangoni 흐름을 포함하기 위해 온도의 선형 함수로 추정되며,σ =1.79-9.90⋅10− 4( 티− 1654케이 )σ=1.79-9.90⋅10-4(티-1654년케이)엔엠− 1-1. 49 ] 계산 영역은 베어 플레이트의 절반입니다(2300 μμ미디엄××250 μμ미디엄××500 μμm) xz 평면 에 적용된 대칭 경계 조건 . 메쉬 크기는 8입니다. μμm이고 시간 단계는 0.15입니다. μμs는 계산 효율성과 정확성 간의 균형을 제공합니다.

결과 및 논의

용융 풀 형태

이 작업에 사용된 5개의 레이저 파워( P )와 6개의 스캐닝 속도( V )는 서로 다른 29개의 용융 풀을 생성했습니다.피- 브이피-V조합. P 와 V 값이 가장 높은 것은 그림 1 을 기준으로 과도한 볼링과 관련이 있기 때문에 본 연구에서는 분석하지 않았다  .

단일 트랙 용융 풀은 그림  1 과 같이 형상에 따라 네 가지 유형으로 분류할 수 있습니다 39 ] : (1) 전도 모드(파란색 상자), (2) 키홀 모드(빨간색), (3) 전환 모드(마젠타), (4) 볼링 모드(녹색). 높은 레이저 출력과 낮은 스캐닝 속도의 일반적인 조합인 키홀 모드에서 용융물 풀은 일반적으로 너비/깊이( W / D ) 비율이 0.5보다 훨씬 큰 깊고 가느다란 모양을 나타냅니다 . 스캐닝 속도가 증가함에 따라 용융 풀이 얕아져 W / D 가 약 0.5인 반원형 전도 모드 용융 풀을 나타냅니다. W / D _전환 모드 용융 풀의 경우 1에서 0.5 사이입니다. 스캐닝 속도를 1200 및 1400mm/s로 더 높이면 충분히 큰 캡 높이와 볼링 모드 용융 풀의 특징인 과도한 언더컷이 발생할 수 있습니다.

힘과 속도의 함수로서의 용융 풀 깊이와 너비는 각각 그림  2 (a)와 (b)에 표시되어 있습니다. 용융 풀 폭은 기판 표면에서 측정되었습니다. 그림  2 (a)는 깊이가 레이저 출력과 매우 선형적인 관계를 따른다는 것을 보여줍니다. 속도가 증가함에 따라 깊이  파워 곡선의 기울기는 꾸준히 감소하지만 더 높은 속도 곡선에는 약간의 겹침이 있습니다. 이러한 예상치 못한 중첩은 종종 용융 풀 형태의 동적 변화를 유발하는 유체 흐름의 영향과 레이저 스캔당 하나의 이미지만 추출되었다는 사실 때문일 수 있습니다. 이러한 선형 동작은 그림 2 (b) 의 너비에 대해 명확하지 않습니다  . 그림  2(c)는 선형 에너지 밀도 P / V 의 함수로서 용융 깊이와 폭을 보여줍니다 . 선형 에너지 밀도는 퇴적물의 단위 길이당 에너지 투입량을 측정한 것입니다. 50 ] 용융 풀 깊이는 에너지 밀도에 따라 달라지며 너비는 더 많은 분산을 나타냅니다. 동일한 에너지 밀도가 준공 부품의 용융 풀, 미세 구조 또는 속성에서 반드시 동일한 유체 역학을 초래하지는 않는다는 점에 유의하는 것이 중요합니다. 50 ]

그림 1
그림 1
그림 2
그림 2

레이저 흡수율 평가

레이저 흡수율은 LPBF 조건에서 재료 및 가공 매개변수에 따라 크게 달라진다는 것은 잘 알려져 있습니다. 31 , 51 , 52 ] 적분구를 이용한 전통적인 흡수율의 직접 측정은 일반적으로 높은 비용과 구현의 어려움으로 인해 쉽게 접근할 수 없습니다. 51 ] 그  . 39 ] 전도 모드 용융 풀에 대한 Rosenthal 방정식을 기반으로 경험적 레이저 흡수율 모델을 개발했지만 기본 가정으로 인해 키홀 용융 풀에 대한 정확한 예측을 제공하지 못했습니다. 40 ] 최근 간 . 53 ] Ti–6Al–4V에 대한 30개의 고충실도 다중 물리 시뮬레이션 사례를 사용하여 레이저 흡수에 대한 스케일링 법칙을 확인했습니다. 그러나 연구 중인 특정 재료에 대한 최소 흡수(평평한 용융 표면의 흡수율)에 대한 지식이 필요하며 이는 CMSX-4에 대해 알려지지 않았습니다. 다양한 키홀 모양의 용융 풀에 대한 레이저 흡수의 정확한 추정치를 얻기가 어렵기 때문에 상한 및 하한 흡수율로 분석 시뮬레이션을 실행하기로 결정했습니다. 깊은 키홀 모양의 용융 풀의 경우 대부분의 빛을 가두는 키홀 내 다중 반사로 인해 레이저 흡수율이 0.8만큼 높을 수 있습니다. 이것은 기하학적 현상이며 기본 재료에 민감하지 않습니다. 5152 , 54 ] 따라서 본 연구에서는 흡수율의 상한을 0.8로 설정하였다. 참고 문헌 51 에 나타낸 바와 같이 , 전도 용융 풀에 해당하는 최저 흡수율은 약 0.3이었으며, 이는 이 연구에서 합리적인 하한 값입니다. 따라서 레이저 흡수율이 스트레이 그레인 형성에 미치는 영향을 보여주기 위해 흡수율 값을 0.55 ± 0.25로 설정했습니다. Vitek의 작업에서는 1.0의 고정 흡수율 값이 사용되었습니다. 3 ]

퓨전 존 미세구조

그림  3 은 200~300W 및 600~300W 및 600~300W 범위의 레이저 출력 및 속도로 9가지 다른 처리 매개변수에 의해 생성된 CMSX-4 레이저 트랙의 yz 단면 에서 취한 EBSD 역극점도와 해당 역극점도를 보여 줍니다. 각각 1400mm/s. EBSD 맵에서 여러 기능을 쉽게 관찰할 수 있습니다. 스트레이 그레인은 EBSD 맵에서 그 방향에 해당하는 다른 RGB 색상으로 나타나고 그레인 경계를 묘사하기 위해 5도의 잘못된 방향이 사용되었습니다. 여기, 그림  3 에서 스트레이 그레인은 대부분 용융 풀의 상단 중심선에 집중되어 있으며, 이는 용접된 단결정 CMSX-4의 이전 보고서와 일치합니다. 10 ]역 극점도에서, 점 근처에 집중된 클러스터⟨ 001 ⟩⟨001⟩융합 경계에서 유사한 방향을 유지하는 단결정 기반 및 에피택셜로 응고된 덴드라이트를 나타냅니다. 그러나 흩어진 곡물은 식별할 수 있는 질감이 없는 흩어져 있는 점으로 나타납니다. 단결정 기본 재료의 결정학적 방향은 주로⟨ 001 ⟩⟨001⟩비록 샘플을 절단하는 동안 식별할 수 없는 기울기 각도로 인해 또는 단결정 성장 과정에서 약간의 잘못된 방향이 있었기 때문에 약간의 편차가 있지만. 용융 풀 내부의 응고된 수상 돌기의 기본 방향은 다시 한 번⟨ 001 ⟩⟨001⟩주상 결정립 구조와 유사한 에피택셜 성장의 결과. 그림 3 과 같이 용융 풀에서 수상돌기의 성장 방향은 하단의 수직 방향에서 상단의 수평 방향으로 변경되었습니다  . 이 전이는 주로 온도 구배 방향의 변화로 인한 것입니다. 두 번째 전환은 CET입니다. FZ의 상단 중심선 주변에서 다양한 방향의 흩어진 입자가 관찰되며, 여기서 안쪽으로 성장하는 수상돌기가 서로 충돌하여 용융 풀에서 응고되는 마지막 위치가 됩니다.

더 깊은 키홀 모양을 특징으로 하는 샘플에서 용융 풀의 경계 근처에 침전된 흩어진 입자가 분명합니다. 이러한 새로운 입자는 나중에 모델링 섹션에서 논의되는 수상돌기 조각화 메커니즘에 의해 잠재적으로 발생합니다. 결정립이 강한 열 구배에서 핵을 생성하고 성장한 결과, 대부분의 흩어진 결정립은 모든 방향에서 동일한 크기를 갖기보다는 장축이 열 구배 방향과 정렬된 길쭉한 모양을 갖습니다. 그림 3 의 전도 모드 용융 풀 흩어진 입자가 없는 것으로 입증되는 더 나은 단결정 품질을 나타냅니다. 상대적으로 낮은 출력과 높은 속도의 스캐닝 레이저에 의해 생성된 이러한 더 얕은 용융 풀에서 최소한의 결정립 핵형성이 발생한다는 것은 명백합니다. 더 큰 면적 분율을 가진 스트레이 그레인은 고출력 및 저속으로 생성된 깊은 용융 풀에서 더 자주 관찰됩니다. 국부 응고 조건에 대한 동력 및 속도의 영향은 후속 모델링 섹션에서 조사할 것입니다.

그림 3
그림 3

응고 모델링

서론에서 언급한 바와 같이 연구자들은 단결정 용접 중에 표류 결정립 형성의 가능한 메커니즘을 평가했습니다. 12 , 13 , 14 , 15 , 55 ]논의된 가장 인기 있는 두 가지 메커니즘은 (1) 응고 전단에 앞서 구성적 과냉각에 의해 도움을 받는 이종 핵형성 및 (2) 용융물 풀의 유체 흐름으로 인한 덴드라이트 조각화입니다. 첫 번째 메커니즘은 광범위하게 연구되었습니다. 이원 합금을 예로 들면, 고체는 액체만큼 많은 용질을 수용할 수 없으므로 응고 중에 용질을 액체로 거부합니다. 결과적으로, 성장하는 수상돌기 앞에서 용질 분할은 실제 온도가 국부 평형 액상선보다 낮은 과냉각 액체를 생성합니다. 충분히 광범위한 체질적으로 과냉각된 구역의 존재는 새로운 결정립의 핵형성 및 성장을 촉진합니다. 56 ]전체 과냉각은 응고 전면에서의 구성, 동역학 및 곡률 과냉각을 포함한 여러 기여의 합입니다. 일반적인 가정은 동역학 및 곡률 과냉각이 합금에 대한 용질 과냉각의 더 큰 기여와 관련하여 무시될 수 있다는 것입니다. 57 ]

서로 다른 기본 메커니즘을 더 잘 이해하려면피- 브이피-V조건에서 응고 모델링이 수행됩니다. 첫 번째 목적은 스트레이 그레인의 전체 범위를 평가하는 것입니다(Φ¯¯¯¯Φ¯) 처리 매개 변수의 함수로 국부적 표류 입자 비율의 변화를 조사하기 위해 (ΦΦ) 용융 풀의 위치 함수로. 두 번째 목적은 금속 AM의 빠른 응고 동안 응고 미세 구조와 표류 입자 형성 메커니즘 사이의 관계를 이해하는 것입니다.

그림 4
그림 4

그림  4 는 해석적으로 시뮬레이션된 표류 입자 비율을 보여줍니다.Φ¯¯¯¯Φ¯세 가지 레이저 흡수율 값에서 다양한 레이저 스캐닝 속도 및 레이저 출력에 대해. 결과는 스트레이 그레인 면적 비율이 흡수된 에너지에 민감하다는 것을 보여줍니다. 흡수율을 0.30에서 0.80으로 증가시키면Φ¯¯¯¯Φ¯약 3배이며, 이 효과는 저속 및 고출력 영역에서 더욱 두드러집니다. 다른 모든 조건이 같다면, 흡수된 전력의 큰 영향은 평균 열 구배 크기의 일반적인 감소와 용융 풀 내 평균 응고율의 증가에 기인합니다. 스캐닝 속도가 증가하고 전력이 감소함에 따라 평균 스트레이 그레인 비율이 감소합니다. 이러한 일반적인 경향은 Vitek의 작업에서 채택된 그림 5 의 파란색 영역에서 시뮬레이션된 용접 결과와 일치합니다  . 3 ] 더 큰 과냉각 구역( 즉, 지 /V티G/V티영역)은 용접 풀의 표유 입자의 면적 비율이 분홍색 영역에 해당하는 LPBF 조건의 면적 비율보다 훨씬 더 크다는 것을 의미합니다. 그럼에도 불구하고 두 데이터 세트의 일반적인 경향은 유사합니다.  , 레이저 출력이 감소하고 레이저 속도가 증가함에 따라 표류 입자의 비율이 감소합니다. 또한 그림  5 에서 스캐닝 속도가 LPBF 영역으로 증가함에 따라 표유 입자 면적 분율에 대한 레이저 매개변수의 변화 효과가 감소한다는 것을 추론할 수 있습니다. 그림  6 (a)는 그림 3 의 EBSD 분석에서 나온 실험적 표류 결정립 면적 분율  과 그림 4 의 해석 시뮬레이션 결과를  비교합니다.. 열쇠 구멍 모양의 FZ에서 정확한 값이 다르지만 추세는 시뮬레이션과 실험 데이터 모두에서 일관되었습니다. 키홀 모양의 용융 풀, 특히 전력이 300W인 2개는 분석 시뮬레이션 예측보다 훨씬 더 많은 양의 흩어진 입자를 가지고 있습니다. Rosenthal 방정식은 일반적으로 열 전달이 순전히 전도에 의해 좌우된다는 가정으로 인해 열쇠 구멍 체제의 열 흐름을 적절하게 반영하지 못하기 때문에 이러한 불일치가 실제로 예상됩니다. 39 , 40 ] 그것은 또한 그림  4 의 발견 , 즉 키홀 모드 동안 흡수된 전력의 증가가 표류 입자 형성에 더 이상적인 조건을 초래한다는 것을 검증합니다. 그림  6 (b)는 실험을 비교Φ¯¯¯¯Φ¯수치 CFD 시뮬레이션Φ¯¯¯¯Φ¯. CFD 모델이 약간 초과 예측하지만Φ¯¯¯¯Φ¯전체적으로피- 브이피-V조건에서 열쇠 구멍 조건에서의 예측은 분석 모델보다 정확합니다. 전도 모드 용융 풀의 경우 실험 값이 분석 시뮬레이션 값과 더 가깝게 정렬됩니다.

그림 5
그림 5

모의 온도 구배 G 분포 및 응고율 검사V티V티분석 모델링의 쌍은 그림  7 (a)의 CMSX-4 미세 구조 선택 맵에 표시됩니다. 제공지 /V티G/V티(  , 형태 인자)는 형태를 제어하고지 ×V티G×V티(  , 냉각 속도)는 응고된 미세 구조의 규모를 제어하고 , 58 , 59 ]지 -V티G-V티플롯은 전통적인 제조 공정과 AM 공정 모두에서 미세 구조 제어를 지원합니다. 이 플롯의 몇 가지 분명한 특징은 등축, 주상, 평면 전면 및 이러한 경계 근처의 전이 영역을 구분하는 경계입니다. 그림  7 (a)는 몇 가지 선택된 분석 열 시뮬레이션에 대한 미세 구조 선택 맵을 나타내는 반면 그림  7 (b)는 수치 열 모델의 결과와 동일한 맵을 보여줍니다. 등축 미세구조의 형성은 낮은 G 이상 에서 명확하게 선호됩니다.V티V티정황. 이 플롯에서 각 곡선의 평면 전면에 가장 가까운 지점은 용융 풀의 최대 너비 위치에 해당하는 반면 등축 영역에 가까운 지점의 끝은 용융 풀의 후면 꼬리에 해당합니다. 그림  7 (a)에서 대부분의지 -V티G-V티응고 전면의 쌍은 원주형 영역에 속하고 점차 CET 영역으로 위쪽으로 이동하지만 용융 풀의 꼬리는 다음에 따라 완전히 등축 영역에 도달하거나 도달하지 않을 수 있습니다.피- 브이피-V조합. 그림 7 (a) 의 곡선 중 어느 것도  평면 전면 영역을 통과하지 않지만 더 높은 전력의 경우에 가까워집니다. 저속 레이저 용융 공정을 사용하는 이전 작업에서는 곡선이 평면 영역을 통과할 수 있습니다. 레이저 속도가 증가함에 따라 용융 풀 꼬리는 여전히 CET 영역에 있지만 완전히 등축 영역에서 멀어집니다. CET 영역으로 떨어지는 섹션의 수도 감소합니다.Φ¯¯¯¯Φ¯응고된 물질에서.

그림 6
그림 6

그만큼지 -V티G-V티CFD 모델을 사용하여 시뮬레이션된 응고 전면의 쌍이 그림  7 (b)에 나와 있습니다. 세 방향 모두에서 각 점 사이의 일정한 간격으로 미리 정의된 좌표에서 수행된 해석 시뮬레이션과 달리, 고충실도 CFD 모델의 출력은 불규칙한 사면체 좌표계에 있었고 G 를 추출하기 전에 일반 3D 그리드에 선형 보간되었습니다. 그리고V티V티그런 다음 미세 구조 선택 맵에 플롯됩니다. 일반적인 경향은 그림  7 (a)의 것과 일치하지만 이 방법으로 모델링된 매우 동적인 유체 흐름으로 인해 결과에 더 많은 분산이 있었습니다. 그만큼지 -V티G-V티분석 열 모델의 쌍 경로는 더 연속적인 반면 수치 시뮬레이션의 경로는 용융 풀 꼬리 모양의 차이를 나타내는 날카로운 굴곡이 있습니다(이는 G 및V티V티) 두 모델에 의해 시뮬레이션됩니다.

그림 7
그림 7
그림 8
그림 8

유체 흐름을 통합한 응고 모델링

수치 CFD 모델을 사용하여 유동 입자 형성 정도에 대한 유체 흐름의 영향을 이해하고 시뮬레이션 결과를 분석 Rosenthal 솔루션과 비교했습니다. 그림  8 은 응고 매개변수 G 의 분포를 보여줍니다.V티V티,지 /V티G/V티, 그리고지 ×V티G×V티yz 단면에서 x  FLOW-3D에서 (a1–d1) 분석 열 모델링 및 (a2–d2) FVM 방법을 사용하여 시뮬레이션된 용융 풀의 최대 폭입니다. 그림  8 의 값은 응고 전선이 특정 위치에 도달할 때 정확한 값일 수도 있고 아닐 수도 있지만 일반적인 추세를 반영한다는 의미의 임시 가상 값입니다. 이 프로파일은 출력 300W 및 속도 400mm/s의 레이저 빔에서 시뮬레이션됩니다. 용융 풀 경계는 흰색 곡선으로 표시됩니다. (a2–d2)의 CFD 시뮬레이션 용융 풀 깊이는 342입니다. μμm, 측정 깊이 352와 잘 일치 μμ일치하는 길쭉한 열쇠 구멍 모양과 함께 그림 1 에 표시된 실험 FZ의 m  . 그러나 분석 모델은 반원 모양의 용융 풀을 출력하고 용융 풀 깊이는 264에 불과합니다. μμ열쇠 구멍의 경우 현실과는 거리가 멀다. CFD 시뮬레이션 결과에서 열 구배는 레이저 반사 증가와 불안정한 액체-증기 상호 작용이 발생하는 증기 함몰의 동적 부분 근처에 있기 때문에 FZ 하단에서 더 높습니다. 대조적으로 해석 결과의 열 구배 크기는 경계를 따라 균일합니다. 두 시뮬레이션 결과 모두 그림 8 (a1) 및 (a2) 에서 응고가 용융 풀의 상단 중심선을 향해 진행됨에 따라 열 구배가 점차 감소합니다  . 응고율은 그림 8 과 같이 경계 근처에서 거의 0입니다. (b1) 및 (b2). 이는 경계 영역이 응고되기 시작할 때 국부 응고 전면의 법선 방향이 레이저 스캐닝 방향에 수직이기 때문입니다. 이것은 드라이브θ → π/ 2θ→파이/2그리고V티→ 0V티→0식에서 [ 3 ]. 대조적으로 용융 풀의 상단 중심선 근처 영역에서 응고 전면의 법선 방향은 레이저 스캐닝 방향과 잘 정렬되어 있습니다.θ → 0θ→0그리고V티→ 브이V티→V, 빔 스캐닝 속도. G 와 _V티V티값이 얻어지면 냉각 속도지 ×V티G×V티및 형태 인자지 /V티G/V티계산할 수 있습니다. 그림 8 (c2)는 용융 풀 바닥 근처의 온도 구배가 매우 높고 상단에서 더 빠른 성장 속도로  인해 냉각 속도가 용융 풀의 바닥 및 상단 중심선 근처에서 더 높다는 것을 보여줍니다. 지역. 그러나 이러한 추세는 그림  8 (c1)에 캡처되지 않았습니다. 그림 8 의 형태 요인 (d1) 및 (d2)는 중심선에 접근함에 따라 눈에 띄게 감소합니다. 경계에서 큰 값은 열 구배를 거의 0인 성장 속도로 나누기 때문에 발생합니다. 이 높은 형태 인자는 주상 미세구조 형성 가능성이 높음을 시사하는 반면, 중앙 영역의 값이 낮을수록 등축 미세구조의 가능성이 더 크다는 것을 나타냅니다. Tanet al. 또한 키홀 모양의 용접 풀 59 ] 에서 이러한 응고 매개변수의 분포 를 비슷한 일반적인 경향으로 보여주었습니다. 그림  3 에서 볼 수 있듯이 용융 풀의 상단 중심선에 있는 흩어진 입자는 낮은 특징을 나타내는 영역과 일치합니다.지 /V티G/V티그림  8 (d1) 및 (d2)의 값. 시뮬레이션과 실험 간의 이러한 일치는 용융 풀의 상단 중심선에 축적된 흩어진 입자의 핵 생성 및 성장이 등온선 속도의 증가와 온도 구배의 감소에 의해 촉진됨을 보여줍니다.

그림 9
그림 9

그림  9 는 유체 속도 및 국부적 핵형성 성향을 보여줍니다.ΦΦ300W의 일정한 레이저 출력과 400, 800 및 1200mm/s의 세 가지 다른 레이저 속도에 의해 생성된 3D 용융 풀 전체에 걸쳐. 그림  9 (d)~(f)는 로컬ΦΦ해당 3D 보기에서 밝은 회색 평면으로 표시된 특정 yz 단면의 분포. 이 yz 섹션은 가장 높기 때문에 선택되었습니다.Φ¯¯¯¯Φ¯용융 풀 내의 값은 각각 23.40, 11.85 및 2.45pct입니다. 이들은 그림  3 의 실험 데이터와 비교하기에 적절하지 않을 수 있는 액체 용융 풀의 과도 값이며Φ¯¯¯¯Φ¯그림  6 의 값은 이 값이 고체-액체 계면에 가깝지 않고 용융 풀의 중간에서 취해졌기 때문입니다. 온도가 훨씬 낮아서 핵이 생존하고 성장할 수 있기 때문에 핵 형성은 용융 풀의 중간이 아닌 고체-액체 계면에 더 가깝게 발생할 가능성이 있습니다.

그림  3 (a), (d), (g), (h)에서 위쪽 중심선에서 멀리 떨어져 있는 흩어진 결정립이 있었습니다. 그들은 훨씬 더 높은 열 구배와 더 낮은 응고 속도 필드에 위치하기 때문에 과냉각 이론은 이러한 영역에서 표류 입자의 형성에 대한 만족스러운 설명이 아닙니다. 이것은 떠돌이 결정립의 형성을 야기할 수 있는 두 번째 메커니즘,  수상돌기의 팁을 가로지르는 유체 흐름에 의해 유발되는 수상돌기 조각화를 고려하도록 동기를 부여합니다. 유체 흐름이 열 구배를 따라 속도 성분을 갖고 고체-액체 계면 속도보다 클 때, 주상 수상돌기의 국지적 재용융은 용질이 풍부한 액체가 흐물흐물한 구역의 깊은 곳에서 액상선 등온선까지 이동함으로써 발생할 수 있습니다. . 55] 분리된 수상돌기는 대류에 의해 열린 액체로 운반될 수 있습니다. 풀이 과냉각 상태이기 때문에 이러한 파편은 고온 조건에서 충분히 오래 생존하여 길 잃은 입자의 핵 생성 사이트로 작용할 수 있습니다. 결과적으로 수상 돌기 조각화 과정은 활성 핵의 수를 효과적으로 증가시킬 수 있습니다.N0N0) 용융 풀 15 , 60 , 61 ] 에서 생성된 미세 구조에서 표류 입자의 면적을 증가시킵니다.

그림  9 (a) 및 (b)에서 반동 압력은 용융 유체를 아래쪽으로 흐르게 하여 결과 흐름을 지배합니다. 유체 속도의 역방향 요소는 V = 400 및 800mm/s에 대해 각각 최대값 1.0 및 1.6m/s로 더 느려집니다 . 그림  9 (c)에서 레이저 속도가 더 증가함에 따라 증기 침하가 더 얕고 넓어지고 반동 압력이 더 고르게 분포되어 증기 침강에서 주변 영역으로 유체를 밀어냅니다. 역류는 최대값 3.5m/s로 더 빨라집니다. 용융 풀의 최대 너비에서 yz 단면  의 키홀 아래 평균 유체 속도는 그림에 표시된 경우에 대해 0.46, 0.45 및 1.44m/s입니다.9 (a), (b) 및 (c). 키홀 깊이의 변동은 각 경우의 최대 깊이와 최소 깊이의 차이로 정의되는 크기로 정량화됩니다. 240 범위의 강한 증기 내림 변동 μμm은 그림 9 (a)의 V = 400mm/s 경우에서  발견 되지만 이 변동은 그림  9 (c)에서 16의 범위로  크게 감소합니다.μμ미디엄. V = 400mm/s인 경우 의 유체장과 높은 변동 범위는 이전 키홀 동역학 시뮬레이션과 일치합니다. 34 ]

따라서 V = 400mm/s 키홀 케이스의 무질서한 변동 흐름이 용융 풀 경계를 따라 응고된 주상 수상돌기에서 분리된 조각을 구동할 가능성이 있습니다. V = 1200mm/s의 경우 강한 역류 는 그림 3 에서 관찰되지 않았지만 동일한 효과를 가질 수 있습니다. . 덴드라이트 조각화에 대한 유체 유동장의 영향에 대한 이 경험적 설명은 용융 풀 경계 근처에 떠돌이 입자의 존재에 대한 그럴듯한 설명을 제공합니다. 분명히 하기 위해, 우리는 이 가설을 검증하기 위해 이 현상에 대한 직접적인 실험적 관찰을 하지 않았습니다. 이 작업에서 표유 입자 면적 분율을 계산할 때 단순화를 위해 핵 생성 모델링에 일정한 핵 생성 수 밀도가 적용되었습니다. 이는 그림  9 의 표류 입자 영역 비율 이 수지상정 조각화가 발생하는 경우 이러한 높은 유체 흐름 용융 풀에서 발생할 수 있는 것,  강화된 핵 생성 밀도를 반영하지 않는다는 것을 의미합니다.

위의 이유로 핵 형성에 대한 수상 돌기 조각화의 영향을 아직 배제할 수 없습니다. 그러나 단편화 이론은 용접 문헌 [ 62 ] 에서 검증될 만큼 충분히 개발되지 않았 으므로 부차적인 중요성만 고려된다는 점에 유의해야 합니다. 1200mm/s를 초과하는 레이저 스캐닝 속도는 최소한의 표류 결정립 면적 분율을 가지고 있음에도 불구하고 분명한 볼링을 나타내기 때문에 단결정 수리 및 AM 처리에 적합하지 않습니다. 따라서 낮은 P 및 높은 V 에 의해 생성된 응고 전면 근처에서 키홀 변동이 최소화되고 유체 속도가 완만해진 용융 풀이 생성된다는 결론을 내릴 수 있습니다., 처리 창의 극한은 아니지만 흩어진 입자를 나타낼 가능성이 가장 적습니다.

마지막으로 단일 레이저 트랙의 응고 거동을 조사하면 에피택셜 성장 동안 표류 입자 형성을 더 잘 이해할 수 있다는 점에 주목하는 것이 중요합니다. 우리의 현재 결과는 최적의 레이저 매개변수에 대한 일반적인 지침을 제공하여 최소 스트레이 그레인을 달성하고 단결정 구조를 유지합니다. 이 가이드라인은 250W 정도의 전력과 600~800mm/s의 스캔 속도로 최소 흩어진 입자에 적합한 공정 창을 제공합니다. 각 처리 매개변수를 신중하게 선택하면 과거에 스테인리스강에 대한 거의 단결정 미세 구조를 인쇄하는 데 성공했으며 이는 CMSX-4 AM 빌드에 대한 가능성을 보여줍니다. 63 ]신뢰성을 보장하기 위해 AM 수리 프로세스를 시작하기 전에 보다 엄격한 실험 테스트 및 시뮬레이션이 여전히 필요합니다. 둘 이상의 레이저 트랙 사이의 상호 작용도 고려해야 합니다. 또한 레이저, CMSX-4 분말 및 벌크 재료 간의 상호 작용이 중요하며, 수리 중에 여러 층의 CMSX-4 재료를 축적해야 하는 경우 다른 스캔 전략의 효과도 중요한 역할을 할 수 있습니다. 분말이 포함된 경우 Lopez-Galilea 등 의 연구에서 제안한 바와 같이 분말이 주로 완전히 녹지 않았을 때 추가 핵 생성 사이트를 도입하기 때문에 단순히 레이저 분말과 속도를 조작하여 흩어진 입자 형성을 완화하기 어려울 수 있습니다 . 22 ]결과적으로 CMSX-4 단결정을 수리하기 위한 레이저 AM의 가능성을 다루기 위해서는 기판 재료, 레이저 출력, 속도, 해치 간격 및 층 두께의 조합을 모두 고려해야 하며 향후 연구에서 다루어야 합니다. CFD 모델링은 2개 이상의 레이저 트랙 사이의 상호작용과 열장에 미치는 영향을 통합할 수 있으며, 이는 AM 빌드 시나리오 동안 핵 생성 조건으로 단일 비드 연구의 지식 격차를 해소할 것입니다.

결론

LPBF 제조의 특징적인 조건 하에서 CMSX-4 단결정 의 에피택셜(기둥형)  등축 응고 사이의 경쟁을 실험적 및 이론적으로 모두 조사했습니다. 이 연구는 고전적인 응고 개념을 도입하여 빠른 레이저 용융의 미세 구조 특징을 설명하고 응고 조건과 표유 결정 성향을 예측하기 위해 해석적 및 수치적 고충실도 CFD 열 모델 간의 비교를 설명했습니다. 본 연구로부터 다음과 같은 주요 결론을 도출할 수 있다.

  • 단일 레이저 트랙의 레이저 가공 조건은 용융 풀 형상, 레이저 흡수율, 유체 흐름 및 키홀 요동, 입자 구조 및 표류 입자 형성 민감성에 강한 영향을 미치는 것으로 밝혀졌습니다.
  • 레이저 용접을 위해 개발된 이론적인 표유 결정립 핵형성 분석이 레이저 용융 AM 조건으로 확장되었습니다. 분석 모델링 결과와 단일 레이저 트랙의 미세구조 특성화를 비교하면 예측이 전도 및 볼링 조건에서 실험적 관찰과 잘 일치하는 반면 키홀 조건에서는 예측이 약간 과소하다는 것을 알 수 있습니다. 이러한 불일치는 레이저 트랙의 대표성이 없는 섹션이나 유체 속도 필드의 변화로 인해 발생할 수 있습니다. CFD 모델에서 추출한 열장에 동일한 표유 입자 계산 파이프라인을 적용하면 연구된 모든 사례에서 과대평가가 발생하지만 분석 모델보다 연장된 용융 풀의 실험 데이터와 더 정확하게 일치합니다.
  • 이 연구에서 두 가지 표류 결정립 형성 메커니즘인 불균일 핵형성 및 수상돌기 조각화가 평가되었습니다. 우리의 결과는 불균일 핵형성이 용융 풀의 상단 중심선에서 새로운 결정립의 형성으로 이어지는 주요 메커니즘임을 시사합니다.지 /V티G/V티정권.
  • 용융 풀 경계 근처의 흩어진 입자는 깊은 키홀 모양의 용융 풀에서 독점적으로 관찰되며, 이는 강한 유체 흐름으로 인한 수상 돌기 조각화의 영향이 이러한 유형의 용융 풀에서 고려하기에 충분히 강력할 수 있음을 시사합니다.
  • 일반적으로 더 높은 레이저 스캐닝 속도와 더 낮은 전력 외에도 안정적인 키홀과 최소 유체 속도는 또한 흩어진 입자 형성을 완화하고 레이저 단일 트랙에서 에피택셜 성장을 보존합니다.

References

  1. R.C. Reed: The Superalloys: Fundamentals and Applications, Cambridge University Press, Cambridge, 2006, pp.17–20.Book Google Scholar 
  2. A. Basak, R. Acharya, and S. Das: Metall. Mater. Trans. A, 2016, vol. 47A, pp. 3845–59.Article Google Scholar 
  3. J. Vitek: Acta Mater., 2005, vol. 53, pp. 53–67.Article CAS Google Scholar 
  4. R. Vilar and A. Almeida: J. Laser Appl., 2015, vol. 27, p. S17004.Article Google Scholar 
  5. T. Kalfhaus, M. Schneider, B. Ruttert, D. Sebold, T. Hammerschmidt, J. Frenzel, R. Drautz, W. Theisen, G. Eggeler, O. Guillon, and R. Vassen: Mater. Des., 2019, vol. 168, p. 107656.Article CAS Google Scholar 
  6. S.S. Babu, S.A. David, J.W. Park, and J.M. Vitek: Sci. Technol. Weld. Join., 2004, vol. 9, pp. 1–12.Article CAS Google Scholar 
  7. L. Felberbaum, K. Voisey, M. Gäumann, B. Viguier, and A. Mortensen: Mater. Sci. Eng. A, 2001, vol. 299, pp. 152–56.Article Google Scholar 
  8. S. Mokadem, C. Bezençon, J.M. Drezet, A. Jacot, J.D. Wagnière, and W. Kurz: TMS Annual Meeting, 2004, pp. 67–76.
  9. J.M. Vitek: ASM Proc. Int. Conf. Trends Weld. Res., vol. 2005, pp. 773–79.
  10. J.M. Vitek, S. Babu, and S. David: Process Optimization for Welding Single-Crystal Nickel-Bbased Superalloyshttps://technicalreports.ornl.gov/cppr/y2001/pres/120424.pdf
  11. J.D. Hunt: Mater. Sci. Eng., 1984, vol. 65, pp. 75–83.Article CAS Google Scholar 
  12. M. Gäumann, R. Trivedi, and W. Kurz: Mater. Sci. Eng. A, 1997, vol. 226–228, pp. 763–69.Article Google Scholar 
  13. M. Gäumann, S. Henry, F. Cléton, J.D. Wagnière, and W. Kurz: Mater. Sci. Eng. A, 1999, vol. 271, pp. 232–41.Article Google Scholar 
  14. M. Gäumann, C. Bezençon, P. Canalis, and W. Kurz: Acta Mater., 2001, vol. 49, pp. 1051–62.Article Google Scholar 
  15. J.M. Vitek, S.A. David, and S.S. Babu: Welding and Weld Repair of Single Crystal Gas Turbine Alloyshttps://www.researchgate.net/profile/Stan-David/publication/238692931_WELDING_AND_WELD_REPAIR_OF_SINGLE_CRYSTAL_GAS_TURBINE_ALLOYS/links/00b4953204ab35bbad000000/WELDING-AND-WELD-REPAIR-OF-SINGLE-CRYSTAL-GAS-TURBINE-ALLOYS.pdf
  16. B. Kianian: Wohlers Report 2017: 3D Printing and Additive Manufacturing State of the Industry, Annual Worldwide Progress Report, Wohlers Associates, Inc., Fort Collins, 2017.Google Scholar 
  17. M. Ramsperger, L. Mújica Roncery, I. Lopez-Galilea, R.F. Singer, W. Theisen, and C. Körner: Adv. Eng. Mater., 2015, vol. 17, pp. 1486–93.Article CAS Google Scholar 
  18. A.B. Parsa, M. Ramsperger, A. Kostka, C. Somsen, C. Körner, and G. Eggeler: Metals, 2016, vol. 6, pp. 258-1–17.Article Google Scholar 
  19. C. Körner, M. Ramsperger, C. Meid, D. Bürger, P. Wollgramm, M. Bartsch, and G. Eggeler: Metall. Mater. Trans. A, 2018, vol. 49A, pp. 3781–92.Article Google Scholar 
  20. D. Bürger, A. Parsa, M. Ramsperger, C. Körner, and G. Eggeler: Mater. Sci. Eng. A, 2019, vol. 762, p. 138098,Article Google Scholar 
  21. J. Pistor and C. Körner: Sci. Rep., 2021, vol. 11, p. 24482.Article CAS Google Scholar 
  22. I. Lopez-Galilea, B. Ruttert, J. He, T. Hammerschmidt, R. Drautz, B. Gault, and W. Theisen: Addit. Manuf., 2019, vol. 30, p. 100874.CAS Google Scholar 
  23. N. Lu, Z. Lei, K. Hu, X. Yu, P. Li, J. Bi, S. Wu, and Y. Chen: Addit. Manuf., 2020, vol. 34, p. 101228.CAS Google Scholar 
  24. K. Chen, R. Huang, Y. Li, S. Lin, W. Zhu, N. Tamura, J. Li, Z.W. Shan, and E. Ma: Adv. Mater., 2020, vol. 32, pp. 1–8.Google Scholar 
  25. W.J. Sames, F.A. List, S. Pannala, R.R. Dehoff, and S.S. Babu: Int. Mater. Rev., 2016, vol. 61, pp. 315–60.Article Google Scholar 
  26. A. Basak, R. Acharya, and S. Das: Addit. Manuf., 2018, vol. 22, pp. 665–71.CAS Google Scholar 
  27. R. Jiang, A. Mostafaei, J. Pauza, C. Kantzos, and A.D. Rollett: Mater. Sci. Eng. A, 2019. https://doi.org/10.1016/J.MSEA.2019.03.103.Article Google Scholar 
  28. R. Cunningham, C. Zhao, N. Parab, C. Kantzos, J. Pauza, K. Fezzaa, T. Sun, and A.D. Rollett: Science, 2019, vol. 363, pp. 849–52.Article CAS Google Scholar 
  29. B. Fotovvati, S.F. Wayne, G. Lewis, and E. Asadi: Adv. Mater. Sci. Eng., 2018, vol. 2018, p. 4920718.Article Google Scholar 
  30. P.-J. Chiang, R. Jiang, R. Cunningham, N. Parab, C. Zhao, K. Fezzaa, T. Sun, and A.D. Rollett: in Advanced Real Time Imaging II, pp. 77–85.
  31. J. Ye, S.A. Khairallah, A.M. Rubenchik, M.F. Crumb, G. Guss, J. Belak, and M.J. Matthews: Adv. Eng. Mater., 2019, vol. 21, pp. 1–9.Article Google Scholar 
  32. C. Zhao, Q. Guo, X. Li, N. Parab, K. Fezzaa, W. Tan, L. Chen, and T. Sun: Phys. Rev. X, 2019, vol. 9, p. 021052.CAS Google Scholar 
  33. S.A. Khairallah, A.T. Anderson, A. Rubenchik, and W.E. King: Acta Mater., 2016, vol. 108, pp. 36–45.Article CAS Google Scholar 
  34. N. Kouraytem, X. Li, R. Cunningham, C. Zhao, N. Parab, T. Sun, A.D. Rollett, A.D. Spear, and W. Tan: Appl. Phys. Rev., 2019, vol. 11, p. 064054.Article CAS Google Scholar 
  35. T. DebRoy, H. Wei, J. Zuback, T. Mukherjee, J. Elmer, J. Milewski, A. Beese, A. Wilson-Heid, A. De, and W. Zhang: Prog. Mater. Sci., 2018, vol. 92, pp. 112–224.Article CAS Google Scholar 
  36. J.H. Cho and S.J. Na: J. Phys. D, 2006, vol. 39, pp. 5372–78.Article CAS Google Scholar 
  37. I. Yadroitsev, A. Gusarov, I. Yadroitsava, and I. Smurov: J. Mater. Process. Technol., 2010, vol. 210, pp. 1624–31.Article CAS Google Scholar 
  38. S. Ghosh, L. Ma, L.E. Levine, R.E. Ricker, M.R. Stoudt, J.C. Heigel, and J.E. Guyer: JOM, 2018, vol. 70, pp. 1011–16.Article CAS Google Scholar 
  39. Y. He, C. Montgomery, J. Beuth, and B. Webler: Mater. Des., 2019, vol. 183, p. 108126.Article CAS Google Scholar 
  40. D. Rosenthal: Weld. J., 1941, vol. 20, pp. 220–34.Google Scholar 
  41. M. Tang, P.C. Pistorius, and J.L. Beuth: Addit. Manuf., 2017, vol. 14, pp. 39–48.CAS Google Scholar 
  42. R.E. Aune, L. Battezzati, R. Brooks, I. Egry, H.J. Fecht, J.P. Garandet, M. Hayashi, K.C. Mills, A. Passerone, P.N. Quested, E. Ricci, F. Schmidt-Hohagen, S. Seetharaman, B. Vinet, and R.K. Wunderlich: Proc. Int.Symp. Superalloys Var. Deriv., 2005, pp. 467–76.
  43. B.C. Wilson, J.A. Hickman, and G.E. Fuchs: JOM, 2003, vol. 55, pp. 35–40.Article CAS Google Scholar 
  44. J.J. Valencia and P.N. Quested: ASM Handb., 2008, vol. 15, pp. 468–81.Google Scholar 
  45. H.L. Wei, J. Mazumder, and T. DebRoy: Sci. Rep., 2015, vol. 5, pp. 1–7.Google Scholar 
  46. N. Raghavan, R. Dehoff, S. Pannala, S. Simunovic, M. Kirka, J. Turner, N. Carlson, and S.S. Babu: Acta Mater., 2016, vol. 112, pp. 303–14.Article CAS Google Scholar 
  47. R. Lin, H. Wang, F. Lu, J. Solomon, and B.E. Carlson: Int. J. Heat Mass Transf., 2017, vol. 108, pp. 244–56.Article CAS Google Scholar 
  48. M. Bayat, A. Thanki, S. Mohanty, A. Witvrouw, S. Yang, J. Thorborg, N.S. Tiedje, and J.H. Hattel: Addit. Manuf., 2019, vol. 30, p. 100835.CAS Google Scholar 
  49. K. Higuchi, H.-J. Fecht, and R.K. Wunderlich: Adv. Eng. Mater., 2007, vol. 9, pp. 349–54.Article CAS Google Scholar 
  50. Q. Guo, C. Zhao, M. Qu, L. Xiong, L.I. Escano, S.M.H. Hojjatzadeh, N.D. Parab, K. Fezzaa, W. Everhart, T. Sun, and L. Chen: Addit. Manuf., 2019, vol. 28, pp. 600–09.Google Scholar 
  51. J. Trapp, A.M. Rubenchik, G. Guss, and M.J. Matthews: Appl. Mater. Today, 2017, vol. 9, pp. 341–49.Article Google Scholar 
  52. M. Schneider, L. Berthe, R. Fabbro, and M. Muller: J. Phys. D, 2008, vol. 41, p. 155502.Article Google Scholar 
  53. Z. Gan, O.L. Kafka, N. Parab, C. Zhao, L. Fang, O. Heinonen, T. Sun, and W.K. Liu: Nat. Commun., 2021, vol. 12, p. 2379.Article CAS Google Scholar 
  54. B.J. Simonds, E.J. Garboczi, T.A. Palmer, and P.A. Williams: Appl. Phys. Rev., 2020, vol. 13, p. 024057.Article CAS Google Scholar 
  55. J. Dantzig and M. Rappaz: Solidification, 2nd ed., EPFL Press, Lausanne, 2016, pp. 483–532.Google Scholar 
  56. W. Tiller, K. Jackson, J. Rutter, and B. Chalmers: Acta Metall., 1953, vol. 1, pp. 428–37.Article CAS Google Scholar 
  57. D. Zhang, A. Prasad, M.J. Bermingham, C.J. Todaro, M.J. Benoit, M.N. Patel, D. Qiu, D.H. StJohn, M. Qian, and M.A. Easton: Metall. Mater. Trans. A, 2020, vol. 51A, pp. 4341–59.Article Google Scholar 
  58. F. Yan, W. Xiong, and E.J. Faierson: Materials, 2017, vol. 10, p. 1260.Article Google Scholar 
  59. W. Tan and Y.C. Shin: Comput. Mater. Sci., 2015, vol. 98, pp. 446–58.Article CAS Google Scholar 
  60. A. Hellawell, S. Liu, and S.Z. Lu: JOM, 1997, vol. 49, pp. 18–20.Article CAS Google Scholar 
  61. H. Ji: China Foundry, 2019, vol. 16, pp. 262–66.Article Google Scholar 
  62. J.M. Vitek, S.A. David, and L.A. Boatner: Sci. Technol. Weld. Join., 1997, vol. 2, pp. 109–18.Article CAS Google Scholar 
  63. X. Wang, J.A. Muñiz-Lerma, O. Sanchez-Mata, S.E. Atabay, M.A. Shandiz, and M. Brochu: Prog. Addit. Manuf., 2020, vol. 5, pp. 41–49.Article Google Scholar 

Download references

Nanoparticle-enabled increase of energy efficiency during laser metal additive manufacturing

레이저 금속 적층 제조 중 나노 입자로 에너지 효율 증가

Minglei Quo bQilin Guo a bLuis IzetEscano a bAli Nabaa a bKamel Fezzaa cLianyi Chen a b

레이저 금속 적층 제조(AM) 공정의 낮은 에너지 효율은 대규모 산업 생산에서 잠재적인 지속 가능성 문제입니다. 레이저 용융을 위한 에너지 효율의 명시적 조사는 용융 금속의 불투명한 특성으로 인해 매우 어려운 용융 풀 치수 및 증기 내림의 직접적인 특성화를 요구합니다. 

여기에서 우리는 현장 고속 고에너지 x-선 이미징에 의해 Al6061의 레이저 분말 베드 융합(LPBF) 동안 증기 강하 및 용융 풀 형성에 대한 TiC 나노 입자의 효과에 대한 직접적인 관찰 및 정량화를 보고합니다. 정량 결과를 바탕으로, 우리는 Al6061의 LPBF 동안 TiC 나노 입자가 있거나 없을 때 레이저 용융 에너지 효율(여기서 재료를 용융하는 데 필요한 에너지 대 레이저 빔에 의해 전달되는 에너지의 비율로 정의)을 계산했습니다. 

결과는 TiC 나노 입자를 Al6061에 추가하면 레이저 용융 에너지 효율이 크게 증가한다는 것을 보여줍니다(평균 114% 증가, 312에서 521% 증가). W 레이저 출력, 0.4m  /s 스캔 속도). 체계적인 특성 측정, 시뮬레이션 및 x-선 이미징 연구를 통해 우리는 처음으로 세 가지 메커니즘이 함께 작동하여 레이저 용융 에너지 효율을 향상시킨다는 것을 확인할 수 있었습니다.

(1) TiC 나노 입자를 추가하면 흡수율이 증가합니다. (2) TiC 나노입자를 추가하면 열전도율이 감소하고, (3) TiC 나노입자를 추가하면 더 낮은 레이저 출력에서 ​​증기 억제 및 다중 반사를 시작할 수 있습니다(즉, 키홀링에 대한 레이저 출력 임계값을 낮춤). 

여기서 보고한 Al6061의 LPBF 동안 레이저 용융 에너지 효율을 증가시키기 위해 TiC 나노입자를 사용하는 방법 및 메커니즘은 보다 에너지 효율적인 레이저 금속 AM을 위한 공급원료 재료의 개발을 안내할 수 있습니다.

The low energy efficiency of the laser metal additive manufacturing (AM) process is a potential sustainability concern for large-scale industrial production. Explicit investigation of the energy efficiency for laser melting requires the direct characterization of melt pool dimension and vapor depression, which is very difficult due to the opaque nature of the molten metal. Here we report the direct observation and quantification of effects of the TiC nanoparticles on the vapor depression and melt pool formation during laser powder bed fusion (LPBF) of Al6061 by in-situ high-speed high-energy x-ray imaging. Based on the quantification results, we calculated the laser melting energy efficiency (defined here as the ratio of the energy needed to melt the material to the energy delivered by the laser beam) with and without TiC nanoparticles during LPBF of Al6061. The results show that adding TiC nanoparticles into Al6061 leads to a significant increase of laser melting energy efficiency (114% increase on average, 521% increase under 312 W laser power, 0.4 m/s scan speed). Systematic property measurement, simulation, and x-ray imaging studies enable us, for the first time, to identify that three mechanisms work together to enhance the laser melting energy efficiency: (1) adding TiC nanoparticles increases the absorptivity; (2) adding TiC nanoparticles decreases the thermal conductivity, and (3) adding TiC nanoparticles enables the initiation of vapor depression and multiple reflection at lower laser power (i.e., lowers the laser power threshold for keyholing). The method and mechanisms of using TiC nanoparticles to increase the laser melting energy efficiency during LPBF of Al6061 we reported here may guide the development of feedstock materials for more energy efficient laser metal AM.

Nanoparticle-enabled increase of energy efficiency during laser metal additive manufacturing
Nanoparticle-enabled increase of energy efficiency during laser metal additive manufacturing

Keywords

Additive manufacturing

laser powder bed fusion

energy efficiency

keyhole

melt pool

x-ray imaging

metal matrix nanocomposites

Fig. 1 Multi-physics phenomena in the laser-material interaction zone

COMPARISON BETWEEN GREEN AND
INFRARED LASER IN LASER POWDER BED
FUSION OF PURE COPPER THROUGH HIGH
FIDELITY NUMERICAL MODELLING AT MESOSCALE

316-L 스테인리스강의 레이저 분말 베드 융합 중 콜드 스패터 형성의 충실도 높은 수치 모델링

W.E. ALPHONSO1*, M. BAYAT1 and J.H. HATTEL1
*Corresponding author
1Technical University of Denmark (DTU), 2800, Kgs, Lyngby, Denmark

ABSTRACT

L-PBF(Laser Powder Bed Fusion)는 금속 적층 제조(MAM) 기술로, 기존 제조 공정에 비해 부품 설계 자유도, 조립품 통합, 부품 맞춤화 및 낮은 툴링 비용과 같은 여러 이점을 산업에 제공합니다.

전기 코일 및 열 관리 장치는 일반적으로 높은 전기 및 열 전도성 특성으로 인해 순수 구리로 제조됩니다. 따라서 순동의 L-PBF가 가능하다면 기하학적으로 최적화된 방열판과 자유형 전자코일을 제작할 수 있습니다.

그러나 L-PBF로 조밀한 순동 부품을 생산하는 것은 적외선에 대한 낮은 광 흡수율과 높은 열전도율로 인해 어렵습니다. 기존의 L-PBF 시스템에서 조밀한 구리 부품을 생산하려면 적외선 레이저의 출력을 500W 이상으로 높이거나 구리의 광흡수율이 높은 녹색 레이저를 사용해야 합니다.

적외선 레이저 출력을 높이면 후면 반사로 인해 레이저 시스템의 광학 구성 요소가 손상되고 렌즈의 열 광학 현상으로 인해 공정이 불안정해질 수 있습니다. 이 작업에서 FVM(Finite Volume Method)에 기반한 다중 물리학 중간 규모 수치 모델은 Flow-3D에서 개발되어 용융 풀 역학과 궁극적으로 부품 품질을 제어하는 ​​물리적 현상 상호 작용을 조사합니다.

녹색 레이저 열원과 적외선 레이저 열원은 기판 위의 순수 구리 분말 베드에 단일 트랙 증착을 생성하기 위해 개별적으로 사용됩니다.

용융 풀 역학에 대한 레이저 열원의 유사하지 않은 광학 흡수 특성의 영향이 탐구됩니다. 수치 모델을 검증하기 위해 단일 트랙이 구리 분말 베드에 증착되고 시뮬레이션된 용융 풀 모양과 크기가 비교되는 실험이 수행되었습니다.

녹색 레이저는 광흡수율이 높아 전도 및 키홀 모드 용융이 가능하고 적외선 레이저는 흡수율이 낮아 키홀 모드 용융만 가능하다. 레이저 파장에 대한 용융 모드의 변화는 궁극적으로 기계적, 전기적 및 열적 특성에 영향을 미치는 열 구배 및 냉각 속도에 대한 결과를 가져옵니다.

Laser Powder Bed Fusion (L-PBF) is a Metal Additive Manufacturing (MAM) technology which offers several advantages to industries such as part design freedom, consolidation of assemblies, part customization and low tooling cost over conventional manufacturing processes. Electric coils and thermal management devices are generally manufactured from pure copper due to its high electrical and thermal conductivity properties. Therefore, if L-PBF of pure copper is feasible, geometrically optimized heat sinks and free-form electromagnetic coils can be manufactured. However, producing dense pure copper parts by L-PBF is difficult due to low optical absorptivity to infrared radiation and high thermal conductivity. To produce dense copper parts in a conventional L-PBF system either the power of the infrared laser must be increased above 500W, or a green laser should be used for which copper has a high optical absorptivity. Increasing the infrared laser power can damage the optical components of the laser systems due to back reflections and create instabilities in the process due to thermal-optical phenomenon of the lenses. In this work, a multi-physics meso-scale numerical model based on Finite Volume Method (FVM) is developed in Flow-3D to investigate the physical phenomena interaction which governs the melt pool dynamics and ultimately the part quality. A green laser heat source and an infrared laser heat source are used individually to create single track deposition on pure copper powder bed above a substrate. The effect of the dissimilar optical absorptivity property of laser heat sources on the melt pool dynamics is explored. To validate the numerical model, experiments were conducted wherein single tracks are deposited on a copper powder bed and the simulated melt pool shape and size are compared. As the green laser has a high optical absorptivity, a conduction and keyhole mode melting is possible while for the infrared laser only keyhole mode melting is possible due to low absorptivity. The variation in melting modes with respect to the laser wavelength has an outcome on thermal gradient and cooling rates which ultimately affect the mechanical, electrical, and thermal properties.

Keywords

Pure Copper, Laser Powder Bed Fusion, Finite Volume Method, multi-physics

Fig. 1 Multi-physics phenomena in the laser-material interaction zone
Fig. 1 Multi-physics phenomena in the laser-material interaction zone
Fig. 2 Framework for single laser track simulation model including powder bed and substrate (a) computational domain with boundaries (b) discretization of the domain with uniform quad mesh.
Fig. 2 Framework for single laser track simulation model including powder bed and substrate (a) computational domain with boundaries (b) discretization of the domain with uniform quad mesh.
Fig. 3 2D melt pool contours from the numerical model compared to experiments [16] for (a) VED = 65 J/mm3 at 7 mm from the beginning of the single track (b) VED = 103 J/mm3 at 3 mm from the beginning of the single track (c) VED = 103 J/mm3 at 7 mm from the beginning of the single track. In the 2D contour, the non-melted region is indicated in blue, and the melted region is indicated by red and green when the VED is 65 J/mm3 and 103 J/mm3 respectively.
Fig. 3 2D melt pool contours from the numerical model compared to experiments [16] for (a) VED = 65 J/mm3 at 7 mm from the beginning of the single track (b) VED = 103 J/mm3 at 3 mm from the beginning of the single track (c) VED = 103 J/mm3 at 7 mm from the beginning of the single track. In the 2D contour, the non-melted region is indicated in blue, and the melted region is indicated by red and green when the VED is 65 J/mm3 and 103 J/mm3 respectively.
Fig. 4 3D temperature contour plots of during single track L-PBF process at time1.8 µs when (a) VED = 65 J/mm3 (b) VED = 103 J/mm3 along with 2D melt pool contours at 5 mm from the laser initial position. In the 2D contour, the non-melted region is indicated in blue, and the melted region is indicated by red and green when the VED is 65 J/mm3 and 103 J/mm3 respectively.
Fig. 4 3D temperature contour plots of during single track L-PBF process at time1.8 µs when (a) VED = 65 J/mm3 (b) VED = 103 J/mm3 along with 2D melt pool contours at 5 mm from the laser initial position. In the 2D contour, the non-melted region is indicated in blue, and the melted region is indicated by red and green when the VED is 65 J/mm3 and 103 J/mm3 respectively.

References

[1] L. Jyothish Kumar, P. M. Pandey, and D. I. Wimpenny, 3D printing and additive
manufacturing technologies. Springer Singapore, 2018. doi: 10.1007/978-981-13-0305-0.
[2] T. DebRoy et al., “Additive manufacturing of metallic components – Process, structure
and properties,” Progress in Materials Science, vol. 92, pp. 112–224, 2018, doi:
10.1016/j.pmatsci.2017.10.001.
[3] C. S. Lefky, B. Zucker, D. Wright, A. R. Nassar, T. W. Simpson, and O. J. Hildreth,
“Dissolvable Supports in Powder Bed Fusion-Printed Stainless Steel,” 3D Printing and
Additive Manufacturing, vol. 4, no. 1, pp. 3–11, 2017, doi: 10.1089/3dp.2016.0043.
[4] J. L. Bartlett and X. Li, “An overview of residual stresses in metal powder bed fusion,”
Additive Manufacturing, vol. 27, no. January, pp. 131–149, 2019, doi:
10.1016/j.addma.2019.02.020.
[5] I. H. Ahn, “Determination of a process window with consideration of effective layer
thickness in SLM process,” International Journal of Advanced Manufacturing
Technology, vol. 105, no. 10, pp. 4181–4191, 2019, doi: 10.1007/s00170-019-04402-w.

[6] R. McCann et al., “In-situ sensing, process monitoring and machine control in Laser
Powder Bed Fusion: A review,” Additive Manufacturing, vol. 45, no. May, 2021, doi:
10.1016/j.addma.2021.102058.
[7] M. Bayat et al., “Keyhole-induced porosities in Laser-based Powder Bed Fusion (L-PBF)
of Ti6Al4V: High-fidelity modelling and experimental validation,” Additive
Manufacturing, vol. 30, no. August, p. 100835, 2019, doi: 10.1016/j.addma.2019.100835.
[8] M. Bayat, S. Mohanty, and J. H. Hattel, “Multiphysics modelling of lack-of-fusion voids
formation and evolution in IN718 made by multi-track/multi-layer L-PBF,” International
Journal of Heat and Mass Transfer, vol. 139, pp. 95–114, 2019, doi:
10.1016/j.ijheatmasstransfer.2019.05.003.
[9] S. D. Jadhav, L. R. Goossens, Y. Kinds, B. van Hooreweder, and K. Vanmeensel, “Laserbased powder bed fusion additive manufacturing of pure copper,” Additive Manufacturing,
vol. 42, no. March, 2021, doi: 10.1016/j.addma.2021.101990.
[10] S. D. Jadhav, S. Dadbakhsh, L. Goossens, J. P. Kruth, J. van Humbeeck, and K.
Vanmeensel, “Influence of selective laser melting process parameters on texture evolution
in pure copper,” Journal of Materials Processing Technology, vol. 270, no. January, pp.
47–58, 2019, doi: 10.1016/j.jmatprotec.2019.02.022.
[11] H. Siva Prasad, F. Brueckner, J. Volpp, and A. F. H. Kaplan, “Laser metal deposition of
copper on diverse metals using green laser sources,” International Journal of Advanced
Manufacturing Technology, vol. 107, no. 3–4, pp. 1559–1568, 2020, doi: 10.1007/s00170-
020-05117-z.
[12] L. R. Goossens, Y. Kinds, J. P. Kruth, and B. van Hooreweder, “On the influence of
thermal lensing during selective laser melting,” Solid Freeform Fabrication 2018:
Proceedings of the 29th Annual International Solid Freeform Fabrication Symposium – An
Additive Manufacturing Conference, SFF 2018, no. December, pp. 2267–2274, 2020.
[13] M. Bayat, V. K. Nadimpalli, D. B. Pedersen, and J. H. Hattel, “A fundamental investigation
of thermo-capillarity in laser powder bed fusion of metals and alloys,” International
Journal of Heat and Mass Transfer, vol. 166, p. 120766, 2021, doi:
10.1016/j.ijheatmasstransfer.2020.120766.
[14] H. Chen, Q. Wei, Y. Zhang, F. Chen, Y. Shi, and W. Yan, “Powder-spreading mechanisms
in powder-bed-based additive manufacturing: Experiments and computational modeling,”
Acta Materialia, vol. 179, pp. 158–171, 2019, doi: 10.1016/j.actamat.2019.08.030.
[15] S. K. Nayak, S. K. Mishra, C. P. Paul, A. N. Jinoop, and K. S. Bindra, “Effect of energy
density on laser powder bed fusion built single tracks and thin wall structures with 100 µm
preplaced powder layer thickness,” Optics and Laser Technology, vol. 125, May 2020, doi:
10.1016/j.optlastec.2019.106016.
[16] G. Nordet et al., “Absorptivity measurements during laser powder bed fusion of pure
copper with a 1 kW cw green laser,” Optics & Laser Technology, vol. 147, no. April 2021,
p. 107612, 2022, doi: 10.1016/j.optlastec.2021.107612.
[17] M. Hummel, C. Schöler, A. Häusler, A. Gillner, and R. Poprawe, “New approaches on
laser micro welding of copper by using a laser beam source with a wavelength of 450 nm,”
Journal of Advanced Joining Processes, vol. 1, no. February, p. 100012, 2020, doi:
10.1016/j.jajp.2020.100012.
[18] M. Hummel, M. Külkens, C. Schöler, W. Schulz, and A. Gillner, “In situ X-ray
tomography investigations on laser welding of copper with 515 and 1030 nm laser beam
sources,” Journal of Manufacturing Processes, vol. 67, no. April, pp. 170–176, 2021, doi:
10.1016/j.jmapro.2021.04.063.
[19] L. Gargalis et al., “Determining processing behaviour of pure Cu in laser powder bed
fusion using direct micro-calorimetry,” Journal of Materials Processing Technology, vol.
294, no. March, p. 117130, 2021, doi: 10.1016/j.jmatprotec.2021.117130.
[20] A. Mondal, D. Agrawal, and A. Upadhyaya, “Microwave heating of pure copper powder
with varying particle size and porosity,” Journal of Microwave Power and
Electromagnetic Energy, vol. 43, no. 1, pp. 4315–43110, 2009, doi:
10.1080/08327823.2008.11688599.

Fig 3. Front view of the ejected powder particles due to the plume movement. Powder particles are colored by their respective temperature while trajectory colors show their magnitude at 0.007 seconds.

316-L 스테인리스강의 레이저 분말 베드 융합 중 콜드 스패터 형성의 충실도 높은 수치 모델링

316-L 스테인리스강의 레이저 분말 베드 융합 중 콜드 스패터 형성의 충실도 높은 수치 모델링

M. BAYAT1,* , AND J. H. HATTEL1

  • Corresponding author
    1 Technical University of Denmark (DTU), Building 425, Kgs. 2800 Lyngby, Denmark

ABSTRACT

Spatter and denudation are two very well-known phenomena occurring mainly during the laser powder bed fusion process and are defined as ejection and displacement of powder particles, respectively. The main driver of this phenomenon is the formation of a vapor plume jet that is caused by the vaporization of the melt pool which is subjected to the laser beam. In this work, a 3-dimensional transient turbulent computational fluid dynamics model coupled with a discrete element model is developed in the finite volume-based commercial software package Flow-3D AM to simulate the spatter phenomenon. The numerical results show that a localized low-pressure zone forms at the bottom side of the plume jet and this leads to a pseudo-Bernoulli effect that drags nearby powder particles into the area of influence of the vapor plume jet. As a result, the vapor plume acts like a momentum sink and therefore all nearby particles point are dragged towards this region. Furthermore, it is noted that due to the jet’s attenuation, powder particles start diverging from the central core region of the vapor plume as they move vertically upwards. It is moreover observed that only particles which are in the very central core region of the plume jet get sufficiently accelerated to depart the computational domain, while the rest of the dragged particles, especially those which undergo an early divergence from the jet axis, get stalled pretty fast as they come in contact with the resting fluid. In the last part of the work, two simulations with two different scanning speeds are carried out, where it is clearly observed that the angle between the departing powder particles and the vertical axis of the plume jet increases with increasing scanning speed.

스패터와 denudation은 주로 레이저 분말 베드 융합 과정에서 발생하는 매우 잘 알려진 두 가지 현상으로 각각 분말 입자의 배출 및 변위로 정의됩니다.

이 현상의 주요 동인은 레이저 빔을 받는 용융 풀의 기화로 인해 발생하는 증기 기둥 제트의 형성입니다. 이 작업에서 이산 요소 모델과 결합된 3차원 과도 난류 ​​전산 유체 역학 모델은 스패터 현상을 시뮬레이션하기 위해 유한 체적 기반 상용 소프트웨어 패키지 Flow-3D AM에서 개발되었습니다.

수치적 결과는 플룸 제트의 바닥면에 국부적인 저압 영역이 형성되고, 이는 근처의 분말 입자를 증기 플룸 제트의 영향 영역으로 끌어들이는 의사-베르누이 효과로 이어진다는 것을 보여줍니다.

결과적으로 증기 기둥은 운동량 흡수원처럼 작용하므로 근처의 모든 입자 지점이 이 영역으로 끌립니다. 또한 제트의 감쇠로 인해 분말 입자가 수직으로 위쪽으로 이동할 때 증기 기둥의 중심 코어 영역에서 발산하기 시작합니다.

더욱이 플룸 제트의 가장 중심 코어 영역에 있는 입자만 계산 영역을 벗어날 만큼 충분히 가속되는 반면, 드래그된 나머지 입자, 특히 제트 축에서 초기 발산을 겪는 입자는 정체되는 것으로 관찰됩니다. 그들은 휴식 유체와 접촉하기 때문에 꽤 빠릅니다.

작업의 마지막 부분에서 두 가지 다른 스캔 속도를 가진 두 가지 시뮬레이션이 수행되었으며, 여기서 출발하는 분말 입자와 연기 제트의 수직 축 사이의 각도가 스캔 속도가 증가함에 따라 증가하는 것이 명확하게 관찰되었습니다.

Fig 1. Two different views of the computational domain for the fluid domain. The vapor plume is simulated by a moving momentum source with a prescribed temperature of 3000 K.
Fig 1. Two different views of the computational domain for the fluid domain. The vapor plume is simulated by a moving momentum source with a prescribed temperature of 3000 K.
Fig 2. (a) and (b) are two snapshots taken at an x-y plane parallel to the powder layer plane before and 0.008 seconds after the start of the scanning process. (c) Shows a magnified view of (b) where detailed powder particles' movement along with their velocity magnitude and directions are shown.
Fig 2. (a) and (b) are two snapshots taken at an x-y plane parallel to the powder layer plane before and 0.008 seconds after the start of the scanning process. (c) Shows a magnified view of (b) where detailed powder particles’ movement along with their velocity magnitude and directions are shown.
Fig 3. Front view of the ejected powder particles due to the plume movement. Powder particles are colored by their respective temperature while trajectory colors show their magnitude at 0.007 seconds.
Fig 3. Front view of the ejected powder particles due to the plume movement. Powder particles are colored by their respective temperature while trajectory colors show their magnitude at 0.007 seconds.

References

[1] T. DebRoy et al., “Additive manufacturing of metallic components – Process, structure
and properties,” Prog. Mater. Sci., vol. 92, pp. 112–224, 2018, doi:
10.1016/j.pmatsci.2017.10.001.
[2] M. Markl and C. Körner, “Multiscale Modeling of Powder Bed–Based Additive
Manufacturing,” Annu. Rev. Mater. Res., vol. 46, no. 1, pp. 93–123, 2016, doi:
10.1146/annurev-matsci-070115-032158.
[3] A. Zinoviev, O. Zinovieva, V. Ploshikhin, V. Romanova, and R. Balokhonov, “Evolution
of grain structure during laser additive manufacturing. Simulation by a cellular automata
method,” Mater. Des., vol. 106, pp. 321–329, 2016, doi: 10.1016/j.matdes.2016.05.125.
[4] Y. Zhang and J. Zhang, “Modeling of solidification microstructure evolution in laser
powder bed fusion fabricated 316L stainless steel using combined computational fluid
dynamics and cellular automata,” Addit. Manuf., vol. 28, no. July 2018, pp. 750–765,
2019, doi: 10.1016/j.addma.2019.06.024.
[5] A. A. Martin et al., “Ultrafast dynamics of laser-metal interactions in additive
manufacturing alloys captured by in situ X-ray imaging,” Mater. Today Adv., vol. 1, p.
100002, 2019, doi: 10.1016/j.mtadv.2019.01.001.
[6] Y. C. Wu et al., “Numerical modeling of melt-pool behavior in selective laser melting
with random powder distribution and experimental validation,” J. Mater. Process.
Technol., vol. 254, no. July 2017, pp. 72–78, 2018, doi:
10.1016/j.jmatprotec.2017.11.032.
[7] W. Gao, S. Zhao, Y. Wang, Z. Zhang, F. Liu, and X. Lin, “Numerical simulation of
thermal field and Fe-based coating doped Ti,” Int. J. Heat Mass Transf., vol. 92, pp. 83–
90, 2016, doi: 10.1016/j.ijheatmasstransfer.2015.08.082.
[8] A. Charles, M. Bayat, A. Elkaseer, L. Thijs, J. H. Hattel, and S. Scholz, “Elucidation of
dross formation in laser powder bed fusion at down-facing surfaces: Phenomenonoriented multiphysics simulation and experimental validation,” Addit. Manuf., vol. 50,
2022, doi: 10.1016/j.addma.2021.102551.
[9] C. Meier, R. W. Penny, Y. Zou, J. S. Gibbs, and A. J. Hart, “Thermophysical phenomena
in metal additive manufacturing by selective laser melting: Fundamentals, modeling,
simulation and experimentation,” arXiv, 2017, doi:
10.1615/annualrevheattransfer.2018019042.
[10] W. King, A. T. Anderson, R. M. Ferencz, N. E. Hodge, C. Kamath, and S. A. Khairallah,
“Overview of modelling and simulation of metal powder bed fusion process at Lawrence
Livermore National Laboratory,” Mater. Sci. Technol. (United Kingdom), vol. 31, no. 8,
pp. 957–968, 2015, doi: 10.1179/1743284714Y.0000000728.

Figure 3: 3D temperature contours and 2D melt pool cross-sections where the melt pool is stabilized at x=500 µm from the start of the laser initial location for cases where (a) absorptivity = 0.1, Recoil pressure coefficient B = 1 and laser beam radius = 12 µm, (b) absorptivity = 0.1, Recoil pressure coefficient B = 20 and laser beam radius = 12 µm, (c) absorptivity = 0.1, Recoil pressure coefficient B = 1 and laser beam radius = 18 µm, (d) absorptivity = 0.45, Recoil pressure coefficient B = 1 and laser beam radius = 18 µm, (e) absorptivity = 0.45, Recoil pressure coefficient B = 20 and laser beam radius = 12 µm, (f) absorptivity = 0.45, Recoil pressure coefficient B = 20 and laser beam radius = 18 µm.

MULTI-PHYSICS NUMERICAL MODELLING OF 316L AUSTENITIC STAINLESS STEEL IN LASER POWDER BED FUSION PROCESS AT MESO-SCALE

W.E. Alphonso1, M.Bayat1,*, M. Baier 2, S. Carmignato2, J.H. Hattel1
1Department of Mechanical Engineering, Technical University of Denmark (DTU), Lyngby, Denmark
2Department of Management and Engineering – University of Padova, Padova, Italy

ABSTRACT

L-PBF(Laser Powder Bed Fusion)는 레이저 열원을 사용하여 선택적으로 통합되는 분말 층으로 복잡한 3D 금속 부품을 만드는 금속 적층 제조(MAM) 기술입니다. 처리 영역은 수십 마이크로미터 정도이므로 L-PBF를 다중 규모 제조 공정으로 만듭니다.

기체 기공의 형성 및 성장 및 용융되지 않은 분말 영역의 생성은 다중물리 모델에 의해 예측할 수 있습니다. 또한 이러한 모델을 사용하여 용융 풀 모양 및 크기, 온도 분포, 용융 풀 유체 흐름 및 입자 크기 및 형태와 같은 미세 구조 특성을 계산할 수 있습니다.

이 작업에서는 용융, 응고, 유체 흐름, 표면 장력, 열 모세관, 증발 및 광선 추적을 통한 다중 반사를 포함하는 스테인리스 스틸 316-L에 대한 충실도 다중 물리학 중간 규모 수치 모델이 개발되었습니다. 완전한 실험 설계(DoE) 방법을 사용하는 통계 연구가 수행되었으며, 여기서 불확실한 재료 특성 및 공정 매개변수, 즉 흡수율, 반동 압력(기화) 및 레이저 빔 크기가 용융수지 모양 및 크기에 미치는 영향을 분석했습니다.

또한 용융 풀 역학에 대한 위에서 언급한 불확실한 입력 매개변수의 중요성을 강조하기 위해 흡수율이 가장 큰 영향을 미치고 레이저 빔 크기가 그 뒤를 잇는 주요 효과 플롯이 생성되었습니다. 용융 풀 크기에 대한 반동 압력의 중요성은 흡수율에 따라 달라지는 용융 풀 부피와 함께 증가합니다.

모델의 예측 정확도는 유사한 공정 매개변수로 생성된 단일 트랙 실험과 시뮬레이션의 용융 풀 모양 및 크기를 비교하여 검증됩니다.

더욱이, 열 렌즈 효과는 레이저 빔 크기를 증가시켜 수치 모델에서 고려되었으며 나중에 결과적인 용융 풀 프로파일은 모델의 견고성을 보여주기 위한 실험과 비교되었습니다.

Laser Powder Bed Fusion (L-PBF) is a Metal Additive Manufacturing (MAM) technology where a complex 3D metal part is built from powder layers, which are selectively consolidated using a laser heat source. The processing zone is in the order of a few tenths of micrometer, making L-PBF a multi-scale manufacturing process. The formation and growth of gas pores and the creation of un-melted powder zones can be predicted by multiphysics models. Also, with these models, the melt pool shape and size, temperature distribution, melt pool fluid flow and its microstructural features like grain size and morphology can be calculated. In this work, a high fidelity multi-physics meso-scale numerical model is developed for stainless steel 316-L which includes melting, solidification, fluid flow, surface tension, thermo-capillarity, evaporation and multiple reflection with ray-tracing. A statistical study using a full Design of Experiments (DoE) method was conducted, wherein the impact of uncertain material properties and process parameters namely absorptivity, recoil pressure (vaporization) and laser beam size on the melt pool shape and size was analysed. Furthermore, to emphasize on the significance of the above mentioned uncertain input parameters on the melt pool dynamics, a main effects plot was created which showed that absorptivity had the highest impact followed by laser beam size. The significance of recoil pressure on the melt pool size increases with melt pool volume which is dependent on absorptivity. The prediction accuracy of the model is validated by comparing the melt pool shape and size from the simulation with single track experiments that were produced with similar process parameters. Moreover, the effect of thermal lensing was considered in the numerical model by increasing the laser beam size and later on the resultant melt pool profile was compared with experiments to show the robustness of the model.

Figure 1: a) Computational domain for single track L-PBF which includes a 200 μm thick substrate and 45 μm powder layer thickness b) 3D temperature contour plot after scanning a single track with melt pool contours at two locations along the scanning direction where the green region indicates the melted regions.
Figure 1: a) Computational domain for single track L-PBF which includes a 200 μm thick substrate and 45 μm powder layer thickness b) 3D temperature contour plot after scanning a single track with melt pool contours at two locations along the scanning direction where the green region indicates the melted regions.
Figure 2: Main effects plot of uncertain parameters: absorptivity, recoil pressure coefficient and laser beam radius on the melt pool dimensions (width and depth)
Figure 2: Main effects plot of uncertain parameters: absorptivity, recoil pressure coefficient and laser beam radius on the melt pool dimensions (width and depth)
Figure 3: 3D temperature contours and 2D melt pool cross-sections where the melt pool is stabilized at x=500 µm from the start of the laser initial location for cases where (a) absorptivity = 0.1, Recoil pressure coefficient B = 1 and laser beam radius = 12 µm, (b) absorptivity = 0.1, Recoil pressure coefficient B = 20 and laser beam radius = 12 µm, (c) absorptivity = 0.1, Recoil pressure coefficient B = 1 and laser beam radius = 18 µm, (d) absorptivity = 0.45, Recoil pressure coefficient B = 1 and laser beam radius = 18 µm, (e) absorptivity = 0.45, Recoil pressure coefficient B = 20 and laser beam radius = 12 µm, (f) absorptivity = 0.45, Recoil pressure coefficient B = 20 and laser beam radius = 18 µm.
Figure 3: 3D temperature contours and 2D melt pool cross-sections where the melt pool is stabilized at x=500 µm from the start of the laser initial location for cases where (a) absorptivity = 0.1, Recoil pressure coefficient B = 1 and laser beam radius = 12 µm, (b) absorptivity = 0.1, Recoil pressure coefficient B = 20 and laser beam radius = 12 µm, (c) absorptivity = 0.1, Recoil pressure coefficient B = 1 and laser beam radius = 18 µm, (d) absorptivity = 0.45, Recoil pressure coefficient B = 1 and laser beam radius = 18 µm, (e) absorptivity = 0.45, Recoil pressure coefficient B = 20 and laser beam radius = 12 µm, (f) absorptivity = 0.45, Recoil pressure coefficient B = 20 and laser beam radius = 18 µm.
Figure 4: Validation of Numerical model with Recoil pressure coefficient B= 20, absorptivity = 0.45 and a) laser beam radius = 15 µm b) laser beam radius = 20 µm
Figure 4: Validation of Numerical model with Recoil pressure coefficient B= 20, absorptivity = 0.45 and a) laser beam radius = 15 µm b) laser beam radius = 20 µm

CONCLUSION

In this work, a high-fidelity multi-physics numerical model was developed for L-PBF using the FVM method in Flow-3D. The impact of uncertainty in the input parameters including absorptivity, recoil pressure and laser beam size on the melt pool is addressed using a DoE method. The DoE analysis shows that absorptivity has the highest impact on the melt pool. The recoil pressure and laser beam size only become significant once absorptivity is 0.45. Furthermore, the numerical model is validated by comparing the predicted melt pool shape and size with experiments conducted with similar process parameters wherein a high prediction accuracy is achieved by the model. In addition, the impact of thermal lensing on the melt pool dimensions by increasing the laser beam spot size is considered in the validated numerical model and the resultant melt pool is compared with experiments.

REFERENCES

[1] T. Bonhoff, M. Schniedenharn, J. Stollenwerk, P. Loosen, Experimental and theoretical analysis of thermooptical effects in protective window for selective laser melting, Proc. Int. Conf. Lasers Manuf. LiM. (2017)
26–29. https://www.wlt.de/lim/Proceedings2017/Data/PDF/Contribution31_final.pdf.
[2] L.R. Goossens, Y. Kinds, J.P. Kruth, B. van Hooreweder, On the influence of thermal lensing during selective
laser melting, Solid Free. Fabr. 2018 Proc. 29th Annu. Int. Solid Free. Fabr. Symp. – An Addit. Manuf. Conf.
SFF 2018. (2020) 2267–2274.
[3] J. Shinjo, C. Panwisawas, Digital materials design by thermal-fluid science for multi-metal additive
manufacturing, Acta Mater. 210 (2021) 116825. https://doi.org/10.1016/j.actamat.2021.116825.
[4] Z. Zhang, Y. Huang, A. Rani Kasinathan, S. Imani Shahabad, U. Ali, Y. Mahmoodkhani, E. Toyserkani, 3-
Dimensional heat transfer modeling for laser powder-bed fusion additive manufacturing with volumetric heat
sources based on varied thermal conductivity and absorptivity, Opt. Laser Technol. 109 (2019) 297–312.
https://doi.org/10.1016/j.optlastec.2018.08.012.
[5] M. Bayat, A. Thanki, S. Mohanty, A. Witvrouw, S. Yang, J. Thorborg, N.S. Tiedje, J.H. Hattel, Keyholeinduced porosities in Laser-based Powder Bed Fusion (L-PBF) of Ti6Al4V: High-fidelity modelling and
experimental validation, Addit. Manuf. 30 (2019) 100835. https://doi.org/10.1016/j.addma.2019.100835.
[6] M. Bayat, S. Mohanty, J.H. Hattel, Multiphysics modelling of lack-of-fusion voids formation and evolution
in IN718 made by multi-track/multi-layer L-PBF, Int. J. Heat Mass Transf. 139 (2019) 95–114.
https://doi.org/10.1016/j.ijheatmasstransfer.2019.05.003.
[7] J. Metelkova, Y. Kinds, K. Kempen, C. de Formanoir, A. Witvrouw, B. Van Hooreweder, On the influence
of laser defocusing in Selective Laser Melting of 316L, Addit. Manuf. 23 (2018) 161–169.
https://doi.org/10.1016/j.addma.2018.08.006.

Figure 2: Temperature contours and melt pool border lines at different times for the 50 % duty cycle case: (a) - (c) Δtcycle = 400 μs, (d) – (f) Δtcycle = 1000 μs and (g) – (i) Δtcycle = 3000 μs.

MULTIPHYSICS SIMULATION OF THEMRAL AND FLUID DYNAMICS PHENOMENA DURING THE PULSED LASER POWDER BED FUSION PROCESS OF 316-L STEEL

M. Bayat* , V. K. Nadimpalli, J. H. Hattel
1Department of Mechanical Engineering, Technical University of Denmark (DTU), Produktionstorvet
425, Kgs. 2800, Lyngby, Denmark

ABSTRACT

L-PBF(Laser Powder Bed Fusion)는 다양한 산업 분야에서 많은 관심을 받았으며, 주로 기존 제조 기술을 사용하여 만들 수 없었던 복잡한 토폴로지 최적화 구성 요소를 구현하는 잘 알려진 능력 덕분입니다. . 펄스 L-PBF(PL-PBF)에서 레이저의 시간적 프로파일은 주기 지속 시간과 듀티 주기 중 하나 또는 둘 다를 수정하여 변조할 수 있습니다. 따라서 레이저의 시간적 프로파일은 향후 적용을 위해 이 프로세스를 더 잘 제어할 수 있는 길을 열어주는 새로운 프로세스 매개변수로 간주될 수 있습니다. 따라서 이 작업에서 우리는 레이저의 시간적 프로파일을 변경하는 것이 PL-PBF 공정에서 용융 풀 조건과 트랙의 최종 모양 및 형상에 어떻게 영향을 미칠 수 있는지 조사하는 것을 목표로 합니다. 이와 관련하여 본 논문에서는 CFD(Computational Fluid Dynamics) 소프트웨어 패키지인 Flow-3D를 기반으로 하는 316-L 스테인리스강 PL-PBF 공정의 다중물리 수치 모델을 개발하고 이 모델을 사용하여 열과 유체를 시뮬레이션합니다. 다양한 펄스 모드에서 공정 과정 중 용융 풀 내부에서 발생하는 유동 조건. 따라서 고정된 레이저 듀티 사이클(50%)이 있는 레이저 주기 지속 시간이 용융 풀의 모양과 크기 및 최종 트랙 형태에 미치는 영향을 연구하기 위해 매개변수 연구가 수행됩니다. 더 긴 주기 기간에서 더 많은 재료가 더 큰 용융 풀 내에서 변위됨에 따라 용융 풀의 후류에 더 눈에 띄는 혹이 형성되며, 동시에 더 심각한 반동 압력을 받습니다. 또한 시뮬레이션에서 50% 듀티 사이클에서 1000μs에서 형성된 보다 대칭적인 용융 풀과 비교하여 400μs 사이클 주기에서 더 긴 용융 풀이 형성된다는 것이 관찰되었습니다. 풀 볼륨은 1000μs의 경우 더 큽니다. 매개변수 연구는 연속 트랙과 파손된 트랙 PL-PBF 사이의 경계를 설명하며, 여기서 연속 트랙은 항상 소량의 용융 재료를 유지함으로써 유지됩니다.

English Abstract

Laser Powder Bed Fusion (L-PBF) has attracted a lot of attention from various industrial sectors and mainly thanks to its well-proven well-known capacity of realizing complex topology-optimized components that have so far been impossible to make using conventional manufacturing techniques. In Pulsed L-PBF (PL-PBF), the laser’s temporal profile can be modulated via modifying either or both the cycle duration and the duty cycle. Thus, the laser’s temporal profile could be considered as a new process parameter that paves the way for a better control of this process for future applications. Therefore, in this work we aim to investigate how changing the laser’s temporal profile can affect the melt pool conditions and the final shape and geometry of a track in the PL-PBF process. In this respect, in this paper a multiphysics numerical model of the PL-PBF process of 316-L stainless steel is developed based on the computational fluid dynamics (CFD) software package Flow-3D and the model is used to simulate the heat and fluid flow conditions occurring inside the melt pool during the course of the process at different pulsing modes. Thus, a parametric study is carried out to study the influence of the laser’s cycle duration with a fixed laser duty cycle (50 %) on the shape and size of the melt pool and the final track morphology. It is noticed that at longer cycle periods, more noticeable humps form at the wake of the melt pool as more material is displaced within bigger melt pools, which are at the same time subjected to more significant recoil pressures. It is also observed in the simulations that at 50 % duty cycle, longer melt pools form at 400 μs cycle period compared to the more symmetrical melt pools formed at 1000 μs, primarily because of shorter laser off-times in the former, even though melt pool volume is bigger for the 1000 μs case. The parameteric study illustrates the boundary between a continuous track and a broken track PL-PBF wherein the continuous track is retained by always maintaining a small volume of molten material.

Figure 1: Front and side views of the computational domain. Note that the region along z and from -100 μm to +50 μm is void.
Figure 1: Front and side views of the computational domain. Note that the region along z and from -100 μm to +50 μm is void.
Figure 2: Temperature contours and melt pool border lines at different times for the 50 % duty cycle case: (a) - (c) Δtcycle = 400 μs, (d) – (f) Δtcycle = 1000 μs and (g) – (i) Δtcycle = 3000 μs.
Figure 2: Temperature contours and melt pool border lines at different times for the 50 % duty cycle case: (a) – (c) Δtcycle = 400 μs, (d) – (f) Δtcycle = 1000 μs and (g) – (i) Δtcycle = 3000 μs.
Figure 3: Plot of melt pool volume versus time for four cases including continuous wave laser as well as 50 % duty cycle at 400 μs, 1000 μs and 3000 μs.
Figure 3: Plot of melt pool volume versus time for four cases including continuous wave laser as well as 50 % duty cycle at 400 μs, 1000 μs and 3000 μs.

CONCLUSIONS

In this work a CFD model of the modulated PL-PBF process of stainless steel 316-L is developed in the commercial software package Flow-3D. The model involves physics such as solidification, melting, evaporation, convection, laser-material interaction, capillarity, Marangoni effect and the recoil pressure effect. In the current study, a parametric study is carried out to understand how the change in the cycle period duration affects the melt pool’s thermo-fluid conditions during the modulated PL-PBF process. It is observed that at the pulse mode with 50 % duty cycle and 400 μs cycle period, an overlapped chain of humps form at the wake of the melt pool and at a spatial frequency of occurrence of about 78 μm. Furthermore and as expected, it is noted that the melt pool volume, the size of the hump as well as the crater size at the end of the track, increase with increase in the cycle period duration, as more material is re-deposited at the back of the melt pool and that itself is caused by more pronounced recoil pressures. Moreover, it is noticed that due to the short off-time period of the laser in the 400 μs cycle period case, there is always an amount of liquid metal left from the previous cycle, at the time the new cycle starts. This is found to be the main reason why longer and elongated melt pools form at 400 μs cycle period, compared to the bigger, shorter and more symmetrical-like melt pools forming at the 1000 μs case. In this study PL-PBF single tracks including the broken track and the continuous track examples were studied to illustrate the boundary of this transition at a given laser scan parameter setting. At higher scan speeds, it is expected that the Plateau–Rayleigh instability will compete with the pulsing behavior to change the transition boundary between a broken and continuous track, which is suggested as future work from this study.

REFERENCES

[1] T. Craeghs, L. Thijs, F. Verhaeghe, J.-P. Kruth, J. Van Humbeeck, A study of the microstructural
evolution during selective laser melting of Ti–6Al–4V, Acta Mater. 58 (2010) 3303–3312.
https://doi.org/10.1016/j.actamat.2010.02.004.
[2] J. Liu, A.T. Gaynor, S. Chen, Z. Kang, K. Suresh, A. Takezawa, L. Li, J. Kato, J. Tang, C.C.L. Wang, L.
Cheng, X. Liang, A.C. To, Current and future trends in topology optimization for additive manufacturing,
(2018) 2457–2483.
[3] M. Bayat, W. Dong, J. Thorborg, A.C. To, J.H. Hattel, A review of multi-scale and multi-physics
simulations of metal additive manufacturing processes with focus on modeling strategies, Addit. Manuf.
47 (2021). https://doi.org/10.1016/j.addma.2021.102278.
[4] A. Foroozmehr, M. Badrossamay, E. Foroozmehr, S. Golabi, Finite Element Simulation of Selective Laser
Melting process considering Optical Penetration Depth of laser in powder bed, Mater. Des. 89 (2016)
255–263. https://doi.org/10.1016/j.matdes.2015.10.002.
[5] Y.S. Lee, W. Zhang, Modeling of heat transfer, fluid flow and solidification microstructure of nickel-base
superalloy fabricated by laser powder bed fusion, Addit. Manuf. 12 (2016) 178–188.
https://doi.org/10.1016/j.addma.2016.05.003.
[6] S.A. Khairallah, A.T. Anderson, A. Rubenchik, W.E. King, Laser powder-bed fusion additive
manufacturing: Physics of complex melt flow and formation mechanisms of pores, spatter, and denudation
zones, Acta Mater. 108 (2016) 36–45. https://doi.org/10.1016/j.actamat.2016.02.014.
[7] M. Bayat, A. Thanki, S. Mohanty, A. Witvrouw, S. Yang, J. Thorborg, N.S. Tiedje, J.H. Hattel, Keyholeinduced porosities in Laser-based Powder Bed Fusion (L-PBF) of Ti6Al4V: High-fidelity modelling and
experimental validation, Addit. Manuf. 30 (2019). https://doi.org/10.1016/j.addma.2019.100835.
[8] A. Charles, M. Bayat, A. Elkaseer, L. Thijs, J.H. Hattel, S. Scholz, Elucidation of dross formation in laser
powder bed fusion at down-facing surfaces: phenomenon-oriented multiphysics simulation and
experimental validation, Addit. Manuf. Under revi (2021).
[9] M. Bayat, V.K. Nadimpalli, D.B. Pedersen, J.H. Hattel, A fundamental investigation of thermo-capillarity
in laser powder bed fusion of metals and alloys, Int. J. Heat Mass Transf. 166 (2021) 120766.
https://doi.org/10.1016/j.ijheatmasstransfer.2020.120766.
[10] J.D. Roehling, S.A. Khairallah, Y. Shen, A. Bayramian, C.D. Boley, A.M. Rubenchik, J. Demuth, N.
Duanmu, M.J. Matthews, Physics of large-area pulsed laser powder bed fusion, Addit. Manuf. 46 (2021) https://doi.org/10.1016/j.addma.2021.102186.
[11] M. Zheng, L. Wei, J. Chen, Q. Zhang, J. Li, S. Sui, G. Wang, W. Huang, Surface morphology evolution
during pulsed selective laser melting: Numerical and experimental investigations, Appl. Surf. Sci. 496
(2019) 143649. https://doi.org/10.1016/j.apsusc.2019.143649.
[12] M. Bayat, V.K. Nadimpalli, D.B. Pedersen, J.H. Hattel, A fundamental investigation of thermo-capillarity
in laser powder bed fusion of metals and alloys, Int. J. Heat Mass Transf. 166 (2021).
https://doi.org/10.1016/j.ijheatmasstransfer.2020.120766.

Figure 3.10: Snapshots of Temperature Profile for Single Track in Keyhole Regime (P = 250W and V = 0.5m/s) at the Preheating Temperature of 100 °C

Multiscale Process Modeling of Residual Deformation and Defect Formation for Laser Powder Bed Fusion Additive Manufacturing

Qian Chen, PhD
University of Pittsburgh, 2021

레이저 분말 베드 퓨전(L-PBF) 적층 제조(AM)는 우수한 기계적 특성으로 그물 모양에 가까운 복잡한 부품을 생산할 수 있습니다. 그러나 빌드 실패 및 다공성과 같은 결함으로 이어지는 원치 않는 잔류 응력 및 왜곡이 L-PBF의 광범위한 적용을 방해하고 있습니다.

L-PBF의 잠재력을 최대한 실현하기 위해 잔류 변형, 용융 풀 및 다공성 형성을 예측하는 다중 규모 모델링 방법론이 개발되었습니다. L-PBF의 잔류 변형 및 응력을 부품 규모에서 예측하기 위해 고유 변형 ​​방법을 기반으로 하는 다중 규모 프로세스 모델링 프레임워크가 제안됩니다.

고유한 변형 벡터는 마이크로 스케일에서 충실도가 높은 상세한 다층 프로세스 시뮬레이션에서 추출됩니다. 균일하지만 이방성인 변형은 잔류 왜곡 및 응력을 예측하기 위해 준 정적 평형 유한 요소 분석(FEA)에서 레이어별로 L-PBF 부품에 적용됩니다.

부품 규모에서의 잔류 변형 및 응력 예측 외에도 분말 규모의 다중물리 모델링을 수행하여 공정 매개변수, 예열 온도 및 스패터링 입자에 의해 유도된 용융 풀 변동 및 결함 형성을 연구합니다. 이러한 요인과 관련된 용융 풀 역학 및 다공성 형성 메커니즘은 시뮬레이션 및 실험을 통해 밝혀졌습니다.

제안된 부품 규모 잔류 응력 및 왜곡 모델을 기반으로 경로 계획 방법은 큰 잔류 변형 및 건물 파손을 방지하기 위해 주어진 형상에 대한 레이저 스캐닝 경로를 조정하기 위해 개발되었습니다.

연속 및 아일랜드 스캐닝 전략을 위한 기울기 기반 경로 계획이 공식화되고 공식화된 컴플라이언스 및 스트레스 최소화 문제에 대한 전체 감도 분석이 수행됩니다. 이 제안된 경로 계획 방법의 타당성과 효율성은 AconityONE L-PBF 시스템을 사용하여 실험적으로 입증되었습니다.

또한 기계 학습을 활용한 데이터 기반 프레임워크를 개발하여 L-PBF에 대한 부품 규모의 열 이력을 예측합니다. 본 연구에서는 실시간 열 이력 예측을 위해 CNN(Convolutional Neural Network)과 RNN(Recurrent Neural Network)을 포함하는 순차적 기계 학습 모델을 제안합니다.

유한 요소 해석과 비교하여 100배의 예측 속도 향상이 달성되어 실제 제작 프로세스보다 빠른 예측이 가능하고 실시간 온도 프로파일을 사용할 수 있습니다.

Laser powder bed fusion (L-PBF) additive manufacturing (AM) is capable of producing complex parts near net shape with good mechanical properties. However, undesired residual stress and distortion that lead to build failure and defects such as porosity are preventing broader applications of L-PBF. To realize the full potential of L-PBF, a multiscale modeling methodology is developed to predict residual deformation, melt pool, and porosity formation. To predict the residual deformation and stress in L-PBF at part-scale, a multiscale process modeling framework based on inherent strain method is proposed.

Inherent strain vectors are extracted from detailed multi-layer process simulation with high fidelity at micro-scale. Uniform but anisotropic strains are then applied to L-PBF part in a layer-by-layer fashion in a quasi-static equilibrium finite element analysis (FEA) to predict residual distortion and stress. Besides residual distortion and stress prediction at part scale, multiphysics modeling at powder scale is performed to study the melt pool variation and defect formation induced by process parameters, preheating temperature and spattering particles. Melt pool dynamics and porosity formation mechanisms associated with these factors are revealed through simulation and experiments.

Based on the proposed part-scale residual stress and distortion model, path planning method is developed to tailor the laser scanning path for a given geometry to prevent large residual deformation and building failures. Gradient based path planning for continuous and island scanning strategy is formulated and full sensitivity analysis for the formulated compliance- and stress-minimization problem is performed.

The feasibility and effectiveness of this proposed path planning method is demonstrated experimentally using the AconityONE L-PBF system. In addition, a data-driven framework utilizing machine learning is developed to predict the thermal history at part-scale for L-PBF.

In this work, a sequential machine learning model including convolutional neural network (CNN) and recurrent neural network (RNN), long shortterm memory unit, is proposed for real-time thermal history prediction. A 100x prediction speed improvement is achieved compared to the finite element analysis which makes the prediction faster than real fabrication process and real-time temperature profile available.

Figure 1.1: Schematic Overview of Metal Laser Powder Bed Fusion Process [2]
Figure 1.1: Schematic Overview of Metal Laser Powder Bed Fusion Process [2]
Figure 1.2: Commercial Powder Bed Fusion Systems
Figure 1.2: Commercial Powder Bed Fusion Systems
Figure 1.3: Commercial Metal Components Fabricated by Powder Bed Fusion Additive Manufacturing: (a) GE Fuel Nozzle; (b) Stryker Hip Biomedical Implant.
Figure 1.3: Commercial Metal Components Fabricated by Powder Bed Fusion Additive Manufacturing: (a) GE Fuel Nozzle; (b) Stryker Hip Biomedical Implant.
Figure 2.1: Proposed Multiscale Process Simulation Framework
Figure 2.1: Proposed Multiscale Process Simulation Framework
Figure 2.2: (a) Experimental Setup for In-situ Thermocouple Measurement in the EOS M290 Build Chamber; (b) Themocouple Locations on the Bottom Side of the Substrate.
Figure 2.2: (a) Experimental Setup for In-situ Thermocouple Measurement in the EOS M290 Build Chamber; (b) Themocouple Locations on the Bottom Side of the Substrate.
Figure 2.3: (a) Finite Element Model for Single Layer Thermal Analysis; (b) Deposition Layer
Figure 2.3: (a) Finite Element Model for Single Layer Thermal Analysis; (b) Deposition Layer
Figure 2.4: Core-skin layer: (a) Surface Morphology; (b) Scanning Strategy; (c) Transient Temperature Distribution and Temperature History at (d) Point 1; (e) Point 2 and (f) Point 3
Figure 2.4: Core-skin layer: (a) Surface Morphology; (b) Scanning Strategy; (c) Transient Temperature Distribution and Temperature History at (d) Point 1; (e) Point 2 and (f) Point 3
Figure 2.5: (a) Scanning Orientation of Each Layer; (b) Finite Element Model for Micro-scale Representative Volume
Figure 2.5: (a) Scanning Orientation of Each Layer; (b) Finite Element Model for Micro-scale Representative Volume
Figure 2.6: Bottom Layer (a) Thermal History; (b) Plastic Strain and (c) Elastic Strain Evolution History
Figure 2.6: Bottom Layer (a) Thermal History; (b) Plastic Strain and (c) Elastic Strain Evolution History
Figure 2.7: Bottom Layer Inherent Strain under Default Process Parameters along Horizontal Scanning Path
Figure 2.7: Bottom Layer Inherent Strain under Default Process Parameters along Horizontal Scanning Path
Figure 2.8: Snapshots of the Element Activation Process
Figure 2.8: Snapshots of the Element Activation Process
Figure 2.9: Double Cantilever Beam Structure Built by the EOS M290 DMLM Process (a) Before and (b) After Cutting off; (c) Faro Laser ScanArm V3 for Distortion Measurement
Figure 2.9: Double Cantilever Beam Structure Built by the EOS M290 DMLM Process (a) Before and (b) After Cutting off; (c) Faro Laser ScanArm V3 for Distortion Measurement
Figure 2.10: Square Canonical Structure Built by the EOS M290 DMLM Process
Figure 2.10: Square Canonical Structure Built by the EOS M290 DMLM Process
Figure 2.11: Finite Element Mesh for the Square Canonical and Snapshots of Element Activation Process
Figure 2.11: Finite Element Mesh for the Square Canonical and Snapshots of Element Activation Process
Figure 2.12: Simulated Distortion Field for the Double Cantilever Beam before Cutting off the Supports: (a) Inherent Strain Method; (b) Simufact Additive 3.1
Figure 2.12: Simulated Distortion Field for the Double Cantilever Beam before Cutting off the Supports: (a) Inherent Strain Method; (b) Simufact Additive 3.1
Figure 3.10: Snapshots of Temperature Profile for Single Track in Keyhole Regime (P = 250W and V = 0.5m/s) at the Preheating Temperature of 100 °C
Figure 3.10: Snapshots of Temperature Profile for Single Track in Keyhole Regime (P = 250W and V = 0.5m/s) at the Preheating Temperature of 100 °C
s) at the Preheating Temperature of 500 °C
s) at the Preheating Temperature of 500 °C
Figure 3.15: Melt Pool Cross Section Comparison Between Simulation and Experiment for Single Track
Figure 3.15: Melt Pool Cross Section Comparison Between Simulation and Experiment for Single Track

Bibliography

[1] I. Astm, ASTM52900-15 Standard Terminology for Additive Manufacturing—General
Principles—Terminology, ASTM International, West Conshohocken, PA 3(4) (2015) 5.
[2] W.E. King, A.T. Anderson, R.M. Ferencz, N.E. Hodge, C. Kamath, S.A. Khairallah, A.M.
Rubenchik, Laser powder bed fusion additive manufacturing of metals; physics, computational,
and materials challenges, Applied Physics Reviews 2(4) (2015) 041304.
[3] W. Yan, Y. Lu, K. Jones, Z. Yang, J. Fox, P. Witherell, G. Wagner, W.K. Liu, Data-driven
characterization of thermal models for powder-bed-fusion additive manufacturing, Additive
Manufacturing (2020) 101503.
[4] K. Dai, L. Shaw, Thermal and stress modeling of multi-material laser processing, Acta
Materialia 49(20) (2001) 4171-4181.
[5] K. Dai, L. Shaw, Distortion minimization of laser-processed components through control of
laser scanning patterns, Rapid Prototyping Journal 8(5) (2002) 270-276.
[6] S.S. Bo Cheng, Kevin Chou, Stress and deformation evaluations of scanning strategy effect in
selective laser melting, Additive Manufacturing (2017).
[7] C. Fu, Y. Guo, Three-dimensional temperature gradient mechanism in selective laser melting
of Ti-6Al-4V, Journal of Manufacturing Science and Engineering 136(6) (2014) 061004.
[8] P. Prabhakar, W.J. Sames, R. Dehoff, S.S. Babu, Computational modeling of residual stress
formation during the electron beam melting process for Inconel 718, Additive Manufacturing 7
(2015) 83-91.
[9] A. Hussein, L. Hao, C. Yan, R. Everson, Finite element simulation of the temperature and
stress fields in single layers built without-support in selective laser melting, Materials & Design
(1980-2015) 52 (2013) 638-647.
[10] P.Z. Qingcheng Yang, Lin Cheng, Zheng Min, Minking Chyu, Albert C. To, articleFinite
element modeling and validation of thermomechanicalbehavior of Ti-6Al-4V in directed energy
deposition additivemanufacturing, Additive Manufacturing (2016).
[11] E.R. Denlinger, J. Irwin, P. Michaleris, Thermomechanical Modeling of Additive
Manufacturing Large Parts, Journal of Manufacturing Science and Engineering 136(6) (2014)
061007.
[12] E.R. Denlinger, M. Gouge, J. Irwin, P. Michaleris, Thermomechanical model development
and in situ experimental validation of the Laser Powder-Bed Fusion process, Additive
Manufacturing 16 (2017) 73-80.
[13] V.J. Erik R Denlinger, G.V. Srinivasan, Tahany EI-Wardany, Pan Michaleris, Thermal
modeling of Inconel 718 processed with powder bed fusionand experimental validation using in
situ measurements, Additive Manufacturing 11 (2016) 7-15.
[14] N. Patil, D. Pal, H.K. Rafi, K. Zeng, A. Moreland, A. Hicks, D. Beeler, B. Stucker, A
Generalized Feed Forward Dynamic Adaptive Mesh Refinement and Derefinement Finite Element
Framework for Metal Laser Sintering—Part I: Formulation and Algorithm Development, Journal
of Manufacturing Science and Engineering 137(4) (2015) 041001.
[15] D. Pal, N. Patil, K.H. Kutty, K. Zeng, A. Moreland, A. Hicks, D. Beeler, B. Stucker, A
Generalized Feed-Forward Dynamic Adaptive Mesh Refinement and Derefinement FiniteElement Framework for Metal Laser Sintering—Part II: Nonlinear Thermal Simulations and
Validations, Journal of Manufacturing Science and Engineering 138(6) (2016) 061003.
[16] N. Keller, V. Ploshikhin, New method for fast predictions of residual stress and distortion of
AM parts, Solid Freeform Fabrication Symposium, Austin, Texas, 2014, pp. 1229-1237.
[17] S.A. Khairallah, A.T. Anderson, A. Rubenchik, W.E. King, Laser powder-bed fusion additive
manufacturing: Physics of complex melt flow and formation mechanisms of pores, spatter, and
denudation zones, Acta Materialia 108 (2016) 36-45.
[18] M.J. Matthews, G. Guss, S.A. Khairallah, A.M. Rubenchik, P.J. Depond, W.E. King,
Denudation of metal powder layers in laser powder bed fusion processes, Acta Materialia 114
(2016) 33-42.
[19] A.A. Martin, N.P. Calta, S.A. Khairallah, J. Wang, P.J. Depond, A.Y. Fong, V. Thampy, G.M.
Guss, A.M. Kiss, K.H. Stone, Dynamics of pore formation during laser powder bed fusion additive
manufacturing, Nature communications 10(1) (2019) 1987.
[20] R. Shi, S.A. Khairallah, T.T. Roehling, T.W. Heo, J.T. McKeown, M.J. Matthews,
Microstructural control in metal laser powder bed fusion additive manufacturing using laser beam
shaping strategy, Acta Materialia (2019).
[21] S.A. Khairallah, A.A. Martin, J.R. Lee, G. Guss, N.P. Calta, J.A. Hammons, M.H. Nielsen,
K. Chaput, E. Schwalbach, M.N. Shah, Controlling interdependent meso-nanosecond dynamics
and defect generation in metal 3D printing, Science 368(6491) (2020) 660-665.
[22] W. Yan, W. Ge, Y. Qian, S. Lin, B. Zhou, W.K. Liu, F. Lin, G.J. Wagner, Multi-physics
modeling of single/multiple-track defect mechanisms in electron beam selective melting, Acta
Materialia 134 (2017) 324-333.
[23] S. Shrestha, Y. Kevin Chou, A Numerical Study on the Keyhole Formation During Laser
Powder Bed Fusion Process, Journal of Manufacturing Science and Engineering 141(10) (2019).
[24] S. Shrestha, B. Cheng, K. Chou, An Investigation into Melt Pool Effective Thermal
Conductivity for Thermal Modeling of Powder-Bed Electron Beam Additive Manufacturing.
[25] D. Rosenthal, Mathematical theory of heat distribution during welding and cutting, Welding
journal 20 (1941) 220-234.
[26] P. Promoppatum, S.-C. Yao, P.C. Pistorius, A.D. Rollett, A comprehensive comparison of the
analytical and numerical prediction of the thermal history and solidification microstructure of
Inconel 718 products made by laser powder-bed fusion, Engineering 3(5) (2017) 685-694.
[27] M. Tang, P.C. Pistorius, J.L. Beuth, Prediction of lack-of-fusion porosity for powder bed
fusion, Additive Manufacturing 14 (2017) 39-48.
[28] T. Moran, P. Li, D. Warner, N. Phan, Utility of superposition-based finite element approach
for part-scale thermal simulation in additive manufacturing, Additive Manufacturing 21 (2018)
215-219.
[29] Y. Yang, M. Knol, F. van Keulen, C. Ayas, A semi-analytical thermal modelling approach
for selective laser melting, Additive Manufacturing 21 (2018) 284-297.
[30] B. Cheng, S. Shrestha, K. Chou, Stress and deformation evaluations of scanning strategy
effect in selective laser melting, Additive Manufacturing 12 (2016) 240-251.
[31] L.H. Ahmed Hussein, Chunze Yan, Richard Everson, Finite element simulation of the
temperature and stress fields in single layers built without-support in selective laser melting,
Materials and Design 52 (2013) 638-647.
[32] H. Peng, D.B. Go, R. Billo, S. Gong, M.R. Shankar, B.A. Gatrell, J. Budzinski, P. Ostiguy,
R. Attardo, C. Tomonto, Part-scale model for fast prediction of thermal distortion in DMLS
additive manufacturing; Part 2: a quasi-static thermo-mechanical model, Austin, Texas (2016).
[33] M.F. Zaeh, G. Branner, Investigations on residual stresses and deformations in selective laser
melting, Production Engineering 4(1) (2010) 35-45.
[34] C. Li, C. Fu, Y. Guo, F. Fang, A multiscale modeling approach for fast prediction of part
distortion in selective laser melting, Journal of Materials Processing Technology 229 (2016) 703-
712.
[35] C. Li, Z. Liu, X. Fang, Y. Guo, On the Simulation Scalability of Predicting Residual Stress
and Distortion in Selective Laser Melting, Journal of Manufacturing Science and Engineering
140(4) (2018) 041013.
[36] S. Afazov, W.A. Denmark, B.L. Toralles, A. Holloway, A. Yaghi, Distortion Prediction and
Compensation in Selective Laser Melting, Additive Manufacturing 17 (2017) 15-22.
[37] Y. Lee, W. Zhang, Modeling of heat transfer, fluid flow and solidification microstructure of
nickel-base superalloy fabricated by laser powder bed fusion, Additive Manufacturing 12 (2016)
178-188.
[38] L. Scime, J. Beuth, A multi-scale convolutional neural network for autonomous anomaly
detection and classification in a laser powder bed fusion additive manufacturing process, Additive
Manufacturing 24 (2018) 273-286.
[39] L. Scime, J. Beuth, Using machine learning to identify in-situ melt pool signatures indicative
of flaw formation in a laser powder bed fusion additive manufacturing process, Additive
Manufacturing 25 (2019) 151-165.
[40] X. Xie, J. Bennett, S. Saha, Y. Lu, J. Cao, W.K. Liu, Z. Gan, Mechanistic data-driven
prediction of as-built mechanical properties in metal additive manufacturing, npj Computational
Materials 7(1) (2021) 1-12.
[41] C. Wang, X. Tan, S. Tor, C. Lim, Machine learning in additive manufacturing: State-of-theart and perspectives, Additive Manufacturing (2020) 101538.
[42] J. Li, R. Jin, Z.Y. Hang, Integration of physically-based and data-driven approaches for
thermal field prediction in additive manufacturing, Materials & Design 139 (2018) 473-485.
[43] M. Mozaffar, A. Paul, R. Al-Bahrani, S. Wolff, A. Choudhary, A. Agrawal, K. Ehmann, J.
Cao, Data-driven prediction of the high-dimensional thermal history in directed energy deposition
processes via recurrent neural networks, Manufacturing letters 18 (2018) 35-39.
[44] A. Paul, M. Mozaffar, Z. Yang, W.-k. Liao, A. Choudhary, J. Cao, A. Agrawal, A real-time
iterative machine learning approach for temperature profile prediction in additive manufacturing
processes, 2019 IEEE International Conference on Data Science and Advanced Analytics (DSAA),
IEEE, 2019, pp. 541-550.
[45] S. Clijsters, T. Craeghs, J.-P. Kruth, A priori process parameter adjustment for SLM process
optimization, Innovative developments on virtual and physical prototyping, Taylor & Francis
Group., 2012, pp. 553-560.
[46] R. Mertens, S. Clijsters, K. Kempen, J.-P. Kruth, Optimization of scan strategies in selective
laser melting of aluminum parts with downfacing areas, Journal of Manufacturing Science and
Engineering 136(6) (2014) 061012.
[47] J.-P. Kruth, J. Deckers, E. Yasa, R. Wauthlé, Assessing and comparing influencing factors of
residual stresses in selective laser melting using a novel analysis method, Proceedings of the
institution of mechanical engineers, Part B: Journal of Engineering Manufacture 226(6) (2012)
980-991.
[48] Y. Lu, S. Wu, Y. Gan, T. Huang, C. Yang, L. Junjie, J. Lin, Study on the microstructure,
mechanical property and residual stress of SLM Inconel-718 alloy manufactured by differing
island scanning strategy, Optics & Laser Technology 75 (2015) 197-206.
[49] E. Foroozmehr, R. Kovacevic, Effect of path planning on the laser powder deposition process:
thermal and structural evaluation, The International Journal of Advanced Manufacturing
Technology 51(5-8) (2010) 659-669.
[50] L.H. Ahmed Hussein, Chunze Yan, Richard Everson, Finite element simulation of the
temperature and stress fields in single layers built without-support in selective laser melting,
Materials and Design (2013).
[51] J.-P. Kruth, M. Badrossamay, E. Yasa, J. Deckers, L. Thijs, J. Van Humbeeck, Part and
material properties in selective laser melting of metals, Proceedings of the 16th international
symposium on electromachining, 2010, pp. 1-12.
[52] L. Thijs, K. Kempen, J.-P. Kruth, J. Van Humbeeck, Fine-structured aluminium products with
controllable texture by selective laser melting of pre-alloyed AlSi10Mg powder, Acta Materialia
61(5) (2013) 1809-1819.
[53] D. Ding, Z.S. Pan, D. Cuiuri, H. Li, A tool-path generation strategy for wire and arc additive
manufacturing, The international journal of advanced manufacturing technology 73(1-4) (2014)
173-183.
[54] B.E. Carroll, T.A. Palmer, A.M. Beese, Anisotropic tensile behavior of Ti–6Al–4V
components fabricated with directed energy deposition additive manufacturing, Acta Materialia
87 (2015) 309-320.
[55] D. Ding, Z. Pan, D. Cuiuri, H. Li, A practical path planning methodology for wire and arc
additive manufacturing of thin-walled structures, Robotics and Computer-Integrated
Manufacturing 34 (2015) 8-19.
[56] D. Ding, Z. Pan, D. Cuiuri, H. Li, S. van Duin, N. Larkin, Bead modelling and implementation
of adaptive MAT path in wire and arc additive manufacturing, Robotics and Computer-Integrated
Manufacturing 39 (2016) 32-42.
[57] R. Ponche, O. Kerbrat, P. Mognol, J.-Y. Hascoet, A novel methodology of design for Additive
Manufacturing applied to Additive Laser Manufacturing process, Robotics and ComputerIntegrated Manufacturing 30(4) (2014) 389-398.
[58] D.E. Smith, R. Hoglund, Continuous fiber angle topology optimization for polymer fused
fillament fabrication, Annu. Int. Solid Free. Fabr. Symp. Austin, TX, 2016.
[59] J. Liu, J. Liu, H. Yu, H. Yu, Concurrent deposition path planning and structural topology
optimization for additive manufacturing, Rapid Prototyping Journal 23(5) (2017) 930-942.
[60] Q. Xia, T. Shi, Optimization of composite structures with continuous spatial variation of fiber
angle through Shepard interpolation, Composite Structures 182 (2017) 273-282.
[61] C. Kiyono, E. Silva, J. Reddy, A novel fiber optimization method based on normal distribution
function with continuously varying fiber path, Composite Structures 160 (2017) 503-515.
[62] C.J. Brampton, K.C. Wu, H.A. Kim, New optimization method for steered fiber composites
using the level set method, Structural and Multidisciplinary Optimization 52(3) (2015) 493-505.
[63] J. Liu, A.C. To, Deposition path planning-integrated structural topology optimization for 3D
additive manufacturing subject to self-support constraint, Computer-Aided Design 91 (2017) 27-
45.
[64] H. Shen, J. Fu, Z. Chen, Y. Fan, Generation of offset surface for tool path in NC machining
through level set methods, The International Journal of Advanced Manufacturing Technology
46(9-12) (2010) 1043-1047.
[65] C. Zhuang, Z. Xiong, H. Ding, High speed machining tool path generation for pockets using
level sets, International Journal of Production Research 48(19) (2010) 5749-5766.
[66] K.C. Mills, Recommended values of thermophysical properties for selected commercial
alloys, Woodhead Publishing2002.
[67] S.S. Sih, J.W. Barlow, The prediction of the emissivity and thermal conductivity of powder
beds, Particulate Science and Technology 22(4) (2004) 427-440.
[68] L. Dong, A. Makradi, S. Ahzi, Y. Remond, Three-dimensional transient finite element
analysis of the selective laser sintering process, Journal of materials processing technology 209(2)
(2009) 700-706.
[69] J.J. Beaman, J.W. Barlow, D.L. Bourell, R.H. Crawford, H.L. Marcus, K.P. McAlea, Solid
freeform fabrication: a new direction in manufacturing, Kluwer Academic Publishers, Norwell,
MA 2061 (1997) 25-49.
[70] G. Bugeda Miguel Cervera, G. Lombera, Numerical prediction of temperature and density
distributions in selective laser sintering processes, Rapid Prototyping Journal 5(1) (1999) 21-26.
[71] T. Mukherjee, W. Zhang, T. DebRoy, An improved prediction of residual stresses and
distortion in additive manufacturing, Computational Materials Science 126 (2017) 360-372.
[72] A.J. Dunbar, E.R. Denlinger, M.F. Gouge, P. Michaleris, Experimental validation of finite
element modeling for laser powderbed fusion deformation, Additive Manufacturing 12 (2016)
108-120.
[73] J. Goldak, A. Chakravarti, M. Bibby, A new finite element model for welding heat sources,
Metallurgical and Materials Transactions B 15(2) (1984) 299-305.
[74] J. Liu, Q. Chen, Y. Zhao, W. Xiong, A. To, Quantitative Texture Prediction of Epitaxial
Columnar Grains in Alloy 718 Processed by Additive Manufacturing, Proceedings of the 9th
International Symposium on Superalloy 718 & Derivatives: Energy, Aerospace, and Industrial
Applications, Springer, 2018, pp. 749-755.
[75] J. Irwin, P. Michaleris, A line heat input model for additive manufacturing, Journal of
Manufacturing Science and Engineering 138(11) (2016) 111004.
[76] M. Gouge, J. Heigel, P. Michaleris, T. Palmer, Modeling forced convection in the thermal
simulation of laser cladding processes, International Journal of Advanced Manufacturing
Technology 79 (2015).
[77] J. Heigel, P. Michaleris, E. Reutzel, Thermo-mechanical model development and validation
of directed energy deposition additive manufacturing of Ti–6Al–4V, Additive manufacturing 5
(2015) 9-19.
[78] E.R. Denlinger, J.C. Heigel, P. Michaleris, Residual stress and distortion modeling of electron
beam direct manufacturing Ti-6Al-4V, Proceedings of the Institution of Mechanical Engineers,
Part B: Journal of Engineering Manufacture 229(10) (2015) 1803-1813.
[79] X. Liang, Q. Chen, L. Cheng, Q. Yang, A. To, A modified inherent strain method for fast
prediction of residual deformation in additive manufacturing of metal parts, 2017 Solid Freeform
Fabrication Symposium Proceedings, Austin, Texas, 2017.
[80] X. Liang, L. Cheng, Q. Chen, Q. Yang, A. To, 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 (2018) 471-486.
[81] L. Sochalski-Kolbus, E.A. Payzant, P.A. Cornwell, T.R. Watkins, S.S. Babu, R.R. Dehoff,
M. Lorenz, O. Ovchinnikova, C. Duty, Comparison of residual stresses in Inconel 718 simple parts
made by electron beam melting and direct laser metal sintering, Metallurgical and Materials
Transactions A 46(3) (2015) 1419-1432.
[82] P. Mercelis, J.-P. Kruth, Residual stresses in selective laser sintering and selective laser
melting, Rapid Prototyping Journal 12(5) (2006) 254-265.
[83] N. Hodge, R. Ferencz, J. Solberg, Implementation of a thermomechanical model for the
simulation of selective laser melting, Computational Mechanics 54(1) (2014) 33-51.
[84] A.S. Wu, D.W. Brown, M. Kumar, G.F. Gallegos, W.E. King, An experimental investigation
into additive manufacturing-induced residual stresses in 316L stainless steel, Metallurgical and
Materials Transactions A 45(13) (2014) 6260-6270.
[85] C. Li, J. liu, Y. Guo, Efficient predictive model of part distortion and residual stress in
selective laser melting, Solid Freeform Fabrication 2016, 2017.
[86] Y. Zhao, Y. Koizumi, K. Aoyagi, D. Wei, K. Yamanaka, A. Chiba, Molten pool behavior and
effect of fluid flow on solidification conditions in selective electron beam melting (SEBM) of a
biomedical Co-Cr-Mo alloy, Additive Manufacturing 26 (2019) 202-214.
[87] J.-H. Cho, S.-J. Na, Implementation of real-time multiple reflection and Fresnel absorption of
laser beam in keyhole, Journal of Physics D: Applied Physics 39(24) (2006) 5372.
[88] Q. Guo, C. Zhao, M. Qu, L. Xiong, L.I. Escano, S.M.H. Hojjatzadeh, N.D. Parab, K. Fezzaa,
W. Everhart, T. Sun, In-situ characterization and quantification of melt pool variation under
constant input energy density in laser powder-bed fusion additive manufacturing process, Additive
Manufacturing (2019).
[89] E. Assuncao, S. Williams, D. Yapp, Interaction time and beam diameter effects on the
conduction mode limit, Optics and Lasers in Engineering 50(6) (2012) 823-828.
[90] R. Cunningham, C. Zhao, N. Parab, C. Kantzos, J. Pauza, K. Fezzaa, T. Sun, A.D. Rollett,
Keyhole threshold and morphology in laser melting revealed by ultrahigh-speed x-ray imaging,
Science 363(6429) (2019) 849-852.
[91] W. Tan, N.S. Bailey, Y.C. Shin, Investigation of keyhole plume and molten pool based on a
three-dimensional dynamic model with sharp interface formulation, Journal of Physics D: Applied
Physics 46(5) (2013) 055501.
[92] W. Tan, Y.C. Shin, Analysis of multi-phase interaction and its effects on keyhole dynamics
with a multi-physics numerical model, Journal of Physics D: Applied Physics 47(34) (2014)
345501.
[93] R. Fabbro, K. Chouf, Keyhole modeling during laser welding, Journal of applied Physics
87(9) (2000) 4075-4083.
[94] Q. Guo, C. Zhao, M. Qu, L. Xiong, S.M.H. Hojjatzadeh, L.I. Escano, N.D. Parab, K. Fezzaa,
T. Sun, L. Chen, In-situ full-field mapping of melt flow dynamics in laser metal additive
manufacturing, Additive Manufacturing 31 (2020) 100939.
[95] Y. Ueda, K. Fukuda, K. Nakacho, S. Endo, A new measuring method of residual stresses with
the aid of finite element method and reliability of estimated values, Journal of the Society of Naval
Architects of Japan 1975(138) (1975) 499-507.
[96] M.R. Hill, D.V. Nelson, The inherent strain method for residual stress determination and its
application to a long welded joint, ASME-PUBLICATIONS-PVP 318 (1995) 343-352.
[97] H. Murakawa, Y. Luo, Y. Ueda, Prediction of welding deformation and residual stress by
elastic FEM based on inherent strain, Journal of the society of Naval Architects of Japan 1996(180)
(1996) 739-751.
[98] M. Yuan, Y. Ueda, Prediction of residual stresses in welded T-and I-joints using inherent
strains, Journal of Engineering Materials and Technology, Transactions of the ASME 118(2)
(1996) 229-234.
[99] L. Zhang, P. Michaleris, P. Marugabandhu, Evaluation of applied plastic strain methods for
welding distortion prediction, Journal of Manufacturing Science and Engineering 129(6) (2007)
1000-1010.
[100] M. Bugatti, Q. Semeraro, Limitations of the Inherent Strain Method in Simulating Powder
Bed Fusion Processes, Additive Manufacturing 23 (2018) 329-346.
[101] L. Cheng, X. Liang, J. Bai, Q. Chen, J. Lemon, A. To, On Utilizing Topology Optimization
to Design Support Structure to Prevent Residual Stress Induced Build Failure in Laser Powder Bed
Metal Additive Manufacturing, Additive Manufacturing (2019).
[102] Q. Chen, X. Liang, D. Hayduke, J. Liu, L. Cheng, J. Oskin, R. Whitmore, A.C. To, An
inherent strain based multiscale modeling framework for simulating part-scale residual
deformation for direct metal laser sintering, Additive Manufacturing 28 (2019) 406-418.
[103] S. Osher, J.A. Sethian, Fronts propagating with curvature-dependent speed: algorithms based
on Hamilton-Jacobi formulations, Journal of computational physics 79(1) (1988) 12-49.
[104] M.Y. Wang, X. Wang, D. Guo, A level set method for structural topology optimization,
Computer methods in applied mechanics and engineering 192(1) (2003) 227-246.
[105] G. Allaire, F. Jouve, A.-M. Toader, Structural optimization using sensitivity analysis and a
level-set method, Journal of computational physics 194(1) (2004) 363-393.
[106] Y. Wang, Z. Luo, Z. Kang, N. Zhang, A multi-material level set-based topology and shape
optimization method, Computer Methods in Applied Mechanics and Engineering 283 (2015)
1570-1586.
[107] P. Dunning, C. Brampton, H. Kim, Simultaneous optimisation of structural topology and
material grading using level set method, Materials Science and Technology 31(8) (2015) 884-894.
[108] P. Liu, Y. Luo, Z. Kang, Multi-material topology optimization considering interface
behavior via XFEM and level set method, Computer methods in applied mechanics and
engineering 308 (2016) 113-133.
[109] J. Liu, Q. Chen, Y. Zheng, R. Ahmad, J. Tang, Y. Ma, Level set-based heterogeneous object
modeling and optimization, Computer-Aided Design (2019).
[110] J. Liu, Q. Chen, X. Liang, A.C. To, Manufacturing cost constrained topology optimization
for additive manufacturing, Frontiers of Mechanical Engineering 14(2) (2019) 213-221.
[111] Z. Kang, Y. Wang, Integrated topology optimization with embedded movable holes based
on combined description by material density and level sets, Computer methods in applied
mechanics and engineering 255 (2013) 1-13.
[112] P.D. Dunning, H. Alicia Kim, A new hole insertion method for level set based structural
topology optimization, International Journal for Numerical Methods in Engineering 93(1) (2013)
118-134.
[113] J.A. Sethian, A fast marching level set method for monotonically advancing fronts,
Proceedings of the National Academy of Sciences 93(4) (1996) 1591-1595.
[114] J.A. Sethian, Level set methods and fast marching methods: evolving interfaces in
computational geometry, fluid mechanics, computer vision, and materials science, Cambridge
university press1999.
[115] C. Le, J. Norato, T. Bruns, C. Ha, D. Tortorelli, Stress-based topology optimization for
continua, Structural and Multidisciplinary Optimization 41(4) (2010) 605-620.
[116] A. Takezawa, G.H. Yoon, S.H. Jeong, M. Kobashi, M. Kitamura, Structural topology
optimization with strength and heat conduction constraints, Computer Methods in Applied
Mechanics and Engineering 276 (2014) 341-361.
[117] S. Hochreiter, J. Schmidhuber, Long short-term memory, Neural computation 9(8) (1997)
1735-1780.
[118] A. Krizhevsky, I. Sutskever, G.E. Hinton, Imagenet classification with deep convolutional
neural networks, Advances in neural information processing systems 25 (2012) 1097-1105.
[119] K. Simonyan, A. Zisserman, Very deep convolutional networks for large-scale image
recognition, arXiv preprint arXiv:1409.1556 (2014).
[120] K. He, X. Zhang, S. Ren, J. Sun, Deep residual learning for image recognition, Proceedings
of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770-778.
[121] O. Russakovsky, J. Deng, H. Su, J. Krause, S. Satheesh, S. Ma, Z. Huang, A. Karpathy, A.
Khosla, M. Bernstein, Imagenet large scale visual recognition challenge, International journal of
computer vision 115(3) (2015) 211-252.
[122] S. Ren, K. He, R. Girshick, J. Sun, Faster r-cnn: Towards real-time object detection with
region proposal networks, Advances in neural information processing systems 28 (2015) 91-99.
[123] E.J. Schwalbach, S.P. Donegan, M.G. Chapman, K.J. Chaput, M.A. Groeber, A discrete
source model of powder bed fusion additive manufacturing thermal history, Additive
Manufacturing 25 (2019) 485-498.
[124] D.G. Duffy, Green’s functions with applications, Chapman and Hall/CRC2015.
[125] J. Martínez-Frutos, D. Herrero-Pérez, Efficient matrix-free GPU implementation of fixed
grid finite element analysis, Finite Elements in Analysis and Design 104 (2015) 61-71.
[126] F. Dugast, P. Apostolou, A. Fernandez, W. Dong, Q. Chen, S. Strayer, R. Wicker, A.C. To,
Part-scale thermal process modeling for laser powder bed fusion with matrix-free method and GPU
computing, Additive Manufacturing 37 (2021) 101732.
[127] A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A.N. Gomez, Ł. Kaiser, I.
Polosukhin, Attention is all you need, Advances in neural information processing systems, 2017,
pp. 5998-6008.
[128] J. Devlin, M.-W. Chang, K. Lee, K. Toutanova, Bert: Pre-training of deep bidirectional
transformers for language understanding, arXiv preprint arXiv:1810.04805 (2018).

Figure 2. (a) Scanning electron microscopy images of Ti6Al4V powder particles and (b) simulated powder bed using discrete element modelling

Laser Powder Bed에서 Laser Drilling에 의한 Keyhole 형성 Ti6Al4V 생체 의학 합금의 융합: 메조스코픽 전산유체역학 시뮬레이션 대 경험적 검증을 사용한 수학적 모델링

Keyhole Formation by Laser Drilling in Laser Powder Bed Fusion of Ti6Al4V Biomedical Alloy: Mesoscopic Computational Fluid Dynamics Simulation versus Mathematical Modelling Using Empirical Validation

Asif Ur Rehman 1,2,3,*
,† , Muhammad Arif Mahmood 4,*
,† , Fatih Pitir 1
, Metin Uymaz Salamci 2,3
,
Andrei C. Popescu 4 and Ion N. Mihailescu 4

Abstract

LPBF(Laser Powder Bed fusion) 공정에서 작동 조건은 열 분포를 기반으로 레이저 유도 키홀 영역을 결정하는 데 필수적입니다. 얕은 구멍과 깊은 구멍으로 분류되는 이러한 영역은 LPBF 프로세스에서 확률과 결함 형성 강도를 제어합니다.

LPBF 프로세스의 핵심 구멍을 연구하고 제어하기 위해 수학적 및 CFD(전산 유체 역학) 모델이 제공됩니다. CFD의 경우 이산 요소 모델링 기법을 사용한 유체 체적 방법이 사용되었으며, 분말 베드 보이드 및 표면에 의한 레이저 빔 흡수를 포함하여 수학적 모델이 개발되었습니다.

동적 용융 풀 거동을 자세히 살펴봅니다. 실험적, CFD 시뮬레이션 및 분석적 컴퓨팅 결과 간에 정량적 비교가 수행되어 좋은 일치를 얻습니다.

LPBF에서 레이저 조사 영역 주변의 온도는 높은 내열성과 분말 입자 사이의 공기로 인해 분말층 주변에 비해 급격히 상승하여 레이저 횡방향 열파의 이동이 느려집니다. LPBF에서 키홀은 에너지 밀도에 의해 제어되는 얕고 깊은 키홀 모드로 분류될 수 있습니다. 에너지 밀도를 높이면 얕은 키홀 구멍 모드가 깊은 키홀 구멍 모드로 바뀝니다.

깊은 키홀 구멍의 에너지 밀도는 다중 반사와 키홀 구멍 내의 2차 반사 빔의 집중으로 인해 더 높아져 재료가 빠르게 기화됩니다.

깊은 키홀 구멍 모드에서는 온도 분포가 높기 때문에 액체 재료가 기화 온도에 가까우므로 얕은 키홀 구멍보다 구멍이 형성될 확률이 훨씬 높습니다. 온도가 급격히 상승하면 재료 밀도가 급격히 떨어지므로 비열과 융해 잠열로 인해 유체 부피가 증가합니다.

그 대가로 표면 장력을 낮추고 용융 풀 균일성에 영향을 미칩니다.

In the laser powder bed fusion (LPBF) process, the operating conditions are essential in determining laser-induced keyhole regimes based on the thermal distribution. These regimes, classified into shallow and deep keyholes, control the probability and defects formation intensity in the LPBF process. To study and control the keyhole in the LPBF process, mathematical and computational fluid dynamics (CFD) models are presented. For CFD, the volume of fluid method with the discrete element modeling technique was used, while a mathematical model was developed by including the laser beam absorption by the powder bed voids and surface. The dynamic melt pool behavior is explored in detail. Quantitative comparisons are made among experimental, CFD simulation and analytical computing results leading to a good correspondence. In LPBF, the temperature around the laser irradiation zone rises rapidly compared to the surroundings in the powder layer due to the high thermal resistance and the air between the powder particles, resulting in a slow travel of laser transverse heat waves. In LPBF, the keyhole can be classified into shallow and deep keyhole mode, controlled by the energy density. Increasing the energy density, the shallow keyhole mode transforms into the deep keyhole mode. The energy density in a deep keyhole is higher due to the multiple reflections and concentrations of secondary reflected beams within the keyhole, causing the material to vaporize quickly. Due to an elevated temperature distribution in deep keyhole mode, the probability of pores forming is much higher than in a shallow keyhole as the liquid material is close to the vaporization temperature. When the temperature increases rapidly, the material density drops quickly, thus, raising the fluid volume due to the specific heat and fusion latent heat. In return, this lowers the surface tension and affects the melt pool uniformity.

Keywords: laser powder bed fusion; computational fluid dynamics; analytical modelling; shallow
and deep keyhole modes; experimental correlation

Figure 1. Powder bed schematic with voids.
Figure 1. Powder bed schematic with voids.
Figure 2. (a) Scanning electron microscopy images of Ti6Al4V powder particles and (b) simulated powder bed using discrete element modelling
Figure 2. (a) Scanning electron microscopy images of Ti6Al4V powder particles and (b) simulated powder bed using discrete element modelling
Figure 3. Temperature field contour formation at various time intervals (a) 0.695 ms, (b) 0.795 ms, (c) 0.995 ms and (d) 1.3 ms.
Figure 3. Temperature field contour formation at various time intervals (a) 0.695 ms, (b) 0.795 ms, (c) 0.995 ms and (d) 1.3 ms.
Figure 4. Detailed view of shallow depth melt mode with temperature field at 0.695 ms
Figure 4. Detailed view of shallow depth melt mode with temperature field at 0.695 ms
Figure 5. Melt flow stream traces formation at various time intervals (a) 0.695 ms, (b) 0.795 ms, (c) 0.995 ms and (d) 1.3 ms
Figure 5. Melt flow stream traces formation at various time intervals (a) 0.695 ms, (b) 0.795 ms, (c) 0.995 ms and (d) 1.3 ms
Figure 6. Density evolution of the melt pool at various time intervals (a) 0.695 ms, (b) 0.795 ms, (c) 0.995 ms and (d) 1.3 ms.
Figure 6. Density evolution of the melt pool at various time intervals (a) 0.695 ms, (b) 0.795 ms, (c) 0.995 ms and (d) 1.3 ms.
Figure 7. Un-melted and melted regions at different time intervals (a) 0.695 ms, (b) 0.795 ms, (c) 0.995 ms and (d) 1.3 ms
Figure 7. Un-melted and melted regions at different time intervals (a) 0.695 ms, (b) 0.795 ms, (c) 0.995 ms and (d) 1.3 ms
Figure 8. Transformation from shallow depth melt flow to deep keyhole formation when laser power increased from (a) 170 W to (b) 200 W
Figure 8. Transformation from shallow depth melt flow to deep keyhole formation when laser power increased from (a) 170 W to (b) 200 W
Figure 9. Stream traces and laser beam multiple reflections in deep keyhole melt flow mode
Figure 9. Stream traces and laser beam multiple reflections in deep keyhole melt flow mode
Figure 10. A comparison between analytical and CFD simulation results for peak thermal distribution value in the deep keyhole formation
Figure 10. A comparison between analytical and CFD simulation results for peak thermal distribution value in the deep keyhole formation
Figure 11. A comparison among experiments [49], CFD and analytical simulations for deep keyhole top width and bottom width
Figure 11. A comparison among experiments [49], CFD and analytical simulations for deep keyhole top width and bottom width

References

  1. Kok, Y.; Tan, X.P.; Wang, P.; Nai, M.L.S.; Loh, N.H.; Liu, E.; Tor, S.B. Anisotropy and heterogeneity of microstructure and
    mechanical properties in metal additive manufacturing: A critical review. Mater. Des. 2018, 139, 565–586. [CrossRef]
  2. Ansari, P.; Salamci, M.U. On the selective laser melting based additive manufacturing of AlSi10Mg: The process parameter
    investigation through multiphysics simulation and experimental validation. J. Alloys Compd. 2022, 890, 161873. [CrossRef]
  3. Guo, N.; Leu, M.C. Additive manufacturing: Technology, applications and research needs. Front. Mech. Eng. 2013, 8, 215–243.
    [CrossRef]
  4. Mohsin Raza, M.; Lo, Y.L. Experimental investigation into microstructure, mechanical properties, and cracking mechanism of
    IN713LC processed by laser powder bed fusion. Mater. Sci. Eng. A 2021, 819, 141527. [CrossRef]
  5. Dezfoli, A.R.A.; Lo, Y.L.; Raza, M.M. Prediction of Epitaxial Grain Growth in Single-Track Laser Melting of IN718 Using Integrated
    Finite Element and Cellular Automaton Approach. Materials 2021, 14, 5202. [CrossRef]
  6. Tiwari, S.K.; Pande, S.; Agrawal, S.; Bobade, S.M. Selection of selective laser sintering materials for different applications. Rapid
    Prototyp. J. 2015, 21, 630–648. [CrossRef]
  7. Liu, F.H. Synthesis of bioceramic scaffolds for bone tissue engineering by rapid prototyping technique. J. Sol-Gel Sci. Technol.
    2012, 64, 704–710. [CrossRef]
  8. Ur Rehman, A.; Sglavo, V.M. 3D printing of geopolymer-based concrete for building applications. Rapid Prototyp. J. 2020, 26,
    1783–1788. [CrossRef]
  9. Ur Rehman, A.; Sglavo, V.M. 3D printing of Portland cement-containing bodies. Rapid Prototyp. J. 2021. ahead of print. [CrossRef]
  10. Popovich, A.; Sufiiarov, V. Metal Powder Additive Manufacturing. In New Trends in 3D Printing; InTech: Rijeka, Croatia, 2016.
  11. Jia, T.; Zhang, Y.; Chen, J.K.; He, Y.L. Dynamic simulation of granular packing of fine cohesive particles with different size
    distributions. Powder Technol. 2012, 218, 76–85. [CrossRef]
  12. Ansari, P.; Ur Rehman, A.; Pitir, F.; Veziroglu, S.; Mishra, Y.K.; Aktas, O.C.; Salamci, M.U. Selective Laser Melting of 316L
    Austenitic Stainless Steel: Detailed Process Understanding Using Multiphysics Simulation and Experimentation. Metals 2021,
    11, 1076. [CrossRef]
  13. Ur Rehman, A.; Tingting, L.; Liao, W. 4D Printing; Printing Ceramics from Metals with Selective Oxidation. Patent No.
    W0/2019/052128, 21 March 2019.
  14. Ullah, A.; Wu, H.; Ur Rehman, A.; Zhu, Y.; Liu, T.; Zhang, K. Influence of laser parameters and Ti content on the surface
    morphology of L-PBF fabricated Titania. Rapid Prototyp. J. 2021, 27, 71–80. [CrossRef]
  15. Ur Rehman, A. Additive Manufacturing of Ceramic Materials and Combinations with New Laser Strategies. Master’s Thesis,
    Nanjing University of Science and Technology, Nanjing, China, 2017.
  16. Wong, K.V.; Hernandez, A. A Review of Additive Manufacturing. ISRN Mech. Eng. 2012, 2012, 1–10. [CrossRef]
  17. Körner, C. Additive manufacturing of metallic components by selective electron beam melting—A review. Int. Mater. Rev. 2016,
    61, 361–377. [CrossRef]
  18. Fayazfar, H.; Salarian, M.; Rogalsky, A.; Sarker, D.; Russo, P.; Paserin, V.; Toyserkani, E. A critical review of powder-based additive
    manufacturing of ferrous alloys: Process parameters, microstructure and mechanical properties. Mater. Des. 2018, 144, 98–128.
    [CrossRef]
  19. Everton, S.K.; Hirsch, M.; Stavroulakis, P.I.; Leach, R.K.; Clare, A.T. Review of in-situ process monitoring and in-situ metrology
    for metal additive manufacturing. Mater. Des. 2016, 95, 431–445. [CrossRef]
  20. Sing, S.L.; An, J.; Yeong, W.Y.; Wiria, F.E. Laser and electron-beam powder-bed additive manufacturing of metallic implants: A
    review on processes, materials and designs. J. Orthop. Res. 2016, 34, 369–385. [CrossRef] [PubMed]
  21. Olakanmi, E.O.; Cochrane, R.F.; Dalgarno, K.W. A review on selective laser sintering/melting (SLS/SLM) of aluminium alloy
    powders: Processing, microstructure, and properties. Prog. Mater. Sci. 2015, 74, 401–477. [CrossRef]
  22. Mahmood, M.A.; Popescu, A.C.; Hapenciuc, C.L.; Ristoscu, C.; Visan, A.I.; Oane, M.; Mihailescu, I.N. Estimation of clad geometry
    and corresponding residual stress distribution in laser melting deposition: Analytical modeling and experimental correlations.
    Int. J. Adv. Manuf. Technol. 2020, 111, 77–91. [CrossRef]
  23. Mahmood, M.A.; Popescu, A.C.; Oane, M.; Ristoscu, C.; Chioibasu, D.; Mihai, S.; Mihailescu, I.N. Three-jet powder flow
    and laser–powder interaction in laser melting deposition: Modelling versus experimental correlations. Metals 2020, 10, 1113.
    [CrossRef]
  24. King, W.; Anderson, A.T.; Ferencz, R.M.; Hodge, N.E.; Kamath, C.; Khairallah, S.A. Overview of modelling and simulation of
    metal powder bed fusion process at Lawrence Livermore National Laboratory. Mater. Sci. Technol. 2015, 31, 957–968. [CrossRef]
  1. Gong, H.; Rafi, K.; Gu, H.; Starr, T.; Stucker, B. Analysis of defect generation in Ti-6Al-4V parts made using powder bed fusion
    additive manufacturing processes. Addit. Manuf. 2014, 1, 87–98. [CrossRef]
  2. Frazier, W.E. Metal additive manufacturing: A review. J. Mater. Eng. Perform. 2014, 23, 1917–1928. [CrossRef]
  3. Panwisawas, C.; Qiu, C.L.; Sovani, Y.; Brooks, J.W.; Attallah, M.M.; Basoalto, H.C. On the role of thermal fluid dynamics into the
    evolution of porosity during selective laser melting. Scr. Mater. 2015, 105, 14–17. [CrossRef]
  4. Yan, W.; Ge, W.; Qian, Y.; Lin, S.; Zhou, B.; Liu, W.K.; Lin, F.; Wagner, G.J. Multi-physics modeling of single/multiple-track defect
    mechanisms in electron beam selective melting. Acta Mater. 2017, 134, 324–333. [CrossRef]
  5. Qian, Y.; Yan, W.; Lin, F. Parametric study and surface morphology analysis of electron beam selective melting. Rapid Prototyp. J.
    2018, 24, 1586–1598. [CrossRef]
  6. Panwisawas, C.; Perumal, B.; Ward, R.M.; Turner, N.; Turner, R.P.; Brooks, J.W.; Basoalto, H.C. Keyhole formation and thermal
    fluid flow-induced porosity during laser fusion welding in titanium alloys: Experimental and modelling. Acta Mater. 2017, 126,
    251–263. [CrossRef]
  7. King, W.E.; Barth, H.D.; Castillo, V.M.; Gallegos, G.F.; Gibbs, J.W.; Hahn, D.E.; Kamath, C.; Rubenchik, A.M. Observation of
    keyhole-mode laser melting in laser powder-bed fusion additive manufacturing. J. Mater. Process. Technol. 2014, 214, 2915–2925.
    [CrossRef]
  8. Panwisawas, C.; Sovani, Y.; Turner, R.P.; Brooks, J.W.; Basoalto, H.C.; Choquet, I. Modelling of thermal fluid dynamics for fusion
    welding. J. Mater. Process. Technol. 2018, 252, 176–182. [CrossRef]
  9. Martin, A.A.; Calta, N.P.; Hammons, J.A.; Khairallah, S.A.; Nielsen, M.H.; Shuttlesworth, R.M.; Sinclair, N.; Matthews, M.J.;
    Jeffries, J.R.; Willey, T.M.; et al. Ultrafast dynamics of laser-metal interactions in additive manufacturing alloys captured by in situ
    X-ray imaging. Mater. Today Adv. 2019, 1, 100002. [CrossRef]
  10. Cunningham, R.; Zhao, C.; Parab, N.; Kantzos, C.; Pauza, J.; Fezzaa, K.; Sun, T.; Rollett, A.D. Keyhole threshold and morphology
    in laser melting revealed by ultrahigh-speed x-ray imaging. Science 2019, 363, 849–852. [CrossRef] [PubMed]
  11. Tang, C.; Tan, J.L.; Wong, C.H. A numerical investigation on the physical mechanisms of single track defects in selective laser
    melting. Int. J. Heat Mass Transf. 2018, 126, 957–968. [CrossRef]
  12. Mirkoohi, E.; Ning, J.; Bocchini, P.; Fergani, O.; Chiang, K.-N.; Liang, S. Thermal Modeling of Temperature Distribution in Metal
    Additive Manufacturing Considering Effects of Build Layers, Latent Heat, and Temperature-Sensitivity of Material Properties. J.
    Manuf. Mater. Process. 2018, 2, 63. [CrossRef]
  13. Oane, M.; Sporea, D. Temperature profiles modeling in IR optical components during high power laser irradiation. Infrared Phys.
    Technol. 2001, 42, 31–40. [CrossRef]
  14. Cleary, P.W.; Sawley, M.L. DEM modelling of industrial granular flows: 3D case studies and the effect of particle shape on hopper
    discharge. Appl. Math. Model. 2002, 26, 89–111. [CrossRef]
  15. Parteli, E.J.R.; Pöschel, T. Particle-based simulation of powder application in additive manufacturing. Powder Technol. 2016, 288,
    96–102. [CrossRef]
  16. Cao, L. Numerical simulation of the impact of laying powder on selective laser melting single-pass formation. Int. J. Heat Mass
    Transf. 2019, 141, 1036–1048. [CrossRef]
  17. Tian, Y.; Yang, L.; Zhao, D.; Huang, Y.; Pan, J. Numerical analysis of powder bed generation and single track forming for selective
    laser melting of SS316L stainless steel. J. Manuf. Process. 2020, 58, 964–974. [CrossRef]
  18. Lee, Y.S.; Zhang, W. Modeling of heat transfer, fluid flow and solidification microstructure of nickel-base superalloy fabricated by
    laser powder bed fusion. Addit. Manuf. 2016, 12, 178–188. [CrossRef]
  19. Tang, M.; Pistorius, P.C.; Beuth, J.L. Prediction of lack-of-fusion porosity for powder bed fusion. Addit. Manuf. 2017, 14, 39–48.
    [CrossRef]
  20. Promoppatum, P.; Yao, S.C.; Pistorius, P.C.; Rollett, A.D. A Comprehensive Comparison of the Analytical and Numerical
    Prediction of the Thermal History and Solidification Microstructure of Inconel 718 Products Made by Laser Powder-Bed Fusion.
    Engineering 2017, 3, 685–694. [CrossRef]
  21. Rosenthal, D. Mathematical Theory of Heat Distribution During Welding and Cutting. Weld. J. 1941, 20, 220–234.
  22. Chen, Q.; Zhao, Y.Y.; Strayer, S.; Zhao, Y.Y.; Aoyagi, K.; Koizumi, Y.; Chiba, A.; Xiong, W.; To, A.C. Elucidating the Effect
    of Preheating Temperature on Melt Pool Morphology Variation in Inconel 718 Laser Powder Bed Fusion via Simulation and
    Experiment. Available online: https://www.sciencedirect.com/science/article/pii/S2214860420310149#bb8 (accessed on 30
    April 2021).
  23. Ur Rehman, A.; Pitir, F.; Salamci, M.U. Laser Powder Bed Fusion (LPBF) of In718 and the Impact of Pre-Heating at 500 and
    1000 ◦C: Operando Study. Materials 2021, 14, 6683. [CrossRef] [PubMed]
  24. Ur Rehman, A.; Pitir, F.; Salamci, M.U. Full-Field Mapping and Flow Quantification of Melt Pool Dynamics in Laser Powder Bed
    Fusion of SS316L. Materials 2021, 14, 6264. [CrossRef] [PubMed]
  25. Gong, H.; Gu, H.; Zeng, K.; Dilip, J.J.S.; Pal, D.; Stucker, B.; Christiansen, D.; Beuth, J.; Lewandowski, J.J. Melt Pool Characterization
    for Selective Laser Melting of Ti-6Al-4V Pre-alloyed Powder. In Proceedings of the International Solid Freeform Fabrication
    Symposium, Austin, TX, USA, 10–12 August 2014; 2014; pp. 256–267.
  26. Song, B.; Dong, S.; Liao, H.; Coddet, C. Process parameter selection for selective laser melting of Ti6Al4V based on temperature
    distribution simulation and experimental sintering. Int. J. Adv. Manuf. Technol. 2012, 61, 967–974. [CrossRef]
  27. Guo, Q.; Zhao, C.; Qu, M.; Xiong, L.; Hojjatzadeh, S.M.H.; Escano, L.I.; Parab, N.D.; Fezzaa, K.; Sun, T.; Chen, L. In-situ full-field
  28. mapping of melt flow dynamics in laser metal additive manufacturing. Addit. Manuf. 2020, 31, 100939. [CrossRef]
  29. Messler, J.R.W. Principles of Welding: Processes, Physics, Chemistry, and Metallurgy; John Wiley & Sons: New York, NY, USA, 2008;
  30. ISBN 9783527617494.
  31. Khairallah, S.A.; Anderson, A.T.; Rubenchik, A.M.; King, W.E. Laser powder-bed fusion additive manufacturing: Physics of
  32. complex melt flow and formation mechanisms of pores, spatter, and denudation zones. Acta Mater. 2016, 108, 36–45. [CrossRef]
  33. Ur Rehman, A.; Mahmood, M.A.; Pitir, F.; Salamci, M.U.; Popescu, A.C.; Mihailescu, I.N. Mesoscopic Computational Fluid
  34. Dynamics Modelling for the Laser-Melting Deposition of AISI 304 Stainless Steel Single Tracks with Experimental Correlation: A
  35. Novel Study. Metals 2021, 11, 1569. [CrossRef]
  36. Paul, A.; Debroy, T. Free surface flow and heat transfer in conduction mode laser welding. Metall. Trans. B 1988, 19, 851–858.
  37. [CrossRef]
  38. Aucott, L.; Dong, H.; Mirihanage, W.; Atwood, R.; Kidess, A.; Gao, S.; Wen, S.; Marsden, J.; Feng, S.; Tong, M.; et al. Revealing
  39. internal flow behaviour in arc welding and additive manufacturing of metals. Nat. Commun. 2018, 9, 5414. [CrossRef]
  40. Abderrazak, K.; Bannour, S.; Mhiri, H.; Lepalec, G.; Autric, M. Numerical and experimental study of molten pool formation
  41. during continuous laser welding of AZ91 magnesium alloy. Comput. Mater. Sci. 2009, 44, 858–866. [CrossRef]
  42. Bayat, M.; Thanki, A.; Mohanty, S.; Witvrouw, A.; Yang, S.; Thorborg, J.; Tiedje, N.S.; Hattel, J.H. Keyhole-induced porosities in
  43. Laser-based Powder Bed Fusion (L-PBF) of Ti6Al4V: High-fidelity modelling and experimental validation. Addit. Manuf. 2019,
  44. 30, 100835. [CrossRef]
Fig. 5 Comparison of experimental SEM image and CtFD simulated melt pool with beam diameters of(a)700 μm,(b)1000 μm, and(c)1300 μm and an absorption rate of 0.3. Electron beam power and scan speed are 900 W and 100 mm s-1, respectively

추가 생산용 전자빔 조사에 의한 316L 스테인리스 용융 · 응고 거동

Melting and Solidification Behavior of 316L Steel Induced by Electron-Beam Irradiation for Additive Manufacturing

付加製造用電子ビーム照射による 316L ステンレス鋼の溶融・凝固挙動

奥 川 将 行*・宮 田 雄一朗*・王     雷*・能 勢 和 史*
小 泉 雄一郎*・中 野 貴 由*
Masayuki OKUGAWA, Yuichiro MIYATA, Lei WANG, Kazufumi NOSE,
Yuichiro KOIZUMI and Takayoshi NAKANO

Abstract

적층 제조(AM) 기술은 복잡한 형상의 3D 부품을 쉽게 만들고 미세 구조 제어를 통해 재료 특성을 크게 제어할 수 있기 때문에 많은 관심을 받았습니다. PBF(Powderbed fusion) 방식의 AM 공정에서는 금속 분말을 레이저나 전자빔으로 녹이고 응고시키는 과정을 반복하여 3D 부품을 제작합니다.

일반적으로 응고 미세구조는 Hunt[Mater. 과학. 영어 65, 75(1984)]. 그러나 CET 이론이 일반 316L 스테인리스강에서도 높은 G와 R로 인해 PBF형 AM 공정에 적용될 수 있을지는 불확실하다.

본 연구에서는 미세구조와 응고 조건 간의 관계를 밝히기 위해 전자빔 조사에 의해 유도된 316L 강의 응고 미세구조를 분석하고 CtFD(Computational Thermal-Fluid Dynamics) 방법을 사용하여 고체/액체 계면에서의 응고 조건을 평가했습니다.

CET 이론과 반대로 높은 G 조건에서 등축 결정립이 종종 형성되는 것으로 밝혀졌다. CtFD 시뮬레이션은 약 400 mm s-1의 속도까지 유체 흐름이 있음을 보여 주며 수상 돌기의 파편 및 이동의 영향으로 등축 결정립이 형성됨을 시사했습니다.

Additive manufacturing(AM)technologies have attracted much attention because it enables us to build 3D parts with complicated geometry easily and control material properties significantly via the control of microstructures. In the powderbed fusion(PBF)type AM process, 3D parts are fabricated by repeating a process of melting and solidifying metal powders by laser or electron beams. In general, the solidification microstructures can be predicted from solidification conditions defined by the combination of temperature gradient G and solidification rate R on the basis of columnar-equiaxed transition(CET)theory proposed by Hunt [Mater. Sci. Eng. 65, 75(1984)]. However, it is unclear whether the CET theory can be applied to the PBF type AM process because of the high G and R, even for general 316L stainless steel. In this study, to reveal relationships between microstructures and solidification conditions, we have analyzed solidification microstructures of 316L steel induced by electronbeam irradiation and evaluated solidification conditions at the solid/liquid interface using a computational thermal-fluid dynamics (CtFD)method. It was found that equiaxed grains were often formed under high G conditions contrary to the CET theory. CtFD simulation revealed that there is a fluid flow up to a velocity of about 400 mm s-1, and suggested that equiaxed grains are formed owing to the effect of fragmentations and migrations of dendrites.

Keywords

Additive Manufacturing, 316L Stainless Steel, Powder Bed Fusion, Electron Beam Melting, Computational Thermal
Fluid Dynamics Simulation

Fig. 1 Width, height, and height differences calculated from laser microscope analysis of melt tracks formed by scanning electron beam. Fig. 2(a)Scanning electron microscope(SEM)image and(b) corresponding electron back-scattering diffraction(EBSD) IPF-map taken from the electron-beam irradiated region in P900-V100 sample. Fig. 3 Average grain size and their aspect ratio calculated from EBSD IPF-map taken from the electron-beam irradiated region.
Fig. 1 Width, height, and height differences calculated from laser microscope analysis of melt tracks formed by scanning electron beam. Fig. 2(a)Scanning electron microscope(SEM)image and(b) corresponding electron back-scattering diffraction(EBSD) IPF-map taken from the electron-beam irradiated region in P900-V100 sample. Fig. 3 Average grain size and their aspect ratio calculated from EBSD IPF-map taken from the electron-beam irradiated region.
Fig. 4 Comparison of experimental SEM image and computational thermal fluid dynamics(CtFD)simulated melt pool with a beam diameter of 700 μm and absorption rates of(a)0.3,(b)0.5, and (c)0.7. Electron beam power and scan speed are 900 W and 100 mm s-1, respectively.
Fig. 4 Comparison of experimental SEM image and computational thermal fluid dynamics(CtFD)simulated melt pool with a beam diameter of 700 μm and absorption rates of(a)0.3,(b)0.5, and (c)0.7. Electron beam power and scan speed are 900 W and 100 mm s-1, respectively.
Fig. 5 Comparison of experimental SEM image and CtFD simulated melt pool with beam diameters of(a)700 μm,(b)1000 μm, and(c)1300 μm and an absorption rate of 0.3. Electron beam power and scan speed are 900 W and 100 mm s-1, respectively
Fig. 5 Comparison of experimental SEM image and CtFD simulated melt pool with beam diameters of(a)700 μm,(b)1000 μm, and(c)1300 μm and an absorption rate of 0.3. Electron beam power and scan speed are 900 W and 100 mm s-1, respectively
Fig. 6 Depth of melt tracks calculated from experimental SEM image and CtFD simulation results.
Fig. 6 Depth of melt tracks calculated from experimental SEM image and CtFD simulation results.
Fig. 7 G-R plots of 316L steel colored by(a)aspect ratio of crystalline grains and(b)fluid velocity.
Fig. 7 G-R plots of 316L steel colored by(a)aspect ratio of crystalline grains and(b)fluid velocity.
Fig. 8 Comparison of solidification microstructure(EBSD IPF-map)of melt region formed by scanning electron beam and corresponding snap shot of CtFD simulation colored by fluid velocity
Fig. 8 Comparison of solidification microstructure(EBSD IPF-map)of melt region formed by scanning electron beam and corresponding snap shot of CtFD simulation colored by fluid velocity

References

1) M.C. Sow, T. De Terris, O. Castelnau, Z. Hamouche, F. Coste, R.
Fabbro and P. Peyre: “Influence of beam diameter on Laser Powder

Bed Fusion(L-PBF)process”, Addit. Manuf. 36(2020), 101532.
2) J.C. Simmons, X. Chen, A. Azizi, M.A. Daeumer, P.Y. Zavalij, G.
Zhou and S.N. Schiffres: “Influence of processing and microstructure
on the local and bulk thermal conductivity of selective laser melted
316L stainless steel”, Addit. Manuf. 32(2020), 100996.
3) S. Dryepondt, P. Nandwana, P. Fernandez-Zelaia and F. List:
“Microstructure and High Temperature Tensile properties of 316L
Fabricated by Laser Powder-Bed Fusion”, Addit. Manuf. 37(2020),
101723.
4) S.H. Sun, T. Ishimoto, K. Hagihara, Y. Tsutsumi, T. Hanawa and T.
Nakano: “Excellent mechanical and corrosion properties of austenitic
stainless steel with a unique crystallographic lamellar microstructure
via selective laser melting”, Scr. Mater. 159(2019), 89-93.
5) T. Ishimoto, S. Wu, Y. Ito, S.H. Sun, H. Amano and T. Nakano:
“Crystallographic orientation control of 316L austenitic stainless
steel via selective laser melting”, ISIJ Int. 60(2020), 1758-1764.
6) T. Ishimoto, K. Hagihara, K. Hisamoto, S.H. Sun and T. Nakano:
“Crystallographic texture control of beta-type Ti-15Mo-5Zr3Al alloy by selective laser melting for the development of novel
implants with a biocompatible low Young’s modulus”, Scr. Mater.
132(2017), 34-38.
7) X. Ding, Y. Koizumi, D. Wei and A. Chiba: “Effect of process
parameters on melt pool geometry and microstructure development
for electron beam melting of IN718: A systematic single bead
analysis study”, Addit. Manuf. 26(2019), 215-226.
8) K. Karayagiz, L. Johnson, R. Seede, V. Attari, B. Zhang, X. Huang,
S. Ghosh, T. Duong, I. Karaman, A. Elwany and R. Arróyave: “Finite
interface dissipation phase field modeling of Ni-Nb under additive
manufacturing conditions”, Acta Mater. 185(2020), 320-339.
9) M.M. Kirka, Y. Lee, D.A. Greeley, A. Okello, M.J. Goin, M.T.
Pearce and R.R. Dehoff: “Strategy for Texture Management in
Metals Additive Manufacturing”, JOM, 69(2017), 523-531.
10) S.S. Babu, N. Raghavan, J. Raplee, S.J. Foster, C. Frederick, M.
Haines, R. Dinwiddie, M.K. Kirka, A. Plotkowski, Y. Lee and
R.R. Dehoff: “Additive Manufacturing of Nickel Superalloys:
Opportunities for Innovation and Challenges Related to
Qualification”, Metall. Mater. Trans. A. 49(2018), 3764-3780.
11) M.R. Gotterbarm, A.M. Rausch and C. Körner: “Fabrication of
Single Crystals through a μ-Helix Grain Selection Process during
Electron Beam Metal Additive Manufacturing”, Metals, 10(2020),
313.
12) J.D.D. Hunt: “Steady state columnar and equiaxed growth of
dendrites and eutectic”, Mater. Sci. Eng. 65(1984), 75-83.
13) S. Bontha, N.W. Klingbeil, P.A. Kobryn and H.L. Fraser: “Effects of
process variables and size-scale on solidification microstructure in
beam-based fabrication of bulky 3D structures”, Mater. Sci. Eng. A.
513-514(2009), 311-318.
14) J. Gockel and J. Beuth: “Understanding Ti-6Al-4V microstructure
control in additive manufacturing via process maps”, 24th Int. SFF
Symp. – An Addit. Manuf. Conf. SFF 2013.(2013), 666-674.
15) B. Schoinochoritis, D. Chantzis and K. Salonitis: “Simulation of
metallic powder bed additive manufacturing processes with the finite
element method: A critical review”, Proc. of Instit. Mech. Eng., Part
B: J. Eng. Manuf. 231(2017), 96-117.
16)小泉雄一郎: “計算機シミュレーションを用いたAdditive
Manufacturing プロセス最適化予測”, スマートプロセス学会誌,
8-4(2019), 132-138.
17) Y. Zhao, Y. Koizumi, K. Aoyagi, D. Wei, K. Yamanaka and A. Chiba:
“Molten pool behavior and effect of fluid flow on solidification
conditions in selective electron beam melting(SEBM)of a
biomedical Co-Cr-Mo alloy”, Addit. Manuf. 26(2019), 202-214.
18) C. Tang, J.L. Tan and C.H. Wong: “A numerical investigation on
the physical mechanisms of single track defects in selective laser
melting”, Int. J. Heat Mass Transf. 126(2018), 957-968.
19) Technical data for Iron, [Online]. Available: http://periodictable.com/
Elements/026/data.html. [Accessed: 8-Feb-2021].
20) N. Raghavan, R. Dehoff, S. Pannala, S. Simunovic, M. Kirka, J.
Turner, N. Carl-son and S.S. Babu: “Numerical modeling of heattransfer and the influence of process parameters on tailoring the grain
morphology of IN718 in electron beam additive manufacturing”,
Acta Mater. 112(2016), 303-314.
21) S. Morita, Y. Miki and K. Toishi: “Introduction of Dendrite
Fragmentation to Microstructure Calculation by Cellular Automaton
Method”, Tetsu-to-Hagane. 104(2018), 559-566.
22) H. Esaka and M. Tamura: “Model Experiment Using Succinonitrile
on the Formation of Equiaxed Grains caused by Forced Convection”,
Tetsu-to-Hagane. 86(2000), 252-258.

Fig. 1. Schematic of (a) geometry of the simulation model, (b) A-A cross-section presenting the locations of point probes for recording temperature history (unit: µm).

Laser powder bed fusion of 17-4 PH stainless steel: a comparative study on the effect of heat treatment on the microstructure evolution and mechanical properties

17-4 PH 스테인리스강의 레이저 분말 베드 융합: 열처리가 미세조직의 진화 및 기계적 특성에 미치는 영향에 대한 비교 연구

panelS.Saboonia, A.Chaboka, S.Fenga,e, H.Blaauwb, T.C.Pijperb,c, H.J.Yangd, Y.T.Peia
aDepartment of Advanced Production Engineering, Engineering and Technology Institute Groningen, University of Groningen, Nijenborgh 4, 9747 AG, Groningen, The Netherlands
bPhilips Personal Care, Oliemolenstraat 5, 9203 ZN, Drachten, The Netherlands
cInnovation Cluster Drachten, Nipkowlaan 5, 9207 JA, Drachten, The Netherlands
dShi-changxu Innovation Center for Advanced Materials, Institute of Metal Research, Chinese Academy of Sciences, 72 Wenhua Road, Shenyang 110016, P. R. China
eSchool of Mechanical Engineering, University of Science and Technology Beijing, Beijing, 100083, P.R. China

Abstract

17-4 PH (precipitation hardening) stainless steel is commonly used for the fabrication of complicated molds with conformal cooling channels using laser powder bed fusion process (L-PBF). However, their microstructure in the as-printed condition varies notably with the chemical composition of the feedstock powder, resulting in different age-hardening behavior. In the present investigation, 17-4 PH stainless steel components were fabricated by L-PBF from two different feedstock powders, and subsequently subjected to different combinations of post-process heat treatments. It was observed that the microstructure in as-printed conditions could be almost fully martensitic or ferritic, depending on the ratio of Creq/Nieq of the feedstock powder. Aging treatment at 480 °C improved the yield and ultimate tensile strengths of the as-printed components. However, specimens with martensitic structures exhibited accelerated age-hardening response compared with the ferritic specimens due to the higher lattice distortion and dislocation accumulation, resulting in the “dislocation pipe diffusion mechanism”. It was also found that the martensitic structures were highly susceptible to the formation of reverted austenite during direct aging treatment, where 19.5% of austenite phase appeared in the microstructure after 15 h of direct aging. Higher fractions of reverted austenite activates the transformation induced plasticity and improves the ductility of heat treated specimens. The results of the present study can be used to tailor the microstructure of the L-PBF printed 17-4 PH stainless steel by post-process heat treatments to achieve a good combination of mechanical properties.

17-4 PH(석출 경화) 스테인리스강은 레이저 분말 베드 융합 공정(L-PBF)을 사용하여 등각 냉각 채널이 있는 복잡한 금형 제작에 일반적으로 사용됩니다. 그러나 인쇄된 상태의 미세 구조는 공급원료 분말의 화학적 조성에 따라 크게 달라지므로 시효 경화 거동이 다릅니다.

현재 조사에서 17-4 PH 스테인리스강 구성요소는 L-PBF에 의해 두 가지 다른 공급원료 분말로 제조되었으며, 이후에 다양한 조합의 후처리 열처리를 거쳤습니다. 인쇄된 상태의 미세구조는 공급원료 분말의 Creq/Nieq 비율에 따라 거의 완전히 마르텐사이트 또는 페라이트인 것으로 관찰되었습니다.

480 °C에서 노화 처리는 인쇄된 구성 요소의 수율과 극한 인장 강도를 개선했습니다. 그러나 마텐자이트 구조의 시편은 격자 변형 및 전위 축적이 높아 페라이트 시편에 비해 시효 경화 반응이 가속화되어 “전위 파이프 확산 메커니즘”이 발생합니다.

또한 마르텐사이트 구조는 직접 시효 처리 중에 복귀된 오스테나이트의 형성에 매우 민감한 것으로 밝혀졌으며, 여기서 15시간의 직접 시효 후 미세 조직에 19.5%의 오스테나이트 상이 나타났습니다.

복귀된 오스테나이트의 비율이 높을수록 변형 유도 가소성이 활성화되고 열처리된 시편의 연성이 향상됩니다. 본 연구의 결과는 기계적 특성의 우수한 조합을 달성하기 위해 후처리 열처리를 통해 L-PBF로 인쇄된 17-4 PH 스테인리스강의 미세 구조를 조정하는 데 사용할 수 있습니다.

Keywords

Laser powder bed fusion17-4 PH stainless steelPost-process heat treatmentAge hardeningReverted austenite

Fig. 1. Schematic of (a) geometry of the simulation model, (b) A-A cross-section presenting the locations of point probes for recording temperature history (unit: µm).
Fig. 1. Schematic of (a) geometry of the simulation model, (b) A-A cross-section presenting the locations of point probes for recording temperature history (unit: µm).
Fig. 2. Optical (a, b) and TEM (c) micrographs of the wrought 17-4 PH stainless steel.
Fig. 2. Optical (a, b) and TEM (c) micrographs of the wrought 17-4 PH stainless steel.
Fig. 3. EBSD micrographs of the as-printed 17-4 PH steel fabricated with “powder A” (a, b) and “powder B” (c, d) on two different cross sections: (a, c) perpendicular to the building direction, and (b, d) parallel to the building direction.
Fig. 3. EBSD micrographs of the as-printed 17-4 PH steel fabricated with “powder A” (a, b) and “powder B” (c, d) on two different cross sections: (a, c) perpendicular to the building direction, and (b, d) parallel to the building direction.
Fig. 4. Microstructure of the as-printed 17-4 PH stainless steel fabricated with “powder A” (a) and “powder B” (b).
Fig. 4. Microstructure of the as-printed 17-4 PH stainless steel fabricated with “powder A” (a) and “powder B” (b).
Fig. 5. Simulated temperature history of the probes located at the cross section of the L-PBF 17-4 PH steel sample.
Fig. 5. Simulated temperature history of the probes located at the cross section of the L-PBF 17-4 PH steel sample.
Fig. 6. Dependency of the volume fraction of delta ferrite in the final microstructure of L-PBF printed 17-4 PH steel as a function of Creq/Nieq.
Fig. 6. Dependency of the volume fraction of delta ferrite in the final microstructure of L-PBF printed 17-4 PH steel as a function of Creq/Nieq.
Fig. 7. IQ + IPF (left column), parent austenite grain maps (middle column) and phase maps (right column, green color = martensite, red color = austenite) of the post-process heat treated 17-4 PH stainless steel: (a-c) direct aged, (d-f) HIP + aging, (g-i) SA + Aging, and (j-l) HIP + SA + aging (all sample were printed with “powder A”).
Fig. 7. IQ + IPF (left column), parent austenite grain maps (middle column) and phase maps (right column, green color = martensite, red color = austenite) of the post-process heat treated 17-4 PH stainless steel: (a-c) direct aged, (d-f) HIP + aging, (g-i) SA + Aging, and (j-l) HIP + SA + aging (all sample were printed with “powder A”).
Fig. 8. TEM micrographs of the post process heat treated 17-4 PH stainless steel: (a) direct aging and (b) HIP + aging (printed with “powder A”).
Fig. 8. TEM micrographs of the post process heat treated 17-4 PH stainless steel: (a) direct aging and (b) HIP + aging (printed with “powder A”).
Fig. 9. XRD patterns of the post-process heat treated 17-4 PH stainless steel printed with “powder A”.
Fig. 9. XRD patterns of the post-process heat treated 17-4 PH stainless steel printed with “powder A”.
Fig. 10. (a) Volume fraction of reverted austenite as a function of aging time for “direct aging” condition, (b) phase map (green color = martensite, red color = austenite) of the 15 h direct aged specimen printed with “powder A”.
Fig. 10. (a) Volume fraction of reverted austenite as a function of aging time for “direct aging” condition, (b) phase map (green color = martensite, red color = austenite) of the 15 h direct aged specimen printed with “powder A”.
Fig. 11. Microhardness variations of the “direct aged” specimens as a function of aging time at 480 °C.
Fig. 11. Microhardness variations of the “direct aged” specimens as a function of aging time at 480 °C.
Fig. 12. Kernel average misorientation graphs of the as-printed 17-4 PH steel with (a) martensitic structure (printed with “powder A”) and (b) ferritic structure (printed with “powder b”).
Fig. 12. Kernel average misorientation graphs of the as-printed 17-4 PH steel with (a) martensitic structure (printed with “powder A”) and (b) ferritic structure (printed with “powder b”).
Fig. 13. Typical stress-strain curves (a) along with the yield and ultimate tensile strengths (b) and elongation (c) of the as-printed and post-process heat treated 17-4 PH stainless steel (all sample are fabricated with “powder A”).
Fig. 13. Typical stress-strain curves (a) along with the yield and ultimate tensile strengths (b) and elongation (c) of the as-printed and post-process heat treated 17-4 PH stainless steel (all sample are fabricated with “powder A”).
Fig. 14. (a) IQ + IPF and (b) phase map (green color = martensite, red color = austenite) of the “direct aged” specimen after tensile test at a location nearby the rupture point (tension direction from left to right).
Fig. 14. (a) IQ + IPF and (b) phase map (green color = martensite, red color = austenite) of the “direct aged” specimen after tensile test at a location nearby the rupture point (tension direction from left to right).

References

[1]

P. Bajaj, A. Hariharan, A. Kini, P. Kürnsteiner, D. Raabe, E.A. Jagle

Steels in additive manufacturing: A review of their microstructure and properties

Materials Science and Engineering: A, 772 (2020), Article 138633

ArticleDownload PDFView Record in ScopusGoogle Scholar

[2]

Y. Sun, R.J. Hebert, M. Aindow

Effect of heat treatments on microstructural evolution of additively manufactured and wrought 17-4PH stainless steel

Mater. Des., 156 (2018), pp. 429-440

ArticleDownload PDFView Record in ScopusGoogle Scholar

[3]

Zemin Wang, Xulei Fang, Hui Li, Wenqing Liu

Atom Probe Tomographic Characterization of nanoscale cu-rich Precipitates in 17-4 precipitate hardened stainless steel tempered at different temperatures

Microsc. Microanal., 23 (2017), pp. 340-349

View Record in ScopusGoogle Scholar

[4]

C.N. Hsiao, C.S. Chiou, J.R. Yang

Aging reactions in a 17-4 PH stainless steel

Mater. Chem. Phys., 74 (2002), pp. 134-142

ArticleDownload PDFView Record in ScopusGoogle Scholar

[5]

Hamidreza Riazi, Fakhreddin Ashrafizadeh, Sayed Rahman Hosseini, Reza Ghomashchi

Influence of simultaneous aging and plasma nitriding on fatigue performance of 17-4 PH stainless steel

Mater. Sci. Eng. A, 703 (2017), pp. 262-269

ArticleDownload PDFView Record in ScopusGoogle Scholar

[6]

M.S. Shinde, K.M. Ashtankar

Additive manufacturing–assisted conformal cooling channels in mold manufacturing processes

Adv. Mech. Eng., 9 (2017), pp. 1-14

View Record in ScopusGoogle Scholar

[7]

A. Armillotta, R. Baraggi, S. Fasoli

SLM tooling for die casting with conformal cooling channels

Int. J. Adv. Manuf. Technol., 71 (2014), pp. 573-583

CrossRefView Record in ScopusGoogle Scholar

[8]

Amar M. Kamat, Yutao Pei

An analytical method to predict and compensate for residual stress-induced deformation in overhanging regions of internal channels fabricated using powder bed fusion

Additive Manufacturing, 29 (2019), Article 100796

ArticleDownload PDFView Record in ScopusGoogle Scholar

[9]

K.S. Prakash, T. Nancharaih, V.V. Subba Rao

Additive Manufacturing Techniques in Manufacturing – An Overview

Materials Today: Proceedings, 5 (2018), pp. 3873-3882

ArticleDownload PDFView Record in ScopusGoogle Scholar

[10]

R. Singh, A. Gupta, O. Tripathi, S. Srivastava, B. Singh, A. Awasthi, S.K. Rajput, P. Sonia, P. Singhal, K.K. Saxena

Powder bed fusion process in additive manufacturing: An overview

Materials Today: Proceedings, 26 (2020), pp. 3058-3070

ArticleDownload PDFGoogle Scholar

[11]

L. Zai, Ch Zhang, Y. Wang, W. Guo, D. Wellmann, X. Tong, Y. Tian

Laser Powder Bed Fusion of Precipitation-Hardened Martensitic Stainless Steels: A Review

Metals, 10 (2020), p. 255

CrossRefView Record in ScopusGoogle Scholar

[12]

H. Khalid Rafi, Deepankar Pal, Nachiket Patil, Thomas L. Starr, Brent E. Stucker

Microstructure and Mechanical Behavior of 17-4 Precipitation Hardenable Steel Processed by Selective Laser Melting

J. Mater. Eng. Perf, 23 (2014), pp. 4421-4428

Google Scholar

[13]

A. Yadollahi, N. Shamsaei, S.M. Thompson, A. Elwany, L. Bian

Effects of building orientation and heat treatment on fatigue behavior of selective laser melted 17-4 PH stainless steel

Int. J. Fatigue, 94 (2017), pp. 218-235

ArticleDownload PDFView Record in ScopusGoogle Scholar

[14]

M. Alnajjar, Frederic Christien, Cedric Bosch, Krzysztof Wolski

A comparative study of microstructure and hydrogen embrittlement of selective laser melted and wrought 17–4 PH stainless steel

Materials Science and Engineering: A, 785 (2020), Article 139363

ArticleDownload PDFView Record in ScopusGoogle Scholar

[15]

M. Alnajjar, F. Christien, K. Wolski, C. Bosch

Evidence of austenite by-passing in a stainless steel obtained from laser melting additive manufacturing

Addit. Manuf, 25 (2019), pp. 187-195

ArticleDownload PDFView Record in ScopusGoogle Scholar

[16]

P.D. Nezhadfar, K. Anderson-Wedge, S.R. Daniewicz, N. Phan, Sh Shao, N. Shamsaei

Improved high cycle fatigue performance of additively manufactured 17-4 PH stainless steel via in-process refining micro-/defect-structure

Additive Manufacturing, 36 (2020), Article 101604

ArticleDownload PDFView Record in ScopusGoogle Scholar

[17]

S. Feng, A.M. Kamat, S. Sabooni, Y. Pei

Experimental and numerical investigation of the origin of surface roughness in laser powder bed fused overhang regions

Virtual and Physical Prototyping, 16 (2021), pp. S66-S84, 10.1080/17452759.2021.1896970

CrossRefView Record in ScopusGoogle Scholar

[18]

W. Liu, J. Ma, M. Mazar Atabaki, R. Pillai, B. Kumar, U. Vasudevan, H. Sreshta, R. Kovacevic

Hybrid Laser-arc Welding of 17-4 PH Martensitic Stainless Steel

Lasers in Manufacturing and Materials Processing, 2 (2015), pp. 74-90

CrossRefView Record in ScopusGoogle Scholar

[19]

J.C. Lippold, D.J. Kotecki

Welding metallurgy and weldability of stainless steels

Wiley (2005)

Google Scholar

[20]

M. Shirdel, H. Mirzadeh, M.H. Parsa

Nano/ultrafine grained austenitic stainless steel through the formation and reversion of deformation-induced martensite: Mechanisms, microstructures, mechanical properties, and TRIP effect

Mater. Charact., 103 (2015), pp. 150-161

ArticleDownload PDFView Record in ScopusGoogle Scholar

[21]

S. Kou

Solidification and liquation cracking issues in welding

JOM, 55 (2003), pp. 37-42

CrossRefView Record in ScopusGoogle Scholar

[22]

T.J. Lienert, J.C. Lippold

Improved Weldability Diagram for Pulsed Laser Welded Austenitic Stainless Steels

Sci. Technol. Weld. Join., 8 (2003), pp. 1-9

CrossRefView Record in ScopusGoogle Scholar

[23]

Ch Qiu, M. Al Kindi, A.S. Aladawi, I. Al Hatmi

A comprehensive study on microstructure and tensile behaviour of a selectively laser melted stainless steel

Sci. Rep., 8 (2018), p. 7785

View Record in ScopusGoogle Scholar

[24]

P.A. Hooper

Melt pool temperature and cooling rates in laser powder bed fusion

Addit. Manuf, 22 (2018), pp. 548-559

ArticleDownload PDFView Record in ScopusGoogle Scholar

[25]

T. DebRoy, H.L. Wei, J.S. Zuback, T. Mukherjee, J.W. Elmer, J.O. Milewski, A.M. Beese, A. Wilson-Heid, A. Ded, W. Zhang

Additive manufacturing of metallic components – Process, structure and properties

Prog. Mater. Sci., 92 (2018), pp. 112-224

ArticleDownload PDFView Record in ScopusGoogle Scholar

[26]

S. Vunnam, A. Saboo, Ch Sudbrack, T.L. Starr

Effect of powder chemical composition on the as-built microstructure of 17- 4 PH stainless steel processed by selective laser melting

Additive Manufacturing, 30 (2019), Article 100876

ArticleDownload PDFView Record in ScopusGoogle Scholar

[27]

L. Couturier, F. De Geuser, M. Descoins, A. Deschamps

Evolution of the microstructure of a 15-5PH martensitic stainless steel during precipitation hardening heat treatment

Mater. Des., 107 (2016), pp. 416-425

ArticleDownload PDFView Record in ScopusGoogle Scholar

[28]

C. Cayron, B. Artaud, L. Briottet

Reconstruction of parent grains from EBSD data

Mater. Charact., 57 (2006), pp. 386-401

ArticleDownload PDFView Record in ScopusGoogle Scholar

[29]

R. Bhambroo, S. Roychowdhury, V. Kain, V.S. Raja

Effect of reverted austenite on mechanical properties of precipitation hardenable 17-4 stainless steel

Mater. Sci. Eng. A, 568 (2013), pp. 127-133

ArticleDownload PDFView Record in ScopusGoogle Scholar

[30]

T. LeBrun, T. Nakamoto, K. Horikawa, H. Kobayashi

Effect of retained austenite on subsequent thermal processing and resultant mechanical properties of selective laser melted 17–4 PH stainless steel

Mater. Des., 81 (2015), pp. 44-53

ArticleDownload PDFView Record in ScopusGoogle Scholar

[31]

T.H. Hsu, Y.J. Chang, C.Y. Huang, H.W. Yen, C.P. Chen, K.K. Jen, A.Ch Yeh

Microstructure and property of a selective laser melting process induced oxide dispersion strengthened 17-4 PH stainless steel

J. Alloys. Compd., 803 (2019), pp. 30-41

ArticleDownload PDFView Record in ScopusGoogle Scholar

[32]

Li Wang, Chaofang Dong, Cheng Man, Decheng Kong, Kui Xiao, Xiaogang Li

Enhancing the corrosion resistance of selective laser melted 15-5 PH martensite stainless steel via heat treatment

Corrosion Science, 166 (2020), Article 108427

ArticleDownload PDFView Record in ScopusGoogle Scholar

[33]

H. Kimura

Precipitation Behavior and 2-step Aging of 17-4PH Stainless Steel

Tetsu-to-Hagane, 86 (2000), pp. 343-348

CrossRefView Record in ScopusGoogle Scholar

[34]

G. Yeli, M.A. Auger, K. Wilford, G.D.W. Smith, P.A.J. Bagot, M.P. Moody

Sequential nucleation of phases in a 17-4PH steel: Microstructural characterisation and mechanical properties

Acta. Mater., 125 (2017), pp. 38-49

ArticleDownload PDFView Record in ScopusGoogle Scholar

[35]

J.B. Ferguson, Benjamin F. Schultz, Dev Venugopalan1, Hugo F. Lopez, Pradeep K. Rohatgi, Kyu Cho, Chang-Soo Kim

On the Superposition of Strengthening Mechanisms in Dispersion Strengthened Alloys and Metal-Matrix Nanocomposites: Considerations of Stress and Energy

Met. Mater. Int., 20 (2014), pp. 375-388

CrossRefView Record in ScopusGoogle Scholar

[36]

H. Mirzadeh, A. Najafizadeh

Aging kinetics of 17-4 PH stainless steel

Mater. Chem. Phys., 116 (2009), pp. 119-124

ArticleDownload PDFView Record in ScopusGoogle Scholar

[37]

L.E. Murr, E. Martinez, J. Hernandez, Sh Collins, K.N. Amato, S.M. Gaytan, P.W. Shindo

Microstructures and Properties of 17-4 PH Stainless Steel Fabricated by Selective Laser Melting

J. Mater. Res. Technol, 1 (2012), pp. 167-177

ArticleDownload PDFView Record in ScopusGoogle Scholar

[38]

Y.F. Shen, L.N. Qiu, X. Sun, L. Zuo, P.K. Liaw, D. Raabe

Effects of retained austenite volume fraction, morphology, and carbon content on strength and ductility of nanostructured TRIP-assisted steels

Mater. Sci. Eng. A, 636 (2015), pp. 551-564

ArticleDownload PDFView Record in ScopusGoogle Scholar

Laser powder bed fusion Figure

A study of transient and steady-state regions from single-track deposition in laser powder bed fusion

SubinShrestha KevinChou

J.B. Speed School of Engineering, University of Louisville, Louisville, KY 40292, United States

Abstract

The surface morphology of parts made by the laser powder bed fusion (L-PBF) process is governed by the flow of the melt pool. The nature of the molten metal flow depends on the material properties, process parameters, and powder-bed particles, etc., and may result in potentially significant variations along a laser scanning path.

This study investigates the formation of transient and steady-state zones through a single-track l-PBF experiment using Inconel 625 powder. Single tracks with lengths of 1 mm and 2 mm were fabricated using 195 W laser power and scan speeds of 400 mm/s or 800 mm/s. The surface morphology of the track was analyzed using a white light interferometer (WLI), and an individual single track can be divided into three distinct zones based on the track width and height.

The initial transient region has a wider and taller solidified track geometry, the region near the end of a scan has a tapered profile with a decreasing track width and height, while the steady-state region in the middle has a smaller variation in the width and height.

A mesoscale numerical model was further developed using FLOW-3D to examine the formation of the transient and steady-state zones. At the start of a scan, strong flow occurs outward and backward, leading to the formation of a wider track with a bump. As the scan continues, the thermal gradient stabilizes, leading to a steady state, which resulted in a very small fluctuation in the width. Furthermore, the tapered end of the scan track is due to the half-lemniscate shape of the melt pool during laser scanning.

L-PBF(Laser Powder Bed fusion) 공정으로 만든 부품의 표면 형태는 용융 풀의 흐름에 따라 결정됩니다. 용융 금속 흐름의 특성은 재료 특성, 공정 매개변수 및 분말층 입자 등에 따라 달라지며 레이저 스캐닝 경로를 따라 잠재적으로 상당한 변동이 발생할 수 있습니다.

이 연구는 Inconel 625 분말을 사용하여 단일 트랙 l-PBF 실험을 통해 과도 및 정상 상태 영역의 형성을 조사합니다. 1 mm 및 2 mm 길이의 단일 트랙은 195 W 레이저 출력과 400 mm/s 또는 800 mm/s의 스캔 속도를 사용하여 제작되었습니다. 트랙의 표면 형태는 백색광 간섭계(WLI)를 사용하여 분석되었으며 개별 단일 트랙은 트랙 너비와 높이에 따라 3개의 별개 영역으로 나눌 수 있습니다.

초기 과도 영역은 더 넓고 더 높은 응고된 트랙 형상을 가지며, 스캔 끝 근처의 영역은 트랙 너비와 높이가 감소하는 테이퍼 프로파일을 갖는 반면, 중간의 정상 상태 영역은 너비와 높이에서 더 작은 변동을 가집니다. 신장. 중간 규모 수치 모델은 과도 및 정상 상태 영역의 형성을 조사하기 위해 FLOW-3D를 사용하여 추가로 개발되었습니다.

스캔이 시작될 때 강한 흐름이 바깥쪽과 뒤쪽으로 발생하여 범프가 있는 더 넓은 트랙이 형성됩니다. 스캔이 계속됨에 따라 열 구배가 안정화되어 정상 상태로 이어지며 폭의 변동이 매우 작습니다. 또한 스캔 트랙의 끝이 가늘어지는 것은 레이저 스캔 중 용융 풀의 반-렘니케이트 모양 때문입니다.

A study of transient and steady-state regions from single-track deposition in laser powder bed fusion
A study of transient and steady-state regions from single-track deposition in laser powder bed fusion

Keywords

Additive manufacturing, Laser powder bed fusion, Numerical modelling, Transient region

Figure 6. Evolution of melt pool in the overhang region (θ = 45°, P = 100 W, v = 1000 mm/s, the streamlines are shown by arrows).

Experimental and numerical investigation of the origin of surface roughness in laser powder bed fused overhang regions

레이저 파우더 베드 융합 오버행 영역에서 표면 거칠기의 원인에 대한 실험 및 수치 조사

Shaochuan Feng,Amar M. Kamat,Soheil Sabooni &Yutao PeiPages S66-S84 | Received 18 Jan 2021, Accepted 25 Feb 2021, Published online: 10 Mar 2021

ABSTRACT

Surface roughness of laser powder bed fusion (L-PBF) printed overhang regions is a major contributor to deteriorated shape accuracy/surface quality. This study investigates the mechanisms behind the evolution of surface roughness (Ra) in overhang regions. The evolution of surface morphology is the result of a combination of border track contour, powder adhesion, warp deformation, and dross formation, which is strongly related to the overhang angle (θ). When 0° ≤ θ ≤ 15°, the overhang angle does not affect Ra significantly since only a small area of the melt pool boundaries contacts the powder bed resulting in slight powder adhesion. When 15° < θ ≤ 50°, powder adhesion is enhanced by the melt pool sinking and the increased contact area between the melt pool boundary and powder bed. When θ > 50°, large waviness of the overhang contour, adhesion of powder clusters, severe warp deformation and dross formation increase Ra sharply.

레이저 파우더 베드 퓨전 (L-PBF) 프린팅 오버행 영역의 표면 거칠기는 형상 정확도 / 표면 품질 저하의 주요 원인입니다. 이 연구 는 오버행 영역에서 표면 거칠기 (Ra ) 의 진화 뒤에 있는 메커니즘을 조사합니다 . 표면 형태의 진화는 오버행 각도 ( θ ) 와 밀접한 관련이있는 경계 트랙 윤곽, 분말 접착, 뒤틀림 변형 및 드로스 형성의 조합의 결과입니다 . 0° ≤  θ  ≤ 15° 인 경우 , 용융풀 경계의 작은 영역 만 분말 베드와 접촉하여 약간의 분말 접착이 발생하기 때문에 오버행 각도가 R a에 큰 영향을 주지 않습니다 . 15° < θ 일 때  ≤ 50°, 용융 풀 싱킹 및 용융 풀 경계와 분말 베드 사이의 증가된 접촉 면적으로 분말 접착력이 향상됩니다. θ  > 50° 일 때 오버행 윤곽의 큰 파형, 분말 클러스터의 접착, 심한 휨 변형 및 드 로스 형성이 Ra 급격히 증가 합니다.

KEYWORDS: Laser powder bed fusion (L-PBF), melt pool dynamics, overhang region, shape deviation, surface roughness

1. Introduction

레이저 분말 베드 융합 (L-PBF)은 첨단 적층 제조 (AM) 기술로, 집중된 레이저 빔을 사용하여 금속 분말을 선택적으로 융합하여 슬라이스 된 3D 컴퓨터 지원에 따라 층별로 3 차원 (3D) 금속 부품을 구축합니다. 설계 (CAD) 모델 (Chatham, Long 및 Williams 2019 ; Tan, Zhu 및 Zhou 2020 ). 재료가 인쇄 층 아래에 ​​존재하는지 여부에 따라 인쇄 영역은 각각 솔리드 영역 또는 돌출 영역으로 분류 될 수 있습니다. 따라서 오버행 영역은 고체 기판이 아니라 분말 베드 바로 위에 건설되는 특수 구조입니다 (Patterson, Messimer 및 Farrington 2017). 오버행 영역은지지 구조를 포함하거나 포함하지 않고 구축 할 수 있으며, 지지대가있는 돌출 영역의 L-PBF는 지지체가 더 낮은 밀도로 구축된다는 점을 제외 하고 (Wang and Chou 2018 ) 고체 기판의 공정과 유사합니다 (따라서 기계적 강도가 낮기 때문에 L-PBF 공정 후 기계적으로 쉽게 제거 할 수 있습니다. 따라서지지 구조로 인쇄 된 오버행 영역은 L-PBF 공정 후 지지물 제거, 연삭 및 연마와 같은 추가 후 처리 단계가 필요합니다.

수평 내부 채널의 제작과 같은 일부 특정 경우에는 공정 후 지지대를 제거하기가 어려우므로 채널 상단 절반의 돌출부 영역을 지지대없이 건설해야합니다 (Hopkinson and Dickens 2000 ). 수평 내부 채널에 사용할 수없는지지 구조 외에도 내부 표면, 특히 등각 냉각 채널 (Feng, Kamat 및 Pei 2021 ) 에서 발생하는 복잡한 3D 채널 네트워크의 경우 표면 마감 프로세스를 구현하는 것도 어렵습니다 . 결과적으로 오버행 영역은 (i) 잔류 응력에 의한 변형, (ii) 계단 효과 (Kuo et al. 2020 ; Li et al. 2020 )로 인해 설계된 모양에서 벗어날 수 있습니다 .) 및 (iii) 원하지 않는 분말 소결로 인한 향상된 표면 거칠기; 여기서, 앞의 두 요소는 일반적으로 mm 길이 스케일에서 ‘매크로’편차로 분류되고 후자는 일반적으로 µm 길이 스케일에서 ‘마이크로’편차로 인식됩니다.

열 응력에 의한 변형은 오버행 영역에서 발생하는 중요한 문제입니다 (Patterson, Messimer 및 Farrington 2017 ). 국부적 인 용융 / 냉각은 용융 풀 내부 및 주변에서 큰 온도 구배를 유도하여 응고 된 층에 집중적 인 열 응력을 유발합니다. 열 응력에 의한 뒤틀림은 고체 영역을 현저하게 변형하지 않습니다. 이러한 영역은 아래의 여러 레이어에 의해 제한되기 때문입니다. 반면에 오버행 영역은 구속되지 않고 공정 중 응력 완화로 인해 상당한 변형이 발생합니다 (Kamat 및 Pei 2019 ). 더욱이 용융 깊이는 레이어 두께보다 큽니다 (이전 레이어도 재용 해되어 빌드 된 레이어간에 충분한 결합을 보장하기 때문입니다 [Yadroitsev et al. 2013 ; Kamath et al.2014 ]),응고 된 두께가 설계된 두께보다 크기 때문에형태 편차 (예 : 드 로스 [Charles et al. 2020 ; Feng et al. 2020 ])가 발생합니다. 마이크로 스케일에서 인쇄 된 표면 (R a 및 S a ∼ 10 μm)은 기계적으로 가공 된 표면보다 거칠다 (Duval-Chaneac et al. 2018 ; Wen et al. 2018 ). 이 문제는고형화 된 용융 풀의 가장자리에 부착 된 용융되지 않은 분말의 결과로 표면 거칠기 (R a )가 일반적으로 약 20 μm인 오버행 영역에서 특히 심각합니다 (Mazur et al. 2016 ; Pakkanen et al. 2016 ).

오버행 각도 ( θ , 빌드 방향과 관련하여 측정)는 오버행 영역의 뒤틀림 편향과 표면 거칠기에 영향을 미치는 중요한 매개 변수입니다 (Kamat and Pei 2019 ; Mingear et al. 2019 ). θ ∼ 45 ° 의 오버행 각도 는 일반적으로지지 구조없이 오버행 영역을 인쇄 할 수있는 임계 값으로 합의됩니다 (Pakkanen et al. 2016 ; Kadirgama et al. 2018 ). θ 일 때이 임계 값보다 크면 오버행 영역을 허용 가능한 표면 품질로 인쇄 할 수 없습니다. 오버행 각도 외에도 레이저 매개 변수 (레이저 에너지 밀도와 관련된)는 용융 풀의 모양 / 크기 및 용융 풀 역학에 영향을줌으로써 오버행 영역의 표면 거칠기에 영향을줍니다 (Wang et al. 2013 ; Mingear et al . 2019 ).

용융 풀 역학은 고체 (Shrestha 및 Chou 2018 ) 및 오버행 (Le et al. 2020 ) 영역 모두에서 수행되는 L-PBF 공정을 포함한 레이저 재료 가공의 일반적인 물리적 현상입니다 . 용융 풀 모양, 크기 및 냉각 속도는 잔류 응력으로 인한 변형과 ​​표면 거칠기에 모두 영향을 미치므로 처리 매개 변수와 표면 형태 / 품질 사이의 다리 역할을하며 용융 풀을 이해하기 위해 수치 시뮬레이션을 사용하여 추가 조사를 수행 할 수 있습니다. 거동과 표면 거칠기에 미치는 영향. 현재까지 고체 영역의 L-PBF 동안 용융 풀 동작을 시뮬레이션하기 위해 여러 연구가 수행되었습니다. 유한 요소 방법 (FEM)과 같은 시뮬레이션 기술 (Roberts et al. 2009 ; Du et al.2019 ), 유한 차분 법 (FDM) (Wu et al. 2018 ), 전산 유체 역학 (CFD) (Lee and Zhang 2016 ), 임의의 Lagrangian-Eulerian 방법 (ALE) (Khairallah and Anderson 2014 )을 사용하여 증발 반동 압력 (Hu et al. 2018 ) 및 Marangoni 대류 (Zhang et al. 2018 ) 현상을포함하는 열 전달 (온도 장) 및 물질 전달 (용융 흐름) 프로세스. 또한 이산 요소법 (DEM)을 사용하여 무작위 분산 분말 베드를 생성했습니다 (Lee and Zhang 2016 ; Wu et al. 2018 ). 이 모델은 분말 규모의 L-PBF 공정을 시뮬레이션했습니다 (Khairallah et al. 2016) 메조 스케일 (Khairallah 및 Anderson 2014 ), 단일 트랙 (Leitz et al. 2017 )에서 다중 트랙 (Foroozmehr et al. 2016 ) 및 다중 레이어 (Huang, Khamesee 및 Toyserkani 2019 )로.

그러나 결과적인 표면 거칠기를 결정하는 오버행 영역의 용융 풀 역학은 문헌에서 거의 관심을받지 못했습니다. 솔리드 영역의 L-PBF에 대한 기존 시뮬레이션 모델이 어느 정도 참조가 될 수 있지만 오버행 영역과 솔리드 영역 간의 용융 풀 역학에는 상당한 차이가 있습니다. 오버행 영역에서 용융 금속은 분말 입자 사이의 틈새로 아래로 흘러 용융 풀이 다공성 분말 베드가 제공하는 약한 지지체 아래로 가라 앉습니다. 이것은 중력과 표면 장력의 영향이 용융 풀의 결과적인 모양 / 크기를 결정하는 데 중요하며, 결과적으로 오버행 영역의 마이크로 스케일 형태의 진화에 중요합니다. 또한 분말 입자 사이의 공극, 열 조건 (예 : 에너지 흡수,2019 ; Karimi et al. 2020 ; 노래와 영 2020 ). 표면 거칠기는 (마이크로) 형상 편차를 증가시킬뿐만 아니라 주기적 하중 동안 미세 균열의 시작 지점 역할을함으로써 기계적 강도를 저하시킵니다 (Günther et al. 2018 ). 오버행 영역의 높은 표면 거칠기는 (마이크로) 정확도 / 품질에 대한 엄격한 요구 사항이있는 부품 제조에서 L-PBF의 적용을 제한합니다.

본 연구는 실험 및 시뮬레이션 연구를 사용하여 오버행 영역 (지지물없이 제작)의 미세 형상 편차 형성 메커니즘과 표면 거칠기의 기원을 체계적이고 포괄적으로 조사합니다. 결합 된 DEM-CFD 시뮬레이션 모델은 경계 트랙 윤곽, 분말 접착 및 뒤틀림 변형의 효과를 고려하여 오버행 영역의 용융 풀 역학과 표면 형태의 형성 메커니즘을 나타 내기 위해 개발되었습니다. 표면 거칠기 R의 시뮬레이션 및 단일 요인 L-PBF 인쇄 실험을 사용하여 오버행 각도의 함수로 연구됩니다. 용융 풀의 침몰과 관련된 오버행 영역에서 분말 접착의 세 가지 메커니즘이 식별되고 자세히 설명됩니다. 마지막으로, 인쇄 된 오버행 영역에서 높은 표면 거칠기 문제를 완화 할 수 있는 잠재적 솔루션에 대해 간략하게 설명합니다.

The shape and size of the L-PBF printed samples are illustrated in Figure 1
The shape and size of the L-PBF printed samples are illustrated in Figure 1
Figure 2. Borders in the overhang region depending on the overhang angle θ
Figure 2. Borders in the overhang region depending on the overhang angle θ
Figure 3. (a) Profile of the volumetric heat source, (b) the model geometry of single-track printing on a solid substrate (unit: µm), and (c) the comparison of melt pool dimensions obtained from the experiment (right half) and simulation (left half) for a calibrated optical penetration depth of 110 µm (laser power 200 W and scan speed 800 mm/s, solidified layer thickness 30 µm, powder size 10–45 µm).
Figure 3. (a) Profile of the volumetric heat source, (b) the model geometry of single-track printing on a solid substrate (unit: µm), and (c) the comparison of melt pool dimensions obtained from the experiment (right half) and simulation (left half) for a calibrated optical penetration depth of 110 µm (laser power 200 W and scan speed 800 mm/s, solidified layer thickness 30 µm, powder size 10–45 µm).
Figure 4. The model geometry of an overhang being L-PBF processed: (a) 3D view and (b) right view.
Figure 4. The model geometry of an overhang being L-PBF processed: (a) 3D view and (b) right view.
Figure 5. The cross-sectional contour of border tracks in a 45° overhang region.
Figure 5. The cross-sectional contour of border tracks in a 45° overhang region.
Figure 6. Evolution of melt pool in the overhang region (θ = 45°, P = 100 W, v = 1000 mm/s, the streamlines are shown by arrows).
Figure 6. Evolution of melt pool in the overhang region (θ = 45°, P = 100 W, v = 1000 mm/s, the streamlines are shown by arrows).
Figure 7. The overhang contour is contributed by (a) only outer borders when θ ≤ 60° (b) both inner borders and outer borders when θ > 60°.
Figure 7. The overhang contour is contributed by (a) only outer borders when θ ≤ 60° (b) both inner borders and outer borders when θ > 60°.
Figure 8. Schematic of powder adhesion on a 45° overhang region.
Figure 8. Schematic of powder adhesion on a 45° overhang region.
Figure 9. The L-PBF printed samples with various overhang angle (a) θ = 0° (cube), (b) θ = 30°, (c) θ = 45°, (d) θ = 55° and (e) θ = 60°.
Figure 9. The L-PBF printed samples with various overhang angle (a) θ = 0° (cube), (b) θ = 30°, (c) θ = 45°, (d) θ = 55° and (e) θ = 60°.
Figure 10. Two mechanisms of powder adhesion related to the overhang angle: (a) simulation-predicted, θ = 45°; (b) simulation-predicted, θ = 60°; (c, e) optical micrographs, θ = 45°; (d, f) optical micrographs, θ = 60°. (e) and (f) are partial enlargement of (c) and (d), respectively.
Figure 10. Two mechanisms of powder adhesion related to the overhang angle: (a) simulation-predicted, θ = 45°; (b) simulation-predicted, θ = 60°; (c, e) optical micrographs, θ = 45°; (d, f) optical micrographs, θ = 60°. (e) and (f) are partial enlargement of (c) and (d), respectively.
Figure 11. Simulation-predicted surface morphology in the overhang region at different overhang angle: (a) θ = 15°, (b) θ = 30°, (c) θ = 45°, (d) θ = 60° and (e) θ = 80° (Blue solid lines: simulation-predicted contour; red dashed lines: the planar profile of designed overhang region specified by the overhang angles).
Figure 11. Simulation-predicted surface morphology in the overhang region at different overhang angle: (a) θ = 15°, (b) θ = 30°, (c) θ = 45°, (d) θ = 60° and (e) θ = 80° (Blue solid lines: simulation-predicted contour; red dashed lines: the planar profile of designed overhang region specified by the overhang angles).
Figure 12. Effect of overhang angle on surface roughness Ra in overhang regions
Figure 12. Effect of overhang angle on surface roughness Ra in overhang regions
Figure 13. Surface morphology of L-PBF printed overhang regions with different overhang angle: (a) θ = 15°, (b) θ = 30°, (c) θ = 45° and (d) θ = 60° (overhang border parameters: P = 100 W, v = 1000 mm/s).
Figure 13. Surface morphology of L-PBF printed overhang regions with different overhang angle: (a) θ = 15°, (b) θ = 30°, (c) θ = 45° and (d) θ = 60° (overhang border parameters: P = 100 W, v = 1000 mm/s).
Figure 14. Effect of (a) laser power (scan speed = 1000 mm/s) and (b) scan speed (lase power = 100 W) on surface roughness Ra in overhang regions (θ = 45°, laser power and scan speed referred to overhang border parameters, and the other process parameters are listed in Table 2).
Figure 14. Effect of (a) laser power (scan speed = 1000 mm/s) and (b) scan speed (lase power = 100 W) on surface roughness Ra in overhang regions (θ = 45°, laser power and scan speed referred to overhang border parameters, and the other process parameters are listed in Table 2).

References

  • Cai, Chao, Chrupcala Radoslaw, Jinliang Zhang, Qian Yan, Shifeng Wen, Bo Song, and Yusheng Shi. 2019. “In-Situ Preparation and Formation of TiB/Ti-6Al-4V Nanocomposite via Laser Additive Manufacturing: Microstructure Evolution and Tribological Behavior.” Powder Technology 342: 73–84. doi:10.1016/j.powtec.2018.09.088. [Crossref], [Web of Science ®], [Google Scholar]
  • Cai, Chao, Wei Shian Tey, Jiayao Chen, Wei Zhu, Xingjian Liu, Tong Liu, Lihua Zhao, and Kun Zhou. 2021. “Comparative Study on 3D Printing of Polyamide 12 by Selective Laser Sintering and Multi Jet Fusion.” Journal of Materials Processing Technology 288 (August 2020): 116882. doi:10.1016/j.jmatprotec.2020.116882. [Crossref], [Web of Science ®], [Google Scholar]
  • Cai, Chao, Xu Wu, Wan Liu, Wei Zhu, Hui Chen, Jasper Dong Qiu Chua, Chen Nan Sun, Jie Liu, Qingsong Wei, and Yusheng Shi. 2020. “Selective Laser Melting of Near-α Titanium Alloy Ti-6Al-2Zr-1Mo-1V: Parameter Optimization, Heat Treatment and Mechanical Performance.” Journal of Materials Science and Technology 57: 51–64. doi:10.1016/j.jmst.2020.05.004. [Crossref], [Web of Science ®], [Google Scholar]
  • Charles, Amal, Ahmed Elkaseer, Lore Thijs, and Steffen G. Scholz. 2020. “Dimensional Errors Due to Overhanging Features in Laser Powder Bed Fusion Parts Made of Ti-6Al-4V.” Applied Sciences 10 (7): 2416. doi:10.3390/app10072416. [Crossref], [Google Scholar]
  • Chatham, Camden A., Timothy E. Long, and Christopher B. Williams. 2019. “A Review of the Process Physics and Material Screening Methods for Polymer Powder Bed Fusion Additive Manufacturing.” Progress in Polymer Science 93: 68–95. doi:10.1016/j.progpolymsci.2019.03.003. [Crossref], [Web of Science ®], [Google Scholar]
  • Du, Yang, Xinyu You, Fengbin Qiao, Lijie Guo, and Zhengwu Liu. 2019. “A Model for Predicting the Temperature Field during Selective Laser Melting.” Results in Physics 12 (November 2018): 52–60. doi:10.1016/j.rinp.2018.11.031. [Crossref], [Web of Science ®], [Google Scholar]
  • Duval-Chaneac, M. S., S. Han, C. Claudin, F. Salvatore, J. Bajolet, and J. Rech. 2018. “Experimental Study on Finishing of Internal Laser Melting (SLM) Surface with Abrasive Flow Machining (AFM).” Precision Engineering 54 (July 2017): 1–6. doi:10.1016/j.precisioneng.2018.03.006. [Crossref], [Web of Science ®], [Google Scholar]
  • Feng, Shaochuan, Shijie Chen, Amar M. Kamat, Ru Zhang, Mingji Huang, and Liangcai Hu. 2020. “Investigation on Shape Deviation of Horizontal Interior Circular Channels Fabricated by Laser Powder Bed Fusion.” Additive Manufacturing 36 (December): 101585. doi:10.1016/j.addma.2020.101585. [Crossref], [Web of Science ®], [Google Scholar]
  • Feng, Shaochuan, Chuanzhen Huang, Jun Wang, Hongtao Zhu, Peng Yao, and Zhanqiang Liu. 2017. “An Analytical Model for the Prediction of Temperature Distribution and Evolution in Hybrid Laser-Waterjet Micro-Machining.” Precision Engineering 47: 33–45. doi:10.1016/j.precisioneng.2016.07.002. [Crossref], [Web of Science ®], [Google Scholar]
  • Feng, Shaochuan, Amar M. Kamat, and Yutao Pei. 2021. “Design and Fabrication of Conformal Cooling Channels in Molds: Review and Progress Updates.” International Journal of Heat and Mass Transfer. doi:10.1016/j.ijheatmasstransfer.2021.121082. [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
  • Flow-3D V11.2 Documentation. 2016. Flow Science, Inc. [Crossref], [Google Scholar]
  • Foroozmehr, Ali, Mohsen Badrossamay, Ehsan Foroozmehr, and Sa’id Golabi. 2016. “Finite Element Simulation of Selective Laser Melting Process Considering Optical Penetration Depth of Laser in Powder Bed.” Materials and Design 89: 255–263. doi:10.1016/j.matdes.2015.10.002. [Crossref], [Web of Science ®], [Google Scholar]
  • “Geometrical Product Specifications (GPS) — Surface Texture: Profile Method — Rules and Procedures for the Assessment of Surface Texture (ISO 4288).” 1996. International Organization for Standardization. https://www.iso.org/standard/2096.html. [Google Scholar]
  • Günther, Johannes, Stefan Leuders, Peter Koppa, Thomas Tröster, Sebastian Henkel, Horst Biermann, and Thomas Niendorf. 2018. “On the Effect of Internal Channels and Surface Roughness on the High-Cycle Fatigue Performance of Ti-6Al-4V Processed by SLM.” Materials & Design 143: 1–11. doi:10.1016/j.matdes.2018.01.042. [Crossref], [Web of Science ®], [Google Scholar]
  • Hopkinson, Neil, and Phill Dickens. 2000. “Conformal Cooling and Heating Channels Using Laser Sintered Tools.” In Solid Freeform Fabrication Conference, 490–497. Texas. doi:10.26153/tsw/3075. [Crossref], [Google Scholar]
  • Hu, Zhiheng, Haihong Zhu, Changchun Zhang, Hu Zhang, Ting Qi, and Xiaoyan Zeng. 2018. “Contact Angle Evolution during Selective Laser Melting.” Materials and Design 139: 304–313. doi:10.1016/j.matdes.2017.11.002. [Crossref], [Web of Science ®], [Google Scholar]
  • Hu, Cheng, Kejia Zhuang, Jian Weng, and Donglin Pu. 2019. “Three-Dimensional Analytical Modeling of Cutting Temperature for Round Insert Considering Semi-Infinite Boundary and Non-Uniform Heat Partition.” International Journal of Mechanical Sciences 155 (October 2018): 536–553. doi:10.1016/j.ijmecsci.2019.03.019. [Crossref], [Web of Science ®], [Google Scholar]
  • Huang, Yuze, Mir Behrad Khamesee, and Ehsan Toyserkani. 2019. “A New Physics-Based Model for Laser Directed Energy Deposition (Powder-Fed Additive Manufacturing): From Single-Track to Multi-Track and Multi-Layer.” Optics & Laser Technology 109 (August 2018): 584–599. doi:10.1016/j.optlastec.2018.08.015. [Crossref], [Web of Science ®], [Google Scholar]
  • Kadirgama, K., W. S. W. Harun, F. Tarlochan, M. Samykano, D. Ramasamy, Mohd Zaidi Azir, and H. Mehboob. 2018. “Statistical and Optimize of Lattice Structures with Selective Laser Melting (SLM) of Ti6AL4V Material.” International Journal of Advanced Manufacturing Technology 97 (1–4): 495–510. doi:10.1007/s00170-018-1913-1. [Crossref], [Web of Science ®], [Google Scholar]
  • Kamat, Amar M, and Yutao Pei. 2019. “An Analytical Method to Predict and Compensate for Residual Stress-Induced Deformation in Overhanging Regions of Internal Channels Fabricated Using Powder Bed Fusion.” Additive Manufacturing 29 (March): 100796. doi:10.1016/j.addma.2019.100796. [Crossref], [Web of Science ®], [Google Scholar]
  • Kamath, Chandrika, Bassem El-Dasher, Gilbert F. Gallegos, Wayne E. King, and Aaron Sisto. 2014. “Density of Additively-Manufactured, 316L SS Parts Using Laser Powder-Bed Fusion at Powers up to 400 W.” International Journal of Advanced Manufacturing Technology 74 (1–4): 65–78. doi:10.1007/s00170-014-5954-9. [Crossref], [Web of Science ®], [Google Scholar]
  • Karimi, J., C. Suryanarayana, I. Okulov, and K. G. Prashanth. 2020. “Selective Laser Melting of Ti6Al4V: Effect of Laser Re-Melting.” Materials Science and Engineering A (July): 140558. doi:10.1016/j.msea.2020.140558. [Crossref], [Web of Science ®], [Google Scholar]
  • Khairallah, Saad A., and Andy Anderson. 2014. “Mesoscopic Simulation Model of Selective Laser Melting of Stainless Steel Powder.” Journal of Materials Processing Technology 214 (11): 2627–2636. doi:10.1016/j.jmatprotec.2014.06.001. [Crossref], [Web of Science ®], [Google Scholar]
  • Khairallah, Saad A., Andrew T. Anderson, Alexander Rubenchik, and Wayne E. King. 2016. “Laser Powder-Bed Fusion Additive Manufacturing: Physics of Complex Melt Flow and Formation Mechanisms of Pores, Spatter, and Denudation Zones.” Edited by Adedeji B. Badiru, Vhance V. Valencia, and David Liu. Acta Materialia 108 (April): 36–45. doi:10.1016/j.actamat.2016.02.014. [Crossref], [Web of Science ®], [Google Scholar]
  • Kuo, C. N., C. K. Chua, P. C. Peng, Y. W. Chen, S. L. Sing, S. Huang, and Y. L. Su. 2020. “Microstructure Evolution and Mechanical Property Response via 3D Printing Parameter Development of Al–Sc Alloy.” Virtual and Physical Prototyping 15 (1): 120–129. doi:10.1080/17452759.2019.1698967. [Taylor & Francis Online], [Web of Science ®], [Google Scholar]
  • Le, K. Q., C. H. Wong, K. H. G. Chua, C. Tang, and H. Du. 2020. “Discontinuity of Overhanging Melt Track in Selective Laser Melting Process.” International Journal of Heat and Mass Transfer 162 (December): 120284. doi:10.1016/j.ijheatmasstransfer.2020.120284. [Crossref], [Web of Science ®], [Google Scholar]
  • Lee, Y. S., and W. Zhang. 2016. “Modeling of Heat Transfer, Fluid Flow and Solidification Microstructure of Nickel-Base Superalloy Fabricated by Laser Powder Bed Fusion.” Additive Manufacturing 12: 178–188. doi:10.1016/j.addma.2016.05.003. [Crossref], [Web of Science ®], [Google Scholar]
  • Leitz, K. H., P. Singer, A. Plankensteiner, B. Tabernig, H. Kestler, and L. S. Sigl. 2017. “Multi-Physical Simulation of Selective Laser Melting.” Metal Powder Report 72 (5): 331–338. doi:10.1016/j.mprp.2016.04.004. [Crossref], [Google Scholar]
  • Li, Jian, Jing Hu, Yi Zhu, Xiaowen Yu, Mengfei Yu, and Huayong Yang. 2020. “Surface Roughness Control of Root Analogue Dental Implants Fabricated Using Selective Laser Melting.” Additive Manufacturing 34 (September 2019): 101283. doi:10.1016/j.addma.2020.101283. [Crossref], [Web of Science ®], [Google Scholar]
  • Li, Yingli, Kun Zhou, Pengfei Tan, Shu Beng Tor, Chee Kai Chua, and Kah Fai Leong. 2018. “Modeling Temperature and Residual Stress Fields in Selective Laser Melting.” International Journal of Mechanical Sciences 136 (February): 24–35. doi:10.1016/j.ijmecsci.2017.12.001. [Crossref], [Web of Science ®], [Google Scholar]
  • Mazur, MacIej, Martin Leary, Matthew McMillan, Joe Elambasseril, and Milan Brandt. 2016. “SLM Additive Manufacture of H13 Tool Steel with Conformal Cooling and Structural Lattices.” Rapid Prototyping Journal 22 (3): 504–518. doi:10.1108/RPJ-06-2014-0075. [Crossref], [Web of Science ®], [Google Scholar]
  • Mingear, Jacob, Bing Zhang, Darren Hartl, and Alaa Elwany. 2019. “Effect of Process Parameters and Electropolishing on the Surface Roughness of Interior Channels in Additively Manufactured Nickel-Titanium Shape Memory Alloy Actuators.” Additive Manufacturing 27 (October 2018): 565–575. doi:10.1016/j.addma.2019.03.027. [Crossref], [Web of Science ®], [Google Scholar]
  • Pakkanen, Jukka, Flaviana Calignano, Francesco Trevisan, Massimo Lorusso, Elisa Paola Ambrosio, Diego Manfredi, and Paolo Fino. 2016. “Study of Internal Channel Surface Roughnesses Manufactured by Selective Laser Melting in Aluminum and Titanium Alloys.” Metallurgical and Materials Transactions A 47 (8): 3837–3844. doi:10.1007/s11661-016-3478-7. [Crossref], [Web of Science ®], [Google Scholar]
  • Patterson, Albert E., Sherri L. Messimer, and Phillip A. Farrington. 2017. “Overhanging Features and the SLM/DMLS Residual Stresses Problem: Review and Future Research Need.” Technologies 5 (4): 15. doi:10.3390/technologies5020015. [Crossref], [Web of Science ®], [Google Scholar]
  • Roberts, I. A., C. J. Wang, R. Esterlein, M. Stanford, and D. J. Mynors. 2009. “A Three-Dimensional Finite Element Analysis of the Temperature Field during Laser Melting of Metal Powders in Additive Layer Manufacturing.” International Journal of Machine Tools and Manufacture 49 (12–13): 916–923. doi:10.1016/j.ijmachtools.2009.07.004. [Crossref], [Web of Science ®], [Google Scholar]
  • Shrestha, Subin, and Kevin Chou. 2018. “Computational Analysis of Thermo-Fluid Dynamics with Metallic Powder in SLM.” In CFD Modeling and Simulation in Materials Processing 2018, edited by Laurentiu Nastac, Koulis Pericleous, Adrian S. Sabau, Lifeng Zhang, and Brian G. Thomas, 85–95. Cham, Switzerland: Springer Nature. doi:10.1007/978-3-319-72059-3_9. [Crossref], [Google Scholar]
  • Sing, S. L., and W. Y. Yeong. 2020. “Laser Powder Bed Fusion for Metal Additive Manufacturing: Perspectives on Recent Developments.” Virtual and Physical Prototyping 15 (3): 359–370. doi:10.1080/17452759.2020.1779999. [Taylor & Francis Online], [Web of Science ®], [Google Scholar]
  • Šmilauer, Václav, Emanuele Catalano, Bruno Chareyre, Sergei Dorofeenko, Jérôme Duriez, Nolan Dyck, Jan Eliáš, et al. 2015. Yade Documentation. 2nd ed. The Yade Project. doi:10.5281/zenodo.34073. [Crossref], [Google Scholar]
  • Tan, Pengfei, Fei Shen, Biao Li, and Kun Zhou. 2019. “A Thermo-Metallurgical-Mechanical Model for Selective Laser Melting of Ti6Al4V.” Materials & Design 168 (April): 107642. doi:10.1016/j.matdes.2019.107642. [Crossref], [Web of Science ®], [Google Scholar]
  • Tan, Lisa Jiaying, Wei Zhu, and Kun Zhou. 2020. “Recent Progress on Polymer Materials for Additive Manufacturing.” Advanced Functional Materials 30 (43): 1–54. doi:10.1002/adfm.202003062. [Crossref], [Web of Science ®], [Google Scholar]
  • Wang, Xiaoqing, and Kevin Chou. 2018. “Effect of Support Structures on Ti-6Al-4V Overhang Parts Fabricated by Powder Bed Fusion Electron Beam Additive Manufacturing.” Journal of Materials Processing Technology 257 (February): 65–78. doi:10.1016/j.jmatprotec.2018.02.038. [Crossref], [Web of Science ®], [Google Scholar]
  • Wang, Di, Yongqiang Yang, Ziheng Yi, and Xubin Su. 2013. “Research on the Fabricating Quality Optimization of the Overhanging Surface in SLM Process.” International Journal of Advanced Manufacturing Technology 65 (9–12): 1471–1484. doi:10.1007/s00170-012-4271-4. [Crossref], [Web of Science ®], [Google Scholar]
  • Wen, Peng, Maximilian Voshage, Lucas Jauer, Yanzhe Chen, Yu Qin, Reinhart Poprawe, and Johannes Henrich Schleifenbaum. 2018. “Laser Additive Manufacturing of Zn Metal Parts for Biodegradable Applications: Processing, Formation Quality and Mechanical Properties.” Materials and Design 155: 36–45. doi:10.1016/j.matdes.2018.05.057. [Crossref], [Web of Science ®], [Google Scholar]
  • Wu, Yu-che, Cheng-hung San, Chih-hsiang Chang, Huey-jiuan Lin, Raed Marwan, Shuhei Baba, and Weng-Sing Hwang. 2018. “Numerical Modeling of Melt-Pool Behavior in Selective Laser Melting with Random Powder Distribution and Experimental Validation.” Journal of Materials Processing Technology 254 (November 2017): 72–78. doi:10.1016/j.jmatprotec.2017.11.032. [Crossref], [Web of Science ®], [Google Scholar]
  • Yadroitsev, I., P. Krakhmalev, I. Yadroitsava, S. Johansson, and I. Smurov. 2013. “Energy Input Effect on Morphology and Microstructure of Selective Laser Melting Single Track from Metallic Powder.” Journal of Materials Processing Technology 213 (4): 606–613. doi:10.1016/j.jmatprotec.2012.11.014. [Crossref], [Web of Science ®], [Google Scholar]
  • Yu, Wenhui, Swee Leong Sing, Chee Kai Chua, and Xuelei Tian. 2019. “Influence of Re-Melting on Surface Roughness and Porosity of AlSi10Mg Parts Fabricated by Selective Laser Melting.” Journal of Alloys and Compounds 792: 574–581. doi:10.1016/j.jallcom.2019.04.017. [Crossref], [Web of Science ®], [Google Scholar]
  • Zhang, Dongyun, Pudan Zhang, Zhen Liu, Zhe Feng, Chengjie Wang, and Yanwu Guo. 2018. “Thermofluid Field of Molten Pool and Its Effects during Selective Laser Melting (SLM) of Inconel 718 Alloy.” Additive Manufacturing 21 (100): 567–578. doi:10.1016/j.addma.2018.03.031. [Crossref], [Web of Science ®], [Google Scholar]
Fig. 1. Schematic description of the laser welding process considered in this study.

Analysis of molten pool dynamics in laser welding with beam oscillation and filler wire feeding

Won-Ik Cho, Peer Woizeschke
Bremer Institut für angewandte Strahltechnik GmbH, Klagenfurter Straße 5, Bremen 28359, Germany

Received 30 July 2020, Revised 3 October 2020, Accepted 18 October 2020, Available online 1 November 2020.

Abstract

Molten pool flow and heat transfer in a laser welding process using beam oscillation and filler wire feeding were calculated using computational fluid dynamics (CFD). There are various indirect methods used to analyze the molten pool dynamics in fusion welding. In this work, based on the simulation results, the surface fluctuation was directly measured to enable a more intuitive analysis, and then the signal was analyzed using the Fourier transform and wavelet transform in terms of the beam oscillation frequency and buttonhole formation. The 1st frequency (2 x beam oscillation frequency, the so-called chopping frequency), 2nd frequency (4 x beam oscillation frequency), and beam oscillation frequency components were the main components found. The 1st and 2nd frequency components were caused by the effect of the chopping process and lumped line energy. The beam oscillation frequency component was related to rapid, unstable molten pool behavior. The wavelet transform effectively analyzed the rapid behaviors based on the change of the frequency components over time.

Korea Abstract

빔 진동 및 필러 와이어 공급을 사용하는 레이저 용접 공정에서 용융 풀 흐름 및 열 전달은 CFD (전산 유체 역학)를 사용하여 계산되었습니다. 용융 용접에서 용융 풀 역학을 분석하는 데 사용되는 다양한 간접 방법이 있습니다.

본 연구에서는 시뮬레이션 결과를 바탕으로 보다 직관적 인 분석이 가능하도록 표면 변동을 직접 측정 한 후 빔 발진 주파수 및 버튼 홀 형성 측면에서 푸리에 변환 및 웨이블릿 변환을 사용하여 신호를 분석했습니다.

1 차 주파수 (2 x 빔 발진 주파수, 이른바 초핑 주파수), 2 차 주파수 (4 x 빔 발진 주파수) 및 빔 발진 주파수 성분이 발견 된 주요 구성 요소였습니다. 1 차 및 2 차 주파수 성분은 쵸핑 공정과 집중 라인 에너지의 영향으로 인해 발생했습니다.

빔 진동 주파수 성분은 빠르고 불안정한 용융 풀 동작과 관련이 있습니다. 웨이블릿 변환은 시간 경과에 따른 주파수 구성 요소의 변화를 기반으로 빠른 동작을 효과적으로 분석했습니다.

1 . 소개

융합 용접에서 용융 풀 역학은 용접 결함과 시각적 이음새 품질에 직접적인 영향을 미칩니다. 이러한 역학을 연구하기 위해 고속 카메라를 사용하는 직접 방법과 광학 또는 음향 신호를 사용하는 간접 방법과 같은 다양한 측정 방법을 사용하여 여러 실험 방법을 고려했습니다. 시간 도메인의 원래 신호는 특별히 주파수 도메인에서 변환 된 신호로 변환되어 용융 풀 동작에 영향을 미치는 주파수 성분을 분석합니다. Kotecki et al. (1972)는 고속 카메라를 사용하여 가스 텅스텐 아크 용접에서 용융 풀을 관찰했습니다. [1]. 그들은 120Hz 리플 DC 출력을 가진 용접 전원을 사용할 때 용융 풀 진동 주파수가 120Hz임을 보여주었습니다. 전원을 끈 후 진동 주파수는 용융 풀의 고유 주파수를 나타내는 용융 풀 크기와 관련이 있습니다. 진동은 응고 중에 용접 표면 스케일링을 생성했습니다. Zacksenhouse and Hardt (1983)는 레이저 섀도 잉 동작 측정 기술을 사용하여 가스 텅스텐 아크 용접에서 완전히 관통 된 용융 풀의 동작을 측정했습니다 [2] . 그들은 2.5mm 두께의 강판에서 6mm 풀 반경 (고정 용접)에 대해 용융 풀의 고유 주파수가 18.9Hz라는 것을 발견했습니다. Semak et al. (1995) 고속 카메라를 사용하여 레이저 스폿 용접에서 용융 풀 및 키홀 역학 조사 [3]. 그들은 깊이가 약 3mm이고 반경이 약 3mm 인 용융 풀에서 200Hz의 낮은 체적 진동 주파수를 관찰했습니다. 0.45mm Aendenroomer와 den Ouden (1998)은 강철의 펄스 가스 텅스텐 아크 용접에서 용융 풀 진동을보고했습니다 [4] . 그들은 침투 깊이에 따라 진동 모드 변화를 보였고 주파수는 50Hz에서 150Hz 사이에서 변화했습니다. 주파수는 완전히 침투 된 용융 풀에서 더 낮았습니다. Hermans와 den Ouden (1999)은 단락 가스 금속 아크 용접에서 용융 풀 진동을 분석했습니다. [5]. 그들은 용융 풀의 단락 주파수와 고유 주파수가 같을 때 부분적으로 침투 된 용융 풀의 경우 공정 안정성이 향상되었음을 보여주었습니다. Yudodibroto et al. (2004)는 가스 텅스텐 아크 용접에서 용융 풀 진동에 대한 필러 와이어의 영향을 조사했습니다 [6] . 그들은 금속 전달이 특히 부분적으로 침투 된 용융 풀에서 진동 거동을 방해한다는 것을 보여주었습니다. Geiger et al. (2009) 레이저 키홀 용접에서 발광 분석 [7]. 신호의 주파수 분석을 사용하여 용융 풀 (1.5kHz 미만)과 키홀 (약 3kHz)에 해당하는 진동 주파수 범위를 찾았습니다.&nb