Elsevier

Powder Technology

Volume 363, 1 March 2020, Pages 602-610
Powder Technology

Adhesion effects on spreading of metal powders in selective laser melting

https://doi.org/10.1016/j.powtec.2019.12.048Get rights and content

Highlights

  • Ni alloy powder spreading process in SLM is investigated by DEM.

  • Adhesion effects can reduce powder packing density and smoothness.

  • Rollers can pack powder more densely and smoothly than blades.

  • Adhesion effects in roller system are reduced compared with blade system.

Abstract

Selective Laser Melting (SLM) is a rapidly developing and advanced manufacturing method for fabricating complex products. In SLM, the powder spreading process is crucial to ensure that the right amount of material can be fully melted by a certain laser energy input in order to minimise defects and achieve the desired microstructure. The packing density and homogeneity of the formed powder bed are of interest when comparing melting efficiency and quality of SLM processes with different metal powders or different spreading methods. Particle-based numerical studies are required for identifying the powder bed structure and particle dynamical behaviours which are affected by particle adhesion. In this work, experiments on powder packing density and repose angle for different particle size distributions are carried out. The discrete element method (DEM) model is validated and calibrated based on experimental results. The DEM is then used to examine the powder spreading process, focusing on the effects of particle adhesion and particle-based behaviours. Effects of spreader type, adhesion magnitude and particle size distribution are analysed. The results show that particle adhesion can reduce powder packing density and smoothness of the powder bed surface. Proper adhesion effects can improve powder bed homogeneity. Powder bed structure is determined not only by adhesion effects but also by particle rearrangement during spreading. Regarding spreader type, the roller can spread a better powder bed than rigid blade due to different particle contact force distributions and particle velocities in the powder pile and powder bed, which lead to different particle rearrangements and particle contact conditions.

Introduction

Additive Manufacturing (AM), which is also known as “3-D printing” and “rapid prototyping” [[1], [2], [3], [4]], is a rapidly developing fabrication process. In contrary to traditional subtractive manufacturing such as casting and machining, parts are made by adding and binding materials in layers and each layer is a thin cross-section of the final part [2]. The applications of AM, including direct digital manufacturing and rapid tooling, have been largely expanded due to process improvement and introduction of new processes and materials [2]. The development and application of AM can benefit and change the manufacturing systems [5,6]. For the fabrication of functional metal parts which are widely used in aerospace, biomedical and automotive industries, Selective Laser Melting (SLM) is most suitable due to its high final part quality [[7], [8], [9], [10], [11], [12]]. In the SLM process, the material is supplied as powder, which is spread by a spreader to form a powder bed, and a laser beam is used to melt the loose powders. The metal powders used in SLM are proprietary alloy powders fabricated by gas or plasma atomization which can produce particles with high sphericity [13]. The powder particle size can vary from several microns to over 100 μm. Al alloy, Co–Cr alloy, Ni alloy, Ti alloy and stainless steel are the main materials used in SLM.

The application improvement of SLM depends on energy and material cost, build time, part accuracy and part quality [2]. The process optimization is related to the understanding of the melting process and mechanical behaviour of the powder used in the process. Characterization of powder behaviours based on experiments may be difficult and costly work. Powder bed characteristics, such as packing density, surface roughness and homogeneity, can affect the energy absorption, heat transfer and melt pool characteristics [[14], [15], [16], [17], [18], [19], [20]] in the melting process and thus affect the final part quality [21,22]. Understanding the powder spreading process and its influencing factors can help the quality control of SLM. Powder spreading process is affected by packing and flow properties of the powder, which are proved to be affected by particle size distribution and particle adhesion. The effects of particle size distribution on powder packing and flow properties and final part quality can be combined with powder adhesion and friction effects because smaller particles lead to more obvious adhesion and friction effects, as shown in some studies [[23], [24], [25], [26], [27]]. These studies also indicate that good flowability of fine powders can lead to either higher or lower packing density which depends on the specific particle size distribution and adhesion magnitude.

On the particle-based simulation studies of powder bed in SLM, different models have been used to generate a poured packing structure. Among them, discrete element method (DEM) is more suitable than rain model [28] because it allows the calculation of forces on each particle and particle motion is determined by the forces. Hence the structure is much closer to the reality [29]. Some studies [30,31] use DEM to generate poured particle packing layer by layer to represent the layered packing structure of powder bed. But the effect of particle-spreader interaction is ignored. The importance of such interactions is demonstrated in Mindt et al.'s work [32] which shows that the powder spreading process leads to powder layer inhomogeneity rather than uniform structure by poured packing. In the work of Haeri et al. [33], a roller spreader and a blade spreader are compared regarding powder bed density and surface roughness for polymer rod-like powder particles. The roller is indicated better than the blade. However, coarse non-adhesive particles are used instead of fine adhesive particles. Furthermore, the powder spreading is also studied by DEM without particle adhesion to reveal some features of the spreading process such as layer thickness and homogeneity, and obtain powder bed structure for melting simulation [34,35]. In Parteli et al.'s work [[36], [37], [38]], van der Waals attractive force and JKR adhesion force are calculated. Effects of fine particle adhesion on inhomogeneity are examined by calculating the load on spreader during roller spreading. The simulated particles are polymer particles with non-spherical shapes. The force model is validated by packing density measured in experiments with glass powders. In the work of Han et al. [39] and Meier et al. [40], the surface energy of JKR force model is calibrated by experiments on powder repose angle to determine the powder cohesion magnitude. In Han et al.'s work [39], the model is then used to analyse the effects of powder layer thickness on void distribution in the powder bed.

For metal powders, characterization of particle-based dynamics during spreading process is needed to help understand the effects of spreader type and particle adhesion magnitude. In this work, the spreading of Ni-alloy powder is simulated. Different spreaders (blade and roller) are simulated to explain the difference in powder bed structure. Different adhesion magnitudes (controlled by Hamaker constant which determines the magnitude of van der Waals force among dry metal powders) are simulated to help understand adhesion effects. The real metal powder properties need to be better understood; therefore, packing density and repose angle of metal powder samples are tested in this work. The DEM model is first calibrated by the experimental results and then used to simulate the powder spreading process.

Section snippets

DEM simulation model

The DEM simulation is carried out using the open-source DEM code LIGGGHTS [41]. Forces between particles include normal and tangential viscoelastic contact forces, rolling friction and van der Waals adhesion force, as shown in Fig. 1. The force models are based on the facts that the particles have good sphericity, large stiffness, and obvious adhesion due to van der Waals force attraction. The translational and rotational motions of particles are governed by Newton's second law of motion:midvidt

Flow patterns during spreading process

Rigid blade and counter-rotating roller are different spreaders used in the powder spreading process. In this work, they are simulated to compare the effects of spreader type on powder spreading. The difference is analysed by powder bed packing density, surface roughness and homogeneity, particle force and velocity, and particle contact condition. Fig. 6 shows the velocity vectors of particles and normal contact force chains among particles during powder spreading. It is noted that compared

Conclusions

DEM is used in this work to study Ni alloy powder spreading process in Selective Laser Melting process. The model is calibrated by experimental results on powder packing density and repose angle. Different spreader types, Hamaker constant and particle size distributions are simulated and compared. The following conclusions can be drawn from this work:

  • Hamaker constant and sliding friction coefficient have the main effects on powder packing density and repose angle in this work. Values of several

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

The authors are grateful for the financial support from Australian Research Council Industrial Transformation Research Hubs Scheme (Project Number IH140100035). This research was undertaken with the assistance of resources from the National Computational Infrastructure (NCI), which is supported by the Australian Government.

References (52)

  • V. Seyda et al.

    Investigation of aging processes of ti-6Al-4 V powder material in laser melting

    Phys. Procedia

    (2012)
  • J. Yang et al.

    Dry particle coating for improving the flowability of cohesive powders

    Powder Technol.

    (2005)
  • M. Markl et al.

    Powder layer deposition algorithm for additive manufacturing simulations

    Powder Technol.

    (2018)
  • S. Haeri et al.

    Discrete element simulation and experimental study of powder spreading process in additive manufacturing

    Powder Technol.

    (2017)
  • E.J.R. Parteli et al.

    Particle-based simulation of powder application in additive manufacturing

    Powder Technol.

    (2016)
  • Q. Han et al.

    Discrete element simulation of powder layer thickness in laser additive manufacturing

    Powder Technol.

    (2019)
  • C. Meier et al.

    Modeling and characterization of cohesion in fine metal powders with a focus on additive manufacturing process simulations

    Powder Technol.

    (2019)
  • A. Di Renzo et al.

    An improved integral non-linear model for the contact of particles in distinct element simulations

    Chem. Eng. Sci.

    (2005)
  • J. Ai et al.

    Assessment of rolling resistance models in discrete element simulations

    Powder Technol.

    (2011)
  • H.C. Hamaker

    The London—van der Waals attraction between spherical particles

    Physica

    (1937)
  • A. Yu et al.

    On the relationship between porosity and interparticle forces

    Powder Technol.

    (2003)
  • R.Y. Yang et al.

    Pore structure of the packing of fine particles

    J. Colloid Interface Sci.

    (2006)
  • I. Campbell et al.

    Additive manufacturing: rapid prototyping comes of age

    Rapid Prototyp. J.

    (2012)
  • I. Gibson et al.

    Additive Manufacturing Technologies

    (2014)
  • S.H. Huang et al.

    Additive manufacturing and its societal impact: a literature review

    Int. J. Adv. Manuf. Technol.

    (2013)
  • D.W. Rosen

    A review of synthesis methods for additive manufacturing

    Virtual Phys. Prototyp.

    (2016)
  • Cited by (75)

    View all citing articles on Scopus
    View full text