Comparison of roller-spreading and blade-spreading processes in powder-bed additive manufacturing by DEM simulations

Highlights

• Movement trajectory of particles in roller-spreading is more complicated.

• Particles have relatively weak dispersion in blade-spreading.

• Spreading methods affect particle segregation and powder layer density.

• Higher density can be obtained by roller-spreading at thicker powder layer.

• Blade-spreading limits its application to thicker powder layer.

Abstract

The roller-spreading and blade-spreading are main powder spreading methods in powder-bed additive manufacturing. The discrete element method was introduced to simulate nylon powder spreading by both roller and blade spreaders. The two spreading processes were compared from several aspects including particle flow behavior, particle contact forces, forces exerted on spreaders, particle segregation and powder layer density. It is found that powder spreading methods mainly affect the movement trajectory of particles, particle contact forces and forces exerted on spreaders. Complicated dispersion and circulation movement of particles occur inside the powder pile by roller-spreading, while particles have relatively weak dispersion by the blade-spreading. The normal force applied to the roller introduces a compacting effect on the powder pile and creates strong force chains that distribute uniformly in the powder pile. Therefore, the powder bed with higher density can be obtained by roller-spreading in thicker powder layer due to the compacting effect. The blade spreader sustains tangential force mainly, so the blade-spreading process limits its application to thicker powder layer. As the powder layer thickness increases, the roller-spreading is more sensitive to segregation index than that of the blade-spreading. The comprehensive comparison of two spreading processes provides criteria for selecting spreading methods.

Introduction

The powder-bed based additive manufacturing (AM), such as selective laser sintering (SLS) and selective laser melting (SLM) take powder as feedstock materials to produce complex parts layer-by-layer, which is the most widely used advanced technology in the industry (Ali et al., 2018; Bourell et al., 2017; Cordova, Bor, de Smit, Campos, & Tinga, 2020). Many complex factors affect the quality of fabricated parts in AM processes. Among them, the powder spreading process is a crucial factor, which directly affects the interaction between the powder and the laser heat source (Chen et al., 2019). The high powder-bed quality is a prerequisite for the more stable and continuous laser melting powder processes, also can reduce surface defects and inner porosity of fabricated parts (Yao et al., 2021; Zhang et al., 2020).

Some factors affect the powder-bed quality, such as powder spreading methods (roller-spreading and blade-spreading) (Haeri, Wang, Ghita, & Sun, 2017; Wang et al., 2020), powder spreading process parameters (translational velocity of the spreader and the powder layer thickness) (Han, Gu, & Setchi, 2019) and physical characteristics (particle size distribution, particle shape and particle surface texture) (Ma, Evans, Philips, & Cunningham, 2020; Mussatto et al., 2021). The most widely used powder spreading methods are roller-spreading (He, Hassanpour, & Bayly, 2020; Wang et al., 2020) and blade-spreading (Haeri, 2017; Nan & Ghadiri, 2019) in practice. For the blade-spreading process, the straight movement and mechanical structure of the blade are simple and have better practicability. For the roller-spreading process, the roller includes translation and counter-rotating movements. The roller requires the shape and dimension with high precision. In addition, the movement control strategy of the roller is complicated. Therefore, the processing accuracy and mechanical structure of the roller are more complex than that of the blade.

In recent years, powder spreading processes have attracted attention by physical experiments. The powder spreadability metrics were first tried to establish by Snow, Martukanitz, and Joshi (2019). The percent coverage, deposition rate and rate of avalanche angle change have been proven to assess powder spreadability quantitatively. The powder dynamics during the powder spreading process were studied by high-speed high-energy X-ray imaging (Escano et al., 2018). The avalanche angle, slope surface velocity, slope surface roughness and dynamics of powder clusters at the front of the powder were revealed and quantified. In addition, the influence of powder morphology, spreader speed and the powder layer thickness on the uniformity of powder bed morphology were revealed (Mussatto et al., 2021). It was found that the sphericity and surface texture of particles determine the degree of influence of the spreader speed and powder layer thicknesses on the morphological quality of the powder bed. However, the experiments are time-consuming and expensive, which is also difficult to understand the micro scales mechanism of powder spreading processes. Therefore, there are very few reports on the study of powder spreading processes by experiments.

A few works have been completed using the discrete element method (DEM) to simulate powder spreading in AM processes to overcome the experimental disadvantages (Chen et al., 2019; Haeri, 2017; Nan & Ghadiri, 2019; Parteli & Pöschel, 2016; Zhang et al., 2020). Some works of literature focus on particles dynamics (Han et al., 2019; Ma et al., 2020), powder-bed quality (He et al., 2020) and spreader structural optimization (Haeri, 2017; Wang, Yu, Li, Shen, & Zhou, 2021) to reveal the physical mechanism of the powder spreading process. The powder spreading methods play a significant role in particle flow behavior and powder-bed quality. Haeri (2017) used DEM to compare roller-spreading and blade-spreading processes. The lower edge of the blade contacts the powder. The contact area between the roller and the powder was larger than that of the blade, which helped to obtain higher powder-bed quality. This was consistent with research results in the literature (Wang et al., 2020). However, the layer thickness of polymer powder is set to 1000 μm in the work (Haeri, 2017), which is 10 times the particle size. It did not consider the requirements of a thin powder layer in SLS processes. In actual engineering applications, both roller-spreading and blade-spreading are used for polymer powder spreading in SLS processes (Chatham, Long, & Williams, 2019). What is the difference between roller-spreading and blade-spreading mechanisms? How to choose the appropriate powder spreading methods for powder spreading in AM processes? Unfortunately, it lacks a reasonable explanation and selection basis.

In this paper, DEM was introduced to simulate nylon powder spreading by both roller and blade. The powder flow zone in the powder spreading process was divided into the avalanche-zone, slow-flow zone and quasi-static zone. The two spreading methods were compared from several points including particle flow behavior, particle contact forces, forces exerted on spreaders (roller or blade), particle segregation and powder layer density. Finally, the two powder spreading methods are comprehensively compared in order to provide theoretical guidance for the choice of AM powder spreading methods.