Elsevier

Advanced Powder Technology

Volume 30, Issue 11, November 2019, Pages 2726-2732
Advanced Powder Technology

Original Research Paper
Numerical simulation of high-pressure gas atomization of two-phase flow: Effect of gas pressure on droplet size distribution

https://doi.org/10.1016/j.apt.2019.08.019Get rights and content

Highlights

  • Eulerian-Eulerian formulation is used to capture shape and size of droplets.

  • Trends in simulation and experiments are compared favorably.

  • Rate of the atomization increases with the gas pressure, resulting in higher yield.

  • Higher gas pressures advect droplets at a higher rate than they are atomized.

  • Atomization process evolves continuously along the axial direction.

Abstract

This paper deals with the physics of high-pressure gas atomization in metal powder production. To gain understanding of the effect of gas pressure on droplet size distribution, a numerical two-phase flow study is performed using Eulerian-Eulerian Volume of Fluid (VOF) interface tracking method. Annular-slit, close-coupled gas atomizer is considered to atomize molten aluminum using nitrogen as the atomizing gas. Four cases with different gas pressures are considered, while geometry and other operational parameters are fixed. Characteristics of several interfacial instabilities have been identified at different stages of the atomization process. Despite the increment in the rate of the atomization with the increasing gas pressure, deformation characteristics and the breakup mechanisms remain unchanged. Droplet size and the cumulative volume distributions indicate that the effectiveness of the atomization process increases with the elevating gas pressure. Cumulative volume obtained from the numerical simulations at low gas pressures display similar trends to the experimental results.

Introduction

Metal powder production is in high demand for its applications in rapid prototyping, injection molding, and additive manufacturing. Metal powder used in additive layer manufacturing is required to have precisely tailored metal powder with a certain size, shape, and morphology [1]. Lawley [2] reviewed many existing powder manufacturing technologies and discussed the superiority of gas atomization among other methods considering its controllability over powder size distribution. In gas atomization, a pressurized gas is used to atomize the molten metal. Momentum and thermal energy [3] transferred from the expanding gas to the molten metal facilitates disintegration and atomization. The rate of interfacial momentum and thermal energy transfer depends on the driving potential and the interfacial area. Since the cumulative interfacial area increases with the atomization process, it increases the atomization rate to several orders of magnitude [4]. The main advantage of the gas atomization is the controllability [5] of the powder size, shape, and the morphology. It is important to study the effect of each operating parameter on the atomization process and the powder distributions.

The atomization process can be divided into two regimes such as primary and secondary atomization (Fig. 1a). The primary atomization regime starts when the melt interacts with the gas flow near the melt tip. Then, the melt stream will deform and create unstable wave-like structures, when it breaks up into ligaments and large droplets. The secondary atomization regime begins further downstream when the large droplets and ligaments extend due to Rayleigh-Plataeu instability and break up into smaller droplets. This breakup process will continue until the critical Weber number is reached or till the droplet solidifies [1], [6].

Several numerical simulations have been performed to study different aspects [7], [8], [9] of the atomization process. Some of these early studies could not successfully predict the atomization process as they neglect the strong feedback from the highly dense melt stream [10]. The Eulerian-Lagrangian approach [3], [11], [12], [13] was first used to investigate the molten metal atomization and particularly the secondary breakup. The main drawback of these discrete element methods is their inability to predict the primary atomization [10]. These simulations often used random droplet size distribution [14] (DSD) to initiate the simulation. Considering these drawbacks, Tong and Browne [10] performed the first Eulerian-Eulerian based simulations using the front-tracking method. This and other papers [15], [16], [17] showed that the gas dynamics can be significantly affected by the topological evolution of the melt stream. However, these studies were either conducted in two-dimensional geometries [10], [15] or considered only the early stages of the atomization process [18].

In this study, the gas atomization process of molten aluminum at different gas pressures is studied using the Volume of Fluid (VOF) model [19] together with geometrical reconstruction method in OpenFOAM software. The method is Eulerian-Eulerian and does not involve any semi-empirical correlation or any facilitation of droplet breakup as it occurs naturally without any modeling. The effect of atomization gas pressure on the droplet size distribution is investigated. To obtain the effect of inlet gas pressure, four inlet gas pressures are considered while keeping other parameters such as melt flow rates, melt and gas physical properties the same. This paper deals with both primary and secondary breakup processes in conjunction with the calculation of dropsize distribution to understand the trends in gas pressure. Even though a very fine mesh is required to resolve both types of atomization completely, some characteristics of the secondary breakup process have been identified in the simulation through the calculation of ligament aspect ratio. The objective of this work is to predict the trend of the drop size with pressure and compare this trend with the experimental trend by discriminating dropsize based on aspect ratio.

Section snippets

Computational model

The computational domain (Fig. 1) is designed based on a double induction, discrete nozzle, close-coupled gas atomizer, which consists of 18 circular gas nozzles evenly spaced around the melt tube. The computational domain follows the experimental set up (not shown here) but uses an annular-slit instead of individual nozzles. Since the velocity in the slit is maintained the same as in experiments, the mass flux is different and only trends in pressure can be determined. Once the melt enters the

Results and discussion

Since a direct comparison of numerical and experimental data is not possible for reasons mentioned in the previous section (i.e., an annular-slit is used in computations with the gas velocity maintained at the experimental velocity; the use of incompressible flow; a smaller computational domain), only the trends obtained for different gas pressures are compared. The melt diameter, funnel-length, and gas nozzle angle are maintained the same as in experiments.

Fig. 2 shows a qualitative comparison

Conclusions

In this numerical study, four gas pressures have been considered for gas atomization of aluminum melt. The atomization process is simulated using Eulerian-Eulerian two-phase numerical framework. The rate of atomization is seen to increase with increasing gas pressure. Droplet size distributions indicate a clear improvement in atomization. Higher gas pressures tend to advect droplets at a higher rate than they are atomized, which may necessitate a larger computational domain. Cumulative volume

Acknowledgements

This research was sponsored by the U.S. Army Research Laboratory through cooperative agreement #W911NF-17-2-1072 between the University of Central Florida, United States and the U.S. Army Research Laboratory. The views, opinions, and conclusions made in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the U.S. Army Research Laboratory or the U.S. Government. The U.S. Government is authorized to reproduce

References (27)

  • D.A. Firmansyah

    Numerical simulations of supersonic gas atomization of liquid metal droplets

    Japan. J. Appl. Phys.

    (2014)
  • U. Fritsching et al.

    Hybrid gas atomization for powder production

    Powder Metall.

    (2012)
  • I.E. Anderson et al.

    Feedstock powder processing research needs for additive manufacturing development

    Curr. Opin. Solid State Mater. Sci.

    (2018)
  • Cited by (37)

    View all citing articles on Scopus
    View full text