Speaker
Description
In the domestic metal powder manufacturing industry, there is an ongoing effort to promote the transition to the next-generation smart factory by the digitization of tacit knowledge using artificial intelligence (AI). This successful initiative enables one company to become global competitiveness in high-efficiency metal powder production. Although the previous AI-based works in our laboratory predicted the powder dimension (D50) with a reliability of over 95% in the various inputs and results, the control over target size on-demand still requires the labor-intensive skills and intuitions in diverse situation.
In this study, we experimentally and theoretically suggested that the particle size and the dispersion of metal powder can be controlled through the molten metal level in crucible of a gas atomization system.
1. First, the theoretical analysis was conducted using the Lubanska’s Equation, which predicted the median particle size (D50) of metal powder scaling as D50 ~ (QM)^0.5 ~ (PM)^1/4~ (h)^0.25 ,where D50 is the median of the produced metal particle, QM is the flow velocity of the molten metal, PM is the molten metal pressure in crucible ground and h is the molten metal level in crucible)
2. During the metal powder production in our system, where the tundish is fixed, the metal powder dimension and the dispersion showed the increasement depending on the molten metal level.
3. Based on the above evidences, we introduced a programmable induction dish, which can maintain the molten metal level in tundish by angular control. In lab-scale approach, a silicon oil (0.65cs), which viscosity was similar to ferrosilicon alloys in the temperature of 1500F, was chosen for pre-experimental system. This experiment clarified not only for the viscous flows by the delivery tube from tundish, but also for the droplet formation, which represents the metal powder, in gas atomization system. Using high-speed camera (~10,000fps) in 768mm X 528mm optical area, the clear images of droplets were captured the experimental relationship between the molten metal level and the diameters in the end.
This experimental and theoretical research would pave the way not only for the customer-demand powder dimension in high quality, but also for the cost-effective and the profitable powder production in the relative industry.
References
[1] Umbrello Steven. (2020) "Meaningful Human Control Over Smart Home Systems" HUMANA.MENTE Journal of Philosophical Studies 13 (37): 40–65.
[2] Susanne Vernim, Harald Bauer, Erwin Rauch, Marianne Thejls Ziegler, Steven Umbrello. "A value sensitive design approach for designing AI-based worker assistance systems in manufacturing." Procedia Computer Science (2022) 505-516.
[3] Zhang, L. C., Klemm, D., Eckert, J., Hao, Y. L., & Sercombe, T. B. “Manufacture by selective laser melting and mechanical behavior of a biomedical Ti–24Nb–4Zr–8Sn alloy.” Scripta Materialia, (2011). 65(1), 21-24.
[4] S.M.E. Sepasgozar, A. Shi, L. Yang, S. Shirowzhan, D.J. Edwards “Additive Manufacturing Applications for Industry 4.0:.” a Systematic Critical Review Buildings, 10 (12) (2020), p. 231.
[5] Rottondi, Cristina, et al. “Machine learning-driven optimization in powder manufacturing of Ni-Co based superalloy.” Materials & Design 198 (2021): 109290.
[6] Y.K. Mohanty, B.P. Mohanty, G.K. Roy, K.C. Biswal. “Effect of secondary fluidizing medium on hydrodynamics of gas-solid fluidized bed—statistical and ANN approaches.” Chem. Eng. J., 148 (1) (2009), pp. 41-49
[7] Wen See Tan, Stanislaus Raditya Suwarno, Jia An, Chee Kai Chua, Anthony G. Fane, Tzyy Haur Chong, “Comparison of Solid, Liquid and Powder Forms of 3D Printing Techniques in Membrane Spacer Rabrication,” Journal of Membrane Science (2017) 283–296.
[8] H.Lubanska, "Correlation of Spray Ring Data for Gas Atomization of Liquid Metals", The journal of the Minerals, Metals and Materials Society, (1970). 22, p. 45
| Keywords | Gas atomization, Tundish angular control, Particle size(D50), Molten metal level in crucible |
|---|
