2023 Fiscal Year Annual Research Report
粉体シミュレーションとデータサイエンスの融合による粉体混合メカニズムの解明
| Project/Area Number |
23KJ0594
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
SHI QI 東京大学, 工学系研究科, 特別研究員(DC2)
|
|---|
| Project Period (FY) |
2023-04-25 – 2024-03-31
|
| Keywords | powder mixing / data-driven method / discrete element method |
|---|
| Outline of Annual Research Achievements |
This research project, conducted over the past year, has developed and applied advanced mathematical-analytical methods for a detailed analysis of powder mixing processes. The project successfully established a novel proper orthogonal decomposition (POD)-analysis of variance (ANOVA) framework, tailored for discrete element method (DEM) simulations in Lagrangian particle systems. This innovative approach was crucial in quantifying convection and diffusion in powder mixing processes, leading to a deeper understanding of the complex interactions within various mixing systems.
First, the POD-ANOVA method was rigorously tested and validated using a benchmark rolling drum system. This validation demonstrated the method's effectiveness in simpler systems. The approach was then extended to more complex industrial container blenders, showcasing its versatility and robustness in analyzing complex mixing mechanisms influenced by airflow. The later period focused on establishing a time-localized reduced order modeling (ROM) strategy, which was applied to a benchmark rolling drum system. This strategy led to efficient reconstruction of particle flows (particle positions and velocities) within the rolling mixer. This model significantly improved simulation accuracy while reducing computational costs and time, as compared to the conventional standard ROM approach. The findings and methodologies developed in this study have been published in high-quality international journals and presented at numerous prestigious conferences.
|
Research Products
(6 results)
