[1J01-04] 原子力計算科学技術1

[1J01] On predictability of POD-based ROM in Lagrangian particle methods

*Guangtao Duan¹, Shuo Li¹, Mikio Sakai¹ (1. University of Tokyo)

キーワード：Reduced order model, Proper orthogonal decomposition , Error analysis, Lagrangian, DEM-CFD, Particle methods, Feasibility analysis

The predictability of POD-based ROM for Lagrangian simulations of DEM and MPS particle methods is explored. A new error analysis is performed to investigate the feasibility of POD-based ROM in arbitrary complicated systems. It is found that mapping the Lagrangian information on Eulerian mesh can improve the predictability.