.png){kind=icon}
.png){kind=icon}
Return
Cuda-based parallel dual-grid material point method for simulating bimetallic coining process
.png){kind=icon}
PRE
AI
DOI：10.1007/s40571-025-00955-8.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In the simulation analysis of the bimetallic coin embossing, challenges such as mesh distortion, managing contact between different materials, and the computational demands of intricate relief patterns arise. To address these issues, this paper proposes a dual-grid material point method (MPM), which avoids mesh distortion and decreasing time steps that often occur in finite element methods (FEMs). The method utilizes surface patches for the rigid mold and material points for the blank surface, employing a point-surface contact algorithm to resolve virtual contact. For the contact between the core and the outer ring of the blank, two sets of background grids describe their movements separately until real contact occurs. CUDA parallel acceleration technology is applied to the MPM to manage large-scale computations. The advantages of MPM over FEM are shown through a comparison of calculation results and efficiency using a single-color commemorative coin embossing example. Additionally, embossing results for six different gaps between the core and outer ring of bimetallic coins are compared to optimize the embossing gap, achieving a maximum speedup of 39 times on a single CUDA machine.
Keywords:
Material point method
CUDA parallel
Virtual contact
Bimetallic coin
Journal
IF:
2.8
Papers:
245
Citations:
2.0K