.png){kind=icon}
Return
A novel normalized radial basis function method for data transfer between non-matching meshes
.png){kind=icon}
PRE
AI
DOI：10.1177/09544070251386511.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The coupling analysis often allows different subdomains to be discretized independently for greater computational flexibility, but which may induce non-matching meshes at the common interface. To reduce numerical oscillations caused by mismatched meshes during load transfer, a novel radial basis function method with repeatedly normalized weight matrix (RN-RBF) is proposed in this paper. Compared with the standard RBF and a combined weight functions method in several 2D and 3D examples, the RN-RBF method exhibits superior robustness in improving transfer results, particularly when dealing with significantly finer target meshes than source meshes. The balanced accuracy in load and displacement transfer, along with compliance to the principle of energy conservation, render RN-RBF suitable for two-way coupling analysis as well. A non-intrusive iterative algorithm based on RN-RBF achieves the minimum computation error to the reference solution in a 2D problem of an open-hole plate, surpassing those algorithms combining one-way methods separately focused on enhancing load transfer or displacement transfer. The validated RN-RBF iterative scheme is further applied in the decoupling analysis of biaxial wheel fatigue simulation, and the calculated wheel stresses agree well with the experimental results. This substantiates its applicability for complex engineering problems that require coupled analysis but have poor convergence.
Keywords:
data transfer
non-matching meshes
radial basis function (RBF)
repeated normalization
biaxial wheel fatigue simulation
Journal
P
IF:
1.5
Papers:
593
Citations:
0
Organization
Cited Papers
A partitioned coupling framework for peridynamics and classical theory: Analysis and simulations
OAAI
.png){kind=icon}
OAAIShareSaveShareSaveShareSaveShareSaveShareSave
View more