Hello everyone,
I am currently working with a highly partitioned geometry in Abaqus/CAE (more than 50,000 cells), and I am experiencing significant performance issues when assigning local seeds, as well as when propagating and merging seeds across the model.
In particular, operations such as:
Assigning local seeds to edges
Propagating seeds across partitions
Merging seed definitions
are becoming increasingly slow and difficult to manage as the model size grows.
My objective is to generate a large-scale model with on the order of 20 million elements, including multiple transition zones and sufficient local refinement to avoid distorted elements and ensure good mesh quality.
For reference, my workstation specifications are:
CPU Model: Intel Xeon Gold 6230R
Physical Cores: 52
Logical Processors (Threads): 104
Total RAM: 256 GB
In this context, I would like to better understand:
How the meshing module algorithm handles seed assignment and propagation in large, highly partitioned geometries.
Whether the performance bottleneck is related to topology , dependency tracking between partitions, or another internal mechanism.
If there are recommended best practices to improve performance in such cases.
Whether there are any upcoming developments or improvements planned in Abaqus/CAE to address meshing performance for large-scale models.
From my experience, improving this aspect would be a highly beneficial upgrade for model definition workflows in large Abaqus analyses.
Any insights, explanations, or suggestions would be greatly appreciated.
Thank you in advance!
