PyAnsys - Running PyDPF Script for Large Model

alexdon

Hi,

I have a customer trying to run a PyDPF script on a harmonic analysis model with 4 million nodes and 150 frequency values yielding an rst file 260GB in size. The script is able to extract stresses and accelerations as expected, but when trying to extract element nodal forces, it throws an error: ENF:81<-mapdl::rst::ENF:86<-memory allocation failed, failed to read element nodal forces from rst file. I suggested a workaround of reducing the mesh scoping but that did not help. On the other hand, extracting results for only 10 frequency values at a time seemed to work, though it takes around 10 minutes per named selection. I suspect this may reveal an issue with the element_nodal_forces operator. Customer is using Ansys 2024 R2 and DPF-Core version 13.8 on a computer with 512 GB of RAM.

Thank you for any insights you can provide.

Thanks,
Alex

Rajesh Meena

@alexdon Can you check below things?
1. Are you running out of RAM?
2. Are you able to read forces in MAPDL?
3. Can you try to iterate and find the time point at which it fails?
4. How is user storing results? Using outres and storing on partial domain?
I have seen such issue and mostly it was an issue in rst storing results but it can be something else in this scenario.