Opening problem with Last Iterations

Bug Reports
1

Hello,
I have a small model, 1.29 M nodes, 629 k elements. First increment didn’t converge and the ResultsForLastItreations.frd was 204MB. Reading it into PPM 2.1.9 is surpricingly slow. I waited for four hours, went to sleep and in the morning it had been read. So the reading time was between 4 an 10 hours for one increment’s one iteration only. The model was already open and I only opened the .frd. Is there another route getting the results imported faster?

Environment is Win10, 256GB ram, 32 cores, Raid PCIe M2:s, so it is hardly a hardware problem.

2

Hm, that is strange. Usually, the reading of the .frd files is much faster.

The .frd file we are talking about is the file ResultsForLastIterations.frd or the Analysis.frd?

3

Yes, it is exactly the ResultsForLastIterations.frd, generated by the run.

4

Can you share the .pmx file? Maybe privately? Or can you prepare a simple example that could be shared?

5

The version v2.1.11 fixed this issue.

The problem was that setting the results for the last iteration meant exporting all contact elements as beam elements to the .frd file. Strangely, there were multiple (100s) contact elements between the same nodes. This messed up the PrePoMax part recognition and part geometry division algorithms. So currently, PrePoMax removes all duplicated elements if the file name is ResultsForLastIteration.frd.

6

I have moved this topic to bug reports. Please confirm it is fixed.