hi, after finished running analysis in PrePoMax i opening result as usual but the PC almost hangs and needed to restart. I tried also opening directly in new files and the problem persist, however double click to read by CGX only in second. Model and mesh are not too large but highly nonlinear with incremental times, size about 3.5Gb.
maybe this is a bug but not sure, any hints about the condition? thank you.
i return to last loading setup by reduction, sized of Frd file also smaller about 1.2Gb. Result is successfully displayed even loading still long by several minutes.
Testing on simple case frequency analysis of cylinder shell with layered composite option shown similar behavior in loading results. I only guessed, maybe it is related to shell element expansion and grouping since PrePoMax generate part for each layer thus make additional working load much higher, an option to disable may help to accelerate if the condition is true.
How much RAM do you have? I just opened a large .frd file of one of my projects, about 4.8 GB, it took about 18 GB of my PC RAM to load it, I have 64 GB RAM. Maybe check your task manager for RAM usage when you load your result file.
thanks for hints, loading results is required almost four times bigger in memory of output file sizes. Is that similar in case of shell with composite option and truly solid model, also how about times in loading when memory fit the output size?
Output file is just lines of data albeit millions of lines, you can open and edit them with a text editor. I don’t know how prepomax handle the parsing of the data, but it is nothing to do with type of model as everything is represented in lines and lines of numbers.
indeed, Frd output files are just data but how CGX can be so fast in loading at the same machine? also no limitation of memory maybe.
as i remember PrePoMax changes the behavior of post processing in shell element with layered composite options later, each layer is grouped into part. That’s great feature, so mid surface or layer interest can be isolate to shown individually. However, extra work in grouping may lead to memory consumption i guessed, option to disable i.e single part as in early implementation similar with CGX may help to accelerate the process of loading.
Is it possible to prepare a someple test file where I could see what is going on?
thanks for looking further into the problem, attached link of Frd file.
Thank you for sharing the file. I was able to finde some bottlenecks (they were mosly in reading the file and not geometry processing) and improve the reading speed. I applied some parallel routines where I could. There is a lot of parsing going on so I dont know how to make that faster.
The fix will be released in the next developer version.
Since this is not a bug I am mobing the topic to Genereal Questions.
