Cancel
Showing results for 
Search instead for 
Did you mean: 

Re: SOL111: crashes because of memory issue

Valued Contributor
Valued Contributor

Hi

I am not doing restarts since the time for solving normal modes is only about an hour, and then I can save some disk space which I am running out of.

My constraints on the excitation node are simply

SPC1 1 123456 1000001

 

Peter 

Re: SOL111: crashes because of memory issue

Phenom
Phenom

ditto for me. Not sure if one could save space requirement in the subsequent SOL111. Mind you one needs to store .dball and .master so might make no real difference

 

just did a bit of clean-up on my local disk and also reduced the number of elm for which I want rms stress - Will see

Production: NX.CAE 10.0.2.6
Development: VB.NET (amateur level !)

Re: SOL111: crashes because of memory issue

Valued Contributor
Valued Contributor

I also did a frequency response run (without asking for stresses), and had a look at the number of solution frequencies. Using that info I also reduced my solution frequencies on my FREQ cards - especially at higher frequencies.

Highlighted

Re: SOL111: crashes because of memory issue

Phenom
Phenom

did a test with a reduced elements set(56,000 2D elements only) and sol111 ran succesfully. 236modes all used, 1Hz freq. increment

 

 

Production: NX.CAE 10.0.2.6
Development: VB.NET (amateur level !)

Re: SOL111: crashes because of memory issue

Valued Contributor
Valued Contributor

What changed? How many elements did you ask for previously?

Re: SOL111: crashes because of memory issue

Phenom
Phenom

146,000 2D & 3D

Production: NX.CAE 10.0.2.6
Development: VB.NET (amateur level !)

Re: SOL111: crashes because of memory issue

Phenom
Phenom

@selex_ct wrote:

did a test with a reduced elements set(56,000 2D elements only) and sol111 ran succesfully. 236modes all used, 1Hz freq. increment

 

 


What we typically do is run a SOL103 first to get the MPF, then truncate using MODSEL.  1TB scratch is what we have to use most of the time... We cutoff at 90% MPF in all axes if possible, luckily our modes are not much cross correlated so we have a much smaller truncated set of modes, which helps tremendously. SOL111 is bad, but wait until you do SRS, this leads to even larger scratch files...

Re: SOL111: crashes because of memory issue

Phenom
Phenom

I haven't used the MODSEL (yet) as I wanted to how far I can go/push it with default. In my test all modes (236) were used. When I did the work with Nx Resp. sim I set the treshold on the MEFFMass to 0.1%. This cut the number of modes used a fair bit and on the max stress it made a difference of 0.5MPa. Because sol111 was a lot faster to deal with the rms VM stress (though I still need to see if i can actucally run 56000 shell with Nx Resp. sim in less than 1.5hr) I can actually create a stress plot of the entire part with different max/min stress plot to see the "hot spot" , etc. This was the original aim.

Production: NX.CAE 10.0.2.6
Development: VB.NET (amateur level !)