Inventor Optimization

[tolgacankurt]

Hello,

I have used optimization for some of my assemblies. But now this assemblies file size gone too bigger. about 250mb... I want to delete optimization part of this assembly. Is it possible?

[henderh]

Hi tolgankurt,

  Are you saying the assembly file itself increased by 250 MB after using Optimization and saving?

  Optimization shouldn't induce any file size bloat.  All the "heavy" result files reside in the subfolder next to the model file on disk.

  Which version are you using R2012 or R2013?

Thanks, -Hugh

[Edit: I can reproduce it now with a simple assembly dataset:

Initial size before entering Optimization: 56 KB

Size after setting up Optimization: 64 KB (+14% change from original)

Size after solve completes and save: 80 KB (+43% change from original)

Size of fmsh and fres result files are 46 and 42 KB respectively ]

[tolgacankurt]

I am using R2013, assembly files itself is 250mb

[henderh]

Thank you for the requested info.

  I'm trying to reproduce this on my end, with a larger weldment assembly.  After I save the assembly with Optimization result data downloaded, the file size doesn't increase dramatically (0.999 to 1.041 MB)

  Would it be possible to share your dataset before the bloat occured, so we can easily reproduce it on our end?  There may be one saved in your "old versions" folder.

  If you can share it, please zip up the assembly + components (no result files needed) and send it to hughDOThendersonATautodeskDOTcom

  We appreciate you bringing this to our attention and apologize for the inconvenience this is causing.

-Hugh

[tolgacankurt]

I have uploaded files

[henderh]

Many thanks tolgacankurt,

  I've downloaded the dataset and we'll keep you updated on what we find.

Best regards, -Hugh

[tolgacankurt]

any news?

[henderh]

Hi tolgacankurt,

  The 250 MB assembly file is considered  corrupted by WinRAR, 7-Zip and Inventor R2013.

  Were you able to find a non-bloated version of the assembly file?  (the small file size assembly before you sent the solve off to the cloud?)

  I couldn't reproduce the issue creating a new assembly with the components provided.  The assembly file size went from 85 to 95 KB after the solve and saving.

 If I had a screen capture to reverse-engineer the assembly it would give us another data-point to try and reproduce the issue.

  My sincere apologies for the late reply.

Best regards, -Hugh

[miechh]

I'm also curious to the answer of the initial question: Is it possible to detach the optimization files which are loaded and saved with the part that has been optimized? Now a few extra folders and 12 extra files are added to the Vault and workspace when optimization has run (See attached picture).

[henderh]

Hi tolgacankurt and miechh ,

  I'd like to express my sincere apologies for the lack of timely follow-up and the inconvenience this has caused.

   Indeed we have this issue in both Optimization and Stress Analysis of embedding the result files into the model files in the R2013_RTM build.  We've addressed the issue in Update 2 and the upcoming SP1 so that it will no longer occur with any new simulation created.

  However, for the model files already bloated with result files, you'll need manually purge the embedded result files.

  Bob van Der Donk has posted the workaround in his blog here:  http://beinginventive.typepad.com/being-inventive/stress-analysis/  I've pasted the purging steps below.

Many thanks again for your time, patience bringing this to our attention. 

Best regards, -Hugh

=========== From Bob's Being Inventive Blog========

• Delete the FEA simulation data folders for the affected files.
  These are subfolders in the location where your FEA parts are stored and they have the same name as the FEA parts.

  Example: if your part is stored as c:\workspace\test.ipt , the FEA (.fres) and mesh (.fmsh) files will reside in c:\workspace\test\AIP\FEA

• If you have FEA files in your %temp% folder from a previous solve, delete them.
• Open Inventor and for each existing 'bloated' model file with a simulation, do the following:
  1. Open the file and skip the request to find the missing linked files.
  2. Use the 'Tools > Links' command and delete the OLE links for all stress analysis files ( .fins, .fsat, .ftes, .fwiz, .fmsh, .fres).
  3. Save the file.
  4. Enter the Stress Analysis environment and re-run the simulation – this should take some time to recreate all of the results.
  5. Save the file again.
  6. Check the 'Tools > Links' command again – the 'link size' should now be much smaller

===========

[henderh]

Hi miechh,

  <<Now a few extra folders and 12 extra files are added to the Vault and workspace when optimization has run...>>

  Optimization result files consist of just the <GUID>.fmesh and <GUID>.fres files.  GUID = global unique identifier, which is basically a unique random name.

  We create these files on disk after you save the model file with results present.  If you change the setup and re-save, we don't create new result files and leave the ones from the old result 'orphaned' on disk.  We simply use the same GUID and replace the files.

  I'm not sure how you got multiple files with different GUIDs for the same model / simulation.  I assume you did have the result files checked out of Vault when you saved the new results and then checked them in?  I don't recall seeing this in my Optimization <=> Vault testing, and I cannot reproduce it when manually making the result files read-only and saving with new results.

  We will investigate this behavior ASAP and get to the bottom of why this is happening.

Thanks, -Hugh