How to store FEM results in Windchill

[tleati]

Hello,

I am going to upload some structural FEM results in Windchill and I am just wondering how to store them:

-the complete result file? --> in this case the Simulate saves a back-up of the model and then I will find the back-up in Windchill? (thing that I want to avoid to prevent having double products in the Common Space)

-Or am I forced to load all the folder (with the back-up)?

Anyone has any suggestion on how to do this?

thanks

bye bye

[cc-2]

Hello,

very interesting question. I faced the same problem a few years ago. I am a strong advocate to keep things in one place with the relationship and Mechanica files are one of those.

One thing I still do not like is that you can't create your constraint etc... directly into the Creo file without creating a new iteration so we ended up creating a new model inherited from the original one for obvious reasons.

PDMLink will take easily all the files created in mechanica to define the conditions of the analysis and create the relations with the model.

All those files are quite small size.

However the results files can be very large and therefore, we decided not to bother with it. WHat was important was to be able to recalculate with the same conditions at any time in the future (ok our calculation lasted only a few hours so not a big deal if needed to recalculate).

The report created from the results was also uploaded in PDMLink and nicely tied up with the rest of the documentation.

Hope this helps

[tleati]

Hi Chris,

thanks for your reply. When you wrote about creating a new iteration, were you meaning for already checked-in models?

I have just made some more specific attempts, and I came to these conclusions:

- from Analysis and Design Studies window I can set as results (temporary and final) storage directory neither the workspace, nor the common space.

- from the workspace I can import neither the results folder, nor a simple .rwd file, since the import interface does not recognize it.

- The same of the precedent point is valid for the common space

- zip files cannot be imported in the workspace

I have also found this article which points out uncompatibility with windchill:

http://www3.eng.cam.ac.uk/DesignOffice/cad/proewild3/usascii/proe/promec/online/database.htm#windchill_support

so the only way to store the results would be to upload a zip file of the entire results folder in the common space.

But this would just be equal to double the space occupied by the folders, might as well keeping them where they are, outside Windchill, on a regularly back-upped server...

It would be much more convenient, instead of making the simulate back-up files in the result folder, to have the results directly linked to the corresponding version in the common space...

and with report of the result what do you mean? is it your own internal procedure or is it a function of simulate?

thanks

bye

[ChrisPLM]

Hi

as said, we did not store the results file. We only stored associated with the version of the model used the conditions so we could reproduce the exact same calculation in the future (of course if you had the result files, it would be quicker, just open the result file but they are too large and our calculation only last a few hours).

For the reports, our own report.

Best regards

[HugoHermans]

Hi Tommaso,

We also only store the input, never the output.  We have several engineers working daily with FEM, for years now, and we've never had bad experiences by not storing these results in Windchill.  We keep them for a while on network drives, organized based according Windchill data, so that there is a quick and conprehensive path from the Windchill repository to these results.

Hugo.

[ChrisPLM]

Hello Hugo

long time not seen.

I am glad you have the same experience that I have. You keep everything so the exact same calculation can be made (ie giving the same results files) but you do not keep that huge file.

Best regards

[tleati]

Hello Hugo and Chris,

the problem in your procedure of storing only the input files is that in my country, according both to international standards and national laws, we must keep the FEM calculation results stored in order to, for instance, promptly demonstrate the quality of the designed product in case, for example, of a controversy due to a contigent accident (fingers-crossed ) --> if we are subject to an ispection by authorities we cannot tell them to wait for the calculations to be done , even if they last a few hours (which is not our case however) therefore the results must be there ready to be opened.

Bye bye

[ChrisPLM]

If by law you must keep those files, good luck. I wonder in which country and industry you work. I have never heard of that before.

If your IT department is happy for you to upload all those files in to PDMLink……

Otherwise you could create a Document Link that will point to where those files lives outside PDMLink, on a shared drive. 

That document link could be added to your WTpart.

[tleati]

Maybe it's you who is working in a strange country, there are international standards (ISO) and every industry which certificates its products according to certain specifications should have all the necessary calculation results as proof of compliance to those same standards.

But this It's basic don't you think so??

For me there is not a problem of loss of data, I store the FEM results on our server and stop, but it seems very bizarre that the user cannot have these files available in windchill so that he can:

- use them in the workspace while designing without having to surf in the windows folders

- neatly store them in the respective folder in the Common Space and readily open them.

bye bye

[ChrisPLM]

Hello Tommaso

maybe you are right   Germany and England can be strange countries sometimes.

More seriously,

There is no right or wrong answer. We are just sharing with you our experience.

If your IT department is happy to load files over GB big, that is fine.

I personally believe that the files that create the conditions to calculate and therefore produce the results files are more important than the results files themself.

and I think PTC may think the same since you can automatically upload the files that defines the constraints etc...... and not the results file. You can still upload manually into whatever WTdocument Subtype you want and then associate them with your other document or WTpart.

The good thing is that you should not really have any iteration of those big files.

I wish you all the best with whatever route you take.

Best regards

[GregoryPERASSO]

Looks like the same kind of discussion for "Long Term Archiving" of CAD data models ....

If the 3D model is used as unique reference for MBD,  certification regarding laws etc ...

The fact that the CAd sofware is enhanced accross his release lifecycle , can change the 3D modelling core engine ... and change the resulting 3D model ... (I remember some huge bugs in Creo when PTC introduce for example unit management in parameters ....  all ou familly tables have their size multiplicated by 1000 !)

I assume it is the same problematic for FEM....

have to garantee that the results will be the same .... 10 years later or more in the latest release of FEM software ... if the FEM software still exists  

Storing results as "neutral" references is probably more reliable, even if I'am agree not the typical use case

regards

Gregory

[tleati]

Hi Greg,

yes in fact as I just wrote to Chris the problem is that the FEM results must also be dated in correspondence to the product design period, for legal reasons too.

Bye bye

[tleati]

Yes it's not a problem of my IT department, we have memory space and it can be augmented too if we want, the problem is as I wrote is the prompt availability of the FEM results directly in Windchill, so that I can have only one platform where I can work on my CAD models and their calculations.

And yes, I also think the input files in themselves are more important than the results, since from the former you can easily get the latter, but not so easily the contrary (from FEM results to the model input files).

I thank you for your replies and support but, what maybe it's not clear to you is that we need to have the FEM results stored because they are a document, related to the product designed, that must be present in our database: for example, in case of a controversy (due, for example, to an accident that, God forbid, may happen ) we would have to immediately proof that at the moment we designed that product, the product has been conceived and developed according to the state of the art and in the scope of guaranteing a correct working order during its employment and in total safety for the people and the environment envolved in its utilization.

If something happens and a law technical consultant (I don't know if it's called this way in english) comes to us and ask for all the documents related to the product/s under investigation, we must have immediately all the calculations (including FEM results) that can proof we designed the product in good faith (bona fide) and according to the stardards requested (or also simply suggested) for that kind of product --> they have a legal value!!

We simply cannot perform the calculations just at the time the results are requested, and for two reasons:

1) they may take hours! and if there are many products under investigation?? Should we say "oh wait just for some days for the software to calculate"?

2) The calculation results must be dated BEFORE the actual put in production of the product -> if not, one might think we fixed (rigged) the FEM input files (without saving the model) at the moment of the investigation in order to make the results in compliance with the standards (while before they weren't).

anyway, I regret to say I think it's quite a big miss by PTC, and it's not the first time: for instance some days ago I realized that Creo View 3.0 Lite/Express does not have a tool for measuring the volume (and the mass) of a model (tool that is preesent instead in previous versions) --> for a software paid tens of thousands of euros one expects to have at least all the basic functions!

thanks

Best regards