Zero Stress On Shell Element in a shell/solid assembly

[Anonymous]

Hi Everyone,

I am running a FE analysis of a structure, the shell part where the load is applied on, shows zero stress. Could anyone let me know what is the reason or what I am doing wrong?

Any comment is appreciated in advance. The shell and solid Von Mises results are presented in these photos.

[Roelof.Feijen]

Did you check the material properties of the shell idealization. Be sure not to use the Generic material property. The pictures shows a min / max value. Does the shell results maybe show a stress singularity on one node for example?

[Anonymous]

@Roelf.Feijen Thanks for your comment. I don't use generic material. a snapshot of shell material is added.

[Roelof.Feijen]

Hi @Anonymous,

It is hard to tell what is going on, based on some images.

Could you share your model?

[Anonymous]

Hi @Roelof.Feijen ,

Not really. If you have such a problem in an analysis which areas do you check? what do you think could be the reason of this.

any thoughts?

Kind Regards,

Ali

[Roelof.Feijen]

It is not so easy to give advise if I don’t understand your problem.

From the title of this post I thought that all shell results are zero. But I am right that there is only one node that has a zero value?

Also I don’t see a displacement value at the summary at the bottom of your results plots. Did you check the displacement results?

If it comes to stress note that by default we are looking at the average nodal results. Is your stress plot showing the average, the min, the max?

[Anonymous]

hi @Roelof.Feijen 

The displacement is:

Stress distribution definition:

Thanks in advance,

Ali

[John_Holtz]

Hi @Anonymous 

Are you sure that the 0 stress is in the shell? From your title, you have solids and shell elements. If I remember correctly (I am answering from home and do not have access to the software), I think the minimum value of 0 is in the solid, not the shell element.

In other words, it is possible that In-CAD displays a value of 0 in all of the solids because the solids do not have a "shell stress". If you edit the contour plot, choose "Part View", and select only the shell parts, then the minimum marker will show the true minimum value in the displayed parts.

[Anonymous]

Hi @John_Holtz 

I checked part view, but I don't know why I can not pick shells!!!. So that, I just used the probe to pick a node stress. The stress on shell element could be seen in the following snapshot which is almost 0:

Thanks for you comment John,

Ali.

P.S: Just in case if it helps, this is the version of my Nastran:

[shigeaki.k]

Hello @Anonymous ,

currently "part view" only works for parts with solid idealization. An idea has already been submitted to the Product Team to allow for the selection of shell and beam parts in "part view" as per,

Ability to select shell or beam meshed parts in "part view" for result plot

Are you also able to share the model?

Regards,

Shigeaki K.

[Anonymous]

hi @shigeaki.k,

I can not share it publicly, but I can do it privately for Autodesk support team.

Thanks,

Ali

[shigeaki.k]

Hello @Anonymous ,

I have sent a private message with regards to uploading the file(s).

Regards,

Shigeaki K.

[shigeaki.k · Accepted solution]

Hello @Anonymous, thank you. We have received the file. I am wondering if your analysis completed successfully. When I ran the last analysis, I noticed that it was struggling to converge. I removed the "arc length method", and all the contact definitions that you had. I then set up a single "solver" contact with activation distance of 50mm and the analysis seems to be chugging along nicely. I will need to wait until the analysis completes to check the shell von mises results. Regards, Shigeaki K.

[Anonymous]

hi @shigeaki.k 

Unfortunately it didn't converge. I am going to update my model to match your suggestions and check the analysis.

Thanks for your try.

Ali

[shigeaki.k]

Hello @Anonymous ,

the model completed successfully. In the meantime, could you send @John_Holtz  or myself a link to a file sharing site you are using via private message for this forum. 

Regards,

Shigeaki K.

[John_Holtz]

Hi @Anonymous 

I wanted to follow up with you to find out where you stand with the analysis. Did you get it to work using Shigeaki's suggestion? Did you provide the updated model that is not working? (Sorry if I missed a message somewhere. )

[Anonymous]

Hi @John_Holtz,

Thanks for your follow up. Yes, I got the analysis working. I used @shigeaki.k comments and it worked.

His comments were:

• use solver contact for contact defenition,
• avoid "arc mesh method".

I provided the model and he worked on that, too.

Regards,

Ali