Some questions (and suggestions needed) with shell analysis

[Anonymous]

I'm analyzing a thin structure, which I wish to model with shell elements. In reality, it's a plastic bag consisting of a composite material, but since Algor crashes when I use composite shell elements - and I wish to include plasticity anyway - I thought to calculate some kind of effective stiffness and utse regular shell elements instead. And since it's actually only the membrane behaviour I'm interested in (for some reason, membrane elements does not support plasticity material models?) I thought to scale the thickness of the shells down a certain amount (and scale the E-modulus 1/that downscaling) to basically remove any bending stiffness of the shell elements while retaining the membrane stiffness. Then I would be able to just look at the membrane stresses of the shell elements...I hope.

I have one or two questions that I would need help with in this, however. First, is there any way in Algor to define a plane stress condition for these shell elements (setting the out-of-plane stress for the shell elements to zero)? Otherwise, I'll be getting far too large stresses obviously, and the membrane stresses will be incorrect.

Also, how do I activate the results window to see the output of membrane stresses? Is this something only available in the linear solver?

Thanks

Björn

[xli]

Since the MES shell expert who should be responseible to your question no longer answer 

so that I am trying to share what I know of shell elements in MES to see if it helps.

Your are trying a hard problem for FEA for sure. Shell probably is the most complicated

element in MES. By "advanced:element formulation" in element definition screen you can 

find three shells:

(1) general, also called original-shell: it is a thick shell and is very close to ADINA's 

    shell formula if they are not exactly same. It has most material models support.

(2) Co-rotational, also called advanced-shell or composite shell, based on thick thoery,

    but it only support linear materials, which means no plasicity as you expect in task;

(3) thin-shell, which is relatively newly impletemented (last year?), it does support some 

    nonlinear materials.

We actually have 4 kinds if count of membrane which is a simplfied variation of shell (2).

But it might crash too :-(. 

Before you get that crash-fixed version, you might try thin-shell which also probably will

support plasticity as wished. I myself have no experience with this shell yet, as I said it is new one.

We do not have user-input plane stress condition as you asked except only in linear thin composite

plate we have some user input materials pattern called flexure control, I am not sure if it is what you asked.

Since the MES shell expert who should be responsible to your question no longer answer so that I am trying to share what I know of shell elements in MES to see if it helps.

Your are trying a hard FEA problem for sure here. Shell probably is the most complicatedelement in MES. By "advanced:element formulation" in element definition screen you can find three shells:

(1) general, also called original-shell: it is a thick shell and is very close to ADINA's shell if they are not exactly same. It has most material models support.

(2) Co-rotational, also called advanced-shell or composite shell, based on thick theory but it only support linear materials, which means no plasicity as your task;

(3) thin-shell, which is relatively newly impletemented (last year?), it does support some nonlinear materials.

We actually have 4 kinds if we also count of membrane element which is a simplfied variation of shell (2).But it might crash too :-(.  Before you get that crash-fixed version, you might try thin-shell which also probably will support plasticity as wished. I myself have no experience with this shell yet, as I said it is new one.

We do not have user-input plane stress condition as you asked except only in linear thin compositeplate we have some user input materials pattern called flexure control, I am not sure if it is what you asked. If that is, you may consider to do a linear model first.

good luck!

-xli

[Anonymous]

Xli,


Thank you very much for your answers. It seems that the maker of the product has determined that the raw material itself does not experience yielding, so I think my best bet might be to use a thin composite shell with the flexural controls. I have never used flexural material properties before, but as I have understood it, the flexural modulus is simply determined from some kind of bending test, to determine the material's stiffness in bending?


Also, do you know where I can activate the "plate results" options in the results environment? What I really want to investigate now that I am using shell elements is the membrane stresses, but these are greyed out and I have not been able to activate it.


Best regards

Björn

[xli]

The flexural modulus input was asked by some customers for direct controls on those flexural modulus instead of calculated from material properties in standard cases, which I am not sure how much it help you. The inoputted values will be directly used in bending. My understand it is sort of providing a way users can use values obtained from their own experiences directly here or make up some numbers for testing purposes. It in fact is also available for MES composite shell (I did not mention).

Both composite shell elements in Algor simulation are linear material which you are now interested. I was a little bit curious of how do you study plasticity in using composite before. You now may try LSS composite first to wait that MES crash problem resolved.

As far as I know there is no simply direct way to activate plate/shell option in results environments. It is decided by program with situation you are displaying, i.e. if it is grey out that it means the model you are looking either does not have data or could not be avaliable for membrane stress or calculation is too complicated for this stress and not supported yet (very small chance). I am not expert of results environment and situations cause the option is grey-out could be complicated. I can not list that combinations, sorry about this. You may need play it a little bit to see if can "figure out" what is preventing the option available.

-xli