Show membrane only from Von Mises

[ptm0270]

I have a question on how to extract membrane only stresses. Please correct me if any statements I make are incorrect. I'm using I-DEAS NX5 and occasionally NASTRAN from an I-DEAS .dat export. The element I'm using is a thin-shell to represent typically angles and plates less than 1". In I-DEAS, when I post-process Von Mises stress, I get "top" and "bottom" surface stresses. It is my understanding that this is considered "Membrane+Bending" stress. I want to also view "Membrane only" stress, which I believe is the stress at the neutral axis of the section. It is my understanding that the "middle" surface in I-DEAS is just (top+bottom)/2 or the average, but this is not what I'm trying to extract.

To sum things up, in I-DEAS how would you extract "Membrane only" stresses using Von Mises stress display for a thin shell element.

Thanks

 

[ESPcomposites]

I'm not an IDEAS user, but I am curious about this:

"It is my understanding that the "middle" surface in I-DEAS is just (top+bottom)/2 or the average, but this is not what I'm trying to extract."

Are you saying that IDEAS has the option to use the "middle" surface, but that you don't believe it would be correct? If the N.A. was at the mid-surface of the shell, then the formula you provided would work.

Brian

http://www.espcomposites.com

 

[ptm0270]

How could (top+bottom)/2 = membrane stress when viewing Von Mises stress? Von Mises stress averaging of top and bottom will not have bending stress equal to zero at the neutral axis doing it the way shown above.

 

[ESPcomposites]

I suppose if it is implemented exactly like what you are saying, it might not work.

But I would have thought it would first average the raw stress components and then create a Von Mises stress. For example, if it is pure bending, you would have compression on one side and tension on the other. The provided equation would cancel these out and leave you with a zero stress, as expected. The same approach could be said for the pure axial load or other cases.

Does IDEAS do it this way or the other way (i.e do you have some test case results)? I have no way of checking IDEAS, but it would seem odd (to me at least) if it is actually doing what you are saying.

Brian

http://www.espcomposites.com

 

[weab]

I am still confused. Are you trying to determine the inplace tension/compression forces or stresses? If so, then it is (top + bottom)/2. Middle is your neutral axis and doesn't have bending in it. The top and bottom have in plane plus +bending and -bending respectively. Average cancels out the bending.

 

[rb1957]

in FeMap (ont beig an IDEAS user) you can get top surface stresses, bottom surface stresses, in-plane load and in-plane moments. i wouldn't try to average top and bottom surface stresses ... too much work (even though it's not That much in excel) ... look for in-plane loads.

 

[trainguy]

In FEMAP, you could also override the definition of where the stresses are extracted and set it to mid thickness. Then the "surface" stress plots will actually display stresses at the NA, where you want them.

Don't know if you can do that with Ideas.

tg

 

[ptm0270]

I'm not trying to get top and bottom. What I'm trying to get is middle fiber stresses (membrane) in I-DEAS using Von Mises criterion. Theoretically for Von Mises, this should only include axial and shear.