von Mises stress larger than maximum stress defined in material stress-strain curve

[SuG89]

Hello,

I am a beginner in Femap. I have defined the following stress vs strain curve for a steel material:

Strain Stress
0 0
0.002 (0.2% ) 300 MPa
0.2 (20% ) 500 MPa

I had expected that the von Mises stress results would in any case be less than (or equal to) 500 MPa for the above definition which is not the case in my model.

On another public forum, I had posted the same query and got a response from a very advanced Femap user that this could be due to extrapolation of stresses from the integration points to the grids or element centroid. He suggested to look at gauss point stresses rather than the nodal values. Could anyone tell me steps to look at the Gauss point stresses in Femap?

Best regards,
Su

 

[rb1957]

if principal stresses have different signs, then von mises will be higher than the largest principal.

another day in paradise, or is paradise one day closer ?

 

[SuG89]

Hello!

Please refer to the attached test model I made for my understanding of the material behaviour. I expected the stresses to never exceed the values defined in the material stress-strain curve. But that's not the case here. Am I missing something basic here?

Thanks,
Su

 

[rb1957]

sorry, I'm not going to open a zip attachment.

are you doing a material non-linear FEA? This would (maybe) limit max principal to your Ftu; or maybe it'd limit von mises to your ftu (though that'd be much harder to calculate).



another day in paradise, or is paradise one day closer ?

 

[SuG89]

yes I am doing non-linear FEA.

Are you willing to look at it if I share the Nastran file with you (not zipped) on dropbox?

 

[rb1957]

not really !? how have you defined failure or, if you don't explicitly define it, how does the FEA determine failure ?

how non-linear is your material curve ?

have you tried an example problem ? have you asked the help desk (they can be very helpful, sometimes) ?

another day in paradise, or is paradise one day closer ?

 

[SuG89]

hi, yes I tried an example problem. Actually that's what I had shared in the zipped folder - a simple cube with loads, boundary conditions and non-linear material curve.
I didn't define failure explicitly. I thought the non-linear material curve definition would take care of this as after 500 MPa (ultimate stress) the stress vs strain curve is a horizontal line.

Stress-strain curve:
Stress [Mpa] Strain [%]
0 0
300 0,2
500 20
500 100

 

[rb1957]

to me an example problem is one with a known solution, ie a worked problem, so you can see what you expect to get.

I think it's important to understand how it's interpreting your material curve. Maybe it's saying no principal stress can exceed the given value, maybe you can tell it to use von mises.

another day in paradise, or is paradise one day closer ?

 

[SuG89]

ok, I will try to figure out how the software is interpreting the material curve. Thanks!