*shell section

[mambo5]

i am using S4R shell element modelling I-beam under hot temperature.

I had my *.inp file (1),

*Shell Section, elset=_PickedSet4, material=STEEL
15., 5
............
*Temperature
_PickedSet4, 500

then i try another *.inp file (2):

*Shell Section, elset=_PickedSet4, material=STEEL, temperature=5
15., 5
............
*Temperature
_PickedSet4, 500, 500, 500, 500, 500

then i try another *.inp file (3):

*Shell Section, elset=_PickedSet4, material=STEEL, temperature=3
15., 5
............
*Temperature
_PickedSet4, 500, 500, 500


ABAQUS analysed both *.inp (1) and (2), they both run ok, and gave very close result.
But ABAQUS stop half way during the analysis of *.inp (3), it gave the following error:

***NOTE: THE SOLUTION APPEARS TO BE DIVERGING. CONVERGENCE IS JUDGED UNLIKELY.
***ERROR: TOO MANY ATTEMPTS MADE FOR THIS INCREMENT

wondering which is the correct way to do thermal analysis for shell element? please advice.

 

[corus]

My guess is that as you have 5 section points through the shell thickness then you need to define 5 temperatures. In case 3 you have only defined 3 and therefore have a temperature gradient through the thickness of 500 degrees, whereas in cases 1 and 2 you have a zero temperature gradient.

corus

 

[mambo5]

thanks corus,

does it really matter to indicate the 'temperature=n', in the line of *shell section?

as i tried not to indicate in my *.inp file (1) (attached RECHEAT.inp), and put it in another *.inp file (2) (attached RECHEAT1.inp), where temperature=5. ABAQUS completed the analysis for case (1), but stop half way for case (2).
I notice that if I not indicate the 'temperature=n', ABAQUS would only considered NT11 and NT12, which i found from the *.dat file after I ran *.inp (1).

please advice.

 

[mambo5]

attached input file RECHEAT1.inp