I am using Nastran Code. I have done two calculation cases of analysis (Elasticity and Plasticity) with the same load. This load remain in the material elastic domain (that is to say the loaded material should meet not plasticity for both cases). The model is also the same mesh and geometry.
I performed :
- the elasticity case is analysed in (1.Static)
- the plasticity case is analysed in (10.NL Static) with the curve function dependent (4. vs Stress)
I observe that the two cases result in different stresses states. I verified also my Stress-Strain curve function between the first point (origin point) and the second point (elastic limit point) is in the elastic domain with the same Young Modulus
My question is: why do I obtain different results for the two cases when I apply the same load in the elastic domain ?
Thank you fou your answer.
The answer to this will require investigating elements used, property settings, analysis control options, etc. I would suggest contacting GTAC and providing example models for them to look at.
It is also possible that your model is geometrically nonlinear, or did you properly rule that out, like you did material nonlinearity?
Linear solutions assume the small displacement assumption holds, which could introduce some inaccuracies to the linear solution, which the nonlinear solution could account for every time it recalculates the stiffness matrix.
Thank you for your anwser.
Please what do you mean by "Geometrically non linear model" ?
Is it notched sample model or a specimen with some stress concentration factor ?
Thank you for your explanation.
One of the fundamental differences between a nonlinear and linear solution is that a nonlinear solution can recalculate the stiffness matrix at increments, whereas the linear solution only does it once at the beginning. This makes the linear solution inaccurate as displacements become larger. For very small displacements, this is completely acceptable. But the larger you go, the more inaccurate the stiffness matrix becomes in representing the deformed shape. Please also look up membrane forces that comes into play at large deflections.
Take for example a cantilever beam loaded with a downward force at the one end. After a couple of degrees of deflection it becomes important to note the difference between whether the force was initially loaded in the negative Z-axis, or whether it was loaded perpendicular to the top face of the beam. The former will have the force remain directed in the same direction, regardless of deflection and the latter will have the force direction change with deflection of the beam. That is geometric nonlinearity.
Thank you very much j_coetzer
I have a tension loading case and I have some little difference in stresses between elasticity and elasto-plasticity calculations for a loading that is strictly under the yield stress. I understand now why I have this little difference even if I have not large deformation.