Necking analysis

[barney75]

Hi all,
I'd like to know if someone have ever tried to simulated FEM analyses with Abaqus far beyond elstic limit. I mean, in the analyses I'm trying to simulate the phenomenon of necking in particular my question is this: assuming a material stress-strain curve with a uniform elongation of 10% (this for what concern the engineering ss curve) and I give in input to Abaqus the corrispondent true ss curve that extends up to, let say, 30% (this, if I'm not wrong, should be still correct because in principle what is important is the maximum stress at UE in eng. curve) will the thickness of my component loaded with tensile force to reduces due to necking? (an easy examble should be a round bar loaded with tensile force as specimens used for material testing). Thanks for help me. Bye

 

[gurmeet2003]

barney,

I have thought of trying to model this effect in a uni-axial case, but have not done so far. I can tell you what I think might happen. I have been told that necking begins to occur, because dislocation come togather at some place. Therefore in order to simulate necking you will have to introduce some kind of defect in the sample. Otherwise a uniaxial test bar will yield uniformly through the section.

Gurmeet

 

[barney75]

hi gurmeet,
yes you are right, necking start as soon as dislocations accumulate excessively at the grains boundary due to the ongoing deformation process and as a consequence you can see that engineering ss curve reaches a maximum and then has a negative slope. In my opinion when you give as input a true ss curve beyond the maximum point on engin. curve you are just take into account this phenomenon. So I would expect that because of the fact that the 2 curve (true and engin.) are corrispondent, FEM software, in general, even if it works with true ss curve in some way can take into account the fact that if during loading in some location of my component the maximum of the engin. curve is reached, there a necking of the thickness or at least a plastic flow should start to take place. Is this right?

 

[gurmeet2003]

Barney,
Engineering stress strain curve is based on a constant refernce area (intial undeformed area). In actual tension test the area begins to change after yielding and changes drastically once the necking starts. As you indicated above necking is a type of instablity cuaused by some defects.
Once necking begins engineering stress strain curve drops due to drastic reduciton in area after necking. However true stress strain curve does not drop because it is beased on the actual area of sample and continues rising.

The droop in engineering curve does not reflect plastic flow, but occurs mostly because of necking. If plastic flow was to occur without necking, the true stress strain curve will still remain the same. However engineering stress strain curve will change drastically.

Gurmeet