Hi

When you extract the stress-strain graph from the experimental test, the machine calculates this equation:

Stress=Force/sample area

Strain= displacement/initial length

Moreover, these calculations are based on the whole sample. So, yes, the machine gives the average of all the sample for the stress-strain graph.

And about Abaqus simulation, you can not valid your experimental tests with Abqus, because you must put some data into this software that was extracted from your experimental text. It is a loop, you will obtain a graph that you gave it to the Abaqus. In the property modules, you must put elastic and plastic property that you take them from your tests, then Abaqus will show your data exactly.

@ Meysam

First of all, thanks for your answer. Actually, we see in some journal papers that they somehow validate their experimental results (stress-strain graph) with the Abaqus or any other commercial software package. As you said, it would not be accurate but it would be precise enough (matched well) that we can validate it. This is why I asked how they do that? I am somewhat sure that they do not put any data from experimental results since this might deviate from the validation essence.

@ sebastian

Thank you. However, I think it makes it complicated. I looked somewhere they used stress-strain curve inside abaqus. i am not sure which element I should choose or is there any difference? Also, which load I should input in my model? For example, I test 2mm/min tensile loading and I have the stress strain plot and also force in each point. Which force should I give to my model?

Thanks

You should do what you do in the experiment: Calculate the displacmement between two points to get the strain (use the same points as in your experiment) and the total force to get at the stress (divide in the same way as you do in experiment).