Simulation of load-unload test in Abaqus

[Faezeh]

Hello,

I am trying to simulate the load-unload test in Abaqus under the following loading schedule using my UMAT (which has already been validated).

As you see, the load-unload test is pure tensile with positive displacements; however as I modelled it in Abaqus the force-displacement curve as well as the stress-strain curves give me some negative forces (stresses), while theoretically all the stresses and strains should be positive. Here is the numerical load-displacement curve which is obtained from Abaqus:
View attachment 343408
This is the experimental result (Stress-strain):

As you see there shouldn't be any negative forces (stresses) in the model. I don't know what is the problem with the model. it should be noted that I tried the model which is in this YouTube video ()

 

[MMTUser1]

Hi,

I have encountered this issue in my cyclic loading tensile test simulation but only my stresses were negative. This usually occurs when you have defined a displacement boundary condition using amplitude to replicate the loading. Since we implement a defined displacement at a defined time step/increment, the stress tends to "overshoot" into the negative region to compensate for this definition as well as calculations done for the material by Abaqus.

I fixed it by increasing the resolution of the amplitude of loading i.e., more points in the amplitude (increasing increment from 0.001 to 0.0001 for example)

Also, I am unable to see your Abaqus result.

Hope this helps. Good luck.

 

[FEAnalyst]

Try with a single hex element first (like the examples in the Abaqus verification guide). You will find some tips (like gradually simplifying the model, checking incrementation and amplitude settings, making sure the results are properly read and so on) here: https://www.physicsforums.com/threads/simulation-of-cyclic-loading-in-tensile-test.1061252/