Unexpected faulty results after solution convergence

[abc123ali]

Respected ABAQUS Users, I am simulating cohesive zone modeling of with the input files attached here with one as a smooth surface and the other as a rough grooved surface. The output reaction force should be greater for the rough surface due to large area of contact. The model is developed using contact traction seperation rather than cohesive element deletion. Why it is happening? The file are attached here.
Kindly assist
Thanks for cooperation

 

[abc123ali]

the second file is here

 

[FEA way]

How do you measure the reaction force ? You don't have any history outputs requested (apart from IRF/IRM used for inertia relief load so why is it here ?). Do you derive it from field output ?

You should also request and compare outputs related to contact.

 

[abc123ali]

I am measuring the reaction force at the reference point created where the load is applied. I think so it should work well as I measure the output for both smooth and grooved surfaces at the same point requesting filed output in the step module. Kindly assist.
Thanks

 

[FEA way]

And what are the final values ? When I check it, it’s 530 N for the grooved case and -210 N for the flat case. Are you comparing proper components ? The models are oriented differently.

 

[abc123ali]

For that I have attached the output graphs for both of cases. The value for smooth peaks in negative direction just only because of the loading direction while for the rough surface is positive but the value of smooth is greater than rough as it is 2KN while for rough it just peaks at 1.2KN at the reference point. The graphical results are attached here. Also in both cases I applied displacement of 1mm but for the groove rough case it just goes to 0.045mm! why?

Thanks for cooperation

 

[FEA way]

So those are different models than what you shared before or they were modified ?

The plots have very different X axes - one (smooth surface) is from 0 to 3 s and the other (rough surface) is from 0 to around 0.14 s. So either the range was modified or the results are incomplete (maybe because the analysis wasn't finished or the .odb file wasn't reloaded and is not up to date).

 

[abc123ali]

Even if it is not complete for the groove surface for time upto 0.14s even than the output is stable and not increasing and tending to decrease at peak of 1.2KN. The files are not modified and are same. Why this results shows less output reaction for groove case? The graph is almost bilinear and hence even if the result is not complete for groove case it is simply expecetd the peak at 1.2KN because if it increase further it will show trilinear or bend upwards which is also not expected so this may surely be expected as a peak.

Thanks for cooperation