ABAQUS convergence problems

[dog2k5]

Hello all:

I am running a cyclic displacement loading on a steel I-beam. I put the loading protocol in the amplitude manager. I have displacements of -.275, .275, -.5, .5 ,-.156, .156 etc (the displacements are all in inches). I run the analysis. when I plot the displacement vs. the time step history my displacements never reach the peak displacements I have input. For example it only makes it to about .47 inches. why won't ABAQUS converge to what I tell it to? I have a friend running the exact same model and his works really well. Thanks for the help.

Tyler

 

[rstupplebeen]

Tyler,

Do you get any error messages? You said that your friend runs it fine. This might be due to memory resources or a setting in the environment file. With more information we may be more help. Hope this helps.

Rob

 

[xerf]

It is not very clear what exactly the problem you encounter is, but:

A. If you get an error due to solution non-convergence, try to post the error message. If you displacements are large , set NLGEOM=ON. Try to reduce the time increment.


B. If your solution converged but your problem is that ABAQUS does not apply the displacements according to the data you input in the amplitude manager then...

If you put the entire loading path (by using an amplitude curve) in a single step and use automatic incrementation you might not be able to get the values at the peak displacements.

Try to:
1. partition the loading path in a series of step such that the peak displacement is at the end of a step.
2. use fixed time incrementation (if the problem is not very non-linear )