ABAQUS EXPLICIT DYNAMIC

[StefBene]

Hello,
I'm running a simulation of a bending test of a particular beam, I have a problem with the analysis time. The analysis used is the explicit dynamic, the test is simulated in about 5 seconds otherwise it would take too long. The kinetic energy is less than the internal one so there should be no relevant dynamic phenonemes, the problem is that the various outputs show some fluctuations in the trend. There is a method to eliminate these fluctuations that does not lead to an increase in analysis times which are already long. I put the output load - vertical displacement so you can see the problem.
thank you so much

 

[FEA way]

You can use antialiasing filters for history output in Explicit. They can be applied either before the analysis or in postprocessing.

 

[StefBene]

OK thank you very much. I have another question please. In my model there are connectors to which a displacement load law is defined. The analysis consists of various steps and in the displacement output or force output of the connector between the two steps there is a spike/jump. What can this be due to?
Plase see the attachment for understand better

Thank

 

[FEA way]

It may be some numerical noise too. They are common in Explicit analyses. You can try applying filter to it as well.

 

[StefBene]

ok, I'll try soon.
thanks

 

[Mustaine3]

Maybe the issue disappears when you run the analysis in double precision.

Otherwise you should check the definition of your displacement. Be aware that loads and non-zero BC change immediately in dynamic steps. It is recommended to use a total time amplitude to have full control over the value at all time.