How can I correctly model the damping behavior in ABAQUS for unidirectional composites to achieve realistic oscillations that decay over time?

[Alex a]

I am investigating the vibration properties of unidirectional composite materials using ABAQUS. My aim is to compare the damping behavior in both the longitudinal and transverse directions. However, my current results show unrealistic behavior: the oscillations decrease at first and then start increasing, which is non-physical.
I suspect that the issue lies in the setup of the boundary conditions, or step definition. The model was subjected to a 1% strain along its length, then released abruptly to observe free vibration.
What I aim to achieve:
1. Faster amplitude decay when damping is added.
2. Higher frequency and smaller amplitude in the longitudinal direction(parallel to the fiber).
3. Lower frequency and larger amplitude in the transverse direction.
The figures below describe the steps I followed in setting up the simulation. Any suggestions to resolve this issue?
Thank you in advance for your help.

 

[FEA way]

Which type of damping did you choose - how did you define it in Abaqus ?

 

[Alex a]

Which type of damping did you choose - how did you define it in Abaqus ?

I applied structural damping, but even after removing it from the material properties, the frequencies remained unchanged.

 

[SWComposites]

why are you modelling at the individual fiber level?? I doubt that such a microscale model will give any useful insight for macro scale structural response.

 

[Alex a]

Which type of damping did you choose - how did you define it in Abaqus ?

I applied structural damping f, but even after removing it from the material properties, the frequencies remained unchanged.

why are you modelling at the individual fiber level?? I doubt that such a microscale model will give any useful insight for macro scale structural response.

Modeling at the fiber level helps us understand the material's behavior more accurately, especially when microscale interactions affect the overall response. But even when I simplify the model to a cube of one material (e.g., epoxy), the oscillations are still not correct. Do you have any suggestions to resolve this?

 

[SWComposites]

sorry, i've never modelled damping in Abaqus specifically.

a guess is that increase in response is due to some sort of resonance.
or perhaps, since your output pt is on the matrix, that the fiber and matrix respond in different frequencies, leading to a complex displacement pattern.
have you run a simple static analysis of this model and review the displacements?
suggest outputting displacements vs time at different locations on your unit cell model.

Modeling at the fiber level helps us understand the material's behavior more accurately

well, that's highly debatable, and highly dependent on the definition of "accuracy". maybe the question should be, what are you going to do with the results of this unit cell model? (presumably this is some sort of research project and not a practical industry problem).