| Hi I'm trying to simulate a DMA test analysis but I've encountered a big issue. I want to know if it is possible to use Nonlinear viscoelasticity in an *steadystatedynamics,direct step or in a *Visco Step. If I use any of the mentioned steps with linear viscoelasticity (prony) I obtain similar results both in time and frequency domain. The material shows a viscoelastic behavior. But if I use a PRF model (law=strain) the material does not shows any viscoelastic phenomena. In other words using PRF the Loss modulus of the materials shows a value of zero (imaginary part of stress in SSD analysis = 0). In the file (please change the extension from .pdf to .zip) attached you will find the two approaches of analysis (*visco and *steadystate) and the input of the material properties. Thank you in advance |
The PRF model should work with *VISCO. I have used many different versions of the PRF model that way, and it does dissipate energy.
I don't recall if the PRF model works with *SteadyStateDynamics. I ran your test case using Abaqus 2019, and I get the warning:
***WARNING: VISCOELASTIC MATERIAL PROPERTIES ARE DEFINED IN THE MODEL. THE MATERIAL RESPONSE IN A STATIC ANALYSIS WILL BE BASED ON THE INSTANTANEOUS ELASTIC MATERIAL PROPERTIES.
Suggesting that it probably does not work (not surprising).
I also noticed that your material model contains an exceptionally many networks (20), and that the Sratio for about half the number of networks is basically 0. I would get rid of those networks.
-Jorgen
Thank you Jorgen for your response!
I tried with my material and others from the documentation and the results are the same.
If I use *visco step I don't have dissipation, stress and strain are in the same phase.
I've delete the networks with sr=0 as you suggested.
Attached are the inp files (change .pdf by .zip)