Modeling cure in viscoelastic materials

[sampath.vanimisetti]

Hi All,

Has anyone used Abaqus to model mechanical behavior during cure of (viscoelastic) thermosetting polymers? Time-cure-temperature superposition method has been developed and used by many researcher. However, implementation in Abaqus is not a very staightforward task. I am specifically refering to Eom et.al.'s work (PES, V40, N6, 2000, pp.1281), who have presented a phenomenological T-C-T model for thermosetting polymers. Any pointers would be greatly appreciated! Further, any expert comments on this method's usefulness in adhesive cure modeling are also welcome.

Thanks in advance!
(Sam)path V.

[Jorgen]

Hi (Sam)path V,

What journal is PES?

I am familiar with the problem that you are describing, but unfortunately I have not had time work on it much myself. From what I have seen, however, you should be able to write a suitable user-material subroutine (UMAT) that captures the T-C-T behavior. The challenge is to have a suitable material model, and then to implement the model as a UMAT subroutine.

- Jorgen

[sampath.vanimisetti]

Jorgen,

Thanks for your reply. Pardon my unscientific abbreviations. I was hinting at Polymer Engineering & Science when I wrote PES. Most thermosetting polymers (including adhesives) gradually evolve their viscoelastic properties in a cure cycle. There are many approaches to describe this phenomenon in a physics based framework (time-cure-temperature superposition being one). I have tried to incorporate TCT model (as described in my previous post) using in-built Abaqus viscoelastic model in conjunction with the UTRS user-routine. I keep track of the state of cure (%conversion) using a state variable. The evolution of cure is goverend by an Kamal's autocatalytic expression. Since cure has sequential dependence on temperature, I can readily compute the shift factors in a UTRS routine and pass it back to Abaqus. However, I have to feed the cure-shifted master curve to Abaqus. I have successfully implemented this in Abaqus.

Now coming to the reason why I posted this thread here. The next level of complexity is that the Prony series terms themselves are a function of cure, i.e., beyond a critical conversion, the viscoelastic parameters change. One cannot implement this in the above framework. The only way would be to write a viscoelastic UMAT with (3D large strain implementation) coupled with TCT shifts.

I want to know if someone has already tried something as audacious (or rather overkill!)? Where could I find a full numerical UMAT implmentation (Abaqus like) of viscoelasticity, so that I can make necessary change with least amount of effort!

I hope this gives you comprehesive overview of what I wish to achieve.

Thanks in advance,

Warm regards,
Sampath

[Jorgen]

Thanks for your very nice summary. That really sounds like an interesting but challenging problem.

About your question, I do not think that you will find a full implementation of the linear viscoelasticity model :(
If anyone knows of any full implementation please let us all know...

As you know, it takes significant effort to develop any serious UMAT, and the people who have done it (or know how do it), typically do not release their code for free.
My Parallel Network Model (http://polymerfem.com/index.php?pageid=umats) (PNM) contains linear viscoelasticity as a special case, but the PNM is only commercially available.

Best of luck,
Jorgen

[Jorgen]

I just want to add that if you have the time, then I think it would be a great exercise for you to implement linear viscoelastic UMAT from scratch.

Best of luck,
- Jorgen

[vamsi295]

Hi Sampath,

I was also currently trying to model the cure part of viscoelastic model but I was unable to figure out how we incorporate into abaqus.You said that you had already worked on the code.Can you please tell me how can we feed abaqus with cure shifted master curve.
Thanks in advance,
Vamsi krishna Kanchumarthy