Search
Close this search box.
Notifications
Clear all

Transient analysis - Help

7 Posts
2 Users
0 Likes
433 Views
Posts: 124
Topic starter
(@bw_composite)
Estimable Member
Joined: 17 years ago

Good day all,

I have a query regarding a transient viscoelastic analysis.

I have modelled a component using Reinf 265 elements in ANSYS. The base material is linear isotropic as well as the fibre material.

So far, everything was good.

I would like to introduce now viscoelastic behaviour in the fibres, which I did with a command as follows

tb,prony,1,,1,shear !define viscosity parameters (shear)

tbdata,1,0.1,2.0

I am struggling now in achieving convergence, the analysis is stopped due to an error in element formulation (after a few warnings on elements becoming excessivley distorted).

Fair enough, what puzzles me is that if I modify the shear relaxation properties as

tbdata,1,0.0001,2.0

de facto making any relaxation negligible, still the same warnings and convergence problem remain.

Could anyone help me on this one? Any tip or advice would be the most precious.

Why should not the model converge as in the elastic case, when visocsity is set to such a small value?

Thank you

Kindest Regards

Muzialis

Topic Tags
6 Replies
Posts: 3998
(@jorgen)
Member
Joined: 5 years ago

Hello Muzialis,

Good question. Can you remind me what the different values in the dbdata, ... lines stands for?

How large strains can you reach?

-Jorgen

Topic Tags
6 Replies
Posts: 124
Topic starter
(@bw_composite)
Estimable Member
Joined: 17 years ago

Hi Jorgen,

thanks for the reply.

the tbdata command in question determines:

first value: how much, in percentage, is lost (shear modulus wise) in the relaxation time

second value: relaxation time

For this model I am just trying to achive convergence so I am using only one pair of values.

Strains in the region of 30% are achieved.

All the best

Muzialis

Reply
Posts: 3998
(@jorgen)
Member
Joined: 5 years ago

OK, I see. I agree that you should not get any problem using that little relaxation.

I can think of different possible reasons your simulation does not work as intended:
1) Do you use too large time increments?
2) Do you use an accurate enough viscoelastic integration procedure (e.g. Abaqus allows you to specify the max creep strain / increment. I am not sure what tolerances you can specify in Ansys)?
3) Is there any other complicating factor such as high pressure?

-Jorgen

Reply
Posts: 124
Topic starter
(@bw_composite)
Estimable Member
Joined: 17 years ago

Jorgen,

thanks for your reply.

I have done some trials tuning the time steps, with little luck.

Regretfully I am not wawre of options to improve the integration as the one you mention for Abaqus, I am going through th manuals at the moment.

But as far as your last question is concerned, yes there is presure applied to the model.

does this explain anything to you?

Thank you

All the Best

Muzialis

Reply
Posts: 3998
(@jorgen)
Member
Joined: 5 years ago

Hmm, I am not sure what is causing your problem 🙁

Are you using an almost incompressible material model?

-Jorgen

Reply
Posts: 124
Topic starter
(@bw_composite)
Estimable Member
Joined: 17 years ago

Hi Jorgen,

just to wrap this one up (an alternative solution was found), I wanted to answer your query about incompressibility.

The base element material was linear isotropic, Poisson ration of 0.41. I am not sure if this falls in the almost imcompressible range.

Thank you for the support anyhow

Muzialis

Reply
Share: