material parameters in druckerprager model?

[jayliu]

Dear experts,
Now I only have uniaxial tension and compression test data of PP, I found it is difficult to do FEA work using drucker-prager model in ABAQUS.
Do I have to do triaxial test, tension and compression?
Thanks

[Jorgen]

Hi Jayliu,

What kind of problems do get with the Drucker-Prager model? How did you determine the material parameters? Can you predict the unixial tenson and compression data that you used for calibration? What temperature are you trying to simulate?

It shouldn't be necessary to perform triaxial tests. What other models have you tried to use?

Thanks,
Jorgen

[jayliu]

Dear Dr. Jorgen,

What I am trying to do is indentation simulation for PP. and I only have uniaxial tension and compression test data in hand. As we know, the PP material has viscoelastic and rate-dependent plastic behaviors.
I select two models provided by ABAQUS for simulating. one is two-layer model which can simulate elastic, viscous, plastic behaviors, another is drucker-prager model which can give hydrosatic pressure's influences.
I do not know if there is a better model than these two models.

Problem for two-layer model is the parameter A and n in Norton-hoff rate law,because our uniaxial tension stress-strain curves do not appear hardening parts, they are rather flate in fact. the hardening modulous thus is 0. I wonder if it is ok.

Problems for linear drucker-prager model are: the parameters K( ratio of triaxial tension yield stress to triaxial compression yield stress, according to the ABAQUS documents), the friction angle belta, (if I use uniaxial tension data, I get a consant value of friction angle using the expressions in ABAQUS t=..., p=... ), and the dilation angle, (I do not know how to get it ). I think If I have triaxial data, then I can choose Exponent drucker-prager model , and abaqus will decide related parameters automatically.

I am sorry I am a newer in Polymer and ABAQUS. I have strong desire to get your directions.

Thanks!

Jay


What kind of problems do get with the Drucker-Prager model? How did you determine the material parameters? Can you predict the unixial tenson and compression data that you used for calibration? What temperature are you trying to simulate?

It shouldn't be necessary to perform triaxial tests. What other models have you tried to use?

Thanks,
Jorgen

[Jorgen]

Jay:

You are using two material models that are not commonly used to represent the behavior of polymers. As you probably know, neither the two-layer model or the drucker-prager model was developed to capture the large strain viscoplastic response of polypropylene. I am therefore somewhat hesitant to recommend these models. What you need to do, if you want to use either of these models, is to find the needed material parameters by calibrating to a certain set of experimental data (the calibration set). And then verify that the model predicts the appropriate response for another set of experimental tests (the validation data set). If you can pass these tests then feel free to use the models.

What I have seen in the past is that people sometimes use *Elastic with *Plastic to represent the behavior of polymers. This approach, however, is not very good since polymers have very different flow characteristics than metals. Hence, the standard plasticity approach often give vere poor predictions in multiaxial loading conditions.

The approach that I use and that I typically recommend is to use a material model specifically developed for polymers. Currently ABAQUS does not have any built-in models for thermoplastics that can capture the large deformation viscoplastic response of these materials. There are a number of advanced and accurate models, fortunately, that have been implemented as external usermaterial subroutines (UMATS/VUMATS). For example, see the user suborutines (http://www.polymerfem.com/modules.php?name=User_Subroutines) section on this website.

Jorgen

[jayliu]

Dr. Jorgen
Thanks for your suggestion. Actually I know these two models I used are not good for polymer. but it is hard for me to develop right UMAT in limited time.

I wonder if you have a right model for polypropylene(PP) or PP+PE(polyethylene). and if so, what is the price or the rent?

Thanks,
Jay

[Jorgen]

There are many different models that you potentially could use. Which model is best for you will depend on the specifics of your case.

I will send you a private message with pricing and license information for the UMATS that I have developed.

- Jorgen

[jayliu]

Dear Dr. Jorgen,
In fact I have been waiting for your private Email about the UMAT. At the moment, I think maybe you have not my e-mail address. for any further message, pls send to: jayliu@mail.tongji.edu.cn
Thank you very much!
Jay


There are many different models that you potentially could use. Which model is best for you will depend on the specifics of your case.

I will send you a private message with pricing and license information for the UMATS that I have developed.

- Jorgen

[swet_chandan]

Dear Dr. Jorgen,

I am using two layer viscoplastic model to find material parameters of
epoxy (Epon) using indentation. It turns out that this model does not give good result. Can you please tell me if you have UMAT for viscoelastic-plastic models?

Swet