Login Using Old Accounts All old passwords were reset when this site was migrated to WordPress. You can change your password by clicking on the "Login" item in the "Support" menu, and then selecting "Lost your password".

Notifications

Clear all

Apply constraints to material model's parameters in MCalibration

Last Post

RSS

Posts: 2

wldeng

Topic starter

Jun 16, 2021 6:07 pm

(@wldeng)

New Member

Joined: 4 months ago

Hi Jorgen,

I wonder if there is a way to apply inequality constraints to material model's parameters in MCalibration during data fitting. For example, considering the Mooney-Rivin model, the Drucker stability condition requires that

C10 + C01 > 0.

I looked up the forum and only found this page for applying Parameter Connections (equality constraints).

Thanks!

Topic Tags

MCalibration Constraint

2 Replies

1 Reply

wldeng

(@wldeng)

Joined: 4 months ago

New Member

Posts: 2

Jun 16, 2021 6:28 pm

Reply to

wldeng

Posted by: @wldeng

Hi Jorgen,

I wonder if there is a way to apply inequality constraints to material model's parameters in MCalibration during data fitting. For example, considering the Mooney-Rivin model, the Drucker stability condition requires that
C10 + C01 > 0.
I looked up the forum and only found this page for applying Parameter Connections (equality constraints).

Thanks!

I guess the way to apply the inequality constraints, such as C10 + C01 > 0, is

put '-C10' in the lower bound for parameter 'C01'???

Posts: 3924

Jorgen

Moderator

Jun 16, 2021 6:34 pm

(@jorgen)

Member

Joined: 2 years ago

There are 2 ways to do that. The first is to use an equation-based constrains on the parameters. In this case I used a "penalty" value of 10 and the step-function to impose the inequality constraint. This approach is similar to a Lagrange multiplier.

Attachment : Constraint1.png

Another approach is to set a fitness weight for the Drucker stability. That can be specified for each load case.

Attachment : Constraint2.png

/Jorgen