Note : This post describes a new feature that becomes available in the material calibration app in R2024x FD01, on Feb 10, 2024.

Back in 2019 I showed a calibration using some test data for butyl rubber.  The test data was a family of rate curves.  Here is the post (you would need to then scroll down):

Calibration App in the 3DExperience Platform - 2019x FP1923

There are a variety of reasons that you might want to establish a relationship between two or more material model parameters.  These relationships may be equality relationships, or inequality relationships.  A few examples:

1) In a Prony series, the pairs are often thought of as an "ordered set".  From decades of experience with stress relaxation test data, there is a common understanding that as the tau parameters increase, the g parameters will decrease.  This is just the nature of commonly observed stress relaxation test data.

2) In metal plasticity, for isotropic tabular hardening, the hardening terms are commonly thought of as an ordered set.  Abaqus requires that the peeq values start at zero and monotonically increase.  It is commonly thought that the corresponding yield_stress values will monotonically increase as well.

3)  There are some aspects of material models that contain exponents, and as one makes these a tabular function of temperature, the exponents ought to be held constant.  An example is *Rate Dependent :  here the exponent "n" at each temperature is best constrained to be the same value. 
 

In the R2024x FD01 release (Feb, 2024) we have added a way to construct these types of relationships between parameters.  This functionality is not available in Analytical mode.  Where this functionality is available, you will see this icon:

As a first release, we have tried to put some reasonable limitations in place to avoid optimization failures.  One example is that only parameters of the same unit may be related.  Another example is that the same parameter cannot be used in multiple relationships.  For this butyl rubber example, our starting point is the ending point of an earlier calibration:
 

The above calibration was performed in analytical mode, then we changed to numerical mode in order to continue the work of adding some parameter relationships.  This state was saved as "Butyl_Rubber_Start.3dxml" (see attached zip file).  In general, this represents a good calibration result.  For the 5 test datasets, there is an R2 error norm of 0.993 or better.  However, as we look at the Prony "g" terms (the "taus" start at  1e-3 and end at 10), they do not monotonically decrease.
 

While we do not have any stress relaxation test data for this butyl rubber, we can use the RO capability to see what the stress relaxation behavior would look like for this hyper-Prony material model. In this plot window, I have shown two stress relaxation loadings, I want to focus on the red line.
 

Here is a short video showing the creation of the relationship :  g1>g2> g3> g4>g5
 

(insert link to 3DX docs after Feb 10)

About Material Parameter Relationships

This post contains the butyl rubber test data:

This attached zip file contains the file Butyl_Rubber_Start.3dxml