Several years ago we added the ability to generate a Master Curve from DMA (storage & loss) test data. I had some real DMA test data for a PVC material and created a DMA-based Master Curve post here: Generate Master Curve from DMA test data
That exercise created the following Abaqus material model for PVC:
Using that Abaqus material model, I ran a number of separate Abaqus unit-cube analyses to generate some synthetic test data for Stress Relaxation (each run at a different temperature). That synthetic test data is shown in the image above. In the image's legend the abbreviation STSR stands for Simple Tension Stress Relaxation and the values after that are the temperature in degrees C.
In general, in the Material Calibration App, we would like to operate on raw test data, that is time-domain test data that starts from the < 0, 0, 0 > time, strain, stress point. We would prefer NOT to create special forms of "processed" data. In the Stress Relaxation and Creep engineering domains, it is somewhat common to see people talk about (and plot) the "Creep Modulus" and the "Stress Relaxation Modulus". when they do this they also typically delete the loading phase of the raw test data. In our implementation, we expect/require the raw test data. The new feature that creates the Master Curve deletes the loading phase data prior to stitching together all the individual datasets into a Master Curve.
For R2026x FD01 we will release an enhancement to the existing "Generate Master Curve" feature to be able to create a master curve from either Stress Relaxation test data, or from Creep test data. This post speaks to just the stress relaxation part of things.
The video below uses a development version of the 3DExperience platform to demonstrate this new feature. The final released version may look a little bit different. This video was created on Oct 30, 2025.
After performing that calibration, I could double check things a bit by importing the material model that was used to generate all the synthetic data. The image below shows that the result is a near perfect match-up of the responses to the test data.
The zip file below contains the files used in the video.
