I use the keyword *PARAMETER quite often in material model calibration work, and I also use .psf scripts to drive parametric studies. I used to use these capabilities much more before I learned how to do similar things in Isight. But they are still valuable to know about.
The Parametric Study capability is described in the Abaqus Analysis User's Guide in section 20.1.1 Scripting parametric studies. If you learn by doing, you might be interested in the zip file attached to this post. The zip file contains just 4 files:
_README.txt describes a bit about the files and what they are used for
relax.psf This is the parametric study file
unit_cube.inp This is a small "parameterized" unit cube input file. Set up to run a simple tension stress relaxation analysis.
Relaxation_PPC3TF2_2010-09-21_STSR.xlsx This Excel file holds some test data for 4 tests and the matching analysis results.
Some common abbreviations I tend to use:
ST = Simple Tension
STSR = Simple Tension Stress Relaxation
PSF = Parametric Study File
If you are interested, you can download the zip file, read the README file, and pretty much see what is happening. Explore the Excel file a bit. In this example, I just needed to run 4 simulations, 4 stress relaxation simulations at 4 separate strain levels. But I was running them over and over, each time changing my material parameters. So the PSF file just automates running all 4 runs and extracting the history results and writing them to the Excel file. The PSF file also allows you to make some post-processing calculations if needed. The use of this PSF file automated a lot of work and let me focus on the material modeling and calibration task.
You will see in this example that I use an *AMPLITUDE definition for applying the strain history. The history is pretty simple, so I've idealized it to linear ramp and constant hold. I've included in the Excel file a comparison of the idealized *AMPLITUDE definition and the real test history of the applied strain. In other cases, with more complicated load histories, use a *AMPLITUDE and just use the actual test data to drive the unit cube. Of course, this use of unit cubes assumes that the stress/strain state in the test is homogeneous within the strain measuring region (for instance, not necking). Another nice feature, not shown in this example, is using the *TIME POINTS feature in Abaqus to output the results only at points that match up with the experiment. This way you can output to Excel and even do error norm calculations between test & simulation responses at matching points in time.
Back to: Sharing Abaqus test specimen models
Back to: Material Modeling and Calibration - An Overview and Curriculum
