Learning Forum

[zjuv9021]

Hi all,

I'm bending and compressing a hyperelastic tubing in ANSYS, and have available multiple strain % experimental data available (e.g. 15% strain, 68% strain, 89 % strain in uniaxial, biaxial, planar shear stress).

What is the best way to determine what strain % data set to use? For example, I want to rotate 1 cm of tubing at most 60 degrees and then push/compress it towards a fixed support.

 

I've seen at MOST ~0.05 in/in or ~5% strain, but am using the higher strain rate experimental data currently.

How do I assess, given the physical boundaries and forces/displacements I would like to apply, what strain % experimental data to use to get the most accurate representation of the real world physical problem?

If I'm using the high strain % experimental data, and only seeing ~5% strain occurring with my tubing, am I losing some material characteristics at these lower strains or are my strains low BECAUSE of me using a high % strain experimental data?

Any insight on best practices here are greatly appreciated.

Regards,

Zach

[Sandeep Medikonda]

Hi Zach,

  It is very rare that people have experimental data from all three tests. So you are in a good place to begin with.

  If you are observing a certain strain range, it would help to have experimental strain data capture that range as you can control the fits and use an appropriate strain energy density/material model.

  The last part of the question is not quite clear to me. Do you not have enough data points in the low-strain region? I don't expect this to be a problem if you are observing a good fit to the experimental data.

  If you can fit your experimental data in Engineering Data first and share images of the curve fits, you can probably get more insights/recommendations here?

Regards,
Sandeep

[peteroznewman]

Hi Zach,

Since you have data at multiple values of strain, you put them in a table of Experimental Uniaxial or Biaxial or Shear Test Data, then let ANSYS curve fit to the data you have in one of the Hyperelastic material models.

Here is an example of six data points in one table.

If you have lots of points, you might choose to limit the strains to 2X the range you expect in order to improve the fit in the low end.

Regards,
Peter