failure for TNV and Three-Networks models from the PolyUMod library

[Valery Anpilov]

Dear colleagues,

I'm solving problems involving the deformation of parts made of hyperelastic materials with viscous properties.

I'm using TNV and Three-Networks models from the PolyUMod  library.

Has anyone used these models to simulate the failure of parts made of these materials?

Regards Valery

[James Kennedy]

Dear Valery,

 

Perhaps these may be of some help:

How to Use the PolyUMod Library with Ansys LS-DYNA in Workbench

 

https://www.youtube.com/watch?v=D5stZQ6gjGo

 

Calibrate a Temperature-Dependent TNV Model using MCalibration

 

https://www.youtube.com/watch?v=8FknxM2p_nA

 

Sincerely,

James M. Kennedy

KBS2 Inc.

December 16, 2025

--
You received this message because you are subscribed to the Google Groups "LS-DYNA2" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ls-dyna2+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ls-dyna2/c3f527ec-2ec5-48fd-a830-50c7473d6784n%40googlegroups.com.

[Valery Anpilov]

Dear  James,

Thank you for your letter.

I've seen the videos you mentioned, as well as all the other videos by Dr. Jorgen Bergstrom.

The most helpful video for my question is this one:
https://www.youtube.com/watch?v=Qd1Pvqh20vE 

For the last nine years, I've been trying to use the models from the Polyumod library.

The creator of this library is Dr. Jorgen Bergstrom - a very competent engineer, or rather scientist.

While this company was independent, the users of Polyumod library received highly qualified assistance,
and the respected Jorgen Bergstrom published many very useful videos and blogs.

From 2022 Jorgen Bergstrom stopped publishing his very useful videos and blogs, as this company was acquired by ANSYS.

[Valery Anpilov]

I seem to have solved a simple test problem—uniaxial tension of a flat specimen with failure—using the aforementioned method.

 

*** I'm using LS-Dyna and Polyumod 2022.

 

I'm seeing some issues:

 

1. When using the TNV model, LS-Dyna doesn't record, or records incorrectly, deformations and displacements for the entire model.

That is, zero values ​​are returned if I try to plot graphs and then write the values ​​to a file.

 

2. Visualization issues:

 

- Visualization of the deformed model is possible before resolution, but if I try to display deformations or displacements for a single point, all values ​​will be zero.

 

- After failure begins, visualization of the ENTIRE model is impossible.

 

Despite the lack of deformation and displacement results in the database, the calculation itself—up until failure—results in a realistic model (like a full-scale experiment and the results of calculating and calibrating the constants).

 

*** I have some assumptions/hypotheses:

 

When solving problems, LS-Dyna incorrectly records the values ​​of some variables for this model in the database or records them in a way that the postprocessor "doesn't see."

 

The stress graphs constructed from LS-Dyna calculations are very similar to reality.

 

I solved the same model using an abacus—the calculation results are very similar.

when the destruction of elements begins, the integration time step decreases by 15 orders of magnitude

[Valery Anpilov]

on the forum
https://innovationspace.ansys.com/forum/forums/topic/failure-tnv-and-three-networks-models-from-the-polyumod-library/#post-449080

I was advised:
By default, all models from the PolyUMod Library, including TNM and TNV, do not provide failure modeling.

To enable failure capture, MAT_ADD_GISSMO or MAT_ADD_EROSION must be used in conjunction with these models.

But I haven’t used MAT_ADD_GISSMO or MAT_ADD_EROSION for TVM model before.

I would be very grateful if you could recommend blogs and/or videos on how to use MAT_ADD_GISSMO or MAT_ADD_EROSION specifically for TVM model.

I would be happy to confirm that the mentioned MAT_ADD_GISSMO or MAT_ADD_EROSION functions correctly when using the TNV model.

[Valery Anpilov]

Quick Calibration of a PolyUMod Failure Model to Experimental Data
https://www.youtube.com/watch?v=Qd1Pvqh20vE

[Valery Anpilov]

I've run several test simulations using MAT_ADD_EROSION (the SIGP1 Maximum Principal Stress at Failure failure criterion).

The crack formation locations and shape are very similar to the results of the full-scale experiments.

However, the subsequent crack propagation and failure of the part are completely different from the full-scale experiment.


I would be grateful for your advice and comments.