Regarding plastic material % of elongation

31 views
Skip to first unread message

PRATHAMESH DEHADRAY

unread,
Aug 26, 2025, 12:08:52 AMAug 26
to LS-DYNA2
Hello all, 

I am working on simulation of plastic material which has % of elongation or elongation at break is 800 %. 

Which material card is appropriate to proceed as limited data is available for the study ? 

Kindly refer the image shown below. 

Is there any conversion factor to make 800% to 0-1 which can be input for MAT_024 "FAIL" option ? 

Thank you.
image.png

l...@schwer.net

unread,
Aug 26, 2025, 10:48:31 AMAug 26
to PRATHAMESH DEHADRAY, LS-DYNA2

I typically use the “elongation to failure” to provide an estimate of the tangent modulus

 

ETAN = (Tensile strength – Yield strength)/elongation

 

You could try setting FAIL = elongation, but this is only true for a quasi-static uniaxial stress states.                  --len

--
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-dyna2/51161314-0fbc-43b9-b8c1-46cc0e15dd06n%40googlegroups.com.

Message has been deleted

Mark Warner

unread,
Aug 29, 2025, 7:27:00 PMAug 29
to LS-DYNA2
The plastic manufactures like to tell you that the plastic can deform with a very high elongation (800%). You see this with simple uniaxial tensile tests of plastics. I have seen plastics fail at very low triaxial strains. If you look closer to the uniaxial data, the material transforms from a elastic/plastic material in to flowing (viscoelastic/plastic) material with high elongation using a simple uniaxial tensile test. They use uniaxial tension to make high strength fibers and ropes using pultrusion from plastic. (Bailing twine)

I typically assume plastic failure between 15% and 20% elongation in FEA models. This is about the extent of the elastic/plastic models and material capabilities. After that it needs to change to a viscoelastic/plastic model which is harder to model. Especially since the only testing is uniaxial stress/strain. This data is only one dimensional and most FEA  models are 3 dimensional. You are lucky to get any biaxial tensile data, let alone any triaxial data.

Plastic flows, creeps, and is strain rate and temperature dependent, plus many more other dependencies. Plastic is not elastic/plastic, which us metal mongers have studied for a good century and understand.

Cheers,


Mark Warner
Reply all
Reply to author
Forward
0 new messages