Equivalent plastic strain vs tensor plastic strain

Alberto Campagna

This document wants to clarify some doubts regarding the different plastic strain outputs created by OptiStruct in a non-linear analysis and how they can be correlated to stresses. Some math computation is presented together with a simple test model subjected to plastic deformation and recovery.

LouisS

Hello Alberto,

Thank you for this document ! After reading it, I still have trouble understanding the physical meaning of equivalent plastic strain, and its difference with tensor plastic strain and "element strain". I have read on Altair help pages that Equivalent Plastic Strain "takes into account the z-component of strain tensor for shells. On the other hand, strain tensor output (even with plastic strain tensor) does not output z-component of strain tensor" (source : STRAIN, comment 17). However, I am surprised at how much equivalent plastic strain can be greater than element strain, as in my mind, it is supposed to represent the state of the material's plastic strain : in some of my simulations going up to burst, the equivalent plastic strain was more than twice the value of element strain ! Can taking into account the z-component have so much effect that a plastic strain indicator is that much greater than what I understand to be VM total strain ?

Also, I was wondering which one of these three (element strain, element plastic strain tensor and equivalent plastic strain) has to be monitored to predict burst in relation with the A% of the material.

Thanks for your help !

Alberto Campagna

Hi,

equivalent plastic strain is a scalar quantity that is calculated by the integration over time of the strain rate tensor and gives an idea of the model plasticization. The maximum value is reported, for example, in the last column of the .out file during non-linear iterations. A warning is also printed if you are running a small displacement non-linear analysis that shows equivalent plastic strain exceeding the small displacement theory.

Element plastic strain is the plastic part of the total element strain and it is a tensor. Total strain can be also separated into deviatoric (that is responsible for the plastic deformation of the material) and hydrostatic that will either shrink or expand the volume uniformly.

Then you can select the criterion. For ductile materials you can use VM that is a maximum distortional energy criterion. For brittle materials you can choose a max principal criterion.

For plane stress in 2D shells eps_zz is output as 0. This should not be theoretically correct, but it seems to be done to align to other solvers convention.

For more accuracy I suggest to request stress and strain at gauss integration points.

Hope this helps.

Alberto