Plastic strain vs Stress curve for multi linear isotropic hardening

[IRUDAYA JEYAPRAKASH]

I am using PP material for my projects. In stress strain curve of PP, Yield strength value 50 MPa and UTS value 20 MPa. Stress strain curve is downward after Yield value.

in this case How to load the plastic strain vs Stress curve for multi linear isotropic hardening?

 

[NRP99]

I guess this is engineering stress-strain curve. What you need is true stress-true plastic strain curve for the input. Convert the engineering stress and strain to true stress and strain. Engineering stress strain curve is not appropriate for FEA analysis as it is not true measure of material behavior post yield.

Refer following links-
Link-1
Link-2

 

[IRUDAYA JEYAPRAKASH]

I have created True stress vs True strain curve. It is also downward after yield strength.

 

[rb1957]

if UTS is 20 MPa, and YTS is 50 MPa, then nonlinear behaviour would not typically matter … the material would behave in a brittle sense … elastic to yield then rapid onset failure.

another day in paradise, or is paradise one day closer ?

 

[DanStro]

Sorry if this is a dumb question but is it possible for the ultimate stress to be lower than the yield? I understand that if the material is brittle may not yield (or yield very little) before it breaks but I can't see how it would yield after it breaks. Am I missing something? Are those values the same stress type (i.e. tensile, compressive...)?

 

[realtime2]

Probably your "ultimate" stress is the same as "Yield" stress.
These plastics often neck down a lot and the neck propagates along
the length of the gauge section and has a completely different
cross sectional area after yield/ultimate. Was the cross-section area
measured after yield? or is the "true" stress based on S *(1+engStrain) ?

 

[rickfischer51]

PP can draw out forever, elongation values over 500 for unfilled homopolymer are not uncommon. First your true stress/strain curve does not look right to me. I think the slope should be continuous. But the point to consider is what is the loading and how much strain is expected. No point putting in a curve good to 200% if you never go beyond 30% The generic curve from your first plot looks like some kind of hyperelastic/Mooney curve. You will never get the entire range of that with a MISO material model. Also, I think the multilinear curve cannot have a negative slope. So if you're limited to MISO, then you will have to deviate from your curve somewhat. What I do is first determine a reasonable maximum strain. I do this with a bilinear curve with a zero( or at least very low) secant modulus and yield at the approximate average of the stress above 10% strain. This should give a reasonable estimate of maximum strain. Then build your MISO model up to the strain level. Try and keep the area under the MISO curve close the the area under your test data curve. Remember that the area under the stress strain curve is proportional to strain energy density. I think that if the two curves have similar area, the you will get a fairly reasonable estimate of yield. Hope this makes sense.

Rick Fischer
Principal Engineer
Argonne National Laboratory