Modeling Plasticity - Tabulating Plastic Stress / Strain from Empirical Data

[TheHumbleStudent]

Attached is a true stress / true strain plot of polycarbonate at some elevated temperature. The true stress / strain was calculated using the engineering stress / strain data and the measured Poisson's ratio.

I'm curious how members of this forum would tabulate plastic stress / strain data points for use in structural FEA? I know for a fact that my FEA program does not allow "negative" slopes when tabulating plastic strain / plastic stress. The yield stress according to manufacturer is around 7 ksi.

So, specifically:
1.) Where would you assume yield (and why), and
2.) How would you tabulate plastic stress / strain in a material model, knowing that stress must always stay the same or increase as strain is increased (and why)?

I'm not looking to make conservative (or non-conservative) assumptions - just the closest possible model to fit the data.

Thanks in advance!

 

[AeroStructAnalyst]

I think I would like to convince myself that my stress-strain data obtained via test follow the trend expected. So my question is, does your stress strain data as obtained represent the stress-strain data obtained elsewhere?

 

[rb1957]

does poisson's ratio account for necking (as seen in compact tension specimens) ?

the behaviour at yield does look odd ... no?

Quando Omni Flunkus Moritati

 

[TheHumbleStudent]

I've looked into this and the data does in fact follow trends I've found elsewhere. The best explanation I've read is that, under load, polymer chains stretch, rotate, slide, and disentangle to cause permanent deformation. Initially the chains may be highly tangled and intertwined but when the stress is sufficiently high, the chains begin to untangle and straighten (orienting themselves parallel to the direction of loading). Eventually the chains become almost parallel and close together, and stronger Van der Waals bonding between the more closely aligned chains require higher stress to complete the deformation and fracture process. This is from a textbook sitting on my desk "The Science and Engineering of Materials" by Donald Askeland...

Another explanation:

http://alexandria.tue.nl/repository/books/653970.pdf

 

[SWComposites]

So was this a load controlled or displacement controlled test?

 

[TheHumbleStudent]

Dispacement

 

[SWComposites]

The drop in stress is an artifice of the displacement control, since when the material yields the stiffness is reduced, and the stress appears to drop because the displacement (strain) is increasing at a fixed rate. In a load controlled test, and in most real loading conditions, there will not be a load drop but a large increase in strain after yield. Suggest "correcting" your stress strain curve by replacing the "dip" in the curve with a horizontal line from the point of maximum stress before the "dip" over to continuation of the stress-strain curve.

 

[TheHumbleStudent]

I appreciate the responses. If the test was load controlled, the strain rate would accelerate when the stiffness begins to decrease in order to maintain or increase load(?). If so, it becomes a matter of different strain rates exhibiting different stiffnesses. Someone please confirm my understanding of SWComposite's response.