Stress-Strain curve into FEA

[mmar2087]

Hi.
Perhaps a stupid question, but here I go....

How do I input stress/strain data from a tensile test into an elasto-plastic material model in FEA?
The program states that the starting point should be the yield point.

Attached You'll find the results and graphs from 2 tests bars.

Result on testbar 215:
Rp0,2: 182 MPa
Rm: 229,1 MPa
At: 0,927

The data file, and plotted curve, shows 120 MPa at 0,2% strain.
So in the results file, the initial slope (E modulus) is parallel moved to 0,2% to hit the curve at 182 MPa.

How do I handle this in my FEA program material model to capture the elasto plastic behaviour?
Should i modify all the strain values so 182 MPa corresponds to 0,2%, instead of about 0,5% for the moment or should my start point be 182/0,5?

Thank you in advance for your help.

http://files.engineering.com/getfil...-ed5e-49fb-95c9-12364816945f&file=Results.PNG

http://files.engineering.com/getfile.aspx?folder=4e2d959c-9ce4-437a-bf3e-61f58ef1b8e0&file=Graph.PNG

 

[rb1957]

piece-wise linear ... you should be able to define the stress/strain curve from a bunch of points.

personally i'd be worried by the scatter you're seeing ... maybe use the minimum curve (215?) ... i'm expecting a smaller Ftu is more critical (than a higher Fty).

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

 

[mmar2087]

rb1957:
You're right, I have all information that I need for the SS curve, but I don't understand how to use it correctly in FEA.

The first point in my material file should be "yield" point, so 0,002 for strain and 182 MPa for stress according to result from the tensile test for sample 215.
But in the data table (and graph) for 215, 0,002 strain corresponds to 120 MPa so if I import these values as they are, it will assume a much lower "yield" point.

The "scatter" comes from that sample 216 is heat treated and 215 isn't and the question is more of principle type and not specific for these SS curves.

 

[rb1957]

you should be able to define your material. my FEA (FeMap) allows several non-linear material input definitions.

which FEA are you using ?

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

 

[rb1957]

reading your post again, you say "The program states that the starting point should be the yield point." ... ok, the program is assuming linear behaviour between zero and your first point, so your first point can be where the material departs of linearity. if i was doing a bi-linear material curve, I'd probably extend the linear portion a little to better match the plastic portion, for both materials something like (0.2, 140) looks reasonable.


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

 

[petb]

Also, find out if your solver expects true stress-strain relationship or engineering stress-strain relationship. Some solver expects you to supply plastic strain vs. This document Link shows some methods to "tune" your curve, it is a bit ANSYS-focused, but the general method applies for tuning the curve applies for any situation.

hth

petb

 

[mmar2087]

Thanks for your answers!

rb1957:
I made a mistake in my thought, where I considered the reported Rp0,2 as the yield point (and start point in my FEA material model)
I importing the SS curve as it was, removing all values below 120MPa that is "linear", and then added load up to 180MPa, the strain after unload was 0,2% as it should.

petb:
Thanks for the heads up. For small strains the program required true stress and engineering strain. Your link was very helpful to convert the stress.