Material model validation in FEA 1

[feaplastic]

Hello Users,

I have a report which validates particular material model available in FEA by doing FEA tensile test against experimental tensile test data of the material.

FEA tensile test process:

Create specimen model - mesh - apply load - calculate displacement - calculate strain and stress using mathematical relationship equations which is inbuilt in FEA software(like for linear steel, stress = E*Strain)

My question: During this simulation, experimental stress - strain diagram values are given as input.

How these input is used in the above process. As per above process, we do not need these values as input to find out stress and strain.

Could someone outline the process in FEA during this simulation. Thanks in advance for all help.

 

[GBor]

Can you provide a link to the report? A stress strain curve would be used to calculate E in a non-linear material model.

 

[feaplastic]

Thanks GBor. Unfortunately, i cant provide this report. But, It is nothing unusual than steel material validation report except it is a polymer.

I overlooked E value. Thanks again for pointing out.

 

[mechengdude]

Sorry to answer your question with a question but:
1. Why would one do an FEA of a tensile test if you have test results unless the exercise is purely academic?
2. If the analysis is non-linear, doesn't the FEA code need to have the curve from an actual test to describe the non-linear behavior?

 

[feaplastic]

mechenddude,

1. to know whether in-built material model developed by software vendors is suitable to simulate your material or not.

2. You are right. We need this curve to find E value as mentioned by GBor.

 

[truckcab]

As far as I understand, you want to see if the material model implemented in your FE-code can simulate the behaviour of your material in a proper manner? I think it is a wise thought if you are not sure if it does. My point is that any material model isotropic linear elastic, Drucker-Prager etc. is our way to fit a model to the nature. The nature itself dont care, it works as it wants to... and always have.

Anyway;
if you have a tension test of your material that is displacement controlled and from which you can extract a stress strain curve, then you can get your answer by simulating the test by means of a displacement controlled analysis using your measured data. If your material model is ok then your computed stress-strain, force-displacement should be similar to your test.

An extreme example would be if you simulated the test using a linear isotropic material model only with E of your tested material as input. In such a case your computed stress-strain curve would incorrect since material yield is not taken into account in that material model. I.e. that model is not good for that application! You can of course do the same test for other material models elastic-ideal plastic etc. using your measured data as input and check if discrepancies between measured an computed data is ok. I do not know i this helps, but its the way I would do.

Live Long And Prosper