Plastic materials and FEM simulation

[dragonix]

Hi Guys!

I'm working with plastic materials (such as Polyamide or PET) reinforced with glass fibers (from 30% to 50%) and I observed a huge difference between the material properties specified in the datasheets and the material properties calculated from lab tests. This is mainly due to random fiber orientation in the molding.

Now the problem is how to simulate this kind of materials in a FEM code (Presently I use Ansys Workbench); in fact is very difficoult to specify a correct Young modulus value.
I made some comparison between FEM and test results and the difference between the datasheet Young modulus and the evaluated modulus is about 30-40% (the real modulus is always less then the datasheet modulus).

My questions are:

1) Has anyone ever had experiences with this kind of materials?

2) Are there any suggestions how to model this matierial?

3) For what concern the failure simulation, how to model this kind of materials, as elasto-plastic, perfectly elastic, or like rubbers using Mooney-Rivlin coefficients?

Thanks to everyone

 

[SWComposites]

"observed a huge difference between the material properties specified in the datasheets and the material properties calculated from lab tests"
> not surprising - vendor data is typically from molded coupons with highly aligned fiber distributions; vendor data is typically not representative of real part data.


2) Are there any suggestions how to model this matierial?
> use properties from your tests, which ideally have similar fiber distributions are your real parts. You may have to use a code such as Moldflow to predict the fiber orientations and volumes in your parts and than tailor the properties in your FEM to match. Not typically trival

3) For what concern the failure simulation, how to model this kind of materials, as elasto-plastic, perfectly elastic, or like rubbers using Mooney-Rivlin coefficients?
> linear elastic should be fine for a fiber reinforced plastic.