Unexpected Decrease in Total Contact Force and Discrepancies in Flexural Strength Predictions in Three-point Bending Simulation

Qiujun Wang

Hello,

I am conducting a three-point bending simulation of a composite material using Radioss. However, I am encountering an issue where the total contact force consistently begins to decrease at a fixed time, regardless of the failure criteria employed. While the predicted flexural modulus aligns with experimental results, the flexural strength is significantly lower than anticipated.

Additionally, I have observed that inputting a lower E22 value results in a higher predicted flexural strength, and delays the decrease in total contact force.

I am confused by this problem and would appreciate any insights or assistance you can provide. Thank you!

Polyvios Romanidis

Hi Qiujun,

There are somethings that need improvement in your model.

1. The components you are using have defined thicknesses in a way that initial penetration exists.  So, you have to move them away from one another to prevent this.  You can always check the parts contact with the module you can see in the next figure.
   
   
2. Additionally, the /INTER/TYPE7 you define is not the one recommended for these kind of applications.  You can find the proposed parameters in the following figure.
   
   
   
3. Additionally, I strongly recommend to use Ishell formulation equal to 24 as a best practice and because it is more suitable for more of the material laws, especially composites.  For composites, I would say that Ishell equal 12 might be even better.
   
   
4. Finally, the material properties you have defined are a bit strange (not neccessarily wrong).  For this kind of materials we usually have lower Poisson Ration values (around 0.05) and greater difference between yield and maximum strength stress (you used 1000 MPa for both).
   
   

I attach a corrected model as well and some corresponding results.

Polyvios