hexagon logo

Hello I am trying solid composite. For have bending results, I need 3 elements covering the thickness. Is there a way to apply 1 comp material to all the elements or do I have to create a material with 1/3 of the plies to have the total number of plies?

I know solid shell elements are better for this application but I am trying to compare the results in both cases
  • If you use the assumed strain formulation you can have good out-of-plane bending results with a single layer of elements. Please review example 7 in the Implicit Nonlinear Demonstration Problem's Manual
     
    image
     
    Regarding the original question, it will depend on the layup. If you had 9 layers 0/90/0/0/90/0/0/90/0 then the composite could be 0/90/0 for each element. but for typical layups that would not be possible
     
  • Thank you for your answer. It seems my results highly underestimate the stresses when I run a LNSTATICS analysis (when I select linear in subcase parameter for sol400) but not when I select non linear (NLSTATICS). Why is that? What do I do wrong? I can't seem to have the same results as example 7 cited above. Thank you for your help!
  • I think you are comparing the results at the centroid of the element with the values in the table in the manual. In the manual, they took the stress results from the centroid of elements 79 and 80 and then extrapolated them to the symmetry line. It's a bit of a hack but technically accurate.
     
    600.8+(25-23.125)*(657.2-600.8)/(24.375-23.125)=685.4 MPa
     
    You will be able to better compare the centroid stress if you refine the mesh in the x-direction. I verified this in Marc (image below)
     
    image.png