Hello everyone,
I’ve been investigating a discrepancy between the in-plane Young’s modulus (E_1) obtained from SwiftComp 2D (Kirchhoff–Love plate) and 3D homogenization.
For a single-ply plain-weave laminate, the results differ by approximately 16%:
-
3D SwiftComp: (E_1 = 81.86) GPa
-
2D SwiftComp: (E_1 = 68.70) GPa
Using the 3D engineering constants, I calculated the bending stiffness using classical laminate theory:
[D_{11}=\frac{E_1t^3}{12(1-\nu_{12}\nu_{21})}]
For a thickness of 0.0667 mm, this gives approximately 2.03 N·mm, which is close to the 2.17 N·mm obtained from an independent FEM analysis.
However, the 2D SwiftComp homogenization predicts (D_{11}=0.55) N·mm, which is significantly lower.
Interestingly, as I increase the number of plies, the 2D and 3D predictions begin to converge. The 2D in-plane modulus (E_1) approaches the 3D value, and the bending stiffness (D_{11}) from the 2D analysis approaches the FEM results.
Has anyone encountered similar discrepancies when homogenizing thin plain-weave composites using SwiftComp? Is this expected behavior due to the different homogenization assumptions, or could there be an issue with my model setup?
Any insights would be appreciated.
Thank you!