In order to investigate Lamb waves in a carbon-fiber composite plate (size: 0.5m x0.5m), which comprises unidirectional, biaxial and twill layers, I homogenized the 3D properties of each layer. Then, I modeled the layers individually using C3D8R-elements.
I find that wave propagation of the Lamb waves is significantly slower in the simulation than it was in experiments (Time of Flight to travel horizontally across half of the plate in the experiment was 4e-5s, but is 6e-5s in the simulation).
My assumption now is that the top layers (twills) greatly impact the wave propagation velocities, but given the fact that I homogenized the twill properties, the Young's moduli and densities of the homogenized layer are much lower than they are locally in the yarns in the experiment.
Is this plausible and how can I go about modelling the local properties for a model of this size? Would it, for example, be realistic to model fibers and matrix as two separate layers for the weave?
As long as the wavelength is much larger than the size of your unit cell of your woven microstructure, homogenization should work. However, you should use a good theory to carry out the homogenization to obtain equivalent effective stiffness. I will recommend you to take a look at TexGen4SC on cdmhub.
- Badges
- Science topic