A Simple Material Model for Composite Based on Elements with Realistic Stiffness

The obvious question is how to combine models and data to create a virtual prediction tool? There is a long history with measured data to adjust finite element models representing the geometry and material properties of a tested component. It seems to be a good starting point to represent the initial geometry with correct stiffness and challenge the element formulations to maintain realistic stiffness even when the elements are severely deformed. A proper discretisation is crucial when building a finite element model, and most metallic parts are represented as continuums although we know about the elementary particles. It is likely that this simplification is acceptable as long as the strains are small and the strain hardening is sufficient to compensate for local variation in material properties, but remember that brittle behaviour may be the result when the elastic energy stored in the component is allocated into the first local area that fails.