There have been a lot of failed simulations based on continuum mechanics. Last decade, a number of numerical techniques based on discontinuum representation of cracks, delamination, etc., have emerged.
I think size effect will be an increasingly important topic. Present strength-based approaches fail in prediction of this phenomenon since they lack of the energetical-statistical formulation to deal with it.
Size effect has been experimentally reported for composites which are quasi-brittle by nature. This will urge for suitable approaches to address this problem especially if one considers that models are usually calibrated/validated on laboratory-scale specimen and than applied to study large structures (e.g. aircraft components or wind turbine blade).
Size effect is bot only the problem faced in composite material as well, it is related to the continuum mechanics itself in some sence. In fact, every material has its structural feature in different scale, the more you take the structural feature into your consideration, the nearer it describe the reality.
Size effect is not only the problem faced in composite material, it met the similar problem in "homogenoius material" as well. It is related to the continuum mechanics itself in some sence. In fact, every material has its structural feature in different scale, the more you take the structural feature into your consideration, the nearer it describe the reality.
We had a go on simulation of size effect in laminates subjected to transversal crack and delamination. Although all simulation is virtual, the results with XFEM are encouraging to tackle both discontinuity and size effect if adequate constitutive laws are also in place. I have attached our paper below
Article Transversal crack and delamination of laminates using XFEM