Due to the growing demand for offshore renewable energy, the development of durable submarine power cables is critical. Submarine power cables are expected to have a service life of over 20 years. However, it has been shown that these cables suffer from water-tree flaws that progressively extend to conductors and corrode copper, which may lead to premature failure. Water treeing is caused by the of interconnection of voids (of a few nanometers) that are present in the insulator after manufacturing or formed during operation. The economic consequences of a breakdown can be drastic due to the heavy maintenance required. In the current study, the insulator is modelled as cubic unit cells containing water voids in the form of ellipsoids. The displacement field of ellipsoids is found to be dependent on its distribution in the cubic cell and on the applied electric field. Von Mises stress and effective plastic strain at the tips of the ellipsoid are found to be significant when either the relative distance between the two ellipsoids is short or the applied electric field is high. The proposed model is intended to provide insights into the ageing of cross-linked polyethylene (XPLE), which is extremely difficult to predict experimentally due to the excessive time needed to achieve coalescence of voids.
Compared to conventional projection-based model-order-reduction, its neural-network acceleration has the advantage that the online simulations are equation-free, meaning that no system of equations needs to be solved iteratively. Consequently, no stiffness matrix needs to be constructed and the stress update needs to be computed only once per increment. In this contribution, a recurrent neural network is developed to accelerate a projection-based model-order-reduction of the elastoplastic mechanical behaviour of an RVE. In contrast to a neural network that merely emulates the relation between the macroscopic deformation (path) and the macroscopic stress, the neural network acceleration of projection-based model-order-reduction preserves all microstructural information, at the price of computing this information once per increment.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.