| Abstract |
Computational material design can be seen as a powerful virtual testing laboratory which will progressively replace expensive and time consuming experimentation techniques in order to devise the materials of the future. In this scenario, the essential physics involved in the service of the new material need to be captured through reliable simulation strategies.
These techniques could be applied, for instance, to the simulation of the new generation fiber reinforced concrete utilized in civil engineering structures, such as dams and bridges, under severe seismic loading. To control the brittleness and extend the service timespan of the structures one needs to capture the interactions of the underlying material components with the local fields and their effect at a higher scale. Obviously, a full fine scale analysis would result prohibitive from a computational point of view and, for this reason, the development of robust and objective multiresolution simulation techniques is seen as an attractive remedy.
This proposal is conceived as a complement to the European project: COMP-DESMAT Advanced tools for computational design of engineering materials funded by the European Research Council advanced grant scheme. The idea is to contribute to the computational design technology through the development and assessment of multiresolution techniques in time and space. In particular, DYNAMOS focuses in the use of multiresolution techniques or concurrent multiscale methods where lower and upper material level discretizations are resolved simultaneously. Spatial multiresolution analysis can be achieved by means of an adaptive refinement at particular regions of interest or hot spots. For instance we are interested in zooming in at those places where fracture growth and coalescence takes place. An adaptive time discretization is also key to correctly reproduce the dynamics of fracture processes and study propagation phenomena and crack tip velocities which highly determine the overall brittleness of the structure. We also propose other promising computational schemes such as Modal Order Reduction in order to capture the essential mechanical features of different heterogeneous domains with the use of well chosen global basis functions. This strategy can be specially attractive to significantly accelerate the computation time.
The expected outcome of the project is a set of general tools with a broad applicability in many dynamic simulations. The range of studies can vary within different loading regimes, e.g. mechanical analysis of civil engineering structures submitted to dynamic excitations and seismic actions or impact analyses giving rise to fragmentation and fast crack propagation processes. The developed set of tools is envisaged to be a key ingredient to unravel dynamic fracture mechanisms such as unstable crack propagation and branching and contribute to the material design concept providing an accurate description of the underlying physics. Therefore, we expect to contribute to the Leadership in Enabling and Industrial Technologies (LEIT) and, more precisely, to the Widening materials models strategy which is in deep connection with most of the challenges in our society. |
