4. Conclusions
Stringer run-out damage mechanisms have been simulated using a small scale, strap lap specimen in order to assess the potential for self-healing using an in-situ vascular network. The introduction of a non-optimised vascular network (in terms of location and orientation) had marginal influence on the global mechanical properties: a reduction of approximately 15% in initial debonding strength and 10% in final failure load. By reducing the diameter of the vascule, changing its position or locally aligning with the host ply, there is scope to further reduce, or eliminate, the network's effect on the mechanical performance. In terms of property recovery, the initial stiffness was fully restored and the strength, in terms of initial disbonding, was increased due to the localised improvement in fracture toughness. In order to demonstrate the potential to transfer these findings to larger, more realistic composite assemblies, two stringer run-out panels were manufactured, one as an unmodified baseline and another containing a vascular network. Both configurations behaved similarly and effective healing was observed. A similar damage pattern was observed as found for the simpler strap lap specimen, thus it can be surmised that a similar magnitude of healing recovery could be achieved. Future work will focus on optimising the vascular approach for the stringer run-out assembly, and consideration given as to how an autonomous sensing-healing system could be incorporated to increase the reliability and robustness of this technology and drive it towards industrial exploitation.