Identification of Effective Healing Locations for Recovering Compressive Strength

This paper investigates the effective distribution of self-healing resin within a delamination area in a fibre reinforced composite material. The study aims to deliver information on the healing resin volume required for achieving mechanical performance recovery across a delamination plane. Targeted healing agent infusion of critical delaminations within the entire damaged volume is critical to minimise localised stress concentrations within the delamination plane during polymerisation, especially when optimised for healing under operational loading. A genetic algorithm (GA) [1] has been applied to perform a bi-objective optimisation [2,3] with respect to (a) maximum compressive strength recovery and (b) minimal infused healing resin volume. In order evaluate the first objective of maximal compressive strength, a finite element model is repetitively solved using LS-DYNA [4] incorporating the methodology illustrated in figure 1. Here a layer of cohesive elements is located in the mid-plane between two sub-laminates. The GA individually controls the state of each cohesive element within the zone of potential delaminations, located in the middle section of the specimen (see figure 1b). Through the implementation of a user defined cohesive elements [5], a deactivated element, representing a delamination, numerically functions as a contact definition. Hence, the status of an element (active cohesive formulation or delamination) can be used to generate the genetic representation of the solution domain. The state of all elements within the region of potential delaminations is encoded as a chromosome (bit string containing 1 and 0’s). To ensure an acceptable run-time of the finite element model with respect to the optimisation process, a slice model with a single element thickness is used. (continued)