Abstract
Composite patch repair has emerged as an effective technique for extending the service life of cracked metallic structures by reducing crack-tip stress intensity and avoiding costly component replacement. However, identifying an optimal repair configuration remains challenging because minimizing both the stress intensity factor (SIF) and patch area involves competing design objectives, while the integrity of the adhesive layer must also be maintained. This study presents a constrained finite element-based optimization framework for the design of double-sided composite patch repair applied to cracked AL7075-T6 plates. Four repair parameters, namely patch width, patch length, adhesive thickness, and patch thickness, are simultaneously optimized using two distinct approaches: a weighted-sum Particle Swarm Optimization (PSO) algorithm and the Non-dominated Sorting Genetic Algorithm II (NSGA-II). An interfacial shear stress constraint is imposed to preserve adhesive layer integrity throughout the optimization process. The parametric analysis highlights the individual influence of each repair parameter but does not account for their interactions or combined effects on the competing objectives. The weighted-sum PSO framework directly yields a targeted design compromise by prioritizing SIF reduction while avoiding excessive patch dimensions. Conversely, NSGA-II generates a broad set of non-dominated solutions without a priority preference, providing a wider range of trade-offs for post-optimization decision-making. Compared with conventional parametric design, the optimized solutions reduced the required repair area by approximately 82% while maintaining acceptable repair performance and satisfying the adhesive shear stress constraint. The proposed constrained multi-objective framework provides an effective computational tool for designing lightweight and mechanically reliable composite repairs for cracked metallic structures.