Cohesive failure analysis of carbon-fiber composite patch repair in perforated steel tubular elements

The use of composite materials to repair corroded steel structural components has expanded in the offshore oil and gas sector. The effective use of composites requires a meticulous process for choosing appropriate constituent elements. In this regard, numerical modeling is an efficient tool to better understand the impact of associated variables and enables the identification of failure causes and simulation of failure propagation. This study aims to assess the mechanical behavior of the adhesive layer at the steel - composite material interface in the repair of perforated tubular structures under bending, compression, and bending-compression. A numerical model is developed simplifying the examination and characterization of the cohesive failure mechanisms and modes occurring between the laminate patch repair and the perforated steel tubular element. The stress distribution and cohesive failure at the interface are then thoroughly investigated. The impacts of the inclusion of fillets at the patch repair ends to smoothen the transition between the composite material and the tube and the repair length are also evaluated. A Python subroutine is developed to identify the failure mode mechanisms to characterize the adhesive areas subjected to peel, shear, or mixed modes. Results suggest that pure shear mode is predominant in most of the repair locations across the three types of loading. For bending loads, the cohesive failure close to the cutout shows prominent shear mode. In the other loading conditions, the adhesive nearly fails in the vicinity of the cutouts (i.e., damage values are close to 1) in a predominant near pure shear failure mechanism. The use of fillets displaces the failure location for bending load, hence increasing the failure load level. Repairs were more extensive under Bending Moment load conditions compared to other load types. The mode ratio analysis and methodology presented in this work offer valuable tools for understanding the causes of failure modes.