Study of adhesively bonded repairs in aircraft CFRP primary structures
|Abstract:||This master’s degree thesis focuses on testing and modeling the bonded scarf-stepped composite joint performance under uniaxial tensile loading as part of the effort to develop testing protocols and analytical tools for the design of scarf bonded repair for primary aeronautical composite structures. First, plain weave (PW) and 8-harness satin (8-HS) carbon fiber-reinforced plastic (CFRP) pre-preg composite materials were tested under uniaxial tensile loading at room temperature dry (RTD) and elevated temperature dry (ETD) conditions. The gathered characterization data was later used for the numerical modeling of the repairs. Furthermore, smoothed parent laminate bonding surface and stepped patch laminate bonding surface (scarf-stepped repairs were performed using both PW and 8-HS composite materials with matching quasi-isotropic ([+45°/0°/-45°/90°]2s) ply stacking sequence between the parent and the patch. The adhesive film that was used is the Cytec FM300-2M. The effects of environmental conditions and the influence of the scarf angle (i.e. 3˚, 5.5° and 7.5˚) on the performance of the bonded repairs were investigated. The tensile test results revealed that the scarf angle has a significant impact on the failure mode of the repaired composite part. While substrate failure occurred with a 3˚ scarf angle, cohesive shear failure was observed for the 5.5° and 7.5˚ angles. This change in failure mode is consistent both at RTD and ETD. When compared with the pristine laminate, an insignificant drop in stiffness was found regardless of the scarf angle. Although, increasing the scarf angle led to a significant drop in strength restitution in comparison with the pristine laminate. This indicates the importance of the scarf angle on the structural integrity of a scarf-step bonded repair. The tensile test results in ETD conditions suggest a slight decrease in stiffness and strength for both materials at ETD. Eventually, 2D through-thickness finiteelement analyses were also conducted using both ABAQUS Standard and Explicit. An elastic analysis was first performed to predict the distribution of normalized shear and peel stresses in the middle of the adhesive along the bondline for three different joint geometries (scarf-scarf, scarf-step and step-step). As opposed to the uniform stress distributions found along the bondline of the scarf-scarf configuration, high and frequent peaks of peal and shear stresses were found for both scarf-step and step-step configurations. These observations led to the conclusion that one must be particularly cautious when modeling a scarf joint bonded repair. Assuming that the patch laminate bonding surface is smoothed (i.e., scarf-scarf configuration) while it is actually stepped (i.e., scarf-step configuration) can lead to overestimating the overall repair strength since the high stress peaks caused by the geometric irregularities of the stepped patch laminate bonding surface would then be ignored. Furthermore, an elastic-plastic analysis was conducted using the already implemented plasticity and shear damage models in ABAQUS. These plasticity and damage models were used for the adhesive film only. The composite material was supposed to behave linear-elastically up to failure. The maximum strain criterion was used to predict the first ply failure in the composite. The predictions obtained with the model correlated very well with the experimental results.|
|Document Type:||Mémoire de maîtrise|
|Open Access Date:||9 October 2018|
|Collection:||Thèses et mémoires|
All documents in CorpusUL are protected by Copyright Act of Canada.