University of Manchester

Damage Mechanisms in Bonded Scarf Repairs in Composite Structures

Kali Babu Katnam
Project Completed

Scientific Case

The proposed scans will be on carbon-fibre-epoxy laminates with well-defined stepped laminae discontinuities, which represent a bonded repair scenario. This is part of an EPSRC-funded project (EP/N021975/1 - Towards Tailored Composite Bonded Repairs). To minimize the removal of undamaged material from the damaged component during bonded composite aircraft repairs, the design of stepped scarf repairs should be tailored for a given repair condition. The approach presented in this work, i.e. with non-linear stepped scarf bondlines and non-conventional patch geometries, is a potential solution to reduce undamaged material removal from parent laminate as well as to improve joint efficiencies. In this context, to study the behaviour of such tailored repairs, representative bonded stepped scarf joints with non-linear bondlines are fabricated with carbon/epoxy laminates following three repair fabrication methods: (1) a soft-patch with resin infusion and discontinuous fabric layup; (2) a hard-patch with secondary adhesive bonding, where laminates are manufactured and CNC machined to produce non-linear tapered edges; and (3) an in-situ infusion of the repair laminated together with a cured and machined laminate. In all these cases, composite laminates are manufactured using vacuum infusion with a stacking sequence of [02/902]2S. For fabricating hard-patches, a structural epoxy film adhesive for bonding and a chemical wipe for surface preparation are used. The objectives of the scans are to identify: (1) manufacturing and repair fabrication defects in the vicinity of the stepped bondlines from untested specimens; and (2) sub-surface matrix-cracking and delamination regions from partially damaged (subjected up to 75% tensile failure load) specimens.

Experiment Design

The aim of the scans is to identify manufacturing defects and correlate them with damage initiation/propagation regions. Specimen dimensions: 25 mm (width) x 100 mm (length) x 5 mm (thickness). The objective is to scan relatively large volumes for identifying features (voids, pre-cracks, matrix-cracks or delamination) > 50 microns. A voxel size of 15 micron or above could be acceptable. Untested (i.e. not subject to any loads) and tested (subject to 75% tensile failure load to introduce damage/cracks; not in-situ loading) specimens will be scanned.
Scanners and Rigs
Nikon Custom 320kV Bay
Not Required
Not quite sure about the number of scans and days. Scans up to 4-6 and days up to 3-4.

Sample & Safety

Specimen dimensions: 25 mm (width) x 100 mm (length) x 5 mm (thickness). This could be made smaller if required. The specimens consist of carbon fibres, epoxy resin and a film adhesive (epoxy veil). We had a discussion with Serafina Garcea. No hazards were identified.
Low Hazard

Scan Records

Project Report