In an attempt to find a way of improving the damage tolerance of composite bonded repairs to metallic aircraft structures, the effect of using conventional crack growth retardation techniques in conjunction with bonded repairs was experimentally investigated. Hence, an experimental test program was set up to determine whether fatigue crack growth under bonded repairs is retarded further by giving the crack to be repaired a crack growth retardation treatment prior to repair patch application. In addition, it was set up to determine the influence of a bonded repair on the effectiveness of a crack growth retardation method. Centrally cracked aluminium plates were used. Stop drilling followed by cold hole expansion and the application of single overloads were selected as retardation treatments. Two patch materials were considered; boron/epoxy and Glare 2. Further test variables were the aluminium alloy and the plate thickness. Fatigue testing was carried out under constant amplitude loading and baseline results were determined first. In addition to optically monitoring the crack growth, local and global out-of-plane deformations were visualised with holographic interferometry and shadow moire??. Furthermore, the stress intensity factors under the repair patch were examined with strain gauges and measurement of the central crack opening displacement. Disbonds and fracture surfaces were studied after residual strength tests. The crack growth results obtained showed that retardation treatments decrease crack growth rates under a repair patch and that the effectiveness of a retardation treatment is increased by the patch. Although identical crack growth rates were observed under boron/epoxy and Glare 2 patches, the reinitiation period after the retardation treatment lasted longer when Glare 2 patches were applied. Analytical predictions of the extent of retardation based on existing models showed that the conjunctive effect of retardation treatments and bonded repairs was underestimated. A sustained reduction in crack growth rates was observed under bonded repairs with a prior overload retardation treatment. It was concluded that the damage tolerance of bonded repairs is increased by the application of a crack growth retardation treatment because the crack growth is retarded further. These findings indicate that the range of cracks in aircraft for which bonded repairs can be considered is expanded and that economic benefits can be obtained.

In an attempt to find a way of improving the damage tolerance of composite bonded repairs to metallic aircraft structures, the effect of using conventional crack growth retardation techniques in conjunction with bonded repairs was experimentally investigated. Hence, an experimental test program was set up to determine whether fatigue crack growth under bonded repairs is retarded further by giving the crack to be repaired a crack growth retardation treatment prior to repair patch application. In addition, it was set up to determine the influence of a bonded repair on the effectiveness of a crack growth retardation method. Centrally cracked aluminium plates were used. Stop drilling followed by cold hole expansion and the application of single overloads were selected as retardation treatments. Two patch materials were considered; boron/epoxy and Glare 2. Further test variables were the aluminium alloy and the plate thickness. Fatigue testing was carried out under constant amplitude loading and baseline results were determined first. In addition to optically monitoring the crack growth, local and global out-of-plane deformations were visualised with holographic interferometry and shadow moire??. Furthermore, the stress intensity factors under the repair patch were examined with strain gauges and measurement of the central crack opening displacement. Disbonds and fracture surfaces were studied after residual strength tests. The crack growth results obtained showed that retardation treatments decrease crack growth rates under a repair patch and that the effectiveness of a retardation treatment is increased by the patch. Although identical crack growth rates were observed under boron/epoxy and Glare 2 patches, the reinitiation period after the retardation treatment lasted longer when Glare 2 patches were applied. Analytical predictions of the extent of retardation based on existing models showed that the conjunctive effect of retardation treatments and bonded repairs was underestimated. A sustained reduction in crack growth rates was observed under bonded repairs with a prior overload retardation treatment. It was concluded that the damage tolerance of bonded repairs is increased by the application of a crack growth retardation treatment because the crack growth is retarded further. These findings indicate that the range of cracks in aircraft for which bonded repairs can be considered is expanded and that economic benefits can be obtained.

A Primary objective of the present research is the development of robust CAE (Computer-Aided Engineering)-based approaches for designing armour plates subjected to ballistic impact by small-calibre hardened peojectiles with or without a protective sheath. Amongst the challenges in simulation is the capturing of target plate material behaviour at high strain rates with possibilities of adiabatic heating. A comprehensive numerical study carried out has resulted in the identification of simulation guidelines using a commercially available explicit finite element anlaysis solver (viz. LS_DYNA). The interferences thus drawn in terms of modeling approach 9I.e. shell, solid or axisymmetric or a mixed representation). Mesh density and element type, contact condition, and constitutive model 9I.e. discrete strain-rate based, Cowper-Symonds, or Johnson-Cook) with failure criteria are verifiable and greatly beneficial for armour plate design. Confidence in the suggested procedures has been obtained through extensive correlation of numerical results with experimental residual velocities and ballistic limits as well as projectile and target plate failure modes. A wide range of impact velocities has been considered (from a low velocity of about 5m/s to an ordnance range velocity of 800+ m/s). Target plates made of variants of mild steel and aluminium alloys have been studied. The simulation approaches have been applied to single-layered as well as multi-layered target plates. Although a majority of the comparisons has been made against published test results, a new ballistic impact testing facility has been set up in course of the current research and excellent correlation of numerically predicted residual velocities and failure modes has been obtained against the tests carried out for aluminium plate using the latter facility. A unique feature of the current experimental effort is the capturing of the complete trajectory of projectile beginning with oblique impact through subsequent perforation/ricochet. Furthermore, projectiles of various nose-shapes such as ogival, conical, hemispherical and blunt have been employed. The power of simulation has been demonstrated with the help of a number of parametric studies with variables such as plate thickness and material properties, as well as projectile mass and diameter, and obtaining physically consistent results. Additionally, existing semi empirical models for residual velocity and ballistic limit prediction have been reviewed, and new user-friendly models have been proposed based on energy conservation and predominant shear plugging failure mode of target plate. Finally, the goal of applying the present research work as a design tool can be well-served by packaging the knowledge gathered here in the form of a user-friendly guide with a graphical user interface(GUI). To this end, an application using MS windows VC++ utilities has been created with the functionalities of: (a) viewing reference LS-DYNA input data files for selecting typical problems of impact on steel and aluminium plates; (b) computing complete lists of strain rate-based material quantities required in LS-DYNA material models like discrete strain rate-based, Cowper-Symonds and Johnson-Cook by specifying the minimum number of easily available quasi-static properties (such as elastic modulus, yield and ultimate strengths, etc.), and (c) estimating residual velocities using the semi-empirical relations for steel and aluminium plates derived in the current work.