No description available.

This body of work has sought to investigate the capacity of sandwich panels containing networks of sintered fibres in the core, to absorb energy when struck at normal incidence by spherical projectiles. Experimental impact tests have been conducted on sandwich panels and monolithic steel plates, over a large range of impact velocity – 80 ≤ <i>V</i>, ≤ 600 m s^-1. The absorbed energies have been calculated from measured incident and residual velocities. The experimental impact tests were then simulated using the explicit finite element code in ABAQUS/CAE. Single faceplate FE models and full sandwich panel models were developed. The faceplates were modelled as isotropic, strain and strain-rate dependent shells, using the constitutive plasticity models of Johnson and Cook and von-Mises. Failure of the faceplates was simulated using a strain rate-dependent, critical plastic strain fracture model. In the sandwich panel model, core material plasticity was modelled using a VUMAT sub-routine. The sub-routine considered the plastic compression of an anisotropic crushable continuum. Failure of the core material was simulated using a quadratic shear stress failure criterion. Experimental tests and numerical simulations on single faceplates indicate a transition in material behaviour for strain rates exceeding ~10⁴ s^-1. This has been interpreted as a transition in the rate-hardening mechanism from conventional dislocation effects to dislocation-drag phenomena. Minimum perforation energies are closely predicted by the empirical model of Hill, whilst a modified version of the analytical model of Teng and Wierzbicki exhibits close agreement with experimental data at high impact velocities. The numerical model successfully predicts the plastic strain fields and failure modes over the experimental velocity range.

No description available.

This thesis investigates the low-cycle fatigue behaviour of CMS-4. The focus is on R=0 load controlled fatigue, with an emphasis on deformation at 750°C. A particular aim of this research is to identify how fatigue proceeds at stress concentrators. Here, high stresses in plain bars have been compared with similar maximum stresses around the known stress concentration of a Kt-2 notch. Fatigue tests were conducted on both plain and notched specimens, across a range of stresses and temperatures, and the deformation has been imaged using TEM. The plain bars show a range of deformation mechanisms, which have been described and analysed. Of particular interest is the presence of widely spaced dipoles of single dislocations, previously noted in tensile studies. These have been imaged on slip planes, and it is shown that they are not formed by the mutual attraction of two independent dislocations, as has been previously thought. Rather, they are two sides of an á110ñ loop, separated by a region of APB. These loops are expected to expand and contract within the precipitates, and it is shown that this could account for a significant degree of plastic strain within the plain bars. Although the fracture patterns and lifetimes are very similar to those seen in the plain bars, no microstructural evidence of significant deformation is seen from TEM examination of the notched bars. Visible dislocation density is consistently extremely low, and it is seen that there is no correlation between deformation in the bulk of the plain bars, the notched bars, and failure. Hence, a new mechanism of initiation is proposed for failure below 950°C. It is proposed that initiation below 950°C is related to the extrusion of g channels at the surface of internal porosity, and the concomitant formation of subsurface voids.

No description available.

No description available.

This dissertation describes work on the thermomechanical fatigue (TMF) behaviour of coated single crystal, nickel-based superalloys. During the course of study, the TMF facility at Cambridge was developed to enhance the reliability of the test technique and to enable the testing of thin ceramic shell TBC test pieces. Benchmark TMF tests were carried out under strain control on a series of different coatings to evaluate the performance of two new coat systems developed at Rolls-Royce that incorporate a platinum surface modification. Cyclic life to failure suggested that the surface modification had minimal effect on TMF performance. The results compared favourably with data from different research centres as well as the lifting model used at Rolls-Royce to predict high temperature coating mechanical integrity (the T<SUB>crit</SUB> model). From this model, the effect of different TMF cycle directions was considered. Experimental tests on aluminide and overlay coated specimens did not support the T<SUB>crit</SUB> predictions in terms of cyclic life to failure, although crack density analysis would seem to provide a more accurate correlation between the TMF cycle experienced and the coating performance. A separate finite element (FE) approach to predict the stress/strain history in an overlay coated specimen was developed in order overcome some of the limitations of the T<SUB>crit</SUB> method. Incorporating developments in the experimental technique, TMF tests on the modified overlay bond coat with and without a thin shell ceramic top coat, were performed. The FE model was extended to include this ceramic in order to investigate the constraint of the bond coat. The level of constraint generated does not seem to support the effect that a thin shell ceramic has on suppressing the surface undulations and crack densities observed in TMF testing.

Hot-wet ageing was performed on three different adhesive joint systems for over 10,000 hours. Joint strengths were found to decrease with ageing time and failure was near-interfacial in the adhesive. There was no evidence of an adhesion type failure mechanism. A novel video imaging system was developed for one of the joint systems which was transparent, and this gave new, in-situ information on damage initiation and propagation in joints. The material properties of the acrylic and epoxy adhesives used for the joints were needed as input for the theoretical studies. Tensile properties were determined for adhesive samples that had been hot-wet aged and which had a range of water contents. Increased water content caused significant reductions in tensile strength and failure strain for both acrylic and epoxy. The reduction in modulus was only significant for the acrylic. A new shear test procedure using a bonded tube-plate sample was developed for adhesives with high failure strains. Accurate and reliable shear properties were measured for the acrylic adhesive. A thickness effect was observed with thin adhesive layers being stronger than thick layers. An analysis was derived, using observations from video images of the fracture process during the shear tests, that helped explain the observed thickness-strength relationship. The theoretical model for adhesive joint durability was developed based on changes to the mechanical properties of the adhesive due to water up-take. The adhesive joints were modelled using crack-bridging concepts not previously applied to this type of problem. The crack-bridging model uses large-scale bridging (LSB) conditions, which requires the material behaviour to be coupled to the component geometry, and this was implemented in a finite element model. The LSB model was found to give accurate results and to give useful predictions of joint performance in response to ageing. The LSB model was able to simulate damage development occurring in joints under load. Failure surfaces for the three joint types were developed and the most appropriate was found to be related to the shear-normal stress ratio in the adhesive.

No description available.

The fracture behaviour of polymer composite-metal joints has been investigated within the context of two real joint applications. A joint intended for the joining of a glass-fibre reinforced vinylester composite (GRP) superstructure to the steel deck of a naval vessel was the starting point for the project. The novel technique lies in the manufacture of the joint in that a steel adherend is incorporated into the superstructure panel during the manufacture of the panel itself by a resin infusion moulding process. One end of this adherend protrudes from the finished panel which can then be welded to the steel deck. The vinylester matrix resin thereby acts as the sole source of adhesion at the composite-metal interface. Characterisation of the mechanical response of this joint highlights the influence of the toughness of the GRP on the joint behaviour. The fracture behaviour of this interface is then investigated to establish the effects of steel surface roughness and heating. Accelerated ageing of the joint in salt and distilled water is carried out and the effect of the GRP mechanical properties and joint mechanical response evaluated. A second application considered has been the joining of a titanium adherend to a carbon-fibre reinforced epoxy composite adherend for use in a racing car suspension system. A comparison of the effectiveness of different titanium surface treatments is made in terms of the surface profile and oxide thickness. A range of adhesive layer thicknesses are used in the manufacture of the joints and the effects of this and the surface treatment on the fracture behaviour are determined. The main conclusions drawn are that the principal factors affecting fracture behaviour of metal-composite joints are adherend surface chemistry, moisture absorption, exposure to elevated temperature, and joint design.