All weld metal test pieces were produced from two commercial and six experimental 7016 type electrodes, each having different contents of manganese and nickel. These were tested in three point bending over a range of depressed temperatures in a cryostatic cell of the author’s devising. Notched samples were used to obtain values for the cleavage fracture strength. This was calculated by reference to slip line field theory from the fracture load at the test temperature where fracture was first coincident with general yielding. The commercial electrodes were first tested in the as welded condition and all eight electrodes were then tested after a heat treatment designed to give a microstructure similar to that found in re-heated weld metal. Notched and pre-cracked samples were then tested in the heat treated condition to determine KI values. Fracture surfaces were examined by scanning electron microscope and grain size measurements made by optical microscopy. Lower shelf impact energy measurements were also made. The results show that both manganese and nickel individually increased the cleavage fracture strength. In combination the best cleavage properties were given by alloying contents of 1.0% manganese and 1.5% nickel increasing alloying beyond this level had a detrimental effect. The mechanism for this was a combination of grain size effects, changes in the slip characteristics of the material, the size, distribution and nature of inclusions and the effects of alloying on grain boundary and matrix/inclusion cohesion. In re-heated material at test temperatures studied the plane strain fracture toughness is influenced by similar factors to those which determine cleavage fracture strength.

The purpose of the study was to investigate the metallurgical interactions that occur during the brazing of zirconium with Ag-28Cu filler, with particular reference to their effect on mechanical properties. The ultimate aim was to provide information that would aid in the production of tougher brazed zirconium joints than are currently being obtained. The investigation was carried out in a number of stages: 1. A general literature survey of the metallurgy of brazed joints. 2. Zirconium brazed lap joints were made using either vacuum furnace or high frequency induction brazing. Interalloying was studied by optical and electron microscopy, X-ray diffraction and electron microprobe analysis. Interface layers were produced consisting of Cu4Zr, CuZr2, AgZr2 and Cu, distributed in a mixture between two sub-layers. Induction brazing produced high quality joints with interface layers thinner than those in vacuum brazed joints; a link between heat-up rate and wetting was observed. 3. Butt joints for mechanical testing were produced to optimum conditions. Instrumented impact testing showed that the toughest joints were those brazed with a reactive copper filler, although in all joints most of the energy absorbed was due to fracture initiation energy. 4. The greater toughness of the reactive copper brazed joints was due to a lack of flaws of any kind in their microstructure, which consisted of a eutectic of CuZr and CuZr2. In contrast, interface layer Ag-28Cu brazed joints contained shrinkage cracks and interfacial voids, along which fracture occurred.

Derivatives of 4,4'-azobis-4-cyanopentanoic acid were prepared using its diacid chloride and various hydroxyl and amino compounds. The objective was to obtain substances that would act both as initiators for different modes of free-radical polymerisation and also as stabilisers for emulsion polymerisation reactions carried out in the absence of added conventional surfactant. During the preparation of the above compounds, it was discovered that 4,4'-azobis-4-cyanopentanoyl chloride appears to exist in two different forms called I and II. These forms yielded two different forms of esters when reacted with straight-chain aliphatic alcohols. These esters are called 'esters A' and 'esters B', respectively. Surprisingly, the esters A were found to act as effective initiator/stabilisers for the emulsion polymerisation of styrene in the absence of conventional surfactant. Unfortunately, during the course of the investigation, it was suddenly found to be impossible to prepare further batches of acid chloride I, and most subsequent attempts to prepare acid chloride I met with failure. The efforts which had been made to try to prepare acid chloride I and to elucidate possible structural differences between I and II that might account for the observed phenomena are described in detail. The rest of this work is concerned with the study of the behaviour of esters derived from acid chloride II and a range of fatty-alcohol ethoxylates. Results are given for the use of these derivatives as both initiators and stabilisers for the emulsion polymerisation of styrene in the absence of added conventional surfactant. The effects of temperature, initiator concentration and the total volume of water upon the rate of polymerisation were studied. Also, the effects of these variables upon the shapes of the conversion-versus-time curves obtained when the esters were used as initiator/stabilisers were investigated. The surfactant properties of some of the initiator/stabilisers have also been investigated. The kinetics of the decomposition of these initiator/stabilisers is also included.

A fracture mechanics study has been carried out on the fracture of rubber by cylindrical indentors. Several phenomena have been identified which affect the puncture process. It has been observed that a ring crack forms on the rubber surface before puncture occurs. By treating this crack as a starter crack, an equation has been derived for the puncture energy. The values of puncture energy so obtained were found to agree well with the catastrophic tearing energy obtained from the trouser tear test. The energy stored in the rubber influenced the value of the puncture energy calculated. The energy beneath the indentor was determined using a model experiment based on the biaxial stretching of rubber by inflation. The energy stored in the rubber surrounding the indentor was calculated. The contribution of these energies was shown to be small. The puncture test was used to study the strength of thick rubber blocks which had been exposed to long-term ageing and to high temperatures. These rubber blocks were taken from rubber bearings of 20 and 96 years of age respectively. A study was also carried out on a natural rubber tyre after forty two years immersion in sea water.

Produced by existing size reduction technologies, rubber crumb is the most valuable type of recycled material. However, only a limited amount of untreated crumb can be used in new tyres due to significant property loss. The de-vulcanisation process offers the potential for providing high quality recycled material that can be used as a substitute for virgin rubber. Rubber materials produced during industrial trials that formed the development stages of the novel DeVulCO”2 de-vulcanisation technology, were used in this work. A practical way to characterise de-vulcanised and re-vulcanised waste tyre materials and blends with virgin compound was established. This work showed that simple to obtain and understand parameters, acquired from tests that are commonly used in the rubber industry, such as Mooney viscosity, Mooney relaxation and curing can be successfully used to indicate the efficiency of the de-vulcanisation process and hence, to assist the research and development stages of the de-vulcanisation process and to assure quality control. Good correlation was evident between these parameters and the network characterisation parameter (Vr) obtained from the swelling measurements that are known to be a reliable guide but are more time consuming to perform. In this work, the Mooney relaxation test was successfully extended to characterise some de-vulcanised compounds (automotive weather strip and clothing tape). In addition, the swollen-state 1H-NMR technique was successfully applied to characterise de-vulcanised and subsequently re-vulcanised “closed-loop” tyre materials. A new tool to measure the efficiency of the de-vulcanisation process has been devised, based on a comparison of the peak areas in the olefinic region of rubber extracts analysed by the liquid 1H-NMR technique. This technique is capable of characterising different components of a rubber blend (e.g. NR, BR) and is suitable for carbon black filled compounds. The results of this work have confirmed the potential of the DeVulCO”2 technology and have indicated that up to 20% of de-vulcanisate can be blended with virgin tyre compound without significant deterioration of the key performance properties.