Thesis (M.S.) in Mechanical Engineering--University of Maine, 2002. / Includes vita. Includes bibliographical references (leaves 96-97).

Thesis (Ph. D.)--University of Washington, 2002. / Vita. Includes bibliographical references (leaves 125-133).

Commonly encountered accidental impact, e.g. due to roadway stone hitting, is detrimental not only because it can produce apparent surface defects, but also because barely visible impact damage (BVID) can be induced inside the material, which is not easy to detect by routine inspection. Reliable prediction of the amount of damage of this type induced under known service conditions is particularly important. Therefore, this type of impact was chosen as the focus of the present investigation. A combination of experimental techniques and finite element modelling was used to explore the behaviour of a fibre reinforced natural gas powered vehicle (NGV) pressure cylinder subjected to a low energy impact. In order to identify the modes of failure and understand the structural response, quasi-static indentation tests were carried out on sections of composite pipes and of a composite pressure cylinder. Delamination and matrix cracking were established to be the two major failure modes induced by indentation. Experimental findings were used as a basis for assessing the validity of the modelling approach. Thick shell and three dimensional finite element models were developed using PAFEC, a general purpose finite element code for dynamic and static analysis. It established that the composite pressure cylinder under this type of impact behaves quasi-statically, i.e. the impact phenomenon predominately excites low frequency response. Repeated impact was considered in order to extend the study to include the impact behaviour of a cylinder with pre-existing damage. It was found that a bulging effect was produced in the pressure cylinder at the impact site, where a weak spot was created due to fibre breakage. A fully three dimensional finite element model with static analysis was developed to investigate the damage and material degradation during the BVID phenomenon. The contact pressure distribution based on the Hertzian contact' relationship was applied. Failure mode identification criteria proposed by Hashin (1980) and Chang and Springer (1986) were used to establish the mode and extent of damage in the composite cylinder under quasi-static loading. The predicted failure modes agreed well with the experimental results. Finally, the present study sets out the methodology allowing systematic design of structures having optimal impact tolerance. Based on the findings of this project, suggestions for the improvement of impact resistance of NGV cylinders were given in Chapters ix.

The corrosion of reinforcing steel has emerged as the most prevalent factor causing deterioration of reinforced concrete structures in aggressive exposures, such as the Middle East and Arabian Gulf States. Recently, the effect of using embedded steel bars passing through different types of concrete has been found to exacerbate the corrosion activity. This may occur with a steel bar passing from a foundation manufactured with sulphate resistance Portland cement (SRPC) into a beam or column manufactured with ordinary Portland cement (OPC). This thesis aims at elucidating the behaviour of reinforcing steel (RC) embedded in concrete mixes of different types by incorporating both OPC and SRPC and blends with ground granulated blast furnace slag and pulverised fuel ash. Two different types of RC specimens were used: Single bars in a single mix and electrically connected bars in same or different mix types sharing a common electrolyte in the absence and presence of a specific interface. The RC specimens were exposed to wet-dry cycles of chloride and sulfate solutions. The results of steel bars that were electrically connected, and were embedded in concrete cubes manufactured with different cement types and without a direct interface exhibited reduced times to onset of corrosion when subjected to chloride ion ingress. In the worst case, the connection of the a bar in OPC (0.45) concrete and one in SRPC (0.45) concrete caused the bar in the SRPC (0.45) mix to start corroding after only 3 cycles (6 weeks) compared to 15 cycles (30 weeks) for a single unconnected bar in the same SRPC (0.45) concrete. Moreover, measurements of the half-cell potential, polarisation resistance and corrosion current, all confirm that a single bar passing the interface between two different concretes manufactured with different cement types can suffer the formation of anode and cathode regions that depend on the local environment. For example, a steel bar passing from an OPC (0.45) concrete to a SRPC (0.45) concrete will preferentially form an active anode within the SRPC concrete. This has implications for concrete structures manufactured using SRPC foundations and OPC superstructure, as often used in the Arabian Gulf States

A novel technique for the assessment of dispersion quality in polymeric composites containing high levels of mineral filler or mixtures of different fillers is described. A comprehensive survey of the literature indicates that current methods of assessment using a combination of microscopy and image analysis are reliant on object specific measures which require good contrast between filler particles and the surrounding polymeric matrix. Such methods are inappropriate for the investigation of highly filled systems due to the presence of touching particles and poorly defined feature boundaries. The method described in this report uses characteristic X-rays or backscattered electron (BSE) imaging for the visualisation of filler dispersion patterns. The X-ray technique is especially useful for the discrimination of individual fillers in a multicomponent system. Multifractal analysis of X-ray or BSE images provides a quantitative description of dispersion quality using a numerical indicator derived from the width of the multifractal spectrum. The technique is successfully demonstrated on a series of flame retardant filler-EVA composites containing different amounts of ATH and Mg(OH)2. Linear models were constructed to illustrate the dependence of composite tensile strength and elongation at break on compound formulation and dispersion quality. Optimum material properties were observed for composites displaying good dispersion quality and predicted mechanical properties were compared against experimentally measured data. A positive correlation was found between real test data and predicted values. This technique provides a unique method for the dispersion assessmenot f mineral filled systems and can be performed on any SEM equipped with an EDX spectrometer and access to a PC running image analysis software. It is a useful addition to the current procedures available for the measurement of dispersion quality and should prove invaluable for the study of filled polymer systems.

No description available.

A comparison was made between a number of laboratory and in-situ concrete permeability test methods. The laboratory tests used measured air, water, and water vapour permeability, whilst the in-situ tests used were the Initial surface absorption test, the Figg air and water tests, and a new in-situ method called the Egg test; a non-destructive surface test which measures air permeability properties. An initial set of tests were carried out on six concrete mixes with water: cement(w/c) ratios between 0.3 and 0.8. These tests showed that problems existed with both the laboratory and in-situ test methods. Some of these problems arose from the preparation of specimens or test procedures and these were overcome with practice or by modifying the test methods. However, it was found that a major problem is moisture in the concrete which decreases it's measured permeability. Further tests were carried out on a second set of concrete specimens with the same mix proportions as the first and a set of mortar specimens with w/c ratios ranging from 0.3 to 1.1 and cement: sand (c: s) ratios from 1: 1 to 1: 5. Results from tests on oven dry specimens were used to compare the different methods and showed that few simple relationships existed between the different methods. Comparing the test results with the mix proportions showed that in the majority of cases, the measured permeability values increased as the w/c ratio increased, but the relationships between the tests results and c: s ratio were more complicated. After these tests had been completed, specimens from twenty six of the mixes were retested after being conditioned to various different moisture contents. The results of these tests showed that in most cases there was a rapid increase in measured permeability as the specimens dried, followed by a slower increase (in some cases a decrease) as the specimens approached an oven dry condition. To complement this study a number of methods were examined for measuring in-situ moisture content. The most promising of these was a non-destructive method which operated by measuring the electrical permittivity of the material it was placed against. Because the electrical permittivity varies with the amount of water in the concrete, it is largely independent of the type material being tested. Results from this test showed a shallow linear drop from saturated to approximately half of the saturated moisture content, followed by a steep drop towards the oven dry condition.

No description available.

No description available.

No description available.