Thesis (Ph. D.)--University of Washington, 1999. / Vita. Includes bibliographical references (leaves 139-150).

Thesis (Ph. D.)--University of Wisconsin--Madison, 1983. / Typescript. Vita. Description based on print version record. Includes bibliographical references (leaves 209-218).

Th. doct. ing.--Besançon, 1978. N°: 88.

Une nouvelle approche intitulée «Théorie Variationnelle des Rayons complexes est appliquée au calcul de vibrations forcées de plaques faiblement amorties dans le domaine des moyennes fréquences. Pour chaque fréquence de la bande considérée, la résolution d'un système linéaire de petite dimension donne les grandeurs effectives (énergie cinétique, intensité vibratoire) ; cette approche ne fait pas intervenir une discrétisation fine «éléments finis» de la structure. Le premier élément caractérisant cette approche est l'exploitation d'une nouvelle formulation variationnelle du problème à résoudre qui a été élaborée afin d'autoriser des approximations non compatibles, c'est à dire ne vérifiant pas a priori les conditions de transmission aux interfaces entre les sous-structures, tant en déplacement qu'en contrainte. Ces conditions sont contenues dans la formulation variationnelle. Le deuxième élément définissant cette approche est l'introduction d'approximations à fort contenu mécanique : la solution est supposée bien décrite localement au voisinage d'un point, comme la superposition de modes de vibration locaux en nombre infini. Ces modes de base, ou «rayons complexes», vérifient les lois de la dynamique.

This dissertation is an analytical study that compares the South African Transport Services (SATS) and Eurocode (EC) live load models for railway bridges. The study is specifically concerned with the critical load effects of shear and bending moment. The load models are simulated as moving loads over the full length of simply supported and continuous railway systems with speeds not exceeding 180km/h. The study is limited to short to medium spans ranging from 5m – 40m analysed in increments of 5m. The position of the maximum load effects for simply supported systems was determined using the frame analysis module in Prokon. Maximum load effects were determined using the influence line method. Maximum load effects for the continuous systems were determined using the moving load option in STRAP. It was found that SATS live load models imposed on single span railway bridges, produce conservative load effects for short span bridges but become over conservative with an increase in span, when compared with characteristic values of the EC load model 71 (LM71). For heavy loads (α = 1,10) in LM71, there is a good comparison with that of the EC for static and design moment (for a track with standard maintenance) with values of 5% lower at 10m but become moderately conservative (2% - 5%) with an increase in span. In the case of design bending moment (for a carefully maintained track) the SATS code is moderately conservative (6% - 8%) over the full range of spans for a carefully maintained track. For heavy loads (α = 1,10) in LM71, there is a good comparison with that of the Eurocode for static and design shear (for a carefully maintained track) with values of 4% lower at 10m but becoming moderately conservative (1% - 5%) with an increase in span. In the case of design shear (for a track with standard maintenance) the SATS code compares well with that of the EC, with values of 5% lower at 10m but becoming moderately conservative (4% - 13%) with an increase in span. Live traffic loads imposed on equal span (limited to 2) continuous railway bridges, produce conservative static and design shear load effects (for a carefully maintained track) in the mid-range of spans but become moderately conservative with increase in span for heavy loads (α = 1,10) for load model SW/0. There is a good comparison with that of the EC for design shear force (for a carefully maintained track) with moderately conservative (1% - 9%) for short span and long span systems for heavy loads (α = 1,10) for load model SW/0. A similar comparison occurs for heavy loads (α = 1,21) for SW/0 for static and design shear for a carefully maintained track. Live traffic loads imposed on equal span (limited to 2) continuous railway bridges produce over conservative static bending moment load effects for short span and long span bridges (2 x 30m – 2 x 40m) for characteristic values and heavy loads (α = 1,10 and α = 1,21) for load model SW/0. Generally, there is not a good comparison with that of the EC for static and design bending moment, for two span continuous railway bridges. Live traffic loads imposed on equal span (limited to 3) continuous railway bridges produce moderately conservative static shear force effects for heavy loads (α = 1,10 and α = 1,21) for load model SW/0. The only significant value is at the 3 x 5m span (21% higher) and the 3 x 15 – 3 x 20m range of spans (9% - 10% lower) for heavy loads (α = 1,10) and (α = 1,21) respectively. A similar comparison is observed for design shear effects for both types of track for heavy loads (α = 1,10) and (α = 1,21) for a carefully maintained track. Generally, there is not a good comparison with that of the Eurocode for static and design bending moment, for three span continuous railway bridges.

Abstract Salt water environments are very harsh on materials that are used within them. Many issues are caused by either corrosion and/or internal degradation to the materials themselves. Composites are better suited for this environment due to their high strength to weight ratios and their corrosion resistance, but very little is known about the fracture mechanics of composites. The goal of this study is to gain a better understanding for the behavior of a composite boat hull under a shear loading, similar to the force water applies on the hull as the boat moves through the water; then attempt to strengthen the composite sandwich panel against the shear loading. A parametric study was conducted to investigate monotonic in-plane shear loading for composite sandwich panels used in commercial naval vessels. In order to model a conventional composite boat hull, test specimens were composite sandwich panels made of a Divinycell H100 foam core with four layers of fiberglass on both sides of the core. Specimens were tested under a monotonic loading with a rate of 0.2 in/min, and tested until complete failure using the standard test. Seawater specimens were manufactured in the same manner as the original test specimens, but then were submersed in either filtered seawater or the ocean. The differences between the filtered pieces and the ocean allowed us to determine if any changes found in the composite sandwich panels were related to environment conditions or if the changes were related to the saltwater interaction itself. To create these different environments the seawater specimens were taken to the Avila pier where 36 specimens were placed in a tub that was fed filtered saltwater, while 30 specimens were placed in a plastic mesh with weights and lowered to a depth of approximately 30 ft. in the ocean. Three specimens were then removed at monthly intervals from both filtered and ocean environments. Shear Keys were created as a method to strengthen the composite sandwich panels against the shear force that the previous specimens had been tested to. Eight Shear Keys were then placed into groves cut into the foam core (four on each side) and the four fiberglass layers were laid on top. Testing showed that the seawater did have an initial effect on the composite sandwich panels. The filtered pieces showed a decrease in yield strength and stiffness the longer they were subjected to the seawater. The raw unfiltered pieces placed in the ocean saw an even higher decrease in their yield strength and decrease in stiffness. However, for both the unfiltered and raw specimens there was an increase in the ultimate strength and fracture point of the specimens. The effects of the sea water seemed to taper off after the 3rd month however. The Shear Key specimens were tested with a 4mm and an 8mm Shear Key. The 8mm Shear Keys showed a decrease in shear strength, which was primarily due to removing too much material from the core and weakening the specimen. It was concluded that the decrease in area created a force concentration at the deepest part of the Shear Key causing the premature failure. The 4mm Shear Key showed an increase in the yield strength, ultimate strength, and fracture point. A finite model was built to simulate the original test specimen along with the 4mm and 8mm Shear Key cases, and the results were compared to the experimental results. The numerical results showed that it was possible to relate the experimental results to the linear or elastic portion of the plots. There was a difference between the maximum displacement of the model and the actual specimens, but this was attributed to potential inaccurate comparison of the loading on the model compared to the actual specimens. The correlation between the model itself and the experimental data was close enough to conclude that it could be used for predicting baseline trends. Further investigation of the specimens should include looking into the effects of a cyclic shear loading on the specimens. This combined with the seawater element used in this thesis would provide further insight to the initial degradation seen in the seawater specimens, and could potentially provide a closer relation to current hull failures. In addition to including a cyclic loading another numerical model should be created. A model that could be constrained both locally and globally would provide more accurate results. The FEM should also include the ability to run a crushable foam core model within the solver which would also increase the accuracy of the numerical solution.

Over the last three decades, the aerospace industry has gradually shifted from metals to composites in many different applications due to the lightweight properties of composite materials. Some benefits of composites include higher strength-to-weight ratio and corrosion resistance. At this point in time, the composite industry researchers are focusing on renewable and sustainable materials (bio-composites). By understanding the structural capabilities of bio-composites that have been used for centuries, new developments of sustainable materials will spark more interest throughout the industry. Bio-composites include fibers such as hemp, bamboo, flax, etc. The high demand for bio-composites in composite structures can also reduce raw material costs. This study investigated, through experimental and numerical analysis, the mechanical behavior of sandwich panels under edgewise compressive loading. The first task of the study was to use four different facesheet materials and the same Nomex honeycomb core. The number of facesheet layers consecutively increased from one layer to four layers on each side of the core for each material. The facesheet materials used were Hexply AGP280-5H Carbon Fiber Pre-Preg, B601 Plain Weave Hemp, D118DKBR Split Herringbone Weave Hemp, and NB308T 7725 Texalium Fiberglass Pre-Preg. The sandwich panels were cured using a composite heat press and followed the recommended cure cycle for the material’s resin matrix. The variation of the facesheet materials while keeping the core consistent showed how the edgewise strength and displacement of the composite sandwiches were affected under compressive loading. The second task of the study was to try and create a multifunctional hybrid composite sandwich with two different facesheet materials; using one hemp material and one pre-preg material. The goal of this task was to try and minimize the damage occured upon failure. Being that the pre-preg materials are more brittle than the hemp material, the hybrid composite sandwiches can potentially create a superior composite structure. The sequence of stacking of the facesheet materials was manipulated to study how changing the outer and inner layers affected the results. All the specimen were loaded at a rate of 0.05 in/min in a steel jig specifically made per ASTM C364 standard using an Instron 8801 to determine the mechanical behavior. These experimental results combined with results from theoretical and finite element analysis using Matlab and Abaqus, respectively, were used to compare composite sandwich designs under compressive loadings. Failure mode comparison between the individual material composites and the hybrid composites were also discussed.

The goal of this effort was to build an analysis tool for aircraft wings and incorporate it into a program that optimizes the wing structure for the lowest weight for the conceptual design phase. The analysis tool calculates the internal stresses in primary load carrying members using established analysis methods for semimonocoque beams. Via a Graphical User Interface (GUI) built in MATLAB, the user can define the structural layout and material properties of the load carrying members. The program requires a degenerate geometry model built with Vehicle Sketch Pad (OpenVSP) to define the outer mold line (OML) of the wing, and a section of the program in MATLAB calculates the geometric parameters of the wing structure based on the model and user inputs in the GUI. Another section generates a lift curve using a Schrenk distribution, the vehicle weight, and load factors as defined by the user. The GUI also allows the user to define other external loads in addition to the aerodynamic loads. With the loads and structural model defined, the program uses the analysis tool to find a minimum structural weight while maintaining positive structural margins for all structural members. The analysis tool was compared against examples in structural analysis books from Bruhn and Peery to validate the method. The average relative difference between the normal and shear stresses calculated by the tool and the sources was 1.6%. To test the program, a Cessna 210G wing was modeled in the program and using Finite Element software. The comparison showed the tip deflection of the MATLAB model was 1.4 times that of the Finite Element Model. When the areas of the structural members were multiplied by 1.4, the normal stress in the stiffeners had an average difference of 5.8 ksi and the shear stresses in the webs had an average difference of 0.33 ksi. The program estimated the weight to be 198 lbs, underestimating the weight when compared to other existing methods.

Note:

No description available.