Search

In this dissertation, the design, implementation, and educational applications of a real-time water and weather monitoring system, developed to enhance water sustainability education and research, are discussed. This unique system, called LabVIEW Enabled Watershed Assessment System (LEWAS), is a real- world extension of various data acquisition modules that were successfully implemented using LabVIEW into a freshman engineering course (Engineering Exploration, ENGE 1024) at Virginia Tech. The outdoor site location measures water quality and quantity data including flow rate, pH, dissolved oxygen, conductivity, and temperature -- as indicators of stream health - for an on-campus impaired stream in real-time. In addition, weather parameters (temperature, barometric pressure, relative humidity and precipitation) are measured at the LEWAS outdoor site. The measured parameters can be accessed by remote users in a real-time through a web-based interface for education and research. LEWAS is solar powered and uses the campus wireless network through a high-gain antenna to transmit data to remote clients in real-time. Its power budget consisting of consumption (14 W), electrical storage, and generation (80 W, peak) is balanced to enable 24/7 operation regardless of weather conditions. An embedded computer with low power consumption and modules for communicating and storing data are installed in the field and it is programmed to process measured environmental parameters to be delivered to remote users. This computer is programmed both using a field programmable gate array (FPGA, for low power consumption and robust operation) and traditional microprocessor programming (for more flexibility). The environmental sensors of the system are routinely calibrated using established procedures. A LEWAS Development Platform was established to develop and test the system and to train and mentor several undergraduate and graduate students who helped in its implementation. A number of design and implementation challenges were overcome including extending campus Internet access to a location not included on the network and integrating hardware and software from three different sensor manufacturers into a unified software platform accessible over the Internet. To study the educational applications of LEWAS, an observational study was conducted as the system was gradually introduced to students in ENGE 1024 between 2009 and 2011. Positive student attitudes on the role of LEWAS to enhance their environmental awareness informed an experimental design implemented to study the motivational outcomes associated with the system. Accordingly, appropriate educational interventions and a hands-on activity on the importance of environmental monitoring were developed for both control and experiment groups, with only the latter given access to LEWAS to retrieve the environmental parameters for the activity. An instrument was developed on the theoretical foundation of the expectancy value theory of motivation and was administered to control and experimental groups in ENGE 1024. Altogether, 150 students participated in the study. Exploratory Factor Analysis (EFA) was applied which resulted in factors that group questions together based on interest, importance, real-time access, and cost (feasibility of monitoring). After conducting parametric and nonparametric statistical analyses, it was determined that there exists a statistically significant difference between control and experimental groups in interest, real-time, and cost factors. This finding implies that providing real-time access to environmental parameters can increase student interest and their perception of feasibility of environmental monitoring -- both major components of motivation to learn about the environment. Future extensions and applications of the system at Virginia Tech and beyond are discussed. Ph. D.

Engineering is widely accepted as a creative discipline. However, research focused on assessment of student creativity in engineering studies is lacking. Creativity by definition encompasses both novelty and value and has been approached through investigations of person, product, process and environment. Contemporary socio-cultural theories of creativity recognize the subjective nature of creativity in terms of a person, domain, and field. Domain specific viewpoints recognize the necessity of specialized skills and knowledge beyond the scope of general creativity to attain advanced levels of creative achievement within a given domain. Given the breadth of these theoretical perspectives, the overarching goal of this research was to initiate a foundational understanding of student creativity through a sustainable energy themed engineering design project developed specifically for an introductory engineering course at Virginia Tech. This embedded mixed-methods study approached creativity through assessment as well as significance of correlations between students' creative thinking abilities, their creative performance and their perceptions of the design experience in terms of known creativity factors on the 8-week long open-ended engineering design project. The study is comprised of two research themes; the investigation begins in theme one with a focus on individual creativity leading to investigation of team creativity in theme two with data collected over two successive semesters (i.e., spring â 07 and fall â 07). Creativity assessment measures included a test of creative thinking abilities, the Abbreviated Torrance Test for Adults (ATTA) as well as application of the Consensual Assessment Technique (CAT), a widely used creativity measurement technique, with graduate student judges assessing creativity of student's brainstorming ideas and final designs. Student surveys were administered for each theme to assess students' perceptions of the design project. Results reveal a significant correlation (p < 0.05) between students' ATTA scores and perceptions of the project as open-ended and interesting. Additional factors significantly correlated (p < 0.05) with students' perceptions of the design experience as creative included finding the project meaningful, interesting, exciting, enjoyable and surprising. Results of inter-rater reliability analysis followed by a two-way ANOVA illustrated difficulties in establishing consistency of judges' assessment of technical brainstorming ideas and final designs. Theme one findings with students repeatedly discussing the importance of their team in the creative process on the design project were further substantiated in theme two, bringing fresh insights concerning the role of collaboration in student creativity in the introductory engineering course. Implications for future research begin with the necessity for establishing reliable judging criteria and training as well as determining appropriate judges used in the assessment of students' creative performance on the design project. This research provides an essential starting point for researchers interested in developing engineering design curriculum to foster creativity. Ph. D.

The structure of education has changed dramatically in the last few decades. Despite major changes in how students are learning, there has not been as dramatic of a shift in how student learning is assessed. Standard letter grades are still the paradigm for evaluating a student's mastery of course content and the grade point average is still one of the largest determining factors in judging a graduate's academic aptitude. This research presents a modern approach to modeling knowledge and evaluating students. Based upon the model of a closed-loop feedback controller it considers education as a system with an instructor determining the set of knowledge he or she wishes to impart to students, the instruction method as a transfer function, and evaluation methods serving as sensors to provide feedback determining the subset of the information students have learned. This method uses comprehensive concept maps to depict all of the concepts and relationships an educator intends to cover and student maps to depict the subset of knowledge that students have mastered. Concept inventories are used as an assessment tool to determine, at the conceptual level, what students have learned. Each question in the concept inventory is coupled with one or more components of a comprehensive concept map and based upon the answers students give to concept inventory questions those components may or may not appear in a student's knowledge map. The level of knowledge a student demonstrates of each concept and relationship is presented in his or her student map using a color scheme tied to the levels of learning in Bloom's taxonomy. Topological principles are used to establish metrics to quantify the distance between two students' knowledge maps and the distance between a student's knowledge map and the corresponding comprehensive concept map. A method is also developed for forming aggregate maps representative of the knowledge of a group of students. Aggregate maps can be formed for entire classes of students or based upon various demographics including race and gender. XML schemas have been used throughout this research to encapsulate the information in both comprehensive maps and student maps and to store correlations between concept inventory questions and corresponding comprehensive map components. Three software packages have been developed to store concept inventories into an XML Schema, to process student responses to concept inventory questions and generate student maps as a result, and to generate aggregate maps. The methods presented herein have been applied to two learning units that are part of two freshman engineering courses at Virginia Tech. Example student maps and aggregate maps are included for these course units. Ph. D.

Designers of engines have always looked for ways to improve the power to weight ratio of mobile internal combustion engines. This was especially true in aircraft engine design and engines for various forms of racing. Today designers are looking for ways to make everything from cars to road tractors to farm tractors lighter and thereby more efficient. In addition, in many cases these vehicles only need the maximum power that an engine can produce for a small amount of time. What is needed is a small, lightweight engine with the ability to produce a large amount of power for a short duration. The work here describes one possible method for constructing just such a type of engine. By adding a combustion chamber in the exhaust flow between the engine exhaust valve and the turbine inlet on a turbocharged diesel engine, it should be possible to increase the turbine temperature. This will in turn allow the turbine to deliver more power to the compressor and create a higher inlet pressure and allow the engine to create more power. This paper describes both a computer simulation and an engine with this combustion chamber installed. There were however, problems with both the simulation as well as the test engine. While no quantitative data was obtained from the test engine, some valuable observations were made. The computer simulation yielded results and from these results and observations made while testing the engine with the combustion chamber installed it was determined that this design shows promise of creating an engine with higher specific power. Master of Science

The assessment of the shear strength of soil deposits after the occurrence of large strains is an important issue for geotechnical engineers. One method for doing so, the steady state approach, is based on the assumption that the steady state undrained shear strength is a unique function of the in situ void ratio and effective stress. This method, which has been applied to liquefaction and flow failures, has been criticized because it may overestimate the in situ shear strength. The key to the steady state approach is accurate determination of the relationship between void ratio and effective stress at steady state. This is typically accomplished using conventional isotropically consolidated undrained (ICU) triaxial tests. The triaxial test was developed for measuring peak strengths, which typically occur at small strains, but steady state conditions typically occur at much larger strains. At large strain levels, the suitability of conventional triaxial testing procedures and error corrections is uncertain. The measured response at large strains may be inaccurate due to the influence of various testing errors. Furthermore, the true material response in the test specimen at large strains may not accurately represent in situ material behavior at large strains. This research effort consisted of an experimental and analytical study to examine the validity of steady state undrained shear strength determination using conventional ICU triaxial tests. The analytical study addressed triaxial testing errors and conventional corrections that are applied to test data and their influence on the measured steady state parameters. Finite element analyses were conducted to investigate the influence of variations in restraint at the end platens on stress distributions in the sample and measured stress-strain response. The finite element analyses incorporated axisymmetric interface elements to model the friction characteristics between the end platens and the specimen ends. The experimental study focused on several sands that are susceptible to liquefaction. An interface direct shear test program was conducted in order to evaluate various schemes for reducing end platen friction. ICU triaxial tests were conducted on each material using both conventional and lubricated end platens. Ph. D.

Curved panels constructed of laminated graphite-epoxy composite material are of potential interest in airframe fuselage applications. An understanding of structural response at elevated temperatures is required for anticipated future high speed aircraft applications. This study concentrates on the response of unstiffened, curved composite panels subjected to combinations of thermal and mechanical loading conditions. Mechanical loading is due to compressive end-shortening and thermal loading is due to a uniform temperature increase. Thermal stresses, which are induced by mechanical restraints against thermal expansions or contractions, cause buckling and postbuckling panel responses. Panels with three different lamination sequences are considered, including a quasi-isotropic laminate, an axially soft laminate, and an axially stiff laminate. These panels were chosen because they exhibit a range of stiffnesses and a wide variation in laminate coefficients of thermal expansion. The panels have dimensions of 10 in. by 10 in. with a base radius of 60 in. The base boundary conditions are clamped along the curved ends, and simply supported along the straight edges. Three methods are employed to study the panel response, including a geometrically nonlinear Rayleigh-Ritz solution, a finite element solution using the commercially available code STAGS, and an experimental program. The effects of inplane boundary conditions and radius of curvature are studied analytically, along with consideration of order of application in combined loading. A substantial difference is noted in the nonlinear load vs. axial strain responses of panels loaded in end-shortening and panels loaded with uniform temperature change, depending on the specific lamination sequence, boundary conditions, and radius of curvature. Experiments are conducted and results are presented for both room temperature end-shortening tests and elevated temperature tests with accompanying end-shortening. The base finite element model is modified to include measured panel thicknesses, boundary conditions representative of the experimental apparatus, measured initial geometric imperfections, and measured temperature gradients. With these modifications, and including an inherent end displacement of the panel present during thermal loading, good correlation is obtained between the experimental and numerically predicted load vs. axial strain responses from initial loading through postbuckling. Ph. D.

General unsymmetric laminates exhibit large natural curvatures at room temperature. Additionally, inherent to most unsymmetric laminates is the presence of two stable configurations. Multiple configurations and stability issues arise because of the geometric nonlinearities associated with the large curvatures. The laminate can be changed from one stable configuration to the other by a simple snap-through action. This situation offers the opportunity to use shape memory alloys (SMA) attached to the laminate to generate the snap-through forces and change the shape of the laminate on command. Presented is a model which can predict SMA-induced deformations in general unsymmetric laminates and, particularly, the occurrence of the snap through. First, a methodology is developed to predict the deformations of flat general unsymmetric epoxy-matrix composite laminates as they are cooled from their elevated cure temperature. Approximations to the strain fields are used in the expression for the total potential energy, and the Rayleigh-Ritz approach is used to study equilibrium. To further study the laminate deformations, finite-element analyses are performed. Experimental results are presented which confirm the predictions of the developed theory and the finite-element analyses regarding the existence of multiple solutions and the magnitude of the deformations. Results are compared with those of several other investigators. Next, the deformation behavior of general unsymmetric laminates subjected to applied forces is studied. The principle of virtual work is used to derive the equilibrium equations relating the laminate deformations to the applied forces. By solving the equilibrium equations as a function of the force level, relations between the laminate deformations and the applied force are derived, and the force level at which the laminate changes shape is determined. Finally, an existing SMA constitutive model is implemented into the developed theory to predict the deformations of simple structures to SMA-induced forces. Experiments on a narrow aluminium plate with an externally attached SMA actuator are conducted. The experimental results show good agreement with the predictions from the developed theory. Next, the deformation behavior of general unsymmetric laminates subjected to SMA actuators is predicted using the developed theory. Experiments using SMA actuators to generate the snap through of nsymmetric laminates are conducted. Good correlation with the developed theory is obtained. Ph. D.

A finite element based model for interrogation of the elastic and progressive failure response of textile based composite materials has been developed and implemented. The geometrically thorough textile geometry model is used as the basis for creation of the geometry used for the finite element analysis. A detailed assessment is made for the linear elastic response of a plain weave textile composite subjected to various boundary conditions and loadings. Hashin's failure criterion and a maximum stress criterion are employed to predict the failure of the yam and matrix material, respectively. A material degradation scheme is implemented to account for the decrease in the elastic constants of the material. The initiation and progression of failure within a plain weave textile composite has been predicted. Ph. D.

In the present research project, four braided composite architectures consisting of graphite fibers in an epoxy matrix were tested under static and fatigue loading conditions to determine damage mechanism types and progressions. The braided architectures consisted of straight axial fiber bundles, which were surrounded by braider fiber bundles oriented at ±a 0 with respect to the axial fiber bundles. Static tension and compression testing was completed first to determine material strengths and basic damage modes for each of the architectures. Under static tension loading, cracking in the braider fiber bundles occurred first, and was followed by splitting in curved regions of the axial fiber bundles. Matrix cracking and kink band formation were found to occur under static compression loading. Ph. D.

Accurate models are necessary in critical applications. Key parameters in dynamic systems often change during their life cycle due to repair and replacement of parts or environmental changes. This dissertation presents a new approach to update system models, accounting for these changes. The approach uses frequency domain data and a neural network to produce estimates of the parameters being updated, yielding a model representative of the measured data. Current iterative methods developed to solve the model updating problem rely on minimization techniques to find the set of model parameters that yield the best match between experimental and analytical responses. Since the minimization procedure requires a fair amount of computation time, it makes the existing techniques infeasible for use as part of an adaptive control scheme correcting the model parameters as the system changes. They also require either mode shape expansion or model reduction before they can be applied, introducing errors in the procedure. Furthermore, none of the existing techniques has been applied to nonlinear systems. The neural network estimates the parameters being updated quickly and accurately without the need to measure all degrees of freedom of the system. This avoids the use of mode shape expansion or model reduction techniques, and allows for its implementation as part of an adaptive control scheme. The proposed technique is also capable of updating weakly nonlinear systems. Numerical simulations and experimental results show that the proposed method has good accuracy and generalization properties, and it is therefore, a suitable alternative for the solution of the model updating problem of this class of systems. Ph. D.

A hybrid method is presented for the analysis of layers, plates, and multi-layered systems consisting of isotropic and linear elastic materials. The problem is formulated for the general case of a multi-layered system using a total potential energy formulation and employing the layerwise laminate theory of Reddy. A one-dimensional finite element model is used for the analysis of the multi-layered system through its thickness, and integral Fourier transforms are used to obtain the exact solution for the in-plane problem. Explicit expressions are obtained for the fundamental solution of the typical infinite layer, which are applied in the two-dimensional boundary integral equation model to produce the integral representation of the solution. The boundary integral equation model is two-dimensional, displacement-based and assumes piecewise continuous distribution of the displacement components through the system's thickness. The developed model describes the three-dimensional displacement field, the stress field, the strains and the interlaminar stresses over the entire domain of the problem as continuous functions of the position. This detailed three-dimensional analysis is achieved by incorporating only contour integrals. The boundary integral equations are discretized using the boundary element method and a numerical model is developed for the single numerical layer (element). This model is extended to the case of a multilayered system by introducing appropriate continuity conditions at the interfaces between the layers (firmly bonded layers, or separation, slip and friction between the layers). Assembly of the element matrices yields the global system of equations, which can be solved via iterative techniques. In addition, numerical techniques are developed for the evaluation of the boundary and domain integrals involved in the construction of the element matrices. The singular boundary integrals are computed using a special coordinate transformation, along with a subdivision of the boundary element and a transformation of the Gauss points. The domain integrals (regular, singular or near-singular) are transformed to regular definite integrals along the boundary through a semi-analytical approach. The proposed method provides a simple, efficient, and versatile model for a three-dimensional analysis of thick plates or multilayered systems. It can also be used to study plates resting on elastic foundations or plates with internal supports. The proposed method can be applied in an obvious manner to anisotropic materials and vibration problems. Ph. D.

When discrete piezoelectric actuator patches bonded on structures are used for active shape, vibration, and acoustic control, the desired deformation field in the structure is obtained through the application of localized line forces and moments generated by expanding or contracting bonded piezoelectric actuators. An impedance-based model to predict the dynamic response of cylindrical shells subjected to excitation from surface-bonded induced strain actuators is presented. The essence of the impedance approach is to include the actuator/structure impedance ratio in the calculations of the actuator forces applied to the structure, which will retain the dynamic characteristics of the actuators. The appropriate representation of the loading due to in-phase and out-of-phase actuation is discussed. Due to the curvature of the shell, the representation of the in-phase actuation with an equivalent in-plane line force applied along the edge of the actuator results in the application of erroneous rigid-body transverse force. To avoid these rigid body forces, the action of the actuator needs to be represented by am equivalent in-plane force and a transverse distributed pressure applied in the region of the actuator patch to maintain the structure self-equilibrium. A full derivation of the impedance model is included, taking great care in the structural and actuator impedance definition. It is found that the actuator's output dynamic force in the axial and tangential direction are not equal. Various case studies of a cylindrical thin shell are performed to illustrate the capabilities of the developed impedance model. The in-phase and out-of-phase actuation authority of induced strain actuators bonded to the surface of a shell is compared. It is shown that out-of-phase actuation has better authority in exciting the lower order bending modes, while in-phase actuation has better authority in exciting the higher order circumferential modes. Dynamic finite clement analysis has been performed using piezoelectric elements available in ANSYS 5.0. The good correlation between the finite clement results and the impedance model confirms the analytical solution. Experimental data of a circular ring actuated in-phase and out-of-phase by a piazoelectric material (PZT) were also compared to the derived impedance model. Ph. D.

The objective of this study was to develop a two-dimensional computer model for the simulation of the resin transfer molding/resin film infusion processing of advanced composite materials. This computer simulation model is designed to provide aircraft structure and tool designers with a method of predicting the infiltration and curing behavior of a composite material component. For a given specified cure cycle, the computer model can be used to calculate the resin infiltration, resin viscosity, resin advancement, heat transfer within the component/tool assembly during processing and preform compaction. Formulations of the resin flow problem are given using the finite element/control volume technique based on Darcy's Law of flow through porous media. This technique allows for the efficient numerical calculation of the advancing resin front within the preform materials. The heat transfer in the fabric preform and tooling is analyzed using a transient finite element method which included the effects of convection on the tooling surfaces. Compaction behavior of the tooling assembly is analyzed using a simplified isotropic form of the plane elasticity equations. All of these solutions were coupled together in a quasisteady state non-linear fashion inside the computer code. Master of Science

A detailed study of the feasibility of anchoring fiber reinforced plastic wires for civil engineering applications is presented. An experimental investigation using tensile testing machines is performed, testing anchorages of 1, 7 and 19 wires. Conventional strain gage and acoustic emission measurement techniques are used. The tests are essential in determining the failure load and failure mode. However, the experimental data alone do not provide enough information about the behavior of the anchorage to be used exclusively in the design process. The results are used to modify the design of the anchorage system. It is shown that for a successful anchorage system the choice of material for the load transfer medium is crucial. A solution is presented to overcome the high stress concentration at the load entry area of the wire into the anchor. A finite element analysis of the anchors for 1 and for 19 wires is used to assess the stress and strain fields inside the anchor, to validate the analytical model, and to determine locations of possible high stress concentrations. Three-dimensional and one-dimensional models, that utilize axisymmetry, are evaluated. The results of the numerical analysis are used to demonstrate the improvement as a result of a change in material choice or design of the anchor. It is shown that the modification of the load transfer medium results in a decrease of 30 % of the average stress level. In the analytical investigation, several common models are introduced that describe the fiber pullout behavior. Based on a recent treatment by C. H. Hsueh, a model is developed that describes the anchorage of a wire in a conical shaped socket using orthotropic materials. This model includes boundary conditions that are similar to the ones observed in the experiments. A parametric study is performed to obtain information on the ideal geometry of the anchor system. The results and predictions of the applied techniques, i. e. analytical description, finite element method and experimental investigation, are compared and contrasted. Based on the analytical, numerical and experimental results, recommendations for improving the design of the anchor system are made. Subsequently, a modified anchor system is proposed that utilizes the properties of a load transfer medium that has a variable stiffness. The inclusion of a pure resin collar and supporting wires is suggested. For a successful completion of this project, ideas are proposed and suggestions made for future work. Ph. D.

A micro mechanical finite element model to compute the overall instantaneous stiffness of fiber-reinforced composites in elastic-plastic response is presented. The model is applicable to a periodic diamond array of elastic circular fibers embedded in an elastoplastic matrix subjected to a plane stress loading. This model enforces symmetry and anti-symmetry conditions isolating the smallest unit cell and should greatly increase the speed of doing "built-inn micromechanics within a larger finite element program because of the small number of degrees of freedom (12 to 14 d.o.f.). The matrix plastic behavior is modeled using the endochronic theory without a yield surface. Various off-axis elastoplastic characteristics predicted by the mini grid for a boron/aluminum composite are presented. Comparison with experimental data and a fine grid finite element solution shows very good agreement and demonstrates the effectiveness of the mini model presented. Master of Science

This study looks at the feasibility of using structural optimization techniques to address the problem of designing composite fuselage frames for crashworthiness. A key feature of any optimization strategy for increasing structural crashworthiness is a progressive failure analysis. Currently, the most widely used analysis methods for progressive failure of composite structures are considered too expensive computationally for practical optimization in today's computing environment. Developing an efficient analysis method for progressive failure of composite frames is a first step in the optimization for crashworthiness. In the current work a progressive failure analysis for thin-walled open cross-section curved composite frames is developed using a Vlasov type beam theory. A curved thin-walled composite beam theory is developed and a finite element implementation of the beam theory is used for progressive failure analysis. The accuracy and limitations of this analysis method are discussed. A model for progressive failure of the composite fuselage frame is developed from an extension of the laminate progressive failure analysis of Tsai-Wu. Comparisons based on a limited amount of available experimental data are encouraging. The first major failure event is captured by the theory, and the prediction of total energy absorbed follows the trend of the experimental data. It is believed that this accuracy is sufficient for preliminary design and optimization for crashworthiness. This progressive failure analysis is then incorporated into a frame optimization for crashworthiness based on the genetic algorithm method. The optimization methodology is demonstrated analytically to obtain frame designs with substantially increased crashworthlness. Laminate stacking sequence and cross-section shape are design variables for optimization Ph. D.

Shape memory alloy (SMA) materials such as nitinol have unique properties associated with the shape recovery effect and the material’s phase changes that have been used in a variety of actuator and sensing applications. By embedding SMA elements into host composite materials, control or modification of the SMA hybrid composite’s structural properties can be accomplished in-service, thereby increasing the hybrid composite’s structural functionality. Previous studies addressed increasing composite materials’ functionality by enabling in-service control of their dynamic response. Utilizing the SMA’s substantial recovery stress and capacity to dissipate strain energy to increase the hybrid composite’s static functionality is addressed herein. Specific applications for SMA hybrid composites include improving composite material’s impact damage resistance and composite cylinder stress and deflection control. In stress and deflection control of cylindrical structures, SMA actuators are placed within the composite cylinder to form an active compound cylinder. The active SMA elements can significantly reduce the internal pressure-induced radial dilation and creep so that under severe loading, piston to cylinder tolerances may be maintained. Similar to a conventional metallic compound cylinder, the active compound cylinder also reduces peak cylinder hoop stresses. Hybridizing composites with nitinol improves their impact resistance because of nitinol’s tremendous capacity to absorb impact strain energy through the stress-induced martensitic phase transformation. The amount of impact damage is reduced and the material’s resistance to impact perforation at various velocities is improved. The experimental response of nitinol hybrid composites and the associated mechanics are presented. The unique toughness and resistance to permanent deformation that is a result of the stress-induced martensitic phase transformation enables the nitinol to absorb on the order of 4 times the strain energy of high alloy steel and 16 times that of many graphite/epoxy composites. In most static applications where SMA elements are used for reinforcement, maintaining the integrity of the interface between the SMA elements and the host polymeric matrix composite material is critical to operation. The relationship between preparation of SMA elements for hybrid composite fabrication and interfacial bond strength is presented to address this issue. The mechanics of interfacial shear failure between SMA element and composite is also presented. Ph. D.

Minimum mass designs are obtained for a simple box beam structure subject to bending, torque and combined bending/torque load cases. These designs are obtained subject to point strain and linear buckling constraints. The present work differs from previous efforts in that special attention is payed to including the effects of subcomponent panel interaction in the optimal design process. Two different approaches are used to impose the buckling constraints. When the global approach is used, buckling constraints are imposed on the global structure via a linear eigenvalue analysis. This approach allows the subcomponent panels to interact in a realistic manner. The results obtained using this approach are compared to results obtained using a traditional, less expensive approach, called the local approach. When the local approach is used, in-plane loads are extracted from the global model and used to impose buckling constraints on each subcomponent panel individually. In the global cases, it is found that there can be significant interaction between skin, spar, and rib design variables. This coupling is weak or nonexistent in the local designs. It is determined that weight savings of up to 7% may be obtained by using the global approach instead of the local approach to design these structures. Several of the designs obtained using the linear buckling analysis are subjected to a geometrically nonlinear analysis. For the designs which were subjected to bending loads, the innermost rib panel begins to collapse at less than half the intended design load and in a mode different from that predicted by linear analysis. The discrepancy between the predicted linear and nonlinear responses is attributed to the effects of the nonlinear rib crushing load, and the parameter which controls this rib collapse failure mode is shown to be the rib thickness. The rib collapse failure mode may be avoided by increasing the rib thickness above the value obtained from the (linear analysis based) optimizer. It is concluded that it would be necessary to include geometric nonlinearities in the design optimization process if the true optimum in this case were to be found. Master of Science

Accurate mechanical properties are critical to the design and use of composite material structures. Due to the available processing methods, the properties of ceramic matrix materials are especially sensitive to the geometry of the component and how it is made. A method is presented by which the ply-level elastic properties of a composite material can be obtained for a common structure; a thick, laminated tube. The mechanical and thermal response of the tubes is modeled by a planar cylindrical elasticity solution. Properties are determined from surface strain measurements of a thick tube subject to axial, torsional, pressure, and thermal loads. All elastic properties (including thermal expansion coefficients) can be obtained except the out-of-plane shear moduli (G13, G23) which are not involved in the planar elasticity solution employed. The ply-level properties are estimated by inversion of the elasticity solution in terms of the global strain measurements. A Least Squares optimization approach is used for the inversion of the elasticity solution. Application of the method for a filament wound aluminum oxide-aluminum oxide tube is presented. Advantages and limitations of the method are identified. Master of Science

A numerical and experimental investigation of the bending behavior of six eight-ply graphite-epoxy circular cylinders is presented. Bending is induced by applying a known end-rotation to each end of the cylinder, analogous to a beam in bending. The cylinders have a nominal radius of 6 inches, a length-to-radius ratio of 2 and 5, and a radius-to-thickness ratio of approximately 160. A [±45/0/90]S quasi-isotropic layup and two orthotropic layups, [±45/02]S and [±45/902]s, are studied. A geometrically nonlinear special-purpose analysis, based on Donnell's nonlinear shell equations, is developed to study the prebuckling responses and gain insight into the effects of non-ideal boundary conditions and initial geometric imperfections. A geometrically nonlinear finite element analysis is utilized to compare with the prebuckling solutions of the special-purpose analysis and to study the buckling and postbuckling responses of both geonletrically perfect and imperfect cylinders. The imperfect cylinder geometries are represented by an analytical approximation of the measured shape imperfections. Extensive experimental data are obtained from quasistatic tests of the cylinders using a test fixture specifically designed for the present investigation. A description of the test fixture is included. The experimental data are compared to predictions for both perfect and imperfect cylinder geometries. Prebuckling results are presented in the form of displacement and strain profiles. Buckling end-rotations, moments. and strains are reported, and predicted mode shapes are presented. Observed and predicted moment vs. end-rotation relations, deflection patterns. and strain profiles are illustrated for the postbuckling responses. It is found that a geometrically nonlinear boundary layer behavior characterizes the prebuckling responses. The boundary layer behavior is sensitive to laminate orthotropy, cylinder geometry, initial geometric imperfections, applied end-rotation, and non-ideal boundary conditions. Buckling end-rotations, strains, and moments are influenced by laminate orthotropy and initial geometric imperfections. Measured buckling results correlate well with predictions for the geometrically imperfect specimens. The postbuckling analyses predict equilibrium paths with a number of scallop-shaped branches that correspond to unique deflection patterns. The observed postbuckling deflection patterns and measured strain profiles show striking similarities to the predictions in some cases. Ultimate failure of the cylinders is attributed to an interlaminar shear failure mode along the nodal lines of the postbuckling deflection patterns. Ph. D.

Inner ear hair cells have been identified as the sites of mechanoelectrical transduction from a mechanical event (e.g. hearing, motion) to an electrical event (e.g. neural response). Deflection of bundles of hair-like stereocilia extending from these cells has been associated with the transduction process. Stereocilia bundle structure and stiffness controls deflection and thus the fundamental sensitivity of the transduction process. The finite-element method was used along with analytical techniques to characterize individual stereocilium and stereocilia bundle stiffnesses. A three ‘stack’ bundle with a Young’s modulus of 3 GPa (F-actin protein) and Poisson’s ratio of 0.4 (nearly incompressible) resulted in a stiffness of K = 2.1 x 10⁻³ N/m. This value is within the range of experimentally determined stiffmesses. Tip-link and subapical band interconnecting structures each contribute significantly to bundle stiffness and each could act as the gating-spring in transduction models, which propose gating structures as a means of regulating ionic activity and therefore neural activity. Stiffness depends most strongly on individual stereocilium geometry and material description, tip-link orientation and material description, and stereocilia bundle width. Stiffness depends least on stereocilia height variations and subapical bands configuration. Linear analysis was reliable up to deflections of 3.5 um, the upper limit of physical response. Preliminary dynamic response indicates a natural frequency of 382 kHz for the vibration mode resembling physical deformation behavior. Future models should include hexagonal bundle arrangements, transversely isotropic stereocilia material descriptions, and viscoelastic tip-link behavior. Master of Science

Presented is a model for predicting SMA-induced deformations in an unsymmetric cross ply laminate. A previously developed theory is used to predict the room-temperature shape of the cross-ply laminate by minimizing its total potential energy. Then, using the principle of virtual work, equations relating the shape of the laminate to a force applied on supports fastened to the laminate are derived. Induced strains and displacements are predicted as a function of the applied force. Experiments where the force is generated by known weights are conducted. Good correlations are established between the experimental results and the predictions. The developed theory is able to predict with good accuracy the shape, strains and, displacements of an unsymmetric cross-ply laminate to the force applied on the laminate. This theory is then used to develop a model relating the laminate response to forces produced by a SMA actuator, the actuator being a SMA wire. To describe the mechanics of the SMA actuator, constitutive equations derived by other researchers are used. These constitutive equations relate the temperature of the wire to forces generated in it. Experiments where a SMA wire is used as an actuator are conducted. These experiments consist of resistively heating a SMA wire attached to supports fastened to the laminate. During these experiments, laminate deformations are measured as a function of the applied voltage. Comparisons with the temperature-based constitutive model predictions are not made since the relation between the applied voltage and the SMA temperature is very difficult to establish. However, the experiments show that a SMA used in conjunction with cross-ply unsymmetric laminates can induce very large changes in the laminate shapes. Thus, the concept of using a SMA actuator to control the shape of cross-ply unsymmetric laminates is validated. Master of Science

The influence of crenulated noncircular fibers on the micromechanical stress states due to a transverse strain and to a temperature change in carbon/carbon composites is examined using the finite element method. Stresses at the interface of both fully bonded and fully disbonded fibers having two crenulation amplitudes and with two fiber volume fractions are presented. In each case, these interface stresses are compared to stresses at the interface of circular fibers which have the same degree of disbond and fiber volume fraction and are under the same loading conditions. For the disbonded cases, deformed meshes showing locations of fiber/matrix contact are also included. In addition to the interface stress states, selected composite properties are also computed and compared in each case examined. Interest in studying noncircular fibers stems from a desire to increase the transverse properties of carbon/carbon by introducing a mechanical interlocking between the fiber and the matrix. Results presented here indicate that this interlocking does in fact occur. Evidence from the interface stress data suggests, however, that any possible advantage of this interlocking may be outweighed by the disadvantage of stress concentrations which arise at the interface due to the crenulated geometry of the fibers Master of Science

A micromechanics-based model for progressive failure analysis of laminated composite shell structures has been developed and implemented. The model couples a finite element model for nonlinear analysis of laminated shells with a micro mechanics elasticity solution for predicting failure and effective composite properties. The nonlinear laminate theory and finite element model are based on a third-order expansion of displacements through the thickness of the shell, thus allowing for both transverse normal and shearing deformations. The initiation and evolution of damage in the shell is modeled at the constituent (i.e., fiber, matrix, interphase) level using an elasticity solution for quasi-three-dimensional hygro-thermo-mechanical loading of continuous fiberreinforced composites. The model is used to predict failure in laminated composite structures and to study the effects of constituent properties (stiffness and strength) on structural behavior. Ph. D.

The influence of layer waviness in thick cross-ply composite cylinders subjected to hydrostatic pressure is investigated. The cylinders considered are graphite-epoxy with a 2: 1 ratio of circumferential to axial layers. All cylinders considered contain 104 total layers with a layup of [90/(90/0/90h71s, where a '0° 1 layer is taken to be in the axial direction. The influence of a single isolated group of wavy layers in an otherwise perfect cylinder is evaluated. Layer waviness in only the circumferential direction is considered, and the analysis is assumed to be valid only away from the cylinder ends. A parametric investigation is performed to determine the combined influence of wave location, wave amplitude, and cylinder geometry on hydrostatic response of the cylinder, particularly the stresses generated in and around the wave. The wave is assumed to be located either at the inner or the outer radius of the cylinder. Three wave amplitudes, 0, are considered: 1/2, 1, and 2 layer thicknesses. Only waves with a half wave length of 10 layer thicknesses are considered. Three cylinder geometries are considered, specifically ones with radius to thickness ratios of 5, 10, and 20. Finite element analysis is used to determine the stress state within the imperfect, i.e., wave included, cylinders. Based on a maximum stress failure criterion, failure pressures are determined for each of the various wave and cylinder geometries. Failure pressures for the imperfect cylinders are compared with those for a perfect cylinder to determine the failure pressure reduction ratios due to fiber waviness. It is shown that pressure capacity reductions of approximately 50% are possible for the range of parameters studied. Failure is primarily due to fiber compression, though interlaminar shear and interlaminar tension are a factor. Finite element analysis is also used to deter ine the failure pressure of the perfect cylinder due to buckling. This is done to determine whether failure due to buckling may overshadow material failure due to fiber waviness. It is shown that buckling is a factor in only one of the cylinder geometries considered, and only in the cases of mild layer waviness. In addition to results, details about the finite element model are presented. These details include geometry of the wave, changes in material properties due to local fiber rotation and local volume fraction changes, boundary conditions, and justifications for modeling simplifications that were made in an effort to reduce computational costs and analysis times. Master of Science

The ability to predict failure of laminated composites in compression has been doggedly pursued by researchers for many years. Most have, to a limited extent, been able to predict failure for a narrow range of laminates. No means, as yet, exist for predicting the strength of generic laminates under various load conditions. Of primary concern has been the need to establish the mode at failure in compression. Even this has been known to vary for fiber and matrix dominated laminates. This study has been carried out to analyze the failure of specimens with a hole made of laminates with various quasi-isotropic stacking sequences. Different stacking sequences are achieved by rotating a [±45/90/0]s stacking sequence laminate as a whole with respect to the loading axis of the specimens. Two- and three-dimensional finite element models, using commercial packages, were generated to evaluate the stresses in the region of the hole. Two different compressive failure prediction techniques based on distinctly different failure modes have been used. The validity of these techniques was measured against experimental data of quasi-isotropic specimens tested. To investigate the applicability of the failure criteria for different laminated composite plates, analyses were repeated for specimens with different stacking sequences resulting from the rotation of the laminate. The study shows the need for the use of three-dimensional analysis of the stress state in the vicinity of the hole in order to be able to accurately predict failure. It also shows that no one mode of failure is responsible for limiting the strength for all laminate orientations but rather the mode changes with change in stacking sequence. The failure of the laminate with a hole was seen to be very sensitive to the stacking sequence. Experimental data presented also shows that the peak strength obtainable from the laminate analyzed, [±45/90/0]s, is going to be in the off-axis configuration rather than on-axis placement of the stacking sequence with respect to the loading direction. Master of Science

A computationally efficient analysis approach is developed to predict buckling load of geodesically stiffened composite panels under in-plane loads. The analysis procedure accounts for the contribution of the in-plane extensional and out-of-plane bending stiffnesses of the stiffeners through the use of Lagrange multipliers in an energy method solution. The analysis is used to isolate the effect of various stiffener deformation modes on the buckling load and skin deformation patterns of geodesically stiffened panels under various load combinations. The analysis routines are then coupled with the numerical optimizer ADS to create a package for the design of minimum-mass stiffened panels, subject to constraints on buckling of the panel assembly and material strength failure. Material failure in the skin and stiffeners are estimated using a maximum strain criterion. The design variables that can be used for optimization include thickness of the skin laminate, stiffener thickness and height, and positions of straight stiffeners. Applied loads are uniaxial compression, pure shear, and combined compression-shear. Master of Science

This study reports on the development and application of the peninsula blister test to quantitatively measure the adhesion of various adhesively bonded joints. Analytical results reveal that this peninsula-like geometry benefits from both a constant strain energy release rate over the major portion of the debond length and a high strain energy release rate at any given pressure. Applications of this technique to several adhesion systems were conducted. Although some of these systems haven't been successfully tested due to various reasons, experimental results from systems of PSA tapes and thin polyimide films bonded on aluminum substrates were promising. The agreement of the bond strength in terms of strain energy release rates obtained from both experimental and analytical methods from the last two systems indicates the feasibility of this technique. Primary studies on the stress analysis for several thin film adhesion tests suggest that the high ratio of strain energy release rate to applied pressure offered by this modified blister geometry may not be able to overcome the tensile strength limitations of thin film adhesion testing. Further studies need to be conducted in order to understand if the strain energy release rate can be raised without the increase of membrane stresses by altering specimen geometries. In conclusion, although this modified blister is not an universal adhesion test for every adhesion system, the attractive nature of the constant strain energy release rate produced by the peninsula blister specimen warrants further investigations and wider applications. Master of Science

While the study of compression after impact of laminated composites has been under consideration for many years. the complexity of the damage initiated by low velocity impact has not lent itself to simple predictive models for compression strength. The damage modes due to non-penetrating. low velocity impact by large diameter objects can be simulated using quasistatic three-point bending. The resulting damage modes are less coupled and more easily characterized than actual impact damage modes. This study includes the compression testing of specimens with well documented initial damage states obtained from three-point bend testing. Compression strengths and failure modes were obtained for quasi-isotropic stacking sequences from 0.24 to 1.1 inches 'thick with both grouped and interspersed ply stacking. Initial damage prior to compression testing was divided into four classifications based on the type. extent, and location of the damage. These classifications are multiple through-thickness delaminations, isolated delaminations. damage near the surface. and matrix cracks. Specimens from each classification were compared to specimens tested without initial damage in order to determine the effects of the initial damage on the final compression strength and failure modes. A finite element analysis was used to aid in the understanding and explanation of the experimental results. It was found that specimens with multiple through-thickness delaminations experienced the greatest reduction in compression strength, from 50 to 75% below the strength of undamaged specimens. All the sublaminates formed by the delaminations failed at the same time. Individual sublaminate buckling was observed for isolated delaminations near 'the surface of the laminate. Delaminations far from the specimen surface had little effect on the final compression strength. Damage occurring in the outside 00 plies caused a 10 to 200/0 strength reduction according to both analytical and experimental results. The effects of increased interlaminar stresses near the specimen edges caused a reduction in undamaged strength of [05/455/-455/905]55 specimens, while having little effect on the [Osl60sl-605]75 specimens. Master of Science

The strain induced into isotropic flat plates by surface bonded and embedded piezoceramic patches was investigated. The free piezoceramic patch voltage/strain response was characterized using moire interferometry and Twyman-Green interferometry techniques. Moire interferometry experiments were conducted to investigate the induced strain of actuation in a flat plate with two surface bonded piezoceramic patches and a flat plate with one embedded patch. Three-dimensional finite element models of these two specimens were constructed and analyzed. The measured in-plane voltage/strain response of a free piezoceramic patch was found to be nearly linear and of a much higher magnitude than the published data. Similarly, the out-of-plane voltage/strain response was found to be much higher in magnitude than the published data as well as being non-linear. Correlation between the moire experiment concerning the surface bonded patch specimen and the corresponding finite element model was good, indicating the validity of the modeling technique. Possible depoling damage of the embedded patch caused the correlation between the experiment and analysis of this specimen to be poor. Both finite element models predicted stresses that are not statically significant but may represent problems over many loading cycles. Master of Science

The feasibility study of subscale notched specimens to determine the shear response of composites is presented. The investigation consists of finite element analyses, conventional strain-gaged testing, and photomechanics experiments. Several notch geometries of the subscale specimens are studied, the standard 900 V -notch, U-notch, and circular notch. The investigation is performed on two different material systems, a standard high performance graphite/epoxy (AS4/3501-6) material and an SMC R-28 material reinforced with 28% volume fraction strand glass fiber. The moduli obtained from the subscale specimens are compared with those obtained from the standard specimens. Different degrees of twisting observed in testing the sub scale specimens are discussed. Numerical and experimental results of the SMC R-28 materials are presented. The dependence of the measured shear modulus on the relative orientation of the specimen in the panel is identified. The application of the subscale, circular notched specimens to obtain the shear modulus of the SMC material is discussed. Master of Science

Two comprehensive composite fabrication simulation computer codes have been written. These codes when coupled together have the capability to model the filament winding and curing processes for composite structures. The "Filament Winding Cure" (FWCURE) code is a thermo-kinetic model. FWCURE models the resin viscosity, percent of cure, temperature, resin flow, and layer location. As these characteristics change, they have an effect on the fiber tension within the composite. The "Winding and Curing Stress Analysis Finite Element" (WACSAFE) code models the filament winding process and predicts manufacturing stresses and strains based on material properties, lay-down tension and wind angle. The permeability model in FWCURE requires constants that are found experimentally. The WACSAFE code requires an input tension that is equivalent to the initial spool tension minus the instantaneous tension losses. The permeability constants and the instantaneous tension losses were found experimentally. The codes were then used to predict fiber tension, tension losses and mandrel strains for experimental test cylinders. The predictions were compared to test data. Master of Science

An optimization system based on the finite element code CSM Testbed and the optimization program ADS is described. The optimization system can be used to obtain minimum-weight designs of composite stiffened structures. Ply thicknesses, ply orientations, and stiffener heights can be used as design variables. Buckling, displacement, and material failure constraints can be imposed on the design. The system is used to conduct a design study of geodesically stiffened shells. For comparison purposes, optimal designs of unstiffened shells and shells stiffened by rings and stringers are also obtained. Trends in the design of geodesically stiffened shells are identified. An approach to include local stress concentrations during the design optimization process is then presented. The method is based on a global/local analysis technique. It employs spline interpolation functions to determine displacements and rotations from a global model which are used as "boundary conditions" for the local model. The organization of the strategy in the context of an optimization process is described. The method is validated with an example. Ph. D.

Braided composites find many diverse applications in modern technology and tailoring the mechanical properties of these structures has become increasingly important. This thesis will examine one class of circular braids encompassing an elastic core. By hypothesizing four modes of operation and incorporating primary influences, the mechanical response of the composite is predicted based on its initial parameters and material properties. The ability to model the yarns that constitute the braid as nonlinear materials enables the simulated response to span finite deformations. A scheme for nondimensionalizing the model parameters and governing equations for each mode of operation is also proposed and implemented. In an effort to validate the assumptions underlying the model's formulation, a series of experimental trials are documented that verify the fundamental braid mechanics. A wide variety of analytical cases are also introduced to investigate the influences of various model parameters. Possible extensions for the existing model are also noted. Master of Science

Interlaminar stresses are estimated from plate theories by equilibrium. The elasticity equations of equilibrium are integrated with respect to the thickness coordinate z using the linear distribution in z of the in-plane stresses. This procedure, for example, requires fourth order derivatives of the out-of-plane displacement w with respect to the in-plane coordinates x and y to compute the interlaminar normal stress. Since compatible elements for the plate bending problem at most require the displacement and its first derivatives to be continuous across element boundaries, low degree interpolation polynomials are used. Thus, fourth order derivatives of the finite element polynomials are either meaningless, or at least inaccurate. In order to compute high order derivatives, an approximate polynomial solution of high degree to the governing partial differential equation for w(x,y) is determined using the finite element solution as a first approximation. A rectangular subdomain that may consist of several elements is selected from the finite element model. The displacement w(,y) over the subdomain is expanded in a Chebyshev series. Then collocation is used to determine the unknown Chebyshev coefficients such that the Chebyshev series matches displacement w and its normal derivative from the finite element solution at discrete points on the boundary of the subdomain, and the partial differential equation is enforced at discrete points within the subdomain. Interlaminar shear and normal stresses are computed from the third and fourth derivatives, respectively, of the Chebyshev series at the collocation points. Master of Science

The feasibility of replacing currently widely used sandwich construction with helically stiffened construction for a Formula racing car structural shell is studied. The torsional deformation behavior of circular and square cross-section shells is analyzed as an approximation to the real car structure. Shells with different sandwich and helically stiffened configurations are analyzed with finite elements. For closed square and circular cross-section shells, the highest torsional stiffness is obtained with helical stiffening. For circular cross-section shells with the cockpit cutout and reinforcements usually present in real racing car structural shells, ±4S' helically stiffened shells are 200% stiffer in torsion than sandwich shells. For square cross-section shells, the torsional stiffness improvement obtained with the helical stiffening is only 27%. The cross-sectional shape of the shell, cockpit opening, and different type of reinforcements (present in a real car structure) affects the selection of the best stiffening for torsional stiffness. The role of the terms of the stiffness matrix of the helically stiffened configuration in the torsional behavior of the shells is studied. The 0/90 waffle stiffening is more efficient than the helical stiffening for the square cross-section shells with the cutout and reinforcements. In the case of circular cross-section shells, the 0/90 waffle stiffening yields approximately the same results than the helical stiffening. The skin-stiffener configuration for maximum torsional stiffness depends on the crosssectional shape of the shell. The advantages of the ±45° helical stiffening over the sandwich construction depend on the cross-sectional shape of the shell and on the way the cutout region is reinforced. Master of Science

Shape memory alloy (SMA) hybrid adaptive composites are a class of materials which combine the strain recovery and elastic properties transformation capabilities of SMA fibers with the structural characteristics of advanced composite materials. This study utilizes the Rayleigh-Ritz method and finite panel acoustic radiation theory to investigate the use of SMA hybrid composite materials for adaptive structural acoustic control by active structural tuning. Analytical models are formulated considering classical laminated plate theory (CLPT) and first-order shear deformation theory (FSDT), to predict modal and structural acoustic response to incident low frequency plane wave acoustic excitation. The analysis is further developed to consider simply supported adaptive panels that are tuned by local fiber activation, such that a panel composed of elastically uniform sections can be evaluated in a piece-wise fashion. Master of Science

Actuator placement is a major concern in control system designs. Utilizing piezoelectric actuators increases the complexity of actuator designs, because both actuator location and dimensions need to be considered. A comprehensive study was conducted in this dissertation on the optimization of piezoelectric actuator designs for vibration suppression of flexible structures. The investigation on the optimal piezoelectric actuator designs were grouped into two parts. Part one covered actuator designs when the same number of actuators as the controlled modes are used. Approaches were formed to optimally design piezoelectric actuators which requires least control efforts. In part two of this dissertation, a method named the Weighted Pseudoinverse Method was introduced to deal with the cases in which fewer actuators than the controlled modes are utilized. The weighted pseudoinverse method yields a optimal transformation from modal control forces into the actuator-moments in physical space. Based on the Weighted pseudoinverse method, the piezoelectric actuator designs were optimized to ensure least-control-effort actuator designs. A simply-supported beam was used as an example to demonstrate the effectiveness of the design methods proposed in this dissertation. However, the design methods are applicable to general cases. Ph. D.

New materials and manufacturing processes, and the quest for economy and user safety, have necessitated the development of nondestructive testing methods to quantify the life and reliability of a product during manufacture and service. Described herein, is a testbed to be used in the research and development of these testing methods. A brief motivation for using ultrasonics applied to nondestructive evaluation is followed by a chapter on the feasibility of using a unique testing method and animated data presentation on advanced composite materials. This testing method, conceived by the author, utilizes oblique injection of ultrasound into the specimen. Several cycles of the ultrasonic waveform radiated from the specimen downstream of the injection area is digitized and recorded. The data has three independent dimensions; cartesion location and time. The time variable is the key to the presentation of the data as an animated two dimensional image. It was this work that illustrated the need for a flexible scanning imaging research testbed, not only for the discussed method, of which it is an integral part, but for advanced development of other techniques. Software development and integration of off -the-shelf parts into a unified computer controlled testing facility is the contribution by the author in the second phase of this research. Chapters on the description of the system, an example showing the capabilities of the system analogous to traditional ultasonic C-scanning, accomplishments, and a look to the future conclude this thesis. The appendices include listings of the programs developed for the system, a manufacturer address list. A videotape of the animation data presentation is included as a second volume of this thesis. Master of Science

This study developed and analyzed a modification of the blister test permitting nearly constant strain energy release rate testing of adhesive bonds. The work consisted of three parts; (1) development of the testing technique to evaluate strain energy release rate and to record the time dependent nature of the fracture process, (2) numerical analysis of the constrained blister test to determine the applicability of an approximate solution for several materials, and (3) development of an. analytical technique to evaluate the strain energy release rate for relatively stiff specimens. Master of Science

Composite materials are considered here as structural materials of highway bridge superstructures. Bridge deck designs can be done according to AASHTO¹ specification and elastic design concepts. In order to evaluate the feasibility of composites as structural materials of highway bridge superstructures, composite materials are compared not only to composite materials themselves but also to the most popular bridge structural materials, which are reinforced concrete and structural steel. The AASHTO¹ HS2O-44 truck load is selected as the standard loading condition of all designs. Loads other than dead load and live load are not considered. Configurations of the bridges are different. Appropriate cross-section of girders are selected according to the material types. For fiber reinforced composite materials, box girder is used, for reinforced concrete, T-beam is selected; in addition, steel concrete composite section is another case. Design methods are different from material to material. Reinforced concrete T-beam design is based on the 'Ultimate Strength Design' method. Steel concrete composite sections are designed according to the 'Load & Resistance Factor Design'. For composite materials, 'Elastic Design' is selected. The results derived are as expected. Substantial weight saving is achieved by simply replacing concrete or steel with composite materials. This also results in many other advantages such as construction time, cost, foundation settlement and support requirements. Master of Science

A Global-Local finite element approach was used to investigate the interlaminar stresses in laminated composite plates with a central circular hole. Detailed solutions were sought for the interlaminar normal stress distributions close to the free straight edge of the plate as well as around the edge of the hole. The Global model was analyzed as a two-dimensional problem. The displacements obtained a distance away from the regions of interest in the two-dimensional model were used as imposed boundary conditions to the three-dimensional models of the edges. The results obtained were found to be accurate, thus demonstrating the validity and strength of the Global-Local technique. The results further concluded that for symmetric cross-ply laminated plates with large central circular holes, the interlaminar normal stresses at the free edges are affected to a small degree by the size of the hole. The CSM Testbed and ANISAP were the two finite element analysis programs used throughout this investigation. The CSM Testbed element library was augmented with 16, 20, 24 and 32 node displacement formulation based elements which were implemented as Experimental elements. Master of Science

A first-order shear deformation plate theory is used to predict free vibration frequencies in rotating laminated composite plates. The theory accounts for geometric non-linearity in the form of von Karman strains. The plate is permitted to have arbitrary orientation and offset from the axis of rotation. A finite element model is developed to obtain a solution to the problem. The model is validated by comparing the results for free vibration of non-rotating plates for various boundary conditions and material properties with the exact results based on the classical plate theory and the first-order shear deformable plate theory. Results are presented for free vibration of isotropic and laminated composite plates rotating at different angular velocities. A study has also been made on the change in the free vibration frequencies of the plate with angular velocity for different plate thicknesses and for different modulus ratios. Master of Science

Adhesives and adhesively bonded structures are being considered as a viable alternative to conventional fastening methods. In order to gain wider acceptance, it is essential to address the issue of the mechanical characterization of adhesive materials and its implementation in the design of bonded joints. While measuring the in-situ properties of the adhesive in a joint is a difficult task, characterizing its bulk properties is a relatively simpler undertaking. The objective of this study was to propose and verify an experimental procedure that would allow the analytical prediction of the viscoelastic behaviour of a bonded joint, using bulk adhesive properties. The Arcan joint geometry was chosen because of the simple state of stress within the adhesive. Master of Science

In support of crashworthiness studies of composite airframes, the present study was undertaken to understand the large deflection flexural response and failure of graphite-epoxy laminated beams. The beam specimens were subjected to eccentric axial impact loads and to static eccentric axial loads, in order to assess the damage caused by impact. A geometrically and materially nonlinear analysis of the response and failure of the static test specimens is presented. The analysis employed an incremental, noniterative finite element model based on the Kantrovich method and a corotational solution technique. Width-wise effects are included by assuming specific forms of the displacements across the width, with length-wise variation introduced as a degree of freedom. This one-dimensional, 22 degree of freedom finite element accurately predicted the load-deflection and strain-deflection responses of the static test specimens. Inclusion of nonlinear material behavior was found to be important in correctly predicting load-deflection response of uniaxial materials, while inclusion of width-wise effects was determined to be more important for laminates with off-axis plies due to the existence of coupling between bending and twisting curvatures (D16and D26). Once material nonlinearity begins to occur in flexure, even symmetric laminates exhibit bending-stretching coupling due to different material response in tension and compression. Master of Science

The need for better impact resistant composites has resulted in the development of many toughened resin systems. A combination of a tougher resin system along with higher strength fibers increases the impact resistance of the composite. The use of an adhesive layer between two plies of the improved prepreg system has been found to considerably increase the impact resistance. This concept is known as "Interleafing." This investigation studies the response of the interleaf materials to instrumented drop weight impact as compared with the response of non-interleaved materials. Two non-destructive quality evaluation techniques, namely, ultrasonics and eddy currents, are used to qualitatively evaluate the damage developed in the specimens. Several different energy levels of damage are studied. The interleaved laminate had significantly better impact response than the non-interleaved laminate for the same impact energy. The onset of delamination was delayed by the use of the interleaf. Whereas damage could be detected at an impact energy as low as 1.75 ft-lb in the baseline laminate; the interleaved laminate did not show any ultrasonic C-scan indication up to an impact of 2.45 ft-lb. The increase of delamination with increasing impact energy was slower in the interleaved specimen. The eddy current method is not effective in detecting damage in the interleaved laminate because of the shielding effect of the interleaf. Compression Strength After Impact (CSAI) could not be used for the test laminates in this project, but the Tensile Strength After Impact test provided useful results. The tensile strength after impact of the interleaved specimen was between 20%-80% more than the baseline laminate up to impact energy of 10 ft-lb. The advantage of the interleaved specimen reduced at higher energy levels of impact. Master of Science

An investigation of the behavior of shear deformable plate finite elements is conducted to determine why and under what conditions these elements lock, or become overly stiff. For this purpose, a new analytical technique is developed to derive the exact form of the shear constraints which are imposed on an element when its side-tothickness ratio is large. The constraints are expressed in terms of the nodal degrees of freedom, and they are easily interpreted as being either the proper Kirchhoff constraints or spurious locking constraints. Moreover, the technique is applicable to any displacement-based shear deformable beam, plate or shell element regardless of the shear deformation theory or the order of the Gauss-Legendre integration rule which is used to numerically evaluate the stiffness coefficients. To gain a better understanding of locking phenomena, the constraints which arise under full and reduced integration are derived for various Mindlin and Reddy-type beam and plate elements. These analytical findings are then compared with numerical results of isotropic and laminated composite plates, verifying the role that shear constraints play in determining the behavior of thin shear deformable elements. The results of the present study lead to definitive conclusions regarding the origin of locking phenomena and the effect of reduced integration. Master of Science

Experiments were conducted to measure the three-dimensional static and free vibrational response of two graphite-epoxy, thin-walled, open section frames. The frames are semi-circular with a radius of three feet, and one specimen has an I cross section and the other has a channel cross section. The flexibility influence coefficients were measured in static tests for loads applied at midspan with the ends of the specimens clamped. Natural frequencies and modes were determined from vibrational tests for free and clamped end conditions. The experimental data is used to evaluate a new finite element which was developed specifically for the analysis of curved, thin-walled structures. The formulation of the element is based on a Vlaysov-type, thin-walled, curved beam theory. The predictions from the finite element program generally correlated well with the experimental data for the symmetric I-specimen. Discrepancies in some of the data were found to be due to flexibility in the â clamped' end conditions. With respect to the data for the channel specimen, the correlation was less satisfactory. The finite element analysis predicted the out-of-plane response of the channel specimen reasonably well, but large discrepancies occurred between the predicted in-plane response and the experimental data. The analysis predicted a much more compliant in-plane response then was observed in the experiments. Master of Science