Finite Element Modeling of Ballistic Impact on a Glass Fiber Composite Armor Dan Davis Experiments measuring the ballistic performance of a commercially available fiberglass armor plate were used to guide the development of constitutive laws for a finite element model of the impact. The test samples are commercially available armor panels, made from E-glass fiber reinforced polyester rated to NIJ level III. Quasi-static tensile tests were used to establish material properties of the test panels. These properties were then used to create models in the explicit finite element code LSDYNA. Ballistic impact testing of the panels was conducted using a compressed gas gun firing spherical steel projectiles oriented normal to the test panel surface. The V50 ballistic limit of these panels was found to be approximately 560 m/s. Tuning parameters in the finite element models were adjusted to match the experimentally measured penetration depths and ballistic limits. Models were created in LSDYNA by adjusting the available material library types 3 and 59 for the target, and material type 15 for the projectile. Type 3 models are isotropic, and resulted in shear punch-out type failures of the plate that poorly replicated the test results. Type 59 takes orthotropic properties into consideration, and can analyze delamination when used with solid elements. Results with model type 59 were significantly better than those using type 3, however, this model was found to vastly underestimate the impact resistance of the plate. With significant adjustments to the material properties in the type 59 model, the LSDYNA simulations were found to better replicate the experimentally observed response of the panels. However, these deformations are questionable since they required quite unrealistic adjustments to the material properties.

Composite lattice structures known as the IsoBeam™ made with unidirectional carbon/epoxy were manufactured and tested in shear-dominated bending. The manufacturing process consisted of placing tows of carbon fiber pre-impregnated with epoxy resin onto a pin-type mandrel to create members with interwoven joints. The members were consolidated with a half spiral aramid sleeve. The IsoBeam structure consists of two main types of members: longitudinal and diagonal members measuring nominally 0.4 in. (10.2 mm) and 0.2 in. (5.1 mm) in diameter, respectively. The hand-manufactured specimens measured nominally 6 in. (152.4 mm) high by 3 in. (76.2 mm) wide by 2 ft (0.61 m) long with 4 bays, each 6 in. (152.4 mm) long. The beams weighed between 1.82-1.86 lbs (8.09-8.27 N). A finite element analysis of the IsoBeam was compared to the experimental results. The IsoBeam specimens were tested in four-point or three-point bending but were dominated by shear due to short-beam bending because of the low length/height aspect ratio. After testing to failure, individual members that were lightly loaded and appeared to be undamaged were removed and tested in axial compression. The void percentage and fiber volume fraction were also measured. The average maximum strength of the IsoBeam structure was 4.11 kips (18.3 kN), yielding an equivalent shear of 2.06 kips (9.15 kN) and bending moment of 20.2 kip-in (2.29 kN-m). This strength was lower than expected and is attributed primarily to low material quality, insufficient consolidation of members, and inadequate tension on the tows during manufacturing. The structure exhibited ductile behavior absorbing considerable energy after initial failure, as well as exhibiting damage tolerance due to the inherent structural redundancy. The inner diagonal members which are inherently stiffer exhibited higher strains than the side outer diagonal members after initial failure. The members removed and tested exhibited an average compression strength of 86.9 ksi (599 MPa) and compression modulus of 17.8 Msi (122 GPa) which are both lower than observed in members tested in past research. The diagonal members had a higher strength of 111 ksi (767 MPa) than the longitudinal member's compression strength of 62.5 ksi (431 MPa). Most members were seen to have a high percentage of voids with an average of 4.3% for diagonal members and 6.4% for longitudinal members. The average fiber volume fraction content of members was very low at 38%. The linear finite element analysis of the IsoBeam structure predicted failure at a load of 34 kips (151 kN). Without considering buckling, the first member predicted to fail was a vertical outer diagonal. This research demonstrates that increasing the manufacturing quality should yield an IsoBeam structure that is strong, ductile and damage tolerant.

The enhancement of node projection techniques in sweeping is the subject of this thesis. Sweeping is a method used to produce all-hexahedral finite element meshes on certain classes of geometry. The placement of nodes in the interior of the geometry during the sweeping process remains a difficult problem. This thesis presents advancements in this area which improve the speed of the algorithm and the resulting element quality. A comparison of existing projection methods was performed. The existing Faceted projection sweeping method was extended to be applicable to more general classes of sweepable geometry. This comparison and extension of node projection algorithms led to the development of a new node projection technique: the SmartAffine method. This method builds on previous techniques and is characterized by its speed. Finally, a technique for coupling node projection techniques is presented. This technique characterizes the complexity of the sweepable geometry and applies the most appropriate node projection scheme. This is accomplished without user interaction and improves the speed of the sweeping algorithm and the quality of swept meshes.

This thesis investigates the suitability of thermo-kinetically recycled plastics for use in commercial product applications using finite element analysis and statistics. Different recycled material blends were tested and evaluated for their use in commercial product applications. There are six different blends of thermo-kinetically recycled plastics used for testing and CATIA is used for finite element analysis. The different types of thermo-kinetically recycled plastics blends are: pop bottles made of PolyethyleneTeraphthalate (PET), milk jugs made of High-Density Polyethylene (HDPE), Vinyl seats made of Poly Vinyl Chloride (PVC) and small amount of Polypropylene (PP) and Urethane, electronic scrap made of engineering resins like Acrylo-Nitrile-Butadiene Styrene (ABS), Polystyrene (PS) and Polycarbonate (PC), agriculture waste consisting of Low Density Polyethylene (LDPE), industrial waste consisting of Nylon (PA66) and PolyethyleneTeraphthalate (PET), household waste consisting of Polystyrene (PS). The methods employed during the study include three phases for each of six blends available: 1.Density, tensile and impact testing of each blend 2.Correlation of mechanical properties to blend 3.Finite element analysis of the service performance of a product made from each thermo-kinetically recycled plastic blend This thesis shows that some of the recycled plastics materials that were tested are qualified to be used in the pallet. Those materials that qualified were Industrial waste consisting of Nylon and PolyethyleneTeraphthalate, household waste consisting of Polystyrene.

Non-uniform rational B-splines (NURBS) methodology is presented, on which the isogeometric analysis is based. T-splines are also introduced as a surface design methodology, which are a generalization of NURBS and permit local refinement. Isogeometric analysis using NURBS and T-splines are applied separately to a structural mechanics problem. The results are compared with the closed-form solution. The desirable performance of isogeometric analysis using T-splines on engineering analysis is demonstrated.

Ductile braced frames are often used to resist lateral earthquake loads in steel buildings; however the presence of a brace element can sometimes interfere with architectural features. One common type of ductile braced frame system sometimes used to accommodate architectural features is the eccentrically braced frame (EBF). In order to dissipate seismic forces, EBF beam regions (called links) must sustain large inelastic deformations. EBF links with column connections must transmit large moments and shear forces to facilitate link rotation. Experiments have shown that welded link-to-column connections tend to fracture in the link flange prior to large link rotations. This study investigated methods for improving EBF link-to-column connection performance, and proposed an alternative ductile braced frame system for accommodating architectural features. Several EBF links with reduced web and flange sections were analytically investigated using validated finite element models in ABAQUS. Results indicated that putting holes in the link web reduced stress and strain values in the link flanges at the connection, but increased the plastic strain and stress triaxiality in the web at the edges of holes. Removing area from the link flanges had little effect on connection stresses and strains. Thus, the reduced web section and reduced flange section methods are not a promising solution to the EBF link-to-column connection problem. The alternative braced frame system proposed in the dissertation used ductile beam splices and buckling-restrained braces in eccentric configurations (BRBF-Es) to accommodate architectural features. Design considerations for the BRBF-Es were determined and dynamic BRBF-E performance was compared with EBF performance. BRBF-E system and component performance was determined using multiple finite element methods. Inter-story drifts and residual drifts for the BRBF-Es were similar to those for EBFs. Results indicated that BRBF-Es are a viable alternative to the EBF, and may result in better design economy than EBFs. With the BRBF-E, damage was isolated within the brace, and in the EBF, damage was isolated within the link, indicating simpler repairs with the BRBF-E. Shop welding of BRBF-E members may replace the multiple field welds required in EBF construction.

The purpose of this work was to determine the possible effects of isolated spinal ligament transection on the biomechanics of the lumbar spine. A finite element model of a lumbar spine was developed and validated against experimental data. The model was tested in the primary modes of spinal motion in the intact condition, followed by comparative analysis of isolated removal of each spinal ligament. Results showed that stress increased in the remaining ligaments once a ligament was removed, potentially leading to ligament damage. Results also showed changes in bone remodeling "stimulus" which could lead to changes in bone density. Isolated ligament transection had little effect on intervertebral disc pressures. All major biomechanical changes occurred at the same spinal level as the transected ligament, with minor changes at adjacent levels. The results of this work demonstrate that iatrogenic damage of spinal ligaments disturbs the load sharing within spinal-ligament complex and may induce significant clinical changes in the spinal motion segment.

The purpose of this research was to investigate the influence of applying an intrinsic decompressive pre-load to a particular dynamic stabilization device on the biomechanical response of the lumbar spine. The FlexSPAR, which supports this ability, was used as a test case. A finite element model of a full lumbar spine was developed and validated against experimental data, and tested in the primary modes of spinal motion. The model was used to compare five lumbar spine test cases: healthy, degenerate, implanted with a pre-loaded device, implanted with a device without a pre-load, and implanted with rigid fixators. Results indicated that a pre-loaded FlexSPAR led to improved disc height restoration and segmental biomechanics. Results also showed that a pre-loaded FlexSPAR led to less change in bone remodeling stimulus in comparison to the device without a pre-load and rigid fixators. This work shows that there is a potential to improve the performance of posterior dynamic stabilization devices by incorporating a pre-load in the device.

The purpose of this research was to design, develop, and test coronary stent designs composed of carbon-infiltrated carbon nanotubes (CI-CNTs). Coronary stents currently have two major complications: restenosis and thrombosis. CI-CNT stents have potential to address both of these issues, and therefore may provide improved clinical outcomes. CI-CNT stent geometry is patterned using high-resolution photolithography that provide advantages in design possibilities.To develop a coronary stent, a standard design process was followed including: background, design specifications, concept generation, development, analysis, and testing. Background research was first completed and general design specifications for coronary stent performance were compiled. Multiple design concepts were generated, evaluated, and finally a design was selected. This stent design was further developed and optimized using analytical tools along with finite element analysis. This stent design used tapered struts in repeating segments to reduce stress and improve radial force. The design was modeled and analyzed as both a flat geometry as well as in a cylindrical configuration. Mechanics of materials equations and geometry specific finite element analysis were used to guide the final coronary stent design. The stent design was tested mechanically, and additional tests were performed to verify the blood compatibility of the CI-CNT material. The flat version of the stent design was manufactured and mechanically tested to verify performance. The performance of the cylindrical stent configuration was analyzed using an FE model of an atherosclerotic artery. This arterial FE model was created and validated by analyzing balloon angioplasty of a common stainless steel stent. The biocompatibility of CI-CNTs was explored and studied. Blood compatibility testing of CI-CNT samples was performed with results comparable in performance to stainless steel. A method of stent deployment was planned, and several other stent design concepts were analyzed. This research demonstrates that a functioning coronary stent can be manufactured from CI-CNTs. The optimized design has potential to address problems currently associated with stents. However, a major challenge for CI-CNT stent designs is meeting the design requirement of sufficient radial force. CI-CNT stents also need to have excellent blood compatibility to justify being used in stent applications.

With the desire for conventional coal-fired power plants to perform flexible operations, the impact of this operation has become important to the field of steam turbine modelling. This study sought to develop a computationally inexpensive turbine model with minimal OEM intervention in order to predict the internal clearances of high-pressure and intermediate-pressure turbines from Eskom's current turbine fleet. The study saw the utilisation of the Nozzle Analogy theory to develop a 1D multistage turbine thermofluid model as well as the development of a representative 1D turbine process model in order to predict the internal temperature gradients promoted within a steam turbine during transient operation. From this model a further 3D FEA turbine model of both the HP and IP turbine units were developed from simple turbine diagrams to apply the predicted temperature boundaries and predict the thermal and structural response of turbine components during transient loading during a full Cold Start procedure. The result of this study was the successful validation of the 1D and 3D Turbine models against plant data from the candidate unit. This was in the form of known process data of unit performance, as well as thermocouple and differential expansion data taken from sensors housed on the turbine unit itself. Through the validation of these parameters, various calibrations techniques were developed over the course of the study with these techniques allowing investigators to gain insight into turbine aging, operator intervention as well as brought turbine component response. The successful establishment of the paired turbine model allowed investigators to evaluate the cold clearances defined during construction and maintenance of these turbine units in industry, which contributes greatly to the availability and efficiency of the unit during these transient operations. Additionally, the establishment of this model allowed for the investigation of the role that start up speed has on turbine component response. This study demonstrated that the development of such a modelling methodology was possible and yielded results with were accurate and insightful in understanding turbine component responses which are otherwise impossible to measure during real-world operation.