This report summarizes computational efforts to model interfacial fracture using cohesive zone models in the SIERRA/SolidMechanics (SIERRA/SM) finite element code. Cohesive surface elements were used to model crack initiation and propagation along predefined paths. Mesh convergence was observed with SIERRA/SM for numerous geometries. As the funding for this project came from the Advanced Simulation and Computing Verification and Validation (ASC V&V) focus area, considerable effort was spent performing verification and validation. Code verification was performed to compare code predictions to analytical solutions for simple three-element simulations as well as a higher-fidelity simulation of a double-cantilever beam. Parameter identification was conducted with Dakota using experimental results on asymmetric double-cantilever beam (ADCB) and end-notched-flexure (ENF) experiments conducted under Campaign-6 funding. Discretization convergence studies were also performed with respect to mesh size and time step and an optimization study was completed for mode II delamination using the ENF geometry. Throughout this verification process, numerous SIERRA/SM bugs were found and reported, all of which have been fixed, leading to over a 10-fold increase in convergence rates. Finally, mixed-mode flexure experiments were performed for validation. One of the unexplained issues encountered was material property variability for ostensibly the same composite material. Since the variability is not fully understood, it is difficult to accurately assess uncertainty when performing predictions.
Non-destructive detection methods can reliably certify that gas transfer system (GTS) reservoirs do not have cracks larger than 5%-10% of the wall thickness. To determine the acceptability of a reservoir design, analysis must show that short cracks will not adversely affect the reservoir behavior. This is commonly done via calculation of the J-Integral, which represents the energetic driving force acting to propagate an existing crack in a continuous medium. J is then compared against a material's fracture toughness (J{sub c}) to determine whether crack propagation will occur. While the quantification of the J-Integral is well established for long cracks, its validity for short cracks is uncertain. This report presents the results from a Sandia National Laboratories project to evaluate a methodology for performing J-Integral evaluations in conjunction with its finite element analysis capabilities. Simulations were performed to verify the operation of a post-processing code (J3D) and to assess the accuracy of this code and our analysis tools against companion fracture experiments for 2- and 3-dimensional geometry specimens. Evaluation is done for specimens composed of 21-6-9 stainless steel, some of which were exposed to a hydrogen environment, for both long and short cracks.
The microcombustor described in this report was developed primarily for thermal management in microsystems and as a platform for micro-scale flame ionization detectors (microFID). The microcombustor consists of a thin-film heater/thermal sensor patterned on a thin insulating membrane that is suspended from its edges over a silicon frame. This micromachined design has very low heat capacity and thermal conductivity and is an ideal platform for heating catalytic materials placed on its surface. Catalysts play an important role in this design since they provide a convenient surface-based method for flame ignition and stabilization. The free-standing platform used in the microcombustor mitigates large heat losses arising from large surface-to-volume ratios typical of the microdomain, and, together with the insulating platform, permit combustion on the microscale. Surface oxidation, flame ignition and flame stabilization have been demonstrated with this design for hydrogen and hydrocarbon fuels premixed with air. Unoptimized heat densities of 38 mW/mm{sup 2} have been achieved for the purpose of heating microsystems. Importantly, the microcombustor design expands the limits of flammability (Low as compared with conventional diffusion flames); an unoptimized LoF of 1-32% for natural gas in air was demonstrated with the microcombustor, whereas conventionally 4-16% observed. The LoF for hydrogen, methane, propane and ethane are likewise expanded. This feature will permit the use of this technology in many portable applications were reduced temperatures, lean fuel/air mixes or low gas flows are required. By coupling miniature electrodes and an electrometer circuit with the microcombustor, the first ever demonstration of a microFID utilizing premixed fuel and a catalytically-stabilized flame has been performed; the detection of -1-3% of ethane in hydrogen/air is shown. This report describes work done to develop the microcombustor for microsystem heating and flame ionization detection and includes a description of modeling and simulation performed to understand the basic operation of this device. Ancillary research on the use of the microcombustor in calorimetric gas sensing is also described where appropriate.