Publications

Forward-to-reverse bias step-recovery measurements were performed on In.07Ga.93N/GaN and Al.36Ga.64N/Al.46Ga.54N quantum-well (QW) light-emitting diodes grown on sapphire. With the QW sampling the minority-carrier hole density at a single position, distinctive two-phase optical decay curves were observed. Using diffusion equation solutions to self-consistently model both the electrical and optical responses, hole transport parameters tp = 758 {+-} 44 ns, Lp = 588 {+-} 45 nm, and up = 0.18 {+-} 0.02 cm2/Vs were obtained for GaN. The mobility was thermally activated with an activation energy of 52 meV, suggesting trap-modulated transport. Optical measurements of sub-bandgap peaks exhibited slow responses approaching the bulk lifetime. For Al.46Ga.54N, a longer lifetime of tp = 3.0 us was observed, and the diffusion length was shorter, Lp = 280 nm. Mobility was an order of magnitude smaller than in GaN, up = 10-2 cm2/Vs, and was insensitive to temperature, suggesting hole transport through a network of defects.

We report micro-Raman studies of self-heating in an AlGaN/GaN heterostructure field-effect transistor using below (visible 488.0 nm) and near (UV 363.8 nm) GaN band-gap excitation. The shallow penetration depth of the UV light allows us to measure temperature rise ({Delta}T) in the two-dimensional electron gas (2DEG) region of the device between drain and source. Visible light gives the average {Delta}T in the GaN layer, and that of the SiC substrate, at the same lateral position. Combined, we depth profile the self-heating. Measured {Delta}T in the 2DEG is consistently over twice the average GaN-layer value. Electrical and thermal transport properties are simulated. We identify a hotspot, located at the gate edge in the 2DEG, as the prevailing factor in the self-heating.

Abstract not provided.

Abstract not provided.

Order-of-accuracy verification is necessary to ensure that software correctly solves a given set of equations. One method to verify the order of accuracy of a code is the method of manufactured solutions. In this study, a manufactured solution has been derived and implemented that allows verification of not only the Euler, Navier-Stokes, and Reynolds-Averaged Navier-Stokes (RANS) equation sets, but also some of their associated boundary conditions (BC's): slip, no-slip (adiabatic and isothermal), and outflow (subsonic, supersonic, and mixed). Order-of-accuracy verification has been performed for the Euler and Navier-Stokes equations and these BC's in a compressible computational fluid dynamics code. All of the results shown are on skewed, non-uniform meshes. RANS results will be presented in a future paper. The observed order of accuracy was lower than the expected order of accuracy in two cases. One of these cases resulted in the identification and correction of a coding mistake in the CHAD gradient correction that was reducing the observed order of accuracy. This mistake would have been undetectable on a Cartesian mesh. During the search for the CHAD gradient correction problem, an unrelated coding mistake was found and corrected. The other case in which the observed order of accuracy was less than expected was a test of the slip BC; although no specific coding or formulation mistakes have yet been identified. After the correction of the identified coding mistakes, all of the aforementioned equation sets and BC's demonstrated the expected (or at least acceptable) order of accuracy except the slip condition.

Abstract not provided.

Rate constants for the thermal dissociation of Si{sub 2}H{sub 6} are predicted with a novel transition state model. The saddle points for dissociation on the Si{sub 2}H{sub 6} potential energy surface are lower in energy than the corresponding separated products, as confirmed by high level ab initio quantum mechanical calculations. Thus, the dissociations of Si{sub 2}H{sub 6} to produce SiH{sub 2} + SiH{sub 4} (R1) and H{sub 3}SiSiH + H{sub 2} (R2) both proceed through tight inner transition states followed by loose outer transition states. The present 'dual' transition state model couples variational phase space theory treatments of the outer transition states with ab initio based fixed harmonic vibrator treatments of the inner transition states to obtain effective numbers of states for the two transition states acting in series. It is found that, at least near room temperature, such a dual transition state model is generally required for the proper description of each of the dissociations. Only at quite high temperatures, i.e., above 2000 K for (R1) and 600 K for (R2), does a single fixed inner transition state provide an adequate description. Similarly, only at quite low temperatures (below 100 and 10 K for (R1) and (R2), respectively) does a single outer transition state provide an adequate description. Pressure dependent rate constants are obtained from solutions to the multichannel master equation. These calculations confirm that dissociation channel (R2) is negligible under conditions relevant to the thermal chemical vapor deposition (CVD) processes. Rate constants for the chemical activation reactions, SiH{sub 2} + SiH{sub 4} {yields} Si{sub 2}H{sub 6} (R-1) and SiH{sub 2} + SiH{sub 4} {yields} H{sub 3}SiSiH + H{sub 2} (R3), are also evaluated within the dual transition state model. It is found that reaction R3 is the dominant channel for low pressures and high temperatures, i.e., below 100 Torr for temperatures above 1100 K.

In this article, we discuss in detail the addition of hydrogen atoms to diacetylene and the reverse dissociation reactions, H + C{sub 4}H{sub 2} {leftrightarrow} i-C{sub 4}H{sub 3} (R1) and H + C{sub 4}H{sub 2} n-C{sub 4}H{sub 3} (R2). The theory utilizes high-level electronic structure methodology to characterize the potential energy surface, Rice-Ramsperger-Kassel-Marcus (RRKM) theory to calculate microcanonical/J-resolved rate coefficients, and a two-dimensional master-equation approach to extract phenomenological (thermal) rate coefficients. Comparison is made with experimental results where they are available. The rate coefficients k{sub 1}(T, p) and k{sub 2}(T, p) are cast in forms that can be used in chemical kinetic modeling. In addition, we predict values of the heats of formation of i-C{sub 4}H{sub 3} and n-C{sub 4}H{sub 3} and discuss their importance in flame chemistry. Our basis-set extrapolated, quadratic-configuration-interaction with single and double excitations (and triple excitations added perturbatively), QCISD(T), predictions of these heats of formation at 298 K are 130.8 kcal/mol for n-C{sub 4}H{sub 3} and 119.3 kcal/mol for the i-isomer; multireference CI calculations with a nine-electron, nine-orbital, complete-active-space (CAS) reference wavefunction give just slightly larger values for these parameters. Our results are in good agreement with the recent focal-point analysis of Wheeler et al. (J. Chem. Phys. 2004, 121, 8800-8813), but they differ substantially for {Delta} H{sub f 298}{sup 0}(n-C{sub 4}H{sub 3}) with the earlier diffusion Monte Carlo predictions of Krokidis et al.

Abstract not provided.

Abstract not provided.

Many accelerators at Sandia National Laboratories utilize the Rimfire gas switch for high-voltage, high-power switching. Future accelerators will have increased performance requirements for switching elements. When designing improved versions of the Rimfire switch, there is a need for quick and accurate simulation of the electrical effects of geometry changes. This paper presents an advanced circuit model of the Rimfire switch that can be used for these simulations. The development of the model is shown along with comparisons to past models and experimental results.

The Sandia Lightning Simulator at Sandia National Laboratories can provide up to 200 kA for a simulated single lightning stroke, 100 kA for a subsequent stroke, and hundreds of Amperes of continuing current. It has recently been recommissioned after a decade of inactivity and the single-stroke capability demonstrated. The simulator capabilities, basic design components, upgrades, and diagnostic capabilities are discussed in this paper.

An experimental investigation is made into the fluid mechanics and heat transfer of a circular cylinder immersed in a wall-bounded turbulent mixed-convection flow of water. The cylinder is oriented spanwise to the forced channel flow and within the thermal boundary layer of the heated lower wall. The flow channel is capped with a cold, near-adiabatic upper wall producing a fully turbulent gap Rayleigh number of 10{sup 8}. A low-speed crossflow is applied to advect the turbulent thermal plumes over the cylinder surface. We present spatially resolved cylinder-surface heat-flux data alongside 2-D PIV imaging of the streamwise and wall-normal velocity components for two flow conditions in the mixed-convection heat-transfer regime. The measured cylinder-wake flowfield reflects the complex coupling between the separated wake flow, the highly turbulent freestream and the buoyant wall and cylinder boundary layers. A method for measurement of spatially resolved surface heat fluxes based on the measured cylinder-surface temperature distribution and a well-posed two-dimensional solution to the conduction problem in the cylinder wall is presented. The resulting spatially resolved flux measurements show enhanced surface heat transfer, which results from the intense buoyancy generated free-stream turbulence and mixing in the cylinder wake. This work extends the literature on thermal convection with crossflow well into the turbulent regime and is, to our knowledge, the first investigation of surface heat-transfer to an object of engineering importance placed in this type of turbulent mixed-convection flowfield. The data are currently being utilized for validation of mixed convection turbulence models at Sandia and comparisons between the computational and experimental results are presented.

Abstract not provided.

The WLF equation is typically used to describe the dependence of polymer mobility on temperature at atmospheric pressure. Tests at different pressures would at least require different WLF parameterization. Completely different tests, for example, probing the temperature dependence of mobility at constant density, would require even greater modifications. By performing molecular dynamics simulations on simple chain molecules equilibrated at different thermodynamic states, we have shown that the mobility depends in a more general sense on the potential energy density of the system. That is, mobilities for any equilibrated state collapse onto one master curve when plotted against the potential energy density. Moreover, this relationship can be fit by either a 'generalized' WLF equation or by a power-law relationship observed in critical phenomena. When this mobility relationship is used within a rheologically simple, thermodynamically consistent, viscoelastic framework, quantitative agreement is seen between experimental data and theoretical predictions on a range of tests covering enthalpy relaxation to mechanical yield to physical aging.

Abstract not provided.

Abstract not provided.

In this study, we describe the extension of the 2-d preliminary design bluff body drag estimation tool developed by De Chant1 to apply for 3-d flows. As with the 2-d method, the 3-d extension uses a combined approximate Green's function/Gram-Charlier series approach to retain the body geometry information. Whereas, the 2-d methodology relied solely upon the use of small disturbance theory for the inviscid flow field associated with the body of interest to estimate the near-field initial conditions, e.g. velocity defect, the 3-d methodology uses both analytical (where available) and numerical inviscid solutions. The defect solution is then used as an initial condition in an approximate 3-d Green's function solution. Finally, the Green's function solution is matched to the 3-d analog of the classical 2-d Gram-Charlier series and then integrated to yield the net form drag on the bluff body. Preliminary results indicate that drag estimates computed are of accuracy equivalent to the 2-d method for flows with large separation, i.e. less than 20% relative error. As was the lower dimensional method, the 3-d concept is intended to be a supplement to turbulent Navier-Stokes and experimental solution for estimating drag coefficients over blunt bodies.

The Method of Nearby Problems is employed to generate exact solutions to equations 'nearby' the steady and unsteady Burgers equation. Burgers equation is chosen because of the existence of exact solutions, and these exact solutions are discussed. Legendre polynomials are used to derive the exact solutions to the nearby problems, and the application of Legendre polynomials for both 1D and 2D problems is also discussed. Results are presented for the steady-state Burgers equation corresponding to a viscous shock wave for Reynolds numbers of 8, 16, and 512. The low Reynolds number cases are well approximated by 10th order Legendre polynomial fits, while the high Reynolds number case is not. The unsteady Burgers equation corresponding to coalescence of two viscous shock waves at a Reynolds number of 8 is also examined. Preliminary results indicate that further investigation is required to accurately capture this 2D solution.

Abstract not provided.

We performed calculations to investigate the classical theories of chain branching and thermal--run--away that lead to the rapid oxidation of fuels. Mathematically, both theories infer the existence of eigenvalues with positive real parts i.e., explosive modes. We found in studies of homogeneous hydrogen--air and the methane--air mixtures that when ignition is initiated by a sufficiently high initial temperature, the transient response of the system exhibits two stages. The first stage is characterized by the existence of explosive modes. The ensuing second stage consists of fast exponential decay modes that bring the system to its equilibrium point. We demonstrated with two examples that the existence of explosive modes is not a necessary condition for the existence of a premixed flame. Homogeneous ignition calculations for mixtures with an initial concentration of radical species suggest that the diffusive transport of radical species is probably responsible for the lack of explosive modes in premixed flames.

Modal analysis of three-dimensional structures frequently involves finite element discretizations with millions of unknowns and requires computing hundreds or thousands of eigenpairs. In this presentation we review methods based on domain decomposition for such eigenspace computations in structural dynamics. We distinguish approaches that solve the eigenproblem algebraically (with minimal connections to the underlying partial differential equation) from approaches that tightly couple the eigensolver with the partial differential equation.

Abstract not provided.

Low-temperature combustion concepts that utilize cooled EGR, early/retarded injection, high swirl ratios, and modest compression ratios have recently received considerable attention. To understand the combustion and, in particular, the soot formation process under these operating conditions, a modeling study was carried out using the KIVA-3V code with an improved phenomenological soot model. This multi-step soot model includes particle inception, surface growth, surface oxidation, and particle coagulation. Additional models include a piston-ring crevice model, the KH/RT spray breakup model, a droplet wall impingement model, a wall heat transfer model, and the RNG k-{var_epsilon} turbulence model. The Shell model was used to simulate the ignition process, and a laminar-and-turbulent characteristic time combustion model was used for the post-ignition combustion process. A low-load (IMEP=3 bar) operating condition was considered and the predicted in-cylinder pressures and heat release rates were compared with measurements. Predicted soot mass, soot particle size, soot number density distributions and other relevant quantities are presented and discussed. The effects of variable EGR rate (0-68%), injection pressure (600-1200 bar), and injection timing were studied. The predictions demonstrate that both EGR and retarded injection are beneficial for reducing NO{sub x} emissions, although the former has a more pronounced effect. Additionally, higher soot emissions are typically predicted for the higher EGR rates. However, when the EGR rate exceeds a critical value (over 65% in this study), the soot emissions decrease. Reduced soot emissions are also predicted when higher injection pressures or retarded injection timings are employed. The reduction in soot with retarded injection is less than what is observed experimentally, however.

Abstract not provided.