Explosion hazards exist in many industrial sectors including chemical processing, mining, nuclear power, and aviation. Thermal ignition is the name given to the particular case where the initiation energy is supplied via thermal heating of a gas. The critical conditions leading to thermal ignition are in general highly configuration dependent and require a broad set of experimentation to investigate the influence of wide ranging physical processes on ignition. To aid this effort the present work comprises three main experiments covering a range of thermal ignition hazards. First, a heated atmosphere test with fuel injection (ASTM-E659) was implemented to enable the study of heavy hydrocarbon fuels such as Jet A and multicomponent surrogates. This approach showed the existence of cool flame ignition modes near the ignition thresholds for most fuels. The autoignition temperature (AIT) of commodity Jet A was found to be reasonably reproducible by most alkane fuels including n-hexane. Multicomponent surrogates were also able to match the cool flame ignition regimes reasonably well.
Next, ignition using a vertical heated surface in a cold reactive atmosphere was studied in the laminar flow regime. The effects of dilution with nitrogen and reduced pressure were explored for n-hexane/oxygen/nitrogen mixtures. Results found a modest dependence of minimum ignition temperatures on pressure and nitrogen fraction however, with a significant reduction in explosion severity as measured by the maximum overpressure and transient duration. At sufficiently reduced oxygen concentrations, localized weakly propagating flames were found to form in the thermal layer near the surface and produce sustained puffing flame instabilities. One-dimensional flame simulations with detailed kinetics were conducted to supplement and aid in interpretation of the experimental measurements for diluted mixtures. Correlation of ignition thresholds were found to be possible using simplified flame properties and laminar natural convection boundary layer theory.
Finally, a novel experiment was designed to explore the effects of turbulent transition and confinement of large heated surfaces on ignition thresholds. Modeling of the energy balance for resistive heating showed that cylinders up to 36 in. long could be heated using modest power supplies. Six cylinder sizes of varying length were chosen based on this analysis to explore laminar, transitional, and turbulent flow regimes. A large scale flow visualization system was created to study these flow regimes and found that turbulent transition occurred for cylinders as small as 10 in. long for wall temperatures of 1000 K. A study of the transitional dependence on temperatures for large temperature difference (T = 555–1140 K), highly non-Boussinesq conditions found that the transitional Rayleigh number decreased by two orders of magnitude in this regime. The thermal layer thickness at the transition height was estimated in order to obtain a relevant length scale to the boundary layer transition problem. Using this a more consistent transition criteria was obtained (Ra using the thermal thickness length scale) and found to vary by only a factor of two in the high temperature cases studied.
The implementation of these cylinders in ignition testing revealed that there was a strong influence of heating rate due to confinement. The use of absorption spectroscopy showed that for low heating rates the fuel was mostly consumed in low temperature reactions prior to or in place of rapid ignition. This resulted in larger ignition temperatures and weak flames which propagate only in the thermal boundary layer. This effect was explained as a consequence of reduced flow recirculation times due to confinement. A strong influence of turbulence was also found for ignition thresholds when compared with other data for ignition by vertical hot surfaces in the laminar regime. Turbulence was also found to strongly influence the explosion properties due to turbulent flame acceleration. This resulted in larger explosion pressures, shorter transients, and faster flames.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/9g5j-2b97, author = {Jones, Silken Michelle}, title = {Thermal Ignition by Vertical Cylinders}, school = {California Institute of Technology}, year = {2021}, doi = {10.7907/9g5j-2b97}, url = {https://resolver.caltech.edu/CaltechTHESIS:12182020-055522985}, abstract = {Accidental thermal ignition events present a significant hazard to the aviation industry. There is scarcity of experimental data on ignition by external natural convection flows for surface areas larger than 10 cm². In this work, thermal ignition of external natural convection flows by vertical cylinders is investigated. The effect of geometry is studied by resistively heating stainless steel cylinders of various sizes in a stoichiometric n-hexane and air mixture at 298 K and 1 bar. Cylinder lengths range from 12.7 to 25.4 cm, and cylinder surface areas vary from 25 to 200 cm². Logistic regression is used to provide statistical information about the ignition threshold (50% probability of ignition). The maximum ignition threshold found is 1117 K for a cylinder 12.7 cm long and 50 cm² in surface area. The minimum ignition threshold found is 1019 K for a cylinder 25.4 cm long and 200 cm² in surface area. The maximum uncertainty on these ignition thresholds is ±29 K, which comes from the maximum uncertainty on the pyrometer measurement used to record cylinder surface temperatures. The dependence of ignition threshold on both surface area and length of a cylinder is found to be minor. High speed visualizations of ignition indicated that ignition occurs near the top edge of all cylinders.
The entire experimental setup is heated to allow for ignition tests with multi-component, heavy-hydrocarbon fuels including Jet A and two surrogate fuels, Aachen and JI. The cylinder used for all testing of heavier fuels is 25.4 cm long and 200 cm² in surface area. Hexane is also tested with the heated vessel to investigate the effect of ambient temperature on ignition. At an ambient temperature of 393 K, the ignition threshold of hexane is 933 K. Aachen has an ignition threshold of 947 K at an ambient temperature of 373 K. JI has an ignition temperature of 984 K at an ambient temperature of 393 K. Jet A has an ignition temperature of 971 K at an ambient temperature of 333 K. The maximum uncertainty on these thresholds is ±29 K. JI is found to be the most appropriate surrogate for Jet A.
From the experiments, two main conclusions are reached. Ignition threshold temperatures in external natural convection flows are very weakly correlated with surface area. The observed ignition thresholds do not show the drastic transition of ignition temperature with surface area that is observed in internal natural convection situations. Observed ignition thresholds for comparable surface areas (100 to 200 cm²) are 500 to 600 K higher for external natural convection than internal natural convection. Hexane was found to be a reasonable surrogate for Jet A (38 K difference in ignition threshold) in external natural convection ignition testing. The more complex multi-component JI surrogate, while having an ignition threshold more comparable to Jet A (13 K difference in ignition threshold), requires heating the experimental apparatus and associated difficulties of fuel handling as well as the soot generation by combustion.
Two simplified models of ignition are explored. The first is an investigation of ignition chemistry using a zero-dimensional reactor and a detailed kinetic mechanism for hexane. The temperature history of the reactor is prescribed by an artificial streamline whose rate of temperature increase is parametrically varied. The results from the zero-dimensional reactor computation reveal that a gradually heated streamline exhibits two-stage ignition behavior, while a rapidly heated streamline only experiences one ignition event. The second model of ignition is a one-dimensional simulation of ignition adjacent to a cylinder at a prescribed temperature. The formulation included diffusion of species and thermal energy as well as chemical reaction and employed Lagrangian coordinates. The chemistry is modeled with a reaction mechanism for hydrogen to reduce numerical demand. Heat flux and energy balance are analysed to gain insight into the ignition dynamics. Initially, heat transfer is from the wall into the gas, and a mostly nonreactive thermal boundary layer develops around the cylinder. As reaction in the gas near the surface begins to release energy, the heat transfer decreases, and, near the critical temperature for ignition, the direction of heat flux reverses and is from the gas into the wall. In a case where ignition takes place, there is rapid rise in temperature in the gas within the thermal layer, and a propagating flame is observed to emerge into surrounding cold gas. The heat transfer from the hot combustion products results in a continuous heat flux from the gas into the wall. In a case where ignition does not take place, no flame is observed and the heat flux at the wall is slightly positive. For the critical condition just below the ignition threshold, a balance between energy release and diffusion in the adjacent gas results in a small temperature rise in the thermal layer, but a propagating flame is not created. The Van’t Hoff ignition criterion of vanishing heat flux at the ignition threshold is approximately but not exactly satisfied. Contrasting the two modeling ideas, we observe that modeling adiabatic flows along computed nonreactive streamlines is useful in examining the role of detailed chemistry but lacks important diffusion effects. Including mass and thermal transport provides more insight into important ignition dynamics but comes at the expense of increased computational complexity.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/VJSH-TF65, author = {Veilleux, Jean-Christophe}, title = {Pressure and Stress Transients in Autoinjector Devices}, school = {California Institute of Technology}, year = {2019}, doi = {10.7907/VJSH-TF65}, url = {https://resolver.caltech.edu/CaltechTHESIS:02172019-174051312}, abstract = {The viscosity of drug solutions delivered parenterally has been increasing over the years. Injecting viscous drug solutions using spring-actuated autoinjector devices is challenging due to a number of technical and human factor constraints. Some of the related challenges are investigated in this thesis.
Actuation of autoinjector devices powered using stiff springs can create deleterious pressure and stress transients which are not needed to achieve the normal functions of the device. Experimental measurements have shown that peak pressures and stresses substantially larger than what is needed to achieve the normal device function can occur during the actuation phase, creating unnecessary potential for device failure.
The acceleration of the syringe during actuation can be very large, often creating transient cavitation in the cone region. The occurrence or absence of cavitation is determined by the relative timing of syringe pressurization and syringe acceleration, which is affected by several factors such as the presence, location, and size of an air gap inside the syringe, and the friction between the plunger-stopper and the syringe.
Experiments and numerical simulations have shown that sharp pressure waves traveling inside the syringe can be amplified within the cone terminating the syringe. Despite the potential for shock focusing, the impulsive pressurization and the rapid deceleration of pre-filled syringes create a potential for failure which is localized in the syringe shoulder and at the junction between the flange and the barrel, not inside the cone. The cavitation events, on the other hand, create a potential for failure which is limited to a region in close proximity of the bubble upon collapse. The collapse of cavitation bubbles located within the syringe cone can be enhanced due to geometrical effects, and the resulting stresses can be large enough to cause syringe failure.
This thesis demonstrates that static and quasi-static analyses do not provide accurate estimates of the peak pressures and stresses occurring within the device. The pressure and stresses created by the highly dynamic events occurring during actuation need to be accounted for during device design in order to improve device reliability, the user’s experience, and patient’s adherence to prescribed treatments. The findings discussed in this work provide insights and guidance as to how the transient events can be mitigated.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/Z93X84M6, author = {Schmidt, Bryan Eric}, title = {On the Stability of Supersonic Boundary Layers with Injection}, school = {California Institute of Technology}, year = {2016}, doi = {10.7907/Z93X84M6}, url = {https://resolver.caltech.edu/CaltechTHESIS:05252016-141702166}, abstract = {The problem of supersonic flow over a 5 degree half-angle cone with injection of gas through a porous section on the body into the boundary layer is studied experimentally. Three injected gases are used: helium, nitrogen, and RC318 (octafluorocyclobutane). Experiments are performed in a Mach 4 Ludwieg tube with nitrogen as the free stream gas. Shaping of the injector section relative to the rest of the body is found to admit a “tuned” injection rate which minimizes the strength of shock waves formed by injection. A high-speed schlieren imaging system with a framing rate of 290 kHz is used to study the instability in the region of flow downstream of injection, referred to as the injection layer. This work provides the first experimental data on the wavelength, convective speed, and frequency of the instability in such a flow. The stability characteristics of the injection layer are found to be very similar to those of a free shear layer. The findings of this work present a new paradigm for future stability analyses of supersonic flow with injection.}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/Z9W37T9X, author = {Coronel, Stephanie Alexandra}, title = {Thermal Ignition Using Moving Hot Particles}, school = {California Institute of Technology}, year = {2016}, doi = {10.7907/Z9W37T9X}, url = {https://resolver.caltech.edu/CaltechTHESIS:06032016-210051818}, abstract = {In this work, ignition of n-hexane-air mixtures was investigated using moving hot spheres of various diameters and surface temperatures. Alumina spheres of 1.8-6 mm diameter were heated using a high power CO2 laser and injected with an average velocity of 2.4 m/s into a premixed n-hexane-air mixture at a nominal initial temperature and pressure of 298 K and 100 kPa, respectively. The 90% probability of ignition using a 6 mm diameter sphere was 1224 K. High-speed experimental visualizations using interferometry indicated that ignition occurred in the vicinity of the separation point in the boundary layer of the sphere when the sphere surface temperature was near the ignition threshold. Additionally, the ignition threshold was found to be insensitive to the mixture composition and showed little variation with sphere diameter.
Numerical simulations of a transient one-dimensional boundary layer using detailed chemistry in a gas a layer adjacent to a hot wall indicated that ignition takes place away from the hot surface; the igniting gas that is a distance away from the surface can overcome diffusive heat losses back to the wall when there is heat release due to chemical activity. Finally, a simple approximation of the thermal and momentum boundary layer profiles indicated that the residence time within a boundary layer varies drastically, for example, a fluid parcel originating at very close to the wall has a residence time that is 65x longer than the residence time of a fluid parcel traveling along the edge of the momentum boundary layer.
A non-linear methodology was developed for the extraction of laminar flame properties from synthetic spherically expanding flames. The results indicated that for accurate measurements of the flame speed and Markstein length, a minimum of 50 points is needed in the data set (flame radius vs. time) and a minimum range of 48 mm in the flame radius. The non-linear methodology was applied to experimental n-hexane-air spherically expanding flames. The measured flame speed was insensitive to the mixture initial pressure from 50 to 100 kPa and increased with increasing mixture initial temperature. One-dimensional freely-propagating flame calculations showed excellent agreement with the experimental flame speeds using the JetSurF and CaltechMech chemical mechanisms.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/Z9Q23X5Z, author = {Bitter, Neal Phillip}, title = {Stability of Hypervelocity Boundary Layers}, school = {California Institute of Technology}, year = {2015}, doi = {10.7907/Z9Q23X5Z}, url = {https://resolver.caltech.edu/CaltechTHESIS:06052015-111128842}, abstract = {The early stage of laminar-turbulent transition in a hypervelocity boundary layer is studied using a combination of modal linear stability analysis, transient growth analysis, and direct numerical simulation. Modal stability analysis is used to clarify the behavior of first and second mode instabilities on flat plates and sharp cones for a wide range of high enthalpy flow conditions relevant to experiments in impulse facilities. Vibrational nonequilibrium is included in this analysis, its influence on the stability properties is investigated, and simple models for predicting when it is important are described.
Transient growth analysis is used to determine the optimal initial conditions that lead to the largest possible energy amplification within the flow. Such analysis is performed for both spatially and temporally evolving disturbances. The analysis again targets flows that have large stagnation enthalpy, such as those found in shock tunnels, expansion tubes, and atmospheric flight at high Mach numbers, and clarifies the effects of Mach number and wall temperature on the amplification achieved. Direct comparisons between modal and non-modal growth are made to determine the relative importance of these mechanisms under different flow regimes.
Conventional stability analysis employs the assumption that disturbances evolve with either a fixed frequency (spatial analysis) or a fixed wavenumber (temporal analysis). Direct numerical simulations are employed to relax these assumptions and investigate the downstream propagation of wave packets that are localized in space and time, and hence contain a distribution of frequencies and wavenumbers. Such wave packets are commonly observed in experiments and hence their amplification is highly relevant to boundary layer transition prediction. It is demonstrated that such localized wave packets experience much less growth than is predicted by spatial stability analysis, and therefore it is essential that the bandwidth of localized noise sources that excite the instability be taken into account in making transition estimates. A simple model based on linear stability theory is also developed which yields comparable results with an enormous reduction in computational expense. This enables the amplification of finite-width wave packets to be taken into account in transition prediction.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/Z9H9935V, author = {Jewell, Joseph Stephen}, title = {Boundary-Layer Transition on a Slender Cone in Hypervelocity Flow with Real Gas Effects}, school = {California Institute of Technology}, year = {2014}, doi = {10.7907/Z9H9935V}, url = {https://resolver.caltech.edu/CaltechTHESIS:05292014-220110640}, abstract = {The laminar to turbulent transition process in boundary layer flows in thermochemical nonequilibrium at high enthalpy is measured and characterized. Experiments are performed in the T5 Hypervelocity Reflected Shock Tunnel at Caltech, using a 1 m length 5-degree half angle axisymmetric cone instrumented with 80 fast-response annular thermocouples, complemented by boundary layer stability computations using the STABL software suite. A new mixing tank is added to the shock tube fill apparatus for premixed freestream gas experiments, and a new cleaning procedure results in more consistent transition measurements. Transition location is nondimensionalized using a scaling with the boundary layer thickness, which is correlated with the acoustic properties of the boundary layer, and compared with parabolized stability equation (PSE) analysis. In these nondimensionalized terms, transition delay with increasing CO2 concentration is observed: tests in 100% and 50% CO2, by mass, transition up to 25% and 15% later, respectively, than air experiments. These results are consistent with previous work indicating that CO2 molecules at elevated temperatures absorb acoustic instabilities in the MHz range, which is the expected frequency of the Mack second-mode instability at these conditions, and also consistent with predictions from PSE analysis. A strong unit Reynolds number effect is observed, which is believed to arise from tunnel noise. NTr for air from 5.4 to 13.2 is computed, substantially higher than previously reported for noisy facilities. Time- and spatially-resolved heat transfer traces are used to track the propagation of turbulent spots, and convection rates at 90%, 76%, and 63% of the boundary layer edge velocity, respectively, are observed for the leading edge, centroid, and trailing edge of the spots. A model constructed with these spot propagation parameters is used to infer spot generation rates from measured transition onset to completion distance. Finally, a novel method to control transition location with boundary layer gas injection is investigated. An appropriate porous-metal injector section for the cone is designed and fabricated, and the efficacy of injected CO2 for delaying transition is gauged at various mass flow rates, and compared with both no injection and chemically inert argon injection cases. While CO2 injection seems to delay transition, and argon injection seems to promote it, the experimental results are inconclusive and matching computations do not predict a reduction in N factor from any CO2 injection condition computed.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/39VT-GB93, author = {Moeller, Robert Carlos}, title = {Current Transport and Onset-Related Phenomena in an MPD Thruster Modified by Applied Magnetic Fields}, school = {California Institute of Technology}, year = {2013}, doi = {10.7907/39VT-GB93}, url = {https://resolver.caltech.edu/CaltechTHESIS:01252013-171305685}, abstract = {This work investigated the effects of tailored, externally-applied magnetic fields on current transport and near-anode processes in the plasma discharge of a magnetoplasmadynamic thruster (MPDT). Electrical and plasma diagnostics were used to determine whether applied magnetic fields could mitigate the effects of the “onset” phenomena, including large-amplitude terminal voltage fluctuations and high anode fall voltages associated with unstable operation and anode erosion. A new MPDT was developed and operated with quasi-steady 1 ms pulses from 36 kW to 3.3 MW with argon propellant. Three magnetic configurations studied included self-field operation (without external electromagnets) and two applied poloidal magnetic fields. One configuration used magnetic field lines tangential to the anode lip (and intersecting the anode further upstream) and the other created a magnetic cusp intersecting the anode downstream.
The influence of the applied fields on the discharge current streamlines, current densities, and key plasma properties (electron temperature, number density, and plasma potential) was studied. Key findings included that the current pattern and current densities redistributed to follow the applied magnetic field lines. Also, the anode fall voltage was substantially reduced with both applied fields over a large range of currents (and eliminated at 8 kA). These results occurred because applied magnetic field lines intersecting the anode provided a high conductivity path and reduced the local electric field required to sustain the radial current densities. The applied fields reduced the amplitude and frequency of the terminal voltage fluctuations (up to 49%) over a broad range of currents and also decreased transients in the ion saturation current, which suggest reduction of current filamentation and surface-eroding anode spots. Additionally, the cusp field reduced mean terminal voltages over the entire range of discharge currents (up to 31%), and the tangential field lowered terminal voltages below 10.7 kA. These significant reductions in onset-related behaviors should lead to improved thruster lifetime and increased efficiency. These results suggest a distinctive and more effective approach to influencing the near-anode phenomena and mitigating the effects of onset with appropriately designed applied magnetic fields that differ from those used in the vast majority of conventional, so-called “applied-field MPD thrusters.”
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/4QW7-TK55, author = {Damazo, Jason Scott}, title = {Planar Reflection of Gaseous Detonation}, school = {California Institute of Technology}, year = {2013}, doi = {10.7907/4QW7-TK55}, url = {https://resolver.caltech.edu/CaltechTHESIS:06112013-153305610}, abstract = {Pipes containing flammable gaseous mixtures may be subjected to internal detonation. When the detonation normally impinges on a closed end, a reflected shock wave is created to bring the flow back to rest. This study built on the work of Karnesky (2010) and examined deformation of thin-walled stainless steel tubes subjected to internal reflected gaseous detonations. A ripple pattern was observed in the tube wall for certain fill pressures, and a criterion was developed that predicted when the ripple pattern would form. A two-dimensional finite element analysis was performed using Johnson-Cook material properties; the pressure loading created by reflected gaseous detonations was accounted for with a previously developed pressure model. The residual plastic strain between experiments and computations was in good agreement.
During the examination of detonation-driven deformation, discrepancies were discovered in our understanding of reflected gaseous detonation behavior. Previous models did not accurately describe the nature of the reflected shock wave, which motivated further experiments in a detonation tube with optical access. Pressure sensors and schlieren images were used to examine reflected shock behavior, and it was determined that the discrepancies were related to the reaction zone thickness extant behind the detonation front. During these experiments reflected shock bifurcation did not appear to occur, but the unfocused visualization system made certainty impossible. This prompted construction of a focused schlieren system that investigated possible shock wave-boundary layer interaction, and heat-flux gauges analyzed the boundary layer behind the detonation front. Using these data with an analytical boundary layer solution, it was determined that the strong thermal boundary layer present behind the detonation front inhibits the development of reflected shock wave bifurcation.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/KZJ1-Y009, author = {Parziale, Nicholaus J.}, title = {Slender-Body Hypervelocity Boundary-Layer Instability}, school = {California Institute of Technology}, year = {2013}, doi = {10.7907/KZJ1-Y009}, url = {https://resolver.caltech.edu/CaltechTHESIS:05312013-164534236}, abstract = {With novel application of optical techniques, the slender-body hypervelocity boundary-layer instability is characterized in the previously unexplored regime where thermo-chemical effects are important. Narrowband disturbances (500-3000 kHz) are measured in boundary layers with edge velocities of up to 5~km/s at two points along the generator of a 5 degree half angle cone. Experimental amplification factor spectra are presented. Linear stability and PSE analysis is performed, with fair prediction of the frequency content of the disturbances; however, the analysis over-predicts the amplification of disturbances. The results of this work have two key implications: 1) the acoustic instability is present and may be studied in a large-scale hypervelocity reflected-shock tunnel, and 2) the new data set provides a new basis on which the instability can be studied.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/JF8P-5495, author = {Thomas, Vaughan Lamar}, title = {Particle-Based Modeling of Ni-YSZ Anodes}, school = {California Institute of Technology}, year = {2012}, doi = {10.7907/JF8P-5495}, url = {https://resolver.caltech.edu/CaltechTHESIS:03302012-133547448}, abstract = {In this work we examine the performance of particle-based models with respect to the Ni-YSZ composite anode system. The conductivity and triple-phase boundary (tpb) of particle-based systems is estimated. The systems considered have mono-dispersed particle size distributions, bi-modal particle size distributions with a YSZ:Ni particle size ratio of 1:0.781, and particle size distributions based on experimental measurements. All three types of systems show qualitative behavioral agreement in terms of conductivity, with clear transition from non-conducting behavior to high conducting behavior over a small transition regime which varied from a nickel phase fraction of .22-.28 for the mono- dispersed cases, 0.19-.0.25 for the bimodal cases, and 0.19-0.30 for the experimentally based cases. Mono-dispersed and simple-polydispersed particle size distribution show very low variation from case to case, with σ/μ ≤ 0.04. Cases based on empirical particle size distribution data demonstrated significantly higher variances which varied over a very large range, 0.3 ≤ σ/μ ≤ 1.1. With respect to the calculations of the TPB length, we find that the same pattern of variance in the measure of the triple-phase boundary length. The TPB length for the mono-dispersed and simple poly-dispersed systems was in the range of 3 × 1012 –4 × 1013 m/m3 . For empirical particle size distribution data the TPB length density was in the range of 8×109–2×1011 m/m3. The variance of the TPB length density follows the same pattern as the conductivity measurements with very low variance for the mono-dispersed and simple poly-dispersed systems and much larger variance for the empirically-based systems. We also examine the association between the TPB length and the availability of conducting pathways for the participating particles xv of individual TPBs. The probability of a TPB having a conducting pathway in the gas phase is essentially 100% in all cases. The probability of an individual tpb section having conducting pathways in either of the solid phases is directly related to percolation condition of that phase.
We also considered a particle-based composite electrode realization based on a three- dimensional reconstruction of an actual Ni-YSZ composite electrode. For this model we used particles which vary in nominal size from 85–465 nm, with size increments of 42.5 nm. We paid particular attention to the coordination numbers between particles and the distribution of particle size interconnections. We found that homogeneous inter-particle connections were far more common than would occur using a random distribution of particles. In particular we found that for a random collection of particles of similar composition the likelihood Ni-Ni particle connections was between 0.18–0.30. For the reconstruction we found the likelihood of Ni-Ni particle connections to be greater than 0.56 in all cases. Similarly, the distribution of connections between particles, with respect to particle size of the participating particles, deviated from what would be expected using a random distribution of particles. Particles in the range of 85–169 nm showed the highest coordination with particles of the same size. Particles in the range of 211–338 nm have the highest coordination with particles of radius 169 nm with very similar distributions. Particles with radius greater than 338 nm represented only 7.2 × 10−3 % of the particles within the reconstruction, and showed the highest coordination with particles of radius of 211 nm, but the distributions vary widely.
In the final chapter, we build a model which can account for mass transfer, hetero- geneous chemistry, surface chemistry, and electrochemistry within a porous electrode. The electric potential is calculated on a particle basis using a network model; gas phase concentrations and surface coverages are calculated with a one-dimensional porous me- dia model. Properties of the porous media are calculated via a TPMC method. TPB electrochemistry is calculated at individual triple phase boundaries within the particle xvi model, based on local gas phase concentrations, surface coverages and particle poten- tials, and then added to the porous media model. Using this tool we are able to calculate the spatial distribution of the Faradaic current within the electrode, and variation in gas phase concentrations within the porous media.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E. and Goodwin, David G.}, } @phdthesis{10.7907/7TDQ-DR81, author = {Capece, Angela Maria}, title = {Plasma-Surface Interactions in Hollow Cathode Discharges for Electric Propulsion}, school = {California Institute of Technology}, year = {2012}, doi = {10.7907/7TDQ-DR81}, url = {https://resolver.caltech.edu/CaltechTHESIS:05312012-113856351}, abstract = {Electric thrusters generate high exhaust velocities and can achieve specific impulses in excess of 1000 s. The low thrust generation and high specific impulse make electric propulsion ideal for interplanetary missions, spacecraft station keeping, and orbit raising maneuvers. Consequently, these devices have been used on a variety of space missions including Deep Space 1, Dawn, and hundreds of commercial spacecraft in Earth orbit. In order to provide the required total impulses, thruster burn time can often exceed 10,000 hours, making thruster lifetime essential.
One of the main life-limiting components on ion engines is the hollow cathode, which serves as the electron source for ionization of the xenon propellant gas. Reactive contaminants such as oxygen can modify the cathode surface morphology and degrade the electron emission properties. Hollow cathodes that operate with reactive impurities in the propellant will experience higher operating temperatures, which increase evaporation of the emission materials and reduce cathode life. A deeper understanding of the mechanisms initiating cathode failure will improve thruster operation, increase lifetime, and ultimately reduce cost.
A significant amount of work has been done previously to understand the effects of oxygen poisoning on vacuum cathodes; however, the xenon plasma adds complexity, and its role during cathode poisoning is not completely understood. The work presented here represents the first attempt at understanding how oxygen impurities in the xenon discharge plasma alter the emitter surface and affect operation of a 4:1:1 BaO-CaO-Al2O3 hollow cathode.
A combination of experimentation and modeling was used to investigate how oxygen impurities in the discharge plasma alter the emitter surface and reduce the electron emission capability. The experimental effort involved operating a 4:1:1 hollow cathode at various conditions with oxygen impurities in the xenon flow. Since direct measurements of the emitter surface state cannot be obtained because of the cathode geometry and high particles fluxes, measurements of the emitter temperature using a two-color pyrometer were used to determine the oxygen surface coverage and characterize the rate processes that occur during poisoning.
A model describing the material transport in the plasma discharge was developed and is used to predict the barium and oxygen fluxes to the emitter surface during cathode operation by solving the species continuity and momentum equations. The dominant ionization process for molecular oxygen in the plasma gas is resonant charge exchange with xenon ions. Barium is effectively recycled in the plasma; however, BaO and O2 are not. The model shows that the oxygen flux to the surface is not diffusion limited.
Experimental results indicate that the oxygen poisoning rate is slow and that the oxygen poisoning coverage on the emitter surface is less than 3%. A time-dependent model of the reaction kinetics of oxygen and barium at the tungsten surface was developed using the experimental results.
The experiments and kinetics model indicate that the dominant processes at the emitter surface are dissociative adsorption of O2, sputtering of the O2 precursor, and desorption of O. Ion sputtering of the weakly bound O2 precursor state limits the poisoning rate and yields low oxygen coverage. Removal of chemisorbed atomic oxygen is dominated by thermal processes. Based on the low oxygen coverage and long poisoning transients, plasma cathodes appear to be able to withstand higher oxygen concentrations than vacuum cathodes.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, } @phdthesis{10.7907/ZKW8-ES97, author = {Ziegler, John Lewis (Jack)}, title = {Simulations of Compressible, Diffusive, Reactive Flows with Detailed Chemistry Using a High-Order Hybrid WENO-CD Scheme}, school = {California Institute of Technology}, year = {2012}, doi = {10.7907/ZKW8-ES97}, url = {https://resolver.caltech.edu/CaltechTHESIS:12302011-185742249}, abstract = {A hybrid weighted essentially non-oscillatory (WENO)/centered-difference (CD) numerical method, with low numerical dissipation, high-order shock-capturing, and structured adaptive mesh refinement (SAMR), has been developed for the direct numerical simulation (DNS) of the multicomponent, compressive, reactive Navier-Stokes equations. The method enables accurate resolution of diffusive processes within reaction zones. This numerical method is verified with a series of one- and two-dimensional test problems, including a convergence test of a two-dimensional unsteady reactive double Mach reflection problem. Validation of the method is conducted with experimental comparisons of three applications all of which model multi-dimensional, unsteady reactive flow: an irregular propane detonation, shock and detonation bifurcations, and spark ignition deflagrations.
The numerical approach combines time-split reactive source terms with a high-order, shock-capturing scheme specifically designed for diffusive flows. A description of the order-optimized, symmetric, finite difference, flux-based, hybrid WENO / centered-difference scheme is given, along with its implementation in a high-order SAMR framework. The implementation of new techniques for discontinuity flagging, scheme-switching, and high-order prolongation and restriction is described. In particular, the refined methodology does not require upwinded WENO at grid refinement interfaces for stability, allowing high-order prolongation and thereby eliminating a significant source of numerical diffusion within the overall code performance.
A minimally reduced irregular detonation mixture mechanism (22 species and 53 reversible reactions) is developed and combined with the WENO-CD numerical method to accurately model two-dimensional hydrocarbon (propane) detonations with detailed chemistry and transport. First of its kind, resolved double Mach reflection (DMR) detonation simulations with a large hyrdocarbon mixture are presented. Detailed discussions and comparisons of the influence of grid resolution, lower-order numerical methods, and inviscid approximations are made in addition to the detailed presentation of fluid dynamics found in an unsteady, highly unstable, reactive DMR simulation. Also conducted are direct experimental comparisons to soot foils and schlieren images with an unresolved large-scale propane detonation channel simulation.
The numerical method is also applied to the DNS of two other problems, detonation/shock bifurcations and spark ignited deflagrations. Through the resolution of viscous/diffusive scales, new insights into how a bifurcated foot develops after a detonation end wall reflection, and how geometry can influence the development of a flame kernel after spark ignition are found.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, } @phdthesis{10.7907/H2W9-ZK95, author = {Boettcher, Philipp Andreas}, title = {Thermal Ignition}, school = {California Institute of Technology}, year = {2012}, doi = {10.7907/H2W9-ZK95}, url = {https://resolver.caltech.edu/CaltechTHESIS:05162012-131336010}, abstract = {
Accidental ignition of flammable gases is a critical safety concern in many industrial applications. Particularly in the aviation industry, the main areas of concern on an aircraft are the fuel tank and adjoining regions, where spilled fuel has a high likelihood of creating a flammable mixture. To this end, a fundamental understanding of the ignition phenomenon is necessary in order to develop more accurate test methods and standards as a means of designing safer air vehicles. The focus of this work is thermal ignition, particularly auto-ignition with emphasis on the effect of heating rate, hot surface ignition and flame propagation, and puffing flames.
Combustion of hydrocarbon fuels is traditionally separated into slow reaction, cool flame, and ignition regimes based on pressure and temperature. Standard tests, such as the ASTM E659, are used to determine the lowest temperature required to ignite a specific fuel mixed with air at atmospheric pressure. It is expected that the initial pressure and the rate at which the mixture is heated also influences the limiting temperature and the type of combustion. This study investigates the effect of heating rate, between 4 and 15 K/min, and initial pressure, in the range of 25 to 100 kPa, on ignition of n-hexane air mixtures. Mixtures with equivalence ratio ranging from 0.6 to = 1.2 were investigated. The problem is also modeled computationally using an extension of Semenov’s classical auto-ignition theory with a detailed chemical mechanism. Experiments and simulations both show that in the same reactor either a slow reaction or an ignition event can take place depending on the heating rate. Analysis of the detailed chemistry demonstrates that a mixture which approaches the ignition region slowly undergoes a significant modification of its composition. This change in composition induces a progressive shift of the explosion limit until the mixture is no longer flammable. A mixture that approaches the ignition region sufficiently rapidly undergoes only a moderate amount of thermal decomposition and explodes quite violently. This behavior can also be captured and analyzed using a one-step reaction model, where the heat release is in competition with the depletion of reactants.
Hot surface ignition is examined using a glow plug or heated nickel element in a series of premixed n-hexane air mixtures. High-speed schlieren photography, a thermocouple, and a fast response pressure transducer are used to record flame characteristics such as ignition temperature, flame speed, pressure rises, and combustion mode. The ignition event is captured by considering the dominant balance of diffusion and chemical reaction that occurs near a hot surface. Experiments and models show a dependence of ignition temperature on mixture composition, initial pressure, and hot surface size. The mixtures exhibit the known lower flammability limit where the maximum temperature of the hot surface was insufficient at igniting the mixture. Away from the lower flammability limit, the ignition temperature drops to an almost constant value over a wide range of equivalence ratios (0.7 to 2.8) with large variations as the upper flammability limit is approached. Variations in the initial pressure and equivalence ratio also give rise to different modes of combustion: single flame, re-ignition, and puffing flames. These results are successfully compared to computational results obtained using a flamelet model and a detailed chemical mechanism for n-heptane. These different regimes can be delineated by considering the competition between inertia, i.e., flame propagation, and buoyancy, which can be expressed in the Richardson number.
In experiments of hot surface ignition and subsequent flame propagation a 10 Hz puffing flame instability is visible in mixtures that are stagnant and premixed prior to the ignition sequence. By varying the size of the hot surface, power input, and combustion vessel volume, we determined that the instability is a function of the interaction of the flame with the fluid flow induced by the combustion products rather than the initial plume established by the hot surface. The phenomenon is accurately reproduced in numerical simulations and a detailed flow field analysis revealed a competition between the inflow velocity at the base of the flame and the flame propagation speed. The increasing inflow velocity, which exceeds the flame propagation speed, is ultimately responsible for creating a puff. The puff is then accelerated upward, allowing for the creation of the subsequent instabilities. The frequency of the puffing is proportional to the gravitational acceleration and inversely proportional to the flame speed. We propose a relation describing the dependence of the frequency on gravitational acceleration, hot surface diameter, and flame speed. This relation shows good agreement for lean and rich n-hexane-air as well as lean hydrogen-air flames.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/GTKC-FY91, author = {Karnesky, James Alan}, title = {Detonation Induced Strain in Tubes}, school = {California Institute of Technology}, year = {2010}, doi = {10.7907/GTKC-FY91}, url = {https://resolver.caltech.edu/CaltechTHESIS:05142010-174001426}, abstract = {When a detonation wave propagates through a piping system, it acts as a traveling pressure load to the pipe wall. The detonation wave must be followed by an expansion wave in order to bring the combustion products to zero velocity at the ignition end. When it reaches a closed end-wall, a reflected shock is formed which propagates back into the tube with a decaying pressure. The present study aims to develop predictive models for the stresses and strains produced in such a situation. To this end, two series of experiments are discussed. The first series used strain gauges and a laser vibrometer to measure the elastic response of the tube to the incident detonation in thin aluminum tubes. The second series used strain gauges and high speed video to measure the plastic response of steel tubes to incident detonations and reflected shocks. In these experiments a novel mode of plastic deformation was discovered in which the residual plastic deformation in the tube wall had a periodic sinusoidal pattern.
A semi-empirical model of the pressure history was developed for use as a boundary condition in models of the mechanical response of the tube. This model was tested against experiment, and it was found that the pressure and arrival time could not be simultaneously predicted from the simple model. This and the general form of the pressure traces in the experiment seem to suggest an interaction between the reflected shock and the boundary layer behind the detonation resulting in a possible bifurcation in the reflected shock wave.
With these considerations in mind, the model was applied to single degree of freedom and finite element models of the tube wall. The ripples observed in the experiment were present in the 1-D single degree of freedom models, indicating that they are a result of the interaction of the reflected shock wave with the elastic oscillations set in motion by the detonation wave. Strain-rate hardening was found to be an important consideration under detonation loading conditions. With proper consideration of rate hardening, a single material model may be used to arrive at reasonable predictions the plastic strains resulting from detonations and reflections at initial pressures of 2 and 3 bar initial pressures.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/NRY2-2K63, author = {Hanna, Jeffrey}, title = {Solid-Oxide Fuel Cell Electrode Microstructures: Making Sense of the Internal Framework Affecting Gas Transport}, school = {California Institute of Technology}, year = {2010}, doi = {10.7907/NRY2-2K63}, url = {https://resolver.caltech.edu/CaltechTHESIS:06072010-090420975}, abstract = {Optimal electrodes for solid-oxide fuel cells will combine high porosity for gas diffusion, high phase connectivity for ion and electron conduction, and high surface area for chemical and electrochemical reactions. Tracer-diffusion simulations are used to gain a better understanding of the interplay between microstructure and transport in porous materials. Results indicate that the coefficient of diffusion through a porous medium is a function of the details of the internal geometry (microscopic) and porosity (macroscopic). I report that current solid-oxide fuel cell electrodes produced from high-temperature sintering of ceramic powders severely hinder gas transport because the resulting structures are highly tortuous, complex three-dimensional networks. In addition, poor phase connectivities will assuredly limit ion and electron transport, as well as the density of active sites for power-producing reactions. With new access to a wide range of technologies, micro- and nano-fabrication capabilities, and high-performance materials, there is a new ability to engineer the fuel cell electrode architecture, optimizing the physical processes within, increasing performance, and greatly reducing cost per kilowatt. Even simple packed-sphere and inverse-opal architectures will increase gas diffusion by an order of magnitude, and provide a higher level of connectivity than traditional powder-based structures.}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/W1NB-5W06, author = {Bane, Sally Page Moffett}, title = {Spark Ignition: Experimental and Numerical Investigation With Application to Aviation Safety}, school = {California Institute of Technology}, year = {2010}, doi = {10.7907/W1NB-5W06}, url = {https://resolver.caltech.edu/CaltechTHESIS:05272010-173243262}, abstract = {Determining the risk of accidental ignition of flammable mixtures is a topic of tremendous importance in industry and aviation safety. The concept of minimum ignition energy (MIE) has traditionally formed the basis for studying ignition hazards of fuels. However, in recent years, particularly in the aviation safety industry, the viewpoint has changed to one where ignition is statistical in nature. Approaching ignition as statistical rather than a threshold phenomenon appears to be more consistent with the inherent variability in the engineering test data.
Ignition tests were performed in lean hydrogen-based aviation test mixtures and in two hexane-air mixtures using low-energy capacitive spark ignition systems. Tests were carried out using both short, fixed sparks (1 to 2 mm) and variable length sparks up to 10 mm. The results were analyzed using statistical tools to obtain probability distributions for ignition versus spark energy and spark energy density (energy per unit spark length). Results show that a single threshold MIE value does not exist, and that the energy per unit length may be a more appropriate parameter for quantifying the risk of ignition than only the energy. The probability of ignition versus spark charge was also investigated, and the statistical results for the spark charge and spark energy density were compared. It was found that the test results were less variable with respect to the spark charge than the energy density. However, variability was still present due to phenomena such as plasma instabilities and cathode effects that are caused by the electrodynamics.
Work was also done to develop a two-dimensional numerical model of spark ignition that accurately simulates all physical scales of the fluid mechanics and chemistry. In this work a two-dimensional model of spark discharge in air and spark ignition was developed using the non-reactive and reactive Navier-Stokes equations. One-step chemistry models were used to allow for highly resolved simulations, and methods for calculating effective one-step parameters were developed using constant pressure explosion theory. The one-step model was tuned to accurately simulate the flame speed, temperature, and straining behavior using one-dimensional flame computations. The simulations were performed with three different electrode geometries to investigate the effect of the geometry on the fluid mechanics of the evolving spark kernel and on flame formation. The computational results were compared with high-speed schlieren visualization of spark and ignition kernels. It was found that the electrode geometry had a significant effect on the fluid motion following spark discharge and hence influences the ignition process.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/8F2Y-NM32, author = {Sullivan, Regina Mariko}, title = {The Physics of High-Velocity Ions in the Hall Thruster Near-Field}, school = {California Institute of Technology}, year = {2010}, doi = {10.7907/8F2Y-NM32}, url = {https://resolver.caltech.edu/CaltechTHESIS:04022010-134100215}, abstract = {
A study of the physics underlying high velocity ion trajectories within the near-field region of a Hall thruster plume is presented. In this context, “high velocity” ions are ions that have been accelerated through the full potential drop of the thruster (sometimes referred to as “primary energy” or “primary beam energy” ions). Results from an experimental survey of an SPT-70 thruster plume are shown, along with simulated data from a Hall thruster code and from a plasma sheath model. Two main features are examined: the central jet along the Hall thruster centerline, and the population of high velocity ions at high angles.
In the experimental portion of the investigation, three diagnostic instruments were employed: (1) a Faraday probe for measuring ion current density, (2) an ExB velocity filter for mapping ions with the primary beam energy, and (3) a Retarding Potential Analyzer (RPA) for determining ion energy distributions. In the numerical portion, two codes were employed: (1) a hybrid-PIC Hall thruster code known as HPHall, and (2) a model of the plasma sheath near the exit plane of the thruster, which was developed by the author.
A comparison between the measured and simulated data sets is made, to analyze the degree to which different mechanisms are responsible for the evolution of the thruster plume in the near-field region. This analysis shows that the central jet is both a function of symmetric expansion of the ion beam as well as asymmetry in the internal potential field of the thruster. Additionally, it is suggested that high energy, high angle ions could be generated given a specific internal electric field configuration, while oscillations are ruled out as the cause of these ions. The results from the sheath model show that while the sheath can change trajectory angles by 10 to 20 degrees, it can not fully explain the presence of high angle ions with high energies.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/H8JN-VS03, author = {Kao, Shannon Theresa}, title = {Detonation Stability with Reversible Kinetics}, school = {California Institute of Technology}, year = {2008}, doi = {10.7907/H8JN-VS03}, url = {https://resolver.caltech.edu/CaltechETD:etd-06022008-170629}, abstract = {Detonation propagation is unsteady due to the innate instability of the reaction zone structure. Up until the present, investigations of detonation stability have been exclusively concerned with model systems using the perfect gas equation of state and primarily single-step irreversible reaction mechanisms.
This study investigates detonation stability characteristics with reversible chemical kinetics models. To allow for more general kinetics models, we generalize the perfect gas, one-step irreversible kinetics, linear stability equations to a set of equations using the ideal gas equation of state and a general reaction scheme. We linearly perturb the reactive Euler equations following the method of Lee and Stewart (1990) and Short and Stewart (1998). Our implementation uses Cantera (Goodwin, 2005) to evaluate all thermodynamic quantities and evaluate generalized analytic derivatives of quantities dependent on the kinetics model.
The computational domain is the reaction zone in the shock-fixed frame such that the left boundary conditions are the perturbed shock jump conditions which we have derived for a general equation of state and implemented for an ideal gas equation of state. At the right boundary, the system must satisfy a radiation condition requiring that all waves travel out of the domain. Unlike the case of a single reversible reaction, in a truly multistep kinetics model, the radiation boundary condition cannot be solved analytically. In this work, we provide a general methodology for satisfying the appropriate boundary condition.
We then investigate the effects of reversibility on the characteristics of the instability in one and two dimensions. These characteristics are quantified by the unstable eigenvalues as well as the shape of the base flow and eigenfunctions. We show that there is an exchange of stability as a function of reversibility. To confirm the results our work, we have performed unsteady calculations. We show that we can match the frequencies predicted by our linear stability calculations near the stability threshold.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/W996-M220, author = {Morris, Bradford S.}, title = {Charge-Exchange Collision Dynamics and Ion Engine Grid Geometry Optimization}, school = {California Institute of Technology}, year = {2007}, doi = {10.7907/W996-M220}, url = {https://resolver.caltech.edu/CaltechETD:etd-02282007-154751}, abstract = {The development of a new three-dimensional model for determining the absolute energy distribution of ions at points corresponding to spacecraft surfaces to the side of an ion engine is presented. The ions resulting from elastic collisions, both charge-exchange (CEX) and direct, between energetic primary ions and thermal neutral xenon atoms are accounted for. Highly resolved energy distributions of CEX ions are found by integration over contributions from all points in space within the main beam formed by the primary ions.
The sputtering rate due to impingement of these ions on a surface is calculated. The CEX ions that obtain significant energy (10 eV or more) in the collision are responsible for the majority of the sputtering, though this can depend on the specific material being sputtered. In the case of a molybdenum surface located 60 cm to the side of a 30 cm diameter grid, nearly 90% of the sputtering is due to the 5% of ions with the highest collision exit energies. Previous models that do not model collision energetics cannot predict this. The present results agree with other models and predict that the majority of the ion density is due to collisions where little to no energy is transferred.
The sputtering model is combined with a grid-structure model in an optimization procedure where the sputtering rate at specified locations is minimized by adjustment of parameters defining the physical shape of the engine grids. Constraints are imposed that require that the deflection of the grid under a specified load does not exceed a maximum value, in order to ensure survivability of the grids during launch. To faciliate faster execution of the calculations, simplifications based on the predicted behavior of the CEX ions are implemented. For diametrically opposed sputtering locations, a rounded barrel-vault shape reduces the expected sputtering rate by up to 30% in comparison to an NSTAR-shaped grid.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/9JZE-X524, author = {Lieberman, Daniel Howard}, title = {Detonation Interaction with Sharp and Diffuse Interfaces}, school = {California Institute of Technology}, year = {2006}, doi = {10.7907/9JZE-X524}, url = {https://resolver.caltech.edu/CaltechETD:etd-11172005-092205}, abstract = {Detonation interaction with an interface was investigated, where the interface separated a combustible from an oxidizing mixture. The ethylene-oxygen combustible mixture had a fuel-rich composition to promote secondary combustion with the oxidizer in the turbulent mixing zone that resulted from the interaction. Both sharp and diffuse interfaces were studied.
Diffuse interfaces were created by the formation of a gravity current using a sliding valve that initially separated the test gas and combustible mixture. Opening the valve allowed a gravity current to develop before the detonation was initiated. By varying the delay between opening the valve and initiating the detonation it was possible to achieve a wide range of interface conditions. Sharp interfaces were created by using a nitro-cellulose membrane to separate the two mixtures. The membrane was destroyed by the detonation wave.
The interface orientation and thickness with respect to the detonation wave have a profound effect on the outcome of the interaction. Diffuse interfaces result in curved detonation waves with a transmitted shock and following turbulent mixing zone. Sharp interfaces result in an interaction occurring at a node point similar to regular shock refraction (Henderson, 1989). The impulse was measured to quantify the degree of secondary combustion accounting for 5-6% of the total impulse. A model was developed that estimated the volume expansion of a fluid element due to combustion in the turbulent mixing zone (Dimotakis, 1991) to predict the impulse in the limit of infinite Damkohler number.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/YSG0-TH85, author = {Pintgen, Florian Peter}, title = {Detonation Diffraction in Mixtures with Various Degrees of Instability}, school = {California Institute of Technology}, year = {2005}, doi = {10.7907/YSG0-TH85}, url = {https://resolver.caltech.edu/CaltechETD:etd-02072005-173741}, abstract = {Planar laser induced fluorescence (PLIF) is widely used in combustion diagnostics but has only recently been successfully applied to detonation. The strong spatial variations in temperature, pressure, and background composition under these conditions influence the quantitative link between OH-number density and fluorescence intensity seen on images. Up to now, this has lead to uncertainties in interpreting the features seen on PLIF images obtained in detonations. A one-dimensional fluorescence model has been developed, which takes into account light sheet attenuation by absorption, collisional quenching, and changing absorption line shape. The model predicts the fluorescence profile based on a one-dimensional distribution in pressure, temperature, and mixture composition. The fluorescence profiles based on a calculated ZND detonation profile were found to be in good agreement with experiments.
The PLIF technique is used to study the diffraction process of a self-sustained detonation wave into an unconfined space through an abrupt area change. Simultaneous schlieren images enable direct comparison of shock and reaction fronts. Two mixture types of different effective activation energy [theta] are studied in detail, these represent extreme cases in the classification of detonation front instability and cellular regularity. Striking differences are seen in the failure mechanisms for the very regular H2-O2-Ar mixture ([theta] ~ 4.5) and the highly irregular H2-N2O mixture ([theta] ~ 9.4). Detailed image analysis quantifies the observed differences. Stereoscopic imaging reveals the complex three-dimensional structure of the transverse detonation and its location with respect to the shock front. The study is concluded by using the experimentally-obtained shock and reaction front profiles in a simplified model to examine the decoupling of the shock from the chemical reaction. The rapid increase in activation energy for the H2-O2-Ar mixtures with decreasing shock velocity is proposed as an important new element in the analysis of diffraction for these mixture.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/MKP3-VC84, author = {Jackson, Scott Irving}, title = {Gaseous Detonation Initiation Via Wave Implosion}, school = {California Institute of Technology}, year = {2005}, doi = {10.7907/MKP3-VC84}, url = {https://resolver.caltech.edu/CaltechETD:etd-05242005-151253}, abstract = {Efficient detonation initiation is a topic of intense interest to designers of pulse detonation engines. This experimental work is the first to detonate propane-air mixtures with an imploding detonation wave and to detonate a gas mixture with a non-reflected, imploding shock. In order to do this, a unique device has been developed that is capable of generating an imploding toroidal detonation wave inside of a tube from a single ignition point without any obstruction to the tube flow path. As part of this study, an initiator that creates a large-aspect-ratio planar detonation wave in gas-phase explosive from a single ignition point has also been developed.
The effectiveness of our initiation devices has been evaluated. The minimum energy required by the imploding shock for initiation was determined to scale linearly with the induction zone length, indicating the presence of a planar initiation mode. The imploding toroidal detonation initiator was found to be more effective at detonation initiation than the imploding shock initiator, using a comparable energy input to that of current initiator tubes.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/2NXT-SE76, author = {Wintenberger, Eric}, title = {Application of Steady and Unsteady Detonation Waves to Propulsion}, school = {California Institute of Technology}, year = {2004}, doi = {10.7907/2NXT-SE76}, url = {https://resolver.caltech.edu/CaltechETD:etd-04222004-121013}, abstract = {The present work investigates the applications of steady and unsteady detonation waves to air-breathing propulsion systems. The efficiency of ideal detonation-based propulsion systems is first investigated based on thermodynamics. We reformulate the Hugoniot analysis of steady combustion waves for a fixed initial stagnation state to conclude that steady detonation waves are less desirable than deflagrations for propulsion. However, a thermostatic approach shows that unsteady detonations have the potential for generating more work than constant-pressure combustion. The subsequent work focuses on specific engine concepts. A flow path analysis of ideal steady detonation engines is conducted and shows that their performance is limited and poorer than that of the ideal ramjet or turbojet engines. The limitations associated with the use of a steady detonation in the combustor are drastic and such engines do not appear to be practical. This leads us to focus on unsteady detonation engines, i.e., pulse detonation engnes. The unsteady generation of thrust in the simple configuration of a detonation tube is first analyzed using gas dynamics. We develop one of the first models to quickly and reliably estimate the impulse of a pulse detonation tube. The impulse is found to scale directly with the mass of explosive in the tube and the square root of the energy release per unit mass of the mixture. Impulse values for typical fuel-oxidizer mixtures are found to be on the order of 160 s for hydrocarbon-oxygen mixtures and 120 s for fuel-air mixtures at standard conditions. These results are then used as a basis to develop the first complete system-level performance analysis of a supersonic, single-tube, air-breathing pulse detonation engine. We show that hydrogen- and JP10-fueled pulse detonation engines generate thrust up to a Mach number of 4, and that the specific impulse decreases quasi-linearly with increasing flight Mach number. Finally, we find that the performance of our pulse detonation engine exceeds that of the ramjet below a Mach number of 1.35.}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/TEZP-YC46, author = {Chao, Tong Wa}, title = {Gaseous Detonation-Driven Fracture of Tubes}, school = {California Institute of Technology}, year = {2004}, doi = {10.7907/TEZP-YC46}, url = {https://resolver.caltech.edu/CaltechETD:etd-04062004-165940}, abstract = {An experimental investigation of fracture response of aluminum 6061-T6 tubes under internal gaseous detonation loading has been carried out. The pressure load, with speeds exceeding 2 km/s, can be characterized as a pressure peak (ranging from 2 to 6 MPa) followed by an expansion wave. The unique combination of this particular traveling load and tube geometry produced fracture data not available before in the open literature. Experimental data of this type are useful for studying the fluid-structure-fracture interaction and various crack curving and branching phenomena, and also for validation for multi-physics and multi-scale modeling.
Axial surface flaws were introduced to control the crack initiation site. Fracture threshold models were developed by combining a static fracture model and an extensively studied dynamic amplification factor for tubes under internal traveling loads. Experiments were also performed on hydrostatically loaded preflawed aluminum 6061-T6 tubes for comparison. Significantly different fracture behavior was observed and the difference was explained by fluid dynamics and energy considerations. The experiments yielded comparison on crack speeds, strain, and pressure histories.
In other experiments, the specimens were also pre-torqued to control the propagation direction of the cracks. Measurements were made on the detonation velocity, strain history, blast pressure from the crack opening, and crack speeds. The curved crack paths were digitized. The Chapman-Jouguet pressure, initial axial flaw length, and torsion level were varied to obtain different crack patterns. The incipient crack kinking angle was found to be consistent with fracture under mixed-mode loading. High-speed movies of the fracture events and blast wave were taken and these were used in interpreting the quantitative data.
Numerical simulations were performed using the commercial explicit finite-element software LS-Dyna. The detonation wave was modeled as a traveling boundary load. Both non-fracturing linear elastic simulations and elastoplastic simulations with fracture were conducted on three-dimensional models. The simulated fracture was compared directly with an experiment with the same conditions. The overall qualitative fracture behavior was captured by the simulation. The forward and backward cracks were observed to branch in both the experiment and simulation.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/97GS-4N79, author = {Cooper, Marcia Ann}, title = {Impulse Generation by Detonation Tubes}, school = {California Institute of Technology}, year = {2004}, doi = {10.7907/97GS-4N79}, url = {https://resolver.caltech.edu/CaltechETD:etd-05252004-164627}, abstract = {Impulse generation with gaseous detonation requires conversion of chemical energy into mechanical energy. This conversion process is well understood in rocket engines where the high pressure combustion products expand through a nozzle generating high velocity exhaust gases. The propulsion community is now focusing on advanced concepts that utilize non-traditional forms of combustion like detonation. Such a device is called a pulse detonation engine in which laboratory tests have proven that thrust can be achieved through continuous cyclic operation. Because of poor performance of straight detonation tubes compared to conventional propulsion systems and the success of using nozzles on rocket engines, the effect of nozzles on detonation tubes is being investigated. Although previous studies of detonation tube nozzles have suggested substantial benefits, up to now there has been no systematic investigations over a range of operating conditions and nozzle configurations. As a result, no models predicting the impulse when nozzles are used exist. This lack of data has severely limited the development and evaluation of models and simulations of nozzles on pulse detonation engines.
The first experimental investigation measuring impulse by gaseous detonation in plain tubes and tubes with nozzles operating in varying environment pressures is presented. Converging, diverging, and converging-diverging nozzles were tested to determine the effect of divergence angle, nozzle length, and volumetric fill fraction on impulse. The largest increases in specific impulse, 72% at an environment pressure of 100 kPa and 43% at an environment pressure of 1.4 kPa, were measured with the largest diverging nozzle tested that had a 12 degree half angle and was 0.6 m long. Two regimes of nozzle operation that depend on the environment pressure are responsible for these increases and were first observed from these data. To augment this experimental investigation, all data in the literature regarding partially filled detonation tubes was compiled and analyzed with models investigating concepts of energy conservation and unsteady gas dynamics. A model to predict the specific impulse was developed partially filled tubes. The role of finite chemical kinetics in detonation products was examined through numerical simulations of the flow in nonsteady expansion waves.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/MAGN-R628, author = {Arienti, Marco}, title = {A Numerical and Analytical Study of Detonation Diffraction}, school = {California Institute of Technology}, year = {2003}, doi = {10.7907/MAGN-R628}, url = {https://resolver.caltech.edu/CaltechETD:etd-02122003-152525}, abstract = {An investigation of detonation diffraction through an abrupt area change has been carried out via two-dimensional, parallel simulations. The existence of critical conditions for successful diffraction is closely related to the occurrence of localized re-initiation mechanisms, and is relevant to propulsion and safety concepts concerning detonation transmission. Our analysis is specialized to a reactive mixture with perfect gas equation of state and a single-step reaction in the Arrhenius form. The concept of shock decoupling from the reaction zone is the simplest idea used to explain the behavior of a diffracting detonation front. Lagrangian particles are injected into the flow in order to identify the dominant terms in the equation that describes the temperature rate of change of a fluid element, expressed in a shock-based reference system. Conveniently simplified, this equation provides an insight into the competition between the energy release rate and the expansion rate behind the diffracting front. We also examine the mechanism of spontaneous generation of transverse waves along the front. This mechanism is related to the sensitivity of the reaction rate to temperature, and it is investigated in the form of a parametric study for the activation energy. We study in detail three highly resolved cases of detonation diffraction that illustrate different types of behavior, super-, sub-, and near-critical diffraction. We review the applicability of existing shock dynamics models to the corner-turning problem. Numerical results from the parametric study are compared with predictions from these theories in the attempt to find a formula for shock decay in a quenching detonation. This estimate is then used in the simplified temperature rate of change equation to provide a relation between critical channel width and activation energy. We conclude this study by examining the spontaneous formation of transverse waves along the wavefront of a successfully transmitted detonation. The problem is simplified to a planar CJ detonation moving in a channel over a small obstacle to investigate how acoustic waves propagate within the reaction zone. Depending on the reaction kinetics, we show that such waves may be amplified due to feedback between the chemical reaction and fluid motion. The amplification can lead to shock steepening and formation of transverse detonation waves.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/H5GV-PV33, author = {Hung, Patrick Hin Fun}, title = {Algorithms for Reaction Mechanism Reduction and Numerical Simulation of Detonations Initiated by Projectiles}, school = {California Institute of Technology}, year = {2003}, doi = {10.7907/H5GV-PV33}, url = {https://resolver.caltech.edu/CaltechETD:etd-05302003-142744}, abstract = {The evolution of a homogeneous, chemically reactive system with n species forms a dynamical system in chemical state-space. Under suitable constraints, unique and stable equilibrium exists and can be interpreted as zeroth-dimensional (point like) attractors in this n-dimensional space. At these equilibrium compositions, the rates of all reversible reactions vanish and can, in fact, be determined from thermodynamics independent of chemical kinetics.
Generalizing this concept, an m-dimensional Intrinsic Low Dimensional Manifold (ILDM) represents an m-dimensional subspace in chemical state-space where all but the m-slowest aggregate reactions are in equilibrium, and these aggregate reactions are determined by eigenvalue considerations of the chemical kinetics. In this context, a certain composition is said to be m-dimensional if it is on an m-, but not an (m-1)-, dimensional ILDM.
Two new algorithms are proposed that allow the dimensionality of chemical compositions be determined simply. The first method is based on recasting the Maas and Pope algorithm. The second, and more efficient, method is inspired by the mathematical structure of the Maas and Pope algorithm and makes use of the technique known as arc-length reparameterization. In addition, a new algorithm for the construction of ILDM, and the application of these ideas to detonation simulations, is discussed.
In the second part of the thesis, numerical simulations of detonation waves initiated by hypervelocity projectiles are presented. Using detailed kinetics, only the shock-induced combustion regime is realized as simulating the conditions required for a stabilized detonation is beyond the reach of our current computational resources. Resorting to a one-step irreversible reaction model, the transition from shock-induced combustion to stabilized oblique detonation is observed, and an analysis of this transition based on the critical decay-rate model of Kaneshige (1999) is presented.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/X7YH-T687, author = {Austin, Joanna Maria Karol}, title = {The Role of Instability in Gaseous Detonation}, school = {California Institute of Technology}, year = {2003}, doi = {10.7907/X7YH-T687}, url = {https://resolver.caltech.edu/CaltechETD:etd-05292003-150534}, abstract = {In detonation, the coupling between fluid dynamics and chemical energy release is critical. The reaction rate behind the shock front is extremely sensitive to temperature perturbations and, as a result, detonation waves in gases are always unstable. A broad spectrum of behavior has been reported for which no comprehensive theory has been developed. The problem is extremely challenging due to the nonlinearity of the chemistry-fluid mechanics coupling and extraordinary range of length and time scales exhibited in these flows. Past work has shown that the strength of the leading shock front oscillates and secondary shock waves propagate transversely to the main front. A key unresolved issue has emerged from the past 50 years of research on this problem: What is the precise nature of the flow within the reaction zone and how do the instabilities of the shock front influence the combustion mechanism?
This issue has been examined through dynamic experimentation in two facilities. Key diagnostic tools include unique visualizations of superimposed shock and reaction fronts, as well as short but informative high-speed movies. We study a range of fuel-oxidizer systems, including hydrocarbons, and broadly categorize these mixtures by considering the hydrodynamic stability of the reaction zone. From these observations and calculations, we show that transverse shock waves do not essentially alter the classic detonation structure of Zeldovich-von Neumann-Doring (ZND) in weakly unstable detonations, there is one length scale in the instability, and the combustion mechanism is simply shock-induced chemical-thermal explosion behind a piecewise-smooth leading shock front. In contrast, we observe that highly unstable detonations have substantially different behavior involving large excursions in the lead shock strength, a rough leading shock front, and localized explosions within the reaction zone. The critical decay rate model of Eckett et al. (JFM 2000) is combined with experimental observations to show that one essential difference in highly unstable waves is that the shock and reaction front may decouple locally. It is not clear how the ZND model can be effectively applied in highly unstable waves. There is a spectrum of length scales and it may be possible that a type of “turbulent” combustion occurs. We consider how the coupling between chemistry and fluid dynamics can produce a large range of length scales and how possible combustion regimes within the front may be bounded.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/83ME-8076, author = {Eckett, Christopher Adam}, title = {Numerical and Analytical Studies of the Dynamics of Gaseous Detonations}, school = {California Institute of Technology}, year = {2001}, doi = {10.7907/83ME-8076}, url = {https://resolver.caltech.edu/CaltechETD:etd-11122003-143255}, abstract = {This thesis examines two dynamic parameters of gaseous detonations, critical energy and cell size. The first part is concerned with the direct initiation of gaseous detonations by a blast wave and the associated critical energy. Numerical simulations of the spherically symmetric direct initiation event with a simple chemical reaction model are presented. Local analysis of the computed unsteady reaction zone structure identities a competition between heat release rate, front curvature and unsteadiness. The primary failure mechanism is found to be unsteadiness in the induction zone arising from the deceleration of the shock front. On this basis, simplifying assumptions are applied to the governing equations, permitting solution of an analytical model for the critical shock decay rate. The local analysis is validated by integration of reaction zone structure equations with detailed chemical kinetics and prescribed unsteadiness. The model is then applied to the global initiation problem to produce an analytical equation for the critical energy. Unlike previous phenomenological models, this equation is not dependent on other experimentally determined parameters. For different fuel-oxidizer mixtures, it is found to give agreement with experimental data to within an order of magnitude. The second part of the thesis is concerned with the development of improved reaction models for accurate quantitative simulations of detonation cell size and cellular structure. The mechanism reduction method of Intrinsic Low-Dimensional Manifolds, originally developed for flame calculations, is shown to be a viable option for detonation simulations when coupled with a separate model in the induction zone. The agreement with detailed chemistry calculations of constant volume reactions and one-dimensional steady detonations is almost perfect, a substantial improvement on previous models. The method is applied to a two-dimensional simulation of a cellular detonation in hydrogen-oxygen-argon. The results agree well with an earlier detailed chemistry calculation and experimental data. The computational time is reduced by a factor of 15 compared with a detailed chemistry simulation.}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/96F1-QR61, author = {Schultz, Eric}, title = {Detonation Diffraction Through an Abrupt Area Expansion}, school = {California Institute of Technology}, year = {2000}, doi = {10.7907/96F1-QR61}, url = {https://resolver.caltech.edu/CaltechETD:etd-11122003-180459}, abstract = {The problem of a self-sustaining detonation wave diffracting from confinement into an unconfined space through an abrupt area change is characterized by the geometric scale of the confinement and the reaction scale of the detonation. Previous investigations have shown that this expansion associated with a detonation transitioning from planar to spherical geometry can result in two possible outcomes depending upon the combustible mixture composition, initial thermodynamic state, and confining geometry. Competition between the energy release rate and expansion rate behind the diffracting wave is crucial. The sub-critical case is characterized by the rate of expansion exceeding the energy release rate. As the chemical reactions are quenched, the shock wave decouples from the reaction zone and rapidly decays. The energy release rate dominates the expansion rate in the super-critical case, maintaining the coupling between the shock and reaction zone which permits successful transition across the area change. A critical diffraction model has been developed in the present research effort from which the initial conditions separating the sub-critical and super-critical cases can be analytically determined. Chemical equilibrium calculations and detonation simulations with validated detailed reaction mechanisms provide the model input parameters. Experiments over a wide range of initial conditions with single- and multi-sequence shadowgraphy and digital chemiluminescence imaging support the model derivation and numerical calculations. Good agreement has been obtained between the critical diffraction model and experimental results.}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/GBE9-FG37, author = {Kaneshige, Michael Jiro}, title = {Gaseous Detonation Initiation and Stabilization by Hypervelocity Projectiles}, school = {California Institute of Technology}, year = {1999}, doi = {10.7907/GBE9-FG37}, url = {https://resolver.caltech.edu/CaltechETD:etd-11122003-144510}, abstract = {An experimental investigation of gaseous detonations initiated and stabilized by high-speed spherical projectiles has been carried out. Detonation initiation by projectiles is closely related to propulsion concepts such as the ram accelerator and the oblique detonation wave engine, in which, theoretically, rapid combustion occurs in detonation waves stabilized on solid objects. The criteria for initiation and stabilization by projectiles are also related to other initiation and propagation criteria such as blast initiation and failure of diffracting detonations. Experimental data of this type are useful for identifying relevant assumptions and important processes, and for providing validation for computational and analytical models.
Experiments were performed in the Caltech T5 shock tunnel laboratory. T5 was used in a shock-compression light gas gun mode, with 25.4-mm diameter nylon spheres and velocities around 2300 m/s. Gaseous mixtures studied included 2H₂+O₂+βN₂ (1 ≤ β ≤ 3.76), C₂H4+3O₂+5N₂, and C₂H₂+2.5O₂+9.4N₂ at initial pressures of 0.08 - 2.56 bar. Flow visualization results obtained by differential interferometry, shadowgraphy, and intensified CCD imaging were augmented by wall pressure records.
A wide variety of results were observed, including non-detonative shock-induced combustion, unstably initiated detonations, stabilized prompt initiations, and stabilized delayed initiations. These results can be roughly correlated in terms of the ratio of projectile velocity to mixture Chapman-Jouguet detonation speed, and the ratio of projectile diameter to detonation cell size or reaction zone thickness, although the effects of confinement and unsteadiness complicate this categorization.
Two basic approaches to modeling the results have been attempted. In the first, a global model for initiation is based on an existing blast-initiation model using the hypersonic blast-wave analogy. This model is simple, and roughly predicts the experimental results, but suffers from a number of assumptions and approximations that restrict its usefulness and accuracy. The second approach, based on the local shock curvature, is not directly capable of predicting global initiation and failure, but illustrates the mechanism responsible for decoupling of the reaction zone from the shock front in cases of detonation failure. Coupled with a separate model for the shock shape, shock-curvature theory can be used for quantitative global predictions.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, } @phdthesis{10.7907/PSTR-T717, author = {Krok, James Christopher}, title = {Jet Initiation of Deflagration and Detonation}, school = {California Institute of Technology}, year = {1997}, doi = {10.7907/PSTR-T717}, url = {https://resolver.caltech.edu/CaltechETD:etd-11122003-181337}, abstract = {We have constructed a facility for the study of jet-initiated deflagration and detonation in hydrogen-air-steam mixtures. The facility is built around two pressure vessels. Mixtures of hydrogen, oxygen and nitrogen are spark-ignited in the driver vessel, generating a hot mixture of combustion products. The pressure rise ruptures a diaphragm, venting the products into the receiver vessel through nozzles of 12.7-92 mm diameter. The receiver is filled with hydrogen-air and hydrogen-oxygen mixtures diluted with either nitrogen or steam.
The deflagration tests studied the lean and maximum-dilution limits of hydrogen-air mixtures ignited by a hydrogen-steam jet. The lean limit of 6% hydrogen was comparable to other studies. The maximum dilution limit for steam was 60%. This is higher than the limit found in spark/glow plug ignition experiments. Shock oscillations in the receiver increased with nozzle size.
Further tests studied the initiation of detonation in both hydrogen-air and stoichiometric hydrogen-oxygen-diluent mixtures. In terms of jet diameter, D, and receiver detonation cell size, λ, we found initiation limits of 2 < D/λ < 7, where other experiments required a D/λ of 11 or more. We propose that the D/λ model does not adequately characterize jet initiation, as it does not reflect the conditions in the driver.
The tests indicated that shock focusing plays an important role, promoting strong secondary explosions with or without prompt initiation of detonation. Mixtures with steam dilution were prone to DDT near the detonation limit, as the slower flame speed allows shock reflection and pressurization to occur before the reactants are consumed. Tests with nitrogen dilution had no DDT regime. Because of DDT and shock focusing, peak pressures were highest in mixtures that were slightly less sensitive than the detonation threshold. Schlieren movies confirmed the formation of a detonation near the nozzle exit.
}, address = {1200 East California Boulevard, Pasadena, California 91125}, advisor = {Shepherd, Joseph E.}, }