Computational fluid dynamics simulations are utilized to study the flow conditions in the loop, design components to meet experimental operation targets, and examine shear stresses on the specimens to inform erosion modeling efforts which can be pursued after the experiments are conducted. Lagrangian-Eulerian coupled simulations are used to study convective mass transport to understand how the convective properties of lead differ from other coolants such as Lead-Bismuth eutectic, sodium, and water. The simulations also explore sensitivity of convective transport to surface roughness of the specimens, temperature of lead, and mean flow velocity. It is then proposed that the coupling between neutronics and mass transport can be leveraged to monitor and study corrosion of cladding materials in reactor conditions. The simulations employed a TRIGA model that was modified to be lead cooled and demonstrated that positive reactivity is added due to mass transfer corrosion which removes nickel and other absorbers from the active core. This proposed approach to be made practical, however, necessitates that reactivity and neutronics contributions from other sources (e.g. temperature distribution changes and burn up) be discerned from relatively small mass transfer contributions which necessitates advanced multi-physics coupling. A platform is developed for geometry-blind, multi-server steady state coupling of prompt neutronics (MCNP6.1) and thermal hydraulics (OpenFOAM/STAR-CCM+) which accounts for effects of power distribution from neutronics on heat transfer in the simulated system and accounts for effects of temperature of densities, surface expansion, and Doppler broadening of cross-sections through MAKXSF. As data on transport coefficients needed for mass transfer modeling and thermal hydraulics simulations, particularly in oxide layers that form with varying thickness and composition, remains scarce, molecular dynamics simulations were proposed as part of the framework. Non-equilibrium molecular dynamics simulations are preferred for their ability to study size effects at the nano and low-micron scales. As an effort to improve the reliability of molecular dynamics simulations, the method of Shannon entropy for convergence assessment of the fission source distribution in Monte Carlo neutron transport in MCNP is adapted and introduced to molecular dynamics. Application of the approach to simulations of thermal conductivity, radiation damage, and fluid flow shows potential for generalization to other areas of molecular dynamics.

]]>Two types of filament wound/CVD carbon carbon composites were studied. A standard filament wound/CVD carbon composite, and a composite similar to the standard processed composite; but having short chopped carbon fibers sprayed on the substrate during the winding process. Volume fraction of open cell porosity was measured, and pore geometry, size, and orientation were determined.

A linear relationship was observed for the logarithm of transverse tensile strength as a function of volume fraction porosity. However, a considerable difference in the slopes of the lines for sprayed and unsprayed materials was observed, indicating the effect of a parameter other than volume fraction porosity. This additional parameter is pore size distribution.

The experimentally obtained data was also compared with theoretical models based on pore geometry, size and orientation and on fracture mechanics.

The agreement between the experimental results and the pore size geometry and orientation model is good. The agreement between experimental data and the fracture mechanics model is poor.

The transverse tensile strength of filament wound/CVD carbon carbon composites is drastically reduced by increasing volume fraction porosity. In addition, parameters such as pore size distribution, orientation, and geometry are also important, and should be considered when studying porosity effects.

]]>Recent thermal-hydraulics studies have demonstrated that iron-chromium-aluminium (FeCrAl) alloys have the thermal priorities over zircaloys and other commercial alloys including critical heat flux and heat transfer coefficient. However, it is found in our experimental results that FeCrAl-C26M, and FeCrAl-C36M have higher critical heat fluxes and heat transfer coefficients than zircaloys and stainless steels while FeCrAl- B126Y and FeCrAl-B136Y have lower critical heat flux than zircaloys. This speaks to the possibility that the superiority of accident tolerant fuel claddings depends on which alloys they belong to. However, the difference gap of critical heat flux and of heat transfer coefficient between various cladding materials can be suppressed by high mass flux and/or high inlet subcooling. The one possible mechanistic rationale behind this is that the material-side factors related with near-field mechanisms such as surface wettability/roughness, and thermal-physical properties, compete with the far-field heat convection mechanisms dominated by mass flux and inlet subcooling. Besides the steady-state boiling experiments, the FeCrAl and zircaloy claddings are subjected to the power transient heat inputs of linear ramp and Fuchs-RIA. In comparison with the steady-state flow boiling, the power transient maximum heat flux is larger than the steady-state critical heat flux and so is the power-transient heat transfer coefficient. As a result, it implies that the power transient state of light water reactors gives higher thermal safety margin than their nominal steady states. The experimental results of Fuchs-reactivity initiated accident power transient across a wide range of inlet subcoolings and mass fluxes give a solid confirmation to that the difference gap of power transient maximum heat flux between various transient timescales and cladding materials can be suppressed by the increasing of mass flux and inlet subcooling. This connotes that the power-transient boiling heat transfer may be dominated by two completely distinct mechanisms, heat conduction between cladding solid and water coolant competing with heat convection that are contributed by mass flux and inlet subcooling.

]]>A pressure transducer is used to monitor the pressure in a sealed graphite-lined volume containing the fuel sample during an ACPR transient. The energy input history to the sample is derived from the reactor power history, fission product inventory measurements, and neutron transport calculations. Energy losses from the sample are determined by linear inverse heat conduction techniques using temperature histories derived from thermocouples located in the walls of the confining pressure cell. Because of nonuniformities in the distribution of fission energy deposition within the sample and because of the energy losses from the sample, it is not possible to define a unique energy content history. It is possible, however, to define bounds on the energy content. Bounds on the vapor pressure (formulated as pressure as a function of energy content) are derived through a point by point comparison of the pressure history with the bounds on energy content.

The EEOS technique was used to determine the vapor pressure of high purity uranium dioxide with an oxygen-to-metal ratio of 2.08. Energy depositions of up to 2720 J/g relative to room temperature were obtained, resulting in measured pressures in excess of 38 MPa. Bounds on the vapor pressure of UO_{2.08} were found to be:

P(MPa) = exp [(10.004 ± 1.497 - (13.365 ± 2.535)/E(kJ)]

(upper pressure bound for 1.4 kJ ≤ E ≤ 1.9 kJ), and

P(MPa) = exp [(10.894 ± .198 - (18.298 ± .386)/E(kJ)]

(lower pressure bound 1.6 kJ ≤ E ≤ 2.0 kJ ). These results bound previous high temperature data reported by other investigators.

]]>The model is investigated for three reactor-reflector configurations; the bare reactor, the reactor with a radially symmetric steel reflector, and a reactor with a radially symmetric polyethylene reflector. This investigation includes the results of a calculational study using Monte-Carlo techniques and an experimental program using Rossi-alpha, pulsed neutron, and reactor burst techniques on the Sandia Pulsed Reactor SPR II.

A Type I (all counts and all times) multichannel time analyzer, which was designed and built to carry out the Rossialpha experiments, is described. Results of a supplementary experiment show the biasing of the prompt neutron decay constant when measured with a Type III (count-to-count) multichannel time analyzer.

Rossi-alpha measurements at subprompt critical reactivities are used with limited success to predict reflector effects on the super prompt critical behavior of a fast burst reactor. It is considered that significant improvements in the predictions should be attainable with additional development of the Rossi-alpha or similar techniques.

]]>In an attempt to provide such a description, that class of high-current, cold-cathode diode distinguished by non-self-convergent electron flow has been studied to (1) define the operative electron emission mechanisms, (2) to determine the dominant plasma phenomena within the interelectrode volume, (3) to classify the modes of electron flow, and (4) to verify the Friedlander beam convergence criterion. A basis for unfolding these elements of observed diode behavior was developed from a detailed analysis of the time dependent, voltage-current characteristics of the diode. Since the electron flow throughout the predominant portion of the accelerating pulse was space-charge limited, separation of the interrelated diode phenomena was facilitated by representing the voltage current data in terms of the diode perveance. To confirm certain of the assumptions fundamental to the interpretation of the perveance data, streak photography of plasma luminosity motion within the diode was combined with measurements of the time-integrated as well as time-dependent current density distribution at the anode plane. When correlated with the theory of high-voltage vacuum breakdown formulated by Dyke, Charbonnier, Mesyats, and others, the results of this diode study provide a complete and self-consistent description of the dominant processes acting within a high-current, cold-cathode diode.

Within this model, six distinct phases of diode response can be identified. If a high-current, relativistic electron accelerator is to generate the high-power electron beam for which it was designed, the diode must evolve through at least the first four of the six phases. In sequential order, these phases are as follows:

1. Stable field emission from microscopic, whisker-like cathode projections

2. Transition to cathode-initiated vacuum breakdown

3. Whisker explosion and the formation of a cathode plasma sheath

4. Space-charge-limited emission from the expanding plasma sheath

5. Transition to self-convergent flow

6. Development of a highly convergent electron flow

]]>We derive space, angle, and time-dependent single chain a source equations for the cumulative energy deposition distribution (the FPDF) in a system via the backward Master equation formulation; from which, equations of the moments are also derived. This new formulation has the benefit of not requiring knowledge of the neutron number distribution. We then compare results of the EBMC method with the direct numerical solution of the moment equations and show excellent agreement. We then show that by altering the induced fission energy deposition distribution, the first four moments are virtually the same for supercritical systems. It is shown that the FPDF itself does indeed have noticeable alterations in the high energy deposition tails of the distribution, suggesting that one may need to consider higher order moments in order to witness a noticeable difference in the respective profile. It was also shown that the multiplicity distribution model being used, where we compared the full distribution with the MCNP mean-preserving model, has an effect on the higher energy deposition region of the single chain FPDF.

Finally, we formulate the Boltzmann Master equation- a novel nonlinear adjoint transport equation satisfied by the neutron number density distribution. In a lumped system setting, we consider several numerical discretization schemes for the number distribution, which show that typical basis and test functions used in transport methods are not robust. We apply the collocation method as well as derive an analytical generalization of Bell's distribution via solution ansatz. We then expand our scope to include space and angle dependence, derived systems of equations for the aforementioned discretization schemes, and compared the results, showing excellent agreement for long enough times in supercritical systems with the Quadratic Approximation applied.

]]>Like most Gen IV reactor concepts, there are a few problems which need to be addressed with the design before it can be licensed. One particular design is the Mark 1 pebble bed FHR, a design put out by the University of California Berkeley. Two problems that need to be addressed are due to characteristics of the choice of molten salt. The main choice is a mixture of lithium fluoride and beryllium fluoride (flibe). This work attempts to analyze the heat transfer performance of a double wall twisted tube heat exchanger which could play a part in the solution to these problems.

Particularly, the focus of the work attempts to understand the performance of a three fluid, parallel stream heat exchanger with two thermal communications with an emphasis on understanding the effect the intermediate fluid and it's flow rate has on the overall effectiveness of the heat exchanger.

]]>Three nuclear rocket designs were used to compare startup control strategies. The three designs were the NERVA (nuclear engine for rocket vehicle application) small engine, wire core, and particle bed reactors. These three reactors were chosen to represent the spectrum of possible solid core reactor types which had been considered by NASA (National Air and Space Administration) for the Space Exploration Initiative.

Models exist in the literature for both the NERVA small engine and particle bed reactors. A comparable model of the wire core reactor was unavailable. Therefore, prior to assessing the impact of nuclear rocket engine design on control strategies. modeling of the wire core reactor was performed.

Development of a computer code that modeled the physical phenomena of a nuclear rocket engine startup was required to compare startup strategies for the three reactor types. The STRTUP computer code which was developed, modeled the important time-dependent effects that occur in a fission reactor as a result of the physics of fission.

Also modeled by the code were the core heat transfer process, a turbopump. and reactivity feedback effects. In the derivation of the non-dimensional Smith-Stenning equations which were coded into the STRTUP computer code, response time constants were defined for reactor power. delayed neutron precursor density, core outlet temperature, and core inlet pressure. The reactor power and core outlet temperature response time constants exhibited the greatest variance between the three reactors. When comparing the startup simulation results between the three reactor models, it was found that the following parameters varied significantly as a result of differences in engine design and control strategies:

a. Startup response time for power level, reactor outlet temperature, and turbopump outlet pressure

b. Propellant reactivity feedback effects

c. Temperature reactivity feedback effects

d. Control drum span startup requirement

e. Control drum reactivity insertion and removal rate requirement.

However. from a reactor kinetics and engine controllability standpoint there was only a small difference found between the NERVA and particle bed reactors. The wire core exhibited a substantial difference with the other two designs primarily due to variance in the startup response time for power level and propellant reactivity feedback effects.

]]>In this dissertation, two models of a particle-loaded shield are investigated. The first is an artificial material postulated to consist of constant-thickness slabs of infinite lateral extent, which are called "particles", randomly embedded in a matrix material. Precise analytic results are obtained for the case in which particle and matrix are pure absorbers. The second is a more realistic model, consisting of constant radius spheres randomly embedded in a matrix material. A precise treatment of this model is extremely difficult; hence, the author examines two approximate schemes which are advantageous from a computational standpoint. The first approximate scheme has the characteristic that, in the limit of low volume-percent loadings, it approaches the rigorous solution. The second approximate scheme, valid when the first scheme is inadequate, provides a useful method of evaluating the behavior of highly loaded materials. Again, both particles and matrix are taken to be pure absorbers. The sphere-loaded problem is investigated by employing Monte Carlo procedures to simultaneously construct transport particle histories and the structure of the medium encountered during these histories.

For the pure absorption problem, it is found in both of the above models that for shields which are very thin relative to particle dimensions, the average transmission initially decreases in agreement with the homogeneous shield assumption. As the shield thickness increases, however, the average transmission can depart very quickly from the behavior one would calculate by assuming the shield to be a homogeneous mixture of particle and matrix material.

This departure asymptotically results in an exponential behavior governed by an effective constant cross section which is characteristic of the loaded shield and which is lower than the cross section associated with the homogeneous assumption. It is also observed that for the pure absorption problem, the behavior of the expected value of higher moments of the transmission is no more difficult to obtain than the first moment.

Calculational results are presented and are found to compare favorably with the limited amount of available experimental data. The extension of the developed computational techniques to transport problems which include scattering is then outlined. In the final chapter, the author discusses the implication of his results upon both experimental and additional theoretical work on transport theory in realistic particle-loaded media.

]]>The VSLLIM reactor core is loaded with hexagonal assemblies of 19 UN fuel rods clad in HT-9 steel and with scalloped BeO walls, clad also in HT-9 steel. In addition to helping achieve an almost uniform flow distribution in the fuel assemblies, the scalloped BeO walls, together with the BeO axial and radial reflectors, increase the hot-clean reactivity for achieving long full-power operation life, at a relatively low UN fuel enrichment. During nominal operation, the inlet and exit coolant temperatures in the reactor core are maintained at 610 K, and < 820 K to minimize embrittlement and corrosion of the HT-9 steel cladding, core structure, and reactor vessel by liquid sodium.

This research conducted neutronic and thermal-hydraulics analyses to investigate and quantify the passive operation and safety features of the **V**ery-**S**mall, **L**ong-**LI**fe, and **M**odular (VSLLIM). The research tasks carried out include:

(a) Performing neutronic analyses of the VSLLIM to investigate the effects of several design and material choices on the cold and hot-clean reactivity, for achieving long operation life without refueling. Also calculate the cold-clean reactivity shutdown margins of the emergency shutdown system (ESS) and reactor control (RC), and the beginning of life (BOL) hot-clean reactivity. Investigated are the effect of UN fuel enrichment, and the material of the axial and radial reflectors. These analyses calculated the neutron energy spectrums and the radial and axial fission power distributions. These are in addition to determining the temperature reactivity feedback effects due to the UN fuel, sodium coolant, HT-9 steel cladding and core structure, BeO in the driver core and axial radial reflectors, and the Doppler broadening of neutron cross-sections. To estimate the full-power operation lives of the VSLLIM reactor at different thermal powers, fuel depletion calculations are carried out at hot-clean operation condition.

(b) Performing thermal-hydraulic and computational fluid dynamics (CFD) analyses during nominal reactor operation and after shutdown. These analyses estimated the friction number for the liquid sodium flow in the core hexagonal UN fuel assemblies, with scalloped wall as a function of the flow Reynolds number. The flow and temperature distributions in the UN fuel assemblies are calculated, at different reactor thermal powers and inlet and exit core temperatures of 610 K and < 820 K, respectively.

(c) Performing CFD analyses to quantify the passive decay heat removal by natural circulation of ambient air along the outer surface of the reactor guard vessel. Both, after reactor shutdown and in the case of a malfunction of the in-vessel Na-Na HEX.

The performed neutronics and fissile depletion analyses of the VSLLIM confirmed that a UN fuel enrichment of 13.76% is sufficient for achieving high enough hot-clean excess reactivity for operating VSLLIM reactor without refueling for ~92 and 5.8 full power years (FPY) at 1.0 and 10 MW_{th, }respectively. Results confirmed sufficient cold-clean reactivity shutdown margins using either the reactor control (RC) or the emergency shutdown system (ESS), independently. In addition to having two independent reactor shutdown systems, results show that the negative temperature reactivity feedback is capable of shutting down the reactor with a modest increase in the temperatures of the UN fuel and the in-vessel liquid sodium. Results also show that the neutron energy spectrum in the VSLLIM reactor core is hard, which reduces the inventory of minor actinides in UN fuel during reactor operation. Because of its low operating temperatures, < 812 K at 10 MW_{th} and UN fuel low average power density (< 23.47 MW_{th}/m^{3}), the fuel in the VSLLIM core experiences practically no swelling and retains practically all fission gas release.

The performed CFD-thermal-hydraulics analyses investigated the effects of using metal fins along the outer surface of the guard vessel wall and changing the width of the cold air intake duct on the decay heat removal rate and the time after shutdown for cooling the in-vessel liquid sodium to 400 K. Results show that without metal fins, the heat removal rate of the decay heat is 244 kW_{th} immediately after shutdown. However, within 8 minutes after shutdown, natural circulation of ambient air along the outer surface of the guard vessel removes more heat than is being generated by radioactive decay in the core. Consequently, the average in-vessel sodium temperature drops from 682 K to 400 K, within 120 hrs after shutdown.

Using metal fins along the outer surface of the guard vessel increases the total area for heat transfer and the decay heat removal rate by 13.5%, reducing the time for the average temperature of the in-vessel liquid sodium to decrease to 400K in ~100 hrs. Without metal fins, reducing the width of the cold air intake duct by 50% decreases the decay heat removal rate by 35%, increasing to ~ 346 hrs the cooling time of the in-vessel sodium to 400 K. These results demonstrate that the decay heat removal by natural circulation of ambient air from the outer surface of the reactor guard vessel wall, with and without metal fins, is quite effective. Additionally, the results show that the temperature of the in-vessel sodium after shutdown remain safely, ~470 K below its boiling point (~1156 K at 0.1 MPa).

The performed CFD analyses investigated the friction factor for laminar, transition, and turbulent flows in hexagonal bundles of bare tube and also investigated with flat and scalloped walls. The results for the bundles with flat walls are in good agreement with previously reported experimental data by others. The CFD results and the reported experimental data for bundles with flat walls and various numbers of tubes, (7, 19, 37, and 61), in a triangular lattice with 1.07 < P/D < 2.416, are used to develop a continuous correlation of the friction factor as a function of P/D and Re_{in}. The developed correlation, for P/D up to 3.0, and a wide range of tubes in the bundles (N = 7 – 331), spans all three flow regions (10^{2} < Re_{in} < 10^{6}) and is in excellent agreement with the compiled numerical and experimental database. The results also validated the applicability of the developed friction factor correlation to the VSLLIM reactor hexagonal bundles with scalloped walls.

The developed continuous friction factor correlation is within 5% - 8% of the CFD data generated for the scalloped walls bundles with 19 and 37 rods or tubes. Results also showed that scalloped walls reduce the bypass flow next to the wall, while increasing the flow in the interior subchannels in the bundles, thus improving heat transfer. Higher flow in interior subchannels enhances the thermal-hydraulics in the VSLLIM reactor core and reduces the maximum temperature at a given Re_{in}.

Thin target excitation functions were determined for (^{3}He,p) reactions with carbon and oxygen over a ^{3}He energy range from 2.5 to 9.0 MeV. These data were used to calculate the activation curves for various combinations of incident particle energy, impurity distributions, and material.

Information on several computer codes used during this study is presented.

]]>