Novel Design Integration for Advanced Nuclear Heat-Pipe Systems

Abstract

The research performed here is focused on heat-pipe modeling and simulation and performance predictions. Heat-pipes are analyzed with axisymmetric assumptions that limit the validity towards geometries that cannot be reduced to these geometries. This reduces confidence from engineers designing these products when the geometry is more novel or exotic. The research investigated the prevailing limits for a novel design configuration for a heat-pipe cooled fuel-element. This includes a more general formulation for the prevailing limits within heat-pipes. A review was performed of prior heat-pipe modeling and simulation efforts investigating the cost and benefits of various heat-pipe models. Using that review, a new method capable of quickly solving full core heat-pipe simulations with reasonable accuracy is developed. More general characteristic limits that show dependence on geometric properties such as that described in the novel design are developed. The design is analyzed showing an increase in fuel density of 17% when comparing to a specific design present in the literature. Compared to that design the novel integration approach did not significantly harm maximum performance of the heat pipe and provided support that non-standard heat-pipe geometries could provide boosts to performance that are being missed because the analytic tools are not present or are not rapid. The merits of a 3D conduction model for approximating heat-pipe solutions in a transient scenario were briefly investigated. Conduction models tend to perform well if the thermal capacity of the whole system is adequately accounted for. A review of the current modeling and simulation approaches for heat-pipes is performed. This explores the benefits and associated costs of each modeling paradigm so that readers can better inform their analytic approach. Recommendations are included to inform the use cases of these models to ensure efficient analysis. Some models miss valuable information that when each and every bit of performance is desired can cause a designer or analyst to improperly select a design or improperly evaluate safety criteria. For example, in conduction models sonic limiting behavior and capillary limiting behavior is missed because flow is entirely ignored. In standard operating regimes this may be desirable but in casualty events these limits may become important to the operation of the reactor. Some models are simply too slow to use in certain scenarios, such as design studies. Rapid design iterations require fast solutions so that many designs can be evaluated in short order. A modeling scheme was developed that could bridge the gap between speed and accuracy. Using this research a development effort resulted in a new model that can obtain accurate results for full core heat-pipe configurations quickly. A network analysis method for obtaining 3D full core temperature solutions from a simple linear system of equations is developed. The theory and performance of the method are discussed and bench marked against OpenFOAM solutions of the conduction problem. The method creates a linear system of equations based on a unit-cell configuration which allows the network approach to work. Then resistances between the points of interest and the interfaces of the neighbors are generated. Then the conservation laws are applied which are discussed in detail. These geometries being a standard cylindrical geometry from literature and the second being similar to the novel geometry described in this dissertation. What was shown is that this method provides accurate solutions and for certain geometries comes within 1% of the high-fidelity solution but in general was shown that this solution is within about 7% of the high-fidelity simulation. This is acceptable because standard experimental uncertainties are of larger order than that for thermal-hydraulic phenomena. This is for steady-state analysis but in literature, network based approaches have been shown to perform well in transients and to add transient behavior would be simple. The completion of these objectives gives a new set of possibilities for heat-pipe design in nuclear systems. Novel geometries with fast approximate simulation methods will accelerate the creation of better, more cost effective nuclear systems to compete with existing technologies. It will give additional confidence and flexibility to designers that analyze these systems.