Developing a Three-Dimensional Bubble Dynamics Model for Centrifugal Nuclear Thermal Propulsion

Abstract

Nuclear thermal propulsion (NTP) is a highly efficient type of propulsion system that could yield a specific impulse up to 800 s. High-performance NTP allows for possibilities of various mission types from flyby to interplanetary flight, manned or unmanned. Centrifugal nuclear thermal propulsion (CNTP) is a type of NTP that uses liquid fissile material instead of a solid core in a standard reactor configuration. CNTP requires an interaction between liquid fissile fuel and gaseous propellant to perform heat transfer and generate thrust in a spinning enclosure. Because of the liquid state of the fissile fuel, it is essential to monitor bubble behavior and trajectory. For example, the bubble dwell time within the liquid and hydrogen flow rates can influence the liquid volume fraction and volume, which in turn can affect reactivity. A model is developed and presented that, as it matures, will predict the bubbles’ behavior based on the various types of liquid and gas and their corresponding properties as well as assist in cross-validation with the concurring experiment. The model primarily utilizes the smooth particle hydrodynamics (SPH) approach to simulate the liquid and gas interaction within the design space. The physics include buoyancy and drag on the bubbles. The pressure field within the liquid is modeled using hydrostatics, in which centrifugal motion introduces a radial gradient. Boundary conditions are devised to confine the liquid and gas within the enclosure. In this paper, we present the progress to date in two centrifugal fuel element–relevant subscale devices, the so-called “Antfarm” and “Blender II.” Antfarm is a static stage in a rectangular enclosure where buoyancy is purely gravity driven. Blender II refers to a rectangular enclosure that spins at up to 7000 rpm. Both Antfarm and Blender II are experiments that provide important data for validation against our SPH model. Two liquid and gas combinations are modeled in the present study using the Antfarm setup: water-air, and Galinstan-Nitrogen. The bubbles’ behavior was comparable to the experiment, and the velocity was at the same order of magnitude. The liquid simulation performed for the Blender II model shows that the pressure gradient within the liquid in the radial direction matches that predicted analytically from the centrifugal acceleration. The Blender II liquid model awaits experimental data for validation and verification.