Development of a mechanistic model for forced convection subcooled boiling

Abstract

The complete model is then applied to simulations of subcooled boiling in nuclear reactor subchannels consistent with the operating conditions of the AP1000 pressurized water reactor. The effects of both axial and lateral nonuniform power distributions inside reactor fuel elements are accounted for. Boiling flows are simulated for three different computational domains of increasing complexity: a quarter-subchannel bordering a single fuel pin, two subchannels surround by an array of 2 by 3 fuel pins, and in four subchannels surrounded by an array of 3 by 3 fuel pins. The predicted behavior is consistent with expectations. In the 3 by 3 array, the two-phase coolant is predicted to flow from the hot channels to the cold channels, enhancing heat exchange between subchannels. This, in turn, demonstrates that the new model is capable of capturing the turbulence- and buoyancy-induced coolant mixing across the neighboring channels.

The focus of this work is on the formulation, implementation, and testing of a mechanistic model of subcooled boiling. Subcooled boiling is the process of vapor generation on a heated wall when the bulk liquid temperature is still below saturation. This is part of a larger effort by the US DoE's CASL project to apply advanced computational tools to the simulation of light water reactors.

Building on the foundation of the interfacial force model, a mechanistic model of forced-convection subcooled boiling is proposed. This model uses the heat flux partitioning concept and accounts for condensation of bubbles attached to the wall. This allows the model to capture the enhanced heat transfer associated with boiling before the point of net generation of vapor, a phenomenon consistent with existing experimental observations. The model is compared to four different experiments encompassing flows of light water, heavy water, and R12 at different pressures, in cylindrical channels, an internally heated annulus, and a rectangular channel. The experimental data includes axial and radial profiles of both liquid temperature and vapor volume fraction, and the agreement can be considered quite good.

To support this effort, the formulation of the dispersed field model is described and a complete model of interfacial forces is formulated. The model has been implemented in the NPHASE-CMFD computer code with a K-epsilon model of turbulence. The interfacial force models are built on extensive work by other authors, and include novel formulations of the turbulent dispersion and lift forces. The complete model of interfacial forces is compared to experiments for adiabatic bubbly flows, including both steady-state and unsteady conditions. The same model is then applied to a transient gas/liquid flow in a complex geometry of fuel channels in a sodium fast reactor.