Nuclear Science and Engineering / Volume 150 / Number 2 / June 2005 / Pages 182-220
Technical Paper / dx.doi.org/10.13182/NSE05-A2510
Advanced reactor system designs are being considered with liquid-metal cooling connected to a steam power cycle. In addition, current reactor safety systems are considering auxiliary cooling schemes that assure ex-vessel debris coolability utilizing direct water injection into molten material pools to achieve core quenching and eventual coolability. The phenomenon common in both applications is direct contact heat exchange. The current study focuses on detailed measurements of liquid-metal/water direct contact heat exchange that is directly applicable to improvements in effective heat transfer in devices that are being considered for both of these purposes.
In this study, a test facility was designed at the University of Wisconsin-Madison to map the operating range of liquid-metal/water direct contact heat exchange. The test section (184-cm height, 45.75-cm width, and 10-cm depth) is a rectangular slice of a larger heat exchange device. This apparatus was used not only to provide measurements of integral thermal performance (i.e., volumetric heat transfer coefficient), but also local heat transfer coefficients in a bubbly flow regime with X-ray imaging based on measured parameters such as bubble formation time, bubble rise velocity, and bubble diameters.
To determine these local heat transfer coefficients, a complete methodology of the X-ray radiography for two-phase flow measurement has been developed. With this methodology, a high-energy X-ray imaging system is optimized for our heat exchange experiments. With this real-time, large-area, high-energy X-ray imaging system, the two-phase flow was quantitatively visualized. An efficient image processing strategy was developed by combining several optimal digital image-processing algorithms into a software computational tool written in MATLAB called T-XIP. Time-dependent heat transfer-related variables such as bubble volumes and velocities, were determined. Finally, an error analysis associated with these measurements has been given based on two independent procedures. This methodology will allow one to utilize X-ray attenuation for imaging vapor bubbles with acceptable errors (bubbles ~1 to 5 cm ± 5 to 20%).
Subcooled water (Tsat - Twater [approximately equal to] 10°C) was brought into contact with liquid lead (or lead alloys) at an elevated temperature (Tlm = 500°C and Tlm - Tmelting [approximately equal to] 200°C). The study was conducted over a range of ambient pressures (1 to 10 bar) with four different water injection rates (1.5 to 8 g/s; 0.1 to 1 kg/m2s). The results showed that the system pressure has a slight effect on volumetric heat transfer coefficient, the bubble formation time, and the bubble rise velocity. Increasing the system pressure, however, resulted in an increase in the bubble average heat transfer coefficient. Increasing the water injection rate directly had only a small effect on the bubble rise velocity or formation rate. Increasing the water injection rate resulted in a decrease in the local bubble heat transfer coefficient.
Direct contact heat transfer also has some key disadvantages; e.g., flow instabilities caused by local vapor explosion is one of the issues related to direct contact heat exchange, particularly for liquid/liquid exchange with high temperature differences. In this study, the region of stable heat transfer was mapped and the effects of the liquid metal temperature, the water injection rate, and the operating pressure were investigated. The pressure required to stabilize the heat exchange process was found to be a function of the water injection rate but generally increasing the system pressure helped stabilize the system. It was also found that the larger the injection rate, the higher the pressure required to stabilize the system.