Research Projects

The safety of nuclear power plants is of utmost importance due to the potential consequences of a nuclear accident. The release of radioactive materials can have severe and long-lasting effects on the environment, public health, and the economy, as seen in the examples of the Chornobyl and Fukushima disasters. The Passive Residual Heat Removal System (PRHR) is a crucial safety feature in nuclear power plants. In the event of a loss of coolant accident, such as a significant pipe break or a loss of power to the cooling system, the PRHR system is designed to extract avialble heat from the core of reactor, preventing damage to fuel rods and potentially catastrophic failure of the containment system. The PRHR system operates without needing external power or operator intervention, instead relying on natural phenomena such as gravity, convection, and radiation to transfer heat from the reactor. Therefore, the present project proposal addresses how to increase the boiling heat transfer on a circular vertical tube and its bundle by controlling the vapour bubble generation and implementing of machining learning approach for the same. It is vital to understand that the phenomena of vapour bubbles movement on the vertical tube, i.e., sliding behavior of vapour bubbles, is important, which significantly influences the heat emission at the top part of the tube bundle because of the vapour bubble coalesces. Hence, a novel technique, i.e., bi-philic coating (the combination of the hydrophobic and hydrophilic surface), will be applied over the tube bundle to reduce the vapour bubble coalescence at the top. Further, the bubble dynamics over the vertical tube bundle along the height is also essential for the present scenario to provide the root level knowledge about the departure characteristics such as initiation, growth, and merging, particularly for the proposed novel technique and its participation to enhance the boiling performance over single as well as the vertical tube bundle. This will be possible to introduce the most sophisticated methodology, i.e., versatile machining learning techniques. In two phase heat transfer (nucleate boiling), a machining learning technique can provide more accurate and efficient process behavior predictions by overcoming traditional modeling approaches' limitations. The complexity, as well as the involvement of large number of variables, have made predicting the two-phase process parameters challenging. Nevertheless, machine learning techniques can be applied to large datasets to detect patterns or correlations that are difficult to identify by humans. By forming machine learning models on large experimental or simulation datasets, it is possible to develop precise models to predict pool boiling parameters. It is then possible to use these models to optimize the variables influencing boiling phenomena, increase heat transfer efficiency, and reduces the risk of overheating and other safety issues.