LUE - Joint Analysis for Characterization of Kinematics of droplets Superheated Patterns using iA– advanced Research on Refloodin

This doctoral project, funded by Lorraine University of Excellence (LUE), is part of a research collaboration between the LEMTA laboratory (Université de Lorraine, France) and the GOTAS research team at the São Carlos School of Engineering (University of São Paulo, Brazil), which has been ongoing since 2021. The PhD student will be enrolled at the SIMPPÉ doctoral school of Université de Lorraine and will be supervised by Prof. Michel Gradeck (UL – France), co-supervised by Dr. Guillaume Castanet (CNRS – France) and Dr. Arthur Oliveira (EESC-USP, Brazil).
The thesis focuses on understanding the transient cooling of a nuclear fuel assembly during reflooding after a Loss of Coolant Accident (LOCA). In such an event, water is injected into the reactor core, causing intense boiling and generating large amounts of steam, leading to the Dispersed Film Flow Boiling (DFFB) regime. Although extensively studied, the behavior of droplets in the post-dryout phase remains poorly understood, especially in terms of heat transfer and reactor safety margins. Developing a more reliable model for droplet-wall interactions is crucial to ensure that the reactor core does not melt, even in degraded conditions.
The objective of this thesis is to improve the understanding and modeling of droplet mass flux to the wall. This will be achieved through three main tasks:
1. Building an experimental database: A variety of droplet impacts in the Leidenfrost regime will be studied using an existing experimental setup at LEMTA. Key parameters (Weber number, droplet diameter, tangential velocity) will be analyzed to determine their influence on the thermal field of the wall.
2. Developing a deep learning algorithm: A deep learning code will be created to identify droplet-wall interaction characteristics from thermal field modifications. This will rely on well-resolved experimental data from the first task and use open-source machine learning libraries like Anomalib.
3. Application in a complex DFFB regime: A new experimental setup will be developed to observe droplet impacts on a heated TiAlN-coated wall, enabling infrared visualization. This experiment will measure droplet distribution (size and velocity) and impact characteristics using high-speed cameras and dual PDA systems, helping to estimate droplet mass flux and refine predictive models.
The thesis timeline spans three years:
• Year 1: Building the database and developing the deep learning code.
• Year 2: Setting up the new experimental system and analyzing droplet mass flux.
• Year 3: Developing a predictive model and writing scientific publications and the PhD dissertation.
The PhD candidate will join a dynamic research team in France and Brazil specializing in heat and mass transfer. Prof. Gradeck's team at LEMTA is involved in national and international programs focused on energy transition and nuclear safety. The Brazilian GOTAS team, led by Dr. Oliveira at the São Carlos School of Engineering, contributes actively to these research efforts.
The outcomes of this thesis will significantly improve the modeling of nuclear core cooling during LOCA scenarios, with potential applications in various thermal power plants.