Department of Engineering, King's College London
Academic Advisor: Dr Yutaku Kita (yutaku.kita@kcl.ac.uk)
This project investigates how the spatial distribution, orientation, and spacing of wet clothes on a standard household drying rack affect local evaporation rates. The student will build a hardware prototype to map the micro-climate surrounding the rack over time, determining the most efficient hanging configuration to reduce indoor drying times.
Hardware prototyping, sensor calibration, data logging, statistical data analysis, and empirical heat/mass transfer mapping
Basis knowledge of thermodynamics, heat and mass transfer (phychrimetrics and evaporation) and introductry coding logic.
A companion computational study to the experimental project. The student will use computational fluid dynamics to simulate natural convection and species transport (water vapor mass fraction) around simplified geometries representing wet hanging fabrics. The objective is to evaluate how localized air stagnation zones form and validate optimal spacing strategies numerically.
Advanced CFD mesh generation, multi-component fluid flow modeling, species transport simulation, and numerical validation methods.
Heat and mass transfer dundamentals, familiarity with partial differential equations, and introductory experience with CAD/CFD tools.
This project explores transient heat conduction within complex, asymmetrical geometries using an agar-gel food mock-up (representing steak/meat). The gel will be embedded with thermochromic liquid crystal (TLC) paint. When subjected to a boundary heat flux via a hot plate, the real-time color transitions will allow the student to map the progression of the thermal front and calculate thermal diffusivity.
Optical thermal diagnostics, image processing, material characterization, and analytical transient conduction modeling.
Core heat transfer fundamentals (Fourier's Law, Biot and Fourier numbers) and laboratory safety awareness.
Droplet impingement on high-temperature surfaces is a fundamental process of spray cooling for power devices, steelmaking, and emergency core cooling. This project utilizes Volume of Fluid (VOF) modeling to capture the precise phase boundary interfaces of a single fluid droplet hitting a hot solid surface. Specifically focusing on the transient thermal response of the solid substrate, the student will evaluate the effects of the Weber number, Reynolds number, substrate thermal properties on the local Nusselt number.
Optical thermal diagnostics, image processing, material characterization, and analytical transient conduction modeling.
Core heat transfer fundamentals (Fourier's Law, Biot and Fourier numbers) and laboratory safety awareness.
The katana is a legendary sword that was first used by the Japanese samurai worriers several centuries ago. It is perhaps most recognisable for its curved shape and remarkably sharp single edge. One of the secrets that make katanas unique is the coating of clay over the blade before thermal treatment: this controls the boiling transition, providing differential cooling rates between the blade and spine. Building on a COMSOL Blog post, the student will develop a high-fidelity numerical model that solves transient conduction through the clay layer, coupled with dynamic boiling heat transfer coefficient boundary conditions (switching from film boiling to nucleate boiling based on empirical correlations) at the clay-water interface.
Modeling the Differential Quenching of a Katana | COMSOL Blog
Equation-based modeling, phase-change boundary layer modeling, programming conditional logical functions within FEA software, and metallurgical thermo-mechanical analysis.
Deep understanding of the boiling heat transfer (Leidenfrost effect, nucleate boiling), basic metallurgy, and prior introductory exposure to COMSOL or FEA workflows.
This project bridges thermofluid physics and software engineering to develop an open-access, browser-based educational web application. The application will serve as an interactive visual toolkit for undergraduate engineering students. It will consist of two distinct core computation modules:
Full-stack scientific software deployment, graphical user interface (GUI/UX) design, numerical discretization of partial and ordinary differential equations, implementation of numerical stability criteria (CFL condition checking), vectorization techniques for real-time mathematical animations, and code version control (Git/GitHub).
Proficient programming foundation in Python (variables, loops, arrays, functions), strong theoretical knowledge of heat transfer, and a solid mathematical understanding of finite difference numerical schemes.
Driven by recent extreme heatwaves and shifting UK climate frameworks, this project investigates the viability of using residential rooftop solar PV to directly power domestic air conditioning (AC). The student will model the thermodynamic alignment between summer solar irradiance profiles and the cooling loads of a typical UK residential dwelling. The study will assess if peak solar generation removes the need for expensive battery storage or night-time grid reliance, factoring in the UK's unique, rapid evening temperature drops.
Solar microgeneration modeling, building transient heat load forecasting, Lifecycle Cost Analysis (LCCA), Net Present Value (NPV) calculation, and policy-level engineering strategy.
Solid understanding of solar thermal systems and building heat gains, paired with standard engineering economics knowledge.
As data centres transition from air to direct-to-chip liquid cooling to manage dense AI compute loads, infrastructure managers face immense financial and thermal trade-offs. The student will develop a system-level thermodynamic loop model of a facilities-level cooling system to analyze heat rejection rates and pumping penalties. The student will also explore the feasibility of emerging two-phase flow cooling (flow boiling, sprays).
Closed-loop thermodynamic cycle design, heat exchanger effectiveness evaluation ($\epsilon$-NTU), power usage effectiveness (PUE) mapping, and operational infrastructure asset management.
Advanced thermodynamics, heat transfer (convective heat transfer coefficients, fluid friction factors) and basic system modeling principles.
Wire Arc Additive Manufacturing (WAAM) enables the production of large-scale structural metal components. However, its economic viability is bottlenecked by process downtime. To prevent excessive heat accumulation and subsequent geometric distortion, previously deposited material must cool below a critical interlayer threshold before the next pass can begin. Under standard ambient air-cooling conditions, this waiting time can account for up to 90% of the total manufacturing cycle time, drastically inflating production costs.
This project uses computational simulation to model and evaluate advanced cooling strategies designed to mitigate this bottleneck. The student will develop a transient thermal finite element model of a multi-layer WAAM deposition process featuring a moving volumetric heat source. Using this numerical baseline, the student will evaluate and compare the thermal dissipation performance of standard natural convection (ambient air) against forced cooling techniques, specifically focusing on high-efficiency spray cooling. The spray cooling will be implemented by programming spatially and temporally varying convective heat transfer coefficients ($h$) that track behind the moving heat source.
Advanced transient finite element thermal modeling, implementation of moving heat sources, modeling of multiphase convective heat transfer coefficients ($h$), cycle-time estimation, and manufacturing production cost-benefit forecasting.
Strong foundations in transient conduction and forced convection heat transfer, and previous introductory exposure to FEA software environments.
The impingement of liquid droplets onto heated substrates is a foundational mechanism in high-flux two-phase cooling techniques such as spray cooling. The physical outcomes of these impacts depend on a complex parameter space, including fluid-solid thermophysical properties, droplet diameter, impact velocity, and initial surface temperature. Researchers traditionally map these outcomes onto a Weber number-versus-temperature phase diagram spanning distinct regimes: single-phase cooling, nucleate boiling, the Leidenfrost state, inertial breakup, and thermal atomisation. However, visual classification by human operators is inherently subjective, particularly near regime boundaries.
This project aims to develop an objective, machine learning-driven framework to automate and quantify regime identification. Utilizing an existing repository of high-speed experimental footage, the student will preprocess video frames to train a convolutional neural network (CNN) or a deep-learning video classifier. Rather than outputting a binary classification, the model will be designed to yield a probabilistic matching percentage (e.g., 70% nucleate boiling, 30% Leidenfrost) at transition boundaries. The final framework will process unseen experimental data to generate a quantitative, probability-contoured regime map.
Advanced computer vision, deep learning architecture implementation, image/video preprocessing pipelines, multi-class probabilistic classification, and the application of data science to multiphase thermofluid analysis.
Strong Python programming background (prior exposure to data science libraries like NumPy/Pandas is highly beneficial), a clear understanding of multiphase boiling curves (heat transfer), and basic knowledge of image processing.
When a liquid droplet impinges upon a heated solid surface, its hydrodynamic and thermal behaviors, ranging from nucleate boiling and splashing to the Leidenfrost state, are dictated by a complex interplay of kinetic, surface, and thermal energies. While substantial literature exists, the experimental datasets remain decentralised. The research group currently utilises a structured system of individual JSON files to store dataset variables (temperatures, velocities, Weber numbers, liquid types, and substrate properties) for individual papers, alongside a Python script to compile and map the results.
This postgraduate project focuses on scaling this existing infrastructure into a deployment-ready web application and expanding the data library. The student will write a data ingestion pipeline that aggregates the decentralized JSON files into a unified, queryable data framework. To streamline data expansion from new literature, the student will develop a semi-automated digitizing tool to extract data coordinates directly from published graphs and output them into the group's standardized JSON format. Finally, the student will build an interactive web-based dashboard using Streamlit/Plotly. Users will be able to filter by fluid-substrate combinations and dynamically select their axes (e.g., $We$ vs. $T$, or $Re$ vs. $T$) to generate custom droplet impact regime maps instantly.
Data engineering and serialization pipeline design (JSON manipulation), GUI/UX product development, software asset management, technical documentation and user guide writing, automated data harvesting/image digitization, and multi-variable thermodynamic system analysis.
Strong programming foundation in Python (specifically dictionary manipulation, file I/O, and plotting), literature survey and analysis, understanding of fluid mechanics and heat transfer, and an interest in clean software formatting.
This slot is reserved for highly motivated students who wish to pursue a specific research question, computational study, or software tool development of their own design. The proposed topic must align tightly with the supervisor’s core expertise in thermofluids, heat transfer, multi-phase flows, or thermal energy management.
To be mutually agreed upon during the scoping phase, adhering strictly to the department's assessment criteria.
The project must utilize existing institutional software licenses (Ansys, COMSOL, MATLAB) or open-source coding frameworks (Python). No bespoke hardware fabrication will be funded.
Dependent on the approved topic.
Dependent on the approved topic.