| A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | PROPOSAL | |||||||||||||||||||||||||
2 | Thematic | Title | Short description and objectives | Contact person | ||||||||||||||||||||||
3 | Proposal research area | The thesis aims at developing an optimized sample holder for testing PEM/AEM low-temperure electrolysis cell component. | HyRES contact person | |||||||||||||||||||||||
4 | ||||||||||||||||||||||||||
5 | Testing / Engineering | Design and Development of a AEM Cell Holder for Material Testing in real operating conditions | The thesis aims at teveloping an optimized smaple holder to test AEM/PEM low-temperature electrolysis cell components. The activities will include: research of the state-of-the-art for PEM/AEM electrolysis cell for testing activities; identification of the electrochemical and mechanical requirements for the prororype development; cell design and realization of a CAD model; (depending on the time available) testing of the realised cell. | Simone Zottele | ||||||||||||||||||||||
6 | Modelling | Dynamic Modelling in Modelica of the Coupling between a Reversible Solid Oxide Cell (rSOC) System and a Small Modular Reactor (SMR): Development and Experimental Validation within Open-Source FBK Libraries [DOLOMIA] | This thesis addresses the dynamic modelling of the coupling between a reversible Solid Oxide Cell (rSOC) system and a Small Modular Reactor (SMR) within a nuclear Hybrid Energy System (HES), using the open-source DOLOMIA Modelica library developed at FBK. Building on the existing High Temperature Steam Electrolysis model, the work models the complete rSOC system rather than the stack alone, including heat exchangers, blowers, recirculation strategy, steam generator and condenser. It supports full reversible operation across SOFC and SOE modes and across nominal, partial-load, standby and night-mode conditions, with metal hydride storage implemented to increase the flexibility of hydrogen production. The stack incorporates activation and concentration overvoltage losses with mode-dependent parameterisation, while a PID-based control architecture acting on the entire Balance of Plant ensures safe operating conditions for pressure, temperature and hydrogen recirculation, so as to prevent degradation phenomena. A dedicated experimental campaign at FBK's hydrogen electrolyser facility will provide the data required to validate the rSOC model, including operational transients and steady-state operation in the endothermic and thermoneutral regions, with particular attention to reversible operation. The validated model will support the analysis of transient behaviour and flexibility constraints of the integrated system, assessing the capability of open-source Modelica libraries to represent the dynamic coupling of hydrogen production with SMR thermal output in realistic HES configurations. | |||||||||||||||||||||||
7 | Modelling | Model validation with MiL approach | This thesis unlocks accurate models for fuel cells and electrolyzers, crucial for clean energy's future. Leveraging MIL, SIL, and HIL approaches, it delves into: • Developing high-fidelity dynamic models using cutting-edge tools. • Integrating models with control algorithms in simulated environments (SIL) for control performance assessment. • Coupling models with real hardware (HIL) for true-to-life validation under dynamic conditions. | Michele Bolognese/ Pietro Iob | ||||||||||||||||||||||
8 | Experimental Characterization of 6 cell stack of DA-SOFC for stationary application | Within AMON project framework, the master student will conduct a series of experimental tests on a 6cell stack of direct ammonia fuel cell to characterize the performance map from static to dynamic operation. The work will cover the design of experiment and data postprocessing and analysis of 6-cell with NH3. | ||||||||||||||||||||||||
9 | Modelling | Modeling and simulation of SOE and thermochemical compression by metal hydirde | This master's thesis presents the development of a numericla model for the simualtion and coupling of a SOE for high efficient Hydrogen production and a metal hydride system for thermo-mecchanics compression | Michele Bolognese | ||||||||||||||||||||||
10 | Optimization of the formulation and deposition process of anodic and cathodic catalytic inks (PEM/AEM electrolysis). | The first phase of the work concerns the optimization of the recipes for the production of PEM/AEM catalytic inks starting from commercial catalytic powders up to catalysts produced internally. Subsequently, the ink deposition techniques (drop casting, Mayer rod and spin coating) on substrate (e.g. glassy carbon) will be optimized by SEM-EDX measurements, profilometry and spot tests. The second phase of the work involves the evaluation of the electrochemical performances (activity and stability of the catalysts) of the deposited catalysts by characterization in a three-electrode cell (flat electrode) through linear sweep voltammetry, cyclic voltammetry, chronopotentiometry and electrochemical impedance spectroscopy. | Domenico Dalessandro | |||||||||||||||||||||||
11 | Engineering | Development of Optimized Control Solutions for a Hydrogen Test Bench for Electrolysis | This master’s thesis focuses on the development and implementation of an optimized control system for a test bench designed to evaluate the performance of Anion Exchange Membrane (AEM) electrolyzers. The project aims to enhance the efficiency, reliability, and automation of hydrogen electrolysis testing by integrating advanced control strategies. This thesis offers a unique opportunity to work at the intersection of hydrogen technology, control engineering, and system optimization, contributing to the advancement of green energy solutions. | Pietro Iob / Alberto Vacilotto | ||||||||||||||||||||||
12 | Testing (high temperature) | Electrochemical characterization of reversible SOFC / SOEC | The thesis work focuses on the characterization of SOFCs / SOECs both in terms of performance evaluation and durability test. Reversible Solid Oxide Cells (fuel cells SOFCs; electrolyzers SOECs) represent one of the most promising technologies for power-to-power applications thanks to their ability to work at high temperature where the efficiency of the electrochemical process is maximized. However, their harsh working environment can lead to long-term degradation due to the combination of high temperatures and reactive atmospheres. Assessing the durability of cell prototypes at the laboratory scale and, in case, identifying the degradation mechanisms and their possible solutions are key factors towards the broad scale application of Solid Oxide Reversible Cells. | Emanuele De Bona | ||||||||||||||||||||||
13 | Material Science / Materials Development | Plasma Vapour Deposition-Assisted Fabrication of Miniaturized Thin Film Components for Solid Oxide Electrolysis (SOECs) | High temperature electrolysis is nowadays regarded as one of the most promising technologies for power-to-power applications. The high working temperatures lead to optimal electrochemical efficiency but also accelerated degradation of the cells. One of the main causes of degradation is the interdiffusion of elements between the steel and the ceramic components. Such process can both lead to cell "poisoning" (i.e., decreased reactivity and performance degradation) and steel damaging (embrittlement, thermophysical properties degradation). The main solution is the application of a protective layer to prevent diffusion without affecting the electrochemical properties of the whole composite. Such layer should ideally be as thin as possible, and is currently applied through a variety of techniques such as screen printing or spray coating. Physical Vapor Deposition (PVD) techniques, such as plasma sputtering, offer precise tuning of thin-film properties (structural, electrical, thermal, etc.), allowing precise control over the film's growth and composition. | Emanuele De Bona + Sandro Zorzi | ||||||||||||||||||||||
14 | Control | Automatic Generation of JSON-Based Hydrogen Control Strategies from Natural Language Requirements | The objective of this thesis is to investigate a methodology for the automatic generation of JSON-based control and automation strategies from functional requirements expressed in natural language. The work will focus on hydrogen control systems in which automation logic, interlocks, and operational strategies are described through structured strings configuration files validated by a JSON schema and executed by a backend control engine. The thesis will address the creation of a supervised dataset composed of requirement–configuration pairs, the formalization of the target JSON structure, and the generation of both manually defined and synthetic examples. Based on this dataset, statistical and NLP-based models will be trained or evaluated to generate valid control strategy configurations from human-readable specifications. The generated configurations will be assessed through syntactic and structural validation against the JSON Schema, as well as through functional evaluation to verify the consistency between the input requirement and the resulting control behavior. The expected outcome is a prototype configuration assistant capable of supporting non-expert users in the definition of industrial automation strategies, reducing configuration errors, development time, and dependency on direct source-code modifications. | Alberto Vacilotto | ||||||||||||||||||||||
15 | Material Science / Materials Development | Reduction of Platinum Group Material Through Atomic Layer Deposition for Sustainable Green Hydrogen Production | The hydrogen economy requires efficient, durable, and affordable electrocatalysts for water splitting. PEM electrolysis relies on scarce, expensive platinum group metals (PGMs) like iridium and platinum, making PGM load reduction a central challenge. Atomic Layer Deposition (ALD), a vapor-phase technique enabling conformal, atomic-scale thin-film growth on complex 3D geometries, remains underexplored for electrocatalyst fabrication despite its precision and versatility. ALD allows precise nanostructuring of PGMs and their heterostructuring with protective oxides/nitrides (TiOx, TiN), preventing degradation mechanisms such as aggregation, coalescence, and dissolution. This project leverages FBK's ALD infrastructure to deposit catalytically active Ir, Pt, and/or oxide coatings heterostructured with protective materials on porous substrates, targeting high-performance OER and HER electrodes, with results aimed at peer-reviewed publication. | Matteo Bordin | ||||||||||||||||||||||
16 | Material Science / Materials Development | Physical Vapor Deposition of High-Entropy Oxide and Alloy Thin Films for Electrocatalytic Hydrogen Production | High-entropy materials (HEMs), alloys (HEAs) and oxides (HEOs) are multi-principal-element compounds where five or more near-equimolar elements within a single-phase structure generate configurational entropy stabilization, enhanced corrosion resistance, and synergistic catalytic effects absent in conventional binary or ternary systems. These properties make HEMs attractive next-generation electrocatalysts for water splitting, where activity and stability under harsh electrochemical conditions are essential. RF magnetron sputtering offers precise compositional control and flexible nanostructure tailoring for HEA/HEO thin-film synthesis. FBK's multi-cathode PVD system — supporting co-deposition from independent elemental targets, reactive sputtering in oxidizing atmospheres, and plasma pre-treatment for surface activation — enables broad exploration of the synthesis parameter space toward exotic electrocatalytic morphologies. This project will synthesize and characterize high-entropy thin-film electrocatalysts via PVD, investigating their structural stability and OER/HER activity, iteratively optimizing deposition toward high-performance, CRM-lean electrodes, with results targeting a peer-reviewed publication. | Sandro Zorzi | ||||||||||||||||||||||
17 | Modelling | Reduced Order Model of a multispecies mixer for a turbine. | This thesis focuses on the development of a Reduced Order Model derived from high-fidelity CFD simulations of a multispecies mixer, with particular emphasis on transport, mixing efficiency, and parametric variability. The work will start from the analysis and preprocessing of CFD simulation data describing the multispecies transport within the mixer under different operating conditions. On this basis, reduced-order modeling techniques will be investigated and implemented, this may include more classical projection-based approaches such as Proper Orthogonal Decomposition, Radial Basis Functions, and Dynamic Mode Decomposition, as well as data-driven and physics-informed (scientific machine learning) methods based on neural networks. The objective of this work is to construct a computationally efficient surrogate model that preserves the essential physical behavior of the mixer while enabling fast evaluations across a wide range of operating and geometric conditions. | - Luca Pratticò - Michele Urbani - Gregorio Casagrande | ||||||||||||||||||||||
18 | Modelling | Scientific Machine learning applied to Multiphase Reactive flow through porous media for Low-Temperature Electrolyzers. | This thesis aims to develop scientific machine learning techniques as a surrogate modeling framework to simulate multiphase reactive flow in porous domains. The objective is to combine CFD-generated data and physical conservation laws (mass, momentum, and species transport) to create data-efficient models capable of generalizing across operating conditions. | Luca Pratticò - Gregorio Casagrande | ||||||||||||||||||||||
19 | Modelling | Advanced Domain Adaptation Techniques for Degradation Modelling in PEM Fuel Cells | This project aims to develop advanced machine-learning techniques for modelling PEMFC degradation across heterogeneous operating conditions. Building on a previous internship that introduced metric-based domain adaptation, this work will investigate more robust approaches such as adversarial DA, multi-domain learning, and modern sequence models. The objective is to improve generalization across purge, recirculation, and humidity regimes. The outcome will be a scalable modelling framework suitable for integration in PEFC diagnostic and prognostic workflows. | Luca Pratticò — FBK HyRES Andrea Gobbi — FBK DSIP Gregorio Casagrande — FBK HyRES | ||||||||||||||||||||||
20 | Modelling | Data-Driven Reduced-Order Modeling for CFD-Based Hydrogen Leakage in Open or Closed Environment | This thesis focuses on the construction of reduced-order models derived from high-fidelity CFD simulations of hydrogen leakage and dispersion in open or closed environments. While CFD provides accurate descriptions of species transport phenomena, its computational cost prevents extensive parametric exploration and real-time usage. The overarching objective is to transform detailed CFD datasets, governed by species transport equations, into computationally efficient surrogate models that retain the essential physical behavior while enabling fast evaluation across a wide range of operating and boundary conditions. The modeling focus concern concentration fields, temporal evolution, and the role of boundary conditions. Different reduced-order modeling paradigms may be investigated and chosen, including projection-based, data-driven or physics constrained such as physics informed neural networks (PINNs) that embed governing transport equations into the modeling process. | Luca Pratticò — FBK HyRES Gregorio Casagrande — FBK HyRES | ||||||||||||||||||||||
21 | Modelling | Optimal sensor layout for hydrogen detection | This thesis focuses on the development of a methodology for the optimal placement of hydrogen sensors in open or confined environments, with the aim of enabling early, reliable, and robust detection of accidental leaks. The work assumes the availability of a sufficiently rich database of hydrogen dispersion scenarios, obtained from parametric high-fidelity CFD simulations and/or from associated reduced-order models. The sensor placement problem should be formulated as a mathematical optimization problem, in which candidate sensor configurations are evaluated against quantitative performance metrics. Depending on time availability, several methods may be implemented, investigated and compared from data-driven approaches to optimization algorithms. The expected outcome is a generalizable workflow that, given a set of dispersion data for a target environment, returns optimal sensor layouts and quantifies the trade-offs between detection performance, sensor count, and resilience against leak-scenario variability. | Gregorio Casagrande Paolo Piras Luca Pratticò Michele Urbani | ||||||||||||||||||||||
22 | Modelling (SOFC thermal distribution 1D) | Scientific Machine Learning for 1D Multiphysics Modeling of Solid Oxide Fuel Cells | The focus of this thesis is to develo a 1D multiphysics model of an Solid Oxide Fuel Cell (SOFC) based on its underlying physics and combines it with Scientific Machine Learning techniques to obtain a model that is both physically consistent and able to reproduce real operating behavior. Solid Oxide Fuel Cells (SOFCs) are governed by tightly coupled electrochemical, thermal, and mass transport phenomena that produce significant gradients along the principal flow direction. Capturing this behavior requires a 1D multiphysics description, in which the governing balances are solved together along the stack coordinate. The aim is to obtain a model that is consistent with the governing physical laws while leveraging the modern data-driven frameworks such as Physic Informed Neural Networks (PINNs), Neural Ordinary Differential Equations (Neural ODEs) or operator learning approaches (DeepONet). The simulation results are then compared with experimental data acquired on an SOFC test bench, providing a way to assess whether the simulated behavior is consistent with the real system, and to evaluate the plausibility of the model as a representation of the actual SOFC operation. | Luca Pratticò - Gregorio Casagrande - Emanuele Martinelli | ||||||||||||||||||||||
23 | Modelling | Physics-Constrained Modeling and Parameter Estimation for Health Monitoring of Solid Oxide Electrolysis/Fuel Cells (SOFC/EC) | This thesis addresses the problem of health monitoring and modeling of high-temperature electrochemical cells, focusing on the interplay between physical knowledge and data-driven methods. These systems are governed by coupled electrochemical, thermal, and transport phenomena, described through dynamic models whose parameters represent physical properties such as resistances, kinetic coefficients, and transport limitations. During long-term operation, these parameters may evolve due to aging and degradation, while only a limited subset of variables is directly measurable. Assessing the internal state of the system and relating parameter changes to performance loss therefore remains a central challenge for diagnostics and lifetime management. The thesis explores physics-constrained modeling approaches in which physically motivated dynamic models are calibrated or continuously adapted using operational data. Modern machine-learning techniques that explicitly incorporate physical constraints, such as physics-informed or physics-guided neural networks, are considered alongside more classical reduced or hybrid strategies, exploring different levels of model complexity. The main objective is to investigate how physics-constrained learning frameworks can reconstruct internal states and estimate parameter variations indicative of degradation or performance drift from sparse and noisy measurements, laying the groundwork for future diagnostic and control-oriented applications. | Luca Pratticò - Gregorio Casagrande - Emanuele Martinelli | ||||||||||||||||||||||
24 | Modelling | Coupling dynamic modelling and Bayesian networks for quantitative risk assessment of hydrogen systems | The increasing adoption of hydrogen as a sustainable energy carrier presents significant safety challenges throughout the entire hydrogen value chain, from hydrogen production to storage and utilisation. This thesis aims to develop an integrated model-based framework for quantitative risk assessment (QRA) of hydrogen systems, building upon the toolchain approach proposed by Rogovchenko-Buffoni et al. (https://doi.org/10.1016/j.jocs.2013.08.009) for functional safety analysis. The research will focus on a specific hydrogen system configuration by considering the dynamic modelling activities of FBK for H2 production, storage or utilisation. By leveraging Modelica language multi-domain modelling capabilities, the framework integrates physics-based system models with functional safety requirements expressed through service-oriented components. The methodology extends beyond traditional Failure Mode and Effects Analysis (FMEA) by incorporating probabilistic risk assessment through automatically generated Bayesian Networks. The proposed approach enables early-stage identification of hazardous scenarios, quantification of failure propagation probabilities, and generation of risk priority metrics specific to hydrogen safety concerns such as leakage, combustion, and material compatibility issues. This model-based integration eliminates the semantic gap between system design and safety verification, ensuring compliance with emerging hydrogen safety standards. The thesis demonstrates how functional requirements for hydrogen systems can be formalised within the modelling environment itself, allowing both dynamic verification during simulation and static analysis for dependency mapping. The resulting Bayesian Networks provide insights into component criticality and enable "what-if" scenarios for risk mitigation strategies. Furthermore, the framework generates comprehensive FMEA tables that support design optimisation throughout the development lifecycle. | Michele Bolognese | ||||||||||||||||||||||
25 | Modelling | A quantitative risk assessment study of a hydrogen storage system based on metal hydrides | This master’s thesis project aims to develop and apply a novel quantitative risk assessment (QRA) framework based on a Bayesian Network (BN) on a Metal Hydride (MH)-based hydrogen storage module fed by hydrogen produced via electrolysis, considering realistic perating conditions, potential failure modes, and mitigation strategies. The study integrates hazard identification, consequence analysis, and probabilistic modelling to characterise risks associated with hydrogen absorption, desorption, and storage operations. The Bayesian network approach represents the complex dependencies between states, component failures, and operational factors. The final framework supports decision-making by identifying the dominant risk contributors, quantifying the effectiveness of safety barriers, and optimal mitigation actions to improve the system in terms of safety and reliability. The reference work is that of (https://doi.org/10.1016/j.ress.2025.111959). | Michele Urbani | ||||||||||||||||||||||
26 | Modelling | Techno-economic analysis of safety barriers selection | This master’s thesis aims to develop and apply a complete and adaptable techno-economic analysis (TEA) tool for the optimal selection of actions, safety barriers, procedural measures and emergency response strategies across the hydrogen value chain technologies. The main objective is to provide a structured decision-support tool that quantitatively balances safety performance with economic feasibility of the adopted safety strategies (https://doi.org/10.1016/j.ress.2025.111646,https://doi.org/10.1016/j.ress.2017.05.005). By integrating inputs from HAZOP and Quantitative Risk Assessment (QRA) studies, such as failure probabilities, barrier effectiveness, and consequence severity, the research will identify safety configurations that minimise total cost while meeting predefined risk tolerance criteria. The framework will be technology-driven, designed for application to various systems like electrolysis, compression and storage (metal hydride system). | |||||||||||||||||||||||
27 | Modelling | Optimal sensor layout for hydrogen detection: High-fidelity models of hydrogen leakages | The safety of expanding hydrogen infrastructure, particularly in systems like Electrolysis and Hydrogen Refueling Stations (HRS), is critical. Accidental H2 leakage poses a serious hazard due to its high flammability. This thesis addresses the need for accurate prediction tools to manage this risk. This project aims to develop a CFD model using OpenFOAM to simulate hydrogen leakage scenarios in hydrogen systems. The study will focus on predicting dispersion patterns, identifying concentration gradients, and assessing potential ignition risks, thereby contributing to improved safety measures in hydrogen infrastructure. The reference works for this topical area are Deng et al. (https://doi.org/10.3390/hydrogen5040052), Zhang et al. (https://doi.org/10.3390/en18020228), Wang et al. (https://doi.org/10.1016/j.ijhydene.2025.150365). | Luca Pratticò - Gregorio Casagrande | ||||||||||||||||||||||
28 | Modelling | Optimal sensor layout for hydrogen detection: Model order reduction of CFD leakage models | In the context of hydrogen leakage, high-fidelity Computational Fluid Dynamics (CFD) simulations provide accurate insight into hydrogen dispersion dynamics, capturing complex phenomena. However, their high computational cost limits their applicability for real-time scenario evaluation. This thesis addresses the need for fast and flexible predictive tools by developing a reduced-order model (ROM) derived from CFD simulations of hydrogen leakage. The ROM aims to reproduce the essential spatio-temporal features of hydrogen concentration fields while significantly reducing computational complexity. Different model reduction strategies may be explored, including classical projection-based approaches such as Proper Orthogonal Decomposition (POD) or data-driven techniques such as or physics-informed neural networks (PINNs), as well as possible hybrid combinations of these methods. The resulting reduced-order framework is intended to enable efficient simulation of a wide range of leakage scenarios under varying boundary and operating conditions, supporting downstream sensor layout optimisation. The reference works for this topical area are Deng et al. (https://doi.org/10.3390/hydrogen5040052), Zhang et al. (https://doi.org/10.3390/en18020228), Wang et al. (https://doi.org/10.1016/j.ijhydene.2025.150365). | Luca Pratticò Giulia Fedrizzi | ||||||||||||||||||||||
29 | Modelling | Optimal sensor layout for hydrogen detection: A preliminary study on hydrogen leakage localisation | This master’s thesis proposes a study on the optimisation of hydrogen sensor layouts within a confined environment to enhance the safety and detection efficacy of potential hydrogen leakages. The project is motivated by the safety-critical nature of hydrogen storage and utilisation. The methodology will leverage a large dataset generated by a pre-existing or specially developed reduced-order model. This model simulates various realistic failure scenarios, accounting for factors like ventilation rates, leak location, and leak magnitude, to map the temporal and spatial evolution of hydrogen concentration. The ultimate goal is to apply advanced optimisation techniques to select the minimum number of sensor placements that maximise the probability of early and reliable leak detection across the entire spectrum of simulated scenarios. The reference works for this topical area are Deng et al. (https://doi.org/10.3390/hydrogen5040052), Zhang et al. (https://doi.org/10.3390/en18020228), Wang et al. (https://doi.org/10.1016/j.ijhydene.2025.150365). | Luca Pratticò | ||||||||||||||||||||||
30 | Modelling | Design of a reactor | Feedstock gasification is a key thermochemical pathway for producing renewable syngas, enabling sustainable power generation, hydrogen production, and biofuels synthesis. However, the process involves strongly coupled phenomena—including turbulent flow, heterogeneous reactions, particle–gas interactions, and heat/mass transfer—which make experimental optimisation costly and challenging. Computational Fluid Dynamics (CFD) provides a robust platform for analysing gasification behaviour, predicting syngas composition, and optimising reactor design and operational conditions. Objectives: 1. Develop a CFD model for biomass gasification, including heat transfer, mass transfer, and chemical reactions. 2. Implement heterogeneous and homogeneous reaction kinetics suitable for typical gasifiers. 3. Analyse key process parameters (temperature, equivalence ratio, particle size, steam/oxygen ratios). 4. Provide optimisation guidelines for reactor performance and syngas quality. | Giuseppe Sassone | ||||||||||||||||||||||
31 | Modelling | Accelerating dynamic simulations of gasification reactors using physics-informed deep learning | Gasification is a thermochemical process that converts carbonaceous materials (biomass, coal, or waste) into syngas (a mixture of H2, CO, and other gasses) and is a critical technology for the energy transition. However, optimizing and controlling these reactors is difficult because they involve complex multi-physics phenomena (fluid dynamics, heat transfer, and chemical kinetics). High-fidelity models (such as CFD or detailed 1-D models) are computationally expensive and too slow for dynamic simulations, real-time control, or digital twin applications. This thesis aims to develop a Reduced Order Model (ROM) using Physics-Informed Machine Learning (PIML). Unlike classical "black-box" ML which requires massive datasets, PIML leverages the fundamental laws of physics (Ordinary and Partial Differential Equations) to train the model, ensuring physical consistency even in the absence of extensive experimental data. The developed PIML model is ready to be used in the context of dynamic simulations. | Michele Urbani | ||||||||||||||||||||||
32 | Modelling | Reduced order model of a single stage of the MH-based compressor. | This thesis focuses on the development of a Reduced Order Model derived from high-fidelity CFD simulations of a metal hydride tank developed as a stage for a thermochemical hydrogen compressor. | Luca Prattico | ||||||||||||||||||||||
33 | Modelling | Dynamic Modeling of a Hydrogen Production System (Waste to Chemical) through piro-gasification of multi feedstocks | The activity focuses on the modeling of an innovative piro-gasification plant designed to convert syngas generated from different feedstoks into hydrogen through a series of treatment and purification components. The adopted methodological approach involves studying and implementing the constitutive laws and mass and energy balances for each component of the plant in a simulation environment based on the Modelica language. Particular attention will be dedicated to the analysis of the gasifier model, which will be thoroughly studied and calibrated through the implementation of partial differential equations or by a ROM method. This phase will rely on data from both scientific literature and specific experiments conducted to ensure an accurate representation of the gasification process. Subsequently, the control strategy, including the corresponding control loops, will be designed and implemented to ensure the optimal operation of the plant. The primary objective will be to maximize hydrogen production from the processing of the single feedstock to multi feedstock, thus contributing to environmental sustainability and efficient resource utilization. Throughout the thesis, various modeling and simulation tools will be used to consolidate the accuracy and reliability of the proposed model.. This thesis aims to make a significant contribution to the understanding and optimization of gasification processes for hydrogen production, integrating advanced modeling, experimental data, and control strategies, with a particular emphasis on the accurate representation of the residue and plastics mix. | Michele Bolognese | ||||||||||||||||||||||
34 | Modelling | Digital twin of a 500 kW Alkaline Water Electrolyser (AWEL) prorotype with a Double Stack Configuration | The dynamic modeling of an alkaline water electrolyser (AWEL) balance of plant (BoP) for a double stack configuration is crucial for optimizing green hydrogen production, the oeverall BoP efficiency and ensuring reliable operation under varying conditions by considering the H2 demand of industrial end-user and the variable RES source. This project focuses on creating a comprehensive model that integrates electrochemical, thermodynamic, and mass/energy balance principles to simulate the behavior of the system. The methodological approach for this project involves a comprehensive and integrated approach to ensure accurate modeling and efficient operation of the alkaline electrolyser balance of plant. To capture the complex interactions within the system, we will utilize advanced multiphysics modeling platforms such as MATLAB/Simulink or Modelica. These tools allow us to dynamically model the integration of electrochemical reactions with thermodynamic processes, providing a holistic view of how different components interact under various operating conditions. To capture the complex interactions within the system, a multiphysics modeling approach such in Modelica will be used. These tool allows to dynamically model the integration of electrochemical reactions with thermodynamic processes, providing a holistic view of how the different components interaction happen under various operating conditions. A crucial step in ensuring the reliability and accuracy of our models is validating them against experimental data from commercial alkaline water electrolyzers. This validation process helps refine our models by aligning simulated results with real- world performance metrics, thereby enhancing their predictive capabilities. Differential algebraic equations (DAEs) in 1-D discretized model will be employed at the component level to accurately describe complex phenomena such as mass transport and chemical reactions. Incorporating mass/energy balances will ensure that all inputs and outputs are accounted for across the system. A proper Heat and Material (H&M) balance for the different operating states will be produced. Moreover, according the definitive P&Id, the control loop architecture will be implemented in the BoP model using traditional PID controllers to establish a baseline performance. Subsequently, a comparison will be made with Model Predictive Control (MPC) approach to evaluate the effectiveness in managing and orchestrate complex interactions during transient operations. MPC controllers will predict future states based on current conditions and adjust control inputs accordingly. Machine learning algorithms can also be integrated if necessary to enhance adaptive capabilities using historical data. An FMU (Functional Mock-up Unit) of the BoP model plant will be exported from the Modelica environment in order to create the MPC. This FMU will include only the balance of plant components without the layer of the PID control, allowing for seamless integration with external MPC tools for simulation and testing purposes. | Michele Bolognese | ||||||||||||||||||||||
35 | Modelling | Dynamic Modeling of a Green Hydrogen BOP of Metal Hydride storage (MH) coupled with 2 MW PEM electrolyzer for Maritime Applications | This activity focuses on the dynamic modeling of an innovative metal hydride storage system capable of storing up to 100 kg of hydrogen, integrated with a 2 MW PEM electrolyzer for green hydrogen production. The primary objectives include optimizing thermodynamic behavior during adsorption and desorption, creating mass and energy balances for all components, and enhancing the process flow diagram (PFD) to achieve a hydrogen delivery pressure of 36 bar while recovering heat from the PEM electrolyzer. Finally a particular focus will be paid in the simulation and implementation of a proper state machine ( State Graph, State Graph2, Petri Nets etc.) both for the operating condition and for the safety. | Michele Bolognese | ||||||||||||||||||||||
36 | Modelling | Entropy Generation Analysis in a Porous Hydrogen Storage Bed | Hydrogen storage in porous media, such as metal hydrides, metal-organic frameworks, or activated carbon, is a promising technology for safe and efficient hydrogen storage. These materials allow reversible hydrogen adsorption and desorption, making them viable for fuel cell applications, hydrogen refueling stations, and portable energy systems. However, during the charging (absorption) and discharging (desorption) processes, significant heat is released or absorbed, leading to temperature gradients and pressure drops within the storage bed, negatively impacting storage performance. One way to quantify these losses is by performing an entropy generation analysis to optimize bed design. | Giuseppe Sassone | ||||||||||||||||||||||
37 | Modelling | CFD Analysis of Multiphase Flows in Porous Media for Hydrogen Production | This thesis aims to conduct a CFD analysis on low temperature electrolyzers for the hydrogen production (e.g. PEM or AEM). The methodology involves the application and implementation of conservation laws, as well as mass and energy balances, using the open-source software OpenFOAM. Special emphasis will be placed on modelling bubble dynamics, considering multiphase flow in porous media at the anode side. The model validation will be addressed using available literature data. Finally, a sensitivity analysis will be conducted to investigate the impact of porous media properties on oxygen bubbles distribution. The ultimate goal of the thesis is to deepen the understanding of the processes involved in hydrogen production technologies and to provide insights for improving their design. | Giuseppe Sassone | ||||||||||||||||||||||
38 | Modelling | Predictive Model for High Temperature Hydrogen Technologies | The target of the thesis is to develop a predicitive modelling tool that addresses degradation analysis in solid oxide technologies (e.g. SOC and PCC). The methodology involves the development of a model that includes both main physics for such technologies (electrochemistry, mass and energy balances) and degradation phenomena (e.g. Ni coarsening, Sr segregation) that are commonly observerd along the lifetime of the electrochemical devices. The model validation will be addressed using available literature data regarding durability curves. The ultimate goal of the thesis is to deepen the understanding of the degradation processes involved in hydrogen technologies. | Giuseppe Sassone | ||||||||||||||||||||||
39 | Modelling | Multiphysics Model for Hydrogen Production | This thesis aims to conduct a CFD analysis on low temperature electrolyzers for the hydrogen production (e.g. PEM or AEM). The methodology involves the application and implementation of conservation laws, as well as mass and energy balances, using the commercial software Ansys Fluent. Special emphasis will be placed on introducing electrochemistry in the available CFD model, which includes multiphase flow in porous media at the anode side. The model validation will be addressed using available literature data. Finally, a sensitivity analysis will be conducted to investigate the impact of porous media properties and operating conditions on oxygen distribution. The ultimate goal of the thesis is to deepen the understanding of the processes involved in hydrogen production technologies and to provide insights for improving their design. | Giuseppe Sassone | ||||||||||||||||||||||
40 | Modelling | Optimization of BoP for a reversible SOC | This master thesis opportunity focuses on the optimization of Balance of Plant (BoP) components for reversible Solid Oxide Cell (SOC) systems, aiming to enhance the overall efficiency and performance of these advanced energy conversion devices. Reversible SOC systems have garnered significant attention for their potential to facilitate efficient and sustainable energy storage and conversion by seamlessly switching between electrolysis mode (producing hydrogen) and fuel cell mode (generating electricity from hydrogen). However, the successful integration of reversible SOC systems into practical applications requires careful design and optimization of the supporting BoP infrastructure to ensure optimal system operation, reliability, and cost-effectiveness. Activity includes: modelica model and validation | Michele Bolognese | ||||||||||||||||||||||
41 | Modelling | Data-driven model of Solid Oxide Stack | This master's thesis is focused on the development of a model based on ML/DL techniques, or other data-driven approaches, for Solid Oxide Cell, working in electrolyzer or fuel cell mode. The model aim to reproduce the dynamic performances of SOC including the degradation of the system. | Luca Pratticò | ||||||||||||||||||||||
42 | Modelling | H2 valley model for TEA and GHG emissions analysis | Developing a model of the hydrogen value chain, from green hydrogen production to end-use, begins with a preliminary definition of various energy sources and electrolyzers. The work will then focus on analyzing the best available land-based transportation systems and the main current and future users. Subsequently, a model will be developed to optimize the entire value chain of interest, ensuring the delivery of green hydrogen to specific end-users in the most cost-effective manner possible. | Paolo Piras Alessandro Saccardi (Michele Urbani x optimization) | ||||||||||||||||||||||
43 | Modelling | Development of a complete dynamic model of a 100kW PEM fuel cell system | In the context of a European project, a complete dynamic model of the 100kW PEMFC system will be developed based on Modelica language software. The scope encompasses capturing the full spectrum of dynamic behaviors, from transient load shifts to temperature fluctuations. | Luca Pratticò/Michele Bolognese | ||||||||||||||||||||||
44 | Modelling | Design a PEMFC dynamic model. | This thesis project focuses on creating a practical and robust model for Proton Exchange Membrane Fuel Cells (PEMFCs) that not only simulates their performance but also predicts degradation using machine learning techniques. By integrating machine learning algorithms into the modeling process, we aim to improve the accuracy of degradation predictions and enable proactive maintenance strategies. | Luca Pratticò/Michele Bolognese | ||||||||||||||||||||||
45 | Modelling | Design a degradation model induced by contaminants for a PEMFC stack. | This thesis project focuses on creating model for Proton Exchange Membrane Fuel Cells (PEMFCs) capable to predict degradation induced by contaminats in the H2 flow using machine learning techniques. Based on dedicated experiments this work aim to develop a degradation model with different contaminants like CO, N2 etc | Luca Pratticò | ||||||||||||||||||||||
46 | Modelling | Optimization of production, distribution and consumption of H2 in a hydrogen valley | This master thesis will begin with the development of models that replicate the behaviour of plants in a simple hydrogen valley using modelica language. Dynamic behaviour is the most important aspect to consider when dealing with production, distribution and consumption. The economic aspect is not neglected, as the aim is the optimization of plants to increase the overall efficiency and reduce operative cost. Modelon integrates a optimization library. The optimization can be handled also with external software/language, e.g. python; a comparison between the two ways can be detailed. What can be the activities carried out after a the optimization obtained with python language? | Emanuele Martinelli | ||||||||||||||||||||||
47 | Modelling | Development of Hydrogen Valley concepts for the design and Multi-Objective Optimization of hydrogen production, distribution, and utilization. | The thesis will include an overview of the various methods of hydrogen production, distribution, and utilization, focusing on the multi-objective optimization of hydrogen management based on the mix of end-users present in the valley (hard-to-abate sectors, mobility, power generation, etc.). The optimization will consider technical, economic, and environmental objective functions (LCOH, LCOT, GHG, etc.) | Alessandro Saccardi | ||||||||||||||||||||||
48 | Modelling | Review and design of Ejector for h2 systems | This thesis focuses on ejector application in h2 systems, in the first part will review state of art of the different applications of ejectror in balance of plant for h2. Then two different use cases will be selected and then analyzed as different option for the procceses modelling, such as number and sizing of the components. A steady state model will be developed along with the selection of technical kpis to evaluate several options of system arrangement. | Luca Pratticò/Giuseppe Sassone | ||||||||||||||||||||||
49 | ||||||||||||||||||||||||||
50 | Modelling | Design and optimization of a sealing system for PEM/AEM cell assembly | This thesis focuses on the development of a modular cell frame to be employed both in PEM and AEM ecosystem. The main goal is to design and test a sealing solutions which is already integrated in the cell's frame. The approach will be to further develop the current frames'design including sealing solutions from main companies (trellenborg/freudenberg...) Progetto simile di Trellenborg+Ensinger+Hycenta | Simone Zottele | ||||||||||||||||||||||
51 | Modelling | Surrogate models based on scientific machine learning for 2D heat conduction problems | Within the context of the AISA project (ESA/FBK-AVIO), which investigates AI-based surrogate models for space thermal engineering, this thesis aims to develop and validate a surrogate model for predicting thermal fields in 2D heat conduction problems with internal heat sources. The work targets a simplified PCB case study as a representative benchmark for conduction-dominated space electronics. Starting from a review of the relevant literature, the candidate will explore scientific deep learning approaches, such as physics-informed neural networks, neural operators, and neural ODEs, to build a framework capable of learning the thermal response of planar geometries under various source configurations and boundary conditions. The surrogate shall demonstrate generalization to parametric configurations unseen during training, including different power levels, source positions, and material properties, and its performance will be quantitatively validated against reference solutions. Potential extensions towards additional heat transfer mechanisms and more complex configurations will follow naturally from the results obtained on the baseline case. | Luca Pratticò | ||||||||||||||||||||||
52 | Modelling | Surrogate models for transient thermal prediction of simplified case study mimicking small satellites in low Earth orbit | Within the context of the AISA project (ESA/FBK), which investigates AI-based surrogate models for space thermal engineering, this thesis aims to develop and validate a surrogate model for predicting the transient thermal behaviour of a simplified case study of a small cubic satellite (CubeSat). The system is characterised by cyclic thermal loads driven by the alternation of sunlight and eclipse phases, with external radiative contributions from solar flux, albedo, and Earth infrared emission. Starting from a review of the relevant literature, the candidate will explore data-driven and hybrid scientific deep learning approaches, such as recurrent architecture, neural ODEs, physics-constrained networks and others, to learn the temporal evolution of the satellite's temperature field across different orbital and design parameters. The surrogate shall demonstrate generalization to unseen orbital scenarios and configurations, and its performance will be quantitatively validated against a high-fidelity reference model. Potential extensions towards more complex satellite geometries or multi-node thermal networks will follow naturally from the results obtained on the baseline case. | Luca Pratticò | ||||||||||||||||||||||
53 | Modelling | Techno-economic Assessment of membrane-based Hydrogen Purification Systems | The widespread use of hydrogen-based applications requires the development of efficient and cost-competitive technologies for the separation of high purity H2 from natural hydrogen and industrial gas streams. In this context, hydrogen-selective membrane-based systems represent a promising solution due to their high perm-selectivity, modularity and potential for reduced energy consumption compared to conventional separation technologies. This thesis will be carried out in collaboration with the Fondazione Bruno Kessler (FBK) and will focus on the techno-economic assessment (TEA) of advanced membrane-based hydrogen purification systems developed within the European project HERMES. The work will combine process modelling and economic evaluation to assess the performance of membrane systems for the hydrogen recovery from different industrial gas streams. The overall purification system will be implemented in a process modelling environment, integrating an existing membrane module model to simulate representative operating conditions and evaluate key performance indicators. Based on the process simulations, a TEA assessment and sensitivity analysis will be performed to estimate the cost of hydrogen purification and evaluate the impact of identified key parameters. The thesis will contribute to identifying critical parameters affecting the feasibility and scalability of membrane-based hydrogen purification technologies. | Luca Pratticò/Giulia Bonamigo | ||||||||||||||||||||||
54 | Testing/fabrication | Anode Electrode Design and Fabrication for AEM Water | Anion Exchange Membrane Water Electrolysis (AEMWE) is an emerging technology that combines the advantages of alkaline and PEM electrolysis, enabling the use of non-precious metal catalysts and reducing system costs. However, achieving high performance with stable, long-term operation remains a challenge, particularly at the anode, where the oxygen evolution reaction (OER) and complex catalyst–ionomer–substrate interactions limit device efficiency. The thesis focuses on the design, fabrication, and electrochemical characterization of high-performance anode electrodes for AEMWE, exploring innovative techniques (ink-based coating and/or physical vapour deposition) as catalyst deposition strategies, providing a comprehensive view of electrode performance from lab scale to device level. | Varun Donnakatte Neelalochan/Giulia Di Gregorio | ||||||||||||||||||||||
55 | ||||||||||||||||||||||||||
56 | ||||||||||||||||||||||||||
57 | ||||||||||||||||||||||||||
58 | ||||||||||||||||||||||||||
59 | ||||||||||||||||||||||||||
60 | ||||||||||||||||||||||||||
61 | ||||||||||||||||||||||||||
62 | ||||||||||||||||||||||||||
63 | ||||||||||||||||||||||||||
64 | ||||||||||||||||||||||||||
65 | ||||||||||||||||||||||||||
66 | ||||||||||||||||||||||||||
67 | ||||||||||||||||||||||||||
68 | ||||||||||||||||||||||||||
69 | ||||||||||||||||||||||||||
70 | ||||||||||||||||||||||||||
71 | ||||||||||||||||||||||||||
72 | ||||||||||||||||||||||||||
73 | ||||||||||||||||||||||||||
74 | ||||||||||||||||||||||||||
75 | ||||||||||||||||||||||||||
76 | ||||||||||||||||||||||||||
77 | ||||||||||||||||||||||||||
78 | ||||||||||||||||||||||||||
79 | ||||||||||||||||||||||||||
80 | ||||||||||||||||||||||||||
81 | ||||||||||||||||||||||||||
82 | ||||||||||||||||||||||||||
83 | ||||||||||||||||||||||||||
84 | ||||||||||||||||||||||||||
85 | ||||||||||||||||||||||||||
86 | ||||||||||||||||||||||||||
87 | ||||||||||||||||||||||||||
88 | ||||||||||||||||||||||||||
89 | ||||||||||||||||||||||||||
90 | ||||||||||||||||||||||||||
91 | ||||||||||||||||||||||||||
92 | ||||||||||||||||||||||||||
93 | ||||||||||||||||||||||||||
94 | ||||||||||||||||||||||||||
95 | ||||||||||||||||||||||||||
96 | ||||||||||||||||||||||||||
97 | ||||||||||||||||||||||||||
98 | ||||||||||||||||||||||||||
99 | ||||||||||||||||||||||||||
100 | ||||||||||||||||||||||||||