PhD Graduate

On the Use of Textile Waste Fibres for Sustainable Solutions in Earth Embankments Repair

The use of soil mixed with fibres from textile waste in embankment improvement has been investigated with the twofold purpose to identify an effective engineering practice and to provide a strategy for the circular economy of textiles. Laboratory tests have been conducted to define the appropriate mixture proportions and to compare physical properties and hydromechanical behaviour of natural soil and soil treated with linen and viscose fibres. The results indicate that with appropriate fibre content, manageable and homogeneous mixtures can be achieved, making this solution suitable for practical applications; however, ensuring applicability at the site scale requires the use of appropriately sized equipment. The presence of the selected textile fibres significantly impacts the hydromechanical behaviour of the soil and the effect of fibres depends on their geometry, properties, and the sample preparation procedure. As the results demonstrate that the geotechnical properties of the treated soil can be engineered to meet specific site requirements, the proposed approach seems a promising solution for earth works and ground improvement, such as embankment repair and restoration.

Road Safety on Italian State Roads through Predictive Approach: Transferability of Existing Models and Development of New Models

Every year, over one million people lose their lives in road crashes worldwide. These alarming figures led to the ambitious international goal of halving road fatalities by 2030 and completely eliminating them by 2050, in line with the so-called “Vision Zero.” This is the framework in which my PhD thesis was developed. Traditionally, road safety was addressed through a reactive approach, which involved interventions only after crashes occurred. However, the real challenge for road engineers is to wonder if it is possible to prevent crashes from happening in the first place. This question lies at the core of the predictive approach, based on regression models to estimate crash risk of various sites based on road geometry, traffic conditions, and other external factors. My research was focused on adapting and evaluating the transferability of existing models to the Italian context, and developing new alternative models. The ultimate goal is to provide a practical tool to support road managers in safety analysis, planning of interventions, and decision-making process.

A PFEM numerical framework for simulating 3D printing of advanced materials

This research introduces a Particle Finite Element Method (PFEM) framework for simulating 3D printing processes involving advanced materials. Leveraging the Lagrangian formulation of PFEM, the approach enables precise tracking of free surfaces and accommodates large deformations through an efficient re-meshing strategy. To capture the complex behaviour of modern printing materials, the framework integrates a range of viscoplastic, viscoelastic, and elasto-viscoplastic constitutive models, accounting for key phenomena such as elastic memory, phase transitions, and thixotropy. Specialized techniques are also developed to simulate extrusion and layer-by-layer deposition, incorporating nozzle motion, inter-layer and substrate contact, as well as strategies for reducing computational cost. The model is validated against experimental data, demonstrating high accuracy in predicting filament geometry and offering insight into the influence of material and process parameters. Finally, an optimized implicit finite element solver is implemented in a Fortran environment, enabling efficient, long-duration simulations of complex printed structures, including detailed infill patterns and parameterized geometries.

From the fault rupture to city seismic response: 3D multi-scale physics-based scenarios of earthquake effects

Earthquakes can be devastating, especially in densely populated urban areas. With urban populations projected to grow, improving seismic hazard and risk evaluation at a site-specific level is essential. While 3D physics-based earthquake simulations (PBS) are reliable for estimating the site-specific ground motion (GM), they are computationally expensive and rarely used for fully coupled earthquake fault-to-structure simulations. In this work, addressing these issues, we validated the accuracy of our open-source code, SPEED (https://speed.mox.polimi.it/), for simulating site-specific broadband ground motion. Then, to perform fully coupled simulations without computational overhead, we developed new tools in SPEED to (1) perform large-scale 3D nonlinear simulations to account for soil response, and (2) couple building response with ground motion at an urban scale.

Development of Models for the Simulation of the Mechanical Behavior of Cords Used in Tires

This study presents a new macroscopic numerical model to simulate the mechanical behavior of polymeric reinforcements used in tires. An extensive experimental campaign on rayon yarns and cords under different loading conditions provided information about material and geometrical nonlinearities. The reinforcement is represented as an equivalent three-dimensional continuum with an anisotropic constitutive model, incorporating viscosity and plasticity components. Local material directions are determined analytically based on twisting and cord construction. The model is calibrated and validated against experimental data, including tensile and bending tests. The proposed approach accurately predicts reinforcement behavior, offering insights for optimizing cord design in tire applications.

Sediment transport: from averages to fluctuations

This thesis explores bedload transport in rivers, focusing on defining, measuring, and modelling solid flowrate. It can be expressed in three ways: as sediment volume crossing a line per unit width and time, as the product of particle activity and velocity, or as the product of entrainment rate and hop length. I show that these definitions are different time/space averages of the same equation describing flux. I then focus on the second definition, highlighting unresolved issues in measuring flowrate, concentration, and velocity. Distinguishing between movement and stillness is key, so I propose three states: stillness, transport, and non-transport. Particles in transport state clearly move downstream, while non-transport involves isotropic jiggling and doesn’t contribute to bedload but affects other transport metrics. Finally, I examine turbulence’s effect on sediment transport using flume experiments with varying turbulence levels. Measuring fluid velocity, shear stress, and sediment motion reveals that increased fluctuations enhance flowrate, concentration, and velocity, with turbulence having a stronger effect on concentration than velocity.

Data-Driven Modeling of Nonlinear Dynamical Systems

The simulation of complex physical phenomena has traditionally relied on solving computationally expensive models derived from first principles. However, the growing availability of rich data enables a new paradigm: discovering governing equations directly from measurements. This approach enhances flexibility and computational efficiency while preserving physical consistency. By combining machine learning with dimensionality reduction, a new generation of reduced-order models has been developed. These methods identify key variables and governing dynamics from high-dimensional data, enabling accurate predictions and efficient simulations. Applications span from structural mechanics to computational biology, demonstrating the versatility and potential of these techniques. This research lays the groundwork for an AI-driven revolution in scientific discovery, accelerating our ability to understand and control physical systems across scales.

Stochastic characterization of reactive processes in porous media

My research is focused on exploring critical elements of the nature of flow and transport processes taking place across porous geomaterials and their interactions with the host solid matrix. In this broad context, dissolution is a key process driving mineral transformations. These transformations, in turn, drive weathering of mineral surfaces, interact with spreading of pollutants in groundwater systems, and favor carbon capture through mineralization. The key research questions tackled in my PhD thesis are related to ( a) enhancing our ability to directly observe and quantify dissolution reaction rates through original nanoscale imaging experiments and (b) providing an interpretation of the observed rates through rigorous stochastic modeling approaches capable of quantifying uncertainty.

The evolution of Self-Healing Performance of Ultra High Performance Concrete under harsh conditions: Experimental Investigations and Machine Learning Modeling Development

Ultra High Performance Concrete (UHPC) is renowned for its durability and mechanical properties. This thesis explores enhancing UHPC’s longevity through self-healing under harsh conditions, focusing on three main objectives: (1) the impact of mechanical loads and harsh environments on self-healing, (2) the repeatability of self-healing under multiple cracking/healing cycles, and (3) developing a predictive model using machine learning. The study finds that UHPC’s strong self-healing ability maintains its high mechanical properties and durability even under harsh conditions. The developed model demonstrates high predictive accuracy for UHPC’s self-healing capacity.

Numerical Investigation of Impacts on/of Granular Masses

Impacts either between flowing granular masses and obstacles or between blocks and still granular strata are characterised by large displacements and large strain rates. Despite the considerable both academic and industrial attention on these processes, their intrinsic complexity makes difficult a full understanding of the mechanisms occurring during these dynamic interactions. Only recently, robust numerical codes, able to deal with large displacements, have allowed the numerical simulation of impacts, opening new frontiers in the study of these processes. In particular, numerical codes can be used for impact forces and granular mass deformation estimation for engineering purposes and to advance our understanding of the processes occurring at the macro and the microscale during impacts, in particular the occurrence of regime transitions (from solid to fluid and viceversa).

Environmental Assessment of Lime-based Innovative Carbon Dioxide Removal and Storage Solutions

Climate change is a global environmental issue, and it is well known that in order to avoid global temperature increase with potential catastrophic effects is necessary to reduce drastically global greenhouse gases emissions and at least for the mid-century remove more CO₂ from the atmosphere than the amount emitted.

There are solutions for removing and storing CO₂ that are an acceleration of the natural process called rock weathering. Some of these processes are based on lime. Lime is a product from an energy intensive process with unavoidable emissions due to the chemistry of the process. However, if lime is produced from renewable energy and CO₂ is stored, it can be used for removing CO₂ from the atmosphere. Some of the studied processes are marine solutions because lime and CO₂ interact with the seawater storing CO₂ in the form of bicarbonates in the sea and counteracting ocean acidification.

The methodology applied to analyse these solutions is Life Cycle Assessment (LCA) a standardized methodology that would assess the potential environmental impacts of a process during its whole lifecycle. The aim is to assess not only the potential beneficial impact on climate change but also to assess the overall impacts generated in the environment.

Harnessing Hydrogenotrophic Methanogens Activity: Experimental and Modelling Insights on H₂ Transfer

Climate change will have negative effects on human health, economics, the environment, flora and fauna presence on the planet. For these reasons, the necessity to find and use renewable energy sources, such as bio e-methane, is compelling. Bio e-methane can be generated by reducing carbon dioxide (CO₂ ) and green hydrogen (H₂ ), through a biological process mediated by hydrogenotrophic methanogens (Archaea). The bio e-methane generated can be directly injected in the national natural gas grid and/or it can be used as a fuel in the transportation sector. During the PhD project, both experimental and modelling activities of this biological process have been conducted. Experimental analysis, carried out in batch and continuous operation, have allowed to identify the microorganisms maximum velocity, methane production rate and the correlation between these aspects. Modelling activities have validated experimental data collected and, helped understanding physical aspect of gas-liquid mass of H₂ . Some noteworthy results achieved include: i) cultivation of hydrogenotrophic methanogens; ii) the development of a method for measuring the specific velocity of microorganisms (SHMA); iii) implementation and validation of a model for the biological process. Looking ahead, further insights into H₂ gas-liquid mass transfer are needed to optimize the process, as it remains a key aspect of this biological process.

Strategies for the Optimization of Resource Recovery from Sewage Sludge Ashes and Other Sludge-Derived Matrices Towards Process Scale-Up

Phosphorus (P) is a critical raw material and its recovery from secondary sources such as sewage sludge and sewage sludge ashes has gained more attention in recent decades. This Ph.D. research aimed firstly at optimizing the recovery of P from sewage sludge ash via wet chemical extraction and secondly at exploring different approaches to recover P from sewage-sludge-derived matrices, also targeting recovery of carbon. Innovative decontamination and precipitation strategies were tested with the aim of producing a high-quality phosphorus-rich precipitate that can be used as fertilizer, reducing pressure on primary sources. The methodology used involved laboratory and small-pilot scale experiments combined with modeling and statistical analyses that allow the selection of the best operational parameters of the recovery process. The results offered valuable insights for refining the sludge treatment line or subsequent sludge disposal methods, such as mono-incineration or hydrothermal carbonization, to improve phosphorus recovery. Further research is needed to ensure the development of the proposed solution to industrial scale.

Fracture Instability In Heated Concrete: A Reassessment Of The Fundamental Mechanisms Behind Explosive Spalling

There is no doubt that concrete will continue to be utilized as a construction material much into the future as it is now the most often used building material. Thanks to its low conductivity, it is known to exhibit good behaviour in fire. However, the fact that concrete is prone to the spalling phenomenon when exposed to fire poses a significant obstacle to its widespread utilization. Explosive spalling keeps being one of the main open issues concerning the fire safety of concrete structures. Though it is recognized that cracks in the exposed cover can be triggered by the combined effect of thermal stress and pore pressure, the energy source and the mechanism behind the violent projection of fragments are not yet fully understood.

A numerical approach to investigate the role of partial saturation in the vibration induced by underground railway traffic

Ground-borne vibration generated by the passage of underground trains may change over time for several reasons, such as increasing weight and speed of trains or ageing of the infrastructure components, as well as because of a variation in the dynamic response of the soil surrounding the tunnel. During my PhD, I mainly focused the attention on the last point, exploring the effect that partial saturation of surface layers may have in the propagation of vibration. Firstly, the case study of the M1 line in Milano was reproduced in Finite Element numerical models. In the models, the results of two geophysical test campaigns conducted in the same place but in different times were exploited to capture the change in the dynamic soil behaviour induced by a different water content. Secondly, a complete framework was developed to take care of variations of conditions of partial saturation in train-induced vibration problems. In the process, special care was devoted to the identification of the hydraulic behaviour of the soil layers starting from the knowledge of basic geotechnical properties. The results of the study provide useful insights in view of the growing attention to quality of life of people in urban areas, together with the increasing concerns about extreme climatic events.

Advancing Seismic Risk Assessment in Large Urban Areas by Means of Regional-Scale 3D Physics-Based Ground Motion Simulations

The main aim of my PhD work is to explore an innovative approach for assessing seismic risk in large urban areas, where ground shaking scenarios are provided by 3D physics-based numerical simulations (PBS) in realistic geological and seismotectonic contexts. The high-performance spectral element code SPEED (https://speed.mox.polimi.it/), developed at Politecnico di Milano, is applied, in combination with an Artificial Neural Network technique to enrich the high-frequency range of PBS. This work was part of the URBASIS Project (https://urbasis-eu.osug.fr/), funded by H2020 – Marie Sklodowska-Curie Actions (MSCA).

This research mainly includes three aspects related to the incorporation of ground shaking scenarios from 3D PBS:

1. spatial correlations of broadband ground motions
2. generation of a suite of seismic shaking scenarios for Thessaloniki region
3. utilization of the simulated ground shaking scenarios in seismic risk assessments

These investigations provide valuable insights into the uncertainties associated with seismic risk estimates, which is crucial for understanding the potential extent of earthquake-induced economic and social losses, especially for cities located close to seismically active faults.

Probabilistic Study of Fluid Migration and Allocation of Underground Energy Resources

In my thesis, I explored key aspects of hydrocarbon production through three distinct studies. Firstly, I investigated gas flow in low permeable materials, a vital indicator of how efficiently conventional underground reservoirs can store gases like methane, carbon dioxide, and hydrogen.

Secondly, I proposed innovative techniques for production allocation, crucial for oil forensics and quantifying fluid contributions when multiple sources share production facilities.

Finally, I introduced a technique for rigorously quantifying groundwater contamination after hydraulic fracturing operations, addressing environmental concerns. These studies offer valuable insights into hydrocarbon production, impacting industry practices.