Higher Institute on Territorial Systems for Innovation
PORTO@iris (Publications Open Repository TOrino - Politecnico di Torino)Not a member yet
146173 research outputs found
Sort by
Silica-alumina aerogels functionalized with amino-groups for the adsorption of CO2
Get PDFTo overcome climate challenges, emerging CO2 capture technologies focus on innovative solid sorbents, like aerogels. Silica aerogels, in particular, have emerged as an exceptional class of materials with unique properties even though their adsorption selectivity towards CO2, and their thermal and mechanical properties are limited. In this research, a hybrid silica-alumina aerogel has been prepared, to merge the elevated surface area typical of silica with the basic character (favorable to CO2 adsorption) and the high thermal and mechanical resistance typical of alumina. For the first time, these silica-alumina aerogels were functionalized with 3-aminopropyltriethoxysilane (APTES) amino groups through a one-pot process. This approach was chosen to directly introduce functional groups during gel formation, reducing the number of post-synthesis steps. The functionalization with amino groups aims to strengthen the interactions with CO2 molecules via acid–base interactions, thus enhancing adsorption capacity and selectivity. Three samples with increasing APTES content as well as a reference without any functionalization were prepared and characterized in terms of physico-chemical and adsorption properties. The results of CO2 and N2 adsorption tests as well as in situ FTIR suggest that low functionalization does not confer a significant advantage in CO2 capture. Higher APTES contents, instead, lead to a significant increase in the total quantity of adsorbed CO2 and in higher selectivity over N2 (calculated according to the Ideal Adsorbed Solution Theory). Moreover, also the strength of interaction increases, since not only physisorption, but also chemisorption takes place
Towards Better Generalization and Interpretability in Unsupervised Concept-Based Models
No full textTo increase the trustworthiness of deep neural networks, it is critical to improve the understanding of how they make decisions. This paper introduces a novel unsupervised concept-based model for image classification, named Learnable Concept-Based Model (LCBM) which models concepts as random variables within a Bernoulli latent space. Unlike traditional methods that either require extensive human supervision or suffer from limited scalability, our approach employs a reduced number of concepts without sacrificing performance. We demonstrate that LCBM surpasses existing unsupervised concept-based models in generalization capability and nearly matches the performance of black-box models. The proposed concept representation enhances information retention and aligns more closely with human understanding. A user study demonstrates the discovered concepts are also more intuitive for humans to interpret. Finally, despite the use of concept embeddings, we maintain model interpretability by means of a local linear combination of concepts
URDICO-Urban Dimension of Cohesion Policy and other EU Programmes: Final Report
No full textDrawing on in-depth analysis of eight European cities – Budapest, Florence, Ghent, Prague, Rotterdam, Strasbourg, Valencia and Warsaw – URDICO examines how cities participated in and influenced Cohesion Policy during the 2014–2020 and 2021–2027 programming periods. The research combines detailed fieldwork, stakeholder engagement and original data collection to assess governance models, funding patterns, institutional capacity and the alignment of EU funds with long-term urban strategies. The findings show that while the urban dimension of Cohesion Policy has been formally strengthened, cities’ real influence, autonomy and access to resources vary widely across Member States. Where cities are entrusted with clear roles, adequate instruments and stable mandates, Cohesion Policy delivers more integrated, place-based and visible results. Conversely, increasing centralisation and fragmented governance risk weakening subsidiarity and diluting the territorial impact of EU investments
Reliability analysis of 1D estimation for TBM operational parameters
Get PDFAccurate TBM performance estimation is essential for effective tunnel design and planning. This study introduces a one-dimensional (1D) estimation model that estimates thrust, torque, power, cutterhead speed, and tool count using only excavation diameter. The model was developed across four TBM types—open, single shield (SS), double shield (DS), and earth pressure balance (EPB)—to isolate the influence of diameter from other variables. Validation against existing models and a 52-case independent dataset confirmed strong correlations: torque scales with the cube of the excavation diameter (R2 = 0.89 for EPB), power grows faster than linearly (R2 = 0.83 for EPB), thrust increases supra-linearly (R2 = 0.79 for EPB), and cutterhead speed decreases with diameter (R2 = 0.87 for open TBM). Tool count grows proportionally. A reliability matrix compares model accuracy and data support, aiding selection based on both fitness and robustness. This 1D model offers fast, consistent estimates for early-stage assessments. While it excludes detailed geological input, it is suited for feasibility studies and preliminary design. Future work will incorporate additional ground and machine parameters and extend validation across a broader range of tunneling conditions to enhance generalizability
Statistical Route Feasibility in Metro-Access Optical Networks for Next-Generation RAN X-Haul
Get PDFThis paper investigates the statistical feasibility of routes in converged metro-access optical networks for next-generation RAN x-haul. Two network topologies are considered: N1 with longer metro spans and N2 with shorter, denser connectivity. Route feasibility is evaluated via Bit Error Rate (BER) profiling under two thresholds (10−3 and 10−2), using two commercially available coherent transceivers: Cassini DCO (DP-QPSK, DP-8QAM, DP-16QAM) and Phoenix DCO (DP-QPSK, DP-16QAM). A Python-based physical-layer simulator models attenuation, ASE, and NLI to derive GSNR and end-to-end SNR per route; BER is then computed via modulation-dependent closed forms. Results show that DP-QPSK provides near-universal feasibility in both networks, while higher-order formats, especially DP-16QAM, are strongly topology and threshold dependent. N2 enables significantly broader applicability of high-order modulation than N1, and Cassini’s intermediate DP-8QAM offers a robust capacity–reliability trade-off absent in Phoenix. These findings highlight the importance of network-aware planning and transceiver flexibility to ensure route viability for RAN transport over converged optical infrastructures
High strain rate testing and modeling of 3D-printed polymeric cellular structures
Get PDFIn the past few years, growing attention has been given to crashworthiness studies of 3D-printed lattice and cellular structures due to their excellent energy-absorbing capabilities and design freedom. The high strain rate behavior of these components has not been fully comprehended and their use is thus still limited. This paper investigates the high strain rate behavior of 3D-printed polymeric cellular structures, varying cell topology, material, density, and launching speed. The specimens are compared by quasi-static and dynamic compression tests, with strain rates varying from 0.01 s-1 to 3000 s-1. A Taylor test is employed to achieve launching speeds up to 200 m/s, which are rarely investigated for such structures. The wide range of strain rates results in notable changes in the collapse mechanisms of the structures and outstanding enhancements in specific energy absorbed. A new specific energy absorption evaluator is developed to account for the different behaviors of the quasi-static and high strain rate cases. Different results are obtained with short carbon fiber reinforced and unreinforced polyamides, while the cell topology is found to influence the failure behavior. The unreinforced material has a greater strain rate sensitivity but shows higher fracturing and catastrophic failure, while the reinforced material behaves more stably. Lastly, a simple finite element model with reduced inputs is developed to reproduce the deformation and the specific energy absorption of the structures. The model is intended to promote the use of simpler yet accurate models in large crashworthiness studies, where 3D-printed polymeric structures can be used
Silver-functionalized polyvinyl alcohol nanofiber membranes: A comparative study of nanoparticle incorporation and coating deposition
Get PDFPreventing contact with pathogens is a critical requirement in filtration applications for both public and private
environments. Developing antibacterial membranes is essential to enhance protection and mitigate contamination risks. In this work, a direct comparative study of two silver-based functionalization strategies applied to
electrospun polyvinyl alcohol (PVA) nanofiber membranes is presented, aiming to identify the most effective
approach for producing durable antibacterial filters. The first method involves in-fiber incorporation of silver
nanoparticles (AgNPs) by electrospinning a colloidal AgNP dispersion blended with the PVA solution, followed
by crosslinking. The second method uses physical vapor deposition (PVD) to deposit a silica- or zirconia-based
composite coating embedding silver nanoclusters onto pre-crosslinked PVA fibers. Both types of membranes
were characterized at each processing stage. Morphological changes were assessed by FESEM and image analysis,
surface wettability by contact angle measurements, and silver content by EDS. Antibacterial activity was tested
against both Gram-positive and Gram-negative bacteria, while silver ion release in water was monitored to
evaluate the ability of the inorganic matrix to control Ag+ leaching.
The results show that, although both strategies confer antibacterial properties, the PVD-based coating provides
a more uniform silver surface distribution, controlled and sustained silver release, and preservation of nanofiber
morphology, making it a promising route for the fabrication of high-performance antibacterial membranes
Advances in Gaming, Media, Artificial Intelligence, and Blockchain for Consumer Technologies
No full textL’internazionalizzazione della formazione in urbanistica. Nota introduttiva
No full textIl testo introduce un volume che esplora come l’internazionalizzazione stia trasformando la formazione urbanistica, nel quadro del più ampio processo che investe lo Spazio europeo dell’istruzione superiore. I saggi sono esito di un un confronto collettivo portato avanti nell’ambito del seminario SIU Intersezioni. La formazione urbanistica di fronte ai mutamenti, coordinato tra il 2022 e il 2025 dalla Commissione Formazione della Società Italiana degli Urbanisti. Questa nota introduttiva illustra i quesiti di fondo e le modalità attraverso la quale si è pervenuti al volume e alla creazione di una commissione permanente della SIU