Doctoral School

Reading date: 23/01/2026

• BARRERA HERRERA, JAVIER ENRIQUE: Improving Time Predictability and Code Coverage of Embedded GPUs for Real-Time Systems
  
  Author: BARRERA HERRERA, JAVIER ENRIQUE
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTER ARCHITECTURE
  Department: Department of Computer Architecture (DAC)
  Mode: Normal
  Deposit date: 07/11/2025
  Reading date: 23/01/2026
  Reading time: 11:00
  Reading place: C6-E101
  Thesis director: CAZORLA ALMEIDA, FRANCISCO JAVIER | KOSMIDIS, LEONIDAS
  Thesis abstract: This dissertation addresses challenges that the adoption of GPUs in Critical Embedded Systems (CES) faces, namely, Time Predictability and Code Coverage. Different domains that deploy CES are constantly adding Artificial Intelligence (AI)-based features, such as autonomous driving, that demand high performance levels. Multi-Processors Sytem-on-Chip (MPSoCs) are widely used to provide said performance levels, as they are equipped with accelerators, among which, Graphics Processing Units (GPUs) are a common choice. However, CES must undergo a rigorous Verification and Validation (V&V) process, in which a certain level of Execution Time Determinism (ETD) must be guaranteed. The use of several tasks to increase the overall utilization introduces contention in shared resources, which induces time variability. To provide the ETD guarantees, the time variability must be either mitigated or tracked and controlled. Another challenge for the adoption of GPUs in CES, is that the V&V process demands evidence of the thoroughness of the testing phase, for which Code Coverage is used as a test quality indicator. However, Code Coverage, as traditionally understood for sequential CES does not cover all possible scenarios in which a GPU thread might execute.For low-criticality and mixed-criticality CES, we contend that we can allow tasks to share the Last Level Cache (LLC) if hardware support for contention tracking is provided. Providing a clear understanding on how tasks contend with each other enables CES developers to balance performance and time predictability. For high-criticality CES, it is a common practice to implement LLC partitioning as it allows tasks to access LLC without suffering from inter-kernel contention, however, tasks may experience a performance loss due to lack of resources. In this Thesis, we propose Demotion Counters, a novel technique that tightly tracks how much each task has been demoted towards eviction in the LLC, thus, effectively quantifying their impact in CES. Additionally, we also assess the use of NVIDIA’s Multi-Instance GPU (MIG) feature as means to improve ETD in high-criticality CES.Code Coverage is used as a test quality indicator to provide evidence of the thoroughness of the testing, as required by the V&V process. However, if applied as traditionally understood, it will ignore the threading dimension of GPUs. Threads have private regions of memory, as well as shared regions at different granularities. This means that errors that are innocuous to one thread are potentially harmful for another, hence, it does not cover all possible cases under which GPU threads might execute. In this Thesis, we propose the use of Per-Thread Statement Coverage (PTSC), which tracks the Code Coverage at thread granularity. In order to mitigate the overheads caused by PTSC, several variants that apply different orthogonal optimizations are also proposed. Finally, we also evaluate the potential benefits of using hardware support for PTSC, mitigating the memory consumption of PTSC, as well as the execution time impact at deployment.Summarizing, this Thesis advances the state of the art in the adoption of GPUs in CES. The proposal of hardware contention tracking support and assessment of NVIDIA’s MIG, as means to improve ETD, effectively tackles the Time Predictability challenge in shared LLC. The proposal of software PTSC allows providing CES designers with the whole picture of the execution in commercially available GPUs. The use of hardware support for PTSC mitigates the overheads of software PTSC in deployment, while the different compression techniques reduce the volume of information during testing phase without losing data. Therefore, this Thesis provides means to face the Time Predictability and Code Coverage challenges of GPUs in CES.
  

• MINGOT BEJAR, JULIA: Applications of Poly(N-isopropylacrylamide)-based Hydrogels in Chemical Engineering
  
  Author: MINGOT BEJAR, JULIA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN POLYMERS AND BIOPOLYMERS
  Department: Department of Chemical Engineering (EQ)
  Mode: Normal
  Deposit date: 21/11/2025
  Reading date: 23/01/2026
  Reading time: 11:00
  Reading place: Campus universitari Diagonal-Besòs Av. d'Eduard Maristany, 6-12, 08930 Sant Adrià de Besòs, Barcelona Edifici I planta 0, espai I.0.1 (Polivalent)
  Thesis director: ARMELIN DIGGROC, ELAINE APARECIDA | LANZALACO, SONIA
  Committee:
       PRESIDENT: MECERREYES MOLERO, DAVID
       SECRETARI: MAS MORUNO, CARLOS
       VOCAL: CAPEZZA, ANTONIO
  Thesis abstract: This doctoral thesis explores the multifunctionality of poly(N-isopropylacrylamide) (PNIPAAm)-based hydrogels as a platform for biomedical and environmental applications. By exploiting the thermoresponsive properties of PNIPAAm and its copolymers, the research demonstrates how this material can be engineered to perform in distinct technological domains.In the biomedical field, inert polypropylene surgical meshes, commonly used for hernia repair, were functionalised with gold nanoparticles and a Raman reporter, converting their surface into a SERS-active platform. Covalent grafting of PNIPAAm-based copolymers onto the plasmonic substrate imparted thermoresponsive behaviour, resulting in an implantable device capable of simultaneous SERS detection and thermal response. In vitro assays with fibroblast cells confirmed the biocompatibility and stability of the device, highlighting its potential for minimally invasive diagnostics and post-surgical monitoring.A complementary theranostic approach was applied to the modification of 3D polyurethane sponges, used in endoluminal vacuum-assisted therapies, with PNIPAAm hydrogel and metallic nanoparticles. Functionalisation with gold and silver nanoparticles, stabilised by biopolymer shells, endowed the modified sponges with antibacterial properties. Photothermal activation under Raman laser irradiation resulted in significant antimicrobial activity against Escherichia coli and Staphylococcus aureus, offering new prospects for infection detection and treatment in implantable devices.In the environmental section, the thermoresponsive behaviour of PNIPAAm hydrogels was exploited for solar-driven water desalination and sustainable energy generation. A PNIPAAm-alginate-PEDOT:PSS system exhibited enhanced water evaporation rates potentiated by the consecutive surface contraction of the hydrogel (“pudding effect”). Further developments involved PNIPAAm-gelatine hydrogels incorporating carbon black as photothermal absorber, achieving stable desalination performances under real conditions (outdoor sunlight), with demonstrated durability and reusability.Finally, PNIPAAm-based matrices were employed to fabricate hydrogel thermal electricity generators. This combination of PNIPAAm with doped conductive polymers enabled photothermal-to-electric energy conversion driven by ionic transport within the hydrogel network upon exposure to solar light.Overall, this thesis establishes PNIPAAm hydrogels as a highly adaptable material platform. Their thermoresponsive behaviour, combined with plasmonic or photothermal functionalities, offers potential solutions to challenges in healthcare and resources sustainability.
  

• SLIMANI, MEHDI: Computational strategies for time-accurate simulation of part-scale LPBF
  
  Author: SLIMANI, MEHDI
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN STRUCTURAL ANALYSIS
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Article-based thesis
  Deposit date: 15/12/2025
  Reading date: 23/01/2026
  Reading time: 11:00
  Reading place: Sala Zienkiewich (CIMNE) Building C1, UPC - Campus North Gran Capitan S/N 08034 Barcelona
  Thesis director: CHIUMENTI, MICHELE | CERVERA RUIZ, LUIS MIGUEL
  Thesis abstract: The qualification of MAM (Metal Additive Manufacturing) processes remains a majorchallenge due to the complex thermo-mechanical phenomena involved.The process is driven by a small moving heat sourcethat generateshighly localized, transient thermal gradientsand induces thermal strains.As these strains are constrained bythe surrounding material,residual stresses and warpage develop,causing part distortion or even failure.Accurate modeling is essential for understanding the underlying physics, aswell as for reliable process qualification and parameter optimization.However,such simulations are computationally expensive due to the small size of theheat source, which introduces disparate spatial scales,and its continuous motion, which gives rise to equally disparate temporalscales.The need to simultaneously resolve these scalesrenders high-fidelity part-scale simulations prohibitively expensive.This thesis contributes to the field of MAM modeling on both the appliedand methodological fronts. On the applied side, methods for warpage and stressmitigation are investigated in both DED (Directed Energy Deposition) and LPBF (Laser Powder Bed Fusion) processes, includinga novel substrate design strategy for DED that significantlyreduces residual stresses, and a modeling framework to capturerecoater–induced build failure in LPBF.On the methodological front, the thesis focuses on developing efficientstrategies for high-fidelity part-scale simulations of LPBF processes,with particular emphasis on overcoming the disparity of temporal scales.WhileAMR (Adaptive Mesh Refinement) has become a popular approach to address the challenge of disparatespatial scales, uniform time stepping remains the standard approach in the field.For centimeter-scale parts, this can require hundreds of millions of time-steps,making such simulations computationally unfeasible.Commonly used strategies to alleviate this issue involveextreme simplifications of the thermal model,such as lumping multiple tracks or layersinto a single time-step.Effectively, this eliminates the small scales associated with the moving heat sourcebut compromises the model's predictive accuracy,requiring additional calibration.Two methods are proposed to address the temporal-scale disparity withouteliminating the underlying small scales: the advected subdomain and aRobin–Robin substepping scheme, both designed to preserve modelfidelity while drastically reducing computational cost.The advected subdomain method attaches a moving mesh to the laser. Bysolving the thermal problem in the reference frame of the heat source, thetransient dynamics near the melt pool become quasi-steady, allowing the use ofsignificantly larger time-steps.Substepping divides the domain into regions that evolve with differenttime-steps:finer steps are applied locally around the moving heat source, while larger stepsare used away from it.The developed Robin-Robin coupling scheme proves robust andensures mesh-independent convergence between the regions.These methods and their components are systematically evaluated throughnumerical analysis, benchmarked against standard approaches, and validatedagainst experimental data. Furthermore, they are combined to compound theirrespective benefits.Together, these contributions advance numerical MAM modeling,thereby improving the computational efficiency of high-fidelity simulationsand enabling reliable process qualification and optimization.
  

Reading date: 26/01/2026

• MACIÀ CID, LLORENÇ: Performance Enhancement of a Vacuum Generation Pneumatic Device by Fluid Dynamics Characterization
  
  Author: MACIÀ CID, LLORENÇ
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN MECHANICAL, FLUIDS AND AEROSPACE ENGINEERING
  Department: Department of Mechanical Engineering (EM)
  Mode: Normal
  Deposit date: 19/12/2025
  Reading date: 26/01/2026
  Reading time: 11:00
  Reading place: Sala de Conferències de TR5, planta 1, ESEIAAT
  Thesis director: CASTILLA LOPEZ, ROBERTO | GAMEZ MONTERO, PEDRO JAVIER
  Thesis abstract: In this thesis, behavior analyses and performance improvements, are presented for a supersonic vacuum ejector, key component in industrial automatization tasks. To characterize the performance of an EVKAC180 model ejector, a combination of numerical simulations using the Computational Fluid Dynamics (CFD) OPENFOAM toolbox and experimental measurements on a dedicated test rig was used. Two operation regimes: a supercritical mode, where the secondary flow chokes, and a subcritical mode, where it remains subsonic. And breakpoints were identified. Simulations reproduced this dual behavior with good agreement with experimental data, though some deviations were found at high and low flow rates. In addition to the density-based implicit solver (HiSA), an ex plicit solver (rhoCentralFoam) was also used, confirming consistent results across the flow rates. The polytropic evolution, another key ejector metric, was found to be initially adiabatic, progressively transitioning to isothermal. A one-dimensional model was developed to complement CFD simulations and estimate ejector performance from geometry and operating conditions. The model computes the entrainment ratio and secondary pressures under both critical and subcritical regimes. Its results were validated against experimental and CFD data, showing accurate predictions and low deviations (below 4 %) in critical regimes. It provides a faster, low-cost alternative for early design stages. The ejector performance was improved by analyzing the influence of design parameters through single- and multi-factor analyses. The mixing chamber length proved to be the most impactful factor, leading to a 10 % individual improvement. The fractional factorial multi-factor analysis confirmed this trend and produced the final improved geometry design, referred to as the EDGE ejector, achieving a slightly higher overall performance gain of 10.4 %. The interaction effects among parameters were found to be limited yet important overall. Finally, an empirical model tool for predicting the Total Evacuation Time (TET) was proposed, combining the characteristic and polytropic curves. Several experimental test rigs were used to refine the polynomial fits of the characteristic curves, exhibiting deviations in TET prediction as low as 1.4 %. The validated tool was then applied to the EDGE ejector, achieving a 4 % reduction in TET (a gain of 8 s) compared to the original model, fulfilling the objective of this research. Moreover, the tools developed in this thesis reduce the need for extensive experimental data and enable reliable forecasting for new ejector designs.
  

Reading date: 27/01/2026

• BOSCH PADRÓS, MIQUEL: Optogenetic control of force transmission in puripotent epithelia
  
  Author: BOSCH PADRÓS, MIQUEL
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN APPLIED MATHEMATICS
  Department: School of Mathematics and Statistics (FME)
  Mode: Normal
  Deposit date: 16/12/2025
  Reading date: 27/01/2026
  Reading time: 15:00
  Reading place: Sala d'Actes de l'FME, Edifici U, Campus SudEnllaç MEET: meet.google.com/wda-kfee-saf
  Thesis director: ARROYO BALAGUER, MARINO | TREPAT GUIXER, XAVIER
  Thesis abstract: Development requires a combination of three phenomena: increasing the number of cells, specifying their fates and undergoing morphogenesis, which means acquiring the correct shapes. Apical constriction is an important driving mechanism of morphogenesis, occurring within a cell but bridging with tissular scale to acquire and maintain shape. Apical constriction is well studied at the cellular level and conserved through the animal kingdom, but the forces that need to be generated and transmitted through the tissue in the process have never been measured and described. To fill this gap, we used a novel optogenetic tool to induce apical constriction in human pluripotent stem cells, combined with traction force microscopy to measure the mechanical forces involved in the process. With this techniques, we discovered that constriction creates a consistent but small signature in traction maps, compatible with apical contractility increase and volume conservation. In addition, we subjected regions of a monolayer to apical constriction and revealed that the cellular displacement field obeys a screened Poisson equation in two dimensions, which implies the existence of a lengthscale with a rheological origin and allows to obtain the Green's function of the tissue. While deformations can be tailored in space and time, we also find that jamming transitions cannot be engineered through apical contractility, which exposes a strong unjammed nature of this pluripotent epithelium. These insights reveal key rheological aspects of human pluripotent stem cells at timescales relevant for morphogenesis, inaccessible through other techniques. Because this cells are used around the globe to derive organoids and embryo models but are highly understudied mechanically, this work establishes a key building block for future works that require shape or force control in stem cell-derived tissues.